(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Fishery bulletin"




Fishery Bulletin 



Vol. 82, No. 1 January 1984 

ROPES, JOHN W., DOUGLAS S. JONES, STEVEN A. MURAWSKI, FREDRIC M. 
SERCHUK, and AMBROSE JERALD, JR. Documentation of annual growth lines in 
ocean quahogs, Arctica islandico Linne 1 

BOWMAN, RAY E. Food of silver hake, Merluccius bilinearis 21 

LARSON, RALPH J., and EDWARD E. DeMARTINI. Abundance and vertical distribu- 
tion of fishes in a cobble-bottom kelp forest off San Onofre, California 37 

COYER, JAMES A. The invertebrate assemblage associated with the giant kelp, Mac- 
rocystis pyrifera, at Santa Catalina Island, California: A general description with 
emphasis on amphipods, copepods, mysids, and shrimps 55 

ANTONELIS, GEORGE A., JR., CLIFFORD H. FISCUS, and ROBERT L. 
DeLONG. Spring and summer prey of California sea lions, Zalophus californianus, at 
San Miguel Island, California, 1978-79 67 

GRIS WOLD, CAROLYN A., and THOMAS W. McKENNE Y. Larval development of the 

scup, Stenotomus chrysops (Pisces: Sparidae) 77 

HETTLER, WILLIAM F. Description of eggs, larvae, and early juveniles of gulf 
menhaden, Brevoortia patronus , and comparisons with Atlantic menhaden,B. tyrannus, 
and yellowfin menhaden, B. smithi 85 

BARNETT, ARTHUR M., ANDREW E. JAHN, PETER D. SERTIC, and WILLIAM 
WATSON. Distribution of ichthyoplankton off San Onofre, California, and methods 
for sampling very shallow coastal waters 97 

McGURK, MICHAEL D. Ring deposition in the otoliths of larval Pacific herring, Clupea 

harengus pallasi 113 

MACDONALD, J. STEVENSON, MICHAEL J. DADSWELL, RALPH G. APPY, GARY 
D. MELVTN, and DAVID A. METHVEN. Fishes, fish assemblages, and their seasonal 
movements in the lower Bay of Fundy and Passamaquoddy Bay, Canada 121 

TILSETH, S., and B. ELLERTSEN. The detection and distribution of larval Arcto- 

Norwegian cod, Gadus morhua, food organisms by an in situ particle counter 141 

EWING, R. D., C. E. HART, C. A. FUSTICH, and GREG CONCANNON. Effects of size 
and time of release on seaward migrations of spring chinook salmon, Oncorhynchus 
tshawytscha 157 

CAMPANA, STEVEN E. Interactive effects of age and environmental modifiers on the 
production of daily growth increments in otoliths of plainfin midshipman, Porichthys 
notatus 165 



V 



(Continued on hack cover) 



Seattle, Washington 






Fishery Bulletin 



CONTENTS 



1985 



Vol. 82, No. 1 January 1 984 

ROPES, JOHN W., DOUGLAS S. JONES, STEVEN A. MURAWSKI, FREDRIC M. 
SERCHUK, and AMBROSE JERALD, JR. Documentation of annual growth lines in 
ocean quahogs, Arctica islandico Linne 1 

BOWMAN, RAY E. Food of silver hake, Merluccius bilinearis 21 

LARSON, RALPH J., and EDWARD E. DeMARTINI. Abundance and vertical distribu- 
tion of fishes in a cobble-bottom kelp forest off San Onofre, California 37 

COYER, JAMES A. The invertebrate assemblage associated with the giant kelp, Mac- 
rocystis pyrifera, at Santa Catalina Island, California: A general description with 
emphasis on amphipods, copepods, mysids, and shrimps 55 

ANTONELIS, GEORGE A., JR., CLIFFORD H. FISCUS, and ROBERT L. 
DeLONG. Spring and summer prey of California sea lions, Zalophus californianus, at 
San Miguel Island, California, 1978-79 67 

GRISWOLD, CAROLYN A., and THOMAS W. McKENNE Y. Larval development of the 

scup, Stenotomus chrysops (Pisces: Sparidae) 77 

HETTLER, WILLIAM F. Description of eggs, larvae, and early juveniles of gulf 
menhaden, Brevoortia patronus , and comparisons with Atlantic menhaden, B. tyrannus , 
and yellowfin menhaden, B. smithi 85 

BARNETT, ARTHUR M., ANDREW E. JAHN, PETER D. SERTIC, and WILLIAM 
WATSON. Distribution of ichthyoplankton off San Onofre, California, and methods 
for sampling very shallow coastal waters 97 

McGURK, MICHAEL D. Ring deposition in the otoliths of larval Pacific herring, Clupea 

harengus pallasi 113 

MACDONALD, J. STEVENSON, MICHAEL J. DADSWELL, RALPH G. APPY, GARY 
D. MELVIN, and DAVID A. METHVEN. Fishes, fish assemblages, and their seasonal 
movements in the lower Bay of Fundy and Passamaquoddy Bay, Canada 121 

TILSETH, S., and B. ELLERTSEN. The detection and distribution of larval Arcto- 

Norwegian cod, Gadus morhua, food organisms by an in situ particle counter 141 

E WING, R. D., C. E. HART, C. A. FUSTICH, and GREG CONCANNON. Effects of size 
and time of release on seaward migrations of spring chinook salmon, Oncorhynchus 
tshawytscha 157 

CAMPANA, STEVEN E. Interactive effects of age and environmental modifiers on the 
production of daily growth increments in otoliths of plainfin midshipman, Porichthys 
notatus 165 



(Continued on next page) 

Seattle, Washington 
1984 



For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 
20402 - Subscription price per year $21.00 domestic and $26.25 foreign. Cost per single issue: 
$6.50 domestic and $8.15 foreign. 



( 'ontents — continued 

LOVE, MILTON S., GERALD E. McGOWEN, WILLIAM WESTPHAL, ROBERT J. 
LAVENBERG, and LINDA MARTIN. Aspects of the life history and fishery of the 
white croaker, Genyonemus lineatus (Sciaenidae), off California 179 

MORRIS, PAMELA A. Feeding habits of blacksmith, Chromis punctipinnis , associated 

with a thermal outfall 199 

MYRICK, ALBERT C., JR., EDWARD W. SHALLENBERGER, INGRID KANG, and 
DAVID B. MacKAY. Calibration of dental layers in seven captive Hawaiian spinner 
dolphins, Stenella longirostris, based on tetracycline labeling 207 

ROSS, STEVE W. Reproduction of the banded drum, Larimus fasciatus, in North 

Carolina 227 



Notes 

SCHMITT, P. D. Marking growth increments in otoliths of larval and juvenile fish by 

immersion in tetracycline to examine the rate of increment formation 237 

ENNIS, G. P. Tag-recapture validation of molt and egg extrusion predictions based upon 

pleopod examination in the American lobster, Homarus americanus 242 

ENNIS, G. P. Comparison of physiological and functional size-maturity relationships in 

two Newfoundland populations of lobsters Homarus americanus 244 

ECHEVERRIA, TINA, and WILLIAM H. LENARZ. Conversions between total, fork, 

and standard lengths in 35 species of Sebastes from California 249 



The National Marine Fisheries Service (NMFS) does not approve, recommend or en- 
dorse am 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 ad- 
vertising oi sales promotion which would indicate or imply that NMFS approves, recom- 
mends 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 pro- 
duct to be used or purchased because of this NMFS publication. 



\ 



\ 



\ 






A 



A 



^ 

^ 



\ 



I 



<® 



^ 



% 




Best NMFS Publications for 1982 

The Publications Advisory Committee of the National 
Marine Fisheries Service has announced the best publica- 
tions authored by the NMFS scientists and published in 
the Fishery Bulletin and the Marine Fisheries Review for 
1982. Only effective and interpretive articles which sig- 
nificantly contribute to the understanding and knowledge 
of NMFS mission-related studies are eligible, and the 
following papers were judged as the best in meeting this 
requirement. 




"Development of the vertebral 
column, fins and fin supports, 
branch iostegal rays, and squamation in the 
swordfish, Xiphias gladius" by Thomas 
Potthoff and Sharon Kelley appears in 
Fishery Bulletin 80(2): 161-186. Thomas 
Potthoff, fishery biologist, and Sharon 
Kelley, research assistant, are from the 
Southeast Fisheries Center's Miami 
Laboratory, Miami, Fla. 



"A review of the offshore shrimp 
fishery and the 1981 Texas closure" by 
Edward F. Klima, Kenneth N. Baxter, and 
Frank J. Patella, Jr. appears in Marine 
Fisheries Review 44(9- 10): 16-30. Edward 
F Klima, Director of the Galveston 
Laboratory, Kenneth N. Baxter, 
supervisory fishery biologist, and Frank J. 
Patella, Jr. , fishery biologist, are also from 
the Southeast Fisheries Center but from the 
Galveston Laboratory, Galveston, Tex. 



Bo 




r$ 



£U 



DOCUMENTATION OF ANNUAL GROWTH LINES IN 
OCEAN QUAHOGS, ARCTICA ISLANDICA LINNE 

John W. Ropes, 1 Douglas S. Jones, 2 Steven A. Murawski, 1 
Fredric M. Serchuk, 1 and Ambrose Jearld, Jr. 1 

ABSTRACT 

About 42,000 ocean quahogs, .Arcif'ca islandica Linne, were marked and released at a deep (53 m) oceanic site 
off Long Island, New York, in 1978. Shells of live specimens recovered 1 and 2 years later were radially sec- 
tioned, polished, and etched for preparation of acetate peels and examination by optical microscopy or micro- 
projection; selected specimens were similarly prepared for examination by scanning electron microscopy. 
Specific growth line and growth increment microstructures are described and photographed. An annual 
periodicity of microstructure is documented, providing a basis for accurate age analyses of this commercially 
important species. 



Numerous bivalve species form periodic growth lines 
in their shells (Rhoads and Lutz 1980). Internal 
growth lines found in the shells of ocean quahogs, 
Arctica islandica Linne, have stimulated interest in 
using these markings to determine age and growth 
(Thompson et al. 1980a, b), since fishery exploitation 
has increased significantly within the past decade 
(Serchuk and Murawski 1980 1 ). 

Documentation of age and growth of ocean quahogs 
has been incomplete. Some studies included no 
account of aging methodologies (Thorson in Turner 
1949; Jaeckel 1952; Loosanoff 1953; Skuladottir 
1967); in others, concentric "rings" or "bands" 
formed in the periostracum of small quahogs (<ca. 60 
mm in shell length) were considered annuli, but 
validation of the annual periodicity of these markings 
was not provided (Loven 1929; Chandler 1965; 
Caddy et al. 1974; Chene 1970 4 ; Meagher and Med- 
cof 1972 5 ). Microstructure of ocean quahog shells 
has been studied, but the analyses did not 
specifically distinguish growth lines from growth 
increments (Sorby 1879; Btfggild 1930; Taylor et al. 
1969, 1973; Lutz and Rhoads 1977, 1980). A means 



■Northeast Fisheries Center Woods Hole Laboratory, National 
Marine Fisheries Service, NOAA, Woods Hole, MA 02543. 

department of Geology, University of Florida, Gainesville, FL 
32611. 

3 Serchuk, F.M. and S.A. Murawski. 1980. Evaluation and status 
of ocean quahog, Arctica islandica (Linnaeus) populations off the 
Middle Atlantic Coast of the United States. U.S. Dep. Commer., 
NOAA, NMFS, Woods Hole Lab. Doc. 80-32, 4 p. 

*Ch6ne\ P.L. 1970. Growth, PSP accumulation, and other 
features of ocean quahog (Arctica islandica). Fish Res. Board Can., 
St. Andrews Biol. Stn., Orig. Manuscr. Rep. 1104, 34 p. 

5 Meagher, J.J., and J.C. Medcof. 1972. Shell rings and growth 
rate of ocean clams (Arctica islandica). Fish Res. Board Can., St. 
Andrews Biol. Stn., Orig. Manuscr. Rep. 1105, 26 p. 



Manuscript accepted July 1983. 

FISHERY BULLETIN: VOL. 82, NO. 1, 1984. 



of clearly separating such shell features was 
needed. 

Recent investigators of age phenomena in ocean 
quahogs have microscopically examined the shells 
and acetate peel images produced from sectioned, 
polished, and etched shells. This method greatly 
aided separating the many crowded growth layers in 
the hinge plate and near the ventral valve margin of 
large, old specimens. Lutz and Rhoads (1977) found 
alternating bands of aragonitic prisms and complex- 
crossed lamellar microstructures in the inner shell 
layer of ocean quahog shells that they believed were 
related to periods of aerobic and anaerobic respira- 
tion. Thompson et al. (1980a, b) reported that inter- 
nal growth bands corresponded to external checks on 
the valves and that the internal growth bands were 
formed by successive deposition of two repeating 
growth layers or increments. Jones (1980) labelled 
the growth increments (GI) as GI I and GI II, since 
each was microstructurally distinct, had thickness, 
and was formed within a time frame of several months. 
For these reasons, he considered the GI I layer to be 
unlike minute "growth lines" or "striations" appear- 
ing as subdaily deposits in the shells of other bivalves 
(Gordan and Carriker 1978); the GI II layer became 
thinner and ill-defined from the GI I layer with 
ontogeny. 

Since growth bands in ocean quahog shells seem to 
lack microstructures of possible subannual peri- 
odicities, the definitions of a growth line and growth 
increment formulated by Clark (1974a, b) have 
general application. Clark (1974b:l) defined the for- 
mer as "abrupt or repetitive changes in the character 
of an accreting tissue" and the latter as "the thick- 

1 



ness or volume of tissue formed by accretionary 
growth between successive growth lines." In fact, 
Jones (1980:333) identified the layers as a "consist- 
ently thin, dark gray, translucent increment" of pris- 
matic microstructures which was "easily distin- 
guished from" homogeneous and crossed micro- 
structural layers. 

Assessment research on ages of ocean quahogs 
requires accurate counts and measurements of 
growth increments. Age observations are cus- 
tomarily made of acetate peel images under optical 
microscopes with transmitted light. An important 
assessment requirement is that an annual increment 
has a distinct beginning and end. The concept of a 
growth line forming between successive growth 
increments fulfills that requirement. 

Counts of supposed annual growth bands seen in 
the shells of ocean quahogs by Thompson et al. 
(1980a, b) resulted in slow growth rates and an 
extreme longevity estimate of 150 yr. Slow growth 
rate and suspiciously long life for a bivalve seemed to 
invalidate the thesis of only a single growth line and 
growth increment being formed annually. Supportive 
evidence included finding similar bands in surf 
clams; finding a low number of bands formed during 
the onset of sexual maturity that were not explained 
by less than an annual frequency; finding an expected 
number of bands in small specimens of known age; 
finding an expected number of bands formed 
sequentially in samples taken frequently during 2 yr 
that had only an annual periodicity; finding a line 
deposited during the fall-winter, a period coinciding 
with spawning; and finding ages determined by 
radiometric analyses that were comparable with 
band counting. The latter three types of investigation 
have been expanded by Jones (1980) and Turekian et 
al. (1982) with the same results. As part of the study, 
I. Thompson (pers. comm.) marked and released 
ocean quahogs in the natural environment, but none 
was recovered. Direct and readily comprehended 
observations of shell growth after marking were con- 
sidered to be important additional evidence in sup- 
port of the thesis of an annual periodicity of growth 
line and growth increment deposition. 

In 1978, the National Marine Fisheries Service 
marked large numbers of ocean quahogs for release 
and recovery at a site 53 m deep and 48 km south- 
southeast of Shinnecock Inlet, Long Island, N. Y., (lat. 

10 21'N,long.72 24'W). Detailsof this project have 
been reported in Murawski et al. (1982). Periodicity 
of growth line formation and shell accretion after 
notching of recovered ocean quahogs and the micro- 
structure of unmarked and marked shells are de- 
scribed herein with photographic documentation. 



FISHERY BULLETIN: VOL. 82, NO. 1 

METHODS 

A commercial clam dredge vessel, the MV Diane 
Maria, was chartered for the marking operation dur- 
ing 25 July to 5 August 1978. The knife of the hy- 
draulic dredge was 2.54 m wide, and the cage was 
lined with 12.7 mm square-mesh hardware cloth to 
retain small clams. Ocean quahogs for marking were 
collected within 9 km of the planting site and released 
during a 10-d period (ca. 17,000 on 26 July, 3,000 on 
2 August, and 21,000 on 4 August 1978). Two 0.7 
mm thick carborundum discs, spaced 2 mm apart 
and mounted in the mandrel of an electric grinder, 
produced distinctive parallel, shallow grooves from 
the ventral margin up onto the valve surface (Ropes 
and Merrill 1970). Four operators of grinders marked 
about 1,600 clams/h. Groups (ca. 3,000-8,000) of 
marked clams were released at loran-C coordinates 
within a rectangular area of about 3 by 6 /as. 

Marked clam recoveries were made in conjunction 
with annual clam resource surveys. During recovery 
operations, a Northstar 6000 6 loran-C unit and Epsco 
loran-C plotter aided in a systematic search of the plant- 
ing site. Marked clam recoveries were highly variable. 
On 20 and 2 1 August 1979 and about 387 d after the 
marking operation, 43 hydraulic dredge tows at the 
planting site captured 14,043 ocean quahogs and 74 
(0.57c ) were marked; on 9 September 1980 and 773 d 
after the marking operation, 1,899 ocean quahogs 
were captured in 2 dredge tows and 249 (13.1%) were 
marked. Some marked specimens were damaged, 
but 67 recovered in 1979 and 200 recovered in 1980 
were alive and had intact paired valves. 

Recaptured specimens were frozen to prevent 
periostracum loss from drying and to facilitate open- 
ing without shell damage. Microscopic examination 
of 267 notched shells was made to assess the effects 
of marking and to obtain growth measurements. 
Shell measurements were made to the nearest 0.1 
mm by calipers; growth after marking was measured 
to the nearest 0.01 mm by an ocular micrometer in a 
dissecting microscope. Acetate peels of all marked 
ocean quahogs were prepared by the procedures de- 
scribed by Ropes (198 2 7 ). Briefly, radial sect ions from 
the umbo to ventral margin were produced on left 
valves, oriented to include the broadest surface of the 
single prominent tooth in the hinge. This exposed the 
internal growth lines in the valve and hinge tooth for 
later treatment. The paired notch marks in a valve 



''Reference to trade names does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 

'Ropes, J. W, 1982. Procedures for preparing acetate peels of 
embedded valves of Arctica islandica foraging. U.S.Dep. Commer., 
NOAA, NMFS, Woods Hole Lab., Doc. 82-18, 8 p. 



ROPES KT AL.: GROWTH LINES OF OCEAN Ql'AHOC.S 



were not usually oriented in the above plane of radial 
sections, so additional radial cuts were made to 
expose growth lines in the notch area. Subsequently, 
shells were embedded in an epoxy resin, and the cut 
shell edge ground on wetable carborundum paper 
and polished. Acetate peels were produced after 
etching the shell cut surfaces in a 1 c /\ HC1 solution for 
1 min, followed by microscopic examination of the 
peels that were sandwiched between slides. 

Preparation of polished and etched radial sections 
of marked and unmarked ocean quahog valves were 
further examined by scanning electron microscopy 
(SEM) 8 . These examinations included vertical tran- 
sects from the periostracum to the shell's interior and 
specific sites affected by the marking operation. 
Shell microstructure was diagnosed by using the 
classification scheme of Carter (1980), wherein shell 
microstructures are elucidated on the basis of their 
major (i.e., first-order) structural arrangement, inde- 
pendent of genetic or optical crystallographic criteria. 

RESULTS 
Whole Shells 

Notch marks showed clearly on wet shells but 
periostracum obscured the ventralmost ends extend- 
ing well beyond the ventral valve margin on all 
specimens (Figs. 1-4). Cuts made in the shell-free 
periostracum beyond the ventral margin of some 
large quahogs had not been repaired after 2 yr (Fig. 
4a); small individuals, however, had completely 
formed yellowish-brown periostracum (grayish white 
in the photographs) over new shell growth, which con- 
trasted sharply with darker, earlier deposition (Fig. 
3a). 

In some specimens, the mark formed U-shaped 
notches at the marginal edge of the old shell. New 
shell deposition was obviously disrupted for quahogs 
with deep U-notches, since the marginal shell be- 
tween the notches was outlined in relief over new 
shell (Fig. 3b, c). Faint paired bulges were also found 
on the ventral inner surface of the notched valve of a 
few shells and occasionally the notches extended 
part way onto the opposite valve. An occasional live 
quahog was found with a cracked valve caused from 
handling during the marking operation. The black- 
ened margins of the cracks, suggestive of reducing 
conditions, indicated that the cracks were old. There 



*SEM work was performed on an ETEC Autoscan instrument at 
the Dental Research Center, University of North Carolina. Chapel 
Hill. N.C.: on the JEOLJSM-35 of the Biology Department, Princeton 
University, Princeton, N.J.; and on the ISI 1 200 of the Department 
of Geology, University of Florida, Gainesville, Fla. 



was no evidence of repair by shell covering the cracks 
in quahogs recovered 2 yr after marking. 

Sectioned Shells 

An interruption of shell deposition from notching in 
some sectioned shells was visible without magnifica- 
tion in the cut surfaces. Microscopic examination 
revealed a depression that curved dorsally back into 
the shell from the external surface and became 
increasingly attenuated until it was unrecognizable 
from the usual shell features along the inner margin. 
This type of interruption was greatest in shells of 
small quahogs, probably due to some mantle tissue 
incision. Periostracum penetrated into the interrup- 
tion to a depth of about 1 mm. The thicker and 
tougher periostracum of large clams was less easily 
incised during marking and probably served to 
minimize incision of mantle tissue. 

Acetate Peels of Sectioned Shells 

Acetate peels enhanced detection of interruptions 
in shell deposition due to notching (Figs. Id, 2d, 3d, 
4d). In small clams (<80 mm shell length), the 
interruption was immediately followed by a line 
similar to a succession of lines formed throughout the 
valve before marking. No additional lines were evi- 
dent thereafter to the marginal tip in shells examined 
from the late August 1979 recovery, but in 34 shells 
of small clams recovered in early September 1980, a 
second line occurred about midway to the tip in all 
shells and a third line had formed very near the 
marginal valve tip and along the inner margin in 47% 
(Fig. 3d). These were all considered to be annual 
growth lines for reasons dicussed later. Shell deposi- 
tion between growth lines in the outer layer had a 
granular appearance, which was sometimes broken 
by a faint line of uncertain origin (Figs. Id, 3d). The 
interruption of shell deposition from marking was 
also evident in large ocean quahogs (Figs. 2d, 4d), 
although an infiltration of the depression with perios- 
tracum was not clearly evident. The separation of 
shell deposits was more definite and extended 
deeper into the shell of large clams, sometimes to a 
depth of 2 mm (Figs. 2d, 4d). Growth lines were very 
closely spaced (ca. 100 jura) and the shell 
depositional texture in between lines appeared 
similar to that seen in smaller clams. In large ocean 
quahogs, new shell was formed laterally beyond the 
notch mark and was an indication that the notching 
operation had little effect on shell deposition and 
growth (Figs. 5a, 6a). 

None of the marked quahogs had as severe an 



FISHERY BULLETIN: VOL. 82, NO. 1 




10mm 




5mm 




FIGURE 1.— (a) Right valve of an 1 8-yr-old ocean quahog,Arctica islandica, 71.4 mm in shell length recovered 
on 21 August 1979. Estimated annual growth was 1.3 mm. The notch-mark area is shown before (h) and after 
(c) peeling off the periostracum. (d) Optical photomicrograph (microprojector) of seven repetitive growth 
lines in an acetate peel of the valve margin. An arrow points to an interruption of growth and growth line 
formed at or soon after marking the clam in 1978. An internal line is_evident between the mark-induced line 
and the valve edge. This line is the normally occurring age mark probably formed in late autumn-early winter. 
Scale bars of magnification are included. 



FIGURE 2. — (a) Left valve of an ocean quahog, Arctica islandica, 
about 110 yr old and 100.5 mm in shell length recovered on 20 
August 1979. Estimated annual growth in 1 yr was 0.1 mm. The 
notch-mark area is shown before (b) and after (c) peeling off the 
periostracum. (d) Optical photomicrograph (compound micro- 
scope) of many repetitive growth lines in an acetate peel of the valve 
margin. An arrow points to an interruption of growth and growth line 
formed at or soon after marking the clam. Scale bars of magnification 
are included. 



ROPES ET AL.: GROWTH LINKS OF OCEAN QUAHOGS 




; 



0.1mm d 



FISHERY BULLETIN: VOL. 82, NO. 1 




10mm 




5mm 




FIGURE 3.— (a) Left valve of a 15-yr-old ocean quahog, .4rrf<ea islandica, 61.7 mm in shell length recovered on 9 Sep- 
tember 1980. Estimated 2-yr growth increment was 3.2 mm. The notch-mark area is shown before (b) and after (c) 
peeling off the periostracum. (d) Optical photomicrograph (microprojector) of five repetitive growth lines in an ace- 
tate peel of the valve margin. An arrow points to an interruption of growth and growth line formed at or soon after 
marking the clam. Two additional lines, one midway to the valve tip and one near the valve tip, were formed after 
marking the quahog. Scale bars of magnification are included. 



FIGURE 4. — (a) Left valve of an ocean quahog. Arctica islandica, 
about 95 yr old and 91.7 mm in shell length recovered on 9 Septem- 
ber 1980. Estimated 2-yr growth increment was 0.3 mm. The notch- 
mark area is shown before (b) and after (c) peeling off the 
periostracum. (d) Optical photomicrograph (compound micro- 
scope) of many repetitive growth lines in an acetate peel of the valve 
margin. An arrow points to an interruption of growth and growth line 
formed at or soon after marking the clam. Scale bars of magnification 
are included. 



ROPES ETAL.: GROWTH LINES OF OCEAN QUAHOGS 



a 




10mm 




0.1mm d 



FISHERY BULLETIN: VOL. 82, NO. 1 












Fl( (URE 5 .— (a) Optical photomicrograph (compound microscope) of an acetate peel at a notch-mark in the ocean quahog, 
Arctica islandica, shown in Figure 2 (scale bar = 100 jum). Note the single annual increment of growth formed laterally 
beyond the notch-mark, (b) SEM photomicrograph of the same shell specimen (scale bar = 100 ;iim). Abbreviatons of 
shell microstructural terms in this and subsequent figures are expalined in this text, (c) Transitional CA-CL microstruc- 



ROPES ET Al..: (IROWTH LINKS OF OCEAN QUAHOGS 





ture interrupted by an ISP band that extends from the notch-line seen in (b) (scale bar = 1 ju.m). This photomicrograph 
was taken beyond the field of view of (b), to the lower left and beyond the zone of epoxy penetration, (d) Enlargement of 
epoxy penetration zone from (b) interrupts normal shell microstructure followed by a zone of cavernous, poorly 
organized, disrupted shell growth (scale bar =10 jum). 



FISHERY BULLETIN: VOL. 82, NO. 1 





FIGURE 6.— (a) Optical photomicrograph (compound microscope) of an acetate peel at a notch-mark in a 100.3 mm long 
ocean quahog,Arctua islandica, about 110 yr old recovered on 9 September 1980. Note two annual increments of growth 
formed laterally beyond the notch-mark, estimated to be 0. 1 mm (scale bar = 1 00 /urn), (b) SEM photomicrograph of the 
same specimen (scale bar = 100 fira). The penetration of epoxy medium corresponds to the marking event, (c) Tran- 



10 



ROPES ET AL.: GROWTH LINKS OK OCEAN QIAHOGS 





sitional CA-CL microstructure in the upper left interrupted by the penetration of epoxy into the shell along the line cor- 
responding with the notching event (scale bar = 10 /um). Beyond the zone of epoxy penetration this line is recognized as 
an ISP band. Following this line is a zone of SphP disruption growth which gradually gives way to transitional CA-CL. 
This photomicrograph was taken off the field of view of (b), to the lower left, (d) Disruption growth of SphP micro- 
structure following the line which corresponds to marking of the clam (scale bar = 50 /xm). 



11 



FISHERY BULLETIN: VOL. 82, NO. 1 



interruption in shell deposition as did a 66.4 mm 
shell-length specimen collected during a 1980 winter 
clam survey (Fig. 7 a, b). A depression outlined the 
entire shape of the clam at the time of its formation 
and at 25.9 mm shell length. The shell formed before 
the depression was raised laterally above that formed 
afterwards in a shinglelike fashion. Both valves of the 
clam showed the interruption. The smaller shell 



shape was only slightly atypical for ocean quahogs, 
and no irregularity was found as an indication that an 
injury had occurred. The ratio between the greatest 
shell height (21.4 mm) and shell length (25.9 mm) of 
the smaller shell was 0.826, and for the entire shell 
(50.2 mm shell height; 66.4 mm shell length) was 
0.756. These values suggest that growth after the 
depression departed from the more typical, isomet- 




10mm 



0.1 mm 




FIGURE 7.— (a) Right valve of an ocean quahog, Arctica islandica, 23 yroldand 66.4 mm in shell length collected 
near the marking site during 1980. An obvious interruption of growth (arrow) and radiating lines in the anterior 
half of the valve are shown, (b) Optical photomicrograph (compound microscope) of an acetate peel of the sec- 
tioned valve at the site of growth interruption. Scale bars of magnification are included. 



12 



ROPES ET AL.: GROWTH LINES OF OCEAN Ql'AHOGS 



ric growth reported for ocean quahogs by Murawski 
et al. (1982). Light radial lines extended from the 
umbonal area to valve margin in the periostracum of 
the anterior half of the shell formed after growth had 
been interrupted, but their significance was not 
evident. 

Microstructure of Unmarked Shells 

The ocean quahog shell is entirely aragonitic with an 
inner and outer layer separated by an extremely thin 
prismatic pallial myostracum. The latter is com- 
posed predominantly of irregular simple prisms 
(ISP) and occasionally a few fibrous prisms (FP). 
Both principal shell layers contain two growth sub- 
layers: The thin annual growth line and the wider 
annual growth increment. Significant variations were 
found in the microstructure of each during 
examinations by SEM. 

The distribution of microstructures in a typical 
ocean quahog shell may be seen by considering a 
transect from the exterior to interior depositional 
surfaces. The thick, dark brown or black perios- 
tracum is an obvious exterior surface covering, but it 
is intimately associated with the shell. Some 
aragonitic shell material is invariably removed when 
peeling off the periostracum and granules of 
aragonite were found embedded in it during 
examination of unetched thin sections under the 
crossed nicols of a polarizing microscope. The 
aragonite is dissolved by etching the polished sur- 
faces of sectioned shells, leaving cavities, some with 
angular faces in the periostracum (Fig. 8a). 

Important microstructures for aging purposes are 
found mostly in the outer shell layer. The dominant 
growth increment sublayer beneath the perios- 
tracum exhibits a granular homogeneous (HOM) 
microstructure which is very cavernous and has 
bleblike isolated crystal morphotypes (ICM) (Fig. 8b, 
c). These microstructures typically grade into 
incipient ISP (Fig. 8d). Below the prisms is a layer of 
crossed microstructures which appear to be tran- 
sitional between simple crossed lamellar (CL) and 
crossed acicular (CA) structures (Fig. 8d). The latter 
predominates in the middle portion of the outer shell 
layer with occasional occurrences of fine complex- 
crossed lamellar (FCCL) microstructure. Tran- 
sitional CA-CL microstructures are also seen in 
Figure 8e. 

In the thin growth lines of the outer shell layer, FP 
near the external surface soon give way to very dis- 
tinctive spherultic prisms (SphP) (Fig. 9a-d). These 
SphP themselves grade into composite prisms 
(CompP) which are comprised of first-order prisms 



with the second-order prisms radiating toward the 
depositional surface from a central, longitudinal axis. 
Closer toward the inner shell layer the FP, SphP, and 
CompP microstructures are gradually replaced by 
ISP bands. 

The inner shell layer is characterized by growth 
lines composed of ISP which alternate with growth 
increment bands of FCCL microstructures. The 
hinge plate and tooth region have microstructures 
recognizable as distinct sublayers that are important 
for aging purposes. Here growth lines are construct- 
ed of narrow ISP (Fig. 9e). These alternate with 
growth increment bands that are composed of tran- 
sitional CA-CL, FCCL, and HOM microstructures 
(Fig. 9e). 

In summary, ocean quahog shells are composed 
largely of HOM, CA-CL, and minor amounts of 
FCCL microstructures with prismatic bands of local 
importance. The latter constitute the growth line 
layer; the former the growth increment layer. Figure 
10 is a diagrammatic sketch of the distribution of 
microstructures in the two principal layers of the 
valve of a typical ocean quahog. 

Microstructure of Marked Shells 

Variations in the shell microstructure associated 
with notching and subsequent shell growth of ocean 
quahog specimens were studied by examining the 
ventral margins of six quahogs. The same basic pat- 
tern described for unmarked shells was observed in 
these specimens. Optical and scanning electron 
photomicrographs of two shells illustrate the salient 
features (Figs. 5, 6). 

The notching event in both shells was accompanied 
by a disruption of the normal growth pattern and a 
resumption of shell growth at a new orientation. This 
is seen in Figures 5a and 6a as a prominent flattened 
surface in the exterior shell surface from the marking 
operation followed by a lateral extension of the shell 
margin out beyond the notch mark and old shell sur- 
face. The extension represents renewed growth at a 
new orientation. Retraction of the mantle during the 
marking process and resumption of shell growth at a 
slightly new orientation resulted in a zone of either 
loosely calcified or uncalcified shell paralleling the 
shell margin and extending from the notch inward 
toward the depositional surface. This zone is filled 
with epoxy medium during the embedding process 
and is seen in Figures 5b and 6b, c, and d as the resis- 
tant, unetched material penetrating several milli- 
meters into the outer shell layer. The penetration 
zone disappeared shortly beyond the field of view in 
Figures 5b and 6b. Where this happened (Fig. 5c), a 



13 



FISHERY BULLETIN: VOL. 82. NO. 1 




FIGURE8. — Microstructural variation in the shell of a 97.5 mm long, 92-yr-old ocean quahog, Arctica islandica. The central photograph is of an 
acetate peel from a radial, polished, and etched section through the valve at the ventral margin. Black lines locate SEM photomicrographic 
enlargements of specific polished and etched areas in the section valve (scale bar = 1 mm), (a) Periostracum ("P") with angular cavities and 
HOM microstructure (scale bar = 10 /xm). (b) Cavities in HOM microstructure of outermost valve surface. White rectangle encloses are 
enlarged in (c) (scale bar = 10 /mm), (c) Cavities showing bleblike ICM structures (scale bar = 1 jtim). (d) HOM microstructure at top trending to 
ISP, then to transitional CA-CL toward the bottom. White rectangle encloses the area enlarged in (e) (scale bar= lOjum). (e) Transitional CA- 
CL microstructure at higher magnification (scale bar = 10 ju.m). 



14 



ROPES ET AL.: GROWTH LINES OF OCEAN QUAHOGS 



^A GL 




FIGURE 9. — Microstructural variation in the shell of ocean quahog, Arctica i.slandica, continued from Figure 8. (Central photograph scale bar = 
1 jam), (a) Annual growth line sublayer (region between upper and lower-most dark bands) in outer shell layer, located beneath the CA-CL mi- 
crostructure of Figure 8e. White rectangle encloses the area enlarged in ( b) ( scale bar =10 jam) . (b) The center of the growth line sublayer show- 
ing FP, SphP, and CompP microstructures (scale bar = 10 jam), (c) Transitional CA-CL microstructure of the growth increment sublayer below 
(b) (scale bar = 10 jam), (d) ISP forming growth line sublayer within FCCL microstructure. Photomicrograph from area closer to the 
deposit ional surface than previous photos (scale bar = 1 jam), (e) HOM microstructure interrupted by two prominent ISP bands from a section 
through the hinge plate (scale bar = 10 jam). 



15 



FISHERY BULLETIN: VOL. 82, NO. 1 



Local ICM 



Outer 
Shell 
Layer 



Inner 
Shell 
Layer 




Periostracum 



Pallial 
Myostracum 



FIGURE 10.— Idealized, partial, radial cross section through the shell of a typical ocean quahog, Arctica islandica, 
showing the distribution of shell microstructures. Ventral margin is toward the left. Section is located inside the 
pallial line. Legend of acronyms: ICM= isolated crystal morphotypes; FP = fibrous prisms; SphP = spherulitic 
prisms; CompP = composite prisms; ISP = irregular simple prisms; HOM = granular homogeneous; CA = 
crossed acicular; CL = crossed lamellar; FCCL = fine complex crossed lamellar. 



growth line of ISP continued parallel to the earlier 
growth lines. 

The typical outer shell layer structure formed by the 
time of marking is noted in Figures 5b and 6b. These 
figures clearly show the alternation of the growth line 
and growth increment sublayers. However, im- 
mediately following the marking event all specimens 
showed a disruption in microstructural development, 
especially out near the shell surface. This coincided 
with the presence of loosely organized SphP 
immediately following the marking event line (see 
particularly Fig. 6c, d). The disruption zone con- 
sisted of cavernous, poorly organized, microstruc- 
ture (Fig. 5d). 

In all six shells examined, the growth line associated 
with the marking event continued inward toward the 
shell interior, even beyond the zone of epoxy penetra- 
tion (Figs. 5b, 6b). When this line was traced inward, 
it was indistinguishable from the many earlier formed 
growth lines. Such a view is seen in Figure 5c located 
well off the field of view Figure 5b, to the bottom left. 
Here the growth line consistedof a diagonal ISP band 
bounded on both sides by transitional CA-CL 
microstructure. 

DISCUSSION 

The layering and separation between growth lines 
and growth increments of small ocean quahogs (< ca 



60 mm shell length) are often visible macroscopically 
on the external surfaces of whole valves and in the cut 
surfaces of radial sections. However, macro- or mi- 
croscopic examinations of large ocean quahog 
valves are consistently frustrated by a lack of clear 
differentiation of the same growth phenomena. Pre- 
paration of acetate peels of shell cross sections, as 
has been described and photographically docu- 
mented, greatly enhances discrimination of the lines 
and increments of growth throughout the range of 
shell sizes. 

Past investigators of the microstructure of ocean 
quahog shells described some of the basic com- 
ponents, but did not clearly elucidate differences 
between the lines and increments of growth. Sorby 
(1879: 62) appeal's to have given the first description 
of the structure of the Arctica {= Cyprina) islandica 
shell: "In C'yprina islandica we have another extreme 
case, in which the fibres perpendicular to the plane of 
growth are so short as to appear like granules, though 
the optic axes are still definitely oriented in the nor- 
mal manner." B<2>ggild (1930:286) reported that he 
was unable to confirm Sorby's observations. Instead 
he stated that Arctica islandica belongs to a group of 
species within the Arcticidae (= Cyprinidae) having 
the least visible structure among all the bivalves. He 
terms this structure homogeneous but suggests there 
are small traces of other structures in the shell. 
Boggild (1930) goes on to point out that the lower 



16 



ROPES ET AL.: GROWTH LINES OF OCEAN OIAHOGS 



part of the shell (inner layer) is perhaps more 
"... representative of the common, complex struc- 
ture . . . and . . . there are alternating layers of more 
transparent layers and finely grained ones." More 
recently Taylor etal. (1969, 1973) examined the shell 
microstructure ofArctica islandica, which they adopt- 
ed as their "type species" to illustrate homogeneous 
shell microstructure. Basically, the general picture 
by Btfggild (1930) agrees with that of Taylor et al. 
(1969), who used electron microscopy in their inves- 
tigation. However, they disagreed sharply with 
B^ggild that the inner shell layer was "representative 
of the common complex structure." After examining 
unetched fractured sections and polished and etched 
sections of both shell layers, Taylor et al. (1969, 
1973) concluded that both shell layers in Arctica 
islandica are built of minute, irregular rounded 
granules, quite variable in size (1.5-3 fim across), 
having highly irregular contacts with their neighbors 
and being poorly stacked. Taylor et al. (1969:51) 
further reported: "In peels and sections of the inner 
layer, within the pallial line there is a marked colour 
banding, in greys and browns. The only fine structure 
that can be resolved is a suggestion of minute grains, 
which are most conspicuous in the translucent, 
grey-colourless parts of the shell. These grains are 
arranged in sheets parallel to the shell interior. In the 
outer layer grains can also be resolved, but are 
arranged in sheets parallel to the margin of the shell 
and growth lines." They also noted that these 
features are more clearly seen in the umbonal region 
where the orientation of grains normal to layering is 
very conspicuous. Taylor et al. (1969) suggested that 
the layering is a reflection of repeated (?diurnal) 
deposition of carbonate (a prospect deemed very 
unlikely by Thompson et al. 1980a). Also in the 
umbonal region are thin (2-3 ju.m) prismatic bands 
which parallel the layering. Outside the pallial line, 
Taylor et al. (1969) reported the outer shell layer to 
be very dense and opaque, with the most obvious 
structural features being fine grains arranged in 
sheets giving the layer a finely banded appearance. 
Analyses under SEM of oriented fractured, and 
polished and etched sections of ocean quahog shells 
revealed that microstructural variation is more com- 
plex than had been proposed by Btfggild (1930) or 
Taylor et al. (1969, 1973). Thin sections of isolated 
periostracal fragments examined under crossed 
nicols confirmed the presence of embedded 
aragonite granules in the periostracum of ocean 
quahogs reported for other recent bivalves (Carter 
and Aller 1975). These granules probably form a 
layer like that described for the blue mussel, Mytilus 
edulis, by Carriker (1979). After special treatment of 



the valves for examination by SEM, he found "a thin 
discrete calcareous layer continuous over the outer 
surface of the valves between the periostracum and 
the outermost shell layer." The layer is called mosaio- 
stracum. The shell microstructure in the growth incre- 
ment sublayer beneath the periostracum is HOM, as 
B0ggild (1930) and Taylor et al. (1969, 1973) re- 
ported. The "... minute, irregular, rounded gran- 
ules . . . have highly irregular contacts ..." (Taylor 
et al. 1969:51) that are particularly well exposed in 
fracture sections. An abundant transitional CA- 
CL microstructure was found in the middle portion 
of the outer shell layer and growth increment sub- 
layer. This study confirmed its presence in ocean 
quahogs as reported by Carter (1980). The growth 
line sublayer of the outer shell layer had four 
grades of prismatic structure (FP, SphP, ComP, and 
ISP). Lutz and Rhoads (1977) examined the inner 
shell layer near the umbo of ocean quahogs and found 
bands of simple aragonitic prisms alternating with 
complex-crossed lamellar and homogeneous struc- 
tures. We found similar microstructures in the inner 
shell layer of the valve of ocean quahogs. Our 
analyses identified distinct microstructures, not 
unlike those found in the valve for the growth line and 
growth increment layers in the hinge plate. 

Growth line deposition more nearly approximates 
an annual event than any shorter or longer interval. 
Marked clams recovered in late August 1979 had 
formed only one growth line other than the mark- 
induced check soon after the notching operation in 
1978. They had been free about 22 d longer than a 
calendar year. Those recovered in early September 
1980 all had formed the growth line soon after the 
notching operation, like those recovered in 1979, and 
a second line appeared midway to the ventral valve 
edge, which in all probability had been formed after 
the late August 1979 recovery effort. These clams 
were free about 33 d more than 2 calendaryears since 
the notching operation. A feature of the specimens 
recovered in 1980 was that about half had formed a 
third line very near the ventral valve edge and along 
the inner margin. All of the narrow growth lines were 
separated by relatively even, broad areas of growth 
increment deposits suggestive of no more or less than 
an annual interval for the deposition of growth lines, 
even though the time of formation of such lines may 
not correspond to an exact number of calendar days. 
These observations confirm similar conclusions of an 
annual periodicity of growth line formation by 
Thompson and Jones (1977), Thompson et al. 
(1980a, b), and Jones (1980). 

Radiometric techniques for aging bivalve shells 
have recently been applied to ocean quahogs. 



17 



FISHERY BULLETIN: VOL. 82, NO. 1 



Thompson et al. (1980a) reported that the predicted 
radiometric age of an ocean quahog having 22 bands 
corresponded exactly to 22 yr when aged using 228 Ra. 
Turekian et al. (1982) concluded that age deter- 
minations of ocean quahogs from radiometric 
analyses are compatible with counts of bands formed 
annually. Thus, radiometric studies support the con- 
tention of an annual periodicity of growth lines in 
ocean quahogs. 

Various environmental disturbances have been 
implicated in the formation of shell abnormalities 
and atypical growth lines in other bivalve species 
(Weymouth et al. 1925; Shuster 1957; Merrill et al. 
1966; Clark 1968; Palmer 1980). It is therefore, con- 
ceivable that the stress imposed by dredging, mark- 
ing, and returning the ocean quahogs to the ocean 
floor and their burrowing activities hastened the for- 
mation of a growth line in 1978. Thereafter, natural 
events affecting the metabolism of shell deposition 
are more likely stimuli. Such events apparently did 
not occur during the period after the formation of the 
growth line in 1978 and recovery of clams in late 
August 1979. Instead a growth line that in all prob- 
ability had formed in 1979 was found in the shells of 
clams recovered on 9 September 1980. Its formation 
may have occurred in late August 1979, but the third 
line found in half of the clams recovered on 9 Septem- 
ber 1980 suggests the possibility of its formation in 
early September 1980. By inference, then, growth 
line formation in 1979 and 1980 occurred in 
September. 

The reported life span (150 yr, Thompson et al. 
1980a) of ocean quahogs surpasses similar estimates 
for other bivalves. Age and growth of the far east 
mussel, Crenomytilus grayanus, have been deter- 
mined from examinations of shell structure, an 
oxygen-isotope method, and notching experiments 
(Zolotarev 1974; Zolotarev and Ignat'ev 1977; 
Zolotarev and Selin 1979). These investigations 
indicated that longevity of the mussel may exceed 
100 yr. Turekian et al. (1975) proposed a longevity of 
about 100 yr for a deep-sea nucoloid, Tindaria callis- 
tiformis, after determining ages by radiometric 
means and counting regularly spaced bands in the 
shell of one of the largest (8.4 mm in shell length). It 
seems likely that longevity of ocean quahogs may 
exceed 150 yr. Murawski and Serchuk (1979) report- 
ed a maximum shell length of 131 mm for ocean 
quahogs in extensive collections taken from the Mid- 
dle Atlantic Bight. A specimen of this size is half 
again as large as the 88 mm example of a 149-yr-old 
ocean quahog reported by Thompson et al. 
(1980a). 

In conclusion, the foregoing description of annual 



growth line formation in marked ocean quahogs and 
analyses of growth in the same specimens by 
Murawski et al. (1982) present significant supporting 
evidence for the hypothesis of slow growth and a long 
life span in the species. Ocean quahogs apparently 
live longer than any other bivalve known to man. 

ACKNOWLEDGMENTS 

We thank Brenda Figuerido and John Lamont for 
their assistance in preparing the art work and 
photographs, and Ida Thompson, University of Edin- 
burgh, Department of Geology, King's Building, 
Edinburgh EH 9 3JW, Scotland, for encouragement 
in undertaking the study and helpful comments on 
the manuscript. 



LITERATURE CITED 

B0GGILD, O. B. 

1930. The shell structure of mollusks. K. Dan. Vidensk. 
Selsk. SK (Copenhagen) 2:231-325. 
Carriker, M. R. 

1979. Ultrastructure of the mosaicostracal layer in the shell 
of the bivalve Mytilus cdulis. Veliger 21:41 1-414. 

Caddy, J. F., R. A. Chandler, and D. G. Wilder. 

1974. Biology and commercial potential of several underex- 
ploited molluscs and crustaceans on the Atlantic coast of 
Canada. Proceeding of a Symposium on the Industrial 
Development Branch of Environmental Canada, Mon- 
treal, Feb. 5-7, 1974, 111 p. (Prepared at Fish. Res. 
Board St. Andrews Biol. Stn., N.B.) 

Carter, J. G. 

1980. Guide to bivalve shell microstructures. In D. C. 
Rhoads and R. A. Lutz (editors), Skeletal growth of 
aquatic organisms, p. 645-674. Plenum Press, N.Y. 

Carter, J. G., and R. C. Aller. 

1975. Calcification in the bivalve periostracum. Lethaia 
8:315-320. 

Chandler, R. A. 

1965. Ocean quahog resources of Southeastern Northum- 
berland Strait. Fish Res. Board Can., Manuscr. Rep. 
(Biol.) 828, 9 p. 
Clark, G. R., II. 

1968. Mollusk shell: Daily growth lines. Science (Wash., 

D.C.) 161:800-802. 
1974a. Growth lines in invertebrate skeletons. Annu. Rev. 

Earth Planet. Sci. 2:77-99. 
1974b. The Paleoperiodicity Newsletter. Vol. 1, p. 1-2. 
Gordon, J., and M. R. Carriker. 

1978. Growth lines in a bivalve mollusk: Subdaily patterns 
and dissolution of the shell. Science (Wash., D.C.) 
202:519-521. 
Jaeckel, S., Jr. 

1952. Zur Oekologie der Molluskenfauna in der westlichen 
Ostsee. Schr. Naturwiss. Ver. Schleswig-Holstein 26:18- 
50. 
Jones, D. S. 

1980. Annual cycle of shell growth increment formation in 
two continental shelf bivalves and its paleoecologic 
significance Paleobiology 6:331-340. 



18 



ROPES ET AI. .: GROWTH LINES OF OCEAN QUAHOGS 



LOOSANOFF, V. L. 

1953. Reproductive cycle in Cyprina islandica. Biol. Bull. 
(Woods Hole) 104:146-155. 
LOVEN, P. M. 

1929 Bietrae zur Kenntnis der Cyprina islandica L. in Ore- 
sund (Contributions to the knowledge of ( 'yprina islandica 
L. in the Oresund). K. Fysiogr. Sallsk. Lund Handl. N.F. 
41:1-38. [Transl. by Lang. Serv. Div., Off. Int. Fish., 
NMFS, U.S. Dep. Commer., Wash., D.C.] 
LUTZ, R. A., and D. C. Rhoads. 

1977. Anaerobiosis and a theory of growth line for- 
mation. Science (Wash., D.C.) 198:1222-1227. 

1980. Chapter 6. Growth patterns within the molluscan shell. 
An overview In D. C. Rhoads and R. A. Lutz (editors), 
Skeletal growth of aquatic organisms, p. 203-254. 
Plenum Press, N.Y. 
Merrill, A. S., J. A. Posgay, and F. E. Nichy. 

1966. Annual marks on shell and ligament of sea scallop 
Placopecten magellanicus. U.S. Fish Wildl. Serv., Fish. 
Bull. 65:299-311. 

Mi rawski, S. A., J. W. Ropes, and F. M. Serchuk. 

1982. Growth of the ocean quahog, Arctica islandica, in the 
Middle Atlantic Bight. Fish. Bull., U.S. 80:21-34. 

Ml IRAWSKl, S. A„ AND F. M. SERCHl'K. 

1979. Shell length-meat weight relationships of ocean 
quahogs, Arctica islandica, from the Middle Atlantic 
Shelf. Proc. Natl. Shellfish. Assoc. 69:40-46. 

Palmer, R. E. 

1980. Observations on shell deformities, ultrastructure, and 
increment formation in the bay scallop Argopecten 
irradians. Mar. Biol. (Berl.) 58:15-23. 

Rhoads, D. C, and R. A. Lutz (editors). 

1980. Skeletal growth of aquatic organisms. Plenum Press, 
N.Y., 750 p. 
Ropes, J. W., and A. S. Merrill. 

1970. Marking surf clams. Proc. Natl. Shellfish. Assoc. 
60:99-106. 
Shuster, C. N., Jr. 

1957. On the shell of bivalve mollusks. Proc. Natl. Shellfish. 
Assoc. 47:34-42. 
Skuladottir, U. 

1967. Krabbadyr og skeldyr (Crustaceans and mollusks). 
Radstefna Isl. Verkfraedinga. 52:13-23. Proceedings of 
the Conference of Islandic Professional Engineers. 
[Transl: Transl. Bur., Fish. Res. Board Can., Biol. Stn., St. 
Andrews, N.B., No. 1206.] 

SORBY, H. C. 

1879. Address on the structure and origins of limestones. Q. 
J. Geol. Soc. Lond. 35:56-95. 
Taylor, J. D., W. J. Kennedy, and A. Hall. 

1969. The shell structure and mineralogy of the Bivalvia. 



Introduction. Nuculacea-Trigonacea. Bull. Br. Mus. 
(Nat. Hist.) Zool. Suppl. 3:4-125. 

1973. The shell structure and mineralogy of the Bivalvia. II. 
Lucinacea - Clavagellacea. Conclusions. Bull. Br. Mus. 
(Nat, Hist.) Zool. 22:255-294. 

Thompson, I., and D. S. Jones. 

1977. The ocean quahog, Arctica islandica, "tree" of the 
North Atlantic shelf. [Abstr.| Geol. Soc. Am. 9:1 199. 
Thompson, I., D. S. Jones, and D. Dreibelbis. 

1980a. Annual internal growth banding and life history of the 
ocean quahog Arctica islandica (Mollusca: Bivalvia). 
Mar. Biol. (Berl.) 57:25-34. 
Thompson, I., D. S. Jones, and J. W. Ropes. 

1980b. Advanced age for sexual maturity in the ocean quahog 

Arctica islandica (Mollusca: Bivalvia). Mar. Biol. (Berl.) 

57:35-39. 

Turekian, K. K., J. K. Cochran, D. P. Kharkar, R. M. Cerrato, 

J. R. Valsnys, H. L. Sanders, J. F. Grassle, and J. A. 

Allen. 

1975. Slow growth rate of a deep-sea clam determined by 
228 Ra chronology. Proc. Natl. Acad. Sci. 72:2829-2832. 
Turekian, K. K., J. K. Cochran, Y. Nozaki, I. Thompson, and D. 
S. Jones. 

1982: Determination of shell deposition rates of Arctica islan- 
dica from the New York Bight using natural " 8 RA and 
228 Th and bomb-produced 14 C. Limnol. Oceanogr. 27: 
737-741. 
Turner, H. J., Jr. 

1949. The mahogany quahog resource of Massachusetts. In 
Report on investigations of improving the shellfish re- 
sources of Massachusetts. Commonw. Mass. Dep. Con- 
serv., Div. Mar. Fish., p. 12-16. 
Weymouth, F. W., H. C. Mcmillin, and H. B. Holmes. 

1925. Growth and age at maturity of the Pacific razor clam, 
Siliqua patula (Dixon). U.S. Dep. Commer., Bur. Fish. 
Doc. 984, p. 201-236. 
ZOLOTAREV, V. N. 

1974. Determination of age and growth rate of the far east 
mussel, Crenomytilus grayanus (Dunker), from its shell 
structure. Dokl. Akad. Nauk SSSR 216(5): 1 195-1 197. 
ITransl. by Plenum Publ. Corp., N.Y., 1974, p. 308- 
309.] 

Zolotarev, V. N., and A. V. Ignatev. 

1977. Seasonal changes in the thickness of the main layers 
and temperature growth of the marine molluscan 
shells. Biol. Morya 5:40-47. |Transl. by Pelnum Publ. 
Corp., N.Y., 1978, p. 352-358.] 
Zolotarev, V. N., and N. I. Selin. 

1979. The use of tags of shells to determine growth of the 
mussel Crenomytilus grayanus. Biol. Morya 1:77-79. 
(Transl. by Plenum Publ. Corp., N.Y.. 1979, p. 58-59.) 



19 



FOOD OF SILVER HAKE, MERLUCCIUS BILINEARIS 



Ray E. Bowman 1 



ABSTRACT 

Stomach contents of 2,622 silver hake collected in the Northwest Atlantic have been analyzed. Fish were 
collected on bottom trawl surveys conducted from 1973 to 1976. The mean fish fork length (FL) was 20 cm 
and the average stomach content weight was 1.5 g. Silver hake <20 cm FL prey mostly on amphipods, 
decapod shrimp, and euphausiids. Fish 20 cm FL and longer take increasing proportions of fish and squid as 
part of their diet. Stomach contents of male and female fish of similar size indicate that females eat larger 
quantities of food (particularly more fish) than the males. The females are also, on the average, longer than 
the males. Silver hake feed primarily at night. Feeding begins near dusk and continues until just after mid- 
night. In the spring a second feeding period seems to occur near noon. Silver hake feed intensively during 
spring. Their stomachs contain almost twice as much food in spring as they do in autumn. Significant dif- 
ferences were noted in the intensity of feeding between areas. Stomachs of fish, caught in the Middle Atlantic, 
contain the largest quantities of food. The species of prey taken by silver hake are highly variable and likely 
reflect prey availability during different years and seasons in various areas. When silver hake spawn, their 
dietary intake is reduced. The diet of fish taken in deep water (> 150 m) is mostly euphausiids and squid, and 
the quantity of food found in their stomachs is less than that in stomachs taken from fish collected at depths 
<150m. 



Silver hake, Merluccius bilinearis (Mitchill 1814), is a 
Northwest Atlantic gadiform fish whose range ex- 
tends from continental shelf waters off South Caro- 
lina to the Newfoundland Banks. It is most abundant in 
offshore waters extending from New York to Cape 
Sable, Nova Scotia (Bigelow and Schroeder 1953). 

Previous investigations have shown that large silver 
hake eat mostly fish and/or squid, while smaller silver 
hake feed on euphausiids, amphipods, and decapod 
shrimp. Among the first to report these findings were 
Nichols and Breder (1927), who noted 75 herring 
about 7 cm long in the stomach of a 59 cm fish. 
Bigelow and Schroeder (1953) reported that silver 
hake are extremely voracious and will prey on smaller 
silver hake or any other of the schooling fishes such as 
young herring, mackerel, menhaden, alewives, or silver- 
sides. Evaluation of other studies on the diet of silver 
hake caught in various areas and during different 
years establishes that the prey of silver hake is very 
predictable in that it is usually comprised of a variety 
of fish, squid, and crustaceans (Jensen and Fritz 
1960; Schaefer 1960; Vinogradov 1972; Noskov and 
Vinogradov 1977; Bowman and Langton 1978; Lang- 
ton and Bowman 1980). Investigations by Swan and 
Clay (1979), Edwards and Bowman (1979), and Bow- 
man and Bowman (1980) have shown that silver hake 
feed mostly at night. 

Until recently the potential impact of silver hake on 



'Northeast Fisheries Center Woods Hole Laboratory, National 
Marine Fisheries Service, NOAA, Woods Hole, MA 02543. 



Manuscript accepted July 1983. 

FISHERY BULLETIN: VOL. 82, NO. 1, 1984. 



the Northwest Atlantic ecosystem had not been de- 
termined. Edwards and Bowman (1979) estimated 
the annual consumption of the principal predators in 
the Northwest Atlantic. They concluded that silver 
hake alone could potentially consume almost 10% of 
the standing crop of all fish within the study area an- 
nually, the bulk of which would be small or juvenile 
fish. They suggested that silver hake, more than any 
other species, plays the principal predatory role in 
regulating the Northwest Atlantic ecosystem. The 
purpose of this report is to document the quantities 
and types of food eaten by silver hake during the 
years 1973-76, and further, to identify feeding trends 
which may be of consequence when attempting to 
precisely determine silver hake's impact on other fish 
populations. 

METHODS AND MATERIALS 

A total of 325 samples from 2,622 silver hake 
stomachs was collected during eight MARMAP (Mar- 
ine Resources Monitoring, Assessment, and Predic- 
tion) bottom trawl survey cruises conducted by the 
National Marine Fisheries Service during spring and 
fall 1973-76 (Table 1). The cruise periods were as 
follows: 16 March-15 May 1973; 26 September-20 
November 1973; 12 March-4 May 1974; 20 Sep- 
tember-14 November 1974; 4 March-12 May 1975; 
15 October- 18 November 1975; 4 March-8 May 
1976; 20 October-23 November 1976. On spring 
cruises a two-seam modified Yankee No. 4 1 trawl was 

21 



FISHERY BULLETIN: VOL. 82. NO. 1 



TABLE 1.— Number of silver hake stomachs examined from each 
geographic area by year and season. 









Nurr 


ber examined 










Southern 




Year 


Season 


Middle Atlantic 


New England 


Georges Bank 


1973 


Spring 
Fall 


39 

144 




105 
129 


48 

191 


1974 


Spring 
Fall 


189 

S4 




93 

117 


103 

157 


1975 


Spring 
Fall 


b8 




100 
120 


92 
146 


1976 


Spring 
Fall 


1 1 1 
93 




125 

129 


63 

115 


Totals 




789 




918 


915 



recorded. A stomach was considered empty when no 
food items could be identified and the material found 
in the stomach weighed <0.001 g. Data were ana- 
lyzed with FORTRAN IV programs written for use 
on a Honeywell SIGMA 7 3 computer system located 
in Woods Hole, Mass. 

Food data are presented in terms of the mean 
stomach content weight, adjusted stomach content 
weight (discussed below), and the percentage weight 



'Reference to trade names does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 



fished, and during fall cruises a standard Yankee No. 
36 was used. The cod end and upper belly of both 
trawls were lined with 13 mm mesh netting to retain 
smaller fish. A scheme of stratified random trawling 
was conducted within the study area (Fig. 1), and 
fishing continued over 24 h/d 2 . All tows were 30 min 
in duration at a vessel speed of 3.5 kn in the direction 
of the next station. 

Sampling of stomachs was concentrated in three 
areas: Middle Atlantic, Southern New England, and 
Georges Bank (Fig. 1). Fish within two length groups 
(>20 cm and <20 cm) were randomly selected (50 
fish/group) during each cruise from the bottom trawl 
survey catches in each area. At each station within a 
particular area no more than 10 fish were taken for 
each of the two length groups, and fish were not sam- 
pled at two consecutive stations. The only exception 
to this collection method occurred when it appeared 
(during the cruise) that 50 large or 50 small fish would 
not be collected within a particular area. In this case, 
all fish caught were collected in an attempt to obtain 
the minimum sample size. Stomachs of large fish 
were excised aboard ship; individually wrapped in 
gauze with a label denoting vessel, cruise, species, 
fork length (FL), sex, and maturity; and preserved in 
3.7% formaldehyde (small fish were preserved whole). 
In the laboratory the preserved stomachs were in- 
dividually opened, and their contents emptied onto a 
0.25 mm mesh opening screen sieve to permit wash- 
ing without loss of any food items. The stomach con- 
tents were sorted, identified, counted, and damp 
dried on absorbent paper. Major prey items and com- 
monly occurring but relatively minor prey, in terms of 
weight, were identified to species whenever possible. 
The wet weight of all stomach content groups was 
determined to the nearest 0.00 1 g and all information 



'Further details of the bottom trawling techniques may be obtained 
from the Resource Surveys Investigation, Northeast Fisheries Cen- 
ter Woods Hole Laboratory, National Marine Fisheries Service, 
NOAA, Woods Hole, MA 02543. 



\ 



S 



% 



\ 



■h. 


cO /' V 


Portland ,'/T| 


f GULF OF 


•\ f 


MAINE 


♦o IS 






\ 



FIGURE 1.— Offshore areas sampled during bottom trawl surveys 
conducted by the Northeast Fisheries Center between the years of 
1973 and 1976, inclusive. 



22 



BOWMAN: FOOD OF SILVER HAKE 



each prey group made up of the total stomach con- 
tents weight. All tables follow a standard format to 
aid in making comparisons. In the tables, subtotals of 
the percentage weight of major stomach content 
groups are offset to the left. The minor prey groups 
are discussed in further taxonomic detail in the text. 
Adjusted stomach content weights are weights ad- 
justed by a correction factor which allows direct com- 
parison of the stomach content weights of different- 
sized fish. Adjustment of the stomach content 
weights was necessary, before any quantitative com- 
parisons could be made between variables such as 
sex or area. Observations on stomach tissue weight 
(excluding contents), mean stomach content weight, 
and whole fish weight (Fig. 2) revealed that neither 
the mean stomach content weight nor the stomach 
tissue weight is proportional to the body weight of 
different-sized fish. Stomach tissue weights of 526 
silver hake were gathered during a study jointly con- 
ducted by American and Soviet scientists on Georges 
Bank, September 1978, aboard the Soviet RV Belo- 
gorsk (operated by the Atlantic Research Institute of 
Marine Fisheries and Oceanography, Kaliningrad, 
USSR). Mean stomach content weight data were 
derived from the 1973-76 food data given in this 
report, and the fish body weights were calculated us- 
ing the silver hake length-weight equation described 
bv Wilk et al. (1978). Silver hake weighing < 100 g, or 
>300 g, have larger stomachs (stomach tissue weight 
being an indication of stomach size), and stomachs 



which contain on the average more food in terms of 
percentage body weight, than fish weighing between 
100 and 300 g. Since both the stomach tissue weight 
and the mean stomach content weight were dis- 
proportionate when presented as percentage body 
weight for different-sized fish (but were generally 
proportionate relative to each other), and because 
the mean stomach content weight data was much 
more variable than the stomach tissue weight data, 
the data adjustment was based on stomach tissue 
weight rather than on body weight or mean stomach 
content weight. The following equation was used to 
adjust the stomach content weights: 



A L = 



xl 
wl 



where 



A= Adjusted stomach content value. 
The adjusted stomach content value 
was converted to grams by multiply- 
ing it by the stomach tissue weight of 
a 30 cm FL fish. 

xl = Mean stomach content weight of all 
fish at a given length. 

wl = Mean stomach tissue weight of silver 
hake at a given length. 



The adjusted stomach content data for fish 4 (0.3 g) 
to 15 (21 g) cm FL and 24 (90 g) to 35 (292 g) cm FL 
are presented separately in forthcoming sections. 



3 



2 5 



O i STOMACH TISSUE WEIGHT/ BODY WEIGHT 

EXPONENTIAL CURVE FIT r 2 = 091, =0 107, b *0092 



x STOMACH CONTENT WEIGHT/ BODY WEIGHT 
,2 ., 



EXPONENTIAL CURVE FIT r^ = O 94, a = 006, b = 170 




200 300 

BODY WEIGHT (G) 



400 



500 



FIGURE 2. — Percentage body weight made up by the stomach tissue weight and the stomach content 
weight of different size silver hake. Area enclosed by solid lines represents more than 80% (excluding 
juveniles) of the silver hake population (fish 2-7 yr old), based on survey data. Stomach tissue weight/fish 
length and stomach content weigh t/fish length data were fit to an exponential curve (formjy = ae bx ). The 
data are presented in terms of body weight for illustrative purposes. 



23 



FISHERY BULLETIN: VOL. 82. NO. 1 



These two length groups were chosen because the 
food consumption of fish < 1 yr old (4- 1 5 cm FL) dif- 
fers substantially from the food consumption of older 
fish (evident from Figure 2). In addition, too few fish 
outside these length ranges were sampled to warrant 
inclusion in any of the calculations dealing with com- 
parisons between data sets. An analysis of variance 
(one way) was used to test the observed differences 
among sample means (e.g., between geographic 
areas). 

RESULTS 

The contents of 2,622 silver hake stomachs, of 
which 803 (30.4%) were empty, were analyzed. Fish 
sampled averaged 20 cm FL and had, including the 
empty ones, a mean stomach content weight of 1.5 g. 
Sources of potential variation in the data presented 
below include size, sex, and maturity stage offish, as 
well as the time of day, area, year, season, bottom 
depth, and temperature when or where the fish were 
caught. Each variable considered in this analysis is 
treated separately, i.e., the data were pooled over 
other variables with no attempt to determine the 
possible confounding effects of different variables on 
the results. Dietary trends noted within each par- 
ticular variable examined should be considered only 
as preliminary observations. 

Composition of the Diet 

Overall, in terms of percentage weight, the diet of 
silver hake consists almost entirely of fish (80.0%), 
crustaceans (10.2%), and squid (9.2%), as can be 
seen in Table 2 . The importance of crustaceans to the 
diet is overshadowed by the fish portion because 
large silver hake eat heavier meals consisting pri- 
marily of fish. However, Table 2 is useful because it 
serves as a composite list of the prey types commonly 
found in the stomachs of silver hake. Fish such as 
silver hake, Merluccius bilinearis; Atlantic mackerel, 
Scomber scombrus; butterfish, Peprilus triacanthus; 
herring (Clupeidae); American sand lance, Am- 
modytes americanus; scup, Stenotomus chrysops; At- 
lantic saury, Scomberesox saurus; and longfin hake, 
Phycis chesteri, each make up >0.1% of the stomach 
contents. The "Other Pisces" category, most of 
which could not be identified, accounts for a substan- 
tial portion (52.07c) of the "Pisces" group. Fishes 
which could be identified within this category (all 
contributed <0.1% to the diet) include summer 
flounder, Paralichthys dentatus; redfish, Sebastes 
marinus; codfishes (Gadidae); and flatfishes (Pleuro- 
nectiformes). 



Crustacea in the diet is represented principally by 
euphausiids (mostly Meganyctiphanes norvegica, 3.7%, 
and Euphausia, <0.1%) and decapods such as the 
Crangonidae (mainly Crangon septemspinosa, 1.4%, 
and Sclerocrangon boreas, <0.1%), Pandalidae (al- 
most exclusively Dichelopandalus leptocerus, 2.0%, 
although some Pandalus borealis, <0.1%, was also 
found), Pasiphaeidae (only Pasiphaea multidentata, 
0.1%), and other unidentified decapods (0.4%) which 
were mostly shrimp (0.3%). Amphipods found in the 
stomachs consist primarily of the families Ampe- 
liscidae (<0.1% each of Ampelisca agaxxizi, A. 
spinipes, A uadorum, and Byblis serrata), 
Oedicerotidae (<0.1% of Monoculodes edwardsi and 
M. intermedius), and Hyperiidae (exclusively the 
genus Parathemisto, 0.1%). The remaining crusta- 
cean groups are the Mysidacea (comprised of 
Neomysis americana, 0.7%, and Erythrops, <0.1%), 
Cumacea (mostly Leptocuma, <0.1%, and some un- 
identified diastylids, <0.1%), Copepoda (almost all 
identified as calanoids, <0.1%), and "Other Crus- 
tacea" (all of which was well-digested crustacean 
remains, 0.3%). 

The only other stomach contents identified were 
the cephalopods (Loligo pealei, 4.17c, and Rossia, 

Table 2.— Dietary composition of 2,622 silver hake 
caught in the Northwest Atlantic during the years 
1973-76. (+ indicates <0.1%.) 





Percentage 


Prey 


weight 


Polychaeta 


0.1 


Crustacea 


10 2 


Amphipoda 


1.3 


Ampeliscidae 


1.0 


Oedicerotidae 


0.1 


Hyperiidae 


0.1 


Other Amphipoda 


0.1 


Decapoda 


39 


Crangonidae 


1.4 


Pandalidae 


2.0 


Pasiphaeidae 


0.1 


Other Decapoda 


0.4 


Euphausiacea 


40 


Mysidacea 


0.7 


Cumacea 


+ 


Copepoda 


+ 


Other Crustacea 


0.3 


Cephalopoda 


9.2 


Loligo 


76 


Other Cephalopoda 


1.6 


Pisces 


80.0 


Scomberesox saurus 


1.5 


Clupeidae 


2.7 


Merluccius bilinearis 


9.2 


Phycis chesteri 


02 


Ammodytes americanus 


1.8 


Scomber scombrus 


7.5 


Stenotomus chrysops 


1.6 


Pepnlus triacanthus 


3.5 


Other Pisces 


52.0 


Miscellaneous 


0.5 


No. of stomachs examined 


2.622 


No. of empty stomachs 


803 


Mean stomach content weight (g) 


1.477 


Mean fish FL (cm) 


20.3 



24 



BOWMAN: FOOD OF SILVER HAKE 



<0.1%), Polychaeta, and the "Miscallaneous" 
category, which consisted of small amounts (<0.1%) 
of Echinodermata, Chaetognatha, unrecognizable 
digested matter, and sand. 

The percentage weights of various prey of silver 
hake within specified length groups are listed in Ta- 
ble 3. Silver hake <20 cm FL eat mostly crustaceans 
(>80% on the average), whereas the food of in- 
dividuals >20 cm FL is mostly fish and squid 
(average over 50%). Stomachs of silver hake 3-5 cm 
FL contain the largest percentages of smaller crusta- 
cean forms, such as amphipods and copepods. 
Decapods, euphausiids, and mysids, which are 
generally larger organisms (see Gosner 1971), make 
up the largest percentage of the diet of fish 6-20 
cmFL. 

Diet Differences Between Males and 
Females 

The diet of male and female silver hake differs in 
both quality and quantity of food (Table 4). The 
stomachs of males have the largest percentage of 
crustaceans, while those of females have the largest 
percentage offish and squid. The mean stomach con- 
tent weight of the males is only about one-fifth that of 
the females. Males also occur less frequently in the 
samples (42% of the fish collected were males) and 
are generally smaller than the females (mean FL 
males, 28.4 cm; females, 32.1 cm). Since female fish 
are, on the average, longer than the males, the dif- 
ferences noted above had to be dealt with in con- 
siderably more detail. 

A comparison of the data in Tables 5 (food of males) 
and 6 (food of females) indicates that males and 
females within the same size groupings consume dif- 
ferent types and amounts of food. The same dietary 
patterns noted for male and female fish in the preced- 
ing paragraph can be seen within most of the in- 
dividual length groups in these two tables (e.g., when 
males and females within the same size group are 
compared, the stomachs of the females contain larger 
quantities of food and higher percentages offish and 
squid). The number of males sampled generally ex- 
ceeds the number of females for length groups <30 
cm, while females dominate the length groups >30 
cm. 

A subset of the data were analyzed separately using 
only fish lengths for which 20 or more individuals 
each of males and females were sampled (Fig. 3). This 
group offish (ranging in FL from 24 to 34 cm) is fairly 
representative of the adult silver hake population 
sampled. The mean stomach content weight (Fig. 
3A), percentage crustaceans (Fig. 3B), and per- 



centage fish and squid (Fig. 3C) data presented 
graphically illustrate the differences between the 
diet of male and female silver hake of the same 
length. The stomachs of females contain more food, 
on the average, than those of males; the stomachs of 
males contain higher percentages of crustaceans 
than females; and the stomachs of females contain 
more fish and squid than those of males. Adjustment 
(by stomach tissue weight) of the mean stomach con- 
tent weights given in Figure 3A revealed that the 
stomachs of females contain, on the average, 1.5 
times the quantity of food found in the stomachs of 
males. 






o 



O 
u 

I 
o 

< 

O 






30 

2.0 
10 

100 

80 
60 
40 
20 
100 
80 
60 
40 
20 



1 1 I I 1 1 

MEAN TOTAL CONTENTS ( G ) 



1 T 




CRUSTACEANS 



L 




24 25 26 27 28 29 30 31 32 3 3 34 
FISH LENGTH ( FL IN CM ) 

FIGURE 3. — A) Mean stomach content weight of male and female 
silver hake versus fish length, B) percentage of total stomach con- 
tent weight made up by crustaceans for male and female silver 
hake, C) percentage of total stomach content weight made up by 
fish and squid for male and female silver hake. 



Diurnal Variation in Feeding 
Intensity 

The adjusted mean stomach content weight data 
presented in Figures 4 and 5 indicate the feeding 
periods of silver hake vary by season and size of fish. 
In autumn, the stomachs of larger fish (24-35 cm FL) 
are fullest just after midnight, while smaller fish (4-15 
cm FL) have the fullest stomachs in late afternoon 



25 



FISHERY BULLETIN: VOL. 82, NO. 1 



so 

3! 



J3 
5i 

3 
O 



c 
< 



s 

E 
o 



.S 

3j 



J4 

03 

-C 



•-- 

o 
V 



3 .5 
en 

•- 4- 



-C 



-5 



2P 
S 

is 

C 

c 



o 
o, 

S 
o 



CO 

C 

CD 
U 
tt 
CD 

Oh 

I 



w 
J 

X 

< 

E- 



III +6 | | 



II I I I I I 



I o I I + 



I I I I 111° 



|||+ oo | | 



I I I 



00 o 



111 + 



o 
cn 



w- o cn r~ r- if) r~ 

d | cb ir> | r-~ cb d | d 



+ 11 + 



+ + 



m cn 
d- I 



o n 
6 t 



CO ID CN 

id co *? 



O co | H r^ 



CD 

6 + 





CM 
CD 
CN ^ 


CN 

CN 


CNO^t 
«- CM If) 








CO 

o o> 


CD 

in 


»- cn r*- 

r- CO ^ 








CN 
CO CO 


CO 


CD O CM 
CM «- ** 




CO 



CM *- CO *- 

d d d d 



CO CO CO i- 

^ ^ 6 — 



ID CD CM «- CO 

cb cb ^ co i-^ 



CO (O 

cb d + 



I o 



CN 


CO 


















O 


CN 
























in 


CN 


T 


CN 




,- 


O 








O 


O 


o 


o 




CD 


CM 






CO 










o 










— 










o 







1 

CN 

to 



I - 



I ri I I «i I + I I I « 



CO ID 

d in 



— o 
d od 



CD + CO ID 

d do 



ID «- -<t CM 

d d d d 



.- *- CO CO 

d •<* d <- 



co r^ 
<* o 



r*. i- 

CO CM 



CD CO + | -- 



CO CN CM CD 

1^00+^ 



CO + | 



CO ID 
ID CM 



I I 



I I a I I I l l g 



1111111°) 



r- co «- 



M 11*1 



o 
I I I o> 



r- CO cn 
cb CN d + 



ID CD 




O r- 
00 






cn <t <N Tt 




"tf CO to ^t 



II I I « I <D I I I id 



ib cn 



+ CO 



00 CD *- CN 
CN CO CO 

** — 



ID 

ID CM 

** CN O CO 

V C7) CM 



Tt ID 

CO CO O CO 
SO CM 



CO O 

CD CD O CO 

CO CM i- 



CD CO O CM 



** CO 
O O 



^ CO CM CN 



I I « 



I I 



I I I I I I II 



cn CM 00 r^ CD 

.- *t d .- d 



- > S 3 



■o 

TO - 

"D hi 
, 9 5 



e-S 



y-tcD^^cjirraro*- 
: E<oiO o)c_)o-q-0 
! < Q 



to 

"O 
CD ° 

a> ro a to 

T3 5 2 <J o 

« C "D £ 5 5 

■o o - 5 Q « 

O O) JO CL - D 
a c T - CD <D 

C CO JZ " 

<o ro " 



T3 m 



I — . -J ^ 
i 5 O O 





ra 






Q 




CO 




C 


J; 


E 


E 
































to 





(0 




w 


4) 




<0 


ta 




tj 




a 
o 




CD 

-a 
£ 
a 

3 






5. 



E 




o 

a 

a 
a 


u 




a. 

o 





u 

CD 

.C 


vt 


o 

cj 


5 


CO 

U 


-Q 

E 



u 


(J 


O 


Q 
CD 




O 


M 


»o 


u 


0. 


^ 


10 



J Q. x: » 
S <b •- o 
Co 0. O « 

5 



co 

CO 
O i^) 

t o c> -i 

** CD 

cn 



E 1 ~ 

g|i 

° E ^ -=■ 

U) O c t 

% > ° -J 

e £ E "• 
2 £°f 



u ^ - - 

z z 5 5 



26 



BOWMAN: FOOD OF SILVER HAKE 



and just after midnight (Fig. 4). During springtime, 
large silver hake have substantial quantities of food 
in their stomachs (almost twice as much as during 
autumn) for two time periods, one near dusk and the 
other just before noon. Smaller fish have the most 
food in their stomachs just after midnight during 
spring (Fig. 5). No indication of a particular prey 
being eaten at a particular time of day was noted. 



silver hake within all geographic areas. Silver hake 
caught in the Middle Atlantic have the highest per- 
centage offish in their diet (Middle Atlantic, 87.5%; 
Southern New England, 78.4%; Georges Bank, 
76.4%), but most was unidentified (60.4%). Silver 
hake (20.8%) and herring (Clupeidae, 3.2%) make up 



Diet Within Geographic Areas 

Stomach content data for silver hake collected in 
various geographic areas (i.e., Middle Atlantic, 
Southern New England, and Georges Bank) are pre- 
sented in Table 7. Fish is by far the dominant prey of 



10 

i 

a 

I 

<_) 

I 

9 

I 

o 
(/I 
O 2 

V- 
V) 

\ 

1 



AUTUMN 














108 


61 


LARGE FISH 


- 


64 


87 


98 


49 






(24-35cmFL) 














22 




72 










88 






SMALL FISH 

(4-15cmFL) 




71 




208 


194 




110 " 




























12 1 
NOON 


5 18 
DUSK 


1 24 
MIDNIGHT 


3 06 09 12 

DAWN NOON 



FIGURE 4. — Adjusted mean stomach content weight of large (24-35 
cm FL) and small (4-15 cm FL) silver hake collected in the autumn 
versus time of day. The number of fish sampled in each time period is 
given just above the histogram. 



TABLE 4.— Stomach contents of male and female silver hake 
collected in the Northwest Atlantic during 1973-76. Data are 
expressed as a percentage weight. (+ indicates <0.1%.) 



Prey 



Male 



Female 



Polychaeta 


0.2 




+ 




Crustacea 


35.0 




4.5 




Amphipoda 


06 




0.2 




Ampeliscidae 




0.2 




0.1 


Oedicerotidae 




0.1 




+ 


Hyperudae 




0.2 




0.1 


Other Amphipoda 




0.1 




+ 


Decapoda 


11.9 




23 




Crangonidae 




5.1 




0.6 


Pandalidae 




55 




1 5 


Pasiphaeidae 




— 




+ 


Other Decapoda 




1.3 




02 


Euphausiacea 


18.8 




1.7 




Mysidacea 


2.7 




02 




Cumacea 


+ 




+ 




Copepoda 


+ 




— 




Other Crustacea 


1.0 




0.1 




Cephalopoda 


4.3 




10.4 




Loli go 




3.4 




8.6 


Other Cephalopoda 




0.9 




18 


Pisces 


59.1 




84.6 




Scomberesox saurus 




— 




18 


Clupeidae 




— 




32 


Merluccius bttinearts 




22.6 




76 


Phycis chesten 




— 




0.2 


Ammodytes amencanus 




1.4 




20 


Scomber scombrus 




3.8 




8.4 


Stenotomus chrysops 




— 




1.9 


Pepnlus tnacanthus 




3 3 




3.7 


Other Pisces 




280 




55.8 


Miscellaneous 


1.4 




05 




No examined 


613 




842 




No. of empty stomachs 


252 




354 




Mean stom. cont. wt. (g) 


0.85: 




4204 


Mean fish FL (cm) 


28.4 




32 1 




Length range (cm) 


6-59 




7-64 





o 20 

t~ 
l 
o 

UJ 

fe 1.0 

UJ 

Z 
O 

o 

X 

u 
< 

o 0.5 
o 

UJ 

h- 
in 

§ o 


SPRING 


LARGE FISH 


49 
















(24-35cmFL) 
53 










101 










59 


34 










37 






I SMALL FISH 

- (4-15cmFL) 




83 


83 




23 


90 






10 












Lj 

1 

NO 


1 


b, 


79 








2 15 18 2 
ON DUSK 


1 24 C 
MIDNIGHT 


3 06 
DAWN 


9 12 
NOON 



Figure 5. -Adjusted mean stomach content weight of large (24-35 cm FL) 
and small (4-15 cm FL) silver hake collected in springtime versus time of 
day. The number of fish sampled in each time period is given just above the 
histogram. 



27 



FISHERY BULLETIN: VOL. 82, NO. 1 
TABLE 5.— Composition of the diet of male silver hake in terms of percentage weight versus fish length. (+ indicates <0.1 %.) 



Length group (cm) 



Prey 5-10 



11-15 






16-20 


21-25 






26-30 






31-35 




36-40 




>41 


— 




— 




0.3 




+ 






03 






— 




— 


64 .1 




97 2 


29.3 




73. 


1 




32.7 






38 




1.9 




- 




2 7 

1.9 

0.8 


1.4 


0.6 

0.6 
02 




1.0 


04 
1 
0.4 
0.1 






2 

+ 
0.1 
0.1 

+ 


0.1 


0.1 

+ 

+ 


+ 

+ 


1.9 


1.9 




1.1 

0.6 

0.5 


10.7 


1 5 

H 3 

0.9 




19.1 


7.4 
9.8 

1.9 




15.0 


7 7 
5 7 

1.6 


2.6 


0.9 

1.7 


1.5 

0.1 

1.4 


50.3 






92 7 


14.2 






41.4 






15 9 




+ 




— 


11 9 






— 


— 






10.4 






0.8 




0.2 




0.4 


— 






— 


0.4 






+ 






+ 




— 




— 


— 






0.7 


+ 






— 






— 




— 




— 


— 




— 




2 6 






12 






0.8 




09 




— 



Polychaeta — 

Crustacea 19.2 

Amphipoda — 

Ampeliscidae — 

Oedicerotidae — 

Hyperudae — 

Other Amphipoda — 

Decapoda 0.3 

Crangonidae — 

Pandalidae — 

Pasiphaeidae — 

Other Decapoda 0.3 

Euphausiacea — 

Mysidacea 11.3 

Cumacea — 

Copepoda — 

Other Crustacea 7,6 

Cephalopoda — 4.4 02 83 2.5 

Lohgo — — — — — 8.1 — — 

Other Cephalopoda — — + 4 4 0.2 0.2 2 5 — 

Pisces 714 216 — 64 1 23.6 57 2 93 7 98.1 

Scomberesox saurus — — — — — — — — 

Clupeidae — — — — — — — — 

Merluccius bilmeans — — — 10.0 5 7.7 70.0 66.2 

Phycis chesten — — — — — — — — 

Ammodytes amencanus 

Scomber scombrus 

Stenotomus chrysops 

Pepnlus tnacanthus 

Other Pisces 
Miscellaneous 94 

No. examined 

No. empty 

Mean stom. cont. wt. (g) 

Mean fish FL (cm) 



50.8 






21.6 




— 


— 


— 




3.1 

9 .' 




— 




20.6 
4 


14.3 




+ 


2.8 


+ 


54 1 
1.9 


18 6 
3.1 


1.5 


80 
29.2 




23.7 


3 


12 
4 

0030 
84 




5 


0.435 
13.4 


20 

4 

0.414 
19.1 




119 
50 

0400 
23,7 


248 
109 

0.456 
28 5 


178 
73 

1215 
32 .2 


21 
9 
3 565 

37.1 


8 

3 

7.282 
509 





Table 6.- 


— Composition of the diet of female silver hake in terms of percentage weight versus fish length. (+ indicates <0.1%.) 






Length group (cm) 


Prey 


5-10 11-15 16-20 21-25 26-30 31-35 36-40 >41 



Polychaeta — — — — 0.4 1 + + 

Crustacea 8.7 100 75 2 27.9 39.9 13.0 2 0.2 

Amphipoda 0.3 — 0.3 1.8 1.3 0,8 + + 

Ampeliscidae — — — 07 0.5 02 + — 

Oedicerotidae — — — + 0.3 + + — 

Hyperudae 03 — 01 0.5 0.3 0.4 — — 

Other Amphipoda — — 0.2 6 0.2 0.2 + + 

Decapoda — 95.4 7 2 21.1 20.3 5 9 14 0.1 

Crangonidae — 95.4 18 6 6 5.1 19 0.2 + 

Pandalidae — — 4.7 12.9 13.3 3.1 1.1 0.1 

Pasiphaeidae — — — — — + — 

Other Decapoda — — 7 16 19 9 0.1 — 

Euphausiacea 7.5 4 66 8 3.3 13 5 5 2 0.6 1 

Mysidacea 0.9 — — 0.3 3.8 05 + + 

Cumacea — — — 0.1 + + 

Copepoda — — — — — — 

Other Crustacea — 6 0.9 1.3 1.0 0.6 + + 

Cephalopoda — — — 28.4 61 18.7 15.1 5.9 

Lohgo — — — 27 2 — 16.6 107 5.8 

Other Cephalopoda — — — 1.2 6.1 2.1 4.4 0.1 

Pisces 819 — 22.0 42.9 518 66.7 82 7 93.6 

Scomberesox saurus — — — — — — 6.1 — 

Clupeidae — — — — — 5.4 3 8 2 8 

Merluccius bilmeans — — — 31 9 5 6.6 20 8 — 

Phycis chesten — — — — — — — — 

Ammodytes amencanus 81.9 — — — 1 3 2 0.5 2 7 

Scomber scombrus — — — — — 7.3 9.5 9.3 

Stenotomus chrysops — — — — 16 — — 3.7 

Pepnlus tnacanthus — — — — — — 3.6 5.0 

Other Pisces — — 22,0 11.0 45 1 44,2 38.4 70.1 

Miscellaneous 9.4 — 2.8 0.8 18 1.5 0.2 0.3 

No. examined 9 3 22 113 202 259 126 103 

No. empty 2 3 45 83 1 20 54 47 

Mean stom. cont. wt. (g) 099 152 670 0.571 0.673 1.597 8.185 17 826 

Mean fish FL (cm) f^0 12.0 18.5 23.4 28 32.9 37.7 46 

28 



BOWMAN: FOOD OF SILVER HAKE 



TABLE 7. — Geographic breakdown of the prey found in the stomachs 
of silver hake caught in the Northwest Atlantic during the years 
1973-76. Data are expressed as a percentage weight. (+ indicates 
<0.1%). 





Middli 


5 


Southern 


Georges 


Prey 


Atlanti 


C 


New England 


Bank 


Polychaeta 


0.1 




0.1 


0.1 


Crustacea 


73 




7 3 


16.4 


Amphipoda 


0.5 




0.2 


0.4 


Ampehscidae 




0.1 


0.1 


0.1 


Oedicerotidae 




02 


+ 


0.1 


Hypenidae 




0.1 


0.1 


0.1 


Other Amphipoda 




1 


+ 


0.1 


Decapoda 


49 




26 


6.5 


Crangonidae 




2.4 


1 


1.3 


Pandalidae 




1.8 


1.2 


4.4 


Pasiphaeidae 




0.4 


— 


+ 


Other Decapoda 




0.3 


04 


0.8 


Euphausiacea 


1.2 




3.4 


7.9 


Mysidacea 


0.3 




0.7 


1.2 


Cumacea 


— 




0.1 


+ 


Copepoda 


+ 




+ 


+ 


Other Crustacea 


0.4 




0.3 


0.4 


Cephalopoda 


4.3 




13.7 


6.7 


Loligo 




2 9 


. 13.0 


6.7 


Other Cephalopoda 




14 


7 


+ 


Pisces 


87 5 




78.4 


76.4 


Scomberesox saurus 




— 


— 


6.1 


Clupeidae 




32 


1.3 


5 


Merluccius btlineans 


: 


208 


7.9 


0.4 


Phycis Chester/ 




— 


— 


0.8 


Ammodytes amencanus 




1.7 


0.4 


4.8 


Scomber scombrus 




— 


6.0 


21.1 


Stenotomus chrysops 




— 


4.1 


— 


Pepnlus triacanthus 




1.4 


2.2 


89 


Other Pisces 


i 


30.4 


565 


29.3 


Miscellaneous 


0.8 




0.5 


0.4 


No. of stomach examined 


789 




918 


915 


No. of empty stomachs 


180 




357 


268 


Mean stom. cont. wt. (g) 


1.544 


1.815 


1.080 


Mean fish FL (cm) 


17.5 




22.5 


20.8 


Length range (cm) 


3-57 




3-59 


3-64 



phausiids (3.4%) and decapods (2.6%) account for 
most of the Crustacea. 

The Cephalopoda was the only other prey group 
recognized as an important food of silver hake. Fish 
in Southern New England eat the largest quantities of 
squid (13.7%). Silver hake sampled on Georges Bank 
and in the Middle Atlantic also take fairly large 
amounts of squid as prey (6.7% and 4.3%, 
respectively). 

A comparison between the quantities of food in the 
stomachs of fish from each area revealed that Middle 
Atlantic silver hake have about two to three times 
more food in their stomachs (on the average) than fish 
from Southern New England or Georges Bank. 
Stomach content data for fish 24-35 cm FL from each 
area were adjusted for fish length; the adjusted mean 
stomach content weights were Middle Atlantic, 
1.328 g; Southern New England, 0.593 g; and Georges 
Bank, 0.707 g. The quantity of food in the stomachs 
of Middle Atlantic silver hake is significantly 
different (with 95% confidence) from the quantity 
in the stomachs of fish from Southern New England 
(F = 6.862 exceeds F 005 lj21 = 4.32). The adjusted 
mean stomach content weights of small (4-15 cm FL) 
silver hake from each area were Middle Atlantic, 
0.149 g; Southern New England, 0.198 g; and 
Georges Bank, 0.214 g. 

Yearly and Seasonal Differences 



the majority of the identified fish prey. The stomachs 
of silver hake caught in Southern New England con- 
tain fairly high percentages of silver hake (7.97o), 
Atlantic mackerel (6.0%), and scup (4.1%). Silver 
hake caught on Georges Bank eat mostly Atlantic 
mackerel (21.1%), butterfish (8.97o), Atlantic saury 
(6.1%), herring (Clupeidae, 5.0%), and American 
sand lance (4.8%). Evidence of the cannibalistic na- 
ture of silver hake is seen in all three areas. In addi- 
tion, silver hake taken as prey comprise the highest 
percentage of identified fish in both the Middle 
Atlantic and Southern New England (Table 7). 

Crustaceans are most important in the diet of silver 
hake collected from Georges Bank (16.4%). Eu- 
phausiids (7.9%), decapods (mostly pandalid 
shrimp, 4.4%, and crangonid shrimp, 1.3%), and 
mysids (1.2%) account for the majority of crustacean 
prey consumed on Georges Bank. In the Middle 
Atlantic and Southern New England, Crustacea is of 
equal importance (7.3%) as a food. For Middle Atlan- 
tic fish, decapods (4.9%) and euphausiids (1.2%) 
make up the majority of crustacean prey identified in 
the stomachs. In Southern New England, eu- 



Percentages of various prey categories in the silver 
hake diet between years, seasons, and geographic 
areas indicate the stomach contents are quite vari- 
able (Table 8). For example, in the Middle Atlantic, 
the Crustacea portion of the diet of silver hake varies 
from 3.1% (spring 1973) to 70.0% (fall 1976). Similar 
variability can be seen in the percentages listed for 
most of the prey categories. Much of the observed 
variation is probably due to differences in predator 
lengths (note mean fish FL's given at the bottom of 
Table 8). Only one prey, the American sand lance, 
was noted as being unique in the diet of silver hake. 
American sand lance was only found in the stomachs 
of silver hake collected in the spring during 1975 and 
1976. The largest percentage weights of American 
sand lance were derived from samples collected only 
during the spring of 1976 in all three areas. Another 
observation is that fish sampled in the spring tend to 
be larger (see mean lengths at bottom of Table 8) 
than those collected in the autumn. 

The adjusted stomach content data for large and 
small silver hake from all areas and years combined 
indicate that about twice as much food is found in the 
stomachs during spring than in autumn. The adjust- 



29 



FISHERY BULLETIN: VOL. 82, NO. 1 

Table 8. — Annual and seasonal breakdown of the stomach contents for silver hake collected in the Middle Atlantic, Southern New England, 
and Georges Bank. Data are expressed as a percentage weight for fish collected during the spring (S) and autumn (F) of 1 973-76. (+ indicates 
present but <0.l7r.) 







1973 






1974 








1975 






1976 


Prey 


S 




F 




S 




F 






S 




F 




S 




F 


MIDDLE ATLANTIC 


































Polychaeta 


— 




— 




0.1 




— 




05 






— 




16 




— 


Crustacea 


3.1 




4,2 




9.6 




6.5 




24.7 






4.7 




34.0 




70.0 


Amphipoda 


+ 




0.4 




1.2 




1.2 






1.3 




2 7 




2.1 




15.2 


Ampeliscidae 




— 




0.2 




+ 




0.7 






— 




0.3 




0.6 


1.3 


Oediceroiidae 




— 




+ 




1.1 




— 






04 




— 




9 


— 


Hyperndae 




— 




0.1 




— 




0.5 






+ 




2.1 




— 


12.1 


Other Amphipoda 




+ 




0.1 




0.1 




+ 






0.9 




03 




0.6 


1.8 


Decapoda 


3 1 




3.3 




3.7 




5.1 






89 




0.4 




22 9 




46.6 


Crangonidae 




1 4 




5 




2.0 




44 






59 




03 




11.1 


258 


Pandalidae 




1.0 




2.4 




— 




7 






26 




— 




11.7 


13.3 


Pasiphaeidae 




0.6 




— 




— 




— 






— 




— 




— 


— 


Other Decapoda 




1 




0.4 




1.7 




+ 






04 




0.1 




0.1 


75 


Euphausiacea 


— 




0.2 




4.4 




— 






14.4 




3 




+ 




— 


Mysidacea 


— 




+ 




— 




— 






— 




— 




5.4 




— 


Cumacea 


— 




+ 




+ 




+ 






0.1 




— 




+ 




— 


Copepoda 


— 




— 




+ 




+ 






— 




+ 




— 




— 


Other Crustacea 


+ 




03 




3 




2 






+ 




1 3 




3.6 




82 


Cephalopoda 


— 




14.9 




9 7 




— 




25.2 






— 




6.3 




— 


Loligo 




— 




12.4 




— 




— 






24.9 




— 




— 


— 


Other Cephalopoda 




— 




2.5 




9.7 




— 






0.3 




— 




6.3 


— 


Pisces 


96 5 




809 




79.5 




93.0 




46.6 






93.7 




548 




52 


Scomberesox saurus 




— 




— 




— 




— 






— 




— 




— 


— 


Clupeidae 




— 




— 




— 




91.5 






— 




— 




— 


— 


Merluccius bihneans 




233 




490 




— 




— 






4.0 




— 




— 


— 


Phycis chesten 




— 




— 




— 




— 






— 




— 




— 


— 


Ammodytes amencanus 




— 




— 




— 




— 






10.7 




— 




19.8 


— 


Scomber scombrus 




— 




— 




— 




— 






— 




— 




— 


— 


Stenotomus chrysops 




— 




— 




— 




— 






— 




— 




— 


— 


Pepnlus tnacanthus 




— 




— 




— 




— 






24 4 




— 




— 


— 


Other Pisces 




73.2 




31.9 




79.5 




l 5 






7 5 




93.7 




35.0 


5 2 


Miscellaneous 


0.4 




+ 




1.1 




0.5 




3.0 






1.6 




3 3 




24.8 


No. examined 


39 




144 




193 




54 




67 




91 




1 1 1 




93 


No empty 


1 1 




52 




26 




10 






7 




23 




22 




29 


Mean stom. cont wl (g) 


19 960 


0982 


0466 


793 




1 057 


0.243 


0606 


0075 


Mean fish FL {cm) 


33.9 




180 




14.1 




12 9 




198 




13 5 




21.7 




16.9 


Length range (cm) 


20-53 




4-45 




3-46 




4-37 




5 


-44 




3-40 




8-57 




3-35 


SOUTHERN NEW ENGLAND 


































Polychaeta 


0.1 




— 




+ 




— 




+ 






+ 




02 




+ 


Crustacea 


2 8 




12 5 




3 3 




46 .1 




7 9 






17.0 




19.8 




2.2 


Amphipoda 


+ 




1.7 




+ 




4.0 






1 




0.8 




02 




0.5 


Ampeliscidae 




— 




1 6 




+ 




1.5 






0.1 




0.2 




+ 


+ 


Oedicerotidae 




— 




— 




— 




— 






+ 




— 




+ 


+ 


Hyperndae 




— 




1 




— 




i 9 






+ 




0.5 




0.1 


5 


Other Amphipoda 




+ 




+ 




+ 




6 






+ 




1 




0.1 


+ 


Decapoda 


1 8 




8 4 




1 




13 7 






6 9 




9.7 




5.5 




1.2 


Crangonidae 




02 




0.9 




+ 




4 5 






2.0 




04 




4 7 


2 


Pandalidae 




0.9 




7 3 




— 




7.0 






1.8 




9 1 




8 


1.0 


Pasiphaeidae 




— 




— 




— 




— 






— 




— 




— 


— 


Other Decapoda 




0.7 




2 




1 




2.2 






3.1 




0.2 




— 


+ 


Euphausiacea 


0.5 




0.9 




3 2 




23 5 






0.8 




49 




99 




+ 


Mysidacea 


0.3 




+ 




— 




— 






0.1 




0.9 




3.8 




— 


Cumacea 


+ 




— 




+ 




1.7 






+ 




— 




+ 




+ 


Copepoda 


— 




+ 




— 




+ 






— 




+ 




— 




— 


Other Crustacea 


0.2 




1.5 




— 




3 2 






+ 




0.7 




0.4 




05 


Cephalopoda 


789 




1 6 




03 




— 




20.2 






— 




— 




2 8 


Loligo 




78.2 




— 




— 




— 






20.2 




— 




— 


— 


Other Cephalopoda 




0.7 




1 6 




03 




— 






— 




— 




— 


28 


Pisces 


18 2 




859 




95 6 




45.2 




70.1 






829 




79.8 




94.5 


Scomberesox saurus 




— 




— 




— 




— 






— 




— 




— 


— 


Clupeidae 




— 




— 




— 




— 






— 




31.8 




— 


— 


Merluccius bilineans 




0.2 




07 




— 




2.3 






55 




16 




— 


44.9 


Phycis Chester/ 




— 




— 




— 




— 






— 




— 




— 


— 


Ammodytes amencanus 




— 




— 




— 




— 






16 




— 




1.8 


— 


Scomber scombrus 




— 




— 




15.7 




— 






— 




— 




— 


— 


Stenotomus chrysops 




— 




— 




— 




— 






— 




— 




— 


24.7 


Pepnlus tnacanthus 




14.7 




— 




— 




— 






— 




— 




— 


— 


Other Pisces 




33 




85.2 




79.9 




42.9 






63.0 


. 


49.5 




78.0 


24.9 


Miscellaneous 


+ 




+ 




08 




8 7 




1.8 






0.1 




02 




05 


No. examined 


105 




119 




93 




117 




100 




1 20 




125 




140 


No. empty 


33 




86 




40 




38 




41 




31 




43 




45 


Mean stom. cont. wt. (g) 


2406 


0.401 


6 902 


107 




952 


0581 


2.181 


1 970 


Mean fish FL (cm) 


15.9 




27.5 




31 2 




16.8 




24 4 




18 1 




23.0 




22.9 


Length range (cm) 


6-47 




4-49 




9-59 




4-37 






6-55 




4-55 




3-53 




4-54 



30 



BOWMAN: FOOD OF SILVER HAKE 



Table 8. -Continued 





1973 






1974 






1975 








1976 


Prey 


S 


F 




S 




F 




S 






F 




S 




F 


GEORGES BANK 
































Polychaeta 


— 


— 




— 




— 




+ 




+ 






+ 




— 


Crustacea 


70.8 


15.0 




41.8 




18.2 




10.9 




5.9 




18 7 




60 


Amphipoda 


1.4 


04 




0.3 




13 




02 






04 




09 




0.1 


Ampehscidae 


0.1 




03 




+ 




07 




— 






0.2 




— 


0.1 


Oedicerotidae 


— 




+ 




+ 




0.3 




0.1 






+ 




— 





Hypemdae 


— 




— 




— 




— 




— 






0.1 




0.8 


+ 


Other Amphipoda 


1.3 




0.1 




03 




03 




0.1 






0.1 




0.1 


+ 


Decapoda 


60.7 


13.9 




2 5 




12 6 




1.0 






3.2 




2.8 




45 


Crangomdae 


1 9 




2.0 




1.3 




2.1 




5 






1.6 




0.6 


2.1 


Pandahdae 


44 5 




11.6 




— 




8 3 




— 






1 1 




20 


2.2 


Pasiphaeidae 


— 




— 




0.1 




— 




— 






— 




— 


— 


Other Decapoda 


14.3 




0.3 




1.1 




2.2 




05 






0.5 




02 


0.2 


Euphausiacea 


2.3 


0.2 




31.2 




26 




94 






0.5 




14.8 




+ 


Mysidacea 


— 


0.1 




7 8 




16 




0.2 






1.7 




— 




0.1 


Cumacea 


+ 


+ 




— 




+ 




— 






+ 




— 




— 


Copepoda 


— 


+ 




— 




+ 




— 






+ 




— 




— 


Other Crustacea 


59 


04 




+ 




1 




0.1 






0.1 




02 




1.3 


Cephalopoda 


— 


— 




— 




— 




— 




— 






1 2.8 




56.4 


Loll go 


— 




— 




— 




— 


— 








— 




12.8 


56.2 


Other Cephalopoda 


— 




— 




— 




— 


— 








— 




— 


02 


Pisces 


23.7 


84.9 




57 9 




81 8 




88.1 




94.1 




68.5 




35.8 


Scomberesox saurus 


— 




— 




— 


1 


38.8 




— 






— 




— 


— 


Clupeidae 


— 




— 




— 




— 




— 






39.2 




— 


— 


Merluccius bilmeans 


— 




— 




— 




4.1 




— 






— 




— 


— 


Phycis Chester/ 


— 




— 




— 




— 




3.2 






— 




— 


— 


Ammodytes amencanus 


— 




— 




— 




— 




— 






— 




31.6 


— 


Scomber scombrus 


— 




31.0 




— 




— 




S3 7 






— 




— 


— 


Stenotomus chrysops 


— 




— 




— 




— 




— 






— 




— 


— 


Peprilus tnacanthus 


— 




45.1 




— 




— 




— 






— 




— 


— 


Other Pisces 


23.7 




88 




57.9 




89 




21.2 






549 




36.9 


35.8 


Miscellaneous 


5 5 


0.1 




0.3 




+ 




1.0 




+ 






— 




1.8 


No. examined 


48 


198 




103 




157 




92 




146 




63 




115 


No. empty 


24 


39 




39 




27 




18 






39 




34 




48 


Mean stom. cont. wt. (g) 


0340 


1 029 


0996 


0.577 


2 629 




0906 


2 478 


0.767 


Mean fish FL (cm) 


31.4 


16.6 




24.2 




16.0 




24.5 






18.1 




29.7 




22.3 


Length range (cm) 


27-42 


4-54 




8-49 




4-40 




1 1-54 






4-48 




10-64 


3-55 



ed mean stomach content weights are presented in 
Table 9 for each season, year, and geographic area. In 
almost every year, in all areas, the stomachs of similar- 
sized fish contain larger quantities of food in the spring 



than in the fall. Only two exceptions were noted to this 
trend (for which there is no ready explanation): Large 
fish collected on Georges Bank in 1973 and small fish 
collected on Georges Bank in 1974. 



Table 9. — Annual and seasonal brea 


ikdown of the adju 


sted mean 


i stomach content' 


weight data of large (24- 


35 cm FL) and small (4- 15 cmFL) silver hake gathered from three 


geographical areas in the Northwest Atlan- 


tic during 1973-76. 


(S = spring, F = 


autumn.) 




















1973 


 


974 


1975 


1976 


Averages 


Area 


S 


F 


S 


F 


S 


F 


S 


F 


S 


F 


Middle Atlantic 






















Large fish 






















Adjusted weight (g) 


5.545 


1.081 


0995 


0.325 


2.203 


0.912 


0.936 


0.149 


2.420 


0.617 


Number in sample 


26 


68 


44 


9 


26 


29 


38 


43 






Small fish 






















Ad|usted weight (g) 


— 


0.108 


180 


0.096 


0.148 


0.142 


0.207 


0.155 


0.178 


0.131 


Number in sample 


— 


61 


136 


33 


31 


45 


47 


42 






Southern New England 






















Large fish 






















Adjusted weight (g) 


0.242 


0.122 


0488 


0.303 


0694 


0.657 


0.987 


0.976 


603 


0.515 


Number in sample 


17 


67 


51 


33 


47 


49 


63 


58 






Small fish 






















Adjusted weight (g) 


0256 


0.036 


200 


0.074 


0.414 


184 


0205 


0.149 


0.269 


111 


Number in sample 


73 


15 


4 


49 


35 


62 


39 


58 






Georges Bank 






















Large fish 






















Adjusted weight (g) 


0400 


743 


0.916 


0.576 


1 239 


0506 


0.735 


0.734 


0823 


640 


Number in sample 


43 


58 


50 


53 


32 


57 


27 


51 






Small fish 






















Adjusted weight (g) 


— 


140 


0.321 


0.325 


0566 


0.106 


0473 


117 


0453 


183 


Number in sample 


— 


119 


36 


95 


16 


80 


9 


50 


















Ave 


large fish 


adj. wt. 


1.282 


0.591 














Ave. 


small fish 


adj. wt 


0.300 


0.142 



31 



FISHERY BULLETIN: VOL. 82, NO. 1 



Maturity Stage Versus Diet 

Information on maturity was gathered in conjunc- 
tion with food data for 759 adult silver hake (Table 
10). Gonads were classified as 1) resting - gonad 
small in size and relatively translucent, 2) developing 
- gonad enlarged and either cream (males) or yellow- 
orange (females) colored, 3) ripe - gonad fills most of 
gut cavity, reproductive material either runs freely 
from an incision in the gonad or is extruded with pres- 
sure on abdomen of fish, 4) spent - gonad is flaccid, 
hemorrhaging is often evident. 



depth range (0.1 g). The quantity of food found in 
stomachs of large fish is variable; it steadily de- 
creases between the 27-37 m and 74-110 m depth 
ranges; increases at the 111-146 m range; and from 
1 1 1-146 m to 257-293 m continues to decrease (Ta- 
ble 12). Overall, the trend is for fish sampled at 
deeper depths to have less food, on the average, in 
their stomachs. It should be mentioned here that 
silver hake are known to regurgitate part or all of their 
stomach contents when they are retrieved from deep 
water depths (pers. obs.). Although fish which show 
obvious signs of regurtitation (e.g., everted stomach) 



TABLE 10. — Relationship between the adjusted stomach content weight and 
maturity stage of silver hake. Fish were caught on spring and autumn bottom 
trawl survey cruises conducted in the Northwest Atlantic from 1973 to 1976. 



Stomach content 
data 



Maturity stage Resting 
Ad], weight (g): 826 



Developing 
1.004 



Ripe 
122 



Spent 
1 292 



No. of fish examined 
Mean fish FL (cm) 
Length range (cm) 



379 

286 

24-35 



297 

30.6 

24-35 



29 

31 3 
27-34 



54 

31.2 

25-35 



No particular prey type is found in the stomachs of 
fish in specific maturity stages; all mature silver hake 
eat mostly fish. However, the stomachs of spawning 
(ripe) silver hake contain an average of about nine 
times less food than the stomachs of fish otherwise 
classified (Table 10). During pre- and postspawning 
periods, stomachs contain the largest quantities of 
food (1.0 and 1.3 g, respectively). 

Influence of Depth 

Analysis of samples from silver hake caught at dif- 
ferent bottom water depth ranges (27->365 m) 
revealed that the average length of fish, food type 
consumed, and quantity of food in the stomachs, 
varies with depth (Table 1 1). The majority (69.47c) of 
silver hake were caught at depths between 38 and 
110 m. Considering only the depth ranges where 
more than 50 fish were sampled (i.e., 27-220 m, and 
representing 95.6% of all silver hake collected) the 
mean FL offish increases with an increase in depth. 
Also, the percentage weight of euphausiids and squid 
in the stomachs tends to increase at deeper bottom 
depths, while the percentage weight offish in the diet 
shows a corresponding decrease. The adjusted mean 
stomach content data for both small and large fish are 
given in Table 12. The data are from only those depth 
ranges from which more than 20 fish (within a size 
group) were collected. The adjusted stomach content 
weight of small silver hake steadily decreases from 
the 27-37 m depth range (0.3 g) to the 111-146 m 



are not sampled on survey cruises, some fish may 
regurgitate and not be discernable from those which 
did not, This phenomenon, in part (other factors such 
as the decrease in abundance of typical prey of silver 
hake with an increase in depth or decrease in bottom 
water temperature may also be important in this 
regard, see Williams and Wigley 1977) could explain 
the decrease noted in stomach content weights with 
an increase in water depth. 

DISCUSSION 

The diet of silver hake consists almost exclusively of 
a combination of fish, crustaceans, and squid. The 
relative importance of each particular prey group as a 
food of silver hake is, for the most part, dependent on 
the size of the predator and/or the availability of the 
prey (Bigelow and Schroeder 1953; Jensen and Fritz 
1960; Fritz 1962; Dexter 1969; Vinogradov 1972). 

The composition of the diet of male and female 
silver hake is known to differ (Vinogradov 1972; Bow- 
man 1975). The present investigation confirms 
earlier reports that females feed predominantly on 
fish and that males eat mostly crustaceans. In addi- 
tion, it has been established that the stomachs of 
females contain larger quantities of food than the 
amounts in the stomachs of males of similar size. Since 
the rate of growth in fishes is directly related to their 
dietary intake, it is not surprising that females grow 
faster than males (Schaefer 1960). 

Bowman and Bowman (1980) studied diurnal varia- 



32 



BOWMAN: FOOD OF SILVER HAKE 



CD 

an 



CD 

a 

co 



a 

X 



eo 

a 



CO 

CO 

o 

CN 



C 

'3b 
c 
to 



a. 



1 « 

- S 

2g 

-* V 
eo C 

-C '" 

a, i, 

> 

— — 

<— .5f 

O a> 
C 

CD 

c 
o 
u 



re 

E 
o 



I I o | I 



III MM 

I 



I II I 



m co 

rid I M 



M || || M II 



M M II I II I 



d cd 



I I 



*- in 
d d 






| | do 



I + 



II I I I I II 



II II II II II 



-c 

Cm 
O 
0) 

b£ 

c 

CO 



-a 
>> 



O 

-a 

CO 

a> 

3D 



W 
-3 
X 
< 



I r. 



+ | O 



O | + + 



i- i- CN .- 

d d d d 



co co r- 

difi I - 



I I o 



I I I 



I I I 



CO CO ".' 

«) o -V 
<* CN CO ■" 



M II II I I 



i"*- co oo 

CO O | CN | | 



CN CN 

| oS iri 



I I 



o 
1- 



CN O 

6- I 



+ d 



en co 
d cm 



I 6 



M I I I I -' I I | 



I <* + d | co | 



CN 

O CM 

CO O IT) 

co »— d lo r^- 

01 CO CN if) 



CO CN ** 

O CO CO CH CN 

CN «* CN «- 



CN LD m 
CN CO CN If) CO 



O — 
- O + 



+ O 



oo in 




oS in 
<* 


CN 

co ■<* 
cm r- to 


r- 


CN CM ^ O CO 

m to cm 



"-; + CN — <t in 
d do -^ cn 



d d + + 



o <tf 
CM d 


to CO tJ 

co — in 


cn co en o to 

cm in co ^ to 

in 


to CN 
CO 00 



•- CO 

d c> + 



«-; CM CO O CN 

O in r- .- d 



d 



II S I I I M 



00 CO CO 

co d + + d 



d ** 



CD - — 



re 



"O 
O 

a. (o 

re fl) 



re 



c E S 
E < o 
< 



i o 



° It - v 

OCJ0.0.O 3>-5 o £ ^ o ,r 
c i 



E * 
c B 
to S 



TO ^ 

"* a> re 









O ^ Qj 



<* Co 



2.-B 

* 2 

§ 5! ' 



5 ?' 

•~ co , 
CO 0. ' 







ID 


cn 




o 

rn 


oa 

CO 


— 


r-» 


CO 



E E 
o o 



1*1 



LU OJ 



o-S 



c c 

re re 



Z Z 



IV — - 



OD 1 - 

55J 



id -a 

Ef — 
co o. 



a; c 
a8 
S-c 

CO C 
IB 00 



CO CD """ 



co 
X! 



•2 > 
c 



a ° 

CD ^. 



IB " > 

. <~ c 



-J 



CO 

s 

c 



c 

x 

CD 



CD p 
CO 

is 



E 
S 

o 

X! 



i " 

CC  — 

6 5 



TJ E 

CD CB 

CB T3 

3 C 



CB on J 



M 



E 

cj 



- 
< 



re 
CN 



C '- 

§ -2 

3 CD 

.2 .£ 

CO *-* 
> -C 

e O 



—I 


o 




Li. 






E 
o 


i' 


-C 


E 


Sfl 




3 




4 


Z 






■C 


a^ 




03 












H 


-C 


re 


D 


~ 


E 


6 


53 


CO 


< 


5 



OO ^t if) . 

cn o co cn 

^ tn m 



CM CO tO CO 

in oo co i- 



E «= 



I I I 



m 'j m co ^r 
co co cn cn cn 

co <N 






s s 



3 

■a 



o 


o 


CN 


o 


CO 


CO 


CN 


t 


CN 




10 


nr 


cn 


'23 


CN 


o 


IT' 


CN 


C3) 


CN 


O 


^ 


•- 


O 


r^ 


O 


O 


C 



to CO O CO 

*t co cm cn 

*- r- CM (N 



33 



FISHERY BULLETIN: VOL. 82. NO. 1 



tion in the feeding intensity of silver hake on Georges 
Bank in September 1978. They found that silver 
hake feed more intensively at night than during 
daylight. The findings of the present study are similar 
to those reported earlier (for the same size fish col- 
lected in autumn), but also indicate that an additional 
feeding period may occur around noon during 
springtime. No such pattern of feeding has been 
noted for adult silver hake in the past. 

Differences in the composition and/or quantity of 
food in the stomachs of silver hake collected within 
various geographic areas have been observed pre- 
viously by Schaefer (1960), Vinogradov (1972), and 
Langton and Bowman (1980). Two items are par- 
ticularly noteworthy concerning the diet of silver 
hake in the different geographic areas studied here. 
The first is the large quantity of food in the stomachs 
of silver hake from the Middle Atlantic (on the 
average two or three times more than the quantities 
in the stomachs of Southern New England and 
Georges Bank fish). The second is the high percent- 
age weight (20.8%) of silver hake in the diet of silver 
hake caught in the Middle Atlantic. Of interest is that 
Langton and Bowman (1980) also found that silver 
hake caught in the Middle Atlantic area (during the 
period 1969-72) are more cannibalistic than silver 
hake in other areas of the Northwest Atlantic. 

Vinogradov (1972) concluded that the differences 
he observed in the feeding of silver hake in the 
Northwest Atlantic during 1965-67 were "due to 
variations from area to area in the species composi- 
tion of the fish food and the rate of feeding." 
Vinogradov's mention of "the rate of feeding" 
referred to the variation in feeding intensity of silver 
hake throughout the year. He found silver hake feed 
most intensively in the spring-summer and autumn 
periods. During the summer (when silver hake 
spawn) and winter, he noted that the feeding rate 
diminishes. The data presented here, in conjunction 
with other published and unpublished data, tend to 
corroborate Vinogradov's conclusions. Silver hake 
caught in spring have twice as much food in their 
stomachs as those caught in fall (data from present 
study for 24-35 cm FL fish— 1.3 g, spring; 0.6 g, fall). 
The stomachs of spawning silver hake contain small 
quantities of food (0.1 g) compared with fish with 
developing (1.0 g) or spent (1.3 g) gonads (data from 
present study). Fish > 20 cm FL collected during late 
summer-early autumn have small quantities of food 
(mean stomach content weight of 0.2 g) in their 
stomachs (Bowman and Bowman 1980). The 
stomach contents of silver hake collected on Georges 
Bank during the winter (December-January) of 
1976-77 were analyzed by Bowman and Langton 



(1978). They found the mean stomach content weight 
offish 20 cm FL and larger to be 0.4 g. The stomachs 
of silver hake (all >29 cm FL) collected in February 
(late winter) of 1977 on Georges Bank, by American 
and Polish scientists aboard the Polish RV Wieczno 
(conducting research in conjunction with the Woods 
Hole Laboratory), contained an average of 0.1 g of 
food (unpublished data available from the author). 
The pattern of feeding intensity for silver hake 
throughout the year, based on the above information, 
is intensive feeding in the spring and early summer; 
curtailment of feeding in summer and early autumn 
(during spawning); resumption of feeding in the 
autumn, but to a lesser degree than in the spring; and 
finally a reduction in feeding throughout the winter. 
Somewhat similar feeding patterns have been es- 
tablished for other species of marine fish (Tyler 
1971). 

Grosslein et al. (1980) reported an increase in bot- 
tom trawl survey catches of American sand lance in 
1976 in the Northwest Atlantic. The population up- 
surge of American sand lance combined with the high 
percentage weights of American sand lance found in 
silver hake stomach contents during 1976 is an in- 
dication of silver hake's opportunistic predatory 
behavior. Availability of prey is probably one of the 
most important factors in determining what types 
and how much food silver hake eat. 

ACKNOWLEDGMENTS 

I thank M. Grosslein for his critical review of the 
manuscript; J. Towns, J. Murray, and others for their 
help in analyzing the fish stomach contents and in 
tabulating the data; and especially G. Kelley, 
laboratory typist, for her patience. 



LITERATURE CITED 

BlGELOW, H. B., AND W. C. SCHROEDER. 

1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., 
Fish. Bull. 53, 577 p. 
Bowman, R. E. 

1975. Food habits of Atlantic cod, haddock, and silver hake in 
the Northwest Atlantic, 1969-1972. U.S. Natl. Mar. Fish. 
Serv., Northeast Fish. Cent., Woods Hole Lab. Ref. 75-01, 
53 p. 
Bowman, R. E., and E. W. Bowman. 

1980. Diurnal variation in the feeding intensity and catch- 
ability of silver hake (Merluccius bilinearis). Can. J. Fish. 
Aquat. Sci. 37:1565-1572. 
Bowman, R. E., and R. W. Langton. 

1978. Fish predation on oil-contaminated prey from the re- 
gion of the ARGO MERCHANT oil spill. In In the wake of 
the ARGO MERCHANT, p. 137-141. Univ. R.I. Cent. 
Ocean Manage. Stud. 



34 



BOWMAN: FOOD OF SILVER HAKE 



Dexter, R. W. 

1969. Studies on the food habits of whiting, redfish, and 
pollock in the Gulf of Maine. J. Mar. Biol. Assoc. India 
ll(l&2):288-294. 
Edwards, R. L., and R. E. Bowman. 

1979. Food consumed by continental shelf fishes. In H. 
Clepper (editor). Predator-prey systems in fisheries man- 
agement, p. 387-406. Sport Fish Inst., Wash., D.C. 

Fritz, R. L. 

1962. Silver hake. U.S. Fish Wildl. Serv., Fish. Leafl. 538, 
7 p. 
GOSNER, K. L. 

197 1. Guide to identification of marine and estuarine inver- 
tebrates. Cape Hatteras to the Bay of Fundy. Wiley, N. Y., 
693 p. 
Grosslein, M. D., R. W. Langton, and M. P. SlSSENWINE. 

1980. Recent fluctuations in pelagic fish stocks of the North- 
west Atlantic, Georges Bank region, in relation to species 
interactions. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 
177:374-404. 

Jensen, A. C, and R. L. Fritz. 

1960. Observations on the stomach contents of the silver 
hake. Trans. Am. Fish. Soc. 89:239-240. 
Langton, R. W., and R. E. Bowman. 

1980. Food of fifteen northwest Atlantic gadiform fishes. 
U.S.Dep. Commer.,NOAATech. Rep. NMFS SSRF-740, 
23 p. 
Nichols, J. T., and C. M. Breder, Jr. 

1927. The marine fishes of New York and southern New Eng- 



land. Zoologica (N.Y.) 9:1-192. 
Noskov, A. S., and V. I. Vinogradov. 

1977. Feeding and food chains of the fish of Georges Bank. 
Rbyn. Khoz. 53:19-20. (Can. Fish. Mar. Serv., Transl. 
Ser. 4540. D.O.E. St. John's, Nfld. 1979.) 

SCHAEFER, R. H. 

1 960. Growth and feeding habits of the whiting or silver hake 
in the New York Bight. N.Y. Fish Game J. 7:85-98. 
Swan, B. K., and D. Clay. 

1979. Feeding study on silver hake (Merluccius bilinearis) 
taken from the Scotian Shelf and ICNAF Subarea 5. 
ICNAF Res. Doc. 79/VI/49, 14 p. 
Tyler, A. V. 

1971. Monthly changes in stomach contents of demersal 
fishes in Passamaquoddy Bay, N.B. Fish. Res. Board 
Can., Tech. Rep. 288, 103 p. 

Vinogradov, V. I. 

1972. Studies of the food habits of silver and red hake in the 
Northwest Atlantic area, 1965-1967. ICNAF Res. Bull. 
9:41-50. 

Wilk, S. J., W. W. Morse, and D. E. Ralph. 

1978. Length-weight relationships of fishes collected in the 
New York Bight. Bull. N.J. Acad. Sci. 23(2):58-64. 

Williams, A. B„ and R. L. Wigley. 

1977. Distribution of decapod Crustacea off northeastern 
United States based on specimens at the Northeast Fish- 
eries Center, Woods Hole, Massachusetts. U.S. Dep. 
Commer., NOAA Tech. Rep. NMFS Circ. 407, 44 p. 



35 



ABUNDANCE AND VERTICAL DISTRIBUTION OF FISHES IN 

A COBBLE-BOTTOM KELP FOREST OFF 

SAN ONOFRE, CALIFORNIA 



Ralph J. Larson 1 and Edward E. DeMartini 2 



ABSTRACT 

Using visual belt transects on the bottom and vertically stratified belt transects taken with movie cameras in 
the water column, we assessed the species composition, vertical distribution, and standing stock of fishes in a 
forest of giant kelp and a nearby kelp-depauperate area off San Onofre, California. The volume of water- 
column "cinetransects" was calibrated for water clarity. Species such as garibaldi, blacksmith, and various 
rockfishes, which depend on high-relief rocky substrates, were rare or absent in these low-relief, cobble- 
bottom habitats. The species present in the kelp forest apparently did not depend on high-relief rock, at least 
in the presence of kelp. These species fell into three groups, based upon their vertical distributions: "canopy" 
species (kelp perch, giant kelpfish, and halfmoon), which occurred mainly in the upper water column; "cos- 
mopolites" (kelp bass, white seaperch, and senorita) .which occurred throughout the water column; and "bot- 
tom" species (California sheephead and various seaperches), which occurred mainly near the bottom. 
Despite the absence of reef-dependent species, estimated standing stocks of 388-653 kg/ha in the San 
Onofre kelp forest were as large or larger than estimates made by others in kelp forests located on higher 
relief bottoms. The kelp-forest areas at San Onofre also supported a larger standing stock of fishes (other 
than barred sand bass) than the adjacent area with little kelp. The relatively large standing stock of fishes in 
the kelp forest can be attributed to the presence of kelp and to the depth of the kelp forest. Located in 
relatively deep water (15m), this kelp forest possessed an extensive midwater zone. The attraction of fish in 
moderate densities to the midwater zone of this kelp forest contributed substantially to overall biomass. We 
conclude that kelp per se can enhance the standing stock of fishes on a temperate reef, at least in areas of low 
bottom relief. 



Rocky reef and giant kelp, Macrocystis pyrifera, 
habitats off the coast of southern California support a 
diverse and abundant assemblage of fishes (Lim- 
baugh 1955; Quast 1968 a, b; Feder et al. 1974; Ebel- 
ing et al. 1980 a, b). Much of the richness of this 
ichthyofauna has been attributed to the rocky sub- 
strate; areas with a rugose, rocky bottom and little 
kelp seem to support more fish than areas with a flat 
bottom and dense kelp (Quast 1968 a, b, Ebeling et 
al. 1980a). However, kelp itself also provides a uni- 
que habitat for some fishes (Coyer 1979; Ebeling et 
al. 1980a) and a point of orientation in the water 
column for others (Quast 1968 a, b; Bray 1981). The 
kelp canopy may also serve as a nursery area for some 
species of fish (Miller and Geibel 1973; Feder et al. 
1974; M. Carr 3 Unpubl. data). 

Several approaches have been used to assess the 
influence of habitat on the abundance and composi- 



1 Marine Science Institute, University of California, Santa Barbara, 
Calif.; present address: Department of Biological Sciences, San 
Francisco State University, San Francisco, CA 94132. 

-Marine Science Institute, University of California, Santa Barbara, 
Calif.; present address: Marine Review Committee Research Cen- 
ter, 531 Encinitas Boulevard, Encinitas, CA 92024. 

3 M. Carr, Moss Landing Marine Laboratories, Moss Landing, 
CA 95039. 



tion of fish assemblages in nearshore kelp and rock 
habitats off California. Perhaps the best analytical 
approach is experimental, as employed by Miller and 
Geibel (1973), Bray (1981), and Carr (footnote 3) ; 
however, the comparative approach of Limbaugh 
(1955; also reported in Feder et al. 1974), Quast 
(1968 a, b), and Ebeling et al. (1980a) is also of value. 
Based on observations in a variety of areas, Lim- 
baugh described the habits and habitats of many 
nearshore fishes. Quast and Ebeling et al. employed 
broad-scale quantitative sampling of fish assem- 
blages in different areas. Quast's interpretation of 
data extended Limbaugh's natural history approach, 
and added to it the actual comparison of abundances 
in different habitats. Ebeling etal. (1980a) employed 
a multivariate analysis of habitat characteristics and 
relative abundances of species to define subassem- 
blages of fishes, and also compared abundances in 
areas of different habitat characteristics. 

In this paper we examine the abundance, vertical 
distribution, and species composition of noncryptic 
fishes in a forest of giant kelp near San Onofre, Calif. 
We also report the abundance and species composi- 
tion of fishes in a nearby area with little kelp. This 
study, undertaken initially to predict the effects of a 



Manuscript accepted June 1983. 

FISHERY BULLETIN: VOL. 82, NO. 1, 1984. 



37 



FISHERY BULLETIN: VOL. 82. NO. 1 



possible loss of kelp (Dean 4 ) on the indigenous fish 
fauna, also allowed us to extend the comparative 
approach of Quast and E beling to assess two features 
of kelp-forest fish faunas and to further evaluate a 
sampling technique. 

The portion of the kelp forest we examined was 
located in relatively deep water (15 m) and was 
anchored on a low-relief cobble bottom. Since it 
lacked a highly heterogeneous substrate, we were 
able, by comparison, to further evaluate the effects of 
kelp per se on nearshore fishes. Because the kelp 
forest was in deep water, we also had the opportunity 
to examine the vertical distribution of fishes in 
greater detail than other workers, by sampling four 
vertical strata, rather than the two strata (canopy and 
bottom) sampled by Quast (1968b) and E beling etal. 
(1980a, b). 

Besides visual transects to sample fish on or near 
the bottom, we used underwater movies ("cinetran- 
sects") to estimate the abundance of fishes in the 
water column above the bottom. Alevizon and 
Brooks (1975) and Ebeling et al. (1980b) discussed 
the advantages and disadvantages of cinetransects, 
but provided only rough estimates of the area 
sampled in a cinetransect. In this paper we more 
carefully evaluate cinetransect volume, emphasizing 
the effect of underwater visibility on cinetransect 
width. 

Our objectives in this paper are 1) to estimate cine- 
transect volume as a function of underwater 
visibility; 2) to examine the vertical distribution of 
fishes in a deep-water kelp forest; 3) to estimate the 
overall abundance and biomass of fishes, integrated 
over depth, in this kelp forest; and 4) to evaluate the 
importance of kelp to nearshore fishes, by comparing 
our data from the San Onofre kelp forest with that 
from an adjacent kelp-depauperate area and with 
other published data from kelp forests located on 
more rugose substrates. 

MATERIALS AND METHODS 

Study Areas 

This study was conducted in and near the offshore 
portion of a giant kelp, Macrocystis pyrifera, forest 
near the San Onofre Nuclear Generating Station, 
between San Clemente and Oceanside, Calif. (Fig. 1). 



4 T. A. Dean. 1980. The effects of San Onofre Nuclear Generating 
Station on the giant kelp, Macrocystis pyrifera. Annual report of the 
Kelp Ecology Project, January-December 1979, to the Marine Re- 
view Committee of the California Coastal Commission. Unpubl. 
rep., 189 p. Kelp Ecology Project, Marine Science Institute, 
University of California, Santa Barbara, CA 93106. 



San Onofre kelp (SOK) varied in areal extent from 
<5 to 95 ha during the mid- to late 1970's, and 
covered about 75 ha during the fall of 1979 (Dean 
footnote 4). SOK occupied a shallowly sloping, low- 
relief (< 1 m) cobble and sand substrate between the 
depths of about 10 and 15 m. Two relatively perma- 
nent, offshore portions of SOK, and an area with little 
kelp located <100 m upcoast from SOK, served as 
our study areas. The upcoast (SOK-U) and 
downcoast (SOK-D) areas within SOK, and the kelp- 
depauperate area ("kelpless" cobble), were all about 
15 m deep and 2-3 km from shore. Because of its 
depth, low relief, and periodic inundation by sand, 
the cobble substrate in all areas was relatively bare of 
understory algae and sessile invertebrates. However, 
some stands of the 1 m tall laminarian kelp 
Pterygophora californica were present, especially 
along the fringes of the Macrocystis forest and 
throughout the kelpless cobble area. 

Sampling Methods 

Our general sampling plan was to stratify fish cen- 
suses by depth and to replicate these samples over 
several dates. In the two kelp-forest areas, we cen- 
sused each of three, equally spaced strata in the 
water column, plus a bottom stratum. Only the bot- 
tom stratum was censused at the kelp-depauperate 
area, since few kelp-associated fishes were observed 
above the bottom in this area. Sampling at each 
stratum was replicated hierarchically: A number of 
replicate transects were made within an area on a 
given sampling day, and counts from these transects 
were averaged. This was repeated on 4 or 5 d at each 
site. The daily averages at each stratum and area 
were themselves used as replicates that provided 
reasonably precise estimates of means per stratum 
and that allowed estimates of variability due to sam- 
pling error. Because of time and manpower con- 
straints, the various study areas were usually 
sampled on different dates. All three water-column 
strata in a given area were sampled on the same day; 
the bottom stratum, however, was usually sampled 
on a different day. 

All sampling took place from October through 
December 1979. This time of year offers the most 
consistently clear and calm water conditions. Since 
most migratory and transient species were excluded 
from analysis (see below), our fall study should 
reasonably characterize the general distribution and 
abundance of "resident", kelp-associated fishes at 
SOK. Within this period, sampling was generally 
limited to dates when horizontal visibility exceeded 3 
m. 



38 



LARSON and DeMARTINI: DISTRIBUTION OF FISHES IN KELP FOREST 



120' 



119' 



118' 



117 < 



Santa D . 
Barbara 

$ an Santa 

Mi^el Is. ^gp |s 

h34° 



Santa Rosa Is. 



Los 
Angeles 



San 



Santa Catalina Is. 



Nicholas Is. 



San 

Clemente 

JOceanside 



1-33° 



San V> 
Clemente Is. 



9.1 nMk 
18.1 m x5.5 






FIGURE 1. — Location of areas sampled 
during fall 1979, near San Onofre, Calif.: 
Upcoast (U) and downcoast (D) portions 
of the San Onofre kelp bed, and nearby 
kelp-depauperate area (kelpless cobble). 



San Mateo Creek 

San Onofre Creek 




San 
Kelp Bed 



kelpless 
cobble 

San Onofre ? 
Kelp Bed d 



12 3 4 5 6 



kilometers 



39 



FISHERY BULLETIN: VOL. 82, NO. 1 



In each area, two permanently buoyed stations 
served as foci for sampling. At each station, we deter- 
mined a range of suitable compass headings for tran- 
sects. To assure complete coverage of the area, we 
divided each range of suitable headings into five 
equal subarcs and randomly chose transect headings 
from each subarc. Headings were selected separately 
for each sampling stratum. One transect per subarc 
was made on each sampling day for bottom sampling. 
In the water-column strata, where fish patchiness 
necessitated more samples, we made one transect in 
each subarc and added another transect from one of 
the subarcs (randomly chosen). Thus, five transects 
were usually made from each station per date on the 
bottom, and six at each station and depth stratum in 
the water column. Regardless of sampling method, 
transects began 7-10 m from the station hub. Tran- 
sects were taken from both sampling stations on a 
sampling day. Data from the two stations at an area 
were pooled, since the abundances of major species 
were generally indistinguishable between stations in 
an area on a given date. 

On the bottom, fish sampling was conducted 
visually in 75 m long strip transects. Divers (one per 
station) counted fish in bands estimated to be 3 m 
wide and 1.5 m high, while reeling out 75 m long lines 
along the predetermined compass headings. All non- 



cryptic fishes within this band were identified and 
counted, with separate tallies kept for juvenile, sub- 
adult, and adult members of each species (Table 1). 
All subadult and adult Mac rocystis plants >1 m tall 
(Dean footnote 4) were counted in the same 3 m wide 
band while reeling in the transect line on the 
return trip. 

Transects in the water column at the two kelp-forest 
areas were made with underwater movie strips, using 
Elmo Super 311 Low Light 5 movie cameras (F/l.l), 
Giddings Cine-Mar housings, and Kodak Ekta- 
chrome 164 super-8 film cartridges. At 18 frames/s, 
the transects lasted about 3 min. Divers swam pre- 
determined compass headings and photographed 
fish occurring in a 120° horizontal arc about the tran- 
sect axis and 1.5 m above and below the diver's 
depth. The transect ended when the film cartridge 
was exhausted. Water-column transects were made 
in three depth strata: 3 m, 7.6 m, and 12 m (Table 2). 
Horizontal visibility was measured with each set of 
transects (at a depth on a sampling date), as the dis- 
tance at which an olive-tan colored, 10 cm long float 
("fish mimic") became indistinct. Films were later 
viewed in slow motion by at least two observers, at 



'Reference to trade names does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 



Table 1.— Common and scientific names of fishes observed at the San Onofre kelp bed and adjacent kelpless cobble area during fall 1979 with 
the estimated weight of juveniles, subadults, and adults. Body weights for teleosts were estimated from average observed lengths, converted to 
weights using the length-weight regressions of Quast (1968a: Appendix B), after adjusting for the bias (underestimate) from the use of average 
body length to predict average body weight (see Pienaar and Ricker 1 968). Weights of elasmobranchs were estimated from fishes trapped in the 
intakes of the San Onofre Nuclear Generating Station, Unit 1, during 1976-79. 1 Asterisks indicate species not included among kelp-bed 
"residents." Common names after Robins et al. (1980). 



Weight (g) 



Weight (g) 



Family and species 


Juvenile 


Subadult 


Adult 


Family and Species 


Juvenile 


Subadult 


Adult 


Serranidae 








Scorpaenidae 








Paralabrax clathratus, kelp bass 


7 


200 


1,050 


Scorpaena guttata, California scorpionfish 


— 


— 


550 


Paralabrax nebulifer. barred sand bass 


20 


300 


1,500 


Sebastes rastrelliger, grass rockfish 2 


— 


— 


400 


Embiotocidae 








Sebastes serranoides, olive rockfish 2 


4 


175 


— 


Brachyistius frenatus, kelp perch 


— 


— 


25 


Sebastes spp,, juvenile rockfish 2 


1 


— 


— 


Embiotoca /acksom . black perch 


10 


75 


350 


Sciaenidae 








Phanerodon furcatus , white seaperch 


10 


50 


175 


'Cheilotrema saturnum. black croaker 


— 


— 


225 


Damalichthys vacca . pile perch 


15 


175 


500 


Pristopomatidae 








Rbacochilus toxotes. rubberlip seaperch 


15 


150 


700 


'Xentstius californiensis. salema 


— 


— 


75 


Hypsurus caryi. rainbow seaperch 


10 


60 


150 


Athennidae 








Labndae 








*silversides spp. 


— 


— 


20 


Oxy/ulis califomica . sehonta 


0.5 


5 


55 


Carangidae 








Semicossyphus pulcher. California sheephead 


50 


250 


875 


*Trachurus symmetricus. jack mackerel 


— 


115 


— 


Halichoeres semicmctus. rock wrasse 


25 


100 


250 


Sphyraenidae 








Girellidae 








'Sphyraena argentea. Pacific barracuda 


— 


150 


— 


Giretla nigricans, opaleye 


— 


— 


950 


Carcharhinidae 








Scorpididae 








'Tnakis semifasciata. leopard shark 


— 


— 


2,000 


Medialuna califormensis, halfmoon 


— 


— 


250 


Rhinobatidae 








Pomacentridae 








'Ptatyrhmotdes tnsenata . thornback 


— 


— 


240 


Chromis punctipinms. blacksmith 


2 


— 


— 


Myliobatidae 








Hypsypops rubicundus. garibaldi 


25 


120 


500 


*Myliobatis califomica. bat ray 


— 


— 


6,700 


Clinidae 








Torpedmidae 








Heterostichus rostratus, giant kelpfish 


3 


30 


175 


'Torpedo califomica. Pacific electric ray 


— 


— 


9,450 


Cottidae 
















Scorpaenichthys marmoratus, cabezon 


— 


— 


1,500 











'E. DeMartini and R Larson. 1980. Predicted effects of the operations of San Onofre Nucler Generating Station Units 1, 2, and 3 on the fish fauna of the San Onofre 
region. Report submitted to the Marine Review Committee of the California Coastal Commission. Unpubl. rep., 27 p. Marine Science Institute, University of California, Santa 
Barbara, CA 93106. 

2 Members of the genus Sebastes will be grouped under rockfish spp. in subsequent tables. 



40 



LARSON and DeMARTINI: DISTRIBUTION OF FISHES IN KELP FOREST 



TABLE 2.— Bathymetric sampling strata at the San Onofre kelp bed. 
Weighting factors (WJ are shown for the above-bottom strata and 
for the above-bottom versus bottom strata. 



Sampling 


Depth Range 


Extent of 


W^ (above- 


W h (all 


depth (m) 


represented (m) 


range (m| 


bottom only) 


strata) 


3 


0-5.3 


5.3 


03926 




7.6 


5.3-9.8 


4.5 


03333 


'0.9 


12 


9.8-13.5 


3.7 


0.2741 




15 (bottoml 


13.5-15.0 


1.5 


— 


0.1 




0-15 




1.0 


1.0 



'Weighting factor for above-bottom strata combined. 



which time fish that were distinguishable on film 
were identified, counted, and assigned to maturity 
classes as above. 

Transect Volume 

The volume of visual bottom transects was con- 
sidered to be fixed, and the volume of water-column 
cinetransects to be dependent on underwater 
visibility. The volume of bottom transects was fixed 
at75mX 3mX 1.5 m = 337.5 m\ since the length of 
transects was measured, and the height and width of 
transects were fixed at values less than horizontal 
visibility. Cinetransect length was taken as the 
average distance covered in simulated, 3-min cine- 
transects swum by three divers over a metered line. 
Each diver swam two simulations against the current, 
and two with the current. The cross-sectional area of 
a cinetransect was treated as an ellipse with a minor 
(vertical) axis of 1.5 m, the distance above and below 
the diver that fish were photographed. The major 
axis of the ellipse was a function of camera range, the 
distance at which fish could be distinguished on film. 
The particular function was cos 30° X camera range, 



since divers photographed fish within a 120° arc (60° 
on each side of the transect axis) (Fig. 2). Thus, the 
volume of cinetransects at a given depth on a given 
day was calculated as 

V= 1.5 ttL (cos 30° X CR), 

where V was cinetransect volume in cubic meters; 
1.5, the minor axis of the ellipse; L, the cinetransect 
length as determined above; and CR, the camera 
range at that depth on that day. Camera range itself 
was estimated as a function of the horizontal 
visibility at a depth on a sampling date. 

The relationship between camera range and 
horizontal visibility was estimated empirically under 
different conditions. The main "other condition" 
that we evaluated was the orientation of the camera 
to the sun. In trials run at different visibilities, two 
fish of similar appearance (usually a kelp perch, 
Brachyistius frenatus, and a white seaperch, 
Phanerodon furcatus) were held on a spear by one 
diver and photographed with our usual equipment by 
another diver at distances decremented from the 
limits of horizontal visibility (measured as described 
above). At each visibility, trials were run with the 
camera facing into the sun and with the camera facing 
away from the sun. Two observers viewed the film 
from each trial and determined camera range as the 
greatest distance at which the two fish could be dis- 
tinguished on film. The criteria for distinguishability 
were the same as those used in evaluating whether or 
not to count a fish when we viewed regular 
cinetransects. 

Data for camera range versus horizontal visibility 
were fit to several asymptotic functions. The fitting 



CINETRANSECT VOLUME 



A. CINETRANSECT 
SHAPE 




B. CINETRANSECT 
CROSS SECTION 



Camera Range 



FIGURE 2.— A. Estimated shape of area sampled in under- 
water transects taken with motion pictures (cinetransects). 
The length of 76 m was estimated from simulated tran- 
sects. B. Elliptical cross section of a cinetransect, with 
minor axis (a) of 1.5 m and major axis (b) calculated from 
camera range when divers surveyed a 120° horizontal arc 
about the central axis of the transect. 



FISHERY BULLETIN: VOL. 82, NO. 1 



routine was BMDP program P3R, nonlinear regres- 
sion (Dixon and Brown 1979). The function with the 
smallest residual mean square was selected to repre- 
sent the relation between camera range and horizon- 
tal visibility, and was employed in estimating camera 
range at a depth on a sampling date. 

Data Analysis 

We reduced data into two general forms: densities 
(number or biomass per unit volume) in different 
strata, and abundances integrated throughout the 
entire water column. The first was used to examine the 
vertical distribution of individual species or of the 
entire assemblage and to compare the relative abun- 
dances of species in a stratum. The second was used 
to estimate the overall abundance of the assemblage 
and to compare the overall abundances of different 
species. In both cases, the final point and interval 
estimates were based on the means and variances, 
over dates, of daily means. 

The daily estimate of density (per 1 ,000 m 3 ) for each 
species in a depth stratum was estimated as the mean 
number or biomass per transect on that day, times 
the ratio (1,000/transect volume), where transect 
volume was estimated as above. Biomass of a species 
on a given transect was estimated by counts of 
individuals in different maturity classes, converted 
to wet weights by the key in Table 1. 

Our estimate of a species' density in a depth 
stratum was calculated as the mean of the daily den- 
sity estimates in that stratum. Similar estimates were 
made for the sum of all "resident" teleosts. Excluded 
from the analysis of total fish density and abundance 
were elasmobranchs and certain teleosts (silver- 
sides, jack mackerel, Pacific barracuda, black 
croaker, and salema) that were rare at SOK, are 
seasonal visitors to kelp beds, or are not primarily 
associated with rock reefs and kelp forests (Feder et 
al. 1974). Species such as white seaperch and barred 
sand bass often occur in other habitats, but were 
included in our analysis because they may have at 
least a marginal association with kelp-rock habitats 
and were frequently encountered and abundant in 
our samples. 

By weighting the average density of a species (or the 
assemblage) in a stratum by the volume of water rep- 
resented by samples in that stratum, we were able to 
obtain estimates of abundance integrated from sur- 
face to bottom (Snedecor and Cochran 1980:444). 
The sampling day was an integral component of our 
analysis, but only the above-bottom strata were 
sampled on the same day at a given site. To obtain 
accurate estimates of variance for integrated abun- 



dances, then, we assembled our integrated estimates 
in two stages. We first estimated stratified mean den- 
sity for the above-bottom strata on each day and 
averaged these values over days. We also computed 
mean density (over days) in the bottom stratum. 
Secondly, we computed stratified mean density (and 
its standard error) for the above-bottom and bottom 
strata, using the means and variances calculated 
above. The stratified mean density estimates for the 
entire water column were then scaled to represent 
abundances over 100 m 2 of bottom. 

Samples in each stratum were assumed to represent 
a range of depths extending to the midpoints be- 
tween strata, with the 3 m stratum also extending to 
the surface (Table 2). Weighting factors for the strata 
were determined from the relative extents of the 
depth ranges represented. Among the above-bottom 
strata, relative weighting factors were the vertical 
ranges of these strata divided by 13. 5 m. For the bot- 
tom versus above-bottom strata the depth ranges 
were divided by 15 m. 

Daily estimates of stratified mean density in the 
above-bottom strata were calculated as 

D m = 2 W h D h , 

where D m was the estimate of stratified mean density 
in the 3 m, 7.6 m, and 12 m strata; W,„ the weighting 
factor; and D h , the mean density on that day in 
stratum h (Snedecor and Cochran 1980). The mean 
(D wc ) and variance (S 2 U J of these daily estimates were 
then computed. The mean {D b ) and variance (S 2 6 ) of 
estimated daily densities on the bottom were also 
calculated. 

Stratified mean abundance throughout the entire 
water column was estimated as 



A 



.-( 



1,500 
1,000 



XW h D 

n 



/!> 



where A st was the stratified mean estimate of 
integrated abundance over 100 m 2 of bottom, W h was 
the weighting factor, and D h was the mean density in 
either the above-bottom strata {D wc ) or in the bottom 
stratum (D h ). The term in the summation is the 
estimate of stratified mean density (per 1,000 m 3 ) 
over all strata, and the ratio (1,500/1,000) converts 
this value to abundance over 100 m 2 of bottom. 
The standard error of A st was calculated as 

 4 " v v 1,000 ' h h '' h 

where S 2 h was the variance of daily density estimates 
in either the above-bottom (S 2 ,,,.) or bottom (S 2 b ) 



42 



LARSON and DeMARTINI: DISTRIBUTION OF FISHES IN KELP FOREST 



strata; W h , the weighting factor; and n h , the number of 
days sampled in stratum h. The portion of the for- 
mula included in the summation is the usual estimate 
of variance for stratified means (Snedecor and 
Cochran 1980), and the root of this sum is the stan- 
dard error of mean density (per 1,000 m 3 ) throughout 
the water column. Multiplying by (1,500/1,000) 2 
adjusts the standard error for the larger volume of 
water in the column over 100 m 2 . 

Estimates of integrated abundance at the kelp- 
depauperate site were obtained by converting mean 
density on the bottom to mean density over 100 
m . 

Arithmetic means (of untransformed data) were 
used for all estimates of density and abundance. 
Geometric means (obtained by back-transforming 
the means of log-transformed data) underestimate 
absolute densities in a manner proportional to their 
variances. Adjustments for this underestimation 
(Elliott 1971) are usually based on the assumption of 
log-normal distributions, and we could not make such 
an assumption. However, some statistical com- 
parisons were made with log-transformed data to 
avoid the problem of heterogeneous variances. 
These were comparisons of mean numbers and 
biomass on the bottom, where varying transect 
volume did not confound the calculation of variance. 
Other comparisons, however, were made with 
untransformed data. These included tests for dif- 



ferences in numbers or biomass in the above-bottom 
strata and in the entire water column. When all three 
areas were compared, a one-way ANO VA was used if 
variances were not heterogeneous. T'-tests for un- 
equal variances (Bailey 1959) were used for pairwise 
comparisons of areas when variances were un- 
equal. 

RESULTS 
Cinetransect Calibration 

We estimated cinetransect length to be about 76 m. 
Six down-current trials averaged 78.3 m in length 
(standard error (SE) = 1.5 m, range = 74-82 m), 6 
upcurrent trials averaged 72.8 m in length (SE = 2.3 
m, range = 67-82 m), and the overall average was 
75.6 m (SE = 1.5 m). 

Camera range was an asymptotic function of 
horizontal visibility, with little increase in camera 
range at visibilities beyond 7-9 m (Fig. 3). Camera 
range was appreciably lower when the camera was 
facing the sun than vice versa, particularly at greater 
visibilities. This was reflected in each of the curves fit 
(Table 3). Since divers did not record whether actual 
transects faced into or away from the sun, we used the 
curve fit to all camera range-horizontal visibility 
values to calibrate cinetransect volume. The logistic 
equation provided, by slight margin, the best fit to 



O 3 

< 

2 . 



< 



1 



284 + 1 893(0.582 * 



<8> 



* 




-*e e- 



• INTO SUN 

: AWAY FROM SUN 



1 2 3 4 5 6 7 8 9 10 11 12 13 

HORIZONTAL VISIBILITY (m) 

Figure 3.— Relation of camera range (the distance at which fish could be distinguished on film) 
and horizontal visibility. Points are observations of maximum camera range at different visi- 
bilities with the camera facing into and away from the sun. The equation and line show the logistic 
function fit to these points. 



43 



FISHERY BULLETIN: VOL. 82, NO. 1 



these data (Table 3) and was the one employed in 
calculating cinetransect volume. 

Distribution and Abundance of Fishes 

Five sets of bottom transects were made in each 
study area. Water-column samples were taken on five 
dates at SOK-U and on four at SOK-D. Transect 



Table 3. — Functions fit to camera range (Y) versus horizontal visi- 
bility (X) relationship, and the best fit parameters as determined by 
BMDP program P3R (Dixon and Brown 1979). Also noted are the 
asymptotes calculated for each equation and data set, and the resid- 
ual mean squares. Into= trials made with the camera facing into the 
sun; Away = trials made with the camera facing away from the sun; 
All = curves fit to all data. Pj, P 2 , and P 3 are arbitrary symbols for 
the parameters of each function; there is no implied correspondence 
between the numbered parameters of different functions. 













Asymp- 


Residual 


Function name 


Set of 








tote 


mean 


and formula 


trials 


P , 


? 7 l 


P 3 


(m) 


square 


Logistic 


All 


284 


1.89 


0.582 


3.52 


0369 


Y 


Away 


0.259 


2.63 


560 


3.86 


250 


Y= 1/(P, + P 3 ) 


Into 


317 


1.20 


0618 


3.15 


0355 


Gompertz 


All 


1.27 


-3 19 


0647 


' 3.56 


0370 


Y=e |P . + P . p J 


Away 


1 37 


-3.88 


648 


3 94 


0255 


Into 


1 15 


-2.35 


0.656 


3.16 


0.354 


Von Benalanffy 


All 


360 


0334 


1.43 


3.60 


0.372 




. p , Away 


4.03 


0.301 


1.62 


4.03 


0.261 


Y= P, (1 - e 2 


3 ) Into 


3 17 


0361 


1.07 


3 17 


0353 


Michaelis-Menton 


All 


4.21 


1.92 


203 


421 


0.377 


P, I* - PJ 


Away 


4 94 


1.91 


2.79 


4.94 


0269 


Y- ' 2 


Into 


3.51 


2.01 


1 28 


3 51 


0354 


P +X- P 














3 2 
















AM 


0.194 


1.06 


— 


5.15 


0388 


Beverton-Holt 


Away 


0.158 


1.17 


— 


633 


0284 




Into 


0241 


092 


— 


4.15 


0352 


Y= 1/|P, + P 2 /X) 















number and visibility at depth on each date are 
shown in Table 4. 

Of the 28 species recorded in this study, 19 were 
"resident" teleosts. Of these, 13 species were record- 
ed on more than two transects in the two kelp-forest 
areas (Table 5). These 13 common species could be 
assigned to bathymetric categories, based on their 
vertical patterns of frequency of occurrence (Table 
5) and density (Tables 6, 7) within SOK. 

Kelp perch, halfmoon, and giant kelpfish were most 
common in the upper strata and are designated 
"canopy" species. While halfmoon and giant kelpfish 
were observed in all strata, all three species were 
most abundant in the 3 m stratum. Only halfmoon 
reached moderate abundances at 7.6 m in the SOK-D 
area (Tables 5, 6, 7). 

Sehorita, white seaperch, and kelp bass were com- 
mon throughout the water column (Tables 5, 6, 7) and 
are designated "cosmopolites". These three species 
were among the most common and abundant fishes in 
all strata. The white seaperch was the most cos- 
mopolitan of the three in 1979, its density and fre- 
quency of occurrence on transects varying little with 
depth. The sehorita was the most abundant species 
in nearly all strata. The kelp bass was also abundant 
at all depths. Its numerical density varied little 
among the water-column strata, but was generally 
greater on the bottom. Its biomass was greater in the 
lower strata (Tables 6, 7). Young kelp bass concen- 
trated in the upper water column (Table 8), con- 
tributing to the relatively low biomass per fish for 
kelp bass in the 3 and 7.6 m strata. Our data indicate 



Table 4. — Sampling dates, number of transects, and visibilities measured during fall 1979 sampling in two areas within 
the kelp bed at San Onofre (SOK-U and SOK-D) and in a nearby cobble-bottom area with little kelp (Cobble). Horizontal 
visibility (vis.) measured in meters 











SOK-U 












SOK-D 








Cobble 




3 


m 


7.6 


m 


12 


m 


Bottom 


3 


m 


7.6 m 


12 


m 


Bottom 


Bottom 


Date 


V 


VIS. 


V 


vis. 


V 


vis 


V 


VIS. 


V 


VIS 


V VIS. 


V 


VIS. 


V 


vis. 


V 


VIS. 


10 Oct 














10 


2 95 




















15 Oct. 




























9 


2.14 






17 Oct. 














9 


2.89 
















7 


300 


22 Oct 














10 


2.75 












9 


3.42 






24 Oct. 
































10 


2.60 


26 Oct 


















1 1 


1400 


12 8.50 


1 1 


3.50 










31 Oct. 














10 


3.85 
















10 


5.00 


7 Nov. 




























10 


3.90 






12 Nov. 


12 


7.30 


12 


5 10 


12 


4.75 
























14 Nov 
































10 


5.50 


16 Nov. 




























10 


4.50 


10 


4.85 


21 Nov. 














10 


8.75 




















26 Nov. 


















12 


10.25 


12 700 


12 


4.00 










28 Nov. 




























10 


4.00 






30 Nov 


12 


12.55 


12 


7.05 


12 


3 15 
























5 Dec. 


















12 


16.00 


12 13.75 


12 


7.25 










7 Dec 


12 


10.50 


12 


585 


12 


5.10 
























10 Dec. 


















12 


8.25 


12 7.80 


12 


6.90 










12 Dec. 


12 


9.45 


12 


6.95 


12 


850 
























19 Dec. 


13 


10.50 


12 


8 50 


12 


5.25 
























Total 


61 




60 




60 




49 




47 




48 


47 




48 




47 




Mean 




10 06 




6.69 




5 35 




4.24 




12.13 


926 




541 




3.59 




4.19 



44 



LARSON and DeMARTINI: DISTRIBUTION OF FISHES IN KELP FOREST 



that the upper kelp canopy serves as a nursery for 
young-of-the-year kelp bass, and these cryptic fish 
were probably much more abundant there than 
shown by our counts. We examined vertical segrega- 



tion of size classes only for kelp bass. This is because 
our 1979 data were too few to evaluate vertical 
segregation by size that has since been noted for two 
other species (senorita and blacksmith) in several 



Table 5.— Percent of transects on which species were observed during fall 1979, in two portions of a kelp forest near San 
Onofre, Calif. (SOK-U and SOK-D) and in a nearby kelpless cobble area (Cobble). Species' ranks are shown in 
parentheses. Number of transects is noted in the column heading. 







SOK-U 






SOK 


■D 




Cobble 




3 m 


7 6 m 


12 m 


Bottom 


3 m 


7.6 m 


12 m 


Bottom 


Bottom 


Species 


n=61 


n=60 


n=60 


n=49 


n=47 


n=48 


n=47 


r?=48 


n=47 


kelp bass 


52(3) 


50(3) 


60(1.5) 


61(2.5) 


74(2) 


77(2) 


81(1) 


81(2) 


26(4.5) 


barred sand bass 




2(10) 


8(4.5) 


59(4) 






9(8) 


58(5) 


53(1) 


kelp perch 


59(2) 


13(4) 


2(13) 




49(3) 


10(6.5) 


9(8) 






black perch 






8(4.5) 


41(5) 






9(8) 


65(3) 


26(4.5) 


white seaperch 


41(4) 


58(2) 


60(1.5) 


39(6.5) 


40(4) 


56(3) 


62(3) 


44(7.5) 


15(8) 


pile perch 




2(10) 


3(10) 


20(9) 


4(9) 




17(5) 


42(9) 


11(9) 


rubberlip seaperch 






3(10) 


16(10) 








19(10) 


4(12.5) 


rainbow seaperch 








39(6.5) 








44(7.5) 


19(6.5) 


senorita 


93(1) 


87(1) 


58(3) 


61(2.5) 


96(1) 


94(1) 


66(2) 


63(4) 


43(2) 


California sheephead 






5(7.5) 


58(1) 


2(10.5) 


19(5) 


36(4) 


90(1) 


36(3) 


rock wrasse 








29(8) 




2(9.5) 


4(10) 


46(6) 


6(105) 


opaleye 


3(8) 
















2(15) 


halfmoon 


16(6.5) 


7(6) 


2(13) 


2(14) 


36(5) 


38(4) 


11(6) 


4(12.5) 


19(6.5) 


blacksmith 




2(10) 










2(11.5) 






garibaldi 
















4(12.5) 




giant kelpfish 


24(5) 


8(5) 


7(6) 


3(11) 


21(6) 


10(6.5) 




4(12.5) 




cabezon 








2(14) 








2(16.5) 




California scorpionfish 
















2(16.5) 


4(12.5) 


rockfish spp. 






3(10) 


2(14) 










6(10.5) 


black croaker 
















2(16.5) 




salema 
















4(12.5) 




silversides 


16(6.5) 








19(7) 










jack mackerel 


2(9.5) 


3(7.5) 


5(7.5) 




17(8) 


8(8) 


2(11.5) 






Pacific barracuda 










2(10.5) 


2(9.5) 








leopard shark 








2(14) 












thornback 


















2(15) 


bat ray 








2(14) 








2(16.5) 




Pacific electric ray 


2(9.5) 


3(7.5) 


2(13) 












2(15) 



TABLE 6. — Mean numerical and biomass densities (per 1,000 m 3 ) of fishes observed in n daily samples per depth stratum at the SOK-U area in 
the San Onofre kelp bed during fall 1979. Values are the grand means (± 1 standard error) of the daily means (adjusted for transect volume) over 
transects taken each sampling day. 













SOK-U 












Numerical dens 


ity (no/1. 000 m 3 ) 






Biomass densi 


ty (kg/1.000 m 3 ) 






3 m 


7.6 m 


12 m 


Bottom 


3 m 


7 6 m 


12 m 


Bottom 




l«=5) 


C=5) 


(n=5) 


(n=5) 


(n=5) 


(n=5) 


0=5) 


(r>=5) 


Species 


x SE 


x SE 


x SE 


x SE 


x SE 


x SE 


x SE 


x SE 


kelp bass 


1.57 0.87 


2 67 1 19 


2 48 093 


4.76 1.20 


0.091 0071 


0416 0.171 


0664 0270 


1 372 372 


barred sand bass 





0.02 0.02 


0.13 0.04 


3.30 0.70 





0024 0.024 


0.173 046 


4.434 0930 


kelp perch 


1 39 0.26 


0.23 13 


0.02 0.02 





0.035 0.007 


0006 0.003 


neg. 





black perch 








12 0.07 


2.25 065 








0046 0.028 


0.717 209 


white seaperch 


1.91 1.21 


3.16 1.20 


2.33 0.86 


3.07 59 


0.319 0210 


0491 0.209 


287 0.105 


0.376 0108 


pile perch 





002 0.02 


0.08 0.05 


0.66 0.11 





009 0.009 


0039 0.025 


0263 0.079 


rubberlip seaperch 








0.04 0.03 


1.08 35 








028 0.017 


0634 0.265 


rainbow seaperch 











2.02 0.92 











0.167 068 


senorita 


2695 6.53 


2445 5.78 


4.66 2 22 


14.16 5.95 


0950 0.223 


1.103 0225 


0.241 110 


0566 0.237 


California sheephead 








013 0.06 


4.87 1 16 








0058 0.040 


1.561 0338 


rock wrasse 











1.20 1.24 











0.237 0022 


opaleye 


0.03 0.03 











0033 0.033 











halfmoon 


0.27 0.20 


0.08 0.05 


0.02 02 


006 006 


0068 0050 


020 012 


0006 0.006 


0.015 0.015 


blacksmith 





002 0.02 











neg. 








garibaldi 


























giant kelpfish 


0.35 0.08 


009 0.04 


0.08 0.04 


0.18 0.12 


018 0.007 


0004 0003 


0.015 008 


0.014 0.012 


cabezon 











0.06 0.06 











0089 0089 


Calif, scorpionfish 


























rockfish spp. 








0.04 0.02 


0.06 0.06 








0.003 0.003 


0.024 0.024 


black croaker 


























salema 


























silversides 


4.21 1.54 











0092 0.029 











jack mackerel 


0.09 0.90 


8 77 8.74 


0.50 0.36 





0010 0.010 


1 008 1.005 


0057 0.041 





Pacific barracuda 


























leopard shark 











0.06 0.06 











119 0.119 


thornback 


























bat ray 











006 0.06 











0.397 0.397 


Pacific electric ray 


001 0.01 


003 002 


0.02 0.02 





0136 0136 


320 196 


0.154 0.154 






45 



FISHERY BULLETIN: VOL. 82, NO. 1 



TABLE 7.— Mean numerical and biomass densities (per 1,000 m 3 ) of fishes observed inn daily samples per depth stratum at the SOK-D area in 
the San Onofre kelp bed during fall 1 97 9. Values are the grand means (± 1 standard error) of the daily means (adjusted for transect volume) over 
transects taken each sampling day. 













SOK-D 














Numerical densi 


ty (no/1, 000 m 3 ) 






Biomass density (kg/1 .000 m 3 ) 






3 m 


7.6 m 


12 


m 


Bottom 


3 m 


7.6 m 


12 m 


Bottom 




(n=4) 


(1=4) 


(n = 


=4) 


(n=5) 


(n=4) 


(n=4) 


(1=4) 


(1=5) 


Species 


x SE 


x SE 


* 


SE 


x SE 


. SE 


x SE 


x SE 


x SE 


kelp bass 


4 23 0.63 


4 61 1.09 


4.84 


1.07 


12.87 395 


726 162 


1.101 0.440 


1621 0.672 


2 363 0675 


barred sand bass 
kelp perch 



0.83 0.20 



0.19 0.09 


0.12 
0.11 


0.02 
0.08 


3 14 029 





0.021 005 



0005 0002 


0.178 0.029 
0003 0.002 


3446 0.577 



black perch 
white seaperch 
pile perch 



3.50 2.38 
0.04 0.04 




415 1 52 




18 
4.83 
0.23 


0.11 
94 
0.13 


4 77 0.63 
3.64 148 
1.74 0.18 



0.582 407 
013 0.013 




0681 0269 




040 0.017 
693 180 
0.105 0.068 


1.401 089 
399 0.137 
0.682 0.056 


njbberlip seaperch 
rainbow seaperch 
sehonta 

California sheephead 
rock wrasse 






19 46 2.82 

0.02 02 








2104 3 57 

0.60 0.23 

0.02 0.02 


568 
1.52 
0.06 






1.82 
0.42 
0.03 


064 0.24 

249 0.71 

13.31 7.77 

13.66 1.29 

1.86 0.49 





0569 0.078 
0.017 0.017 







1.039 0.158 
0.181 0.119 
0005 0.005 





0.312 0.100 
0.770 0.386 
0028 0004 


0.447 0.165 
0.238 0.053 
0.435 0.205 
4990 0.322 
0.405 0.110 


opaleye 
halfmoon 



1.09 0.44 



2.92 1.83 


0.35 



0.19 



0.12 0.12 



0.237 1 10 



0730 0.457 



0087 0047 



0.030 0.030 


blacksmith 








0.03 


0.04 











neg. 





garibaldi 
giant kelpfish 
cabezon 




0.28 0.06 






10 0.02 











0.12 0.07 
012 0.07 
007 0.06 




0024 0007 






0008 0004 









0.014 0.009 
0.012 0.010 
0099 0099 


Calif, scorpionfish 
rockfish spp. 
black croaker 








<) 












006 006 



11.85 11.85 



















0.033 0033 


2.667 2.667 


salema 
silversides 
jack mackerel 
Pacific barracuda 




5 99 3.96 

2096 9.05 

0.13 0.13 






19.34 17 69 

0.61 0.61 


3.32 





3.32 



889 5.93 






0.120 0.079 
2.410 1.040 
0.019 0.019 





2.224 2.035 
0.092 0092 






0.381 0.381 




0.667 0.444 





leopard shark 

thornback 

bat ray 

Pacific electric ray 





























0.12 0.12 



























0.794 0.794 





TABLE 8. — Mean numerical densities (per 1,000 m 3 ) of young-of- 
the-year (yoy), all juveniles (including yoy), subadult, and adult kelp 
bass inn daily samples per depth stratum at SOK-U and SOK-D 
during fall 1979. Grand means calculated as in Tables 6 and 7. 









Numerical dens 


ity (no./1,000 m 3 ) 








3 m (i 


n = 5) 


7,6 m 


(n = 5) 


12 m 


|i = 5) 


Bottom (n = 5) 


SOK-U 


X 


SE 


X 


SE 


X 


SE 


< 


SE 


yoy 

all juvs. 
subadults 
adults 


065 
1 23 
0.32 
002 


0.20 
062 
0.25 
0.01 


036 
085 
1.76 
0.05 


0.12 
0.34 
0.94 
04 


0.10 
0.90 
1.18 
0.40 


0.05 
0.38 
0.52 
0.24 


024 
1 36 
2.59 
0.80 


0.24 
0.76 
069 
026 




3 m (, 


n = 4| 


7,6 m 


|i = 4) 


12 m 


(1 = 4) 


Bottom 


(1 = 5) 


SOK-D 


X 


SE 


X 


SE 


X 


SE 


X 


SE 


yoy 

all juvs. 
subadults 
adults 


0.88 
1.50 
2 52 
0.20 


0.42 
62 
0.98 
0.12 


033 
1.24 
2.88 
0.49 


13 
0.19 
0.87 
028 


020 
1.37 
2 39 
1.08 


0.09 
0.50 
0.84 
0.49 


0.12 
5.21 
6 72 
0.94 


0.12 
3.26 
3.05 
18 



kelp beds off northern San Diego County (DeMartini 
et al. 6 ). 

Seven of the 13 common species were most abun- 
dant near the bottom (Tables 5, 6, 7). Rainbow 
seaperch and rock wrasse rarely, if ever, strayed 
above the bottom. Black perch and rubberlip 
seaperch were recorded occasionally at 12 m, but 



6 E. DeMartini, F. Koehrn, D. Roberts, R. Fountain, and K. Plum- 
mer. Variations in the abundances of fishes within and between 
stands of giant kelp {Macrocystis pyrifera) during successive years. 
Manuscr. in prep. Marine Science Institute, University of Califor- 
nia, Santa Barbara, CA 93106. 



were much more abundant on the bottom. Pile perch 
were seen, at one site or the other, in all strata, but 
were most abundant on the bottom and at 12 m. 
Barred sand bass also concentrated on the bottom 
and, to a lesser degree, at 1 2 m. California sheephead 
were observed as shallow as 3 m at SOK-D, but no 
shallower than 12 m at SOK-U. 

Species composition and relative abundance in 
each stratum reflected the distributional patterns of 
the species (Tables 9, 10). The three cosmopolitan 
species were among the three to five most abundant 
species in every stratum, particularly above the bot- 
tom. At 3 and 7.6 m, they made up 89-99% of total 
numerical density. The remaining fish in these strata 
were mainly upper water-column species, with a few 
of the more errant bottom species (such as California 
sheephead and pile perch) entering at 7.6 m. The 
three cosmopolites again dominated the assemblage 
at 12 m, forming 86-94% of fish numbers. A few 
individuals of canopy species were present at 12 m, 
however, and a greater number of bottom species 
were observed. The bottom stratum contained the 
greatest number of recorded species, and individuals 
were distributed more evenly among these species. 
The cosmopolites were still among the most abun- 
dant species on the bottom, but several of the 
bottom-zone species (such as California sheephead, 
black perch, and barred sand bass) were also abun- 



46 



LARSON and DeMARTINI: DISTRIBUTION OF FISHES IN KELP FOREST 

Table 9.— Percent contribution of species to total numerical and biomass density at the SOK-U area of the San Onofre kelp bed during fall 
1979. Percentages are given by stratum and for abundance integrated throughout the water column. Only those species contributing 1% or 
more are listed. Stratum values are based on data in Tables 6 and 7; integrated abundances on Table 1 1. 



3 m 




7.6 m 




12 m 




Bottom 




Integrated 




Species 


% 


Species 


% 


Species 


% 


Species 


% 


Species 


% 


SOK-U Numbers 




















senorita 


83.0 


senonta 


79.5 


senorita 


46 


senorita 


37.5 


senorita 


72.0 


white seaperch 


59 


white seaperch 


10.3 


kelp bass 


24 5 


Calif, sheephead 


12.9 


white seaperch 


9.3 


kelp bass 


4 8 


kelp bass 


8 7 


white seaperch 


230 


kelp bass 


12.6 


kelp bass 


9.1 


kelp perch 


43 






barred sand bass 


1.3 


barred sand bass 


8 7 


kelp perch 


2.1 


giant kelpfish 


1.1 






Calif, sheephead 


1.3 


white seaperch 


8 1 


Calif, sheephead 


2.0 










black perch 


1.2 


black perch 
rainbow seaperch 
rock wrasse 
rubberlip seaperch 
pile perch 


6.0 
54 
3.2 
2.9 

1.7 


barred sand bass 


1.4 


SOK-U Biomass 




















senonta 


62.7 


senorita 


53.2 


kelp bass 


42.6 


barred sand bass 


42.4 


senonta 


30.2 


white seaperch 


21.1 


white seaperch 


23 7 


white seaperch 


18 4 


Calif, sheephead 


149 


barred sand bass 


19.1 


kelp bass 


6.0 


kelp bass 


20 1 


senonta 


15.4 


kelp bass 


13.1 


kelp bass 


1 7.7 


halfmoon 


45 


barred sand bass 


1.2 


barred sand bass 


1 1.1 


black perch 


6.8 


white seaperch 


14.2 


kelp perch 


2.3 


halfmoon 


1.0 


Calif, sheephead 


3.7 


rubberlip seaperch 


6.1 


Calif, sheephead 


6.6 


opal 


2.2 






black perch 


2.9 


senorita 


5.4 


black perch 


3.2 


giant kelpfish 


1.2 






pile perch 


2.5 


white seaperch 


36 


rubberlip seaperch 


2 7 










rubberlip seaperch 


18 


pile perch 


2 5 


pile perch 


1.5 










giant kelpfish 


1.0 


rock wrasse 
rainbow seaperch 


2.3 
1.6 


halfmoon 


1.2 



Table 10. — Percent contribution of species to total numerical and biomass density at the SOK-D area of the San Onofre kelp bed during fall 
1979. Percentages are given by stratum and for abundance integrated throughout the water column. Only those species contributing 1% or 
more are listed. Stratum values are based on data in Tables 6 and 7; integrated abundances on Table 1 1. 



3 m 




7 6m 




12 m 




Bottom 




Integrated 




Species 


% 


Species 


% 


Species 


% 


Species 


'V. 


Species 


% 


SOK-D Numbers 




















senorita 


66.1 


senorita 


62.6 


senorita 


31.6 


Calif, sheephead 


23.3 


senorita 


51 7 


kelp bass 


14 4 


kelp bass 


13.7 


kelp bass 


27.0 


senorita 


22 7 


kelp bass 


174 


white seaperch 


119 


white seaperch 


12 3 


white seaperch 


27.0 


kelp bass 


22.0 


white seaperch 


13.1 


halfmoon 


3.7 


halfmoon 


8 7 


Calif sheephead 


8 5 


black perch 


8.1 


Calif, sheephead 


63 


kelp perch 


2 8 


Calif, sheephead 


1 8 


halfmoon 


1.9 


white seaperch 


6.2 


halfmoon 


4,4 










pile perch 


1.3 


barred sand bass 


5.4 


black perch 


1.7 










black perch 


1 


rainbow seaperch 
rock wrasse 
pile perch 
rubberlip seaperch 


4.2 
3.2 
3.0 
1.1 


kelp perch 
barred sand bass 


1.2 

1.1 


SOK-D Biomass 




















kelp bass 


32.6 


kelp bass 


29.4 


kelp bass 


42 2 


Calif, sheephead 


33 3 


kelp bass 


28.2 


white seaperch 


26.2 


senorita 


27.7 


Calif, sheephead 


20.1 


barred sand bass 


23.0 


Calif, sheephead 


17.2 


senonta 


256 


halfmoon 


19.5 


white seaperch 


1 8.1 


kelp bass 


15 8 


senonta 


14.5 


halfmoon 


12 3 


white seaperch 


18 2 


senonta 


8.1 


black perch 


9.3 


white seaperch 


14.3 


giant kelpfish 


1.1 


Calif, sheephead 


4.8 


barred sand bass 


4.6 


pile perch 


4,5 


barred sand bass 


89 










pile perch 


2 7 


rubberlip seaperch 


30 


halfmoon 


7 8 










halfmoon 


2.3 


senorita 


2 9 


black perch 


3.4 










black perch 


1.0 


rock wrasse 
white seaperch 
rainbow seaperch 


2.7 
2.7 

1.6 


pile perch 
rock wrasse 
rubberlip seaperch 


2.3 

1 1 
1.0 



dant. The gradual change in species composition that 
occurred between the water-column strata became 
more abrupt at the bottom. 

The vertical profile of total numerical density 
reflected changes in the abundance of the most 
numerous species, senorita, and the increase in 
species number on the bottom. Numerical density 
was about the same at 3 and 7.6 m, dropped at 12 m, 
and peaked on the bottom (Fig. 4). Small differences 
in species composition at 3 and 7.6 m led to only small 
differences in the abundances of noncosmopolites, 
and the cosmopolites (particularly senorita) had 
similar densities in these strata (Tables 6, 7). Despite 



increased abundances of bottom species at 12 m, the 
loss of upper water-column species and the decline in 
abundance of senorita led to low overall numerical 
densities in this stratum (Tables 6, 7). Senorita 
became more abundant again in the bottom stratum, 
kelp bass reached peak density, and the bottom 
species became abundant (Tables 6, 7), leading to 
high numerical densities on the bottom (Fig. 4). 

Biomass density did not differ among the water- 
column strata, but reached an exaggerated peak on 
the bottom (Fig. 5). At 12 m, the increase in size of 
kelp bass, and the addition of large-bodied species 
like California sheephead, barred sand bass, and 



47 



FISHERY BULLETIN: VOL. 82, NO. 1 



X 
(- 
Q. 



u - 
















t 




5- 


\ 1 
\l 
\l 

l\ 
1 \ 






• 


SOK U 
SOK D 














I . 








// 












/ / 












/ / 












/ / 












/ / 










io- 


/ / 
/ / 
/ / 
/ / 
/ / 

^ 






















 J 














1 1 1 1 


— i r 






— i 1 



10 20 30 40 50 60 70 80 

MEAN NUMERICAL DENSITY 
(Number/lOOOm 3 ) 



','« 



Q. 
LU 

Q 



0- 














1 \ 

1 \ 
1 \ 


• 


SOK U 
SOK D 






\ \ 
I \ 
1 \ 

\ \ 










111 J 


1 

1 

1 

1 

1 

1 

1 








"***-.«. — ^~_ 


IS- 









5 10 

MEAN BIOMASS DENSITY 
(kg/1000m 3 ) 



FIGURE 4. — Vertical distribution of the numerical densities of all 
resident teleosts in two areas within the San Onofre kelp bed during 
fall 1979. Points are mean densities over sampling dates at each site 
and stratum, and bars represent one standard error of the mean. 



FIGURE 5. — Vertical distributions of the biomass density of all resi- 
dent teleosts in two areas within the San Onofre kelp forest during 
fall 1979. Points are mean densities over sampling dates at each site 
and stratum, and bars represent one standard error of the mean. 



various embiotocids compensated for the decline in 
abundance of senorita (Tables 6, 7). The higher 
numerical densities of these large fishes on the bot- 
tom contributed most to the peak biomass densities 
in this stratum. 

Weighting densities for the size of stratum, we 
estimated that on average about 40 and 46 fish 
occurred over 100 m 2 at SOK-U and SOK-D, respec- 
tively, with corresponding biomass values of 3.9 and 
6.5 kg/100 m 2 (Table 11). About 66% (SOK-D) to 



77% (SOK-U) of all individuals occurred in the upper 
two strata, 9% (SOK-U) to 14% (SOK-D) at 12 m, and 
14% (SOK-U) to 19% (SOK-D) on the bottom. The 
small vertical extent of the bottom stratum 
diminished its contribution to the abundance offish 
integrated over the entire water column. About 44- 
45% offish biomass occurred in the two upper strata, 
15% (SOK-U) to 22% (SOK-D) occurred at 12 m, and 
34% (SOK-D) to 40% (SOK-U) on the bottom. Thus 
much of biomass was near the bottom, but because of 



Table 1 1. — Abundance of resident teleosts, based on densities integrated through the water column over 100 m 2 
of bottom. The standing stock in numbers and biomass is given for each of two areas (SOK-U and SOK-D) within 
the San Onofre kelp bed, and for an adjacent area of cobble bottom with little kelp (Cobble), for samples taken in 
fall 1979. 











Numbers 


1... 


, U)l) m 2 












B 


omass (kg) 


per 100 


m* 








SOK-U 


SOK-D 


C 


Dbble 


SOK-U 


SOK-D 


Cobble 


Species 


X 




SE 


X 




SE 


X 




SE 


X 




SE 


X 




SE 


X 




SE 


kelp bass 


3.66 




1.02 


8.04 




80 


0.25 




14 


0.67 




15 


1.83 




041 


12 




004 


barred sand bass 


0.55 




0.1 1 


0.52 




0.04 


116 




0.37 


0.74 




14 


0.58 




0.09 


1 69 




0.55 


kelp perch 


085 




0.20 


0.57 




05 









0.02 




0.01 


0.01 




0.01 









black perch 


038 




0.10 


0.78 




10 


54 




032 


0.13 




003 


023 




0.02 


0.19 




1 1 


white seaperch 


3 76 




1.43 


6.05 




1 54 


046 




31 


0.55 




0.24 


93 




030 


007 




0.06 


pile perch 


14 




0.03 


0.37 




007 


005 




002 


0.06 




0.02 


0.15 




003 


002 




001 


rubberlip seaperch 


18 




0.05 


0.10 




0.04 


03 




002 


0.11 




0.04 


0.07 




02 


0.02 




0.01 


rainbow seaperch 


030 




14 


037 




1 1 


001 




12 


0.03 




0.01 


04 




001 


001 




001 


senorita 


28 86 




4.63 


2388 




2.06 


2 16 




77 


1.17 




0.21 


0.95 




05 


0.06 




004 


Calif, sheephead 


78 




0.18 


2.89 




0.30 


0.61 




0.20 


0.26 




005 


1.12 




17 


18 




006 


rock wrasse 


18 




0.04 


0.31 




0.08 


0.03 




0.01 


004 




0.01 


0.07 




0.02 


0.01 




0.01 


opaleye 


0.02 




0.02 









001 




0.01 


0.02 




0.02 









0.01 




001 


halfmoon 


0.20 




13 


2 04 




1 03 


11 




005 


0.05 




003 


051 




026 


0.03 




0.01 


blacksmith 


001 




0.01 


001 




0.04 









neg 






neq 











garibaldi 









0.02 




0.01 


















neq 











giant kelpfish 


028 




007 


021 




0.04 









0.02 




001 


0.02 




001 









cabezon 


0.02 




001 


0.01 




0.01 









0.01 




0.01 


0.02 




0.02 









Calif, scorpionfish 









001 




001 


0.02 




001 











neq 




0.01 




001 


rockfish spp 


0.02 




0.01 









0.04 




0.03 


0.01 




001 









0.01 




0.01 


All residents 


40.4 




6.0 


46.2 




4.1 


5.6 




0.94 


3 '1 




0.5 


65 




0.7 


2.4 




0.6 



48 



LARSON and DeMARTINI: DISTRIBUTION OF FISHES IN KELP FOREST 



their more extensive bathymetric ranges, the low 
biomass- density upper strata still contributed nearly 
one-half of total biomass. 

The most abundant species at SOK were the cos- 
mopolites (Tables 9, 10, 1 1). Sehorita, kelp bass, and 
white seaperch comprised 82 and 90% of all 
individuals in the kelp forests at SOK-D and SOK-U, 
respectively. These species also contributed strongly 
to overall integrated biomass, although large species 
like California sheephead, barred sand bass, and 
halfmoon were also important. As a result, the dis- 
tribution of biomass among species was more even 
than the distribution of numbers (Tables 9, 10, 11). 

Two relatively large fishes were more abundant at 
SOK-D than SOK-U during fall of 1 979, contributing 
to the differences (see below) in our estimates of total 
biomass at each site (Table 1 1). The integrated abun- 
dance of kelp bass was significantly higher, or nearly 
so, at SOK-D (Numbers: t = 3.37, df = 7, 0.0K P < 
0.02; Biomass: t = 2.65, df = 4, 0.05 < P < 0.1). 
California sheephead were also more abundant at 
SOK-D, as tested with log-transformed bottom data 
(Numbers: t = 4.81, df = 6, P < 0.01; Biomass: t = 
3.35, df = 5, 0.02 < P < 0.05) and with integrated 
abundances (Numbers: t = 6.03, df = 5, P < 0.01; 
Biomass: t = 4.92, df = 4, P < 0.01). Halfmoon 
seemed to be more abundant at SOK-D, but the dif- 
ference was not significant (Numbers: t = 1.78, df = 
3, P > 0.1; Biomass: t = 1.78, df = 3, P > 0.1). 

At the kelpless cobble site, most fish were bottom 
species and cosmopolites (Tables 5, 11). While 
barred sand bass, black perch, and California 
sheephead were fairly abundant in this area, the 
average abundances of other species were less than 
in the kelp-bed areas. The integrated numerical 
abundance of all fishes was significantly lower in the 
kelpless cobble area (cobble vs. SOK-U: t = 5.71, df 
= 4,P< 0.01; cobble vs. SOK-D: t = 9.42, df = 3,P < 
0.01; SOK-U vs. SOK-D: f = 0.79, df= 7, P> 0.4). A 
one-way ANOVA of log-transformed counts on the 
bottom showed significant differences among the 
three areas (F 2 12 = 9.42, P < 0.01), but an a priori 
comparison of SOK-U and SOK-D versus the cobble 
area was not significant (F x 12 = 1.207, P > 0.25). 
Thus, the lower overall numerical abundance at the 
kelpless cobble area was due largely to the presence 
offish above the bottom at SOK. The integrated total 
biomass of fish did not differ significantly among the 
three areas (F 2>11 = 0.25, P > 0.75), even though the 
point estimate of 2.4 kg/100 m 2 at the cobble area 
was lower than both values at SOK. However, barred 
sand bass made up over 70% of fish biomass in the 
cobble area, so most other species were much less 
abundant there. 



We estimated the density of Macrocystis plants >1 
m tall to be 7.51 ± 0.71 (1 SE) plants/ 100 m 2 at the 
"kelpless" cobble area, 23.11 ± 1.47 plants/100 m 2 at 
SOK-U, and 30.18 ± 1.69 plants/100 m 2 at SOK-D. 
Thus, some kelp was present at the cobble area, but 
the density of subadult-adult plants there was 25- 
32% of density in our kelp-bed areas. 

DISCUSSION 

Sampling 

Regardless of water clarity, our camera and film 
were unable to resolve fish beyond 3-4 m; this set an 
upper limit of just over 1,000 m 3 to cinetransect 
volume. Alevizon and Brooks (1975) noted that in 
very clear, shallow waters, fish seemed difficult to 
distinguish on film beyond 5 m. Ebeling et al. ( 1 980b) 
found camera range to be 3-3.5 m at horizontal 
visibilities of 4 and 15 m, and concluded that there 
was essentially no relation between camera range 
and horizontal visibility. Our data show this to be true 
at visibilities >7-9 m. The fixed focal length of the 
camera, shallow depth of field at maximum aperture, 
and quality of film account for the limited camera 
range, as discussed by Ebeling et al. (1980b). 
However, our data show that camera range decreases 
when visibility decreases to values that approach 
maximum camera range. Corrections for visibility are 
common in terrestrial line transects, whether the 
area of a given transect is taken as fixed throughout or 
as variable (Caughley 1977; Burnham et al. 1980). 
We regarded the volume of a given cinetransect to be 
fixed, its width determined by visibility. 

The relatively low upper limit to camera range may 
help to make cinetransects in the water column more 
accurate than visual censuses. Searching efficiency 
would likely be poorer for broad visual transects 
made to the limits of visibility. Furthermore, it is dif- 
ficult to judge arbitrary smaller distances in open 
water, unless they are only a meter or two on either 
side of the diver. Cinetransects provide an almost 
automatic upper limit to transect width, and this limit 
is wide enough (about 3 m to either side in mod- 
erately clear water) that a substantial volume of 
water is censused. 

We have not verified the exact volume sampled in 
each of our cinetransects, nor are we able to compare 
densities measured in cinetransects with actual den- 
sities (Brock 1982), since the latter have not been 
measured by any method. To our knowledge, only 
Keast and Harker (1977) have actually marked the 
outside boundaries of visual underwater transects. 
However, Terry and Stephens (1976) and Stephens 



49 



FISHERY BULLETIN: VOL. 82, NO. 1 



and Zerba (1981) utilized two divers, swimming 
parallel, unmarked courses and counting fish be- 
tween each other, to sample rocky-reef fishes. 
Perhaps such a method could be used to evaluate 
densities estimated in cinetransects. 

Species Composition, Distribution, 
and Abundance 

The species observed in the San Onofre kelp forest 
were a subset of the species found in other nearshore 
areas of hard substrate and vegetation off southern 
California. Many reef-dependent fishes that are very 
common in other kelp forests were either rare or 
unrecorded at San Onofre. Species such as black- 
smith and opaleye (Ebeling and Bray 1976; Hobson 
and Chess 1976), garibaldi (Clarke 1970), painted 
greenling (DeMartini and Anderson 1979), and some 
species of Sebastes (Larson 1980) depend on rugose 
reefs for shelter or spawning sites. Some turf-grazing 
and otherwise bottom-feeding species of embi- 
otocids also appeared to be less abundant at San 
Onofre than in other areas. Our estimates of 14-37 
kg/ha of pile perch, 38-78 kg/ha of black perch, and 
10-18 kg/ha of rubberlip seaperch were mostly 
smaller than the estimates of Ebeling et al. (1980b) 
off Santa Barbara and Santa Cruz Island. The rarity and 
low abundance of all these species markedly alters 
the character of the fish assemblage at San Onofre. 
The abundant species at San Onofre kelp forest 
either are less dependent on rock reefs (at least, if 
kelp is present) or associate preferentially with low- 
relief substrates. The former group might include the 
canopy species, the cosmopolitan kelp bass and 
sehorita, and perhaps the epibenthic California 
sheephead. The latter group might include barred 
sand bass and white seaperch. These two species 
(and perhaps sehorita) were more common at San 
Onofre than others (Ebeling et al. 1980a, b) have 
reported in kelp forest anchored on high-relief sub- 
strates. Barred sand bass occurred in over half of the 
bottom transects at SOK, but in no more than 12% of 
bottom transects near Santa Barbara (Ebeling et al. 
1980a). We found white seaperch in 40-60% of our 
transects, while Ebeling etal. (1980a) saw them on 7- 
42% of all transects (but 20-42% of "sandy margin" 
transects). Both of these species have been reported 
as associating with sand or the sand-rock interface 
(Quast 1968a; Feder et al. 1974; Ebeling et al. 
1980a). Moreover, barred sand bass have a 
warmwater affinity (Frey 1971) and on average 
should be more abundant farther south in the 
Southern California Bight. The abundance of white 
seaperch at SOK may be unusually high during the 



fall. At this time, white seaperch appear to use the 
SOK habitat for mating as well as feeding. While 
some individuals of white seaperch are found in kelp 
forests all year, much of their populations in kelp 
beds off northern San Diego County move offshore 
after fall (authors' observations). 

The vertical distributions of species present at the 
San Onofre kelp bed were similar to patterns de- 
scribed in other kelp forests. Kelp perch, giant 
kelpfish, and, to a lesser extent, halfmoon have been 
recognized as water-column and canopy species 
(Quast 1968a; Feder et al. 1974; Bray and Ebeling 
1975; Ebeling and Bray 1976; Hobson and Chess 
1976; Coyer 1979; Ebeling et al. 1980a, b). Kelp bass 
and white seaperch have been described as members 
of a vertical "commuter" group of fishes in kelp 
forests near Santa Barbara (Ebeling et al. 1980a). 
The term "cosmopolite" better describes the habits 
of these two fishes. Sehorita also fell into Ebeling et 
al.'s "canopy" group, but its occurrence throughout 
the water column was recognized by Hobson (1971), 
Ebeling and Bray (1975), Bernstein and Jung (1979), 
and others. We feel that it too should be considered a 
cosmopolite. Pile perch and rubberlip seaperch were 
also assigned to the commuter group of Ebeling et al. 
(1980a) and did appear above the bottom at San 
Onofre. However, the dense midwater aggregations 
of these species observed elsewhere were not present 
at San Onofre. Perhaps the relatively low density of 
these species at San Onofre was responsible for the 
absence of these aggregations. On the other hand, 
our fairly frequent observation of California 
sheephead well above the bottom is apparently new. 
Quast (1968a), in fact, noted that sheephead seem 
"reluctant" to leave the bottom. Barred sand bass, 
black perch, rainbow seaperch, and rock wrasse 
occurred almost exclusively on the bottom, and have 
been generally recognized as bottom dwellers. 

Our estimates of vertically integrated standing 
stock were surprisingly high. Most estimates of fish 
biomass on tropical and temperate reefs fall into the 
range of a few to several hundred kg/ha (Brock 1954; 
Bardach 1959; Randall 1963; Quast 1968b; Talbot 
and Goldman 1972; Miller and Geibel 1973; Jones 
and Chase 1975; Russell 1977). It is encouraging that 
our estimates of 3.88-6.53 kg/100 m 2 (388-653 kg/ 
ha) fell within this range. Furthermore, our density 
estimates for fall 1979 are generally similar to subse- 
quent estimates made for canopy and bottom strata 
during the fall periods of 1980 and 1981 (E. DeMar- 
tini 7 Unpubl. data). In particular, the densities of resi- 



7 E. DeMartini, Marine Science Institute, University of California, 
Santa Barbara, CA 93106. 



50 



LARSON and DeMARTINI: DISTRIBUTION OF FISHES IN KELP FOREST 



dent species (kelp bass and California sheephead) 
that contributed most to biomass estimates for fall 
1979 were not consistently larger or smaller, if dif- 
ferent at all, at SOK during fall 1 980 and 1 98 1 . Hence 
we feel that our estimates for fall 1979 are typical for 
SOK during this season. Furthermore, while species 
such as kelp bass and sheephead were most abun- 
dant at SOK-D during fall 1979, this was not always 
true in 1980 and 1981; the site of greater abundance 
switched between SOK-U and SOK-D for many 
species over the period of 1979-81 (DeMartini et al. 
footnote 6). Thus we also conclude that apparent dif- 
ferences between SOK-U and SOK-D during fall 
1979, although perhaps statistically real, are not 
meaningful for our general characterization of stand- 
ing stock at SOK. For this reason, we have provided 
data for the areas separately as brackets for our 
estimates of conditions in the San Onofre kelp bed in 
general, and do not specifically attribute the greater 
abundance of fishes at SOK-D to greater numerical 
density of giant kelp plants > 1 m tall. 

The surprising aspect of our standing-stock 
estimates is that they are as large or larger than those 
of Quast (1968b) in nearshore areas of greater bot- 
tom relief. Subtracting elasmobranchs, "nonresi- 
dent" teleosts, and cryptic bottom species, his 
estimates of standing stock at two sites near San 
Diego were about 366 kg/ha for Del Mar and 299 kg/ 
ha for Bathtub Rock. Thus, even though our areas at 
San Onofre lacked many individuals of such great 
contributors to biomass at Quast's sites as opaleye, 
blacksmith, kelp rockfish, and garibaldi, our brack- 
eted values of biomass were of the same order to 
nearly twice Quast's estimates. Below, we examine 
three possible reasons for this perceived disparity: 
Bias due to sampling methods, bias due to the times 
and places sampled, and the possibility that there 
really was a relatively large standing stock of fishes at 
San Onofre. 

Our sampling methods may have led to over- 
estimates, or Quast's (1968b) to underestimates, of 
standing stock. Quast's quantitative collection at Del 
Mar lacked a wall net, so some fish may have escaped. 
Although he used transect densities for three of the 
abundant species in his corrected estimates, his tran- 
sect method of counting fish to the limits of visibility 
may have led to reduced searching efficiency (as dis- 
cussed above). It is less likely that we counted fish in a 
larger volume than we think. We may have inflated 
our estimates of integrated abundance by sampling 
the bottom stratum on different days than the water- 
column strata, so that the same individuals could 
have figured into average density in more than one 
stratum as distributions changed from day to day. 



Such errors would have been most serious in the cos- 
mopolitan species, and perhaps in large bottom 
species (like California sheephead) that also 
occurred in the water column. However, even in our 3 
m stratum, the average numbers of senorita and 
white seaperch per transect (uncorrected for 
visibility) were greater than similar averages 
obtained by Ebelingetal. (1980a, b) in cinetransects 
off Santa Barbara, implying that these species really 
were abundant during the fall at San Onofre. For kelp 
bass, the average standing stock above the bottom 
was 48 ± 13 (SE) kg/ha at SOK-U and 148+40 at 
SOK-D. These values are large fractions of our total 
respective estimates of about 69 and 183 kg/ha. 
Similarly, our estimates of sheephead biomass on the 
bottom alone were 23 ± 5 kg/ha at SOK-U and 75 ± 5 
kg/ha at SOK-D, compared with our total estimates 
of about 26 and 112 kg/ha at the respective areas. We 
conclude that, while sampling problems may have 
contributed some bias to both our estimates and 
those of Quast's, much of the difference between 
Quast's estimates and ours is real, and fish really 
were relatively more abundant in the areas we sam- 
pled at SOK during the fall. 

Our selection of sampling times and places could 
have led to estimates that are somewhat unrep- 
resentative of conditions in general at San Onofre. 
Seasonal factors might be involved for some of our 
"resident" species. Dense concentrations of some 
fishes (notably white seaperch) may be atypically 
high at SOK and perhaps other kelp beds during the 
fall, when these areas are used for breeding. Many 
species of fish can be found in kelp beds all year, but 
their abundances might nevertheless fluctuate 
greatly as individuals move among areas within kelp 
beds, between different kelp beds, and perhaps be- 
tween different nearshore habitats. We feel that our 
samples accurately characterize the standing stock 
of fishes at San Onofre kelp in the fall, but cannot 
extend our observations to other seasons. 

Horizontal patchiness in the distribution of fish may 
also have affected our estimates. Our kelp-forest 
sampling areas were near the offshore edge of a large 
area of surface canopy, and fish often were quite 
dense at the actual edge of the kelp forest. Limbaugh 
(1955), Quast (1968a), Feder et al. (1974), Hobson 
and Chess (1976), Bray (1981), and others have dis- 
cussed this "edge effect". Although many of our tran- 
sects did not (by chance) sample the edge of the bed, 
the averages we calculated nonetheless may have 
overestimated the density of some species through- 
out the entire bed. However, our estimates of fish 
density at the particular study areas should be 
relatively unbiased. Quast's (1968b) Del Mar collec- 



51 



FISHERY BULLETIN: VOL. 82. NO. 1 



tion was also made at the edge of a kelp forest, so 
comparison with our areas is warranted. 

The comparatively large standing stock of fishes at 
SOK in part reflects the nature of the kelp forest off 
San Onofre. This kelp forest was located in relatively 
deep (15 m) water, and was of moderate (0.1 adult 
plant/m 2 ; Dean footnote 4) kelp density, with a sur- 
face canopy. Both of Quast's (1968b) sites were 
located in relatively shallow (7.6-10.7 m) water. 
Furthermore, Quast's Bathtub Rock site lacked a 
surface kelp canopy. A substantial part of the fish 
biomass we observed at San Onofre was in the exten- 
sive canopy and midwater zones. Nearly half of the 
biomass occurred in the upper two strata at each site, 
and about one-quarter occurred in the midwater (7.6 
m) stratum alone. The contribution of the upper 
water column to overall standing stock is also illus- 
trated by the relative importanceof the cosmopolitan 
species. Ranging throughout the water column, kelp 
bass, white seaperch, and sehorita comprised about 
60% of total biomass at the San Onofre kelp bed. The 
relative contribution of water-column species to 
overall standing stock would be lower in kelp forests 
anchored on high-relief rock, because reef-de- 
pendent species would be more abundant than at San 
Onofre. However, the presence of an extensive 
bathymetric zone from the canopy into midwaters 
provided space, forage, and orientation for a substan- 
tial standing stock of fishes in the San Onofre kelp 
bed. The lack of such an extensive midwater zone 
may have limited the abundance of canopy and cos- 
mopolitan species at Bathtub Rock and Del Mar, 
accounting, in part, for the relatively low estimates of 
standing stock in these areas. 

Our study, then, suggests that kelp per se can 
enhance the potential standing stock of fishes in an 
area. Our kelp-forest areas lacked a high-relief bot- 
tom and the species of fish that depend on it. The 
remaining fish were those that either tolerate or are 
not influenced by a cobble bottom, and those that 
depend intimately on kelp. Yet the standing stock of 
fishes at the San Onofre kelp bed was substantial. 
The reduced numerical abundance of fishes and 
smaller biomass (excluding barred sand bass) in our 
kelp-depauperate area further indicates the impor- 
tance of kelp at San Onofre. Experimental manipula- 
tion of kelp density is probably the best test of the 
influence of kelp on fish abundance (Miller and 
Geibel 1973; Bray 198 1; M. Carr footnote 3). We also 
recognize that large-scale oceanographic factors may 
strongly affect survivorship of planktonic larvae and 
the subsequent abundance of juvenile and adult 
fishes (Stephens and Zerba 1981; Parrish et al. 
1981). However, our comparisons indicate that giant 



kelp, even in only moderate density, was necessary 
for the existence of a large standing stock of diverse 
fishes in cobble-bottom areas. We conclude that, 
while rock reefs enhance the fish fuana of an area 
whether or not there is kelp, the presence of kelp in an 
area of low-relief bottom also augments the abun- 
dance of juvenile and adult fish on a local scale. Kelp 
may also contribute strongly to the standing stock of 
fish in areas of high-relief bottom, but no one to date 
has adequately evaluated this hypothesis. We pre- 
dict that the densities of canopy species and cos- 
mopolites like kelp bass and sehorita will also prove 
to be related to the density of giant kelp on high- 
relief bottoms. 

ACKNOWLEDGMENTS 

We thank Ken Plummer and Mark Wilson for assis- 
tance with filming cinetransects. Sandy Larson and 
Jan Fox typed versions of the manuscript. Diane 
Fenster drafted Figures 2-4. This paper is a result of 
research funded by the Marine Review Committee 
(MRC), Encinitas, Calif. The MRC does not 
necessarily accept the results, findings, or con- 
clusions stated herein. A. W. Ebeling kindly loaned 
the cameras and housings used in the study. 

LITERATURE CITED 

Alevizon, W. S., and M. G. Brooks. 

1975. The comparative structure of two western Atlantic 
reef-fish assemblages. Bull. Mar. Sci. 25:482-490. 
Bailey, N. T. J. 

1959. Statistical methods in biology. English Univ. Press 
Ltd., Lond., 200 p. 
Bardach, J. E. 

1959. The summer standing crop of fish on a shallow Ber- 
muda reef. Limnol. Oceanogr. 4:77-85. 
Bernstein, B. B., and N. Jung. 

1979. Selective pressures and coevolution in a kelp canopy 
community in southern California. Ecol. Monogr. 
49:335-355. 

Bray, R. N. 

1981. Influence of water currents and zooplankton densities 
on daily foraging movements of blacksmith, Chromis 
punctipmnis, a planktivorous reef fish. Fish. Bull., U.S. 
78:829-841. 

Bray, R. N., and A. W. Ebeling. 

1975. Food, activity, and habitat of three "picker-type" mi- 
crocarnivorous fishes in the kelp forests off Santa Barbara, 
California. Fish. Bull., U.S. 73:815-829. 

Brock, R. E. 

1982. A critique of the visual census method for assessing 
coral reef fish populations. Bull. Mar. Sci. 32:269-276. 

Brock, V. E. 

1954. A preliminary report on a method of estimating reef 
fish populations. J. Wildl. Manage. 18:297-308. 
Burnham, K. P., D. R. Anderson, and J. L. Laake. 

1980. Estimation of density from line transect sampling of 



52 



LARSON and DeMARTINI: DISTRIBUTION OF FISHES IN KELP FOREST 



biological populations. Wildl. Monogr. 72, 202 p. 

Caughley, G. 

1977. Analysis of vertebrate populations. John Wiley & 
Sons, N.Y., 234 p. 
Clarke, T. A. 

1970. Territorial behavior and population dynamics of a 
pomacentrid fish, the garibaldi, Hypsypops rubicun- 
da. Ecol. Monogr. 40:189-212. 

Coyer, J. A. 

1979. The invertebrate assemblage associated with Mac- 
rocystis pyrifera and its utilization as a food resource by 
kelp forest fishes. Ph.D. Thesis, Univ. Southern Califor- 
nia, Los Angeles, 364 p. 
DeMartini, E. E., and M. E. Anderson. 

1979. Comparative survivorship and life history of painted 
greenling (Oxylebius pictus) in Puget Sound, Washington 
and Monterey Bay, California. Environ. Biol. Fish. 5:33- 
47. 
Dixon, W. J., and M. B. Brown (editors). 

1979. BMDP-79. Biomedical computer programs, P- 
series. Univ. Calif. Press, Berkeley, 880 p. 
Ebeling, A. W., and R. N. Bray. 

1976. Day versus night activity of reef fishes in a kelp forest 
off Santa Barbara, California. Fish Bull., U.S. 74:703- 
717. 
Ebeling, A. W., R. J. Larson, and W. S. Alevizon. 

1980a. Habitat groups and island-mainland distribution of 
kelp-bed fishes off Santa Barbara, California. In D. M. 
Power (editor), The California Islands: Proceedings of a 
Multidisciplinary Symposium, p. 403-431. Santa Bar- 
bara Museum of Natural History. 
Ebeling, A. W., R. J. Larson, W. S. Alevizon, and R. N. Bray. 
1980b. Annual variability of reef-fish assemblages in kelp 
forests off Santa Barbara, California. Fish. Bull., U.S. 
78:361-377. 
Elliott, J. M. 

197 1. Some methods for the statistical analysis of samples of 
benthic invertebrates. Freshw. Biol. Assoc, Sci. Publ. 25, 
144 p. 

Feder, H. M., C. H. Turner, and C. Limbaugh. 

1974. Observations on fishes associated with kelp beds in 
southern California. Calif. Dep. Fish Game, Fish Bull. 
160, 144 p. 

FREY, H. W. (editor). 

1971. California's living marine resources and their uti- 
lization. Calif. Dep. Fish Game, 148 p. 
Hobson, E. S. 

1971. Cleaning symbiosis among California inshore fishes. 
Fish. Bull., U.S. 69:491-523. 
Hobson, E. S., and J. R. Chess. 

1976. Trophic interactions among fishes and zooplankters 
near shore at Santa Catalina Island, California. Fish. 
Bull., U.S. 74:567-598. 
Jones, R. S., and J. A. Chase. 

1975. Community structure and distribution of fishes in an 
enclosed high island lagoon in Guam. Micronesica 
11:127-148. 



Keast, A., AND J. HARKER. 

1977. Strip counts as a means of determining densities and 

habitat utilization patterns in lake fishes. Environ. Biol. 

Fish. 1:181-188. 
Larson, R. J. 

1980. Competition, habitat selection, and the bathymetric 
segregation of two rockfish {Sebastes) species. Ecol. 
Monogr. 50:221-239. 

Miller, D. J., and J. J. Geibel. 

1973. Summary of blue rockfish and lingcod life histories; a 
reef ecology study; and giant kelp, Macrocystis pyrifera, 
experiments in Monterey Bay, California. Calif. Dep. 
Fish Game, Fish Bull. 158, 137 p. 

Parrish, R. H., C. S. Nelson, and A. Bakun. 

1981. Transport mechanisms and the reproductive success 
of fishes in the California current. Biol. Oceanogr. 1:175- 
203. 

Pienaar, L. V., and W. E. Ricker. 

1968. Estimating mean weight from length statistics. J. Fish. 
Res. Board Can. 25:2743-2747. 
Quast, J. C. 

1968a. Fish fauna of the rocky inshore zone. In W. J. North 
and C. L. Hubbs (editors), Utilization of kelp-bed re- 
sources in southern California, p. 35-55. Calif. Dep. Fish 
Game, Fish Bull. 139. 
1968b. Estimates of the populations and standing crop of 
fishes. In W. J. North and C. L. Hubbs (editors), Utiliza- 
tion of kelp-bed resources in southern California, p. 57- 
79. Calif. Dep. Fish Game, Fish Bull. 139. 
Randall, J. E. 

1963. An analysis of the fish populations of artificial and 
natural reefs in the Virgin Islands. Caribb. J. Sci. 3:31- 
47. 
Robins, C. R., R. M. Bailey, C. E. Bond, J. R. Brooker, E. A. 
Lachner, R. N. Lea, and W. B. Scott. 

1980. A list of common and scientific names of fishes from the 
United States and Canada. 4th ed. Am. Fish. Soc, Spec. 
Publ. 12, 174 p. 
Russell, B. C. 

1977. Population and standing crop estimates for rocky reef 
fishes of north-eastern New Zealand. N.Z. J. Mar. 
Freshw. Res. 11:23-36. 
Snedecor, G. W., and W. G. Cochran. 

1980. Statistical methods. 7th ed. Iowa State Univ. Press, 
Ames, 507 p. 

Stephens, J. S., Jr., and K. E. Zerba. 

1981. Factors affecting fish diversity on a temperate 
reef. Environ. Biol. Fish. 6:111-121. 

Talbot, F. H., and B. Goldman. 

1972. A preliminary report on the diversity and feeding 
relationships of reef fishes of One Tree Island, Great 
Barrier Reef system. Proceedings, Symposium on Corals 
and Coral Reefs, 1969, p. 425-444. 
Terry, C. B., and J. S. Stephens, Jr. 

1976. A study of the orientation of selected embiotocid fishes 
to depth and shifting seasonal vertical temperature 
gradients. Bull. South. Calif. Acad. Sci. 75:170-183. 



53 



THE INVERTEBRATE ASSEMBLAGE ASSOCIATED WITH THE 

GIANT KELP, MACROCYSTIS PYRIFERA, AT SANTA CATALINA 

ISLAND, CALIFORNIA: A GENERAL DESCRIPTION WITH 

EMPHASIS ON AMPHIPODS, COPEPODS, MYSIDS, AND SHRIMPS 1 

James A. Coyer 2 
ABSTRACT 

The motile invertebrate assemblage associated with the giant kelp, Macrocystis pyrifera, fronds was 
examined monthly from June 1975 through December 1976, at Santa Catalina Island, California. Replicate 
samples were collected from each of three vertical zones (canopy |C|, middle [M|, bottom [B]). 

The number of species collected from all zones was 114 and ranged from 51 to 75 for any given month. 
Amphipods, copepods, mysids, and shrimps comprised the majority of invertebrate abundance (86 [C], 92 
[M], 93 f ; |B|) and biomass (90 [C|, 89 [M|, 86 f i [B]). Gammarid amphipods dominated the assemblage in 
numbers (34 [C|, 60 [M], 519? |B]), biomass (34 [C|, 68 [M], 67% [B]), and number of species (20). 

The assemblage displayed three patterns of vertical stratification within the Macrocystis forest: 1) The 
mean number of species progressively decreased from the bottom to the canopy (several species displayed 
zone preferences); 2) more individuals and a greater total biomass were present in the lower zones than in the 
canopy; and 3) the mean lengths of gammarids, mysids, and shrimps were significantly larger and propor- 
tionately greater numbers of large individuals were present in the canopy than in either of the lower 
zones. 



Subtidal forests of giant kelp have long attracted the 
interest of biologists, beginning with Darwin's (1860: 
240) description of the organisms associated with the 
giant kelp forests off Tierra del Fuego. Since the 
advent of scuba techniques in the mid-1950's, 
several studies have examined in detail the attached 
and/or motile species of invertebrates associated 
with surfaces of the giant kelp, Macrocystis pyrifera 
(Limbaugh 1955; Clarke 1971; Ghelardi 1971; Jones 
1971; Wing and Clendenning 1971; Miller and 
Geibel 1973; Lowry et al. 1974; Bernstein and Jung 
1979; Yoshioka 1982 a, b). Few, however, have 
attempted a long-term and comprehensive examina- 
tion of the entire assemblage of small and motile 
invertebrates found with the giant kelp. The as- 
semblage is important for several reasons, notably as 
the major source of food for most fishes residing 
within the kelp forests (see fish diet studies by Quast 
1968; Hobson 197 1; Bray and Ebeling 1975; Hobson 
and Chess 1976). 

The present report examines the composition, pat- 
terns of vertical stratification, and seasonal dynamics 
of the small and motile invertebrate assemblage 



'Contribution No. 37, from the Catalina Marine Science Center. 

: Catalina Marine Science Center (University of Southern Califor- 
nia), Avalon, Calif.; present address: Division of Science and 
Mathematics, Marymount Palos Yerdes College, Rancho Palos Ver- 
des, CA 90274. 



associated with the fronds of M. pyrifera. A general 
overview of the assemblage and a detailed examina- 
tion of the amphipods, copepods, mysids, and 
shrimps are presented. 

STUDY AREA 

The study area was Habitat Reef, located in Big 
Fisherman Cove, Santa Catalina Island, Calif, (lat. 
33°28'N, long. 118°29'W). Habitat Reef is a 
fingerlike extension of bedrock ranging in depth from 
2 to 18 m and is bounded on the three outer margins 
by an expansive area of shelly debris substrate. The 
western and northern sides of the reef slope sharply 
to a depth of 20-25 m, whereas the eastern edge 
slopes gradually to a shallower area ranging from 8 to 
19 m. Water temperatures at Habitat Reef ranged 
from 13.6° to 21.2°C during the study, warmest dur- 
ing July through September and coolest from 
December to February. 

The algal community of the shoreward portion (<3 
m depth) of Habitat Reef was dominated by Phyllo- 
spadix torreyi, Eisenia arborea, Cystoseira neglecta, 
and Sargassum muticum (seasonally) . The outermost 
portion (>3 m depth) was dominated by Macrocystis 
and the understory algae in this area was sparse, 
although small patches of Dictyopteris zonarioides 
and C. neglecta were present in some areas. 



Manuscript accepted August 1983. 
FISHERY BULLETIN: VOL. 82, NO. 1, 1984. 



55-^ 



FISHKRY BULLETIN: VOL. 82, NO. 1 



MATERIALS AND METHODS 
Zonation and Kelp Density 

The kelp forest at Habitat Reef was divided into 
three vertical zones: Canopy (C), middle (M), and 
bottom (B). The canopy extended from the water sur- 
face to a depth of 1 m, the bottom ranged from just 
above the kelp holdfasts to 2 m above the substrate, 
and the middle included the area between the canopy 
and the bottom. Holdfasts were not examined. Kelp 
density was measured by randomly establishing 25 
circular 1 m : plots within the study area during 
November 1975 and October and December 1976. 
The number of enclosed plants and the number of 
fronds/plant were determined. 

Sampling Procedure 

Samples were collected monthly from plants in the 
central portion of the kelp forest (7-9 m depth) during 
tidal heights ranging from +1.0 to +1.3 m mean 
lower low water. From June through September 

1975, three replicate samples were collected from 
each zone; from October 1975 through December 

1976, five replicates were collected. Only one sample 
was collected from any plant, and this sample con- 
sisted of the entire plant portion within the desired 
zone. The middle and bottom zones were collected 
by carefully severing the upper portions and allow- 
ing them to drift away. Disturbance to the lower 
zones during this procedure was negligible. Similar 
amounts of kelp were collected from each zone 
throughout the study (n = 19; kg = 2.5[C] 2.1[M], 
2.3 [B]). 

The kelp-associated invertebrates were collected 
by scuba divers maneuvering a plankton net (1 m 
diameter, 3 m long, 0.33 mm mesh) over the desired 
portion of the plant. This procedure captured most 
motile invertebrates on the kelp, as well as within the 
surrounding water column (1 m diameter). The 
enclosed sample was placed in a large container filled 
with warm freshwater (providing a thermal and 
salinity shock), vigorously agitated, and removed. 
The remaining water was filtered through a 0.25 mm 
sieve and the residue preserved. Thus, the term "in- 
vertebrate" in this investigation refers to all motile 
individuals larger than 0.33 mm (excluding pro- 
tozoans, cnidarians, and nematodes). 

The efficiency of the agitation-freshwater method 
was tested by placing the processed kelp into another 
container of warm freshwater and allowing it to stand 
for 4 h. Subsequent agitation and filtering indicated 
that 96% of all motile invertebrates in each zone were 



removed by the initial agitation-freshwater treat- 
ment. 

Organisms were identified to species (except for 
some juveniles). The wet weight of kelp from each 
sample was measured, and abundances of all taxa 
were expressed as the number of individuals per 
kilogram (wet weight) of kelp. The somewhat uncon- 
ventional normalization of species abundance to unit 
biomass was selected for three reasons. First, struc- 
tural complexity within the kelp forest habitat is 
created by interdigitating kelp blades and stipes and 
is a function (in part) of both kelp surface area and 
biomass. Many kelp-associated species, particularly 
the swarming mysids, may respond primarily to 
structural complexity of the habitat when seeking 
shelter and/or food. Secondly, biomass is much 
easier and faster to measure than is surface area (con- 
version ratios of kelp wet weight to surface area [both 
sides of blades + stipes] and kelp dry weight to wet 
weight are presented in Table 1). Thirdly, unit 
biomass will facilitate comparisons with invertebrate 
associations of other species of marine algae for 
which it is difficult to compute a unit area (i.e., bushy 
reds and browns). 



TABLE 1. — Ratios of kelp wet weight (kg) to kelp surface 
area (rrr) and dry weight (kg) to wet weight (kg). 





Wet 


weight/area 




Dry we 


ight/Wet weight 


Zone 


< 


SD 


n 


SD n 


Canopy 
Middle 
Bottom 


021 
19 
42 


002 
002 
040 


10 
10 
10 


16 
15 
13 


0010 6 
0025 6 
0.042 6 



Determination of Invertebrate 
Lengths and Biomass 

Growth series within the principal taxa were 
established. Individuals (n = 30-94) were measured 
to the nearest 0.04 mm, using a dissecting micro- 
scope and occular micrometer, blotted dry, and 
weighed using an analytical balance to determine 
length-weight relationships. Smaller and/or minor 
taxa (copepods, ostracods, caprellids, molluscs, etc.) 
were assigned constant weights based on the mean 
weight of 20 individuals. 

Vertical patterns of size-stratification were ex- 
amined by measuring the lengths of principal taxa 
within each zone for each quarter from January 
1975 through October 1976. Single samples were 
collected in January and April 1975; subsequent 
samples were replicated (3 or 5). For shrimps and 
mysids, all (January through July 1975) or up to 75 
individuals of each major species were measured 



56 



COVER: INVERTEBRATE ASSEMBLAGE WITH GIANT KELP 



from each replicate of each zone; for gammarid 
amphipods, at least 50 individuals (comprising all 
species) were measured from one randomly selected 
replicate of each zone. Replicates were pooled and 
size-frequency distributions were determined for 
each taxon within each of the zones. The non- 
parametric Kolmogorov-Smirnov (K-S) two-sample 
test (one-tailed) was used to test whether the values 
from one distribution were stochastically larger 
than the values from another distribution (Siegel 
1956). 

The mean weight of an individual within a major 
taxon (shrimps, mysids, gammarids) was determined 
from the mean length and the appropriate length- 
weight formula. The mean weight then was mul- 
tiplied by the mean monthly abundance of the taxon 
to determine the taxon biomass. Quarterly length 
measurements were applied to the month preceding 
and following the measuring month (i.e., April 
measurements were assigned to March and May) for 
biomass measurements. Monthly abundance values 
of the smaller taxa were multiplied by the assigned 
weight to estimate the biomass. 

RESULTS 
Kelp Density 

Macrocystis density at Habitat Reef was high (4.7 
plants/m 2 ) from November 1975 through August 
1976 (Table 2). In late September 1976, density and 
canopy cover were reduced ( 1 .5/nr) and continued to 
decline over the next 4 mo. 



TABLE 2. — Macrocystis density and the number of 
fronds/plant (± width of 959? C.I./2) at Habitat 
Reef. Sample size in parentheses. 



Date 



Density m- 



No. of fronds/plant 



General Taxonomic Composition of 
the Invertebrate Assemblage 

The invertebrate assemblage associated with the 
fronds of Macrocystis was composed primarily of 
amphipods, copepods, mysids, and shrimps (Tables 
3, 4). Mysids and shrimps were among the largest 



TABLE 4.— Mean abundance (no./kg kelp ± width of 95', C.I./2) for 
the major invertebrate taxa within each zone. Parenthetical values 
are the mean length and weight (mm, mg) of each taxon; an asterisk 
indicates that a constant length and weight was used for all zones. All 
values are averaged over the entire 19-mo study. 



Nov. 1975 
Oct 1976 
Dec. 1976 



4.7+2.3 (25) 
1.5±0.7 (29) 
0.7±0.3 (25) 



3.4±1.0(1 18) 
6.7+1.6 (46) 
4.7+2.6 (17) 



Taxon 


Canopy 


Middle 


Bottom 


Gammarid amphipods 


882.4 ±267.0 


4,1 23.0 ±890.2 


3,1 1 7 8 ± 715.3 




(2.8.0.6) 


(1 8,0.4) 


(20.04) 


Copepods 


1,1 28.0 ±370.2 


1 .977.0 ± 540.8 
'(0.8,0.1) 


2,453.1 ±441.0 


Ostracods 


188.2 ± 76.7 


108.8 ± 51.1 

'(0.9.0.1) 


65 7 ± 300 


Echinoids (juv.) 


13 9±25.5 


260.5 ±375 4 
•(0.5,0.1) 


83 0± 119.2 


Mysids 


91 .4 ± 57.8 


151 8±389 


108 ±50. 5 




(6 2.3.5) 


14.7.1.3) 


(4.4.1 2) 


Molluscs 1 shelled ) 


14.8 ±9.0 


98.0 ±21.8 
•(1.3.0.7) 


1684 ±44 9 


Candean shrimps 


136.5 ±48 4 


65 2 ±28.4 


51 4 ± 16 




(7.1,3.8) 


(6.0,2.7) 


(5 4.2 3) 


Platyhelminthes 


31 7 ± 17 1 


36.8 ± 16 2 

'(-.38) 


34.0 ± 17 


Cladocerans 


72.2 ±93.4 


9.2 ±5 .1 
'(0.7,0.1) 


9.3 ±6.9 


Polychaetes 


88± 11 3 


28.0 ±8.0 
"(3 3.0.5) 


17 4 ± 7.2 


Cypris (barnacle) 


13.4 ± 16.6 


24.0 ±22.1 


14. 3± 9 4 


larvae 




'(0.7,0.1) 




Molluscs (nudibranchs) 


10 9 ± 11.3 


13. 4± 11 8 
•|1 3.1.1) 


8 1+75 


Sphaeromatid isopods 


0.1 ±0.1 


0.2 ±0 1 
•(2.4.1.1) 


19 7 ±23.0 


Caprellid amphipods 


4 1 ±20 


2.7 ± 1.0 
'169.0.8) 


1.8 ±1.3 


Idoteid isopods 


3.1 ± 3.2 


0.1 ±0.1 

•(7.2,4,0) 


1 ±<0.1 


Asteroids (juv.) 


0.2 ±0.1 


1.1 ±1.3 
•(2.7.2.0) 


6±08 


Jaeropsid isopods 


0.1 ±0.1 


0.2 ±0.3 
"(2.3.0.3) 


1 4 + 09 


Cumaceans 


— 


02±0 2 


3 ±0.3 


Brachyurans (zoea) 


— 


— 


0.1 ± <0.1 


Ophiuroids (juv.) 


— 


<0.1 ± <0 1 


<0 1 ± <0 1 


Tanaids 


— 


<0 1 ±0.1 


0.1 ±0.1 



TABLE 3. — The mean (± width of 95' i C.I./2) monthly abundance (no. organisms/kg kelp) and 
biomass (mg organisms/kg kelp) for each major invertebrate group associated with the giant 
kelp. Data are averaged over the entire 19-mo study; proportions of total numbers or biomass (all 
species) are presented in parentheses. 



Zone 


Gammarids 


Copepods 


Mysi 


ds 


Shrim 


3S 


Total 


Canopy 


















Numbers 


882 ±267 (33 9) 


1.128±370 


(43.4) 


91 ±58 


(3.5) 


136 ±48 


(5.2) 


2.599 ±580 


Biomass 


589 ±236 (33.8) 


56 ± 18 


(3.2) 


336±255 (19.3) 


583 ± 300 


(33 4) 


1,743 ±765 


Middle 


















Numbers 


4.123 ±890 (59.8) 


1.977 ±541 


(28 7) 


152 ±39 


(2.2) 


65 ± 28 


(0.9) 


6,900 ± 1,382 


Biomass 


1,634 ±359 (68.4) 


99 ±27 


(4.1) 


218±68 


(9.1) 


1 74 ± 71 


(7.3) 


2,387 ± 493 


Bottom 


















Numbers 


3.1 18 ± 71 5 (50.8) 


2,453 ±441 


(39.9) 


1 08 ± 5 1 


(1 7) 


51 ± 16 


(0.8) 


6.153 ±937 


Biomass 


1.388 ±337 (67.4) 


1 23 + 22 


i6.0) 


143 ±83 


(6.9) 


116± 37 


(5.6) 


2.061 ±454 



57 



FISHERY BULLETIN: VOL. 82, NO. 1 



species present, copepods among the smallest 
(Table 4). 

The number of species collected from all zones 
totaled 114, but ranged from 51 to 75 for any given 
month (Fig 1). When ranked by the mean monthly 
abundance, 7 species were dominant (>100/kg), 23 
were common (10-100/kg), 24 were uncommon (1- 
10/kg), and 60 were rare (<l/kg). Crustaceans and 
gastropods had the greatest species representation 
with 63 and 36 species present, respectively. The 10 
most abundant species within the canopy and bottom 
and 9 of the top 10 species in the middle were crus- 
taceans; of the 14 crustacean species represented, 6 
were gammarid amphipods and 4 were harpacticoid 
copepods (Table 5). 

Vertical Patterns of Distribution, 
Abundance, and Sizes 



70 



60 



en 50 



- / A 



o 

UJ 
Q. 
CD 



40 



O 

oo 

2 

z 20 



10 




V / 




A-./N'" 




\ 



/ TV 






b— o 



• CANOPY 
A MIDDLE 
o BOTTOM 
a COMBINED 



_l I I I I I I I I I I I I I I I I L_l 

JJAS0NDJFMAMJJAS0ND 
1975 1976 



Invertebrate numbers and biomass (no./kg kelp, 
mg/kg kelp) were greatest in the middle (6,900, 
2,387) and bottom (6,143, 2,061) zones, lowest in 
the canopy (2,599, 1,743; Table 3). Similarly, the 
number of species was always lowest in the canopy, 
intermediate in the middle, and highest in the bottom 
(Student's £-test,P< 0.05; Fig. 1). 

Gammarid amphipods were the most important 
taxon associated with Macrocystis, dominating the 
invertebrate assemblage within each zone in terms of 
numbers (34-60%) and biomass (34-68%; Table 3). 
Twenty species were collected with fewer species 
present in the canopy (11) than in the middle (16) or 
bottom (18). 

Collectively, Microjassa litotes, Gitanopsis vilordes, 
and Aoroides columbiae comprised 837c by number 
of all gammarids in the canopy, 92% of those in the 
middle, and 70% of the gammarids in the bottom. 
The most abundant gammarid in the canopy was G. 
vilordes (53.0%); M. litotes was most abundant in the 
middle (49.0%) and bottom (33.8%; Table 6, Fig. 2). 
Among the other gammarids present, Ampithoe plea 
and Hyale frequens were much more common in the 



FIGURE 1. — The number of invertebrate species present in each of 
the vertical zones. Many species are present in more than one zone. 
Grand means (± width of 95%C.I./2): 26.7 ± 1.6(C), 33.0 ± 3.3 (M), 
38.9 + 3.2 (B). 



canopy than in the lower zones, whereas Batea 
transversa and Pontogeneia rostrata were abundant 
in the bottom zone and uncommon in the canopy 
(Table 6). 

Numerically, copepods formed a major portion of 
the invertebrate assemblage (29-43%), but con- 
tributed very little to the total biomass (3-6%; Tables 
3, 4). Although numerous in all zones, copepods were 
more abundant in the middle and bottom (Table 3, 
Fig. 2). Most (88% by number) in the middle zone 
consisted o{ Porcellidium viridae, Porcellidium sp. A, 
and Tisbe sp. In the canopy and bottom zones, P. 
viridae, Tisbe sp., and Scutellidium lamellipes 
accounted for 89 and 92%, respectively. 

Mysids and shrimps were minor numerical com- 
ponents of the assemblage (2-3 and 1-5%, respective- 
ly), but formed major proportions of the invertebrate 
biomass (7-19, 6-33%; Tables 3, 4). Each of the three 



Table 5.— The ten most abundant invertebrate species in each zone. Abundances are mean monthly values for the 19-mo study 
(C = copepod, G = gammarid amphipod, M = mysid, O = ostracod, S = shrimp, E = echinoid |urchin|). 



Canopy 


No./kg 

kelp 


Middle 


No./kg 
kelp 


Bottom 


No./kg 
kelp 


Porcellidium viridae |C) 


557 


Microjassa litotes (G) 


2.018 


Porcellidium viridae (C) 


1.768 


Gitanopsis vilordes (G) 
Tisbe spp (C) 


466 
303 


Gitanopsis vilordes (G) 
Porcellidium viridae (C) 


1.551 
1,106 


Micro/assa litotes (G) 
Gitanopsis vilordes (G) 


1.054 
778 


Macrocypnna pacifica (O) 
Aoroides columbiae (G) 


165 
1 64 


Aoroides columbiae (G) 
Tisbe spp. (C) 


600 
358 


Pontogeneia rostrata (G) 
Tisbe spp (C) 


397 

374 


Scutellidium lamellipes (C) 


144 


Porcellidium sp. A (C) 


270 


Aoroides columbiae (G) 


350 


Hippolyte clarki (S| 


137 


Strongyfocentrotus sp. (E) 


260 


Batea transversa (G) 


342 


Microjassa litotes (G) 


103 


Sinella pacifica (M) 


148 


Scutellidium lamellipes (C) 


163 


Ampithoe plea (G) 
Acanrhomysis sculpta (M) 


91 
74 


Scutellidium lamellipes (C| 
Macrocypnna pacifica (O) 


137 
88 


Porcellidium sp. A (C) 
Sinella pacifica (M) 


1 13 
99 



58 



COYER: INVERTEBRATE ASSEMBLAGE WITH GIANT KELP 

TABLE 6. — The ten most abundant gammarid amphipods in each zone. Abundances are mean monthly values for the 19-mo 

study. 





No/kg 




No./kg 




No./kg 


Canopy 


kelp 


Middle 


kelp 


Bottom 


kelp 


Gitanopsis vilordes 


466 


Microjassa litotes 


2.018 


Microjassa litotes 


1.054 


Aoroides columbiae 


164 


Gitanopsis vilordes 


1.183 


Gitanopsis vilordes 


778 


Microjassa litotes 


103 


Aoroides columbiae 


600 


Pontogeneia rostrata 


435 


Ampithoe plea 


91 


Batea transversa 


82 


Aoroides columbiae 


350 


Hyale frequens 


24 


Pontogeneia rostrata 


40 


Batea transversa 


342 


Batea transversa 


2 


Ampithoe plea 


37 


Ampithoe plea 


147 


Pontogeneia rostrata 


1 


Pleustes platvpa 


5 


Pleustes platvpa 


12 


Pleustes platvpa 


1 


Erichthomus braziliensis 


1 


Erichthomus braziliensis 


8 


Erichthomus braziliensis 


1 


Pleusirus secorrus 


1 


Amphilochus sp. 


2 



5 
4 
3 
2 



Copepods 



41 



% 



m 



i: 



It 



i 



I 



if 



M 



k 



& 



i 



if 



i 



k 



it 



i 



£• 



2 " A columbiae 



en 



fpu- M \ 



rti 



tL 



J il l Ji r^ 



rfF n-i 



-*J*1 fr+1 



V\ &\ M 



hW 



* 



^ 



ft 



o 
o 
o 



cc 

LU 



G vilordes 



° 3 



A 



jtt 



i 



Hi 



i* 



iMh 



fa 



ibh 



jtk 



jy 



U 



I 



j£ 



tuty 



ra 



ik 



k 



*i 



* A *i — 1 — L_A 



_ M litotes 



_feL 



A 



* 



ffi 



hd 



km 



1 



A 



m 



j j 

1975 



N D 



M A 



□ Canopy 

□ Middle 

□ Bottom 



jit 



ft 



J 



I 



m 



■h:: 



h 



th. 



£ 



h 



M 



J J 
1976 



A 



N D 



FIGURE 2.— Monthly abundances (mean and 95% C.I.) of copepods (all species) and the gammarid amphipods (Aoroides columbiae, Gitanopsis 
vilordes, Microjassa litotes) in each vertical zone. Each value represents the mean of three lJuly-September 1975) or five (October 1975- 
December 1976) replicate samples. 



59 



FISHERY BULLETIN: VOL. 82, NO. 1 



mysid species present differed in their patterns of 
vertical distribution. Acanthomysis sculpta essen- 
tially was confined to the canopy (no. /kg ± width of 
95% C.I./2 = 73.6 ± 56.4 [C], 4.0 ± 2.2 [M], 7.7 ± 5.2 
[B|), and accounted for 80.57c (by number) of all 
mysids in the zone, whereas S.pacifica was less abun- 
dant in the canopy (17.8 ± 6.3 [C], 147.6 ± 39.1 [M], 
99.2 ± 46.9 [B]), but was dominant in the middle 
(97.2%) and bottom (92.2%). An unidentified 
erythropinid rarely was encountered and, when pre- 
sent, found only in the lower zones (0 [C], 0.2 ± 0.2 
[M|, 0.7 ± 0.5 [B]). Nearly all (99.9%) of the shrimps 
associated with the kelp fronds were Hippolyte clarki, 
and this species was most abundant in the canopy 
(Table 3, Fig. 3). 

Throughout most of the study, gammarid sizes were 
largest in the canopy, intermediate in the bottom, 
and smallest in the middle (K-S test; C-M, C-B, MB: 
P < 0.01; Table 4, Fig. 4). Mysids and shrimps also 
were largest in the canopy, but were smallest in the 
bottom (K-S test; C-M, C-B, M-B:P < 0.001; Table 
4, Fig. 4). Among the mysids, S. pacifica was more 
slender (mm, mg= 6.5,2.9 [C], 4.7, 1.2 [M], 4.5, 1.2 
|B|) thanA sculpta (6.2, 3.7 |C], 4.2, 1.7 [M], 3.1, 1.1 
[B]). Combined size distributions of the four major 
taxa for the 19-mo study (weighted according to 
mean monthly abundance) revealed proportionately 
greater numbers of large individuals present in the 
canopy than in either the middle or the bottom (K-S 
test; C-M, C-B: P < 0.001; M-B: ns; Fig. 5) 

Seasonal Patterns of Species, 
Abundances, and Sizes 

No seasonal patterns were apparent for total num- 
ber of invertebrates in the canopy; however, total 
biomass increased dramatically (from 1,696 to 6,315 
g/kgkelp) during winter 1975-76 (Fig. 6). In the lower 
zones, both numbers and biomass were highest dur- 
ing winter 1975-76 and the following spring (Fig. 6). 

Seasonal patterns of abundance for the major 
species were evident only for the shrimp H. clarki, 
which displayed maximum abundance during both 
winters of the study (Fig. 3). The canopy mysid, A. 
sculpta, was abundant (1 13. 2-395. 5/kg kelp) during 
winter 1975 and early spring 1976, but was uncom- 
mon (6. 5/kg kelp) during the following winter (Fig. 3). 
Single monthly samples collected in winter 1974-75 
also indicated high numbers (79. 2-169. 8/kg kelp) of 
the canopy mysid. Seasonal patterns were not evi- 
dent for the three most common gammarids (Fig. 2). 
As a group, copepods were most abundant during 
winter and early spring in the lower zones, but no 
seasonal pattern was apparent (Fig. 2). 



Seasonal variations in the sizes of gammarids, 
mysids, and shrimps were frequently observed (Fig. 
4). Gammarid sizes were largest during winter and 
spring in the canopy (2.76-3.85 mm), but no seasonal 
patterns were present in the lower zones. Carapace 
lengths of S.pacifica were largest during winter in the 
canopy (1.68-1.89 mm) and middle (1.24-1.45 mm) 
with 1976 measurements greater than 1975. 
Smallest sizes were present during summer in both 
the canopy (1.28-1.39 mm) and middle (0.92-1.22 
mm). No seasonal patterns were present in the bot- 
tom. The shrimp H. clarki was largest in the canopy 
during winter-spring of both years (1.55-2.08 mm) 
and during spring 1 975 and winter 1976 in the middle 
(1.56-1.64 mm) and bottom (1.36-1.43 mm). 
Smallest shrimps were present during fall in all three 
zones (1.30-1.42 mm [C]; 1.01-1.17 mm [MJ; 1.02- 
1.03 mm [B]). No pattern was observed for A. 
sculpta. 

DISCUSSION 
Kelp Density 

Elevated temperatures and/or low nutrients may have 
caused the Habitat Reef kelp forest to decline in late 
1976. Kelp forests in southern California deteriorate 
when the water temperature exceeds 20°C for substan- 
tial periods (North 1971), and high temperatures often 
are associated with low nutrients (Jackson 1977). Signif- 
icantly, temperatures at Habitat Reef did not reach 
20°C in 1975, but exceeded 20~C from mid-June to 
November 1976. During the second half of 1976, other 
areas of southern California also experienced warm 
water and corresponding declines in Macrocystis stand- 
ing crop (Southern California Edison Co. 1978 1 ). 

Preference of Macrocystis 
as a Habitat 

Few of the 114 species associated with Macrocystis 
fronds at Habitat Reef were restricted to the frond 
habitat. Most were present, and many were more 
abundant in Macrocystis holdfasts, understory algae, 
or other habitats within or adjacent to Habitat Reef 
(Hobson and Chess 1976; Hammer and Zimmerman 
1979). A few, however, such as the gammarid M. 
litotes, the shrimp H. clarki, and both species of 
mysids, were more abundant in the Macrocystis 
fronds than in other habitats. 



'Southern California Edison Company. 1978. Annual operating 
report, San Onofre Nuclear Generating Station, Vol. IV. Biological, 
sedimentological, and oceanographic data analyses. Southern 
California Edison Co., Rosemead, CA 91770, 300 p. 



60 



COYER: INVERTEBRATE ASSEMBLAGE WITH GIANT KELP 



£o> E 

o -o 2 

c -o t: 

o - o 

o ^ m 

□ E3 D 



€ 



1 



^ 



I W^: . :.x . ',; , :',: .:j 



-S 



-^^£^2S 



- F??F 



-E 



D 

u 

I/) 



-i 



J I L 




JE 



"1 




o 

o 

Q. 

COI 



-ExTi 



J I L 



i 



-E 



f^rrr 



^E 



& 



lvoS 






Q 



O 




t" 



H ~ , ' ,, ,' , ',',' 



4— 



J 1 L 




O 



O 
O 
m 



O 
O 



o 

8 



di3M 6>i/a3aiAinN 



a 




H 




01 




C 




o 




N 








(0 
















~ 




OJ 




> 




J3 








oa 




<u 




C 


CO 

cu 






__ 


a 


* 


E 


CI 


B 










4) 


O 


;> 


cs 






E 

a. 


?. 


W, 


cu 


&3 


L* 


" 


CC 






r 


Oi 










8 


t- 


CO 


CD 




— 


J3 


r 


— 




~ 


u 


c 


CD 


0! 


Q 






C 


lO 


~ 


r~- 


*-. 


C35 


;- 


I— < 


5 




ft 


CU 


CI 


-O 




o 






. ~ 


o 




_ 


-r 




• 


CD 


a 


> 






w 


— 


a 


O 


y. 






lO 


r 


cn 


p 


i — i 






c 


CD 


"C: 


— 


c 


~ 


^ 








-. 


— 


^~ " 


CD 


— 


X 




>, 


>> 


- 


- 


>-3 










43 
O 


CD 

J3 








— 


 : 


o 


<_> 


c 




01 


b*> 


CD 


05 


£ 


— 


CD 


n 


J3 


3) 


■•^ 


c 


CO 


CO 


C 


o> 


CD 


fci 


CO 
CD 






111 

u 


— 
CD 


en 


CD 


— 


- 


= 


- 


C3 


> 


— 




eg 




>> 








c 








c 




o 




2 
1 




<?6 




w 




« 




_ 




O 





61 



FISHERY BULLETIN: VOL. 82, NO. 1 



I 



91 



^mwW: 



^E 



::62 : 



CI' 



^Y', V ** 



*-:•:•:•:•>:•: -coi 



lE 



M 



J.ZZ 



tmmmi™ 



-t 



mm 



:U 



Zi 



f , , " , ' M< 



$mz$ 



M 



-^ M:;:fr£l 



!■ " ■€ 



crm 



J I 



zv 



jzz 



OIC 



£ 



'SIC 



t ,. .,.,.. ^ 



S.i£ 



r.-x-x <n.i 



mVftV^ 



9fr£ 



•szc 



* ™ 



I 



in 



9S2 



 S/£ 



kkk\w.vw 



m 



6I2 



5 



i^£ 



k\\\\\\\\V£0£ 



■fry;';,'. : >/s 



o 
o 



o 



+^x'^> 



::*t»i 



■/Z2 



- ^ ,,., . .,, ^ , 



J 



;:G12 



-fcv:-.-:-::i/ 

. . '.. . *• 



j i 



J 



[62 



-p ::■■■: 9zz 



fv '^ 



I 



•l££ 



m\\y^ 



6Z£ 



 ;:•■.•'.•:•:• 9cc 



- Ry^^Bze 



* 



■teg;"':""' 

i ' i i 1 1 in 



06l 



<ilt 



m 



00£ 



occ 



■s^^Si 






^NNW ^ 



X 



_L 



+/'.•:•:•'.•*. 96 z 

 



C\J — CO — CM 

(WW) H19N31 33VdVcd\/0 



cu 



Q. — 

O T3 

C -o 

3 s 



E 
o 

o 

QD 



\ <*-'■'■•;;■ , 



IS 



OS 



- ^ . v.v.v:::, 1 ::: . ^ ^ 



-£I 



ED E3 



§f 



CS: 



-g^^^^ns 



■m 



3: 



t>s 



ss. 



 KKVt^^tM^ s^ 




^Satt^^^^*^*^ 



to 

o ^ 

Q. in 

jE .^ 

Q. O 

■D -D 

o 5 

U 

o 



i 



^:'x*:-: 



S¥S¥9S|: 



^^Ms 



-EZ 



s5 



£S 



^^13 



 n 1 : 1 : 1 : 1 .". 



::9s; 



^\\\\\\\^\\\\\\^^ 



-E 



::£S:: 



4^k\\\\\\\\\\\\\\\^^ ^2 



o 
O 



Z3 



Q. 
< 



o 

O 



Q. 
< 



O I s - 



m 



rO 



CM — 



(wuj) H19N31 AQOa 





 M 


n 


CO 




t3 
O 


> 

-C 
- 

1> 


3 
— 

E 
- 


CO 

CC 

u 

"3 

c 






C 


B, 


; 


























< 


a^ 


cc 








/ 






3 
—> 




a> 






-e 


~ 


a 
i> 






«2 


31 


t- 






tj 


~~ 


— 






H^» 


~. 


u 




CL 


-~ 


B 


j 
u 




< 


5 


JZ 

■J: 

— 


CO 






a: 


£ 




CO 




a 

CO 

CO 

— 




£ I s - 




■a 






J5 


> 


~ 




*— 

o 
O 


09 

c 


£ 
o 

■J. 


c 

c 

£ 

c 
— 






l~ 


~ 


i> 






- w 


> 


<a 






e 


4J 


CO 




3 




c 

cu 


3 




-3 


■2 


= 


a 






 


- 


.- 






C 


D 

CO 


T3 






y 


_c 


— 




^_. 




a 


cc 




Q. 
< 


§. 


i) 


£ 

CC 

M 






v. 


a 




m 


ts 


C 




£ I s - 


7 


CO 


Cil 




-^ CD 


i 


0) 

5c 


11 






2 


C^ 


CO 






-C 


X 


— 














^ 


CC 


1 
















_ 




cfl 






- 


- 


u 






cc 


— 


a 






— 


11 








1 

£ 


a 

CO 

a 


cc 

CO 

u 








£ 


S 






1 


- 


















J2 


n. 








X 








a. 


-r 


u 






t 


•_ 






1 — ' 


t~ 






id 


C/) 


<JJ 


c 

£ 




M 


■a 






~ 


£ 


01 


3 




0. 


- 


c^ 




,c 


r/T 


- 




a 


~ 


n 


B 




- 






4-1 




B 


- 




n 




~ 


£ 


m 


-£ 




— 


- 




— 






a! 


u 


-: 




S 


-r 








— 


_ 
- 


u 




a 


e 








— 


T. 


J 

o 


- 
B 

CO 




~ 


11 


"" 




™ 






-c 




CJ 


— 


'— 






..^ 


a 


- 


3 




1C 

- 
— 


< 

— 
c 

cc 




CO 

a 
£ 




to 
c 

CO 


>> 

CO 

3 
C 


3 


- 
— 




X 


1) 


> 






£ 


u 


— 




01 


d 


B 




^ 


o 














it-* 






~- 


11 


O 




c 


co 


c 






CO 


H 








0) 


= 




.- 




ts 


X 


£ 
n 


£ 




| 


> 


3 










B 




■^ , 


1/ 
C 


CO 


"x 






o 


1 




{£ 


N 


a 
£ 


S 

4J 




CJ 


a 


11 




Pu 




CO 


♦J 



62 



COYER: INVERTEBRATE ASSEMBLAGE WITH GIANT KELP 




LENGTH (mm) 



Figure 5. — Combined size-frequency distributions of copepods, 
gammarid amphipods, mysids, and shrimps measured quarterly 
from July 1975 through October 1976 for each of the three vertical 
zones. Copepods were measured during 1 mo only because of their 
small size and variability. After normalization (%), the distributions 
of each taxon were weighted according to mean monthly abundance 
to create the combined distributions. The numbers of each taxon 
measured before weighting are (C, M, B): copepods (54, 54, 55), 
gammarids (308, 323, 317), mysids (2,037, 2,625, 2,500), and 
shrimps 1 1,896. 1,776, 1,561). Statistics determined after weighting 
are displayed in the figure. 



Mysids are remarkably specific in habitat prefer- 
ences. Clarke (1971) found 12-14 species of mysids 
cooccurring in the kelp forests off San Diego and Baja 
California, but only A. sculpta and S.pacifica were 



10 






CANOPY 



/ 






•o-a 



V HjfG 



J — I — I — I — I — I — I — 1 — I I I 1 I I I I  I I 



C7< 



o 
o 
o 




10 



BOTTOM _ 



~&J 



/V<' \ 






W-s-- 



\ 



x><\ 



va 



i i i i i ' ' i i ' i ' I 1 I I I I L 



7 
6 
5 
4 
3 
2 
1 



JJASONDJF MAMJ J AS0ND 
1975 1976 

FIGURE 6. — Monthly variation in numbers and biomass of all inver- 
tebrate taxa (combined) within each vertical zone. Each monthly 
value for the canopy, middle, and bottom represents a mean of three 
(June-September 1975) or five (October 1975-December 1976) 
replicate samples. 



associated with the kelp fronds. Similar patterns 
were observed at Habitat Reef, as both A. sculpta 
and S.pacifica were present in large numbers within 
the kelp fronds, but were rarely observed in Mac- 
rocystis holdfasts or other algal habitats within or 
near Habitat Reef (Hammer and Zimmerman 1979). 
Hobson and Chess (1976) found a few individuals of 
A. sculpta in the water column at night, but most 
remained closely associated with the kelp which was 
utilized as food. In contrast, S.pacifica migrated from 
kelp fronds into the surrounding open water at night 
to capture small plankton (Hobson and Chess 1976). 



63 



FISHERY BULLETIN: VOL. 82. NO. 1 



Vertical Patterns of Species, 
Abundances, and Sizes 

Several of the commonly occurring species within 
the Habitat Reef kelp forest were far more abundant 
in the canopy than in the lower zones. Ampithoe plea, 
Hyale frequens,Acanthomysis sculpta, and Hippolyte 
clarki all displayed this type of distribution, and 
other investigators have noted the canopy prefer- 
ences of these species. Limbaugh (1955) described a 
large canopy-dwelling amphipod {Ampithoe sp.) that 
formed a tube by rolling and "stitching" the edge of a 
Macrocystis blade. Several investigators working in 
kelp forests off San Diego and at Habitat Reef have 
noted the canopy occurrence of Aconthomysis 
sculpta (Limbaugh 1955; Clutter 1967; Clarke 1971; 
Hobson and Chess 1976) and H. clarki (Hobson and 
Chess 1976). Lowry (unpubl., cited in Lowry et al. 
1974) observed large numbers off/, californiensis, a 
close relative of//, clarki, in the canopy of kelp forests 
off central California. 

The canopy contained larger gammarids, mysids, 
and shrimps as well as proportionately greater num- 
bers of large individuals of these groups than in either 
of the lower zones. Size-selective predation by fishes 
frequently has been documented to be a major factor 
in structuring aquatic communities (Brooks and 
Dodson 1965; Archibald 1975; Vince et al. 1976; 
Macan 1977; Nelson 1979) and may account for the 
size distributions of invertebrates observed at 
Habitat Reef. The interdigitating fronds of the 
canopy greatly increase the structural complexity in 
this zone and may offer more spatial refuge for motile 
invertebrates than provided by the middle and bot- 
tom zones. As increased structural complexity has 
been demonstrated to decrease effectiveness of prey 
capture by fishes, particularly larger prey (Vince et al. 
1976; Brock 1979; Coen et al. 1981; Heck and Tho- 
man 1981; Savino and Stein 1982), the canopy com- 
plexity may discourage extensive foraging by 
fishes. 

Relatively few fishes forage within the kelp 
canopies off southern California. The most abundant 
fish is the kelp perch, Brachyistius frenatus, a small 
diurnal species that forages preferentially in the 
canopy and preys extensively on small gammarids 
and copepods (Hobson 1971; Bray and Ebeling 
1975; Hobson and Chess 1976). Other fishes are 
observed in the kelp canopy, but the large-mouthed 
species are much less abundant than the kelp perch 
and forage more often in other areas of the kelp 
forest, and the small-mouthed species capture small 
planktonic prey or utilize small invertebrates 
attached directly to the kelp surfaces (Bray and E bel- 



ing 1975; Hobson and Chess 1976; Bernstein and 
Jung 1979). Consequently, predation pressure on 
larger individuals of motile prey in the canopy may be 
reduced relative to the lower zones, resulting in a pro- 
portionately greater abundance of larger individuals. 
For example, the mysid S. pacifica was much more 
abundant in the lower zones than in the canopy, yet 
the largest individuals consistently were present in 
the canopy. 

Alternate hypotheses may explain the size 
stratification of some species. Intraspecific behav- 
ioral interactions may confine certain size classes to 
specific zones, as demonstrated experimentally for 
an amphipod (Van Dolah 1978). Larger individuals 
may be more abundant in the canopy simply in re- 
sponse to the presence of preferred food types and/ 
or sizes, although this hypothesis has not been 
examined. 

The size distribution of invertebrates in the lower 
zones resembled the size distribution of insects in 
temperate terrestrial forests (Schoener 197 1), in that 
both areas supported large numbers of small, and few 
large, individuals. The size distribution in the 
canopy, however, was somewhat similar to the insect 
size distribution of tropical terrestrial forests where 
there are proportionately greater numbers of large 
insects (Schoener and Janzen 1968; Schoener 1971). 
The presence of larger insects in the tropical forests 
effectively expands the food size dimension relative 
to the temperate forests (assuming equal abundance). 
The expansion has been hypothesized to account for 
some of the increased diversity of bird species in the 
tropics, as much of this increase is due to the addition 
of insectivorous birds adapted to capture large 
insects (Schoener 1971). 

In contrast to the tropical forests, the higher propor- 
tion of large prey items in the Habitat Reef kelp 
canopy apparently did not attract additional species 
of fish predators. Nevertheless, it may be useful to 
examine the size distributions of important prey 
items in other kelp forests to determine whether a 
relationship exists between prey size distributions 
and fish species diversity. 

Seasonal Patterns of Species, 
Abundances, and Sizes 

The kelp-associated invertebrates as a group did 
not exhibit seasonal cycles. Numbers and biomass 
generally were highest during winter 1975, with the 
marked increase in biomass due primarily to 
increased abundances of the relatively large canopy 
mysid A. sculpta and shrimp H. clarki. Gammarid 
amphipods, particularly M. litotes, were largely re- 



64 



COYER: INVERTEBRATE ASSEMBLAGE WITH CIANT KELP 



sponsible for the increased abundances in the lower 
zones during this period. 

Fluctuations in the population size of several 
species may have been associated with changes in 
kelp biomass, particularly the general decline of kelp 
biomass beginning in fall 1976. The canopy mysid 
probably attains its greatest population size during 
winter; however, the canopy was markedly reduced in 
area by winter 1976-77 and the mysid was rare. 
Copepods and gammarids displayed decreased 
canopy abundances during late 1976, and in the 
lower zones, abundances of the gammarid M. litotes 
began to decline as kelp biomass was reduced. As the 
canopy mysid andM. litotes were major components 
of the general invertebrate peak observed during 
winter 1975-76, their reduced abundances in late 
1976 undoubtedly were a major reason for the 
absence of a general invertebrate peak in late 
1976. 

Reduction in kelp biomass, however, did not affect 
H. clorki. Even though the shrimp was most 
numerous in the canopy, its abundance in the 
reduced canopy of late 1976 was similar to levels 
recorded in the larger canopy of late 1975. 

Although the amount of kelp biomass ultimately 
must determine the abundance and occurrence of 
kelp-associated invertebrates, the importance of 
proximal factors remains to be determined. Proximal 
factors may be particularly important in many areas 
of southern California, where the kelp forests are 
characterized by relatively long-term cycles of loss 
and renewal (Rosenthal et al. 1974). In such con- 
ditions of relative biomass constancy, abundances of 
some species may not be correlated with seasonal 
changes (i.e., temperature, day length, nutrients, 
etc.). Additional research is necessary to determine 
the importance of proximal factors such as kelp 
quality (healthy vs. decomposing), inter- and intra- 
specific competition for space and food, and preda- 
tion by fishes and/or motile invertebrates, in 
determining the abundance and occurrence of kelp- 
associated invertebrates. 

ACKNOWLEDGMENTS 

The manuscript was adopted from a portion of a 
doctoral dissertation completed at the University of 
Southern California. I thank my committee, chaired 
by J.N. Kremer, and am grateful to R. L. Zimmerand 
R. R. Given for their support and cooperation at the 
Catalina Marine Science Center. The substantial 
field assistance of J. R. Chess, J. F. Pilger, C. S. 
Shoemaker, and T. E. Audesirk is sincerely appre- 
ciated. Special thanks to D. Cadien, J. R. Chess, G. 



Kramer, B. Myers, J. Soo-Hoo, J. Word, and R. C. 
Zimmerman for assistance with species identifica- 
tion and to G. S. Hageman for help in sorting samples. 
The valuable suggestions of R. J. Schmitt, R. F. 
Ambrose, and two anonymous reviewers improved 
earlier drafts of the manuscript. 

The research was supported in part by the NOAA 
Office of Sea Grant under Grant No. USDC 04-158- 
44881 to the University of Southern California and 
by Sea Grant Traineeships. 

LITERATURE CITED 

Archibald, C. P. 

1975. Experimental observations of the effects of predation 

by goldfish (Carrassius auratus) on the zooplankton of a 

small saline lake. J. Fish. Res. Board Can. 32:1589- 

1594. 
Bernstein, B. B.. and N. Jim; 

1979. Selective pressures and coevolution in a kelp canopy 

community in southern California. Ecol. Monogr. 

49:335-355. 
Bray, R. N.. and A. W. Ebeling. 

1975. Food, activity, and habitat of three "picker-type" mi- 
crocarnivorous fishes in the kelp forests off Santa Bar- 
bara, California. Fish. Bull., U.S. 73:815-829. 

Brock. R. E. 

1979. An experimental study on the effects of grazing by 
parrotfishes and role of refuges in benthic community 
structure. Mar. Biol. (Berl.) 51:381-388. 
Brooks. J. L., and S. I. Dodson. 

1965. Predation, body size, and composition of 
plankton. Science (Wash., D.C.) 150:28-35. 
Clarke, W. D. 

1971. Mysids of the southern kelp region. In W. J. North 
(editor), The biology of the giant kelp beds (Macrocystis I 
in California, p. 369-380. Beih. Nova Hedwigia 32. 
Clutter, R. I. 

1967. Zonation of nearshore mysids. Ecology 48:200-208. 
Coen, L. D.. K. L. Heck, Jr.. and L. G. Abki.e. 

1981. Experiments on competition and predation among 
shrimps of seagrass meadows. Ecology 62:1484-1493. 
Darwin, C. 

1860. The voyage of the Beagle. Anchor Books, Doubleday 
and Co., Garden City, N.Y., 524 p. (1962) 
Ghelardi, R. J. 

1971. Species structure of the animal community that lives in 
Macrocystis pyrifera holdfasts. In W. J. North (editor). 
The biology of giant kelp beds (Macrocystis ) in California, 
p. 381-420. Beih. Nova Hedwigia 32. 
Hammer, R. M., andR. C. Zimmerman. 

1979. Species of demersal zooplankton inhabiting a kelp 
forest ecosystem off Santa Catalina Island, Califor- 
nia. Bull. South. Calif. Acad. Sci. 78:199-206. 
Heck, K. L., Jr., andT. A. Thoman. 

1981. Experiments on predator-prey interactions in 
vegetated aquatic habitats. J. Exp. Mar. Biol. Ecol. 
53:125-134. 

HOBSON. E. S. 

1971. Cleaning symbiosis among California inshore 
fishes. Fish. Bull, U.S. 69:491-523. 

HoBsn\. e. S., and. J. R. Chess. 

1976. Trophic interactions among fishes and zooplankters 



65 



FISHERY BULLETIN: VOL. 82, NO. 1 



near shore at Santa Catalina Island, California Fish. 
Bull., U.S. 74:567-598. 

Jackson, c, a. 

1977. Nutrients and production of giant kelp, Macrocystis 
pyrifera, off southern California. Limnol. Oceanogr. 
22:979-995. 

Jones, l. G. 

1971. Studies on selected small herbivorous invertebrates 
habiting Macrocystis canopies and holdfasts in southern 
California kelp beds. In W. J. North (editor), The biology 
of giant kelp beds (Macrocystis) in California, p. 343- 
367. Beih. Nova Hedwigia 32. 
LlMBAI GH, C. 

1955. Fish life in the kelp beds and the effects of kelp harvest- 
ing. Univ. Calif. Inst. Mar. Resour. Ref. 55-9, 158 p. 
Lowry, L. F., A. J. McElroy, and .J. S. Pearse. 

1974. The distribution of six species of gastropod molluscs in 
a California kelp forest. Biol. Bull. (Woods Hole) 
147:386-396. 
Macan.T.T. 

1977. The influence of predation on the composition of fresh- 
water animal communities. Biol. Rev. 52:45-70. 
Miller, D. J., and J. J. Geibel. 

1973. Summary of blue rockfish and lingcod life histories; a 
reef ecology study; and giant kelp, Macrocystis pyrifera 
experiments in Monterey Bay, California. Calif. Dep. 
Fish Game, Fish Bull. 158, 137 p. 
Nelson, W. G. 

1979. Experimental studies of selective predation on 
amphipods: Consequences for amphipod distribution and 
abundance. J. Exp. Mar. Biol. Ecol. 38:225-245. 
North, W. J. 

1971. Introduction and background. In W. J. North (editor), 
The biology of giant kelp beds [macrocystis) in California, 
p. 1-96. Beih. Nova Hedwigia 32. 
QUAST, J. C. 

1968. Observations on the food of the kelp-bed fishes. In W. 
J. North and C. L. Hubbs (compilers and editors), Utiliza- 
tion of kelp-bed resources in southern California, p. 109- 



142. Calif. Dep. Fish Game, Fish Bull. 139. 
Rosenthal, R. J., W. D., Clarke, and P. K. Dayton. 

1974. Ecology and natural history of a stand of giant kelp, 
Macrocystis pyrifera, off Del Mar, California. Fish. Bull., 
U.S. 72:670-684. 
Sayino, J. F., and R. A. Stein. 

1982. Predator-prey interaction between largemouth bass 
and bluegills as influenced by simulated, submerged 
vegetation. Trans. Am. Fish. Soc. 111:255-266. 
SCHOENER, T. W. 

1971. Theory of feeding strategies. Annu. Rev. Ecol. Syst. 
2:369-404. 
SCHOENER, T. W., AND D. H. JANZEN. 

1968. Notes on environmental determinants of tropical ver- 
sus temperate insect size patterns. Am. Nat. 102:207- 
224. 
SlEGEL, S. 

1956. Nonparametric statistics for the behavioral 
sciences. McGraw-Hill, N.Y., 312 p. 
Van Dolah, R. F. 

1978. Factors regulating the distribution and population 
dynamics of the amphipod (lammaruspalustris in an inter- 
tidal salt marsh community. Ecol. Monogr. 48:191-217. 
Vince, S., I. Valiela, N. Backus, and J. M. Teal. 

1976. Predation by the salt marsh killifish Fundulus 
heternclitus (L.) in relation to prey size and habitat struc- 
ture: Consequences for prey distribution and abundance. J. 
Exp. Mar. Biol. Ecol. 23:255-266. 
Wing, B. L., and K. A. Clendenning. 

1971. Kelp surfaces and associated invertebrates. In W. J. 
North (editor), The biology of giant kelp beds (Macrocys- 
tis) in California, p. 319-341. Beih. Nova Hedwigia 32. 
YOSHIOKA, P. M. 

1982a. Predator-induced polymophism in the bryozoan 
Membranipora membranacea (L.). J. Exp. Mar. Biol. 
Ecol. 61.233-242. 
1982b. Role of planktonic and benthic factors in the popula- 
tion dynamics of the bryozoan Membranipora mem- 
branacea. Ecology 63:457-468. 



66 



SPRING AND SUMMER PREY OF 

CALIFORNIA SEA LIONS, ZALOPHUS CALIFORNIANUS, 

AT SAN MIGUEL ISLAND, CALIFORNIA, 1978-79. 

George A. Antonelis, Jr., Clifford H. Fiscus, and Robert L. DeLong 1 

ABSTRACT 

During the late spring and summer of 1978 and 1979, 224 scats were collected from rookeries of the Cali- 
fornia sea lion, Zalophus californianus , at San Miguel Island for the purpose of identifying prey species. A 
total of 2,629 otoliths and 2,06 1 cephalopod beaks were recovered. The frequency of occurrence for the four 
most commonly identified prey species was 48.7% Pacific whiting, Merluccius productus; 46.7% market 
squid, Loligo opalescens; 35.9% rockfish, Sebastes spp.; and 20.0% northern anchovy, Engraulis mordax. 
Seasonal variability in the frequency of occurrence of these four prey species from late spring to summer 
indicates that California sea lions feed opportunistically on seasonally abundant schooling fishes and squids. 
Five species of fish (California smoothtongue, Bathylagus stilbius; northern lampfish, Stenobrachius leucop- 
sarus; chub mackerel, Scomber japonicus; medusafish, Icichthys lockingtoni; sablefish, Anoplopoma fimbria) 
and one cephalopod (two-spotted octopus, Octopus bimaculatus) were identified as previously unreported 
prey of the California sea lion. 



The California sea lion, Zalophus californianus, is the 
most abundant pinniped inhabiting the coastal 
waters off California (Le Boeuf and Bonnell 1980). 
During the summer most California sea lions are on 
or near their breeding sites which are located on 
islands south of Point Conception, along the coast of 
southern California, Baja California, and into the 
Gulf of California. After the breeding season in the 
summer, a portion of the subadult and adult male sea 
lion populations migrates north of Point Conception 
as far as British Columbia, while the rest of the pop- 
ulation remains off the coasts of southern California 
and Baja California, Mexico (Peterson and Bartho- 
lomew 1967). Numerous studies of the food of 
migrant male California sea lions have been con- 
ducted in the areas north of their traditional breeding 
islands (Briggs and Davis 1972; Jameson and 
Kenyon 1977; Morejohn et al. 1978; Bowlby 1981; 
Everitt et al. 1981; Jones 1981; Ainley et al. 1982; 
Bailey and Ainley 1982), while comparatively little 
information has been reported on the feeding 
behavior of sea lions in areas off the coast of Cali- 
fornia south of Point Conception (Rutter et al. 1904; 
Scheffer and Neff 1948; Fiscus and Baines 1966). 
From the information presented in all of these 
studies, it has been suggested that California sea 
lions feed opportunistically on a variety of prey 



species (Antonelis and Fiscus 1 980) and that "switch 
feeding" is probably an important component of 
their feeding behavior (Bailey and Ainley 1982). 
However, since most of the information on sea lion 
feeding behavior is based on observations north of 
their breeding islands, additional information from 
within their breeding range would allow us to deter- 
mine if similar feeding characteristics can be expect- 
ed in other geographical areas. 

Studies conducted before 1970 usually obtained 
stomach contents for feeding information by killing 
sea lions, while most post- 1970 feeding studies have 
used nonlethal techniques including examination of 
scats and oral rejecta (spewings) and direct 
behavioral observations. Another method was the 
examination of gastrointestinal tracts from animals 
found dead. In this study, prey-species classification 
is based on the identification of fish otoliths and 
cephalopod beaks found in scats collected during the 
spring and summer for two consecutive years on the 
California sea lion rookeries of San Miguel Island, 
Calif. In addition to the identification of prey, we 
calculated the percent frequency of occurrence of 
each prey, compared annual and seasonal differ- 
ences in prey selection, and estimated the lengths 
and weights of the most frequently occurring prey 
species. 



'Northwest and Alaska Fisheries Center National Marine Mammal 
Laboratory, National Marine Fisheries Service, NOAA, 7600 Sand 
Point Way N.E., Seattle, WA 98115. 



Manuscript accepted July 1983. 

FISHERY BULLETIN: VOL. 82, NO. 1, 1984. 



MATERIALS AND METHODS 

Scats were collected from areas utilized exclusively 

67 



FISHERY BULLETIN: VOL. 82, NO. 1 



by California sea lions on the west end of San Miguel 
Island, Calif., during spring (2-3 May 1978; 2 and 16 
May 1979) and summer (3-4 August 1978; 30-3 1 July 
1979). During both sample periods, scats were 
collected from areas where mostly females and 
juveniles of both sexes occurred and relatively few 
(<12% of the total animals censused) adult and sub- 
adult males were present. In order to document the 
occurrence of prey species which were consumed at 
or close to the time of collections, only recent scats, 
which showed no obvious signs of desiccation, were 
collected. Each scat was placed in a plastic bag, 
where it was later soaked in water or a solution of 
about 1 part liquid detergent to 100 parts water for 
about 24 h. Each bag was shaken occasionally to 
facilitate emulsification of the digested organic 
material, and then rinsed with water through three 
nested sieves with screen mesh sizes of 3.35 mm, 
2.00 mm, and 1.00 mm from top to bottom. After 
most of the soft digested organic material was 
washed away, fish otoliths and cephalopod beaks 
were removed and stored in a solution of 70% 
ethanol. Prey totals were determined by using the 
higher number of left or right otoliths and upper or 
lower squid beaks. The otoliths were identified by 
the late J. Fitch, California Department of Fish and 
Game, Long Beach, Calif., the octopus beaks by E. 
Hochberg, Santa Barbara Museum of Natural His- 
tory, Santa Barbara, Calif., and the squid beaks by 
the second author. 

The data for each of the four major prey species 
were summarized by a three-way (2X2X2) con- 
tingency table and tested for independence of 
occurrence by season, year, and both season and year 
(Fienberg 1977). 

Length measurements of these otoliths and squid 
beaks were used to estimate the body lengths or ages 
of the most frequently occurring prey species. 
Although many otoliths and beaks of all sizes were 
recovered from the scats in good condition, some 
were not measured because they were broken or 
showed obvious signs of damage from digestion. We 
assumed that damage to the otoliths and squid beaks 
collected in this study was not dependent on size. 
Lengths of northern anchovy, Engraulis mordax, 
were estimated from a regression equation of fish 
lengths on otolith lengths (Spratt 1975). Length 
information for rockfish, Sebastes spp., was obtained 
from previously reported data (Phillips 1964) for 
specimens (bocaccio, Sebastes paucispinis) of the 
same age as most of the rockfish reported in this 
study. Bocaccio was chosen as the representative 
rockfish because it has been reported as the most 
abundant rockfish in the waters near San Miguel 



Island (Best and Oliphant 1965). The regression 
equation used to estimate the length of Pacific whit- 
ing, Merluccius productus, was derived in this study 
from specimens collected off the coast of southern 
California by the National Marine Fisheries Service 
(NMFS). The Pacific whiting otoliths and the corre- 
sponding length information were provided by K. 
Bailey of the NMFS Northwest and Alaska Fisheries 
Center, Seattle, Wash. Market squid, Loligo opales- 
cens, lengths were estimated from a regression equa- 
tion of dorsal mantle length on upper hood length of 
the beak. Upper hood measurements were chosen for 
the estimation of squid lengths because they were 
reported as having the highest correlation to dorsal 
mantle length (Kashiwada et al. 1979). 

In order to detect changes in the diet which would 
reflect apparent yearly changes in the age and size 
composition of a specific prey-species population, we 
compared the lengths of otoliths for 1978 and 1979 
using the Wilcoxon rank sum test (Hollander and 
Wolfe 1973). 

Weight estimates of the most frequently occurring 
prey species were obtained by using the prey length 
estimate (described above) in regression equations 
of length and weight measurements or by obtaining 
weight data from fish which were the same age as 
those identified in the scats (Phillips 1964; Fields 
1965; Dark 1975; Pacific Fishery Management 
Council 1978). The total estimated weight for each of 
the four major prey species was obtained by mul- 
tiplying the weight of the average-sized prey by the 
number of individuals represented in the scat collec- 
tion. Differences between these estimates could 
not be statistically analyzed because the raw data 
for the growth curves of each species were not 
available. 

The names of fishes follow Fitch and Lavenberg 
(1968) and Robins (1980), and those of cephalopods 
follow Fields (1965) and Young (1972). 

RESULTS 

We collected 224 California sea lion scats on San 
Miguel Island during the spring and summer of 1978 
and 1979. From 195 (87%) of those scats, we 
recovered 2,629 otoliths and 2,061 cephalopod 
beaks. Twenty-nine (13%) scats did not contain 
otoliths or cephalopod beaks. The prey species iden- 
tified in the scats are shown in Table 1 by their per- 
centage of occurrence. The four most frequently 
occurring prey in scats containing otoliths and/or 
cephalopod beaks were Pacific whiting (48.7%), 
market squid (46.7%), juvenile rockfish from 
the Sebastes paucispinis -goodei-jordani complex 



68 



ANTONELIS ET AL.: SPRING AND SUMMER PREY OF CALIFORNIA SEA LIONS 



TABLE 1. — Percentage occurrence of all prey species identified from 
195 California sea lion scats collected on San Miguel Island, Calif., 
spring and summer, 1978-79. 



Prey 




Oc 


currence 


Scientific name 


Common name 


No. 


% 


Merluccius producws 


Pacific whiting 


95 


48.7 


Loligo opalescens 


market squid 


91 


46 7 


Sebastes spp. 


rockfish (juvenile) 


70 


359 


Engrauhs mordax 


northern anchovy 


39 


20.0 


Octopus rubescens 


red octopus 1 


19 


9.7 


Trachurus symmetncus 


jack mackerel 


9 


46 


Onychoteuthis 


nail squid 


9 


46 


boreahjapomcus 








Gonatidae (other than 


squid 


8 


4.1 


Gonatus sp.) 








Scomber japonicus 2 


chub mackerel 


7 


3.6 


Pepnlus similhmus 


Pacific pompano 


5 


2.5 


Symbolophorus 


California lantern- 


5 


25 


californiensis 


fish 






Gonatus sp. 


squid 


2 


1.0 


Microstomus pacificus 


Dover sole 


2 


1.0 


Bathylagus stilbius 2 


California smooth- 
tongue 


2 


1.0 


Senphus pohtus 


queenfish 


2 


10 


Zalembtus rosaceus 


pink surf perch 




0.5 


Anoplopoma fimbria 2 


sablefish 




05 


Poncbthys notatus 


plainfin midshipman 




0.5 


Ictchthys lockingtoni 2 


medusafish 




5 


Stenobrachius leucopsarus 2 


northern lampfish 




05 


Octopus bimacu/atus 2 


two-spotted octopus 




0.5 


1 Pelagic life stage 








2 Not previously reported as prey 


of the California sea lion. 







(35.99c) 2 , and northern anchovy (20.0%). All other 
prey species occurred in <10.0% of the scats. 

Relative length and weight estimates of the four 
major prey species and the information used to calcu- 
late these estimates are shown in Figure 1 and Table 
2, respectively. The length and weight information 
for rockfish is from data reported by Philips (1964) 
for one of the three species (5. paucispinis) repre- 
sented in this study. 

Measurements of otoliths from Pacific whiting and 
northern anchovy provided sufficient information to 
compare changes in the size and age of each prey 
group from 1978 to 1979. For Pacific whiting the 
lengths of otoliths were significantly greater (W* = 



•'About 95Tc of the juvenile rockfish were yearlings and were in- 
cluded in this three-species complex because their otoliths are too 
similar to differentiate. 



5.82, P< 0.0001) in 1979 (x = 7.71 mm,n = 90) than 
in 1978 (x = 6.71 mm, n = 132). From these otolith 
measurements, we estimated the mean length of 
Pacific whiting at 156 mm in 1978 and 176 mm in 
1979. All of the Pacific whiting otoliths were obtained 
from 1- and 2-yr-old fish. The occurrence of 1-yr-old 
fish in the sea lion diet was estimated at 98.5% in 
1978 and 70% in 1979. For northern anchovy, the 
lengths of otoliths were significantly greater (W* = 
4.36,P < 0.0001) in 1978 (j = 3.58 mm, n= 19) than 
in 1979 (x = 3.01 mm, n = 75). For these otolith 
measurements we estimated the mean length of 
northern anchovy at 111 mm in 1978 and 92 mm in 
1979. Although all age classes of northern anchovy 
were recovered from the scats, there was a notable 
change in the percent occurrence of yearling fish 
from 1978 (42%) to 1979 (81%). 

The percentage of occurrence in the four major prey 
species is shown for the spring and summer of 1978 
and 1979 in Figure 2. From the three-way con- 
tingency table analysis, it was determined that 
Pacific whiting occurred significantly more frequent- 
ly in 1978 than in 1979 (P < 0.01), and there was a 
greater percentage of occurrence in spring than in 
summer (P < 0.01). For rockfish, there was no signifi- 
cant difference in occurrence between years; 
however, there was a greater percentage of 
occurrence in the summer than in spring (P < 0.01). 
The percentage occurrence of northern anchovy was 
not significantly different between season, but there 
was a significantly greater occurrence in 1979 than 
in 1978 (P < 0.01). The relative proportion of oc- 
currence for the two seasons for each year was 
significantly different (P < 0.01) for Pacific whiting, 
rockfish, and northern anchovy. Tests of significance 
could not be done for market squid because of the 
strong three-way interaction between occurrence, 
season, and year. It is apparent, however, that the 
percent occurrence of market squid did increase 
from spring to summer during both years of the study 
(Fig. 2). 



Table 2.— Information used in estimating the length of the four major prey species identified from 
the scats of California sea lions, on San Miguel Island, Calif., 1978-79. 



Prey species 



Regression 
equation 



R 2 



Reference 



Market squid 



Y = 0.243 + 0.0481X 



60 0.974 upper hood dorsal mantle Kashiwada 

length (mm) length (mm) etal. 1979 



Pacific whiting Y = 26 2 + 19.38X 84 977 fork otolith This study 

length (mm) length (mm) 

Juvenile rockfish' (') 155 (') (') Phillips 

1964 

Northern anchovy Y = -8 4946 + 33 216X 677 0.774 standard otolith Spratt 

length (mm) length (mm) 1975 

'Length measurements are from yearling bocaccio. Sebastes paucispinis. 



69 



FISHERY BULLETIN: VOL. 82. NO. 1 



MARKET SQUID 

x=127mm (Weight = 47.0 g) 

SD = 17 mm 

Range = 62-185 mm 

n = 76 




PACIFIC WHITING 

x= 166 mm (Weight = 42.6 g) 

SD = 60 mm 

Range = 89-261 mm 

n = 222 




BOCACCIO (Rockfish) 

x=171mm (Weight = 45.4 g) 

SD = 22 mm 

Range = 129-227 mm 

n = 155 







^7*V 




4tZ 



NORTHERN ANCHOVY 

x = 95 mm (Weight = 10.8 g) 

SD = 8 mm 

Range = 55-141 mm 

n = 94 

Fir.i re 1 .—Relative length and weight estimates of the four major prey species identified in California sea lion scats collected on San Miguel 
Island, Calif., spring and summer, 1978-79. Methods used to calculate these estimates are shown in Table 2. 




The number of prey species occurring in individual 
scats changed from spring to summer. For combined 
years, the percentages of scats containing single or 
multiple prey are shown in Figure 3. In the spring, the 
percentage of singly occurring prey species in scats 
was 59.7%; in the summer the percentage dropped 



to 34.6%. Scats containing more than one prey 
increased from 17 species combinations occurring in 
40.3% of the scats in the spring to 23 in 65.4% during 
the summer. 

The percentages of the total estimated weight of the 
four major prey species for spring and summer are 



70 



ANTONELIS ET AL.: SPRING AND SUMMER PREY OF CALIFORNIA SEA LIONS 



Number 


of 




scats 






Spring 


1978 


n=21 




1979 


n=46 


Summer 


1978 


n=43 




1979 


n=85 



1978 



1979 



Northern anchovy 



Juvenile Rockfish 



Market squid 



Pacific whiting 



Spring 
Summer 




n = 39 



n = 18 




n = 26 



-i i i l i i i i l i i    i   '    

10 20 30 40 50 60 70 80 90 100 10 20 30 40 50 60 70 80 90 100 



Occurrence (percent) 



Occurrence (percent) 



FIGURE 2. — Spring and summer occurrence (percentage) of the four major prey species identified in California sea 

lions scats collected on San Miguel Island, Calif., 1978-79. 



Spring, n = 67 
Summer, n = 128 



c 
u 

v- 
<D 
_Q. 

0) 

o 

c 
a> 

k_ 

L. 

D 
U 
O 

O 




2 3 4 5 

Number of different species 



FIGURE 3. — Occurrence of single and multiple prey species in in- 
dividual sea lion scats collected on San Miguel Island, Calif., 1978- 
79. 



shown in Figure 4. The seasonal changes in the per- 
cent of weight for Pacific whiting showed a decrease 
from spring to summer in 1978 and 1979, while an 
increase occurred from spring to summer for market 
squid in 1978 and rockfish in 1979. There was 
relatively little change in the percentage of weight 
between the two seasons for market squid in 1979 
and rockfish in 1978. The northern anchovy also 
showed little difference between the two seasons 
during both years. Additionally, the results from this 
analysis show that market squid made the greatest 
contribution to the total estimated weight of prey in 
the summer of 1978 (71.2%) and for both spring 
(53.9%) and summer (48.7%) of 1979, while Pacific 
whiting made the greatest contribution to the total 
estimated weight only in the spring of 1978 
(87.3%). 

DISCUSSION 

Pacific whiting, market squid, juvenile rockfish, and 
northern anchovy were the four most important prey 
of California sea lions at San Miguel Island during the 
spring and summer of 1978 and 1979. These four 
prey species have also been reported as common 
prey of California sea lions in areas north of Point 
Conception (Morejohn et al. 1978; Everitt et al. 
1981; Jones 1981; Ainley etal. 1982) and exemplify 



71 



FISHERY BULLETIN: VOL. 82, NO. 1 



Estimated weigh 


in kg 






Spring 


1978 
1979 


10.7 kg 
35.0 kg 




Summe 


• 1978 
1979 


29.7 kg 
29.3 kg 




1978 








Northern anchovy 







y 


Spring 
Summer 


Juvenile Rockfish 


- 








"•"""" 


* 


■n 


mam 


 


Pacific whiting 


- 


i i 


i i i 


1 1 


I l 


( 


) 10 


20 30 


40 50 60 


70 80 90 100 








Percentage 







1979 




10 20 30 40 50 60 70 80 90 100 
Percentage 



FIGURE 4.— Percentages of the total estimated weight for the four major prey species in spring and summer, 

1978-79. 



the type of large, dense schooling prey which are 
commonly fed upon by many of the pinnipeds in the 
coastal waters off California (Antonelis and Fiscus 
1980). Furthermore, the variety of food items re- 
ported in this and other studies (Jameson and Kenyon 
1977; Morejohn et al. 1978; Bowlby 1981; Jones 
1981; Ainley et al. 1982) indicates that California sea 
lions are capable of foraging on a wide range of fish 
and cephalopods. 

The range in the average length estimates of the 
four major prey species (95-171 mm) does not exhibit 
a great diversity in size, and may reflect a size pref- 
erence for sea lions feeding in the waters near San 
Miguel Island. Both Pacific whiting and rockfish 
attain a much larger size as adults (Phillips 1964; 
Dark 1974), while the length estimates of northern 
anchovy and market squid are within the size range of 
juveniles and adults (Fields 1965; Spratt 1975). 

As more information is obtained on the prey and the 
foraging behavior of California sea lions, researchers 
will attempt to evaluate the biomass of each prey 
species consumed (Bailey and Ainley 1982). These 
types of studies require information on the variations 
in the diet of California sea lions throughout their 
range. For this reason, we compare the estimated 
length data of market squid and Pacific whiting from 
this study with similar information reported in areas 
north of Point Conception. The estimated lengths 
were similar for market squid which were preyed 
upon by California sea lions in Monterey Bay, Calif. 



(Morejohn et al. 1978) and in the waters near San 
Miguel, with mean values of 130 mm (Morejohn et al. 
1978, estimated from figure 27) and 127 mm, respect- 
ively. California sea lions foraged on all age classes of 
market squid in both areas. For Pacific whiting, 
however, differences between the northern and 
southern range of the California sea lion were 
apparent, with estimated length averages ranging 
from 250 to 360 mm at Southeast Farallon Island, 
Calif. (Bailey and Ainley 1982) compared with an 
average of 166 mm at San Miguel Island. Primarily 1- 
and 2-yr-old fish were preyed upon near San Miguel, 
while 2- and 3-yr-old fish were reported as prey at 
Southeast Farallon. 

From these comparisons, we assume that squid of 
all sizes and age classes will be preyed upon by 
California sea lions, in both their breeding and non- 
breeding ranges. For Pacific whiting, however, there 
are apparent differences in the size and age classes 
consumed by California sea lions in the two areas. 
These differences could be related to three possible 
factors: 1) There could be differential feeding 
according to various age and/or sex classes of sea 
lions which occur in the two areas. When present, 
there are mostly subadult and adult males at 
Southeast Farallon Island, and at San Miguel Island 
there are comparatively fewer subadult and adult 
males and many more females and juveniles of both 
sexes (Peterson and Bartholomew 1967; Le Bouef 
and Bonnell 1980; Ainley et al. 1982). 2) Differences 



72 



ANTONELIS ET AL.: SPRING AND SUMMER PREY OF CALIFORNIA SEA LIONS 



between the two areas may be an artifact of the dif- 
ferent methods used for estimating fish length. 3) 
What appears most probable to us, is the differential 
geographical distribution of Pacific whiting accord- 
ing to age. Generally, the younger fish occur in the 
southern portion of their range, and, although there is 
some overlap in age groups, the age and size of the 
fish increase in a northward direction (Bailey et al. 
1982). 

In cases where sufficient life history information is 
available, seasonal or annual changes in the 
occurrence of the four major prey (Fig. 2) can be 
related to known changes in the prey's relative abun- 
dance and availability to California sea lions. During 
both years of this study, the decrease in the 
occurrence of Pacific whiting in the scats from spring 
to summer appears to reflect known changes in the 
migration pattern of the species when adults and a 
portion of the juvenile population migrate toward 
shore and north of Point Conception (T. Dark'). For 
market squid and juvenile rockfish, however, the 
movement patterns off the coast of California are 
conspicuously different than Pacific whiting. Gener- 
ally, market squid increase in abundance in shallow 
waters (5-50 m depth) near the northern California 
Channel Islands in late spring, and peak numbers 
occur in the early summer during spawning (S.Kato 4 ). 
Inspection of the unpublished data from the 1970-75 
commercial catches of market squid within 30 nmi of 
Point Bennett, San Miguel Island, also indicated that 
peak abundance occurs during the summer months. 5 
Similarly, in spring through summer, juvenile rock- 
fish (S. paucispinis and 5. jordani) from the three- 
species complex identified in this study begin to move 
into more shallow waters (5-50 m depth) as they com- 
plete the pelagic stage of their life cycles (E. Hob- 
son 6 ). In these three instances, seasonal changes in 
the relative availability of Pacific whiting, market 
squid, and juvenile rockfish are reflected in the fre- 
quency of their occurrence in sea lion scats. A similar 
relationship was also suggested by Bailey and Ainley 
(1982), when they observed a seasonal change in the 
prey consumed by California sea lions near the 
Farallon Islands. 

Although the percentage of occurrence of northern 



'T. Dark, Northwest and Alaska Fisheries Center, National Marine 
Fisheries Service, NOAA, Seattle, WA 98112, pers. commun. 
1982. 

4 S. Kato, Southwest Fisheries Center Tiburon Laboratory, Na- 
tional Marine Fisheries Service, NOAA, Tiburon, CA 94920, pers. 
commun. 1981. 

'Data provided bv E. Knaggs, Calif. Dep. Fish and Game, Long 
Beach, Calif., 1982". 

fi E. Hobson, Southwest Fisheries Center Tiburon Laboratory, Na- 
tional Marine Fisheries Service, NOAA, Tiburon, CA 94920, pers. 
commun. 1981. 



anchovy in the scats showed no significant seasonal 
changes from spring to summer, the annual occurrence 
of otoliths from northern anchovies in the sea lion 
scats was significantly greater in 1979 than in 1978. 
Their low numbers in the 1978 scats could be related 
to a decline in the northern anchovy population 
resulting from poor recruitment of the 1974-77 year 
classes (Mais 1981). In 1978, however, the year-class 
recruitment was strong (Mais 1981), and the 
increased abundance appears to be reflected in an 
increased percentage of occurrence in the 1979 
collection. This explanation is corroborated by our 
comparison of the northern anchovy otoliths collect- 
ed during the 2 years, where we found that the 1979 
scats contained significantly smaller fish which were 
mostly (81%) yearlings from the 1978 year class. 

Differences in the annual occurrence of Pacific 
whiting and market squid were also noted in this 
study. For market squid, there was no fishery infor- 
mation available during the time of this study which 
would provide us with a possible explanation for 
these differences. With Pacific whiting, however, the 
decrease in occurrence in the scats from 1978 to 
1979 appears to be related to exceptionally high re- 
cruitment of the 1977 year class which was followed 
by an average, or possibly somewhat less-than- 
average, recruitment in 1978 (T. Dark footnote 3). 
This information is corroborated by a comparison of 
the Pacific whiting otoliths collected during the 2 
years of our study. In 1978, sea lions preyed upon 
significantly smaller fish which were mostly (98.5%) 
yearlings from the 1977 year class. 

Our analysis of the frequency occurrence of prey 
species per individual scat (Fig. 3) suggests that 
California sea lions commonly feed on single prey 
species during the spring and feed more frequently 
on multiple prey species in the summer. This shift 
from single to multiple occurrence of prey species in 
scats could reflect a decrease in the overall 
availability of the potential prey species in the sum- 
mer which may necessitate foraging on a greater 
variety of food items for survival (Morse 1980). Alter- 
natively, numerous potential prey species may 
become more available (Morse 1980) during the 
summer; thus, California sea lions could forage 
opportunistically on a greater variety of schooling 
fishes or squids which concentrate in a comparatively 
small area of high productivity. 

There are, however, a variety of factors which could 
affect prey-species availability. Seasonal migration, 
diel vertical migration, variability in schooling 
behavior, or physiological changes associated with 
spawning (Moyle and Cech 1982) are probably some 
of the more important factors related to prey selec- 



73 



FISHERY BULLETIN: VOL. 82, NO. 1 



tion and preference of California sea lions which 
necessitate additional research. 

Unfortunately, virtually no information has been 
reported on the digestive rates or retention time of 
the prey species' hard parts in California sea lions. 
Therefore, it is presently impossible to ascertain how 
many meals, or portions thereof, are represented in a 
single scat. There is some evidence, however, from 
feeding studies (Pitcher 1980) of harbor seals, Phoca 
uitulina, and (Miller 1978) northern fur seals, 
Callorhinus ursinus , which indicates that cephalopod 
beaks are not readily passed through the intestinal 
tract and are regurgitated. This would result in an 
underrepresentation of cephalopod beak percent- 
age-of-occurrence data from scats as suggested by 
Morejohn et al. (1978). Furthermore, the possible 
occurrence and identification of hard parts of second- 
ary prey (from the stomach of the prey of the marine 
mammal) could bias the results of scat or stomach 
analysis (Perrin et al. 1973). 

Additional information on the feeding habits of 
California sea lions can also be obtained from the 
weight estimates of the four major prey species iden- 
tified in this study. The 1978 and 1979 percentages 
of total weight estimates (Fig. 4) for each major 
species showed seasonal changes that are similar to 
the analysis of percentage of occurrence (Fig. 2), 
although there are a few exceptions. In 1979 the 
market squid weight estimate showed a slight de- 
crease, instead of an increase, from spring to sum- 
mer, however, of more importance, is its relationship 
to Pacific whiting. The estimated weight of market 
squid from the scats clearly exceeded the relative 
weight of Pacific whiting and other prey species con- 
sumed during the spring and summer of 1979. These 
results suggest market squid may be a more impor- 
tant food item than was predicted from the analysis 
of their percent of occurrence. The importance of the 
squid in the diet of the California sea lion during the 
summer months near the northern California Chan- 
nel Islands was also documented by Rutter et al. 
(1904), when they found that 84.6% (n = 13) of the 
animals examined had squid in their stomachs. 

Bailey and Ainley (1982) estimated the spring and 
summer percent (weight) of Pacific whiting in the 
California sea lion diet in the southern region to be 
within a range of 50 to 90%. Yet our estimates fell 
below 40% in the spring of 1979 and below 20% in the 
summer of both 1978 and 1979, and only one 
instance (spring 1978) did our estimates fall within 
the range suggested by Bailey and Ainley (1982). 
Since Bailey and Ainley (1982) based their estimates 
on data from California sea lions in the northern 
region, we assume our estimates more accurately 



represent the percent (weight) of Pacific whiting in 
the diet of California sea lions south of Point Concep- 
tion, and we recommend that additional feeding 
studies of California sea lions be conducted 
throughout their range. 

The percentage of estimated weight results also 
suggests that Pacific whiting was preyed upon more 
heavily in the spring of 1978 than in the spring of 
1979. This is consistent with the exceptionally high 
recruitment of the 1977 year class of Pacific whiting 
(discussed above) which was available as yearlings to 
California sea lions in 1978. 

Although these weight (biomass) estimates are only 
approximate measurements, they appear instructive 
when used in conjunction with percentage-of- 
occurrence data. Unfortunately, there is some uncer- 
tainty as to the accuracy of using estimates of weight 
to estimate consumption. Our ability to make con- 
sumption estimates awaits the resolution of several 
questions: 1 ) What proportion of a given meal is repre- 
sented in a single scat? 2) Are there differential 
digestive rates of fish and squid? 3) Do sea lions of 
different ages and sexes digest food differently? 

The results of this study suggest that the California 
sea lions found on San Miguel Island feed oppor- 
tunistically on prey species of changing availability, 
and we agree with Bailey and Ainley (1982) that they 
are behaviorally flexible enough to switch from one 
major prey species to another, both seasonally and 
annually. This type of flexibility in foraging appears 
to be adaptive and may be a major factor contributing 
to the success of the California sea lion population off 
the coasts of California and Baja California. 

ACKNOWLEDGMENTS 

Permission to work on San Miguel Island was grant- 
ed by the National Park Service in conjunction with 
the U.S. Navy. Logistical assistance was provided by 
the U.S. Coast Guard, U.S. Navy, and the Channel 
Islands National Park. 

Superintendent W. Ehorn and his staff of the Chan- 
nel Islands National Park frequently assisted us dur- 
ing our research activities. Others who volunteered 
their time and assistance during our research 
included L. Antonelis, T. Antonelis, and K. Antonelis 
of Seattle, Wash.; P. Collins of Santa Barbara 
Museum of Natural History, Santa Barbara, Calif.; E. 
Jameyson of MARIS, Seattle, Wash.; R. Morrow of 
Oregon State University, Corvallis, Oreg.; D. Seagars 
of NMFS Southwest Region, Terminal Island, Calif.; 
and B. Steward of Hubbs-Sea World Research In- 
stitute of San Diego, Calif. M. Weber of California 
Marine Mammal Center, Fort Cronkhite, Calif., pro- 



74 



ANTONELIS ET AL.: SPRING AND SUMMER PREY OF CALIFORNIA SEA LIONS 



vided valuable assistance during the entire 1978 field 
season. Advice and aid during statistical analysis of 
the data were given by J. Breiwick, R. Kappenman, R. 
Ryel, and A. York of NMFS Northwest and Alaska 
Fisheries Center, Seattle, Wash. 

The otoliths were identified by the late J. Fitch of 
Long Beach, Calif., and the octopus beaks by F. E. 
Hochberg of Santa Barbara Museum of Natural His- 
tory, Santa Barbara, Calif. Additional consultation 
and assistance in our efforts to understand the forag- 
ing ecology of the California sea lion were given by K. 
Bailey of NMFS Northwest and Alaska Fisheries 
Center, Seattle, Wash; E. Knaggs, and K. Mais of 
California Fish and Game, Long Beach, Calif; A. 
Mearns and M. Perez of NMFS Northwest and 
Alaska Fisheries Center, Seattle, Wash.; and P. 
Smith and G. Stauffer of NMFS Southwest Fisheries 
Center, La Jolla, Calif. 

The senior author (Antonelis) presented a portion 
of this paper at the 4th Biennual Conference on the 
Biology of Marine Mammals in San Francisco, 
December 1981. 



LITERATURE CITED 

AlNLEY, D. G., H. R. HUBER, AND K. M. Bailey. 

1982. Population fluctuations of California sea lions and the 
Pacific whiting fishery off central California. Fish. Bull., 
U.S. 80:253-258. 
Antonelis, G. A., Jr. and C. H. Fiscus. 

1980. The pinnipeds of the California current. Calif. Coop. 
Oceanic Fish. Invest. Rep. 21:68-78. 

Bailey, K. M., and D. G. Ainley. 

1982. The dynamics of California sea lion predation on 
Pacific whiting. Fish. Res. 1:163-176. 
Bailey, K. M., R. C. Francis, and P. R. Stevens. 

1982. The life history and fishery of Pacific whiting, Merluc- 
cius productus. Calif. Coop. Oceanic Fish. Invest. Rep. 
23:81-98. 
Best, E. A., and M. S. Oliphant. 

1965. Evaluation of the fish resources of the Pt. Arguello 
Area. Marine Resources of the Pt. Arguello Area, Part 
II. Bur. Commer. Fish. Contract 14-17-001-995, as part 
of agreement AT (49-7)-2428 for the Div. Biol, and Med.. 
U.S. At. Energy Comm., 129 p 
Bowlby, C. E. 

1981. Feeding behavior of pinnipeds in the Klamath River, 
northern California. M. A. Thesis, Humboldt State Univ., 
Eureka, Calif., 74 p. 

Briggs, K . T., and C. W. Davis. 

1972. A study of predation by sea lions on salmon in Mon- 
terey Bay. Calif. Fish Game 58:37-43. 
Dark, T. A. 

1975. Age and growth of Pacific hake, Merluccius produc- 
tus. Fish. Bull., U.S. 73:336-355. 
Everitt, R. D., P. J. Gearin, J. S. Skidmore, and R. L. 
DeLong. 

1981. Prey items of harbor seals and California sea lions in 
Puget Sound, Washignton. Murrelet 62:83-86. 



Fields, W. G. 

1965. The structure, development, food relations, reproduc- 
tion, and life history of the squid Loligo opalescens 
Berry. Calif. Dep. Fish Game, Fish Bull. 131:1- 
108. 

Fienberg, S. E. 

1977. The analysis of cross-classified categorical data. 
Mass. Inst. Technol. Press, Cambridge, Mass., 151 p. 
Fiscus, C. H., and G. A. Baines. 

1966. Food and feeding behavior of Steller and California sea 
lions. J. Mammal. 47:195-200. 

FITCH, J. E., AND R. J. LAVENBERG. 

1968. Deep-water teleostean fishes of California. Univ. 
Calif. Press, Berkeley, 155 p. 
Hollander, M., and D. A. Wolfe. 

1973. Nonparametric Statistical Methods. John Wiley and 
Sons, Inc., N.Y., 503 p. 
Jameson, R. J., and K. W. Kenyon. 

1977. Prey of sea lions in the Rogue River, Oregon. J. Mam- 
mal. 58:672. 
Jones, R. E. 

1981. Food habits of smaller marine mammals from northern 
California. Calif. Acad. Sci. Proc. 42:409-433. 
Kashiwada, J., C. W. Recksiek, and K. A. Karpov. 

1979. Beaks of the market squid, Loligo opalescens, as tools 
for predator studies. Calif. Coop. Oceanic Fish. Invest. 
Prog. Rep. 20:65-69. 

LE BOEUF, B. J., AND M. L. BONNELL. 

1980. Pinnipeds of the California Islands: Abundance and 
Distribution. In D. M. Power (editor), The California 
Islands: Proceedings of a Multidisciplinary Symposium, 
p. 475-493. Santa Barbara Museum of Natural History, 
Santa Barbara, Calif. 

Mais. K. F. 

1981. Age-composition changes in the anchovy, Engraulis 
mordax, central population. Calif. Coop. Oceanic Fish. 
Invest. Prog. Rep. 22:82-87. 

Miller, L .K. 

1978. Energetics of the northern fur seal in relation to 
climatic and food resources of the Bering Sea. Mar. Mam- 
mal Comm., Wash., D.C., 32 p. (Available U.S. Dep. Com- 
mer., Natl. Tech. Inf. Serv., as PB-275296.) 

Morejohn, G. V., J. T. Harvey, and L. T Krasnow. 

1978. The importance of Loligo opalescens in the food web of 
marine vertebrates in Monterey Bay, California. In C. W. 
Recksiek and H. W. Frey (editors), Biological, 
oceanographic and acoustic aspects of the market squid, 
Loligo opalescens. Berry. Calif. Dep. Fish Game, Fish 
Bull. 169:67-97. 

Morse, D. H. 

1980. Behavioral mechanisms in ecology. Harvard Univ. 
Press, Cambridge, Mass., 383 p. 

MOYLE, P. B., AND J. J. CECH, Jr. 

1982. Fishes: an introduction .to ichthyology. Prentice-Hall, 
Englewood Cliffs, N.J., 593 p. 

Pacific Fishery Management Council. 

1978. Implementation of northern anchovy fishery manage- 
ment plan, Book 2. Fed. Reg. 43(141):31651-31879. 
Perrin, W. F., R. R. Warner, C. H. Fiscus, and D. B. Holts. 

1973. Stomach contents of porpoise, Stenella spp., and 
yellowfin tuna, Thunnus albacares, in mixed-species 
aggregations. Fish. Bull., U.S. 71:1077-1092. 
Peterson, R. S., and G. A. Bartholomew. 

1967. The natural history and behavior of the California sea 
lion. Am. Soc. Mammal., Spec. Publ. 1, 79 p. 



75 



FISHERY BULLETIN: VOL. 82, NO. 1 



Phillips, -J. B. 

1964. Life history studies on ten species of rockfish (genus 
Sebastodes). Calif. Dep. Fish Game, Fish Bull. 126, 70 
P- 

Pitcher, k. W. 

1980. Stomach contents and feces as indicators of harbor 
seal, Pkoca i itulina, foods in the Gulf of Alaska. Fish. 
Bull., U.S. 78:797-798. 
ROBINS, C. R. (chairman). 

i:i,sn. A list of common and scientific names of fishes from the 
United States and Canada. 4th ed. Am. Fish. Soc, Spec. 
Publ. 12, 174 p. 
Ri iter, ('., R. E. Snodgrass, and E. C. Starrs. 

1904. Report on the sea lion investigation, 1901. In H. M. 



Smith, Report on the inquiry respecting food-fishes and 
the fishing-grounds, p. 116-119. U.S. Comm. Fish Fish., 
Part XXVin, Rep. Comm. 1902. 

SCHEFFER, V. B., AND J. A. NEFF. 

1948. Food of California sea-lions. J. Mammal. 29:67-68. 
Spratt, J. D. 

1975. Growth rate of the northern anchovy, Engraulis mor- 
dax, in southern California waters, calculated from 
otoliths. Calif. Fish Game 61:116-126. 

Young, R. E. 

1972. The systematics and areal distribution of pelagic 
cephalopods from the seas off Southern Califor- 
nia. Smithson. Contrib. Zool. 97,159 p. 



76 



LARVAL DEVELOPMENT OF THE SCUP, STENOTOMUS CHRYSOPS 

(PISCES: SPARIDAE) 1 



Carolyn A. Griswold 2 and Thomas W. McKenney 3 



ABSTRACT 

Larval scup, Stenotomus chrysops (Linnaeus 1766), were reared from eggs hatched in an aquarium. 
Measurements of morphological features for 88 specimens from 2.0 to 16.9 mm SL indicate that growth is 
gradual and continual with no well-defined changes in relative body proportions. Twenty-four myomeres are 
present in larvae, agreeing with published reports of vertebrae numbers in adult scup. Ossification begins 
first in the skulls of 6.1 mm SL larvae, and by 7.0 mm SL the vertebrae, neural spines, and fin rays are begin- 
ning to ossify. Ossification is nearly complete in 18.7 mm SL juveniles. Three preopercular spines are present 
in 4.1 mm SL specimens; the numbers of spines increase and by 16.9 mm SL the preopercular margin is 
serrate. Median fin development occurs at 4.1 mm SL, all fins are present in 8.8 mm SL larvae, and a full com- 
plement of rays are observed by 12.8 mm SL. Larvae are completely scaled by 13.0 mm SL. 



Scup, Stenotomus chrysops (Linnaeus 1766), the only 
common sparid in southern New England waters, is a 
popular sport and commercial fish in spring and sum- 
mer. Their range is from South Carolina to Sable 
Island, Nova Scotia, although they are uncommon 
north of Cape Cod (Breder 1948; Bigelow and 
Schroeder 1953; Leim and Scott 1966). Scup move 
inshore in schools in early April in the Chesapeake 
Bay area and in May north to Cape Cod. Most scup 
spend the summer in bays or within 8-10 km of the 
coast where they spawn from May to August with a 
peak in June in Narragansett Bay (Perlmutter 1939; 
Bigelow and Schroeder 1953; Wheatland 1956; Her- 
man 1963). In late October scup begin to move 
offshore to depths of 40-100 m. Commercial catches 
between January and April indicate that many scup 
winter off Virginia and North Carolina (Neville and 
Talbot 1964; Smith and Norcross 1968). 

Despite the commercial importance and abundance 
of this species, only one description of the eggs and 
larvae exists (Kuntz and Radcliffe 1917). This de- 
scription, which has been paraphrased several times, 
and the accompanying illustrations, which have been 
reprinted several times, provide no information on 
osteological development nor do they present meris- 
tic and morphometric data. Consequently we under- 
took to rear larvae from laboratory-spawned eggs to 
provide specimens for a more complete description 
which would be useful for identification of wild 
larvae. 



'MARMAP Contribution MED/NEFC 81-03. 

2 Northeast Fisheries Center Narragansett Laboratory, National 
Marine Fisheries Serivce, NOAA, Narragansett, RI 02882. 

'Northeast Fisheries Center Sandy Hook Laboratory, National 
Marine Fisheries Service, NOAA, Highlands, NJ 07732. 



METHODS 

Adult fish captured by trawl in Narragansett Bay, 
R.I., were held in a 58 m 1 aquarium until they spawned 
naturally. Fertilized eggs were collected from the 
aquarium with plankton nets and incubated in 40 1 
aquaria at 18° and 21C in 3 l%o salinity. Thepostin- 
cubation series for this study was reared at 18°C. 
After hatching, the larvae and juveniles were fed 
zooplankton and brine shrimp nauplii. Larvae were 
removed regularly for our studies and preserved in 
4% buffered Formalin 4 and Formalin with Ionol 
added as a color preservative. Specimens up to 19.5 
mm standard length (SL) are included in this descrip- 
tion, but scup were reared to >40 mm in some of our 
experiments. Eighty-eight larvae from 2.0 to 16.9 
mm SL were measured with an ocular micrometer. 
The data were pooled for all fish of the same SL, and 
all measurements converted into percentages of SL 
and summarized in Table 1. The following 
measurements were made: 

Total length (TL) : Tip of snout to end of caudal fin or 
finfold. 

Standard length (SL): Tip of snout to end of 
notochord in larvae prior to and during notochord 
flexure; tip of snout to base of hypural plate once it 
is formed. All references to length or size in the text 
refer to SL unless otherwise noted. 

Postanal length: Anus to end of notochord mea- 
sured along midline of body. 



Manuscript accepted June 1983. 

FISHERY BULLETIN: VOL. 82, NO. 1. 1984. 



4 Reference to trade names does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 

77 



FISHERY BULLETIN: VOL. 82. NO. 1 



TABLE 1.— Summary of nine morphological features of specimens of Stenotomus chrysops as shown 

by their percentages of standard length. 









Postanal 


Preanal 


Head 


Snout 


Prepectoral 


Prepelvic 


Eve 


Body 


SL 


TL 




length: 


length: 


length: 


length: 


length: 


length: 


diam. 


depth: 


(mm) 


(mm) 


SL:TL 


SL 


SL 


SL 


SL 


SL 


SL 


SL 


SL 


2.0 


2.1 


95.2 


450 


40.0 


10 


2.5 


— 


— 


7.5 


300 


2 1 


2.3 


91.3 


46.7 


40.5 


11.0 


3 8 


— 


— 


7.7 


29.0 


2 2 


: 4 


91.7 


466 


42.7 


1 2 5 


3.4 


— 


— 


7.6 


26.0 


2 3 


2 & 


92.0 


43.5 


38.0 


10.9 


4.3 


— 


— 


7 6 


25.0 




2 8 


92.9 


42 .3 


36 5 


17.3 


38 


19.2 


— 


7.5 


240 




2.9 


93.1 


41.0 


36.6 


18.2 


5.4 


21.5 


— 


7.7 


210 


2 a 


3.0 


93 3 


40.8 


36.0 


18.6 


6 5 


20.8 


— 


6.2 


24.0 


2 y 


3 l 


93 5 


414 


36.6 


18.5 


4.1 


21.6 


— 


6.9 


20.7 


3 


3 2 


93.8 


42.8 


37 8 


17 8 


50 


21.1 


— 


7.6 


224 


3 2 


3.4 


94 1 


43.8 


37 5 


18 8 


4.7 


21.9 


— 


8 1 


22 6 


:•; 4 


3.6 


94.4 


42.6 


38.2 


19.1 


5 1 


23 5 


— 


8 1 


22.6 


3 5 


3 7 


946 


41 4 


37.1 


18.1 


43 


22.4 


— 


8 8 


28.1 


i 6 


3.8 


94.7 


40 7 


37 


18.5 


56 


21 8 


— 


8.8 


20.6 


3 7 


1 ■< 


94 9 


41 9 


37 2 


18.9 


5 4 


21.6 


— 


8 8 


229 


3 9 


4 i 


95 1 


436 


38.5 


17.9 


5 1 


20.5 


— 


8 1 


284 


4.1 


4 3 


95 3 


439 


39.0 


17 1 


4 9 


22.0 


— 


8 3 


24.4 


4 < 


4.6 


93 5 


46.5 


41.9 


209 


4 7 


22.1 


— 


8 2 


29 5 


4 6 


49 


939 


45.7 


41.3 


21 2 


4.9 


23.9 


— 


8 2 


295 


48 


5 1 


94 1 


45.7 


41.7 


20.8 


6 3 


250 





8.9 


30.0 


4 9 


5.2 


942 


44 9 


40.8 


219 


5.6 


25 5 


— 


8.7 


30.4 


c i 4 


5 6 


964 


51.9 


48 1 


22 2 


56 


25.0 


— 


9.5 


23.8 


5 & 


59 


932 


51 9 


46.7 


23.9 


7 3 


27,6 


— 


10.9 


382 


5 6 


62 


90.3 


48 2 


44 6 


232 


5 4 


268 


— 


9.5 


224 


6 1 


69 


884 


53 


48.1 


25.1 


7.7 


30.1 





9.8 


29 5 


>.4 


7 4 


86.5 


56 3 


500 


25.0 


9 4 


29.7 


— 


10.0 


25.8 


66 


6 6 


868 


54.5 


50.0 


25 8 


7.6 


31.8 


— 


9.8 


30.3 


7 1 


8 


88 8 


54.9 


507 


25.4 


7 


31.0 


— 


99 


21 1 


; q 


9.2 


859 


53.2 


494 


24.7 


7 6 


29 7 


— 


9 6 


25 3 


8 6 


10.3 


82.5 


57.1 


52 9 


253 


7 1 


28 8 


35.3 


9.4 


27.1 


9 2 


10.9 


84.4 


55.4 


51 1 


26.1 


7 6 


29.3 


380 


92 


27 2 


9.9 


11.8 


839 


54 5 


50.5 


242 


9 1 


32.3 


33 3 


9.1 


27.3 


10.0 


12.0 


83 3 


58 5 


52.5 


27 5 


9 


30.5 


37.0 


9 5 


260 


12.0 


14.5 


828 


56.7 


508 


258 


9 2 


30.0 


35.8 


9 2 


26.7 


12.6 


154 


81.8 


55 6 


500 


262 


8.7 


31.0 


34.1 


8.7 


25.4 


13 1 


156 


84.0 


565 


51.9 


28.2 


8 4 


32.1 


35.9 


10.7 


30.5 


135 


16.3 


82 8 


563 


52 6 


21.5 


8.1 


304 


34.1 


89 


296 


14.6 


17.0 


85 9 


54.8 


50.0 


25.3 


8.9 


32.2 


35 6 


9 6 


26.0 


14.9 


17.1 


87.1 


54.4 


48 3 


25.5 


8 1 


30.9 


34.2 


8.7 


26.8 


15.9 


18.8 


84.6 


59.1 


54.7 


25.2 


8.2 


31.4 


384 


7.5 


32.7 


169 


20.6 


82.0 


62.1 


59.2 


30.2 


11.2 


35.5 


42.6 


10.7 


33.7 



'Notochord flexion 



Preanal length: Tip of snout to anus measured along 

midline of body. 
Head length: Tip of snout to posterior margin of otic 

capsules in young larvae; tip of snout to cleithrum 

once it is apparent. 
Eye diameter: Horizontal distance between anterior 

and posterior edges of orbit. 
Snout length: Tip of snout to anterior margin of 

eye. 
Body depth: Vertical height of body at pectoral 

axis. 
Prepectoral length: Tip of snout to axil of pectoral 

fin, or its anlage, measured along midline of body. 
Prepelvic length: Tip of snout to axil of pelvic fins, 

measured along midline of body. 
Meristics: Fin rays and spines were counted as they 

became apparent. Myomeres (total, precaudal, and 

caudal) were counted. Seventeen specimens were 

cleared and stained by Hollister's method (Hollister 

1934) to determine the ossification sequence of 

78 



developing skeletal elements to verify counts of 
bony structures. 



DESCRIPTION 

Eggs 

The scup egg is spherical, buoyant, and transparent. 
The shell is unsculptured and the yolk unsegmented. 
It has one gold-colored oil globule that is posterior in 
the yolk sac and bears black pigment. The yolk is 
about 187( and the oil globule about 21% of the egg 
diameter. The average diameter of the 97 eggs we 
measured was 0.93 mm (range 0.81-1.00 mm). They 
hatched in 70-75 h at 18°C and in 44-54 h at 21°C. 
These measurements and the incubation time are 
similar to those found by others for this species 
(Kuntz and Radcliffe 1917; Perlmutter 1939; 
Bigelow and Schroeder 1953; Wheatland 1956). 



GRISWOLD and McKENNEY: LARVAL DEVELOPMENT OF SCUP 



Larvae 

Newly hatched larvae average 2.0 mm SL. The eyes 
are not pigmented and the mouth is not functional. 
The head is bent slightly over the elliptical yolk sac. 
Yolk sac and oil globule are absorbed and gut dif- 
ferentiation occurs between 48 and 72 h after hatch- 
ing at 18°C. During this period the eyes become 
pigmented, the mouth functional, and the larvae 



begin to feed. Larvae ranging in size from 2.0 to 18.7 
mm are shown in Figure 1. 

Yolk Absorption and Gut 
Differentiation 

At hatching the gut is a tube with a constriction at its 
posterior end that extends to the ventral edge of the 
finfold, but by 48 h (2.7 mm) a foregut and hindgut 




A DAY I 2.0 




B DAY 4 2.8 




C DAY 5 3.0 




F DAY 13 5.7 




G DAY 15 



7.3 




H DAY 17 9.4 




D DAY 6 3.4 




I DAY 21 14.9 





E DAY 9 4.2 J DAY 24 18.7 

FIGURE l.— Development of Stenotomus chrysops. Lengths (SL) are in millimeters. 



79 



FISHERY BULLETIN: VOL. 82, NO. 1 



can be distinguished. The hindgut appears to be 
muscular and remains a tube until between day 7 and 
day 9 (ca. 4.0 mm), when a well-defined stomach 
becomes apparent and the hindgut is relatively 
shorter. 

Total Length and Standard 
Length 

Larval growth appears to be gradual and continuous 
with no well-defined changes in relative body propor- 
tions. Apparent slight changes which are noticeable 
after notochord flexion relate to a change in measure- 
ment from an SL which is actually notochord length 
to one which is a true SL. 



Snout Length 

As with eye diameter, there is considerable varia- 
tion among individuals of the same size. At hatching 
the snout length is 2.5% of SL, but this increases 
gradually to 9.4% of SL at 15.9 mm and 11.2% of SL 
in the juveniles. 

Body Depth 

Body depth ranges from 25 to 30% in newly hatched 
larvae, but once the yolk is absorbed it decreases to 
between 21 and 24.4% of SL (with one exception) up 
to 3.9 mm, and then increases to 22.4 to 33.7% of 
SL. 



Postanal Length 

Postanal length remains about 45% from 2.0 to 4.9 
mm SL, when notochord flexion is occurring. A 
gradual increase to 62.1% in juveniles longer than 
16.9 mm SL is concurrent with development of ver- 
tebrae and overall growth of the larvae. 

Preanal Length 

Preanal length increases relative to SL from 36.0 f /f 
at 2.0 mm to 41.9% at 4.9 mm to >59% for juveniles 
longer than 16.9 mm. The lengthening of the body 
cavity during growth accounts for the increase in pre- 
anal length. 

Head Length 

Head length increases relative to SL from an 
average of 11.1% (10-12.5%) in newly hatched larvae 
(2.0-2.3 mm) to 17.3% in 2.6 mm larvae, then 
gradually increases to 30.2% in the largest juvenile 
specimen. In very young larvae the otic capsules are 
the reference structure for head measurements. 
However, once the cleithrum develops it is used as 
the reference structure for subsequent head 
measurements and an increase in head length per- 
centage is observed. 

Eye Diameter 

The ratio of eye diameter to SL in our series is 6.2- 
10.9% of SL. It averages 7.5% SL in 2.0-3.0 mm lar- 
vae, and 8.5% SL (range 8.1-8.9%) in 3.2-4.9 mm 
larvae. In larvae >4.9 mm the average is 9.5% of SL 
(range 7.5-10.9%). Variation in individuals of the 
same size is considerable. 



Prepectoral Length 

Anlagen are present at hatching. Initially prepec- 
toral length is about 19.2% of SL. This increases 
gradually during the larval and postlarval period to 
35.5% of SL in the juvenile. 

Prepelvic Length 

Pelvic fin buds do not appear until the larvae are 
8.0-8.5 mm long. Prepelvic to SL ratio is about 35.6% 
(range 33.3-38.4%) for larvae from 8.5 to 15.9 mm 
SL, but increases to 42.6% of SL in the juvenile. 

MERISTICS 



Scup, being typical of most perciform fish, have 24 
myomeres. This agrees with Miller and Jorgenson's 
(1973) vertebrae numbers for adult fish. 

FIN DEVELOPMENT 

At hatching a finfold extends from the top of the 
head to the visceral sac interrupted only by the anus. 
There are no fin rays. A remnant of this persists be- 
tween the anus and the first anal fin ray in a larva 9.1 
mm long. Fin sequence development is given in Table 
2. 

Anlagen of the pectoral fins are present in most, if 
not all, hatchlings. These are low buds at first, but by 
the time the larvae are about 2.5 mm these fins have 
bases and blades. By removing pectoral fins from one 
side of some of our larvae and flattening them out, we 
could see 13 rays in one 4.9 mm larva and 10 rays in a 
5.7 mm larva. Aside from these two, however, we 
could not see pectoral fin rays, even on cleared and 



80 



GRISWOLD and McKENNEY: LARVAL DEVELOPMENT OK SCUP 



TABLE 2. — Summary of fin development sequence in larvae of 
Stenotomus chrysops. 





Notochord or standard length (mm) 






Buds 


Rays 


Full 


Number of rays 




first 


first 


complement 


in fully 


Fin 


appear 


appear 


of rays 


developed fin 


Dorsal 




5.5-6.0 


10.8 


XII + 12 


Caudal 






10.4-108 


32-34 


Principal 




43 


5.3 


Dorsal 9 
Ventral 8 


Secondary 




53 




Dorsal 7-8 
Ventral 8-9 


Anal 




5.5-6.0 


108 


III + 1 1 


Pelvic 


5 7 


8 8-10.0 


12.8-13.2 


I + 5 


Pectoral 


2 3 


2 9-3.0 


10.4-10.8 


16 



stained specimens until the larvae were about 8.0 
mm, when the larvae had nearly the full complement 
of 15-16 pectoral rays. 

An anlage of the caudal base can be seen in larvae as 
small as 3.4 mm. Some of the principal caudal rays 
are detectable in the finfold of larvae as small as 4.3 
mm, and are the first rays of any fins to appear. 
Notochord flexion in our series begins at 4.7 mm. By 
5.3 mm all of the principal caudal rays are present as 
are some of the secondary ones. Flexion is complete 
at about 8.0 mm and the caudal fin begins to fork at 
about 10 mm. Full complements of caudal rays 
(9-10+9+8+8-10) are present in larvae 14 mm or 
longer. Secondary rays develop in a posterior to 
anterior direction. 

The soft-ray parts of the median fins first develop 
beginning at 5.3 mm in our series. Anal and dorsal 
rays develop together. In both fins, the central soft 
rays develop first. The development of anterior and 
posterior rays follows rapidly so that when the larvae 
in our series are >6.0 mm, full complements of 11-12 
soft rays are present in these fins. Development of 
the spiny rays in these fins is from posterior to 
anterior and follows the soft-ray development. An 
exception is the posteriormost spiny rays in both fins 
that appear first as soft rays. 

The last fins to appear are the pelvics. Anlagen are 
first seen in our series in some larvae at 5.7 mm. 
Other larvae are >7.0 mm long before these anlagen 
are visible. Development thereafter is from the distal 
edge medially. Full arrays of 1 spine and 5 soft rays 
are present in larvae 8.5 mm or longer. 

Adult scup have six pairs of branchiostegals. Five 
pairs of these are present in a 4.2 mm larva of our 
series. They were visible in all of our series that were 
5.0 mm or larger. The sixth pair, the median one, is 
not visible in some of our larvae even at 16.5 mm. The 
first five pairs usually appear simultaneously, but the 
sixth appears later. 



PIGMENT 

Although scup have chromatophores other than 
melanophores, these faded rapidly after preserva- 
tion in Formalin. This account is confined to 
melanophore pigmentation (Fig. 1). Pigmentation 
other than that by melanophores is extensive on 
embryos and early larvae and is described and illus- 
trated by Kuntz and Radcliffe (1917). 

Head Region 

Newly hatched scup have unpigmented eyes. Two 
rows of stellate melanophores, one on either side, 
extend from the snout back over the eyes and con- 
tinue as part of a lateral series on the trunk. At a 
length of about 2.5 mm there is a hiatus in this series 
that extends from mideye level to over the visceral 
sac. 

At 4.0-5.0 mm length, the pattern that will 
culminate in that of the juvenile has begun to appear. 
There are few, usually no, melanophores on the dor- 
sal and lateral parts of the head anterior to the middle 
of the eyes. However, there are several prominent 
melanophores on the posterior midbrain and several 
on the hindbrain. Ventrally there is usually no pig- 
ment on the head. A few of our specimens have one or 
two small melanophores. 

Development beyond this stage consists of a 
gradual increase of pigment on the dorsal and dor- 
solateral parts of the head. Most of it occurs above 
mideye level. A few melanophores appear on the 
snout and below the eye. There is a prominent 
melanophore, sometimes accompanied by one or two 
small ones as well, at the articulation of the lower jaw 
with the quadrate bone. 

Between the head and the trunk pigmentation in the 
occipital region there is a gap in the dorsal pigment 
with relatively little pigment in it . This gap is part of 
the barred pattern of the juvenile. 

Trunk and Tail Region 

At hatching there are two dorsal rows of stellate 
melanophores extending from the head to beyond 
myomere 20. They appear to be between myomeres 
on the myosepta. Occasionally these rows are 
interrupted by "missing" melanophores. When this 
is so, the melanophore is usally lower down on the 
side of the body. A few, usually three or four, 
melanophores occur at various places on the anterior 
part of the yolk sac, and there are one or two on the oil 
globule. Some specimens have widely spaced 
melanophores on the ventral margin of the tail. 



81 



FISHERY BULLETIN: VOL. 82, NO. 1 



This pattern persists until the larvae are about 2.5 
mm long with a gradual increase in the number of 
melanophores along the ventral margin of the tail. 
The melanophores on the yolk sac and oil globule dis- 
appear with the exception of one or two on the mid- 
ventral line of the anterior part of the yolk sac. 

In larvae 4.0-5.0 mm long, most of the pigment is on 
the peritoneum dorsal to the viscera and along the 
midventral line. Anteriorly there is a melanophore at 
the cleithral symphysis and posterior to it a large one 
midventrally on the anterior belly and a smaller one 
on the posterior belly. There is a prominent 
melanophore on the hindgut just anterior to the anus. 
Posterior to this there is a melanophore on most of 
the anal pterygiophores; this pattern is continued 
externally on the ventral myosepta. Dorsally there 
are several melanophores on the posterior 
pterygiophores of the dorsal fin. There are usually a 
few scattered spots on the finfold and, on some 
specimens, a few on the sides. 

The peritoneal pigment becomes denser and more 
prominent in 9-10 mm larvae, however, it is often 
obscured because of the opacity of the thickening 
body musculature in preserved specimens. The 
hindgut is nearly covered by large melanophores. 
The melanophores on the midventral line are still 
present, usually accompanied by two or three smaller 
ones. The melanophore just posterior to the anus is 
still present, but less prominent. 

The trunk and tail pigment is more extensive at 9- 1 
mm. There are many more pigment spots along the 
sides, but these are still widely spaced, especially 
anteriorly. There is pigment along the bases of the 
dorsal and anal fins that continues posterior to them 
to the procurrent caudal rays. A line of melanophores 
runs dorsoventrally at about the juncture of the 
caudal fin rays and caudal bones. Internally there are 
melanophores near the bases of the haemal and 
neural arches. These become increasingly obscure as 
the body musculature thickens. 

Pigment development beyond this size is charac- 
terized by the development of the barred pattern of 
the juvenile accompanied by a general increase in 
pigment everywhere, especially above the mid- 
lateral line. 

OSSIFICATION 

A total of 17 fish were stained with Alizarin Red to 
determine where ossification began and the 
sequence in which the bones ossified. A summary of 
osteological development is presented in Table 3. 

There is no dye uptake in 5.2 or 6.0 mm larvae, 
although the cartilaginous skeleton is easily dis- 



tinguished. Cartilaginous hypural plates are present 
in larvae undergoing notochord flexure (5.4-5.6 mm). 
The first ossification occurs in skulls of 6.1 mm lar- 
vae. The premaxillary, maxillary, dentary, articular, 
and quadrate bones associated with the jaws, the pre- 
operculum, hyomandibular, branchiostegal rays, and 
cleithrum showed varying degrees of dye uptake. 
There is no ossification posterior to the cleithrum. 

By 7.0 mm more ossification of the skull occurs, 
notably the pterygoid, metapterygoid, opercular 
series, supracleithrum, and frontal bones. The cir- 
cumorbitals, as well as the parasphenoid and the 
scapula, show the beginning of dye uptake. Ossifica- 
tion has begun in the first 10 vertebrae, the neural 
spines of the first 4 vertebrae, and the pectoral and 
caudal fin rays. 

In 9.3 mm specimens the skull is further developed; 
teeth are visible and the lachrymal, dermethmoid, 
nasal, prefrontal, and urohyal bones show varying 
degrees of ossification. The postcleithrum is well 
developed and the radials, scapula, and coracoid are 
ossifying. The entire vertebral column is ossified with 
the exception of the ultral centrum and penultimate 
vertebrae. Both haemal and neural spines are 
ossified. The pleural ribs, and the dorsal and anal fin 
rays are beginning to ossify; hypural plates and 
caudal fin rays are partially ossified. 

By 10.8 mm the distal vertebrae (caudal complex) 
have ossified and scales are present. The pelvic fin 
supports and rays show some dye uptake. Pterygio- 
phores are present as cartilage. All the dorsal and 
anal fin rays and spines have ossified. 

Skull development and ossification of most of the 
bones of the skull is complete by 14.5 mm. The 
radials and scapula which were just beginning to 
ossify in the 10.8 mm fish are now complete. Pelvic 
fin supports are complete. The pleural ribs are 



TABLE 3. — Summary of osteological development in laboratory 
reared larvae of Stenotomus chryxops. 







Notochord 


or standard length (i 


nm) 






First 


First evidence 






appearance 


of ossification 


All 




in 


cartilage 


(stain uptake) 


ossifying 


Cartilaginous 










skeleton 




5.2 






Hypural plates 




5.4 


93 


14.5 


Vertebrae and 










neural spines 






7.0 


10.8 


Pectoral girdle 




7,0 


108 


14 5 


Pectoral and caudal 










fin rays 






7.0 


9.3 


Haemal spines 






93 


93 


Pleural ribs 






93 


14.5 


Dorsal and anal 










fin rays 






93 


10.0 


Caudal complex 






108 


10.8 


Pelvic fin supports 






108 


14.5 


Pelvic girdle 






108 


18.7 



82 



GRISWOLD and McKENNEY: LARVAL DEVELOPMENT OF SCUP 



stained as are the pterygiophores. The hypural plates 
are all present and completely ossified; a few dorsal 
plates are still partially cartilaginous. 

By the time scup are 18-19 mm long, they are 
juveniles. Ossification continues in the skull with the 
bones being joined at suture points; the pelvic girdle 
is complete. The pterygiophores and ribs have com- 
pleted ossification. 

PREOPERCULAR SPINES 

Figure 2 shows the development of preopercular 
spines. We saw them first on a 4.1 mm specimen, 
which has three spines on the preopercular margin. 
Thereafter their number increases until there are so 
many on a 16.9 mm specimen that the margin is 
serrate. Specimens larger than about 25 mm have 
nearly smooth preopercular margins. 






and Schroeder 1928) placed S. aculeatus in the 
synonomy of S. chrysops; Robins et al. (1980) did not 
list S. aculeatus. Dahlberg (1975) mentioned young 
stages of S. chrysops with crossbars (i.e., juveniles) in 
his account of Georgia coastal fishes, although it is 
not clear whether he had taken such specimens in 
his collections. 

This issue is further complicated by lack of informa- 
tion about the northern extent of spawning of other 
sparid fishes. If their spawning ranges overlap with 
that of S. chrysops, then the younger larvae of some 
species will probably be confused with scup larvae, at 
least until the dorsal, anal, and pectoral fin rays can 
be counted. Except for the reference to juvenile scup 
on the Georgia coast by Dahlberg, the authors can 
find no references to such an overlap. 

We have seen larval scup in collections misiden- 
tified as Scomber scombrus, the Atlantic mackerel, 
from which they can be separated at all stages by the 
numbers of myomeres (24 in scup and 31 in mack- 
erel). We have also seen larval gerreid fishes misiden- 
tified as scup. Among other characters, scup differ 
from gerreid fishes in lacking the long premaxillary 
spines that extend up between the eyes in gerreids. 





Figure 2.— Development of the preopercular spines of Stenotomus 
chrysops. Standard lengths in millimeters of the specimens are A) 
4.1, B) 5.6, C) 8.3, D) 9.8, and E) 16.9. 



ACKNOWLEDGMENTS 

We are grateful to John Colton, Bernard Skud, 
Donna Busch, Wallace Smith, and Michael Fahay for 
their reviews of the manuscript. We thank Jennie 
Dunnington and Maureen Montone for typing and 
retyping the manuscript. We are particularly indebt- 
ed to Lianne Armstrong who prepared Figure 1 and 
Alyce Wells for help with Figure 2. 



SCALES 

The first scales are seen between 9.9 and 10.8 mm. 
At 12.3-13.0 mm the larvae are completely scaled. 

COMPARISONS 

The geographical extent of spawning of S. chrysops 
is not known. The authors can find no record of it 
spawning south of the New York Bight. At least one 
other species of Stenotomus, S. caprinus (Bean 
1882), occurs in the western North Atlantic. Accord- 
ing to Geohagen and Chittenden (1982), the major 
population of this species is in the northern Gulf of 
Mexico and it occurs only rarely along the east coast 
to North Carolina. A third nominal species, S. aculeatus 
(Valenciennes 1830), said to replace S. chrysops 
south of Cape Hatteras, is of doubtful validity. 
Birdsong and Musik (in 1977 reprint of Hildebrand 



LITERATURE CITED 

BlGELOW, H. B., AND W. C. SCHROEDER. 

1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., 
Fish. Bull. 53, 577 p. 
Bremer, C. ML, Jr 

1948. Field book of marine fishes of the Atlantic coast from 
Labrador to Texas. G. P. Putnam's Sons, N.Y., 349 p. 
Dahlberg, M. D. 

1975. Guide to coastal fishes of Georgia and nearby 
states. Univ. Georgia Press, Athens, 186 p. 
Geohagen, P., and M. E. Chittenden, Jr. 

1982. Reproduction, movements, and population dynamics 
of the longspine progy, Stenotomus caprinus. Fish. Bull., 
U.S. 80:523-540. 
Herman, S. S. 

1963. Planktonic fish eggs and larvae of Narragansett 
Bay. Limnol. Oceanogr. 8:103-109. 
Hildebrand, S. F., and W. C. Schroeder. 

1928. Fishes of Chesapeake Bay. T. F. H. Publications, Inc., 
Neptune, N.J., 388 p. (1 972 reprint with comments by R. 
S. Birdson and J. S. Musik.] 



S3 



FISHERY BULLETIN: VOL. 82, NO. 1 



HOLLISTER, G. 

1934. Clearing and dyeing fish for bone study. Zoologica 

(N.Y.) 12:89-101. 
KlNTZ, A.. AND L. RADCLIFFE. 

1917. Notes on the embryology and larval development of 

twelve teleostean fishes. Bull. U.S. Bur. Fish. 35:89- 

134. 

Leim, a. H.. am) W. B. Scott. 

1966. Fishes of the Atlantic coast of Canada. Fish. Res. 
Hoard Can. Bull. 155.485 p. 

Miller. G. L., and s C. Jorgenson. 

1973. Meristic characters of some marine fishes of the west- 
ern Atlantic Ocean. Fish. Bull., U.S. 71:301-312. 
Neville, W. C, and G. B. Talbot. 

1964. The fishery for scup with special reference to fluc- 
tuations in yield and their causes. LIS. Fish Wildl. Serv., 
Spec. Sci. Rep. Fish. 459, 61 p. 



Perlmitter, A. 

1939. A biological survey of the salt waters of Long Island. 
Section I. An ecological survey of young fish and eggs iden- 
tified from tow-net collections. Suppl. 28th Annu. Rep. 
N.Y. Cons. Dep., Par! 11:1 1-71. 
Robins, C. R., R. M. Bailey, C. E. Bond, J. R. Brooker, E. A. 
Lachner, R. N. Lea, and W. B. Scott. 

1980. A list of common and scientific names of fishes from the 
United States and Canada. 4th ed. Am. Fish. Soc. Spec. 
Publ. 12, 174 p. 
Smith. W. G., and J. J. Norcross. 

1968. The status of the scup (Stenotomus chrysops) in winter 
trawl fishery. Chesapeake Sci. 9:207-216. 
Wheatland, S. 

1956. Oceanography of Long Island Sound, 1952-1954. VII. 
Pelagic fish eggs and larvae. Bull. Bingham Oceanogr. 
Collect., Yale Univ. 15:234-314. 



84 



DESCRIPTION OF EGGS, LARVAE, AND EARLY JUVENILES OF 

GULF MENHADEN, BREVOORTIA PATRONUS, AND COMPARISONS 

WITH ATLANTIC MENHADEN, B. TYRANNUS, AND 

YELLOWFIN MENHADEN, B. SMITHI ] 



William F. Hettler 2 



ABSTRACT 

Morphometric, merist ic, and pigmentation descriptions of laboratory-reared gulf menhaden, Brevoortia pa- 
tronus, and Atlantic menhaden.fi. tyrannus, indicate that larvae of these species can be distinguished from 
each other by the number of myomeres and vertebrae; that Atlantic menhaden can be distinguished from 
yellowfin menhaden, B. smithi, by the number of myomeres and vertebrae, by pigmentation, and by 
morphometries; and that gulf menhaden can be separated from yellowfin menhaden by pigmentation and 
morphometries. Unlike yellowfin menhaden, gulf and Atlantic menhaden lacked paired melanophores along 
the dorsal midline forward of the dorsal fin and along the ventral midline between the paired fins. Compared 
with yellowfin menhaden larvae of equal lengths, gulf menhaden had less body depth, shorter heads and 
snouts, smaller eyes, and longer prepelvic and predorsal distances. Gulf menhaden eggs averaged 1.29 mm in 
total diameter, 0.95 mm in yolk diameter, and 0.20 mm in oil droplet diameter. Twelve-hour-old larvae had a 
snout-notochord tip length of 3.3 mm. Their growth rate averaged 0.30 mm/day through 90 days of rearing at 
20°C. On specimens 6-17 mm the mean number of myomeres was 44.6; on specimens >15 mm the mean 
number of vertebrae was 45.3. Postdorsal-preanal myomeres decreased from 5.3 to 1.8 as the dorsal fin grew 
and the gut shortened during development. Transformation from larva to juvenile in laboratory-reared gulf 
menhaden was completed at a smaller size than reported for field-caught fish (25 vs. 28 mm SL). 



Eggs and larvae of gulf menhaden, Brevoortia pa- 
tronus Goode, have not been described, even though 
this species is the most economically important 
clupeid in the United States. The gulf menhaden 
purse seine fishery landed an average of 660,368 t 
annually from 1977 to 1981, making it the largest 
fishery in the United States (U.S. National Marine 
Fisheries Service 1982). Gulf menhaden, one of three 
species of Brevoortia in the Gulf of Mexico, are found 
from Florida Bay to the Gulf of Campeche, Mexico. 
They spawn in the northern gulf at least as far 
offshore as the 80 m isobath between mid-October 
and late March, with a peak in December (Christmas 
and Waller 1975'); juveniles are estuarine depen- 
dent. Yellowfin menhaden, B. smithi, and finescale 
menhaden, B. gunteri, co-occur with gulf menhaden, 
but contribute <1% to the landings. The Atlantic 
menhaden, B . tyrannus , which supports a large purse 
seine fishery along the U.S. Atlantic coast, is a large- 



■Contribution No. 83-33B of the Southeast Fisheries Center, 
Beaufort Laboratory, National Marine Fisheries Service, NOAA. 

-Southeast Fisheries Center Beaufort Laboratory, National 
Marine Fisheries Service, NOAA, Beaufort, NC 28516-9722. 

'Christmas, J. Y., and R.S. Waller. 1975. Location and time of 
menhaden spawning in the Gulf of Mexico. Unpubl. manuscr., 20 
p. Gulf Coast Research Laboratory, Ocean Springs, MS 39564. 



Manuscript accepted July 1983. 

FISHERY BULLETIN: VOL. 82, NO. 1, 1984. 



scaled cognate of the gulf menhaden, but does not 
occur in the Gulf of Mexico (Hildebrand 1963). 
Distribution of yellowfin menhaden is contin- 
uous around Florida to as far north as North Caro- 
lina. 

Menhaden larvae superfically resemble the larvae 
of other clupeids with which they co-occur and can be 
distinguished from them (Houde and Fore 1973; 
Houde and Swanson 1975), but current descriptions 
(Suttkus 1956; Houde and Fore 1973; Houde and 
Swanson 1975; Jones et al. 1978) are not adequate to 
separate sympatric Brevoortia larvae. Eggs, larvae, 
and juveniles of yellowfin menhaden have been de- 
scribed (Houde and Swanson 1975), whereas the 
early development of finescale menhaden has not. 
Gulf and yellowfin menhaden hybrids in the eastern 
Gulf of Mexico (Hettler 1968; Turner 1969; 
Dahlberg 1970) further complicate separation by 
species. Although gulf and Atlantic menhaden larvae 
cannot be confused in ichthyoplankton collections 
because of their allopatric separation by the Florida 
Peninsula, Atlantic and yellowfin menhaden larvae 
may be confused in collections from the east coast of 
Florida, where both species are known to spawn dur- 
ing the winter (Dahlberg 1970). 

In this paper, I describe the eggs, larvae, and early 

85 



FISHERY BULLETIN: VOL. 82. NO. 1 



juveniles of gulf menhaden spawned and reared in 
the laboratory using morphometries, meristics, and 
pigmentation features, and I compare gulf menhaden 
larvae with yellowfin menhaden larvae described by 
Houde and Swanson (1975). Morphometric and 
meristic data on laboratory-spawned and reared 
Atlantic menhaden are also presented to supplement 
the composite description of this species by -Jones et 
al. (1978) and to aid in the separation of Atlantic 
menhaden and yellowfin menhaden larvae. Charac- 
ters for separating Brevoortia from other clupeids are 
reviewed. 



cleithra, exclusive of the finfold. 
Dorsal and anal fin base lengths — distance from 

anterior to posterior edges of fin base; in larvae with 

incomplete fins, distance from origin of first ray to 

the insertion of the last ray. 
Head length— tip of snout to posterior margin of otic 

capsules in yolk-sac larvae; tip of snout to opercular 

margin in older larvae and juveniles. 
Snout length — tip of snout to anterior margin of 

eye. 
Eye diameter — horizontal distance between anterior 

and posterior edges of fleshy orbit. 



METHODS 

Gulf menhaden were collected as mature adults in 
September 1981 near Gulf Breeze, Fla., tranported 
to the Beaufort Laboratory, and induced to spawn 
with human chorionic gonadotropin (HCG) and carp 
pituitary (Hettler 1983). Spawnings that occurred in 
November 1981 and February 1982 provided a 
developmental series of eggs, larvae, and juveniles 
up to 90 d old, reared at a temperature of 20° ± 2°C 
and a salinity of 30%o. One hundred eggs, preserved 
during the early embryo stage, and 100 live eggs 
were measured. 

Atlantic menhaden were captured as juveniles in 
September 1978 near Beaufort, N.C., and reared to 
sexual maturity in the laboratory for 19 mo. They 
were induced to spawn in April 1980, and the larvae 
were reared at temperatures that began at 15°C and 
increased to 25°C during development (Hettler 
1981). This spawning resulted in a developmental 
series of larvae and juveniles up to 130 d old. 

All specimens were preserved in 2% buffered for- 
maldehyde in seawater before being measured. The 
following morphometic measurements were taken 
with an ocular micrometer in a dissecting microscope 
on 123 gulf menhaden and 196 Atlantic menha- 
den. 

Standard length (SL) — tip of snout to tip of 
notochord before and during notchord flexion; in 
postflexion larvae, tip of snout to posterior margin 
of hypural bones. All references to length in this 
paper are standard length unless otherwise 
stated. 

Preanus length — tip of snout to posterior end of 
anus, measured along midline. 

Predorsal length — tip of snout to anterior edge of 
dorsal fin base, measured along midline. 

Prepelvic length — tip of snout to anterior insertion of 
pelvic fin, measured along midline. 

Body depth — vertical depth at symphysis of the 



Myomeres were counted on semidry specimens (not 
completely immersed) up to 1 7 mm with transmitted 
unpolarized light by adjusting the microscope mirror 
to give maximum contrast between myosepta and 
myomeres. Myomeres were classified as follows: 

Total myomeres — all myomeres between the most 
anterior myoseptum and the most posterior 
myoseptum. 

Preanal myomeres — number anterior to the myo- 
mere in which the anterior ray of the anal fin is 
inserted or to the myomere in contact with the 
downward curve of the dorsal margin of the anus in 
larvae without anal fin rays. 

Postanal myomeres — number posterior to the 
anterior insertion of the anal fin. 

Predorsal myomeres — number anterior to the 
myomere containing the origin of the first dorsal 
fin ray. 

Postdorsal-preanal myomeres — number between 
the myomere connected to the last dorsal fin ray and 
the most posterior preanal myomere. 

Following morphometric measurements on all 
specimens and myomere counts on specimens with 
visible myomeres, the pigment pattern was recorded 
and specimens of gulf menhaden were illustrated 
with a camera lucida. Atlantic menhaden were not 
illustrated as the figures in Jones et al. (1978) are 
adequate. 

Specimens were then used for counts of fin rays, 
pterygiophores, predorsal bones, vertebrae, and 
scutes. Specimens were transferred to 95 /f ethanol, 
stained with alcian blue for cartilage, cleared with 
trypsin, stained with alizarin red S for bone, and 
stored in 100% glycerin 4 . 



"Taylor, W. R., and G. C. Van Dyke. 1978. Staining and clearing 
small vertebrates for hone and cartilage study. Unpubl. manuscr., 
19 p. National Museum of Natural History, Washington, DC 
20560. 



86 



HETTLER: DESCRIPTION OF Ol'LF MENHADEN 



DESCRIPTION 

Embryos 

Gulf menhaden eggs were spherical, and had an 
unsculptured chorion, a faintly segmented yolk, and a 
single oil droplet. Living eggs were buoyant in 
salinities >26%o. Twenty-seven percent had both an 
outer and inner chorionic membrane. This has not 
been reported in wild-caught Brevoortia eggs. This 
inner chorion was not an artifact of preservation, 
since live eggs also contained a double chorion, but 
may have been a result of induced ovulation by HCG 
and carp pituitary. Dimensions of preserved and live 
eggs were the same as maximum sizes given by 
Houde and Fore (1973) for gulf menhaden eggs taken 
in plankton collections (Table 1). At its widest point 
the perivitelline space was 24-28% of the egg 
diameter. Eggs produced during December 1982 by 
another spawning group of gulf menhaden were 
smaller than gulf menhaden eggs produced the year 



TABLEl. Mean diameter (mm) of j-cull menhaden, Hrevnnrtia pa- 
tronus, eggs. Numbers in parentheses are equal to one standard 
de\ iation oi the mean. 



Eggs 



Total 
diameter 



Inner chorion 

diameter (if 

present) 



Yolk 

diameter 

(along axis) 



Oil droplet 
diameter 



Preserved 100 1.29(0.04) 
Live 100 1.30(0.05) 



1.23 (0.04) 
1.25 (0.03) 



0.95 10.05) 
97 (0.04) 



0.20(0.02) 
0.19 (0.01) 



before; total diameter was 1.18-1.22 mm; the yolk 
diameter was 0.66-0.79 mm; the oil droplet was 0.16 
mm. The adults producing these eggs were smaller 
(17.8 cm mean length, 90 g mean weight) than the 
spawners that produced the larger eggs (20 cm, 135 
g) (Table 1). Small adult size may be responsible for 
the small eggs as well as the reduced fecundity. Only 
a few hundred fertilized eggs were collected from the 
December 1982 group of 20 fish. 

Advanced embryos had 30-40 small melanophores 
on each side along the dorsal surface from the pos- 
terior end of the head to the notochord tip (Fig. 1 A). 




1 mm 



V^frA 



FIGURE 1.— Early stages oiBrevoortia patronus. A. Embryo 40 h after fertilization. B. 2.6 mm larva, 5 min after hatching, e. 3.5 mm larva, 1 d 

after hatching. D. 3.9 mm larva, 2 d after hatching. 



87 



FISHERY BULLETIN: VOL. 82, NO. 1 



About 15-20 myomeres were visible in the caudal 
region. The yolk was faintly segmented into irregular 
globules. E ggs hatched in 40-42 h at a water tempera- 
ture of 19°-20°C. 

Atlantic menhaden eggs spawned in the laboratory 
were larger than gulf menhaden eggs in total 
diameter (1.54- 1.64 mm) but similar in yolk diameter 
(0.82-0.95 mm) and oil droplet diameter (0.20- 
0.23). 



Larvae 



Growth 



Gulf menhaden larvae were 2.6-3.0 mm SL 
immediately after hatching (Fig. IB), but within 6 h 
had a mean length of 3.3 mm. The yolk and oil droplet 
were absorbed, the eyes were pigmented, and the 
mouth was functional at a length of 4.5 mm, 4 d after 
hatching. The growth rate of larvae at 20° ± 2°C 
averaged 0.30 ± 0.03 mm/d through 90 d of rearing 
(Fig. 2). Yellowfin menhaden reared for 32 d at 20°C 
grew 0.36 mm/d (Hettler 1970). Yellowfin menhaden 
reared at 26°C grew 0.45 mm/d until the 20th day 
(Houde and Swanson 1975). 

Body Proportions 

For 123 gulf menhaden, 3.1-34.9 mm, body depth, 
head length, prepelvic length, dorsal fin base length, 
anal fin base length, snout length, and eye diameter 
all increased relative to standard length as larvae 
grew, while preanus length and predorsal length de- 



36 r 



32 



28 

e 

I 
| 20 

LLI 

_l 

D 16 
DC 
< 
D 

I 12 

co 
8 



./: 



10 20 30 40 50 60 70 80 90 
DAYS AFTER HATCHING 

FlOlRE 2. — Growth of laboratory-reared larvae of Hrcvaartki pa- 
t ran us. Lines connect means of each age group. 

creased (Table 2). The decrease in predorsal length 
resulted from the forward movement of the dorsal fin, 
and the decrease in preanus length reflected the 
transformation from an elongate clupeiform larva 
shape to the laterally flattened fusiform shape of the 
juvenile. Transformation from the larval to the 
juvenile form in gulf menhaden began at about 19 
mm (Fig. 3C) and was completed at about 25 mm. 
Atlantic menhaden larvae completed transformation 
at about 27 mm. 



TABLE 2. — Proportions of head and body parts of gulf menhaden, Brevoortia patronus, expressed as a percent of stan- 
dard length. Characters were not developed at lengths marked with a dash. 



Length class 


Number of 


Preanus 


Predorsal 


Prepelvic 


Body 


Dorsal fin 


Anal fin 


Head 


Snout 


Eye 


(mm, SL) 


specimens 


length 


length 


length 


depth 


base length 


base length 


length 


length 


diameter 


30-3.9 


( 


840 


_ 


— 











14 1 


2 3 


5,2 


4.0-4.9 


19 


8<> 2 


— 


— 


9.7 


— 


— 


13,5 


1.7 


5,4 


5.0-5.9 


12 


81.4 


— 


— 


9 


6 


— 


— 


15.8 


3.0 


5,2 


6.0-6.9 


B 


82.4 


69.3 


— 


8 


4 


4 4 


— 


155 


3.1 


5,0 


7.0-7.9 


7 


83 


70.2 


— 


8 


2 


5 


— 


15.4 


2 9 


4,9 


8.0-8.9 


4 


832 


67 6 


— 


7 


9 


8 2 


_ 


15 5 


3 1 


4 8 


9.0-9.9 


5 


839 


658 


— 


8 


3 


10.0 


3 8 


16.2 


3 6 


50 


10.0-10.9 


6 


85 5 


656 


— 


8 


3 


1 15 


4 3 


169 


3 7 


5.0 


11.0-11.9 


1 


85 5 


652 


— 


8 


9 


13 2 


5.6 


17 7 


3 9 


5 2 


120-12 9 


2 


83 .1 


63.0 


— 


8 6 


13 3 


6.0 


16.9 


3,6 


4 8 


130-13.9 


3 


84 2 


62 8 


— 


9,9 


15 1 


68 


17.8 


3.7 


4.9 


14.0-14 9 


i 


81.0 


620 


41.5 


100 


14 5 


7 5 


17 


3 5 


50 


15,0-15.9 


i 


82.2 


61 2 


— 


10 7 


15 4 


7,5 


18.2 


3.7 


5.1 


160-169 


4 


79.8 


60.8 


44 2 


108 


15 3 


94 


18.9 


40 


5 5 


17 0-17 9 


3 


79.0 


61 


44 3 


12 4 


14 8 


11.0 


19.4 


4 1 


5.5 


180-189 


2 


760 


570 


47 6 


18.1 


16,8 


12,3 


24 1 


5.0 


7 3 


19 0-19 9 


4 


762 


56.4 


469 


17 8 


16,6 


12 8 


240 


5.1 


6 8 


200-21 9 


B 


71 4 


51.8 


50.0 


25.8 


186 


16,0 


28 1 


60 


8.1 


22.0-23.9 


8 


70.2 


48.6 


49.2 


280 


18 4 


15.7 


28 9 


6,1 


84 


240-25.9 


6 


70 7 


47,9 


49,6 


29,1 


193 


16,0 


29 3 


6.6 


8 5 


26.0-279 


3 


70.6 


44,7 


50.0 


31.6 


19 2 


17 


29 7 


6 9 


8 6 


280-29.9 


1 


70.2 


43.5 


49.4 


30 1 


20 1 


16,4 


27 7 


64 


7.7 


30.0-34 9 


7 


72 7 


47.7 


51 3 


36.0 


19 4 


17,1 


31 5 


7.2 


7 8 



88 



HETTLER: DESCRIPTION OF GULF MENHADEN 






2mm 

FIGURE 3.— Larval Brevoortia patronus: (A) 13.0 mm (28 d after hatching). (B) 16.5 mm (44 d after hatching). (C) 18.9 mm (53 d after 

hatching). 



Gulf menhaden larvae and Atlantic menhaden lar- 
vae could not be separated morphometrically (Table 
3, Fig. 4), but both could be separated from yellowfin 
menhaden larvae between 10 and 20 mm (Houde and 
Swanson 1975) by body depth, prepelvic length, and 
head length. Snout length and eye diameter may be 
useful to distinguish 15-25 mm specimens; snouts 
>7% of SL and eye diameter >9% of SL probably 
identify yellowfin menhaden. 

Myomeres 

The total number of myomeres could be counted 
only on specimens under 17 mm in length. Although 
the preanal myomeres could be easily counted on 
larger specimens, the last few postanal myomeres on 
the peduncle became indistinguishable. The number 
of myomeres (mean = 44.6) did not change 
significantly with length in gulf menhaden and cor- 
responds with the number of adult vertebrae (44-46; 



mean = 44.7 not counting the hypural bones) report- 
ed by Dahlberg (1970). Radiographs of 20 adult gulf 
menhaden spawners used in my study showed that all 
fish had either 45 or 46 vertebrae (counting 
hypurals), with a mean of 45.6. During development 
the dorsal and anal fins moved in relation to the 
myomeres (Table 4). The anterior end of the dorsal 
fin moved from myomere 30 forward to myomere 23, 
numbered from head to tail. The posterior end of the 
dorsal fin remained fixed at myomere 32. The anus 
and the anterior end of the anal fin moved forward 
from myomere 37 to myomere 34. The postdorsal- 
preanal myomere count of 2 or 3 is diagnostic for 
Brevoortia at lengths >14 mm. Atlantic menhaden 
larvae 6-16 mm SL had a mean of 47.2 myomeres, 
with about two more predorsal myomeres and one 
more postanal myomere than gulf menhaden. 
Myomere number and distribution for gulf men- 
haden and yellowfin menhaden (Houde and Swanson 
1975) were so similar that neither were useful for 



89 



Table 3. 



FISHERY BULLETIN: VOL. 82, NO. 1 

-Proportions of head and body parts of Atlantic menhaden, Brevoortia tyrdnnus, expressed as a percent of 
standard length. Characters were not developed at lengths marked with a dash. 



Length class 


Number of 


Preanus 


Predorsal 


Prepelvic 


Body 


Dorsal fin 


Anal fin 


Head 


Snout 


Eye 


(mm. SLJ 


specimens 


length 


length 


length 


depth 


base length 


base length 


length 


length 


diameter 


30-3.9 


4 


85.4 


— 


— 


— 


— 


— 


14 5 


1 9 


6.7 


4.0-4.9 


10 


82 8 


— 


— 


8 3 


— 


— 


12 


1.9 


54 


5.0-5.9 


15 


81 


— 


— 


84 


— 


— 


12.1 


2 2 


50 


6.0-6.9 


7 


81 4 


— 


— 


8 4 


— 


— 


13.3 


2 5 


4.8 


7.0-7.9 


18 


82 3 


71.0 


— 


8 


2 6 


— 


13.7 


2 6 


4 8 


8.0-8 9 


12 


82 7 


696 


— 


7.9 


4 3 


— 


13 9 


2 6 


4 8 


9.0-99 


13 


82 8 


67.3 


— 


8 3 


6 5 


24 


15.0 


3 


5.2 


10.0-10.9 


i 3 


85 6 


669 


— 


8.6 


9.5 


4.2 


16.4 


3.4 


5.3 


11.0-11.9 


8 


85.9 


66.4 


— 


3 7 


10.1 


5 


16.6 


3 5 


5.4 


120-12 9 


10 


84.7 


64 6 


— 


9 1 


1 1 5 


5 5 


17 6 


3 7 


5.6 


13.0-13 9 


10 


83.2 


63.6 


— 


9 4 


13.0 


6 / 


18 2 


4 


6.0 


14.0-14.9 


7 


82.9 


62 7 


45.8 


9 8 


13 6 


7.1 


18.3 


40 


62 


150-15.9 


7 


81 7 


61 9 


45 3 


10 


14.0 


7 8 


18.3 


40 


6 2 


16.0-16.9 


9 


80.8 


62 5 


45 5 


11.4 


14.0 


8.8 


20 2 


4 1 


6 8 


170-17.9 


3 


79 9 


60 2 


47 6 


12 8 


15 2 


10.0 


22.9 


4 5 


7.3 


180-18.9 


>■, 


77.9 


586 


47 


14 2 


15 7 


106 


23.2 


4 4 


7.4 


19 0-199 


9 


76 9 


57 3 


48 


16 1 


16 


11.8 


23 8 


4 6 


7.8 


20.0-21 9 


7 


74 2 


538 


486 


19.8 


17 1 


142 


27.2 


48 


8.0 


220-239 


3 


73.4 


504 


509 


24 7 


17.7 


16 1 


29 8 


5.5 


80 


240-25.9 


2 


72.7 


51.4 


51.3 


25.4 


17 6 


15 9 


31 


5.7 


7 6 


26.0-27 9 


3 


74.7 


49 6 


52 8 


29 1 


19.6 


18.0 


31.1 


7.0 


8 3 


280-29.9 


1 


72.6 


48 9 


51 5 


290 


17.3 


16 3 


31 3 


b 8 


7.8 


30 0-349 


4 


75.4 


49.6 


52.4 


32.6 


20.0 


15.6 


330 


8.2 


88 


350-39.9 


3 


76.0 


49.9 


53 2 


36.5 


20 4 


16.6 


33.6 


7.6 


7.8 


40 0-49.9 


4 


74.9 


49.6 


52 2 


33.5 


20 5 


17.1 


32.2 


1 6 


8 3 


60.0-69.9 


3 


74.8 


48.9 


52 2 


334 


19 5 


16 8 


324 


7.0 


5.3 



TABLE 4. — Number of myomeres relative to dorsal fin and anal locations on gulf menhaden, 

Brevoortia patronus, larvae. 



Length class 
(mm. SL) 




Preana 






Postana 






Predorsa 


1 


Postc 


ursal- 


Preanal 


N 


Range 


Mean 


N 


Range 


Mean 


N 


Range 


Mean 


N 


Range Mean 


<60 


4 


36-37 


36 7 


4 


3 


8.0 


— 


— 


— 


— 


— 


— 


6 1-80 


16 


36-37 


36.7 


3 


7-9 


7 7 


9 


28-30 


28.9 


9 


4-6 


5 3 


8.1-10.0 


9 


35-38 


36 3 


9 


8-10 


8 6 


9 


26-28 


27.3 


9 


8 5 


4 4 


10 1-120 


10 


33-37 


354 


10 


8-10 


9.1 


10 


23-27 


252 


10 


3-4 


3.3 


12 1-140 


4 


33-35 


34.0 


4 


8-10 


9 5 


4 


23-25 


23.7 


4 


2-3 


2 2 


14 1-17 


4 


32-33 


32.5 


— 


— 


— 


-1 


22-23 


22 2 


9 


1-2 


1.8 



separating small larvae of these species. Yellowfin 
menhaden had a mean of 45.7, about one less pre- 
dorsal myomere, and about one to two more postanal 
myomeres than gulf menhaden. Atlantic menhaden 
had about two more preanal myomeres and about 
one more postanal myomere than gulf menhaden at 
each size class (Table 5). 

Meristics 

In gulf menhaden the caudal and dorsal fins were 
the first fins to initiate development and the pectoral 



fins were the last fins to complete development, even 
though they were the first fins to form as nonrayed 
buds (Table 6, Fig. 1C). Two specimens had an extra 
principal ray in both the upper and lower group of 
caudal rays. Vertebrae centra did not first stain with 
alcian blue as did other bony structures. At 13 mm, 
vertebrae first stained with alizarin red S, with the 
staining reaction progressing from the middle of the 
column towards each end as length increased. The 
neural and haemel spines initially stained blue, 
beginning at each end of the column and progressing 
towards the middle. The mean number of vertebrae, 



TABLE 5. — Number of myomeres relative to dorsal fin and anal locations on Atlantic men- 
haden larvae, Brevoortia tyrannus. Myomeres on specimens <6 mm could not be 
accurately counted. 



Length class 

(mm. SLl 




Preana 






Postana 






Predors, 


I 


Postdorsal- 


Preanal 


/V 


Range 


Mean 


N 


Range 


Mean 


N 


Range 


Mean 


N 


Rang 


; Mean 


6.1-8.0 


13 


38-40 


38 7 


13 


8-10 


9.0 


10 


30-31 


30 7 


10 


5-6 


5 7 


8 1-10.0 


16 


37-40 


384 


16 


8-11 


99 


16 


27-30 


29.0 


10 


4-6 


5 2 


10 1-120 


16 


36-37 


36 1 


16 


10-11 


108 


16 


25-28 


262 


16 


3-5 


4 II 


12.1-140 


14 


35-37 


35.6 


10 


10-11 


10 7 


14 


24-26 


25.1 


14 


3-4 


3 2 


14 1-16.0 


2 


35-36 


35 5 


— 


— 


— 


2 


24-25 


24 5 


3 


3 


3.0 



90 



HETTLER: DESCRIPTION OF GULF MENHADEN 



40r 



-I 
CO 



O 

z 

UJ 

-I 
co 

Z 
< 
UJ 

a 



CO 

I 
»- 
o 

z 

UJ 



< 

CO 

a 
O 

o 

UJ 

a 
o. 



90 



80 



70- 



7 Or 



60- 



50- 



55 



O 

Z 

UJ 

> CO 

_l 

UJ 

a. 

UJ 

a 
a. 



* 

-j 

CO 

I 

H 

o 

z 

UJ 



45 



10 



5- 



o 

z 

CO 



x .-«r> 












•v 









N 



•T 
• P 



s. • p 

, . -7 
T"—' / 



■y p 






M* 



# 



CO 



a 

UJ 
Q 

> 
Q 
O 
CO 



30 



20 



10 



40 r 



CO 



O 

z 

UJ 

_l 

Q 
< 

UJ 

I 



30- 



20- 




CO 

a. 

UJ 

< 

Q 

UJ 

> 



10 
15 



10 











- 




». 


.4 


- 




'■m 

■••• •■•'' £— 

// 


-<f 


- 










■i\ 






*— 


^ 






r.<^ 


/ 




— I 1 


i 


1 1 






* S 

T 



►>-v 



^ 



g^ t ^c - 



-- -£-' 



,V 



10 



20 



30 



10 



20 



30 



SIZE CLASS (mm) 



SIZE CLASS (mm) 



FIGURE 4. — Morphometric comparisons as a percentage of standard length of laboratory-reared Brevoortiapatronus (P).ff tyrannus (T),andR 

smithi (S). Yellowfin menhaden data from Houde and Swanson (1975). 



91 



FISHERY BULLETIN: VOL. 82, NO. 1 



TaBI I 6. Meristics in gulf menhaden, Hrcvtmrtia patronus, (35 specimens) and in Atlantic menhaden, li 

tyrannus, (3 I specimens). 





Size (mm 


SL) 


when 


Size (mm 


SL) when 












first stained 




all are stained 


Number in fu 


II complement 


Meristic 


B patronus 




B. 


tyrannus 


B. patronus 


B 


tyrannus 


B patronus 


B tyrannus 


Caudal fin rays 
























Principal 


8 








3 


9 




12 


10-11 


(dorsal) 


10 (dorsal) 




11. 








I i 


18 




20 


9-10 (ventral) 


9 (ventral) 


Procurrent 


















8-9 
7-8 


(dorsal) 
(ventral) 


7-8 (dorsal) 
6-7 (ventral) 


Diiis.iI tm 
























Pterygiophores 


B 








B 


16 




n, 


19-21 




18-19 


Rays 


8 








'< 


19 




17 


21-23 




20-22 


Anal fin 
























Pterygiophores 


9 








10 


16 




15 


17-20 




17-20 


Rays 


10 








12 


17 




1!, 


18-22 




19-21 


Pelvin fin rays 


li, 








!'■ 


18 




18 


7 




7 


Pectoral fin rays 


18 








IK 


21 




.'1 


13-15 




15-17 


Predorsal bones 


pi 








1 J 


21 




21 


9-1 1 




10-12 


Vertebrae 


13 








14 


16 




15 


45-46 




48-49 


Ventral scutes 


21 








21 


31 




27 


29-31 




32-33 



including the hypural bones, was 45.3 counted in 21 
specimens longer than 16 mm SL. The first bones to 
stain with alizarin red S were the dentaries, the max- 
illaries, and the cleithra which occurred in 9 mm 
specimens. 

Only vertebrae and ventral scute counts were useful 
in separating gulf menhaden and Atlantic menhaden; 
other meristics overlapped (Table 6). Yellowfin 
menhaden larvae could not be separated from the 
two large-scaled menhaden by meristics, with the 
possible exception of Atlantic menhaden that had 
47-48 vertebrae and yellowfin menhaden that had 
45-47 (including the hypural bones) (Dahlberg 1970). 

Pigmentation 

Pigmentation of gulf menhaden larvae (Figs. 1,3,5, 
6) was similar, but not identical, to the pigmentation 



described for yellowfin menhaden (Houde and Swan- 
son 1975) and Atlantic menhaden (Jones etal. 1978). 
Gulf menhaden up to 8 mm had 1 melanophore on the 
dorsal side of the notochord tip and 1 or 2 mela- 
nophores on the ventral side of the notochord tip, 
which is diagnostic for the genus Rrevoortia (Figs. 1C, 
D, 5A). Lateral pigmentation, although found on the 
trunk of specimens as small as 4.9 mm, was not found 
on all small specimens. At 10 mm, all specimens had 
5-20 melanophores scattered the length of the trunk. 
Larvae 4-5 mm had 10-20 tiny melanophores on top 
of the head. One 7.8 mm larva had a single stellate 
melanophore on top of the head behind the eyes. One 
single medial melanophore, which enlarged into 
additional melanophores as larvae grew, was present 
along the isthmus (ventral midline forward of the 
cleithrum) on 6 mm and larger larvae. On 8-20 mm 
larvae, 1 or more melanophores occurred along the 





Vrftl 



1mm 



Fit, i RE 5. Larval Brei oortla patronus: (A) 7.2 mm (12 d alter hatching). (B) 9.2 mm (20 d after hatching). 



92 



HETTLER: DESCRIPTION OF GULF MENHADEN 




5mm 



FIGURE 6.— Juvenile Brevoortia patronus 33.8 mm (90 d after hatching). 



cleithrum axis on each side. Along the surface, lateral 
and parallel with the dorsal surface of the foregut, 
there were usually 6-10, but sometimes up to 20, rec- 
tangular melanophores on each side. These paired 
melanophores were positioned anteriad to 2 or 3 
stellate melanophores covering the dorsal surface of 
the gas bladder. A series of 10-18 medial, unpaired 
melanophores occurred between the trunk muscula- 
ture and the dorsal surface of the gut. This series 
merged into 1-3 stellate melanophores projecting 
ventrally over the end of the gut towards the anus. A 
medial string of nearly continuous, thin mela- 
nophores traced the junction of the finfold along the 
ventral surface of the hindgut. Dorsal to the base of 
the anal fin 2 or more melanophores were always pre- 
sent in larvae >5 mm. The caudal fin was pigmented 
by 10 mm, whereas the medial fins, lower jaw tip, 
snout, and nape acquired pigment by 18 mm (Fig. 
3C). Pigment was absent on the surface lateral to the 
ventral portion of the foregut between the distal end 
of the pectoral fin rays and the pelvic fin. 
Melanophores were present on specimens >17 mm 
along the base of the dorsal fin and along the dorsal 
midline between the dorsal and caudal fins. Paired 
melanophores were absent between the head and 
dorsal fin. For pigment descriptions of gulf 
menhaden larvae and juveniles >19 mm, see 
Suttkus (1956). 

Other Structures 

By 4.5 mm, the dentaries, maxillaries, branchial 
arches, cleithra, and hypurals were stained with 
alcian blue, but the first bones to accept alizarin red S 
stain, and thus indicate ossification, were the cleithra 
in 8.5 mm specimens. Flexion of the notochord 



upward to initiate caudal fin development began at 7 
mm. Ossification of the hypural bones began at 10 
mm and was completed at 15 mm. Eight maxillary 
teeth and three dentary teeth on each side were 
observed on 10 mm larvae. Fourteen teeth on each 
maxillary and three teeth on each dentary were still 
visible on 25 mm juveniles. In the oral cavity of 16-24 
mm larvae, one or two teeth projected downward 
from each endopterygoid bone and one or two teeth 
projected upward from the second basibranchial car- 
tilage. These teeth were absent in fully transformed 
juveniles. Scales were first visible along the dor- 
solateral margin of the caudal peduncle and along the 
midline on each side of the trunk at the beginning of 
transformation, which occurred at 19 mm. 

COMPARISON AMONG BREVOORTIA 
AND WITH OTHER CLUPEIDS 

Of the Brevoortia species, eggs and larvae of gulf 
menhaden were the most difficult to distinguish from 
yellowfin menhaden. Gulf menhaden had 44-46 
myomeres, whereas yellowfin menhaden had 45-47 
(Houde and Swanson 1975). Morphometries may be 
useful to distinguish 10-25 mm specimens of gulf 
menhaden from yellowfin menhaden. At equal 
lengths, gulf menhaden had less body depth, a short- 
er head length, a longer prepelvic distance, a longer 
predorsal distance, a shorter snout, and a smaller 
eye. Yellowfin menhaden >17 mm had paired 
melanophores between the head and the dorsal fin 
(Houde and Swanson 1975), whereas gulf menhaden 
did not. Wild specimens of yellowfin menhaden from 
southern Florida also had a double row of 
melanophores along the ventral midline between the 
pectoral and pelvic fins, but neither laboratory- 



93 



FISHERY BULLETIN: VOL. 82, NO. 1 



reared gulf menhaden or wild specimens of gulf 
menhaden collected from four locations along the 
northern Gulf of Mexico had ventral midline pig- 
ment. Gulf menhaden had more dorsal fin rays, but 
both species had an equal number of anal rays. Fer- 
tilized eggs of the two species had the same diameter, 
but gulf menhaden had a larger oil droplet (0.20 vs. 
0.15 mm) than yellowfin menhaden. No description 
of finescale menhaden larvae exists, but presumably 
they have 42-43 myomeres, based on the number of 
vertebrae reported for this species (Dahlberg 1970). 
Although gulf menhaden larvae are geographically 
separated from Atlantic menhaden larvae, they can 
be separated by counting myomeres or vertabrae; 
gulf menhaden, 44-46; and Atlantic menhaden, 47- 
48. Atlantic menhaden and yellowfin menhaden had 
nearly equal dorsal and anal fin ray numbers, but 
Atlantic menhaden had one to four more myomeres 
and lacked dorsal and ventral midline paired 
melanophores anterior to the dorsal and pelvic fins. 
Mophometric differences between Atlantic men- 
haden and yellowfin menhaden are similar to dif- 
ferences between gulf menhaden and yellowfin 
menhaden. 

There are some differences in egg and larval meris- 
tics and morphology data between my study and the 
literature, which may be due to differences between 
laboratory-reared and wild specimens. Houde and 
Fore (1973) reported that gulf menhaden had 45-48 
myomeres (vs. 44-46 that I found for gulf menhaden), 
20-23 anal rays (vs. 19-21), 17-21 dorsal rays (vs. 20- 
22), and reported that pelvic fins in northern gulf 
specimens were not developed until 20 mm (vs. 18 
mm). They also reported that gulf menhaden eggs 
had a diameter of 1.04-1.30 mm (vs. 1.18-1.34 mm), 
an oil droplet of 0.08-0.20 mm (vs. 0.16-0.22 mm), 
and a wide perivitelline space of about 33% (vs. 24- 
28';). Jones et al (1978) reported that Atlantic 
menhaden egg diameter was 1.30-1.95 mm (vs. 1.54- 
1.64 mm that I found for Atlantic menhaden), that 
yolk diameter was 0.90-1.20 (vs. 0.82-.095 mm), and 
that the oil droplet diameter was 0.11-0.17 (vs. 
0.20-0.23). For Atlantic menhaden larvae of 
unspecified lengths they reported 16-18 dorsal rays 
(vs. 20-22), 18-20 anal rays (vs. 19-21), and a body 
depth:standard length ratio of about 0.05 at 23 mm 
total length (vs. about 0.20 I found at the same 
length); however, the body depth ratio is undoubt- 
edly a typographical error. 

Laboratory-reared gulf menhaden and Atlantic 
menhaden both appeared to transform into juveniles 
at a smaller size than wild fish. Morphometric data 
and photographs of specimens of gulf menhaden 
from Louisiana indicated that the juvenile form was 



not reached until about 30 mm SL (Suttkus 1956). 
Lewis et al. (1972) indicated that Atlantic menhaden 
from North Carolina did not complete "prejuvenile" 
growth until about 33 mm SL. Houde and Swanson 
(1975) suggested that tank-reared yellowfin men- 
haden transformed at smaller sizes than did wild fish, 
and I concur. 

Characters useful for separating eggs and larvae of 
Brevoortia from other clupeids have been identified 
(Houde and Fore 1973; Richards et al. 1974; Houde 
and Swanson 1975; Powles 1977). Sardinella and 
Opisthonema have about the same total myomere 
counts as Brevoortia, but usually have 6-9 post- 
dorsal-preanal myomeres. Ktrumcus has the same or 
more total myomeres than Brevoortia, but about 10 
fewer anal rays. The smaller larvae of Sardinella, 
Opisthonema, and Etrumeus have no pigment on the 
dorsal side of the notochord tip, whereas Brevoortia, 
Harengula, and Jenkinsia have this pigment. 
However, Jenkinsia and Harengula have 42 or fewer 
myomeres. The spawning seasons of all these genera 
overlap with the spawning season of Brevoortia 
species (Houde and Fore 1973; Powles 1977; Jones 
et al. 1978). Larvae of Dorosoma and Alosa are not 
normally found in marine waters with Brevoortia. 

ACKNOWLEDGMENTS 

I thank John J. Govoni, William R. Nichols, and 
Allyn and B. Powell of the Beaufort Laboratory for 
reviewing the early drafts of the manuscript; Ed 
Houde of the University of Maryland for his review of 
a later draft; and Thomas Potthoff of the Southeast 
Fisheries Center, NMFS, and G. David Johnson of 
the South Carolina Wildlife and Marine Resources 
Department for their comments on terminology. This 
research was supported by a contract from the Ocean 
Assessments Division, National Ocean Services, 
NOAA. 

LITERATURE CITED 

Dahlberg, M. D. 

1970. Atlantic and Gulf of Mexico menhadens, Genus 

Brevoortia (Pisces: Clupidae). Bull. Fla. State Mus., Biol. 

Sci. 15:91-162. 
Hettler, W. F., Jr. 

1968. Artificial fertilization among yellowfin and Gulf 

menhaden [Brevoortia) and their hybrid. Trans. Am. 

Fish. Soc. 97:119-123. 
1970. Rearing larvae of yellowfin menhaden, Brevoortia 

smithi. Copeia 1 970:775-776. 
1981. Spawning and rearing Atlantic menhaden. Prog. Fish- 
Cult. 43:80-84. 
1983. Transporting adult and larval gulf menhaden and 

techniques for spawning in the laboratory. Prog. Fish- 



94 



HETTLER: DESCRIPTION OF GULF MENHADEN 



Cult. 45:45-48. 

HlLDEBRAND, S. F. 

1963. Family Clupeidae. In H. B. Bigelow (editor), Fishes of 
the western North AtlantT. Par) Three, p. 257-454. 
Mem. Sears found. Mar. Res. Yale Univ. 1. 
HOUDE, E. D., and P. L. Fore. 

1973. Guide to identity of eggs and larvae of some Gulf of 
Mexico clupeid fishes. Fla. Dep. Nat. Resour., Mar. Res. 
Lab., Leafl. Ser. Vol. IV, Pt. 1, No. 23, 14 p. 
Hoi UK, E. D., AND L. J. SWANSON, JR. 

197"). Description of eggs and larvae of yellowfin menhaden. 
Brcvnortia smithi. Fish. Bull.. U.S. 73:660-673. 
Jones, P. W. . F. D. Martin, and J. D. Hardy, Jr. 

1978. Development of fishes of the Mid-Atlantic Bight. Vol. 
1, Acipenseridae through Ictaluridae. U.S. Fish Wildl. 
Serv., Biol. Serv. Program FWS/OBS-78/12, 314 p. 
Lewis, R. M., E. P. H. Wilkens, and H. R. Gordy. 

1972. A description of young Atlantic menhaden, Brevoortia 
tyrannus, in the White Oak River estuary. North Car- 
olina. Fish. Bull., U.S. 70:115-118. 



POWLES, H. 

1977. Description of larval Jenkinsia lamprotaenin 
(Clupeidae, Dussumieriinae) and their distribution off 
Barbados, West Indies. Bull. Mar. Sci. 27:788-801. 
Richards, W. J., R. V. Miller, andE. D. Houde. 

1974. Egg and larval development of the Atlantic thread her- 
ring. Opisthnncma nglinum. Fish. Bull., U.S. 72:1123- 
1 136. 
SUTTKUS, R. D. 

1956. Early life history of the largescale menhaden, Brevoor- 
tia patronis, in Louisiana. Trans. North Am. Wildl. Conf. 
21:390-407. 
Turner, W. R. 

1969. Life history of menhadens in the eastern Gulf of Mex- 
ico. Trans. Am. Fish. Soc. 98:216-224. 
U.S. National Marine Fisheries Service. 

1982. Fisheries of the United States, 1981. U.S. Dep. Com- 
mer., NOAA. Natl. Mar. Fish. Serv., Curr. Fish. Stat. 8200, 
131 p. 



95 



DISTRIBUTION OF ICHTHYOPLANKTON OFF SAN ONOFRE, 
CALIFORNIA, AND METHODS FOR SAMPLING VERY SHALLOW 

COASTAL WATERS 



Arthur M. Barnett, 1 Andrew E. Jahn, 2 Peter D. Sertic, 1 
and William Watson 1 



ABSTRACT 

Spatial abundance patterns of inshore marine fish larvae, together with day-night and ontogenetic changes in 
these patterns, were investigated at a single site off the southern California coast using neustonic, midwater, 
and epibenthic samplers. Fifteen of the nineteen most abundant taxa showed statistically significant abun- 
dance patterns: Five taxa were principally in the inshore (<2 km from shore) epibenthos, one in the inshore 
neuston, two in the neuston and midwater less than about 5 km from shore, three to midwater 2-5 km from 
shore, and four in midwater offshore of about 3.5 km. Abundance patterns for the three most common taxa, 
Engraulis mordax, Genyonemus lineatus, and Seriphus politus, shifted toward shore and toward the bottom 
with increasing larval size. Comparison of E. mordax egg and larval abundances indicated a large excess of 
larvae over eggs nearshore. Only two taxa showed statistically significant day-night pattern changes; both 
were lower in the water column during the day. 

The existence of inshore abundance maxima implies significant survival value in occupying the nearshore 
zone. The shallow waters of the southern California coast may represent a nursery area comparable in impor- 
tance to the estuarine nurseries of the Atlantic coast of North America. 



Through the pioneering California Cooperative 
Oceanic Fish Investigation (CalCOFI) work of the 
late E. H. Ahlstrom and co-workers (Ahlstrom 1959, 
1965), ichthyoplankton of the Southern California 
Bight are generally well known. However, the 
CalCOFI effort was concentrated on species found 
principally offshore of the 100 m isobath, and the lar- 
vae of most inshore fishes are rare or missing in the 
published CalCOFI data. Recent studies of 
ichthyoplankton in the Southern California Bight 
inshore of the 100 m isobath (Brewer et al. 1981; 
Gruber et al. 1982; Brewer and Smith 1982) have 
indicated that many of these larvae are found in the 
relatively shallow waters. 

In this paper we present methods for sampling 
quantitatively the entire water column in shallow 
waters (6-75 m) and describe the spatial abundance 
patterns of the most commonly occurring larval 
fishes. Of particular interest was the distribution of 
larvae in the onshore-offshore vertical plane. 
Ontogenetic pattern changes were investigated for 
three abundant species: Engraulis mordax, Geny- 
onemus lineatus, and Seriphus politus. 



'Marine Ecological Consultants of Southern California, 531 
Encinitas Boulevard, Suite 110, Encinitas, CA 92024. 

2 Marine Ecological Consultants of Southern California, 531 
Encinitas Boulevard, Suite 110, Encinitas, Calif.; present address: 
Los Angeles County Museum of Natural Histoiy, 900 Exposition 
Boulevard, Los Angeles, CA 90007. 



The study was done off San Onofre, Calif., (Fig. 1) 
from September 1977 to September 1979. Unit 1 of 
the San Onofre Nuclear Generating Station, a 500- 
megawatt plant located 1.5 km northwest of the 
sampling area, was operating continuously through- 
out the course of the study. However, this plant has 
been shown to have only very localized effects which 
have not interfered measurably with the results 
reported herein (Marine Review Committee 1979 3 ; 
Bartlet et al. 198 1 4 ). This study was completed prior 
to the beginning of operation of Units 2 and 3 of the 
San Onofre Nuclear Generating Station. 

Our sampling methodology resulted from a pre- 
liminary study in which we found that a combination 
of sampling gear was necessary to estimate nearshore 
larval abundance. The chief purpose of this paper is 
to present these sampling methods. Results are 
shown which verify the effectiveness of these 
methods and further suggest some peculiarities of 
the nearshore habitat. 



'Marine Review Committee. 1979. Interim report of the Marine 
Review Committee to the California Coastal Commission. Part 1: 
General summary of findings, predictions, and recommendations 
concerning the cooling system of the San Onofre Nuclear Generating 
Station. In Marine Review Committee Document 79-02, p. 1- 
20. Marine Review Committee of the California Coastal Commis- 
sion, 631 Howard Street, San Francisco, CA 94105. 

4 Barnett,A.M.,P.D. Sertic, and S.D. Watts. 1981. Final report: 
Ichthyoplankton preoperational monitoring program. Marine 
Ecological Consultants of Southern California, 531 Encinitas 
Boulevard, Encinitas, CA 92024, 8 p. 



Manuscript accepted June 1983. 

FISHERY BULLETIN: VOL. 82, NO. 1. 1984. 



97 



FISHERY BULLETIN: VOL. 82, NO. 1 



San Onofre Nuclear 
Generating Station 




18 m 

37 m 
55m 

74 m 



13 m it- A 

'"oV-iT- San Onofre Kelp 



c 



D 




Long Beoch^ 

N 



16 km 
Pacific Ocean 



FIGURE 1.— Chart of the sampling area and its position off the 
southern California coast. The one- and two-dimensional pattern 
analyses were based on samples taken at a randomly selected 
isobath in each of the five sampling blocks (A-E) on each sampling 
date. The study of daily vertical migration was based on samples 
taken along the 8 and 13 m isobaths (dotted lines). 



size, and sampling time for the ensuing full-scale 
program. The results of this brief study indicated 
that 

1. Filtration efficiency was at least 85% for all nets 
and lengths of tow. 

2. Samples of 400 m 3 were adequate to attain 
asymptotes of numbers of taxa per tow. A sampled 
volume of 400m 3 from the epibenthos was the max- 
imum that could be handled economically. 

3. The 12 most abundant larval fish taxa were 
neither randomly nor evenly distributed with respect 
to the three vertical strata. Half the taxa were prin- 
cipally epibenthic, while 25% were neustonic and 
25% were most abundant in midwater. 

4. Only one of these taxa showed a daily vertical 
migration; Paraclinus integripinnis, not a top-ranking 
species in the ensuing study, tended to descend from 
midwater to the epibenthic layer at night. 

5. Size of individuals and apparent abundance of 
most taxa increased at night, probably because of 
visual avoidance during the day. 

6. Nitex netting of 0.333 mm mesh retained more 
fish eggs and smaller anchovy larvae than did 0.505 
mm mesh. 

From the preliminary results, it was clear that the 
bongo net alone would undersample significant frac- 
tions of many larval populations. Since our goal was 
to estimate the density and distribution of nearshore 
ichthyoplankton, we decided to use all three types of 
gear with 0.333 mm mesh and to filter a target volume 
of 400 m 3 . 



METHODS 
Preliminary Study 

In shallow depths, interfaces at the sea surface and 
seabed comprise a substantial portion of the water 
column. In addition, concentration of a species at 
either interface would necessitate sampling the 
epibenthic and neustonic layers as well as the mid- 
water column to obtain quantitative abundance 
estimates. 

Neustonic, midwater, and epibenthic samplers 
were used in a preliminary study 5 between Septem- 
ber and November 1977, to verify their effectiveness 
and to select mesh size, net design, standard sample 



! Barnett, A. M.,J. M. Leis, and P. D. Sertic. 1978. Report to the 
Marine Review Committee on the preliminary ichthyoplankton 
studies. Marine Ecological Consultants of Southern California, 53 1 
Encinitas Boulevard, Encinitas, CA 92024. 



Sampling Gear 

A bongo net was selected for sampling the midwaters, 
as recommended by Smith and Richardson (1977). 
An opening-closing 71 cm Brown-McGowan bongo 
net (total mouth area = 0.79 m 2 ) was used. A General 
Oceanics 6 (GO) flowmeter was mounted in the star- 
board frame. The bongo net, as conventionally used, 
is placed on the wire some distance above a weight 
and towed astern. The geometry of this arrangement 
and the circular net mouths make the gear ill-suited 
for sampling the plankton in the neustonic and epi- 
benthic strata near the sea surface and seabed, re- 
spectively. Therefore, specially designed samplers, 
described below, were used to sample these 
layers. 

We chose the brown manta net (Brown and Cheng 



6 Reference to trade names does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 



98 



BARNETT ET AL.: DISTRIBUTION OF ICHTHYOPLANKTON OFF SAN ONOFRE, CALIF 



1981) as our neustonic sampler. This net had an 88 
cm wide mouth and fished to a depth of 16 cm. 
Fiberglass-covered styrofoam floats kept the top of 
the net out of water, and a 3 m spar and asymmetrical 
bridle kept the gear outboard of the bow wave. A 
weight suspended from the end of the wire held the 
bridle well below the surface, out of the path of the 
net. The sampler was launched and recovered off the 
quarter by means of a tag line. Both a Tsurumi-Seiki 
(TSK) flowmeter and a GO flowmeter were mounted 
in the mouth of the net. The GO meter served as a 
back-up for the TSK, which sometimes fouled with 
kelp and eelgrass. 

The Auriga net, 7 used to sample the epibenthic 
layer, consisted of a rectangular net frame (0.5 m high 
X 2 m wide) attached to a chassis equipped with a 
pair of side-mounted, 2 m diameter wheels. The 
device rolled on the bottom so that the mouth of the 
net was 10 cm (original design) or 17 cm (later ver- 
sions) above the bottom of the wheels. A series of 12 
cm diameter plastic rollers below the mouth of the 
net helped prevent the sampler from digging into the 
bottom and presumably minimized escapement 
below the net. Both GO and TSK flowmeters were 
mounted within the mouth of the Auriga net. The 
Auriga net was towed off the stern. Divers have 
observed (M. Sowby 8 ) that the mouth of the Auriga 
assumes a horizontal attitude when the wheels are off 
the bottom. We therefore believe that contamination 
of the epibenthic samples by midwater plankton was 
minimal during launch and recovery, when the main 
component of (relative) water movement was across, 
rather than through, the mouth. Any contamination 
that did occur should have been a function of depth, 
which was always <209r of the length of an 
epibenthic tow (this potential source of error has 
been ignored in the density calculations). 

Although serious clogging was not apparent in the 
preliminary study, denser plankton concentrations 
at other times of the year might clog the nets before 
400 m 1 of water could be filtered. Clogging would be 
most serious for oblique bongo tows, because it 
would result in undersampling of the upper part of 
the water column. In anticipation of this possibility, 
the area of mesh in all nets was increased according to 
the criteria suggested by Smith et al. (1968, equation 
5) in order to sample 500 m 3 (bongo), 400 m 3 (Auriga), 
and 200 m 3 (Manta) for "green" coastal waters. The 
filtering ratios (R = mesh pore area/net mouth area) 
of bongo, Auriga, and Manta nets were increased to 



7.8, 6.6, and 1 0.7, respectively, by adding mesh cylin- 
ders ahead of the conical portions of the nets. Exter- 
nal flowmeters were not used in the subsequent 
surveys, but tows were carefully timed. Internal flow- 
meter readings were checked upon recovery, and 
samples were repeated if the readings differed by 
more than 20% from expected values. 

Except for the limited study of daily vertical migra- 
tion, all sampling was done at night. The deck lights 
were always off during the neuston tows. All samplers 
were launched, towed, and recovered with the vessel 
underway at about 1 m/s. For bongo tows, wire was 
paid out (scope about 2:1) until the weight, located 
1.5 below the center of the net frame, bumped the 
bottom. Then the nets were opened, and a stepped 
oblique tow was made consisting of 18 30-s steps. 
The Auriga sampler was towed with a scope of 3:1 
and recovered after 6.5 min on the bottom. With the 
small-mouthed Manta net, the volume of 400 m 3 was 
achieved by towing two nets simultaneously, off port 
and starboard, for 20 min (about 1.4 km). 

Samples were preserved in 5-10% seawater- 
Formalin. 

Sampling Locations and Frequency 

Since we eventually wanted to assess the effects of a 
power plant cooling system, it was necessary to con- 
centrate much of our sampling effort within the depth 
contours encompassing the cooling structures. At the 
same time, in order to estimate the abundance of 
nearshore species, we needed to sample far enough 
from shore to delimit their centers of abundance. We 
decided upon a stratified random sampling design 
(Snedecor and Cochran 1967) wherein, on each sam- 
pling date, the neustonic, midwater, and epibenthic 
layers were sampled along a randomly chosen depth 
contour in each of five blocks (Figs. 1, 2). The five 
blocks were defined by depth contours: A) 6-9 m, cor- 
responding to cooling water intake locations; B) 9- 1 2 
m and C) 12-22 m, both corresponding to future dif- 
fuser discharge locations; D) 22-45 m, corresponding 
to a faunal break between inshore and coast- 
al zooplankton assemblages (Barnett and Sertic 9 ); 
and E) 45-75 m, chosen a priori as the likely offshore 
limit of most nearshore larval fishes. 

The sampling transect thus consisted of 15 strata: 
Three depth layers in each of five blocks (Fig. 2). To 



"Marine Biological Consultants, Inc., 947 Newhall Street, Costa 
Mesa, CA 92627. 

"M. L. Sowby. Marine Biological Consultants, Inc., 947 Newhall 
Street, Costa Mesa, CA 92627, pers. commun. 1979. 



'Barnett, A.M., and P. D. Sertic. 1979. Spatial and temporal pat- 
terns of temperature, nutrients, seston, chlorophyll-a and plankton 
off San Onofre from August 1976 - September 1978, and the 
relationships of these patterns to the SONGS cooling system. In 
Marine Review Committee Document 79-01, p. vii through 9- 
89. Marine Review Committee of the California Coastal Commis- 
sion, 631 Howard Street, San Francisco, CA 94105. 



99 



FISHERY BULLETIN: VOL. 82, NO. 1 



FIGURE 2. — Diagrammatic profile of the 
study transect showing the 15 strata sam- 
pled off San Onofre, Calif. Neustonic and 
epibenthic layers are vertically ex- 
aggerated. 



Distance from Shore (km) 
2 3 4 5 6 7 8 
_1 I I I I 1 I 






- 10 
20 
30 o 

ft) 

40 ? 

- 50 3 

- 60 
70 
80 



avoid the San Onofre kelp bed, some of the tows in 
the B and C blocks were offset by about 1 km. 
VVilcoxon signed rank tests of samples taken from B 
block and B offset (Fig. 1) showed no significant dif- 
ferences in species abundances (P > 0.05) between 
the main block and the offset which could not be 
related to the inshore-offshore patterns discussed 
below. 

The transect was sampled monthly in January and 
February 1978, fortnightly from March through 
August 1978, and again monthly through September 
1979. During each of these 28 sampling periods, the 
five blocks were surveyed once each night for 1-3 
nights, giving a total of 57 sampling dates for the 21- 
mo study. 

As noted above, we chose a standard sampled 
volume of 400 m 3 based on the preliminary study. 
This volume was large enough to assure a representa- 
tion of all abundant species throughout the year. 
Volume was used as the sampling unit, although an 
argument based on the scale of patchiness could be 
made for length of tow (i.e., 400 m in each water layer) 
as the criterion, rather than volume filtered (P. 
Smith 10 ). Most tows were at least 400 m long. 

Laboratory Procedures 

Samples were sorted for fish eggs and larvae under 
dissecting microscopes at 10X magnification. The 
choice of 400 m 3 as the sampled volume was made at 
a time of year when ichthyoplankton abundance was 
low (Walker et al. 11 ); consequently the samples from 
other times of year were larger than necessary to rep- 



,0 P. E. Smith, La Jolla Laboratory, Southwest Fisheries Center, 
National Marine Fisheries Service, NOAA, La Jolla, CA 92038, 
pers. commun. 1979. 

"Walker, H. J., A. M. Barnett, and P. D. Sertic. 1980. Seasonal 
patterns and abundance of larval fishes in the nearshore Southern 
California Bight off San Onofre, California. Marine Ecological Con- 
sultants of Southern California, 53 1 Encinitas Boulevard, Encinitas, 
CA 92024. 



resent the nearshore assemblage. Samples with large 
plankton volumes were subsampled, using a Folsom 
plankton splitter before sorting. The size of the sub- 
sample was set to include at least 100 non-engraulid 
larvae (the mean number of larvae counted per sub- 
sample was 277, of which 62.8% was is. mordax). This 
fraction was usually one-fourth and was seldom 
smaller than one-eighth. Eggs were sorted from 1%, 
5%, or 10 f /r (to get at least 100 eggs) of the residue of 
the fraction sorted for larvae. Sorting efficiency was 
maintained above 90%. 

Some epibenthic samples contained so much sand 
and detritus that it was necessary to clean them 
before sorting, using a flotation technique adapted 
from Ladell (1936). After removal of large fish and 
debris, such a sample was mixed with a 40% MgS0 4 
solution (specific gravity = 1.2) in a large separator 
fashioned from a 19 1 (5-gal) plastic carboy with the 
bottom cut off and the neck fitted with a rubber hose 
and ball valve. Most detritus sank, while plankton 
floated to the top. The heavy material was drained off 
and processed once or twice more to ensure separa- 
tion of the plankton. Checks of the heavy residue of 
three such samples showed that more than 99% of 
the larvae were separated by flotation. 

All larvae were identified to the lowest taxonomic 
category currently possible. Eggs were identified as 
Engraulis mordax or "other". In some larval cate- 
gories (e.g., Atherinidae, Goby Type A), our ability to 
discriminate among species or larval types (sensu 
Richardson and Pearcy 1977) improved as the study 
progressed. However, not all of the old collections 
were reprocessed. When mixed taxa showed 
seasonal and spatial coherence, they were retained 
for the analyses presented here. 

Pattern Analysis 

All counts of eggs and larvae were standardized to 
number/400m-\ Thus the standardized numbers 



100 



BARNETT ET AL.: DISTRIBUTION OF ICHTHYOPLANKTON OFF SAN ONOFRE, CALIF. 



were roughly the same as the actual numbers of eggs 
and larvae caught, a desirable situation for analysis 
with transformed data (Murphy and Clutter 1972). 
These values were transformed by log (X + 1) before 
analysis for offshore and vertical pattern. The results 
were back-transformed, resulting in geometric 
means with asymmetric confidence bounds, and pre- 
sented as number/1 00m 1 . 

To describe the cross-shelf abundance patterns of 
ichthyoplankton, a procedure was adopted involving 
Hotelling's T 2 test and a series of a posteriori '(-tests 
(Morrison 1976) to divide the 15 strata into groups. 
These parametric methods allowed us to detect 
significant differences in mean abundance among 
components of a pattern and to determine con- 
fidence bounds on the means. 

Hotelling's T l test was selected over an analysis of 
variance (ANOVA) because the covariance struc- 
tures in the data tended not to meet the assumptions 
of standard ANOVA models (i.e., errors were not 
independent; the abundances of neighbor strata 
were likely to be correlated). The T 2 -test allows this 
correlation by using the sample covariance matrix, 
rather than (as in ANOVA) assuming a specified 
covariance pattern (Winer 1971; Morrison 1976). 

With a significant T 1 test result obtained (P< 0.05), 
a posteriori multiple (-tests were used to separate 
strata into groups having significantly different 
abundances. The strata were contrasted in a series of 
(-tests using the Bonferroni statistic, ((0.05),,, where 
k — potential number of contrasts, s = number of 
sampling periods — 1, and 0.05 = overall type / (a) 
error. The value of k was set as the number of all poss- 
ible contrasts among m strata plus 5, for further tests 
employing combinations of the initial strata: i.e., 
(m)(m-l) 



k 



+ 5. Bonferroni (-values were taken 



After the initial series of (-tests of all possible com- 
parisons, strata found not to differ significantly were 
pooled into initial groups. The time-averaged abun- 
dance of each stratum was used to calculate the initial 
groups' mean abundance 



Zj = 



z 



Z,/n 



where Z, is the initial group mean, n is the number of 
strata in the initial group, and Z, are the means of 
individual strata. Further (-tests (the total of all tests 
<k) were made to contrast the resulting initial 
groups. If more than one final grouping was possible, 
the final set of groups selected was that which max- 
imized the (-statistic. 

Both the Hotelling T 2 and the (-test assume nor- 
mally distributed data. Excessive zero values in a 
data set violate this assumption in a way that cannot 
be corrected by transformations. The methods used 
here were robust with respect to zero values in 
zooplankton data (Barnett et al. 12 ); nevertheless, 
some sampling dates for 1 2 of the 1 9 ichthyoplankton 
taxa analyzed were deleted in one of two ways in 
order to reduce the number of zero observations. The 
preferred method, useful for eight seasonally abun- 
dant taxa, was to eliminate from analysis all consecu- 
tive samples taken when the annual abundance cycle 
was lowest. In these cases, the number of survey 
dates was <57 (Fig. 3), and means and confidence 
bounds presented (Table 1) apply to the "season of 
abundance". The second method, used for four 
sporadically abundant taxa, was to include only those 



from Myers (1972, table A-12). 



l; Barnett, A. M., A. E. Jahn, and P. D. Sertic. 1980. Long term 
average spatial patterns of zooplankton off San Onofre and their 
relationship to the SONGS cooling system, ME CO 1380994. Marine 
Ecological Consultants of Southern California, 531 Encinitas 
Boulevard, Encinitas, CA 92024. 



Table 1. -Geometric mean abundance (no./100m 3 ) with 96% confidence bounds (C.B.) for the 15 larval fish taxa 
showing statistically significant cross-shelf patterns off San Onofre, Calif. Groups of strata which differ significantly in 
mean abundance are ranked from highest to lowest. Refer to Figure 3 for locations of these groups. 



Mean abundance: 


Highest 






















Lowest 






















Strata groups: 




1 






2 






3 






4 




95% C.B.: 


Lower 


Mean 


Upper 


Lower 


Mean 


Upper 


Lower 


Mean 


Upper 


Lower 


Mean 


Upper 


Gibbonsia sp. A 


0.15 


0.35 


0.66 


0.00 


0.01 


0.03 














Senphus politus 


5.47 


22.71 


91.93 


1.15 


2.67 


5.85 


0.19 


0.67 


1.66 


0.07 


0.24 


0.50 


Gobiesox rhesodon 


0.81 


2.07 


4.84 


0.13 


0.32 


0.60 


0.00 


0.02 


0.05 








Goby Type A 


1.09 


2.70 


6.28 


0.35 


088 


1.88 


0.09 


0.28 


0.56 


001 


0.04 


0.07 


Genyonemus lineatus 


8.36 


37.21 


162.71 


0.84 


2.65 


7.46 


0.17 


0.68 


1.78 


0.04 


0.23 


0.57 


Atherinidae 


9.71 


23.11 


54.54 


1.37 


4.17 


11.83 


0.27 


0.66 


1.32 


0.00 


0.05 


0.11 


Hypsopsetta gutlulata 


0.06 


0.27 


0.63 


0.01 


0.03 


0.08 














Hypsoblennius spp. 


0.44 


1.03 


2.13 


0.03 


0.09 


0.17 














Engraulis mordax 


44.03 


88.49 


177 60 


21.78 


47.81 


104.56 


8.41 


1958 


45.19 


1.67 


393 


8.86 


Paralichthys calilormcus 


0.66 


1.81 


4.40 


0.15 


0.38 


0.75 


0.04 


0.12 


0.23 








Pleuromchlhys verlicalis 


0.04 


0.13 


0.24 


0.00 


0.02 


0.05 














Cithanchthys spp. 


0.06 


0.20 


0.40 


0.00 


0.03 


0.07 














Sebastes spp 


0.42 


1.28 


3.22 


0.07 


0.30 


0.69 


0.00 


0.03 


0.07 








Stenobrachius leucopsarus 


0.24 


0.83 


2.14 


0.01 


0.05 


0.11 















101 



DISTANCE FROM SHORE (km) 
12 3 4 5 6 7 




A. N-57 
Gibbon sia Type A 



J. N=39 

Paralichthys call forme us 



M N = 57 
Sebastes spp 




D. N = 57 
Goby Type A 



12 3 4 5 6 7 

I 1 1 1 1 I r i 



\E2g 




G N = 27 




Hypsopsetta 


guttulata \. 





10 

20 

30 

40 

50 

60 

70 
80 





10 

20 
30 
40 

50 
60 
70 
80 




10 
20 

30 

40 

50 
60 
70 
80 



12 3 4 5 6 7 





DISTANCE FROM SHORE (km) 



FISHERY BULLETIN: VOL. 82, NO. 1 
DISTANCE FROM SHORE (km) 




B. N-45 
Senphus 



3 1 2 3 4 5 


6 


7 




^"•^i^,^ 












E. N= 45 






Genyonemus /meatus 







H. N=57 
Hypsoblennius spp. 



K N = 57 
Pleuronichthys verticalis 




10 

20 
30 
40 
50 
60 
70 
80 





) 1 2 3 4 5 


6 


7 


ill 


\ln^jjjjj™ 






C. N = 43 \ 






Gobiesox rhessodon 







F N=57 
Athennidae 



L N=57 
Cithanchthys spp 



N. N=28 

Stenobrachius leucopsarusV 




1 1 1 1 1 1 ' 1 


^ , Ml Hfc 
















1. N = 57 
Engraulis mordax 






'■ 





1° 




 10 




- 20 


F 


- 30 




- 40 


r 

r- 


- 50 


a. 


- 60 


Q 


- 70 




J 80 




1° 




- 10 




- 20 


? 


- 30 


^^ 


- 40 


X 


- 50 


CL 


- 60 


UJ 
Q 


- 70 




J 80 




1° 




- 10 




- 20 


? 


" 30 




- 40 


X 


- 50 


a_ 


- 60 


UJ 

n 


- 70 




J 80 




1° 




- 10 




- 20 


? 


- 30 


— ' 


- 40 




- 50 


a 




hi 


- 60 


a 


- 70 




J 80 




1° 




- 10 




- 20 


? 


- 30 


^^ 


- 40 


X 


- 50 


a 


- 60 


LlI 

Q 


- 70 




^80 





FIGURE 3.— Cross-shelf abundance patterns for the 1 4 most common larval fish taxa off San Onofre, Calif. Shading indicates relative abun- 
dance in groups of strata differing significantly in mean abundance. Heavier shading indicates higher abundance; the darkest shading ( black) is 
used for centers of abundance with larval densities >3 individuals/400 m 3 (0.75/100 m 3 ). N = Numbers of surveys used in analysis. Geometric 
mean abundances with 957c confidence bounds for each of these groups are given in Table 1. 



dates when a taxon was present. The latter method 
was used only to obtain cross-shelf patterns; in these 
cases, mean abundances in the various parts of the 
pattern are relative numbers, and confidence bounds 
were not calculated (Table 1). 

All testing was done on the basis of abundance 
alone, without regard to the strata being grouped. 
Final groupings of strata are shown in diagrams of the 
cross-shelf transect (Fig. 3). Occasionally, non- 
abutting strata were members of the same statistical 



group. These are depicted as being physically con- 
nected when such an interpretation is reasonable. In 
all cases, shading is used to indicate groups of strata 
which differ significantly. 

RESULTS 
Cross-Shelf Patterns 

The 19 larval taxa analyzed were those which rank- 



102 



BARNETT ET AL.: DISTRIBUTION OF ICHTHYOPLANKTON OFF SAN ONOFRE, CALIF. 



ed among the 10 most abundant in any of the 5 sam- 
pling blocks. Fourteen taxa showed significant 
differences among the strata which were resolved 
into spatial patterns (Table 1, Fig. 3). Taxa with cen- 
ters of abundance nearest shore tended to be concen- 
trated in either the epibenthic or the neustonic layer. 
Of the five epibenthic taxa, four (Gibbonsia Type A, 
Seriphuspolitus,Gobiesox rhessodon, and Goby Type 
A [consisting of Ilypus gilberti and Quietulay-cauda]; 
Fig. 3A-D) had centers of abundance within 2 km of 
shore. The fifth, Genyonemus lineotus, was most 
abundant out to about 4 km (Fig. 3E). Atherinidae 
(Fig. 3F) were neustonic and most abundant within 2 
km of shore. Hypsopsetta guttulata (Fig. 3G) was 
abundant in the neustonic and midwater layers out to 
2 km. It had the most nearshore pattern of any mid- 
water taxon. Hypsoblennius spp. were concentrated 
in the neustonic and midwater layers out to about 5 
km and in the neustonic layer beyond 5 km from 
shore (Fig. 3H). 

The remaining six taxa with discernible patterns 
were all most concentrated in midwater. The centers 
of abundance of Engraulis mordax and Paralichthys 
californicus (Fig. 31, J) extended from 2 to ~5 km 
from shore, while those of Pleuronichthys verticalis, 
Citharichthys spp., Sebostes spp., and Stenobra- 
chius leucopsarus appeared to extend seaward of the 
sampling area (Fig. 3K-N). 

Five taxa (Chromis punctipinnis, Parolobrax spp., 
Porophtys vetulus, Peprilus simillimus, and Pleuro- 
nichthys ritteri) were not shown to have patterns by 
this analysis. 

Vertical Migration 

Because the basic study plan called for nighttime 
sampling, the patterns described would pertain to 
nighttime distributions. The preliminary study found 
little evidence of daily vertical migration; neverthe- 
less, we conducted a further small study of vertical 
migration to test whether the vertical component of 
the patterns remained the same during daylight 
hours. The study was conducted at two inshore 
locations (Fig. 1). A description of the vertical study 
is given in the Appendix. 

There was no indication of vertical migration at the 
8 m station, but at the 13 m station two taxa, Hyp- 
soblennius spp. and Paralichthys californicus, 
showed significant (P < 0.05) vertical shifts 
downward in the water column during the day (Fig. 
4). The low probability (0.055) of the F value for 
Gobiesox rhcssodon (App. Table 2), though higher 
than the customary rejection level of 0.05, suggests a 
daily change in vertical distribution. The data indi- 



cate this species may, like Paraclinus integripinnis in 
the preliminary study, tend to migrate or settle from 
midwater into the epibenthic layer at night. 

Onshore-Offshore Abundance 

The analysis of cross-shelf pattern assumes that lar- 
vae are uniformly distributed throughout each mid 
water stratum, an assumption that becomes in- 
creasingly untenable with depth of stratum. Layering 
of ichthyoplankton within the midwater zone will 
cause an apparent decrease in density in the seaward 
blocks, as more of the volume used in the density 
calculations comes from deeper waters where a 
species may be rare. To eliminate bias in the cross- 
shelf patterns due to inclusion of noncontributing 
depths in the density calculations, one-dimensional 
abundances were calculated based on the estimated 
number of larvae under a unit ( 1 00 m 2 ) of sea surface 
in each offshore block 



jV 



3 

z 



rtjdi 



where n = larvae/ 1 00 m- 1 in stratum i and d — vertical 
thickness of stratum i in meters (0.16 m, neustonic; 
0.50 m, epibenthic; depth of water column — 1 m, 
midwater). 

The one-dimensional patterns, which emphasize 
numbers of larvae (Table 2), provide a useful com- 
parison to the two-dimensional patterns which 
emphasize larval density (Table 1, Fig. 3). All 
epibenthic and neustonic taxa had similar onshore- 
offshore centers of abundance as determined by both 
methods. This was expected, since their cross-shelf 
abundance patterns were essentially one- 
dimensional. Gibbonsia Type A, Seriphus politus, 
Gobiesox rhessodon, Goby Type A, and Atherinidae, 
all with abundance centers within 2 km of shore in the 
two-dimensional analysis (Fig. 3), likewise had one- 
dimensional maxima shoreward of 2 km. With the 
exception of S. politus, these taxa were less than half 
as abundant beyond 2 km. Genyonemus lineatus, 
most concentrated in the epibenthic layer within 
about 4 km of shore, had a one-dimensional max- 
imum at 2-4 km but remained abundant (>V£ max- 
imum) out to ~5 km. 

Of the eight midwater taxa, only two had one- 
dimensional patterns which differed from their two- 
dimensional patterns Engraulis mordax appeared 
more abundant farther offshore in one dimension (cf. 
Table 2 and Fig. 31). The steady increase in abun- 
dance of E. mordax with distance from shore is at 
odds with its two-dimensional pattern (Fig. 31) and 



103 



Hypsoblennius spp. 

24 July 1978 
Mean Number/100 m 
65432 10 I 2345 



Hypsoblennius spp. 
30 August 1978 
Mean Number/ 100 m 
10 5 5 10 



15 25 20 



FISHERY BULLETIN: VOL. 82, NO. 1 

Hypsoblennius spp. 
22 September 1978 

Mean Number/ 100 m 
15 10 5 



I i i l 



T~ V 



1 



5 

~r 



10 




2 
4 



a 



10 
12 











' 

















Day Night 

Paralichthys californicus 
24 July 1978 
Mean Number/100 m 
2 10 12 
I 1 1 1 1 



Day Night 

Paralichthys californicus 

30 August 1978 

Mean Number/IOOm 
'10 12 3 



Day Night 



50 



Paralichthys californicus 
22 September 1978 

Mean Number/IOOm 
40 30 20 10 10 



20 




2 
4 

E . 6 

.c 

QJ 8 
Q 

10 




ILT 



Day Night 



Day Night 



Day Night 



FIGURE 4. —Average vertical abundance profiles of Hypsoblennius spp. andParalichthys californicus during the study of daily vertical migration 
off San Onofre, Calif. The depth ranges of the five sampling strata are the averages (based on four to six profiles) for each sampling period. Note 
that the horizontal (abundance) scale varies. 

TABLE 2.— Numbers of larvae under 100 m 2 of sea surface in the five sampling 
blocks, averaged over 57 cruises, off San Onofre, Calif. 



Sampling block: 


A 


B 


c 


D 


E 


Offshore limits (km): 


0.5-1.1 


1.1-1.9 


1.9-3.7 


3.7-54 


5 4-7 2 


Gibbonsia Type A 


6.4 


103 


1.5 


03 


1.1 


Senphus pohtus 


273.9 


103 9 


217.9 


118.9 


93 7 


Gobiesox rhessodon 


46 


12.1 


5.3 


1.1 


30 


Goby Type A 


24.5 


17.5 


3 5 


2 9 


1.1 


Genyonemus /meatus 


1 32.7 


312.4 


623 3 


5665 


221.1 


Athennidae 


35.7 


28 1 


11.7 


89 


49 


Hypsopsetta guttulata 


3.1 


3 2 


39 


06 


7 


Hypsoblennius spp. 


27.5 


26.9 


48.1 


63.0 


369 


Engrauhs mordax 


9700 


1,833.4 


6.454 4 


9.2502 


10.263.5 


Paralichthys californicus 


4 3 


1 14 


90 


103 2 


42 4 


Pleuronichthys vertical's 


04 


2 3 


13 4 


36.4 


11 7 


Pleuromchthys ntten 


<0.1 


02 


56 


30 9 


13 9 


Cithanchthys spp. 


2 9 


3.5 


99 


17.9 


31 0- 


Sebastes spp. 


<0 .1 


<0 .1 


18 2 


77 7 


5186 


Stenobrachius leucopsarus 


0.1 


04 


4 4 


29 1 


106.1 


Chromis punctipmnis 








08 


66 


53 3 


Paralabrax spp. 


0.1 


08 


34 3 


97.8 


84 .1 


Parophrys vetulus 


5 


03 


0.1 


7.3 


33.6 


Peprilus simillimus 


20 


4 1 


3 6 


10.0 


17.4 



104 



BARNETT ET AL.: DISTRIBUTION OF ICHTHYOPLANKTON OFF SAN ONOFRE. CALIF. 



indicates that this species must be vertically 
stratified beyond the 45 m contour. This agrees with 
the findings of Ahlstrom (1959) in which the majority 
of E. mordax larvae occurred above 50 m. In contrast, 
Pleuronichthys veHicalis peaked in abundance at 4-5 
km rather than extending offshore as in the two- 
dimensional analysis (cf. Table 2 and Fig. 3K). This 
result may have occurred because the tests used in 
the two-dimensional analyses failed to distinguish 
between offshore blocks due to the small number 
(27) of non-zero observations for this species. 

Four of the five taxa lacking statistically significant 
two-dimensional patterns (Chromis punctipinnis, 
Paralabrax spp., Parophrys vetulus, Peprilus siml- 
imus) appeared to be most abundant beyond 4-5 km 
when considered in one dimension (Table 2). The 
fifth, Pleuronichthys ritteri, peaked in abundance at 
4-5 km from shore. 

Ontogenetic Pattern Changes 

Larvae of the three most abundant species were 
divided into size groups, which were analyzed 
separately for spatial pattern. To prevent temporal 
bias in the patterns, only 1978 data were used since 
they covered a full year. Larvae of two sciaenids, 
Genyonemus lineatus and Seriphus politus, were each 
divided into groups corresponding to developmental 
stages. Preflexion larvae, with straight notochords 
and no hypural development, were analyzed 
separately from more fully developed, and pre- 
sumably more mobile, flexion and postflexion larvae. 
Hypural development was found to begin at 3.8 mm 
for G. lineatus and at 4.1 mm for S. politus. Similarly, 
Engroulis mordax larvae were divided into early and 
late developmental stages, but this was done on the 
basis of size alone and did not correspond to flexion 
of the notochord. Early preflexion larvae (<6 mm), 
termed "early stage", were analyzed separately from 



other larvae, termed "late stage". One hundred lar- 
vae or all specimens, whichever was less, were mea- 
sured for each species in each collection. When only 
the first LOO larvae were measured, the proportions 
of the various size classes were applied to the total. 

To examine the ratio of older to younger larvae, the 
total number in each sampling block (Fig. 1) was 
calculated, using a longshore dimension of 1 m, i.e., 
number in block, 

N b = N  L, 

where N is number under 1 00 m 2 of sea surface in the 
block, and L is the onshore-offshore extent of the 
block in hundreds of meters. 

The patterns of all three species were more 
nearshore and epibenthic for older larvae (Table 3, 
Fig. 5). The ratio of older to younger larvae was about 
1:2 for all three species (transect totals, Table 4). 
This ratio increased in the shoreward blocks for G. 
lineatus and 8. politus, reaching maxima in blocks A 
and B. The ratio of older to younger E. mordax larvae 
was maximum in blocks C and D. The remarkable 
aspect of the E. mordax data is that there were far too 
few eggs in the nearshore zone to account for the 
numbers of larvae. The ratio of total E. mordax lar- 
vae to eggs was about 28:1. The median size of the 
larvae was about 6 mm, corresponding to an average 
age of roughly 10 d (Methot and Kramer 1979). 
Zweifel and Lasker (1976) found a time to hatching of 
2.5 d (at about 16°C). The ratio of 10-d-old larvae to 
eggs thus has an upper limit of the order 4:1 in the 
absence of mortality, implying at least a sevenfold 
excess of larvae in these nearshore samples. The 
minimum diameter of E. mordax eggs during the 
months of maximum egg abundance is about twice 
the mesh opening of the plankton nets used, so that 
sampling deficiencies for these immobile objects 
should be negligible. 



Table 3.— Geometric mean abundance (no. 100m 3 ) with 95% confidence bounds (C.B.) for younger and older age 
groups of larvae of EngrauHsm&rdax, Genyonemus lineatus, and Seripkus politus, showing statistically signifi- 
cant cross-shelf patterns off San Onofre, Calif. Groups of strata which differ significantly in mean abundance 
are ranked from highest to lowest. Refer to Figure 5 for locations of these groups. 



Mean abundance" 
Strata groups: 




























1 






2 






3 






4 




95% C.B.: 


Lower 


Mean 


Upper 


Lower 


Mean 


Upper 


Lower 


Mean 


Upper 


Lower 


Mean 


Upper 


Engraults mordax 


























early stage larvae 


2 33 


13.21 


7006 


52 


3.31 


16.24 


022 


1.10 


3.63 








late stage larvae 


23.43 


62 42 


165.65 


5.53 


14.34 


36.60 


0.92 


2.99 


8.72 








Genyonemus lineatus 


























Preflexion stage larvae 


1 55 


7 42 


3240 


73 


3 15 


11 52 


33 


1.11 


2.93 


004 


26 


0.64 


flexion and postflexion 


























stage larvae 


7.46 


3088 


125.51 


53 


1 56 


3 97 


10 


062 


1.90 


02 


0.08 


0.14 


Senphus politus 


























preflexion stage larvae 


58 


1.37 


2.90 


15 


0.49 


1.12 


004 


16 


0.33 








flexion and postflexion 


























stage larvae 


4 31 


2064 


95 57 


0.50 


1 90 


5.86 





0.10 


0.22 









105 



FISHERY BULLETIN: VOL. 82, NO. 1 



DISTANCE FROM SHORE (km) 
12 3 4 5 6 7 




Engrauhs mordax 
eggs 






10 

20 

30 

-|40 
50 
60 
70 

J 80 



DISTANCE FROM SHORE (km) 
12 3 4 5 6 




Engrauhs mordax 
early stage larvae 



DISTANCE FROM SHORE (km) 
12 3 4 5 6 7 
— i 1 1 1 1 1 1 




Engrauhs mordax 
late stage larvae 




10 

20 £ 

30" 

- 40 fE 

^°£ 
60 o 

70 

J 80 




Genyonemus hneatus\ 
preflexion stage larvae 





10 
20 
30 
40 
50 
60 
- 70 
J 80 




Genyonemus hneatus% 
flexion and postflexion 
stage larvae 




10 

20 £ 
30- 

- 40 f 

-50 a- 

60 £ 
H70 

80 



Senphus pohtus 
preflexion stage larvae 




o 

10 
20 
30 
40 

- 50 

- 60 

- 70 
J 80 




Senphus politus 
flexion and postflexion 
stage larvae 




10 

20 E 
30 — 

40 f 

50 Si 

60 q 
70 

80 



FIGURE 5. — Changes with development stage in the cross-shelf abundance patterns of Engraulis mordax, Genyonemus lineotus, and Seriphus 
politus off San Onofre, Calif. Shading indicates relative abundance in groups of strata differing significantly in mean abundance. Heavier shad- 
ing indicates higher abundance; the darkest shading (black) is reserved for densities >3 individuals/400 m 5 (0.75/100 m 3 ). Geometric mean 
abundance and 95' i confidence bounds for each group are given in Table 3. 



Table 4.— Early life stages of Engraulis mordax, Genyonemus lineatus, and Seriphus politus, for 
1978 off San Onofre, Calif. See Figure 1 for description of sampling blocks. 







Sampl 


ng block (avg. no./m 


of coastline) 




Total 


Species 


A 


B 


C 


D 


E 


no 


Engraulis mordax 














eggs 


3,100 


25.334 


85,372 


75.238 


95.782 


284,826 


larvae <6 mm 


85,302 


363.002 


1,387,638 


1,770,549 


1,770.939 


5,377,430 


larvae >6 mm 


42,970 


86,977 


816.941 


1,164,41 7 


607,805 


2,719,1 10 


Genyonemus /meatus 














preflexion larvae 


463 


688 


7.440 


9.290 


3.724 


21,605 


flexion and post- 














flexion larvae 


464 


2,969 


4,198 


2.699 


107 


10,437 


Senphus politus 














preflexion larvae 


592 


490 


4.200 


2.137 


2,103 


9,522 


flexion and post- 














flexion larvae 


2.214 


809 


779 


197 


96 


4,095 



DISCUSSION 

The methods we have employed for sampling very 
shallow inshore waters, though not without short- 
comings, have proven satisfactory in that they clearly 
emphasize the degree to which many larval fishes are 
concentrated in different layers, especially near bot- 
tom. Any quantitative sampling of nearshore fish lar- 
vae over soft bottom (at least) in the Southern 
California Bight must clearly include the epibenthic 
layer. However, our method of doing so may leave 



room for improvement. The Auriga net probably 
does not sample the 17 cm immediately above the 
substrate, unless the rollers induce an avoidance re- 
sponse such that larvae swim upward and into the 
mouth. Moreover, we have not determined the thick- 
ness of the epibenthic microhabitat or whether it is 
the same for all species. The sharpness of some abun- 
dance patterns suggests this layer may be no more 
than 1 m thick (the bongo net tows began about 1 m 
above the bottom), but small errors in this deter- 
mination, and failure to sample obliquely from the 



106 



BARNETT ET AL.: DISTRIBUTION OF ICHTHYOPLANKTON OFF SAN ONOFRE, CALIF. 



top of the range of the epibenthic gear, could make 
large differences (by a factor of 2) in the abundance 
estimates of some taxa. 

Other studies from the Southern California Bight 
have shown cross-shelf patterns similar to those 
which we describe. For example, Gruber et al. (1982; 
sampling neuston and midwater) and Brewer et al. 
(1981; sampling the entire water column) both 
showed vertical and cross-shelf changes in species 
composition. In both studies, atherinid larvae were 
principally neustonic. Brewer et al. (1982) took 69% 
of all larvae on their surveys from the epibenthic 
stratum. Both studies showed that clinids, most 
gobiids, sciaenids, and atherinids were most pre- 
valent nearer shore. Such inshore-offshore patterns 
have also been shown further north along the west 
coast (Pearcy and Meyers 1974; Richardson and 
Pearcy 1977). 

Icanberry et al. (1978) conducted a distributional 
study of ichthyoplankton above the epibenthic 
stratum at two nearshore stations off Diablo Canyon, 
about 100 km northwest of the Southern California 
Bight. Though there is taxonomic overlap between 
their study and ours, their sampling was too 
nearshore to delimit the offshore extent of any 
species in our study. Published data on widely 
(offshore) ranging species are contained in the 
CalCOFI atlas series (Kramer and Ahlstrom 1968; 
Ahlstrom 1969, 1972; Ahlstrom and Moser 1975) and 
complement some of the offshore patterns report- 
ed here. 

Engraulis mordax, one of these widely ranging 
species, spawns principally offshore (Richardson 
1981; Brewer and Smith 1982). The number of 
excess E. mordax larvae (over those which can be 
accounted for by eggs) in the nearshore zone must 
come from outside the sampling area, and these lar- 
vae must begin moving shoreward at an early age. 
Richardson (1981) suggested that currents might be 
a mechanism through which larvae of the northern 
subpopulation of E. mordax are redistributed. We 
presently cannot identify a mechanism for the redis- 
tribution off San Onofre. However, if one assumes it 
involves some behavioral response to environmental 
cues, it is worth considering just how far a larval 
anchovy might swim. Hunter (1972) estimated cruis- 
ing speed on the order of one-half body length/s. At 
this speed, a 6 mm larva would swim about 250 m/d, 
far enough to move several kilometers along an 
environmental gradient during the larval period. Any 
behavior allowing larvae to remain in the nearshore 
zone (e.g., orientation toward the bottom), once 
encountered, could help explain their observed 
concentration. 



The increased concentration of older larvae of E. 
mordax, Genyonemus lineatus, and Seriphus politus 
nearshore and near the bottom is reminiscent of the 
invasion and retention of larval and postlarval fishes 
in estuaries and tidal creeks of the Atlantic coast (cf. 
Chao and Musick 1977; Weinstein et al. 1980). Older 
larvae of Paralichthys californicus, although too rare 
for statistical analysis, also appeared more concen- 
trated nearshore than did the younger larvae. 
Whatever the mechanisms for such ontogenetic 
redistribution, they must be at least partly 
behavioral. Weinstein et al. (1980) found vertical 
movements in response to tides, whereby postlarvae 
became more concentrated near the bottom during 
ebb flows, thus taking advantage of the slower 
seaward current in the boundary layer. In the 
Southern California Bight the mean nearshore flow is 
alongshore, with relatively weak cross-shelf com- 
ponents (Hendricks 1977; Reitzel 1979 11 ; Parrish et 
al. 1981; Winant and Bratkovich 1981). The major 
source of cross-shelf water motion is internal waves 
of tidal frequency (Winant and Olson 1976) which 
propagate toward shore. For these waves to pro- 
pagate, the water column must be stratified. It is not- 
able that larval S. politus, which displayed the most 
intense ontogenetic redistribution, is most abundant 
during late summer-early fall (Walker et al. foot- 
note 11), the season of maximum thermal 
stratification in the Bight (Cairns and Nelson 1970). 
Thus it may be that S. politus and other semi- 
planktonic organisms of the shallow shelf waters take 
advantage of internal tides in somewhat the same way 
that the estuarine fauna use the surface tide to regu- 
late position. It is conceivable that due to dissipation 
of energy, seaward motions in the boundary layer are 
slower than shoreward motions. 

A similar internal wave mechanism for shoreward 
migration has been suggested by Norris (1963). He 
hypothesized that postlarval Girella nigricans might 
swim ahead of the cold waters of the incoming inter- 
nal wave fronts, thus producing the observed early 
shoreward migration of that species. 

Brewer and Smith (1982) estimated that the num- 
bers of E. mordax larvae spawned in the nearshore 
waters were approximately proportional to the area 
the nearshore waters represented in the total waters 
inhabited by the central subpopulation. They con- 
cluded that the nearshore region off southern 



"Reitzel, J. 1979. Physical/chemical oceanography. In Interim 
report of the Marine Review Committee to the California Coastal 
Commission. Part II: Appendix of technical evidence in support of 
the general summary. MRC Document 79-02(11), p. 6-23. Marine 
Review Committee of the California Coastal Commission, 631 
Howard Street, San Francisco, CA 94105. 



107 



FISHERY BULLETIN: VOL. 82, NO. 1 



California was not a preferred habitat for adult 
spawning during 1978-80. Our ratios of E. mordax 
eggs to early larvae support this conclusion. 

On the other hand, larval survivorship may be 
enhanced in these nearshore waters. Hjort (1914), 
Lasker (1975), and Brewer and Smith (1982) pointed 
out that the number of eggs and larvae surviving to 
recruitment may vary independently of spawning 
stock size. Brewer and Smith (1982) indicated that 
the shallow coastal region's importance as a nur- 
seryground for E. mordax is not yet clear. Their pre- 
liminary length-frequency data show relatively high 
numbers of large size classes nearshore, which are 
rare further offshore. Our preliminary length- 
frequency data corroborate this. The onshore 
ontogenetic shift of these larvae is a conspicuous and 
persistent feature of our data set (fig. 5). Thus 
nearshore environmental conditions may enhance 
growth or survivorship or both fori?, mordax larvae as 
well as for other larvae with typically inshore 
patterns. 

The larval taxa discussed in this paper represent 
some 12' ? of the types identified in the course of this 
study. Less common taxa were omitted for statistical 
reasons, but inspection of the data suggests that the 
patterns of abundance shown here are typical. Lar- 
vae of many species found in our study are most 
abundant in shallow water within a few kilometers 
from shore. Laroche and Holton (1979), noting the 
inshore abundance of 0-age Parophrys vetulus off the 
Oregon coast, suggested a nusery function for those 
open, nearshore areas. Concentration of juvenile 
fishes well inshore of adult depth ranges is also well 
known along the southern California coast (Lim- 
baugh 1961; Feder et al. 1974). 

Whether such patterns result from behavioral 
mechanisms leading to nearshore concentration, 
from differential onshore-offshore mortality, or sim- 
ply from random movements away from very 
localized spawning sites, their evolution and main- 
tenance imply significant value in occupying 
nearshore waters. Eppley et al. (1978) found higher 
concentrations of phytoplankton inshore of the 50- 
100 m depth contours, and Lasker (1975, 1978) 
showed that nearshore abundance of suitable-sized 
phytoplankton can be an important determinant of 
year-class strength in E. mordax. Gruber et al. (1982) 
noted that Pacific sardine, Sardinops caeruleus, once 
spawned over wide areas of the California Current 
region, but the reduced stock now concentrates its 
spawning effort nearshore. They suggested the pro- 
ductive nearshore zone may be especially important 
to recovering fish stocks, a situation which might 
apply to northern anchovy at some future date. 



Pearcy and Myers (1974) noted that a number of 
studies found estuaries of northern California and 
Oregon to be important nurseries. However, 
estuaries in the Southern California Bight, as along 
much of the Pacific coast of North America, are small 
and far between. Enhanced productivity in the 
shallow waters of the open coast seems to provide a 
nursery area for many Southern California fishes 
analogous to the estuarine nurseries of other 
regions. 

ACKNOWLEDGMENTS 

This paper is a result of research funded by the 
Marine Review Committee (MRC), Encinitas, Calif. 
The MRC does not necessarily accept the results, 
findings, or conclusions stated herein. 

We are indebted to Jeffrey M. Leis for his important 
contributions to all parts of the preliminary study and 
to the field and laboratory aspects of the main study. 
Susan Watts provided invaluable assistance in the 
computer analysis of the data. Keith Parker and 
Allen Oaten assisted with the statistical problems 
encountered. Paul Smith offered many helpful sug- 
gestions on a manuscript dealing with the pre- 
liminary study. Edward DeMartini, H. J. Walker, Jr., 
and Robert J. Lavenberg read earlier versions of this 
manuscript and offered useful comments. The paper 
has also benefitted from the comments of an 
anonymous reviewer. Judy Sabins, Carolyn Davis, 
and Karen Lee typed the various versions of the 
manuscript. We especially wish to thank the many 
technicians who spent long hours in the collection 
and processing of samples. 

LITERATURE CITED 

Ahlstrom, E. H. 

1959. Vertical distribution of pelagic fish eggs and larvae off 
California and Baja California. Fish Wildl. Serv., Fish. 
Bull, U.S. 60:107-146. 

1965. Kinds and abundance of fishes in the California 
Current region based on egg and larval surveys. Calif. 
Coop. Oceanic Fish. Invest. Rep. 10:31-52. 

1969. Distributional atlas of fish larvae in the California 
Current region: jack mackerel, Trachurus symmetricus, 
and Pacific hake, Merluccius productus, 1951 through 
1966. Calif. Coop. Oceanic Fish. Invest. Atlas 11, xi + 
187 charts. 

1972. Distributional atlas of fish larvae in the California 
Current region: six common mesopelagic fishes — Vin- 
ciguerria lucetia, Triphoturus mexicanus, Stcnobrachius 
leueopsarus, Leuroglossus stilbius, Bathylagus wesethi, and 
Bathylaxus miuilrnsis. 1955 through 1960. Calif. Coop. 
Oceanic Fish. Invest. Atlas 17, xv + 306 charts. 

Ahlstrom, E. H., and h. G. Moser. 

1975. Distributional atlas of fish larvae in the California 



108 



BARNETT ET AL.: DISTRIBUTION OF ICHTHYOPLANKTON OFF SAN ONOFRE, CALIF. 



Current region: flatfishes, 1955 through 1960. Calif. 
Coop. Oceanic Fish. Invest. Atlas 23, xix + 207 charts. 
Brewer, G. D., R. J. Lavenberg, and G. E. McGowen. 

1981. Abundance and vertical distribution of fish eggs and 
larvae in the Southern California Bight: June and October 
1978. In R. Laskerand K. Sherman (editors), Symposium 
on the early life history of fish. Introduction and back- 
ground, Woods Hole, April 1979, Vol. 178, p. 165- 
167. Rapp. P.-V. Reun. Cons. Int. Explor Mer. 

Brewer, G. D., and P. E. Smith. 

1982. Northern anchovy and Pacific sardine spawning off 
southern California during 1978-1980: preliminary' obser- 
vations on the importance of the nearshore coastal re- 
gion. Calif. Coop. Oceanic Fish. Invest. Rep. 23:160-171. 

Brown, D. M., and L. Cheng. 

1981. New net for sampling the ocean surface. Mar. Ecol. 
Prog. Ser. 5:225-227. 

Cairns. J. L., and K. W. Nelson. 

1970. A description of the seasonal thermocline cycle in 
shallow coastal water. J. Geophys. Res. 75:1 127-1131. 
CHAO, L. N., AND J. A. MUSICK. 

1977. Life history, feeding habits, and functional morphology 
of juvenile sciaenid fishes in the York River estuary, 
Virginia. Fish. Bull.. U.S. 75:657-702. 

Eppley, R. W., C. Sapienza, and E. H. Renger. 

1978. Gradients in phytoplankton stocks and nutrients off 
southern California in 1974-76. Estuarine Coastal Mar. 
Sci. 7:291-301. 

Feder, H. M.. C. H. Turner, and C. Limbaugh. 

1974. Observations on fishes associated with kelp beds in 

southern California. Calif. Dep. Fish Game, Fish Bull. 

160, 144 p. 
Gruber, D., E. H. Ahlstrom, and M. M. Mullin. 

1982. Distribution of ichthyoplankton in the Southern 
California Bight. Calif. Coop. Oceanic Fish. Invest. Rep. 
23:172-179. 

Hendricks, T. J. 

1977. Coastal currents. In Southern California Coastal 
Water Research Project, Annual Report, p. 53-62. El 
Segundo, Calif. 

H.JORT, J. 

1914. Fluctuations in the great fisheries of northern Europe 
viewed in the light of biological research. Rapp. P.-V. 
Reun. Cons. Int. Explor. Mer 20:1-228. 
Hunter, J. R. 

1972. Swimming and feeding behavior of larval anchovy 
Engraulis mordax Fish. Bull., U.S. 70:821-838. 
ICANBERRY. J. W., J. W. WARRICK, AND D. W. RlCE, JR. 

1978. Seasonal larval fish abundance in waters off Diablo 
Canyon, California. Trans. Am. Fish. Soc. 107:225-233. 

Kramer, D., and E. H. Ahlstrom. 

1968. Distributional atlas of fish larvae in the California 
Current region: northern anchovy, Engraulis mordax 
Girard, 1951 through 1965. Calif. Coop. Oceanic Fish. 
Invest. Atlas 9, xi + 268 charts. 
Ladell, W. R. S. 

1936. A new apparatus for separating insects and other 
arthropods from the soil. Ann. Appl. Biol. 23:862-879. 
Laroche, W. A., and R. L. Holton. 

1979. Occurrence of 0-age English sole, Parophrys vetulus, 
along the Oregon coast: an open coast nursery area? 
Northwest Sci. 53:94-96. 

LASKER, R. 

1975. Field criteria for survival of anchovy larvae: The rela- 
tion between inshore chlorophyll maximum layers and 
successful first feeding. Fish. Bull., U.S. 73:453-462. 



1978. The relation between oceanographic conditions and lar- 
val anchovy food in the California Current: Identification 
of factors contributing to recruitment failure. Rapp. P.-V. 
Reun. Cons. Int. Explor. Mer 173:212-230. 

Limbaugh, C. 

1961. Life-history and ecologic notes on the black 
croaker. Calif. Dep. Fish Game 47:163-174. 
Methot, R. D., and D. Kramer. 

1979. Growth of northern anchovy, Engraulis mordax, larvae 
in the sea. Fish. Bull., U.S. 77:413-423. 

Morrison, D. F. 

1976. Multivariate statistical methods. 2d ed. McGraw- 
Hill, N. Y., 415 p. 

Mcrphy, G. I., and R. I. Clutter. 

1972. Sampling anchovy larvae with a plankton purse 
seine. Fish. Bull., U.S. 70:789-798. 

Myers, J. L. 

1972. Fundamentals of experimental design. 2d ed. Allyn 
and Bacon, Boston, 465 p. 
Norris, K. S. 

1963. The functions of temperature in the ecology of the per- 
coid fish Girella nigricans (Ayres). Ecol. Monogr. 33:23- 
62. 
Parrish, R. H., C. S. Nelson, and A. Baki n. 

1981. Transport mechanisms and reproductive success of 
fishes in the California Current. Biol. Oceanogr. 1:175- 
203. 
Pearcy, W. G., and S. S. Myers. 

1974. Larval fishes of Yaquina Bay, Oregon: A nursery 
ground for marine fishes? Fish. Bull., U.S. 72:201-213. 
Richardson, S. L. 

1981. Spawning biomass and early life of northern anchovy, 
Engraulis mordax, in the northern subpopulation off 
Oregon and Washington. Fish. Bull., U.S. 78:855-876. 
Richardson, S. L., and W. G. Pearcy. 

1977. Coastal and oceanic fish larvae in an area of upwelling 
off Yaquina Bay, Oregon. Fish. Bull., U.S. 75:125-145. 

Smith, P. E., R. C. Counts, and R. I. Clutter. 

1968. Changes in filtering efficiency of plankton nets due to 
clogging under tow. J. Cons. Int. Explor. Mer 32:232-248. 

Smith, P. E., and S. L. Richardson. 

1977. Standard techniques for pelagic fish egg and larva sur- 
veys. FAO Fish. Tech. Pap. 175, 100 p. 
Snedecor, G. W., and W. G. Cochran. 

1967. Statistical methods. 6th ed. Iowa State Univ. Press, 
Ames, 593 p. 

SOKAL, R. R., AND F. J. ROHLF. 

1969. Biometry. The principles and practice of statistics in 
biological research. W. H. Freeman, San Franc, 776 p. 

Weinstein, M. P., S. L. Weiss, R. G. Hodson, and L. R. Gerry. 

1980. Retention of three taxa of postlarval fishes in an inten- 
sively flushed tidal estuary. Cape Fear River, North Car- 
olina. Fish. Bull., U.S. 78:419-436. 

WlNANT, C. D., AND J. R. OLSON. 

1976. The vertical structure of coastal currents. Deep-Sea 
Res. 23:925-936. 
Winant, C, D., and A. W. Bratkovich. 

1981. Temperature and currents on the Southern California 
shelf: a description of the variability. J. Phys. Oceanogr. 
11:71-86. 

Winer, B. J. 

1971. Statistical principles in experimental design. 2d 
ed. McGraw-Hill, N.Y., 907 p. 
Zweifel, J. R., AND R. Lasker. 

1976. Prehatch and posthatch growth of fishes— a general 
model. Fish. Bull., U.S. 74:609-621. 



109 



APPENDIX 1 



FISHERY BULLETIN: VOL. 82, NO. 1 



On 24 July, 30 August, and 22 September 1978, 
vertically stratified samples were taken at one station 
along the 8 m isobath and at another along the 13 m 
isobath. A sample set, or profile, consisting of five 
strata was sampled at each station: Neustonic, three 
midwater strata, and the epibenthic layer (the mid- 
water strata were chosen with regard to the depths of 
power plant cooling structures). At the 8 m station, 
the midwater strata were 1) the lower 3 m of the water 
column, 2) 3 m above the bottom, and 3) the water 
column above stratum 2. At the 13 m station, the 
lower midwater stratum was the lower 2 m of the 
water column, while the upper two depended on the 
vertical thermal structure. When a thermocline was 
present, as during the September cruise and inter- 
mittently during the August cruise, the middle 
stratum extended from 2 m above the bottom to the 
base of the thermocline, and the upper stratum from 
the top of the thermocline to just below the surface. 
In the absence of a well-defined thermocline, the 
water column above 2 m from the bottom was divided 
into two equal parts. Sample sets were replicated 
four to six times in the day and again at night, result- 
ing in 325 samples in the vertical migration study. 

Data from the two stations were analyzed 
separately, since all sampling depths (except the 
neustonic layer) differed between stations. No 
analysis was done of the effects of the thermocline, 
since its extreme movements with respect to the ver- 
tical scale of interest would require a more intensive 
sampling program. In this analysis nominal sampling 
depths were treated as constants. 

Because of patchy distributions of ichthyoplankton 
and movements of the thermocline (August and Sep- 
tember), inherent variability was expected among 
the sets of profiles taken on a given date. In order to 
separate this variability from variability due to sam- 



pling date (cruise), time of day, and "error", we 
analyzed the data in a repeated-measures type 
analysis of variance design (App. Table 1). In this 
design, the depth effect was contained within the 
fixed-effect time of day and the random-effect cruise. 
The questions addressed were 1) whether there was a 
depth effect, i.e., significant differences among 
strata, within cruise X time-of-day blocks, and 2) if a 
depth effect did exist, whether there was a significant 
depth X time-of-day interaction. This interaction, 
interpreted (when significant) as daily vertical migra- 
tion, was evaluated as the F-ratio of the depth X time 
of day to the depth X time of day X cruise mean 
square errors. When the three-way term was 
insignificant (in this case, P > 0.75), the error sums 
of squares and the three-way sum of squares were 
pooled, and this pooled term was used as the 
denominator in the F-ratio (Sokal and Rohlf 1969: 
266). 

The 10 most frequently occurring taxa were 
analyzed (App. Table 2). (A high frequency of 
occurrence was important to keep cell variances 
relatively homogeneous.) To reduce the effect of day- 
night differences in apparent abundance (most likely 
from visual net avoidance), we reduced each profile 
to a set of differences, or A's between adjacent strata, 
e.g. 

A, = (abundance at depth 1) — (abundance at 

depth 2). 

Abundance was expressed as log,,, (A' +1), where X 
= larvae/1 00m'. Any daily change in the relative 
abundance in two strata would thus be manifest in a 
change in sign and/or magnitude of the correspond- 
ing A. 



Appendix Table 1 .— ANO VA model applied in the analysis of daily vertical 
migration. The last two terms can form the error estimate (e) in Appendix 
Table 2. 



i/ km ' 
where V, 



+ CT M + CD 



I'* 



+ TD m + DP [mlllJk] + CTD (l/k) + e, lkm 



i /km 
M 
c , 

oT 

OP trnt,,lk) 

crb m 

i/km 



Density 

Mean effect 

Sampling date (cruise) effect 

Time-of-day effect (day-night) 

Depth profile within cruise and time-of-day 

Depth effect 

Interaction, cruise X day-night period 

Interaction, cruise X depth 

Interaction, day-night period X depth 

Depth k for profile m within cruise and time-of-day 

Interaction, cruise X day-night period X depth 

Residual error 



110 



BARNETT ET AL.: DISTRIBUTION OF ICHTHYOPLANKTON OFF SAN ONOFRE. CALIF. 

Appendix Table 2. — F-table for the 1 most frequently occurring larval fish taxa off San 
Onofre, Calif.: repeated-measures ANOVA. D = depth, TD = day-night period X depth, 
CTD = cruise X day/night period X depth. When the CTD mean square error (MSE) was 
insignificant (P > 0.75), the CTD and Error (e) sums of squares were pooled. The TD 
interaction term, when significant (*P < 0.05; **P < 0.0 1 ). is interpreted as daily vertical 
migration. Frequency refers to the number of samples in which a taxon occurred out of 
325 total samples. Results are presented for the 13 m station only. 



Taxa 



Freq. 



Source 



df 



MSE 



Engraulis mordax 



Senphus polttus 



Hypsoblenmus spp 



251 



232 



206 



Genyonemus hneatus^ 148 



Cheilotrema salurnum 1 44 



Menticirrhus undulatus 125 



Paralabrax spp. 



122 



Paratichthys califormcus 1 1 9 



Gibbonsta Type A 



Gobiesox rbessodon 



114 



113 



D 


3 


0.79065 


4.101 


0.010 


TD 


3 


18989 


0335 


0.801 


CTD 


6 


056687 


2 940 


0.013 


f 


69 


0.19281 






D 


3 


1.721 10 


8 943 


<o.oor 


TD 


3 


006644 


045 


0.986 


CTD 


6 


1 46510 


7.613 


0.000 


f 


69 


0.19246 






D 


3 


3.93932 


19 586 


<0.001 


TD 


3 


2.37801 


12 344 


0000' 


CTD 


6 


09508 


0473 


0.826 


£ 


69 


201 13 






D 


3 


1 56249 


11 853 


<0.001 


TD 


3 


0.82790 


5 377 


100 


CTD 


3 


15403 


1.168 


0332 


f 


45 


13183 






D 


3 


1 15593 


13 838 


<0.001 


TD 


3 


003663 


0.145 


0.929 


CTD 


6 


025185 


3 015 


011 


f 


69 


008353 






D 


3 


2.52228 


31 968 


<0.001 


TD 


3 


03333 


0294 


0.829 


CTD 


6 


1 1325 


1.435 


0.214 


£ 


69 


007890 






D 


3 


1.86818 


15 383 


<0.001 


TD 


3 


0.60921 


1 802 


0.247 


CTD 


6 


0.33768 


2 781 


0018 


f 


69 


0.12144 






D 


3 


2 97375 


15 609 


<0.001 


TD 


3 


76462 


4.873 


0047 


CTD 


6 


15653 


0822 


0.557 


f 


69 


19051 






D 


3 


1 25337 


25 265 


<0.001 


TD 


3 


020486 


1 742 


0.258 


CTD 


6 


0.11763 


2 371 


0.039 


c 


69 


0.04961 






D 


3 


3 52655 


103 984 


<0.001 


TD 


3 


17248 


4,528 


0055 


CTD 


6 


03806 


1 122 


359 


f 


69 


003391 







1 The analysis of G. /meatus differed from those of other taxa At the 1 3m station, the 24 July cruise was 
eliminated because the extremely low abundance of G lineatus on that date caused the variance to be unac- 
ceptbly heterogeneous (by inspection! 



Ill 



RING DEPOSITION IN THE OTOLITHS OF 
LARVAL PACIFIC HERRING, CLUPEA HARENGUS PALLASI 



Michael D. McGirk 1 



ABSTRACT 

The first normal ring in the sagittae of Pacific herring, Clupea harengus paltasi, larvae is deposited at the age 
of complete yolk absorption. The rates of deposition of subsequent rings in four groups of larvae that were fed 
daily ranged from 0. 1 2 to 0.96 rings per day, and only two of the four groups had a daily pattern. Larvae that 
were starved from hatch deposited one normal ring on day 6 posthatch, but all ring deposition stopped 
thereafter. The starvation of subgroups of larvae after 7 days of feeding and after 25 days of feeding produced 
deposition rates that were not significantly different from those of the parent feeding groups. The average 
rates of normal ring deposition were positively correlated with the average rates of growth in length. Daily 
ring deposition in herring larvae <20 mm long occurs in populations with an average growth rate equal to or 
higher than 0.36 mm per day. 



Rings or increments in the otoliths of fishes have 
been used to age wild larvae of several species 
(Ralston 1976; Kendall and Gordon 1978; Methot 
and Kramer 1979; Townsend and Graham 1981; 
Lough et al. 1982; Victor 1 982). This method has two 
assumptions: 1) The first ring is deposited at a fixed 
age in each species, and 2) the rate of ring deposition 
is constant at 1 ring/d. Evidence from studies of ring 
deposition in enclosure-reared larvae of the Atlantic 
herring, Clupea harengus harengus, (Geffen 1982; 
Lough et al. 1982); northern anchovy, Engraulis mor- 
dax, (Brothers et al. 1976); and English sole, 
Parophrys vetulus, (Laroche et al. 1982) indicates 
that these two assumptions may not be true in first- 
feeding larvae that are starving or growing slowly. 
The deposition of subsequent rings may be 
significantly < 1 ring/d. This paper reports that the 
first ring is deposited at a fixed age in herring larvae 
and that this age is coincidental with the age at com- 
plete yolk absorption. It also confirms that the subse- 
quent rate of deposition is not always daily but that it 
is positively correlated with the rate of growth in 
body length. 

MATERIALS AND METHODS 
Experimental Groups 

The batch experiments reported here were part of a 
research program on culturing Pacific herring larvae. 
Several different container sizes, temperatures, and 
prey types were employed (Table 1). Six groups of 



Table 1.— The experimental groups of Pacific herring larvae and 
their rearing conditions. 







Rearing 








Tank 


Temper- 








volume 


ature 


Feeding 


Food 


Group 


(I) 


(°C) 


treatment 


organisms 


1980A 


50 


12.1 


fed from day 2 


Anemia 


1980B 


50 


12.1 


starved from day 7 


none 


1980C 


50 


12 1 


starved from hatch 


none 


1980D 


50 


7 


starved from hatch 


none 


1981 A 


1.000 


8-9 


fed from hatch 


Artemia. plankton 


1981B 


2.000 


9-10 


fed from hatch 


plankton 


1982A 


25 


8-9 


fed from hatch 


Artemia 


1982B 


25 


8-9 


starved from day 30 


none 



'Institute of Animal Resource Ecology, University of British 
Columbia, Vancouver, B.C., Canada V6T 1W5. 



Pacific herring larvae were reared from the egg: Four 
were fed daily from hatch, one was starved from 
hatch, and one was terminated 3 d after hatch before 
food was offered. Two additional starving groups 
were formed from subgroups that were removed from 
feeding tanks after 7 d of feeding and after 25 d of 
feeding and then starved to death. 

Rearing Conditions 

Three groups, 1980A, 1980C, and 1980D, were 
raised from eggs in 50 1 circular aquaria in April-June 
1980. The eggs were laid on the walls of a holding 
tank by adult herring that had been captured in the 
Strait of Georgia by personnel of the Pacific Biologi- 
cal Station, Nanaimo, B.C. Therefore, the eggs came 
from the lower east coast stock (Taylor 1964). After 
1 4 d incubation at 7°C, the eggs were hatched and the 
larvae of 1980A and 1980C were transferred to the 
rearing aquaria. The mean (±1 SD) temperature of 
these tanks during the rearing period was 12.1° + 
0.9°C. The 1980A group was fed from hatch to the 



Manuscript accepted September 1983. 
FISHERY BULLETIN: VOL. 82, NO. 1, 1984. 



113 



FISHERY BULLETIN: VOL. 82, NO. 1 



end of the experiment, and the 1980C group was 
starved from hatch to death. The 1980D group was 
reared at 7°C for 3 d, before it was accidentally de- 
stroyed. A fourth group, 1980B, was formed from a 
1980A subsample and was placed in its own 50 1 
aquarium after 7 d of feeding and then starved for 
5d. 

Two groups, 1981A and 1981B, were reared in 
April-May of 1981 at the Bamfield Marine Station, 
Bamfield, B.C. The eggs came from spawn laid on 
Fucus spp. in the intertidal zone of Toquart Bay, 
Barkley Sound, B.C. Therefore, the eggs came from 
the lower west coast stock. The first population, 
1981 A, was raised in a 1,000 1 circular, flow-through 
aquarium. The water temperature rose gradually 
from 8° to 9°C over the rearing period. Food was 
added daily. Group 1981B was reared in a culture 
chamber suspended in Bamfield Inlet. The chamber, 
a 2,000 1 circular tank (Marliave 1981), floated at the 
surface of the Inlet. Wild plankton was swept through 
louvres on one side of the chamber by tidal currents 
and was trapped in the chamber where it served as 
food for the larvae. During the first 3 wk , no herring 
larvae was in the plankton from the Bamfield Inlet, so 
the tank population was not contaminated with wild 
herring. The surface water temperature of the Inlet 
over the rearing season was 9°-10°C. 

One group, 1982 A, was reared from eggs in the 
laboratory at the Bamfield Marine Station, April- 
May 1982. The group grew in a 25 1 rectangular 
aquaria cooled to 8°-9°C. The eggs came from spawn 
laid on eelgrass, Zostera marinus, in the intertidal 
zone at the head of Bamfield Inlet and, therefore, 
came from the lower west coast stock. The fish were 
fed from hatch to age 30 d, and then the survivors 
were moved to another tank of the same size where 
they were starved for 8 d. This subgroup was 
named 1982B. 

The lighting for all the laboratory groups was 
fluorescent and was on a 10-h light: 14-h dark cycle. 
This cycle was cued to the natural photoperiod with 
light sensors. Water in all of the tanks, except 1981A 
and 1981B, was gently aerated with an airstone, and 
about one-third of the volume was replaced daily with 
fresh seawater. Dead organisms and feces were daily 
siphoned off the floor in all tanks, except 198 IB 
which did not accumulate wastes because its floor, 
drilled with over 1,000 small holes, was self-cleaning. 

Hatching 

All larvae in any single group were hatched within 
24 h of each other. In 1980, hatching was stimulated 
by scraping the eggs off the wall of a holding tank. In 



1981 and 1982, hatching was stimulated by exposing 
late-stage eggs to air for 1 5 min. The exposure caused 
an explosive hatch when the eggs were returned to 
seawater. The egg masses were removed from the 
tanks <24 h after hatching began. 

Food 

Food for three of the four fed populations consisted 
of freshly hatchedArtemia nauplii. One of the feeding 
groups, 1981B, fed exclusively on wild plankton 
swept into the chamber by tidal currents. Another 
group, 1 98 1 A, was raised on a diet of Artemia nauplii 
supplemented with wild plankton captured with a 
plankton net from the surface of Bamfield Inlet, In all 
feeding groups, food was first supplied either at 
hatch or before the second day after hatch, the day 
when Pacific herring larvae first begin to exhibit feed- 
ing behavior. Both the Artemia nauplii and the wild 
zooplankters were attracted to the overhead light, 
and they tended to cluster in a patch at the surface of 
the water. Enough food organisms were added each 
day to the feeding groups to maintain the patches at 
all times so that the larvae of these groups had the 
opportunity to feed at will at any time. It is not known 
whether the 198 IB larvae in the culture chamber had 
a similar opportunity, but the relatively high growth 
rate of this group indicates that food was 
abundant. 

Absence of food organisms in the water of starving 
groups was ensured by filtering seawater through a 
layer of glass wool before it was added to a tank. Sam- 
ples of filtered water were examined under a micro- 
scope to verify the absence of food organisms. 

Samples 

Samples of 10-18 larvae were taken from each of the 
groups at intervals of 2-20 d. In 1980 the fish were 
frozen at -10°C, and in 1981 and 1982 they were 
preserved in 377c isopropyl alcohol. The standard 
length was measured from the tip of the snout to the 
end of the notochord with the vernier scale of a com- 
pound microscope. Some of the larvae were 
measured live before preservation, stored in- 
dividually, and then measured again 1-6 mo later. 
Freezing caused a mean (±1 SD) percent shrinkage 
of 6.3 + 3.5 (n = 26), and isopropyl alcohol caused a 
mean (± 1 SD) percent shrinkage of 0.04 ± 3.2 (n = 
97) which was not significantly different from 0% 
shrinkage (t = 0.0124, df = 96, P > 0.9). An examina- 
tion of the individual percent shrinkages showed no 
trend with live standard length. Frozen lengths were 
corrected to live lengths by multiplying by the factor 



114 



McGURK: PACIFIC HERRING OTOLITH RINGS 



1.063. Alcohol-preserved lengths did not require 
correction. 

Ring Counting 

After extraction from the skull the sagittae were 
placed on a glass slide under immersion oil; their 
diameters were measured with an ocular micrometer. 
Sagittae are slightly flattened spheroids in young lar- 
vae and tend to become more oval in shape as the fish 
grows. The diameter measured was always the long- 
est axis of the otolith. The sagittae were photo- 
graphed at 400-1, 000X, the developed film was 
projected on a screen, and the rings were counted. A 
single ring consisted of a dark band and an adjacent 
light band. All rings, no matter how faint, were coun- 
ted in order to avoid observer bias towards a daily 
ring pattern. Two classes of rings were observed: 1) A 
group of 1-5 thin, faint rings clustered about the 
nucleus surrounded by 2) wider, darker rings that 
composed the majority of the rings in most larvae. In 
some sagittae the second class of rings were 
separated from the first by a distinct ring which may 
have been a check deposited in response to the 
exhaustion of the yolk. The two classes could not 
always be clearly distinguished, particularly in slow- 
growing fish. The first class corresponds to Geffen's 
(1982) "yolk sac" rings and the second to her "nor- 
mal" or "regular" rings. In this paper the first class 
will be unnamed for two reasons: 1) Most of the rings 
were found in the larvae that had completely 
absorbed their yolk, so they were not exclusively 
yolk-sac rings, and 2) it has not been established that 
the two classes of rings are fundamentally different 
from each other, so the introduction of new terminol- 
ogy is premature. Geffen (1982) defined a "first 
heavy ring" that was found between the outer margin 
of the nucleus and the first normal ring. This term has 
not been used because the first normal ring was not 
always distinguishable from subsequent normal 
rings on the basis of width or darkness. 

Each sagitta was counted three times, and the mean 
of the three counts was taken as the final count of that 
sagitta. The ring count of a fish was the mean of the 
final counts of its two sagittae. The mean (± 1 SD) dif- 
ference in final counts between sagittae from the 
same fish was 1.3 ± 1.4 which was not significantly 
different from zero (t = 0.9028, df = 148, 0.4 > P > 
0.2). The sagittae of 21 large larvae (live length range 
= 14-29 mm, age range = 20-54 d posthatch) se- 
lected at random from several groups were photo- 
graphed and then fixed to a glass slide with 
cyanoacrylate glue and ground to the midplane with 
metallic lapping paper. They were rephotographed 



and recounted. The mean (± 1 SD) difference was 1.1 
± 2.0 which was not significantly different from zero 
(t = 0.5273, df = 20, 0.5 > P > 0.9). Inspection of the 
data revealed no trend of the difference with age or 
with the ring count of the nonground sagittae. 

Data Analysis 

The average rates of ring deposition and of growth 
in length were calculated as the slopes of linear pre- 
dictive regressions of mean ring number and mean 
length on age posthatch. The homogeneity of the 
variances of the means of a group was tested with 
Bartlett's test (Sokal and Rohlf 1969), and, if they 
were found to be heterogenous, each mean was 
weighted with its sample size divided by its variance. 
T-tests were used to test the significance of differ- 
ences between the slope of a regression of mean ring 
number on age and 1 ring/d and ring/d. F-tests were 
used in covariance analyses to test for significant dif- 
ferences between two slopes. 

RESULTS 

Growth in live standard length was positive in all 
groups except 1980C and 1980B, in which the starv- 
ing larvae shrank (Fig. 1). There are indications that 
growth was curvilinear, especially in 1980A and 
198 IB where the growth rates between the two last 
sampling dates in each group were much less than the 
previous growth rates. However, linear growth was 
assumed for the purpose of obtaining average growth 
rates to compare with the average ring deposition 
rates (Table 2). Growth rate was highest in the 2,000 1 
culture chamber and lowest in the 25 1 aquarium, and 
there was a positive but nonsignificant correlation 
between growth rate and container size in the four fed 
groups (n = 4, r = 0.90, 0.05 > P > 0.10). 

Thin, faint rings of the first class were found in the 
otoliths of most of the 1980 fish that were < 14 mm 
long, but were not found in the otoliths of any 1981 
and 1982 fish (Fig. 2). These rings may have been 
deposited at any time between the late embryo and 
the postyolk-sac stage. The only sample of otoliths 



TABLE 2. — Linear regressions of mean standard length on age in ' 
groups of Pacific herring larvae. 





/-intercept 


Slope 


SE of 




No. of 






Group 


(mm) 


(mm/d) 


slope 


r 


means 


n 


df 


1980A 


10 4 


180 


0030 


097 


4 


36 


1.2 


1980B 


13.1 


-0.004 


0.019 


19 


3 


20 


1,1 


1980C 


11.2 


-0.107 


0.031 


0.90 


5 


50 


1.3 


1981 A 


82 


0.231 


0.011 


0.99 


6 


57 


1.4 


1981B 


8.4 


0290 


0.049 


0.96 


5 


60 


1.3 


1982A 


10.6 


0.090 


0.047 


0.89 


3 


38 


1.1 


1982B 


1 1.4 


0.100 


0.035 


0.89 


4 


39 


1,2 



115 



FISHERY BULLETIN: VOL. 82, NO. 1 





30 




26 




22 




ia 




14 


- — « 


10 




b 


■» — ' 


3 


T 




H 


2h 


(3 




z 


22 


HI 




1 






ia 


Q 




CC 


14 


< 




a 


10 


z 




< 


b 


\- 


rn 


in 




1980B 



1980C 




1981B 





40 60 20 



AGE (DAYS) 



Fic.i'RE 1.— Mean (± 1 SD) live standard length at age posthatch for 
seven groups of Pacific herring larvae. See Table 2 for the regres- 
sion equations. 



from yolk-sac larvae was a single sample from 1980D 
that had a mean (± 1 SD) ring count of 5.2 ± 0.8 (n = 
9) on day 1 posthatch. The rings were not observed in 
older, larger larvae; they may have been present but 
obscured by overburden over the nucleus. This 
phenomenon has been observed in the otoliths of lar- 
val largemouth bass, Micropterus salmoides, (Miller 
and Storck 1982). A group of 7-8 "prolarval rings" 
that were clustered about the nucleus at swim-up 
were visible for only 10-15 d afterward, because the 
nucleus became more opaque with age. 

The first normal ring was deposited in all groups 
including 1980C by day 6 posthatch, the day after 
complete yolk absorption. This agrees well with the 
age at first increment of 4.5 (range = 0-9 d) found for 
Atlantic herring by Lough et al. (1982) and with the 
age of 6 d found for the same species by Geffen 
(1982). This indicates that herring larvae of both 
species do have a fixed age at first increment deposi- 
tion and that it coincides with the age at complete 
yolk absorption. 

Rates of subsequent ring deposition for the four fed 




40 60 



AGE (DAYS) 

FIGURE 2.— Mean (±1 SD) ring count at age posthatch for seven 
groups of Pacific herring larvae. Open circles are total rings and 
closed circles are normal rings only. See Table 3 for the regres- 
sion equations. 



groups were not all daily, and they ranged from 0.12 
to 0.96 rings/d (Table 3); only two groups, 1980A and 
198 IB, had rates that were not significantly different 
from 1 ring/d (t = 0.5772, df = 3, 0.5 > P> 0.9 andt = 
2.0142, df = 4, 0.10 > P > 0.20, respectively). The 
1981A group had a rate that was significantly <1 
ring/d (t = 6.3465, df=5, 0.01 >P> 0.001) butalso 
significantly > (t = 10.8062, df = 5,P< 0.001) and 
the 1982 A group had a rate that was significantly < 1 
ring/d {t = 10.0228, df = 2, 0.01 > P> 0.001) andnot 
significantly >0 (t = 1.3667, df = 2, 0.20 > P > 
0.40). 
The rate of ring deposition in 1980C, the group that 
was starved from hatch, was —0.05 ring/d, which was 



Table 3. — Linear regressions of mean normal ring number on age in 
7 groups of Pacific herring larvae. 





^-intercept 


Slope 


SE of 




No. of 






Group 


(mm) 


(nng/d) 


slope 


T 


means 


n 


df 


1980A 


-4 12 


096 


0.06 


99 


4 


36 


1.2 


1980B 


2.06 


023 


0.28 


063 


3 


20 


1.1 


1980C 


2 12 


-0.05 


002 


0.83 


5 


50 


1.3 


1981A 


-931 


0.63 


0.05 


0.99 


6 


57 


1.4 


1981B 


-5 60 


0.83 


08 


099 


5 


60 


1.3 


1982A 


1.45 


12 


0.08 


0.83 


3 


38 


1.1 


1982B 


4.90 


10 


11 


053 


4 


39 


1.2 



116 



McGURK: PACIFIC HERRING OTOLITH RINGS 



not significantly different from (t = 2.2831, df = 4, 
0. 10 > P > 0.20). This indicates that the starvation of 
first-feeding larvae stopped ring production. The 
1980B group had a rate which was not significantly 
different from one of 1 ring/d (t = 2.3397, df = 2, 0.10 
> P > 0.20) and not significantly different from a rate 
of (t = 0.6989, df = 2, 0.50 > P > 0.90) or from the 
rate of its parent feeding group, 1980A (F = 5.9185, 
df = 1.3, 0.25 > P > 0.50). One reason for these 
results is that the 1980B group had only three data 
points for the regression, and the standard error of 
the slope was therefore relatively high: 122% of the 
value of the slope (Table 3). I conclude that starva- 
tion for 5 d after a feeding period of 6-7 d has no effect 
on the rate of ring deposition. The 1982B group had a 
ring depoistion rate that was not significantly dif- 
ferent from (t = 0.7843, df = 3, 0.40 > P> 0.50) and 
which was not significantly different from the rate of 
its parent feeding group, 1982A (F = 0.1352, df = 1, 
3, P > 0.75). I conclude that starvation for 8 d after a 
feeding period of about 25 d has no effect on the rate 
of ring deposition, at least not in 25 1 enclosures. 

The average ring deposition rates were significantly 
positively correlated with the average growth rates (n 
= l,r = 0.83, 0.01 > P > 0.05) (Fig. 3). The regres- 
sion of ring rate on growth rate was: 

Ring rate — 0.14 + 2.40 (growth rate). 



The residuals of this regression were not correlated 
with container size, and there was no obvious 
relationship with prey type. However, there was a 
significiant positive correlation between the 
residuals and the mean rearing temperature (n = 7,r 
= 0.83,0.01 >P> 0.02). The midpoints of the tem- 
perature range were used as an estimate of the mean 
temperature (Table 1). A regression of ring deposi- 
tion rate on growth rate and temperature increase the 
multiple r to 0.99: 

Ring rate = -1.39 + 3.36 (growth rate) 
+ 0.14 (temperature). 

These results confirm the correlation between ring 
deposition rate and growth rate found for Atlantic 
herring larvae by Geffen (1982), who interpreted the 
relationship as being curvilinear and linearized it by 
transforming both variables with logarithms. In order 
to compare the two sets of data the relationship be- 
tween ring deposition rate and growth rate was 
assumed to be linear. A covariance analysis of the 
slopes of the two linear regressions indicated that 
there was no significant difference between them at 
the 0.05 probability level. Data from this study and 
from Geffen's were pooled and a single linear regres- 
sion was calculated (n = 12, r = 0.85, P < 0.001): 

Ring rate = 0.17 + 2.12 (growth rate). 



< 

Q 
U 

z 
cr 

LU 

i- 
< 

DC 

O 

z 

DC 



1.2 



1.0 



0.8 



0.6 



0.4 



0.2 



0.0 




-0.2 

-0.1 0.0 0.1 0.2 0.3 0.4 0.5 

GROWTH RATE (MM/DAY) 

FIGURE 3. — Relationship between the average ring deposition rates 
and the average growth rates of seven groups of Pacific herring lar- 
vae. See text for regression equation. 



The influence of temperature on ring deposition rate 
could not be compared between the two data sets 
because the rearing temperature for Geffen's fish 
was not constant over the rearing period. 

Plots of fish length on otolith diameter for the seven 
populations were curvilinear, and the rate of growth 
of fish length decreased with increasing otolith 
diameter. Transforming otolith diameter with 
logarithms best linearized the data, transforming 
both variable with logarithms produced lower cor- 
relation coefficients in all groups. Thus length was 
regressed on log (otolith diameter) (Table 4, Fig. 4). 
An analysis of covariance that included all seven 
groups indicated that the slopes of the regres- 
sions were significantly different from each 
other at the 0.05 probability level. Inspection of the 
slopes and their standard errors indicated that the 
fed groups and 1980B had slopes of a similar value 
and that 1980C and 1982B had slopes of a similar 
value but that they were much lower than those of the 
fed groups. The two groups were subjected to 
separate covariance analyses, and in each group the 
slopes were found to be not significantly different 
from each other at the 0.05 probability level. The 



117 



FISHERY BULLETIN: VOL. 82, NO. 1 



Table 4.— Linear regressions of fish length on log (oto- 
lith diameter). 





y-intercept 


Slope 


SE of 








Group 


(mm) 


|mm fxm) 


slope 


r 


n 


df 


1980A 


-5.76 


1 1 57 


0.49 


097 


36 


1.34 


1980B 


-4 54 


10 77 


3 17 


0.73 


1.' 


1,10 


1980C 


2 73 


4.40 


2.10 


0.28 


52 


1.50 


1981A 


-7.50 


13.36 


043 


0.97 


57 


1.55 


1981B 


-5.50 


.' 14 


0.46 


096 


i l( ) 


1.58 


1982A 


-7 24 


12.73 


1 45 


083 


<H 


1.36 


1982B 


4.82 


5 94 


1.65 


059 


2 7 


1.25 



Lough et al. (1982); they have also been described in 
the otoliths of larval turbot, Scopthalmus maximus, 
(Geffen 1982) and Arcto-Norwegian cod, Gadus 
morhua, (Gj0saeter and Tilseth 1982). Increments 
have also been found inside the nucleus in Atlantic 
herring (Lough et al. 1982), in three species of the 
genus Lepomis, and in the Mozambique mouth- 
breeder, Tilapia mossambica, (Taubert and Coble 



Fie.l'RE 4. — Relationship of fish length to log (otolith 
diameter) for seven groups of Pacific herring lar- 
vae. Open circles in 1982 are 1982A and solid cir- 
cles are 1982B. See Table 4 for the regression 
equations. 



I 

r- 

o 

LU 

_l 

Q 

< 

Q 
Z 
< 

r- 



25 r 



20 - 



10 



5 

2 5 



20 



5 

25 



20 



10 



1980A 



St" 



1980B 



«r 



_j i i i_ 



1980C 



1981A 




1981B 




1982 




  



20 



5 



100 



250 20 



50 100 250 



OTOLITH DIAMETER (LJM) 



slope of the 1980B group was not significantly dif- 
ferent from either the feeding groups or the starving 
groups because of its high standard error. The four 
fed groups were pooled to give a single regression {n 
= 191, r= 0.95, P< 0.001): 

Fish length = 30.90 + 12.49 log (otolith 
diameter). 

The three starved groups could not be pooled 
because of the different growth and feeding histories 
of each group. 

DISCUSSION 

The first class of thin rings was found in the otoliths 
of Atlantic herring larvae by Geffen (1982) and by 



1977). In at least one species offish, the mummichog, 
Fundulus heteroclitus, these nonregular rings are 
regular daily rings that are deposited before hatching 
(Radtke and Dean 1982). The relationship between 
the number of nonregular rings, the age and size of 
the fish, and rearing conditions can only be deter- 
mined with more experimental work, particularly on 
the sagittae of embryo and yolk-sac herring. 

Presence of the thin rings in the 1980 fish and their 
absence in the 1981-82 fish was not the result of 
genetic differences between the eggs of the lower 
east coast stock and the eggs of the lower west coast 
stock. The sagittae of many small (length range = 9- 
20 mm) wild herring larvae captured from Bamfield 
Inlet in 1981 and 1982 were found to have several 
thin, faint rings around the nucleus (McGurk unpubl. 
data). It seems more reasonable to hypothesize that 



118 



McGUKK: PACIFIC HERRING OTOLITH KINGS 



the difference arose from factors that have already 
been reported to affect the rate of deposition of nor- 
mal rings. These factors include temperature 
(Taubert and Coble 1977; Marshall and Parker 
1982), short-term temperature fluctuations 
(Brothers 1978), and feeding activity (Uchiyama and 
Struhsaker 1981; Neilson and Geen 1982). Lough et 
al. (1982) suggested that the first class of thin rings 
were related to the inability of first-feeding larvae to 
meet their metabolic energy demands during the 
transition from yolk to exogenous food. This argu- 
ment implies that the 1980 herring larvae were less 
able to capture sufficient food during first feeding 
than the 1981-82 larvae. However, this hypothesis 
does not explain the presence of the faint rings in the 
1980C larvae that were starved from hatch. 

Results of this study confirm the observations of 
Geffen (1982) that the rate of normal ring production 
is not always daily in young herring larvae and that it 
is positively correlated with the rate of growth in 
body size. The correlation means that normal rings 
cannot be used with confidence to age wild herring 
larvae less than about 20 mm long, unless the average 
growth rate of the population is known to be higher 
than about 0.36 mm/d (calculated from the regres- 
sion of ring deposition rate on growth rate for Pacific 
herring only). Growth of larval fishes is influenced by 
several factors: temperature (Kramer and Zweifel 
1970), food density (Haegele and Outram 1978), and 
container size (Theilacker 1980). The tendency for 
larger containers to support higher growth rates in 
the four fed groups of this study may explain why only 
two of the four had a daily ring pattern. The correla- 
tion implies that, if the rate of growth is slowed or 
stopped by starvation after a period of feeding, then 
the rate of ring deposition should also slow or stop. 
The two experimental groups that were treated in 
this manner did not produce rings at rates that were 
significantly different from those of their parent 
feeding groups. This suggests that a starvation 
period >5-8 d is necessary in order to demonstrate a 
statistically significant effect. Larger rearing con- 
tainers are also recommended to produce greater 
contrast in growth rates between feeding and starv- 
ing fishes. 

Container size, temperature, or prey size may 
possibly have additional effects on the rate of ring 
deposition apart from that which is explained by 
growth rate. Temperature does explain some of the 
residual variance of the ring deposition rate-growth 
rate regression. However, published evidence on 
effect of constant temperature on ring deposition 
does not support the hypothesis that higher tem- 
peratures produce more increments. For example, 



Neilson and Geen (1982) found no difference be- 
tween the number of increments produced by 
juvenile chinook salmon, Oncorhynchus tshawytscha, 
reared at 5.2°C and at 11.0°C. The effects of such 
environmental factors as light, temperature, and 
prey type on the ring pattern of herring sagittae can 
only be determined with a well-controlled ex- 
perimental study. 

ACKNOWLEDGMENTS 

I gratefully acknowledge the assistance given by the 
staff of the Bamfield Marine Station. I also thank 
Gary Kingston for assistance in rearing fish in 1980, 
Jeff Marliave for advice on rearing healthy marine 
fish larvae, and Steve Campana for discussions on 
grinding and reading fish otoliths. The paper has 
benefited substantially from reviews by two 
anonymous reviewers. This study was funded by a 
G.R.E.A.T. Award from the Science Council of 
British Columbia and by a grant to N.J. Wilimovsky 
from the National Sciences and Engineering 
Research Council of Canada. 

LITERATURE CITED 

Brothers, E. B. 

1978. Exogenous factors and the formation of daily and sub- 
daily growth increments in fish otoliths. Am. Zool. 
18:631. 
Brothers, E. B., C. P. Mathews, and R. Lasker. 

1976. Daily growth increments in ototliths from larval and 
adult fishes. Fish. Bull., U.S. 74:1-8. 
Geffen, A. J. 

1982. Otolith ring deposition in relation to growth rate in her- 
ring (Clupea harengus) and turbot (Scophthalmus max- 
ima*) larvae. Mar. Biol. (Berl.) 71:317-326. 
GJ0SAETER, H., AND S. TlLSETH. 

1982. Primary growth increments in otoliths of cod larvae 
(Gadus morhua L.) of the Arcto-Norwegian cod stock. 
Fiskeridir. Skr. Ser. Havunders. 17(7):287-295. 
Haegele, C. W., and D. N. Outram. 

1978. The effects of diet and ration on the growth and 
survival of Pacific herring (Clupea harengus pallasi) lar- 
vae. Fish. Mar. Serv. Can. Tech. Rep. 767, 41 p. 
Kendall, A. W., Jr., and D. Gordon. 

1978. Growth rate of Atlantic mackerel (Scomber scombrus) 
larvae in the middle Atlantic Bight. Rapp. P.-V*. Reun. 
Cons. Int. Explor. Mer 178:337-341. 
Kramer, D., and J. R. Zweifel. 

1970. Growth of anchovy (Engraulitt mordax Girard) in the 
laboratory as influenced by temperature. Calif. Coop. 
Oceanic Fish. Invest. Rep. 14:84-87. 
Laroche, J. L., S. L. Richardson, and A. A. Rosenberg. 

1982 Age and growth of a pleuronectid, Parophrys vetulus, 
during the pelagic larval period in Oregon coastal waters 
Fish. Bull. U.S. 80:93-104. 
Lough, R. G., M. Pennington, G. R. Bolz, and A. A. 
Rosenberg. 

1982. Age and growth of larval Atlantic herring, Clupea 



119 



FISHERY BULLETIN: VOL. 82, NO. 1 



harengus L., in the Gulf of Maine-Georges Bank region 
based on otolith growth increments. Fish. Bull., U.S. 
80:187-199. 
Marliave, J. B. 

1981. Use of tidal power for culture of marine animals. Proc. 
1981 Conf. Am. Assoc. Zool. Parks and Aquariums, New 
Orleans, La.. Sept. 13-17, 1981, p. 103-105. 

Marshall, S. L„ and S. S. Parkkr. 

1982. Pattern identification in the microstructure of sockeye 
salmon {Oncorhynchus nerka) otoliths. Can. J. Fish. 
Aquat. Sci. 39:542-547. 

Methot, R. D., and D. Kramer. 

1979. Growth of northern anchovy, Engraulis mordax, larvae 
in the sea. Fish. Bull, U.S. 77:413-423. 
Miller, S. J., and T. Storck. 

1982. Daily growth rings in otoliths of young-of-the-year 
largemouth bass. Trans. Am. Fish. Soc. 1 1 1:527-530. 
NEILSON, J. D., AND G. H. GEEN. 

1982. Otoliths of chinook salmon (Oncorhynchus tsha- 
wytscha): Daily growth increments and factors influencing 
their production. Can. J. Fish. Aquat. Sci. 39:1340- 
1347. 
Radtke, R. L., and J. M. Dean. 

1982. Increment formation in the otoliths of embryos, larvae, 
and juveniles of the mummichog, Fundulus hetero- 
clitus. Fish. Bull., U.S. 80:201-215. 
Ralston, S. 

1976. Age determination of a tropical reef butterflyfish utiliz- 
ing daily growth rings of otoliths. Fish. Bull., U.S. 74:990- 
994. 



SOKAL, R. R., AND F. J. ROHLF. 

1969. Biometry. The principles and practice of statistics in 
biological research. W. H. Freeman, San Franc, 776 p. 
Taubert, B. D., and D. W. Coble. 

1977. Daily rings in otoliths of three speices of Lepomls and 
Tilapia mossambica. J. Fish. Res. Board Can. 34:332- 
341). 
Taylor, F. H. C. 

1964. Life history and present status of British Columbia her- 
ring stocks. Fish. Res. Board Can., Bull. 143, 81 p. 
Theilacker, G. H. 

1980. Rearing container size affects morphology and nutri- 
tional condition of larval jack mackerel, Trachurus sym- 
metricus. Fish. Bull., U.S. 78:789-790. 

TOWNSEND, D. W., AND J. J. GRAHAM. 

1981. Growth and age structure of larval Atlantic herring, 
Clupea harengus harengus, in the Sheepscot River 
estuary, Maine, as determined by daily growth increments 
in otoliths. Fish. Bull., U.S. 79:123-130. 

UCHIYAMA, J. H., AND P. STRUHSAKER. 

1981. Age and growth of skipjack tuna, Katsuwonnus 
pelamis, and yellowfin tuna, Thunnus albacares, as 
indicated by daily growth increments of sagittae. Fish. 
Bull., U.S. 79:151-162. 

Victor, B. C. 

1982. Daily growth increments and recruitment in two coral- 
reef wrasses, Thalassoma bifasciatum, and Halichoeres 
bu-ittatus. Mar. Biol. (Berl.) 71:203-208. 



120 



FISHES, FISH ASSEMBLAGES, AND THEIR SEASONAL 

MOVEMENTS IN THE LOWER BAY OF FUNDY 

AND PASSAMAQUODDY BAY, CANADA 



J. Stevenson Macdonald, 1 , Michael J. Dadswell, 2 
Ralph G. Appy, 3 Gary D. Melvin, 2 and David A. Methven 4 



ABSTRACT 

Five fish assemblages, dominated by pleuronectids, cottids, gadids, clupeids, and rajids, were identified 
from collections taken during a 5-year survey in the lower Bay of Fundy region, Canada. Individual assem- 
blages occurred in each of estuarine, beach, pelagic, and offshore hard- and soft-bottom habitats. Species 
and/or age-class components within assemblages varied seasonally but, in general, each assemblage was dis- 
tinct. There was a progressive seaward displacement of these assemblages from shallow, inshore to deeper, 
offshore habitats in winter followed by a reversal during summer. Yearly changes in species occurrence and 
abundance during the study period were predominantly attributable to variation in ocean climate. Long-term 
changes in abundance of two commercial species at one of the sampling sites, since a similar study there in 
1965, appear related to population fluctuations in the Bay of Fundy and the Gulf of Maine. The beach habitat 
apparently served as a major nursery area for juvenile gadids, pleuronectids, and clupeids. 



Although the fish fauna of the Bay of Fundy-Gulf of 
Maine system is well documented (Bigelow and 
Schroeder 1953; Leim and Scott 1966), few studies 
have examined long-term spatial and temporal 
changes or interrelationship among the fish assem- 
blages. Previous studies in this region were con- 
cerned with the biology and seasonal movements of a 
single species (McCracken 1959, 1963; McKenzie 
and Tibbo 1961; Wise 1962) or the occurrence and 
composition of communities at a single site (Bigelow 
and Schroeder 1939; Tyler 1971). 

Moore (1977) and Quinn (1980) have emphasized 
the need for long-term research to establish baseline 
information and estimates of natural variability for 
fisheries assessments and pollution impact studies. 
This is particularly true for inshore regions because 
of their importance as nurseries and feeding grounds 
(Warfel and Merriman 1944; Rauck and Zijlstra 
1978). The increasing interest in trophic rela- 



'Department of Zoology, University of Western Ontario, London, 
Ontario. Canada; present address: Department of Fisheries and 
Oceans, West Vancouver Laboratory, Vancouver, British Columbia, 
Canada V7V 1N6. 

'Fisheries and Environmental Sciences, Department of Fisheries 
and Oceans, Biological Station, St. Andrews, New Brunswick, 
Canada EOG 2X0. 

'Fisheries and Environmental Sciences, Department of Fisheries 
and Oceans, Biological Station, St. Andrews, New Brunswick, 
Canada; present address: Department of Zoology, College of 
Biological Sciences, University of Guelph, Ontario, Canada NIG 
2W1. 

'Fisheries and Environmental Sciences, Department of Fisheries 
and Oceans, Biological Station, St. Andrews, New Brunswick, 
Canada; present address: Department of Biology, Memorial Univer- 
sity, St. John's, Newfoundland, Canada A1B 3X9. 



tionships among entire communities of fishes is 
further reason to document movement, abundance, 
and co-existence of fishes potentially utilizing the 
same food resource (Richards 1963; Keast 1970; 
Tyler 1972; Steiner 1976; Hacunda 1981). 

Long-term changes in fish assemblages have been 
attributed to overexploitation of one or more of the 
species within the assemblage (Brown et al. 1973; 
Burd 1978; Sherman et al. 1981) and climatic 
variations (Dow 1964; Sutcliffe et al. 1977). 
However, it is usually difficult to separate natural 
fluctuations from those caused by imbalance in com- 
petitive and predator-prey relationships due to 
exploitation (Cushing 1980; Daan 1980; Sissenwine 
et al. 1982). With the view in mind of assessing these 
long-term changes to properly assign cause and 
effect, repetitive, in-depth studies of well-known or 
type localities are needed. 

This study examines spatial and temporal variation 
in fish diveristy and abundance over a 5-yr period at 
two offshore stations within Passamaquoddy Bay, 
one offshore station in the Bay of Fundy, and at 
inshore and beach stations in Passamaquoddy Bay. 
One offshore station was the same station sampled 
by Tyler (1971) during 1965-66, allowing documen- 
tation of changes that have occurred over the inter- 
vening 10-15 yr. 

METHODS 

Three offshore stations in the Bay of Fundy (B) and 
in Passamaquoddy Bay (A, C) (Fig. 1) were sampled 



Manuscript accepted August 1983. 
FISHERY BULLETIN: VOL. 82, NO. 1, 1984. 



121 



FISHERY BULLETIN: VOL. 82. NO 1 




FIGURE 1.— Passamaquoddy Bay and the adjacent Bay of Fundy indicating sampling stations 

occupied during the study. 



at approximately monthly intervals over a 5-yr 
period, 1976-81 (Table 1). Station A was the same 
site sampled by Tyler (1971) during 1965-66. Fish 
were collected using a %-35 shrimp trawl (3.8 cm 
stretch mesh nylon; 15.5 m foot rope), similar to the 
%-35 Yankee trawl used by Tyler (1971), towed by 
the 1 50-hp, 14m stern trawler, Fisheries and Oceans' 
RV Pandalus II. Tows at each station were along a 1 .6 
km transect at about 4 km/h. Stations A and B were 
sampled once per trip between 1976 and 1979, and 
station C was sampled sporadically. From 1979 to 
1981, tows at stations A and B were replicated and 
station C was sampled regularly. Captured fishes 
were identified to species, and adults and juveniles 
were categorized by size and enumerated separately. 
During the final year of collecting, fork length of all 
fishes was recorded to the nearest centimeter and 
otoliths were collected from Atlantic cod, ocean pout, 
American plaice, winter flounder, and witch flounder 
for age determination. Atlantic cod otoliths were sec- 



tioned for aging, other species were aged using the 
whole otolith. Results reported are the empirical 
length at age. 

Between June and September 1976, 12 estuarine, 
intertidal, and inshore marine stations were sampled 
within Passamaquoddy Bay and Head Harbour 
Passage (Fig. 1). In addition, station 3 was sampled 
monthly during the period May 1976-November 
1977, station 8 was sampled at approximately weekly 
intervals from May to September 1981, and stations 
1 and 10 were sampled in December 1980 (Table 1). 
Fish were collected using a 9 m, 1.3 cm mesh beach 
seine, a 3.7 m shrimp trawl with a 3 mm cod end 
towed behind a 5 m Boston whaler, or bottom-set gill 
nets with stretched mesh sizes ranging from 7.6 to 
17.8 cm. Standard fishing efforts employed with each 
gear type were shore seine hauls of 5 min during the 
2-h period before and after low water, trawl tows of 10 
min, and overnight gill net sets of 16 h. 

Temperature, salinity, and substrate type were 



122 



MACDONALD ET AL.: FISH ASSEMBLAGES AND THEIR SEASONAL MOVEMENTS 

Table 1.— Physical and chemical characteristics and sampling history of stations in the Bay 
of Fundy and Passamaquoddy Bay. Gear: ST = shrimp trawl; S = seine; GN = gill 
net. Bottom type: M = mud; Sa = sand; Rk = gravel or rock. 











Sampling 


Sampling 










Maximum 




temp. 


salinity 










depth 


Bottom 


range 


range 


Collection 


Sampling 


Station 


Gear 


(m) 


type 


(C) 


(%o) 


period 


trips 


A 


ST 


80 


M-Rk 


0-15 


29 5-32.5 


1976-81 


39 


B 


ST 


80 


M 


1-12 


31 0-32 5 


1976-81 


37 


C 


ST 


20 


M 


0-15 


— 


1978-81 


15 


1 


S 


1.5 


M-Rk 


14.5-20.0 


22.1-260 


06-08/76. 12/80 


3 


2 


S 


1.5 


M-Rk 


15.5-22.5 


26.0-29.5 


05-08/76 


4 


3 


S 


1.5 


Sa-Rk 


0-16.0 


21 0-30.0 


05/76-11/77 


16 


4a 


S 


1.5 


Rk 


12 5 


29.0 


06. 07/76 


2 


4b 


ST 


7 5 


Rk 


12.5 


— 


07/76 


1 


5a 


S 


1.5 


Sa 


14 5 


30.0 


07. 08/76 


2 


5b 


GN 


33 


Sa-M 


— 


— 


08/76 


1 


6 


S 


1.5 


Sa-M 


14.0 


28.0-300 


08-09/76 


2 


7 


GN 


30 


M 


13.5 


28.0 


06/76 


1 


8a 


S 


1.5 


M-Sa 


11.0-18.5 


28.7-30.7 


06. 07/76 
05-09/81 


2 
23 


8b 


ST 


12 


Rk-Sa 


— 


— 


06, 07/76 


2 


9 


S 


1.5 


Sa 


140 


29.5 


06.08/76 


2 


10 


GN 


3 


M-Rk 


13.0 


280 


06. 09/76. 1 2/80 


3 


1 1 


S 


1 5 


M 


— 


— 


07/76 


1 


12a 


s 


1.5 


Sa-Rk 


150 


280 


07, 09/76 


2 


12b 


ST 


15 


Sa-Rk 


— 


— 


07/76 


1 



recorded for most sampling sites (Table 1). Bottom 
temperature and salinity data inside and outside 
Passamaquoddy Bay came from routine monthly 
sampling by the Department of Fisheries and Oceans 
at a site opposite the Biological Station (near Station 
A) and at "Prince 5" 3.2 km south of Bliss Islands in 
the Bay of Fundy (near station B). Temperatures at 
deep stations were taken with a reversing ther- 
mometer attached to a Nansen bottle and at shallow 
stations with a hand thermometer. Salinities were 
determined with a laboratory salinometer from sam- 
ples collected in the field. Substrate samples at deep 
stations were obtained with a PONAR grab. At 
shallow stations, substrate type was assessed 
visually. 

Fishes were identified using Leim and Scott (1966) 
with the exception of red and white hake and redfish, 
which were determined by using Musick (1973) and 
Ni (1982), respectively. Because we were unaware of 
the problem of distinguishing between young Raja 
ocellata and R. erinacea (McEachran and Musick 
1973), these determinations may be incorrect. 

Coefficients of community were calculated using 
the formula: 

X 100 



A + B-C 



where C = number of common species, A = number 
in assemblage 1, and B = number in assemblage 2 
(Jaccard 1932; Kontkanen 1957). An index that com- 
pared presence and absence of species at each sta- 
tion (binary data) was used because species 
abundances among stations were not comparable 
due to different gear used. 



RESULTS AND DISCUSSION 

Station Environmental 
Characteristics 

Temperature and salinity at stations A and B (Fig. 
2) followed the typical, yearly cycle of a cold tem- 
perate sea (Fig. 3). Annual temperature range in the 
Bay of Fundy was less than in Passamaquoddy Bay. 
Summer temperatures at inshore sites were simlar to 
offshore sites with the exception of higher tem- 
peratures at some estuary stations (i.e., 1 and 2) 
(Table 1). Two notable variations occurred: The 
early months of 1977 and August 1978 were abnor- 
mally warm, particularly at station A ( J. Hull 5 ); and 
throughout the study period there was a generalized 
cooling trend. 

Salinities were highest in late summer through the 
fall and lowest in spring at both sites. At all times of 
year, salinities were higher in Bay of Fundy (station 
B) than at station A (Fig. 2). Inshore sites had 
salinities of 1-2 ppt less than station B, and salinities 
at estuarine sites were as low as 2 1 .0 ppt during sum- 
mer (Table 1). 

Substrates of most sites were composed of sand 
and/or mud (Table 1). Station A had the steepest 
slope, about 2: 100 m. Slopes at stations B and C were 
0.4:100 and 0.6:100 m, respectively. Slopes at coastal 
intertidal sites were gradual, about 1:100 m. Estuarine 
stations (1,2, and 10) had extremely soft mud bot- 
toms and station 2 had extensive eel grass beds. 



'Fisheries and Environmental Sciences, Department of Fisheries 
and Oceans, Biological Station, St. Andrews, New Brunswick, 
Canada EOG 2X0. 



123 



FISHERY BULLETIN: VOL. 82, NO. 1 



Figure 2.— Bottom tem- 
perature and salinities at 
station A in Passama- 
quoddy Bay and station B in 
the Bay of Fundy during 
1976-81. 



° 
a 

<D 

a 
E 

a; 



16 

14 

12 

10 

8 

6 

4 

2 



14 
12 
10 
8 
6 
4 
2 




Station A 



Temperature 




tt — n n 1 | i i — n n rrn — n n 1 | I i — n — n MM n n MM 

( Station B 




34 
33 
32 

31 
30 
29 

36 
35 
34 
33 
32 
31 
30 
29 



c 
CO 



TT TT 

M A 

1976 



M 



TT 

A 



1977 



T T 

F 



M 
1978 



I I 

A 

8 


1 
D 


1 I 

F 


M II 
M A 

1979 


! 
D 


I I 

F 


ii ii 
M A 

1980 


I 
D 


M II 
F M 

1981 



3|- 
2 



January-March 



1 


9r 



...^^V/^^-.-c 




12 
11 

9 
8 
7 
6 
8 
7 
6 
5 



October-December 



ucioper-ufluwiiioHP . « 

V „^ :/ -/ V X^A;-, 



8°C 



Average Annual 






cvJUi 



2 "C 



L L 



L 



1960 65 



70 75 



80 



Fishes and Seasonal Occurrence 

Sixty-two species of fish were captured during the 
study period (Tables 2, 3). For those stations sam- 
pled regularly and intensively, residency period and 
abundance are indicated. Fish occurrence is 
expressed as part of the summer component (June- 
October), the winter component (November-May), 
the regular component (caught year round), or the 
occasional component. Fishes classified as 
occasional components show no seasonal abundance 
pattern and were collected <70 times (specimens 
captured X sampling trips) over 5 yr at stations A, B, 
and C, or < 15 times at stations 1-12. We realize that 
some fish at all stations may have been missed 
entirely or are listed as rare simply because they were 
unavailable to the sampling gear used or could avoid 
capture (e.g., mackerel). Abundance of fishes in the 
catch was categorized as rare (1-10 specimens), com- 



FlGURE 3.— Mean surface sea temperatures at 
station B by season during 1960-80 as com- 
pared with the 50-yr mean at this site. 



124 



MACDONALD ET AL.: FISH ASSEMBLAGES AND THEIR SEASONAL MOVEMENTS 



TABLE 2. — Residency and abundance of fishes occurring at deep sampling stations in the Bay of Fundy andPassama- 
quoddy Bay. Residency is R= regular; S= summer; W= winter; = occasional; N= never encountered. Abundance 
is a = abundant; c = common; r = rare. 



Station 



Species 


A 


B 


C 


Mysine glutmosa 


Oc 


N 


N 


Squalus acanthias 


Oc 


Sa 


Oc 


Raja r ad lata 


Sc 


Re 


Or 


Fa; a sent a 


Or 


Re 


Or 


Raja ennacea 


Ra 


We 


Oc 


Raja ocellata 


Re 


Or 


Or 


Raja laevis 


N 


Or 


N 


Acipenser oxyrhynchus 


N 


Wr 


N 


Alosa Aestivalis 


N 


N 


Or 


Alosa pseudoharengus 


Or 


Or 


Sc 


Alosa sapidissima 


Or 


Se 


N 


Clupea harengus 


Wa 


Wa 


Wa 


Osmerus mordax 


Re 


Or 


Sa 


Mallotus vtllosus 


/V 


N 


Or 


Enchelyopus cimbnus 


Sr 


Re 


Or 


Gadus morhua (adult) 


Se 


Wc 


Oc 


G morhua (juvenile) 


Wa 


Wc 


Oc 


Microgadus tomcod 


Or 


Or 


Oc 


Poltachtus virens (juvenile) 


Wa 


Or 


N 


Melanogrammus aeg/efmus (adult) 


Sc 


Oc 


N 


M aeglefmus (juvenile) 


We 


N 


N 


Merluccius btlineans 


Sa 


Sa 


Sc 


Urophycis tenuis 


Sc 


Re 


Oc 


Urophycis chuss 


Sc 


Sr 


Or 


Anarhichas lupus 


Sr 


N 


N 


U/vana subbifurcata 


Wr 


N 


N 


Cryptacanthodes maculatus 


Or 


Or 


N 



Species 



Station 



A 


B 


C 


Or 


Or 


N 


Or 


N 


N 


Ra 


Sc 


Sc 


N 


Or 


N 


Or 


Or 


N 


Wc 


Or 


N 


Wr 


N 


N 


Wc 


N 


N 


Ra 


Wc 


Oc 


Wc 


Or 


N 


Wc 


Or 


N 


Re 


Re 


Or 


Oc 


Or 


N 


Wc 


N 


N 


Re 


N 


N 


Oc 


Or 


Oc 


Sa 


Wa 


Sa 


Wa 


Wr 


Sa 


Or 


Sc 


N 


Or 


Wc 


N 


Rr 


Ra 


Sa 


Sr 


Or 


Or 


N 


N 


Wc 


Wr 


N 


N 


Or 


N 


N 


Or 


Or 


Re 


Sr 


Or 


N 



Lumpenus lumpretaetormis 
Lumpenus maculatus 
Macrozoarces amencanus 
Nezumia bairdi 
Cyclopterus lumpus 
Liparis coheni 
Liparis inquilmus 
Sebasres fasciatus 
Myoxocephalus octodecemspmosus 
Myoxocephalus aeneus 
Myoxocephalus scorpius 
Hemitripterus amencanus 
Tnglops murrayi 
Artedie/lus uncmatus 
Aspidophorotdes monopterygius 
Poronotus triacanthus 
Pseudopleuronectes amencanus 
P amencanus (juvenile/ 
Glyptocephalus cynoglossus (adult) 
G. cynoglossus (juvenile) 
Hippoglossoides platessoides 
Limanda ferrugmea 
Liopsetta putnami 

Hippoglossus hippoglossus (juvenile) 
Parahchthys oblong us 
Scopthalmus aquosus 
Lophius amencanus 



TABLE 3. — Residency and abundance of fishes occurring at estuarine, intertidal, and 

shallow marine sites in Passamaquoddy Bay. Residency is R = regular; S = summer; W 

= winter; = occasional; N = never encountered (station 3 only). Abundance is a = 

abundant; c = common; r= rare. 

Station 

Species 1 23456789 10 11 12 

Squalus acanthias — — N — Sc — Sc — — 

Raja radiata (juvenile) — — Sr — Oc — — — 

Raja ennacea — — Sc — — — — — 

Raja ocellata — — N — — — — 

Alosa aestivalus — — N — — — — Sa — Sa 

Alosa pseudoharengus — — Sr — — — Sa — Sa 

Clupea harengus Sc — OSa — Sa — 

Salmo salar — — S — — — — — — 

Osmerus mordax Re Re Wa — — — Sa — Wa — — 

Mallotus villosus — — N — — — — — — — OcOc 

Fundulus heteroclitus Sa Sc Wc — — — — Or — Sc 

Gasterosteus aculeatus Sc Sa Oc — Sc — Sc 

Gasterosteus wheatlandi Oc Sa Sr — — — — Sr — Sr — — 

Apeltes quadracus — ScN — — — — — — — — 

Pungitius pungitius — SaN — — — — — — — — 

Anguilla rostrata OrScN — — — — — — Sr 

Enchelyopus cimbnus (larvae) — — Sc — — — — — — 

Gadus morhua (adult) — — N — — OSr — — 

G morhua (juvenile) — — Sr — Sc — Sc Sr — — 

Microgadus tomcod Wc Wc Wc — — Sc Wc — 

Pollachius virens (juvenile) OSa — Sa — 

Urophycis tenuis (juvenile) — — Sc — — — — Sc 

Ammodytes amencanus — — N — — — — — — 

Scomber scombrus — — Sr — Sc — Sc — — 

Pholis gunnellus — Re — — — S — 

Ulvaria subbifurcata — — N — — — — — — 

Cyclopterus lumpus (juvenile) — — Sc — — — — — 

Macrozoarces amencanus — — N — — — — — — 

Myoxocephalus octodecemspmosus — — Sr — — 00S — 000 

Myoxocephalus aeneus — — Sc — — — — Sc — — — 

Myoxocephalus scorpius — — — — — Sc — — — 

Hemitripterus amencanus Oc — Sc — — — — — — 

Menidia menidia Sa Sa Wc — — — — — — — — 

Pseudopleuronectes amencanus (adult) Sr — Sr Sa — Sa — — — 

P. amencanus (juvenile) — Sa Sa — Sa Sa — — — 

Liopsetta putnami Sc OSr — — — — — — Sc — — 

Syngnathus fuscus OScN — — — — — — — — — 



125 



FISHERY BULLETIN: VOL. 82, NO. 1 



mon(l 1-100), and abundant (+100). Because of gear 
differences further quantification of catches was 
unjustified. 

Eight species of flatfishes were captured during the 
study period (Table 2). The winter flounder, 
Pseudopleuronectes americanus , was the numerically 
dominant species (Figs. 4, 5). Juveniles (<22 cm, 2 + 
age group; Fig. 6) were abundant in shallow water 
during summer (Fig. 5) and at deep station over hard 
bottom in Passamaquoddy Bay during winter (Fig. 
4). Adult winter flounder were abundant during sum- 
mer in Passamaquoddy Bay, both inshore and 
offshore, but rare at the Bay of Fundy site (Fig. 4). 
During winter they were rare or absent inside 



Passamaquoddy Bay but common to abundant at the 
Bay of Fundy station. This pattern reflects the winter 
movement of this species into offshore water in the 
northern part of its range (Saila 1961; McCracken 
1963; Van Guelpen and Davis 1979), which is prob- 
ably triggered by temperature. Adults were seldom 
present in Passamaquoddy Bay when temperatures 
were below 6°C. McCracken (1963) found a similar 
relationship between minimum flounder catch-per- 
effort and minimum temperature. Surges of adult 
flounder abundance at offshore sites, which coin- 
cided with rapid temperature change in spring and 
fall, were evident in most years (Fig. 4; Tyler 1971) 
and may have been related to rapid onshore or 



300 



250 



200 



! 



500 



400 



o 



a. 

» 
o 

c 



o 

< 



150- 



100 



50- 




50 



i i 

411 



Pseudopleuronectes americanus 




-f^ 



r*i — i — i — r"r 



V- 



■w A + C adult means 
ES3-A* C juvenile means 
czzi-B adult means 

* -A + C no catch 

» -B no catch 



.JUjlp 



— i*^— r 



JU 



1 I I I 



^*T-nn- 4 T^*r-rn*T^4H^ t ¥*r a i^?*-P4*T»T»rT* i i i i P 4 i  P +*P-pJ 



Glyptocophalus cynoglossus / "ISJJJjJJ, 8 



'r ^ 



^JtJVJ^ 



t — rf i — i — n 



150r- 



100- 



50- 




A J A O D 
1977 




Hlppoglossoldes platessoides 



\ a 

T*1 rT^r-r-r 



A J A O D 
1978 



~r 



V I i 
A J A O D| 
1979 



I 



JUu 



F A 



J A O 
1980 



i i » i i 
F A J 

1981 



Winter 

Figure 4. — Seasonal occurrence and abundance of juvenile and adult flatfishes (Pseudopleuronectes aniiricanus, Glyptocephalus cyno- 
glossus, and Hippoglossoides platessoides) at offshore station in the Bay of Fundy and Passamaquoddy Bay. For witch flounder (middle) 
dark bars are adults and lined bars are juveniles at site B. For American plaice (bottom) open bars are site B and lined bars are site 
C (all juveniles). 



126 



MACDONALD ET AL.: FISH ASSEMBLAGES AND THEIR SEASONAL MOVEMENTS 

100 (— 

_ PollachluM vlfns 

50 



JL 



FIGURE 5. — Seasonal occurrence and abundance of 
fishes captured by seine at beach station 8 from May 
to September 1981. 



r 



JjI 



jiL 



No Catch 



10 L Urophyclt fnula 

I A A A A — « 

100- 

_ Pseudopleuronectes amerlcanus 



^h 



3 
 

« 

c 

® 50 
CO 

« 

Q. 

g o 

•S 160,- 

c 

D 

n 

< 120- 



80- 

40 


160 

120 
80 



-a — a- 
I 



-■-A ■-■ A ■- A 11 l| 



I 



Clupea harengus harengus 



"I 



A A— A r-A — A — A-A-AA-* 



L 



-A— 



40 




_ Alosa sp. 



— A A- A i-l A —   !■ 



JkhL 



_ Mlcrogadus tomcod 



-L 



May 



t" 

Jun 



1 



!■ I ill  



July 



August I Sept. 



offshore movement of the population. 

The witch flounder, Glyptocephalus cynoglossus, 
was absent or rare at all times inside Passamaquoddy 
Bay, but was a regular component at the soft-bottom 
Bay of Fundy station (Fig. 4). Catches from June to 
October consisted of large adult witch flounder (30- 
60 cm, >6+ age group), but catches from November 
to May were 6-25 cm juveniles (0-6 yr) (Figs. 6, 7). 
Adult witch flounder on the Scotian Shelf also move 
from intermediate depths (100 m) in summer to 
deeper water in winter (Powles and Kohler 1970). 
Both Powles and Kohler (1970) and Markle (1975) 
reported juvenile witch flounder from deep water 
(150-1,000 m) over hard bottom, quite unlike the 



situation we encountered except for similar tempera- 
ture regimes. Also, replacement of adults by 
juveniles during winter seems peculiar to our study, 
but may have been observed because of year- 
round sampling. 

Juvenile American plaice, Hippoglossoides pla- 
tessoides, were a major summer component of station 
C and a regular component of the Bay of Fundy sta- 
tion (Fig. 4), both soft-bottom habitats, but was only 
occasional at the hard-bottom station (A). Age-2 
plaice (6-14 cm; Fig. 6) were first captured with our 
shrimp net in April. By the following year, recruit- 
ment to the gear appears complete at an average size 
for the age-class of 17 cm (Fig. 7). Juvenile plaice are 



127 



FISHERY BULLETIN: VOL. 82, NO. 1 



60i- 



Glyptocephalus cynoglossus 



Pseudopl»uronect«s 
smerlcanus 



Hlppoglossoldas platessoldus 




Figure 6.— Fish length versus age for five fish species caught at 
stations A and B; December 1980-June 1981. Lines are fitted by 



eye. 



I^U 








** 


100 


" 






Gadus morhua s 

/ 
/ 

/ 


80 








A / 


60 








./ ^^^ 


40 


— 


/ 


/ 
/ 


/ • ,jr Macrozoarcas amerlcanus 


20 

n 


/ 


/ 

i 


1 


,i,i i,l 



6 8 10 

Age (Years) 



12 



14 16 



20r 



10 



Psaudoplauronectat 
amerlcanus 



rffTl 1 1 1 rn rh i 



a n= 83 



20 




1  r> = 2 
. d n = 128 


10 



— .ccaJllL 


I 




Glyptocaphalus 
cynoglossus 
Jun* 23, 1980 



-i 1 1 1 — I i 

Aug. 13, 1980 
 n= 32 



r 




Oct. 9, 1980 

 Dsll 


" 








T 


i , i 



Dec. 13, 1980 

 n = 17 



W — ¥— "I 1 — *i 

Jan. 15, 1981 

-  n - 16 

^ 1 1 1 1 

April 29, 1981 

 n = 2 



Hlppoglossoldas 
platassoidas 

|-|  n - 60 

i 1 — ^^* ^*T 1 1 



X 



 n = 44 
o n = 2 



 n= 27 

o n = 1 



T 1 1 




n= 77 



 n = 30 
a n = 3 



n= 76 
n= 7 



^Afc. 



— i r i 1 1 1 

20 40 60 20 40 60 20 40 60 

Length (cm) 



Figure 7.— Seasonal size distributions of flat- 
fishes (Pseudopleuronectes americanus, Glyp- 
tocephalus cynoglossus, and Hippoglossoides 
platessoides) from offshore stations in the Bay of 
Fundy and Passamaquoddy Bay, 1980 and 
1981. Shaded area is captures at station B; un- 
shaded, station A inside Passamaquoddy Bay. 



128 



MACDONALD ET AL.: FISH ASSEMBLAGES AND THEIR SEASONAL MOVEMENTS 



sedentary, soft-bottom dwellers, that exhibit little 
seasonal movement, and migration from nursery 
ground to adult stock is diffusive (Bigelow and 
Schroeder 1953; Leim and Scott 1966). However, 
some seasonal movement does occur when plaice 
leave soft-bottom, middepth habitat (30 m) for win- 
ter and return in summer (present study). Plaice were 
a regular, low-abundance component at station A in 
1965 (Tyler 1971), but we found they were virtually 
absent between 1976 and 1981. The difference may 
be attributable to the general decline of groundfish 
abundance in the Bay of Fundy after 1970 (Hare 
1977). 
Among other flatfishes, windowpane, Scoph- 
thalmus aquosus, was a regular component at station 
C and the smooth flounder, Liopsetta putnami, was 
common among the inshore-estuarine communities 
during summer (Tables 2, 3). Yellowtail flounder, 
Limanda ferruginea, was a rare member (4-5/tow) of 



the summer assemblage at station A and occasional 
at the other two deep stations. Juvenile Atlantic 
halibut, Hippoglossus hippoglossus, was a low- 
abundance member (2-3/tow) of the winter assem- 
blage at station A. The fourspot flounder, 
Paralichthys oblongus, was captured once at station 

A during the abnormally warm fall of 1978. 

Eight species of gadoid fishes were captured during 
the study (Tables 2, 3). Adult Atlantic cod, Gadus 
morhua, was an abundant member of the summer 
component at offshore sites in Passamaquoddy Bay, 
particularly station A, but was absent from there in 
winter. It was a common member of the early winter 
assemblage in the Bay of Fundy but rare thereafter 
(Figs. 8, 9). During summer, juvenile Atlantic cod 
(10-20 cm) were captured occasionally while seining 
beach sites, but were more common in gill net catches 
at intermediate depth (30 m) inshore (stations 5 and 
7; Table 3). The shallow water abundance maxima of 



400 



300 



200 



100 



Gadus morhua 



 B 

J-r4, AM X i i ^Avflj pM A 

1 R5Q0 ' 



■"■i-A+C adult means 
G5SS3-A + C juvenile means 
I i-B means 

* -A + C no catch 

A -B no catch 



P p , n,n^ , , J j , ^m^rn* p a , p A , 



' A 

E*4 | | | 



I 



200 r 




f-i*V- r A i— I 6 



^ I I I 




100 

50 



Melanogrammus aegleflnus 

-rt, , f , JU q ^ rrfWYf^ r' AM i  | ■*■ i i I'l' V^V t i [ M. i M '*' »'  ■* * i A T*P-r-rAr*i 



ULf 



i»4*+V-t* 




Urophycis tenuis - Urophycls chuss 



50 r 



-Ar-r-AV 



-r4* i'i P 'i' i i  | i*T-r-r-r* 

Enchelyopus cimbrius 




J A O 
1976 



^Tf A J A O W F A J A O D F A J A O D F A J A O D^ F A J A 
1 1977 ' 1978 ' 1979 ' 1980 1981 



Winter 

FIGURE 8.— Seasonal occurrence and abundance of gadoids at offshore stations in the Bay of Fundy and Passamaquoddy Bay, 1976-81. 

129 



FISHERY BULLETIN: VOL. 82. NO. 1 



10r 



Gadus morhua 

June 23, 1980 

 n - Bay of Fundy 

- n -. 6 Passamaquoddy Bay 



,n 



-i 1 M " l " . 1 1" 1 1 1 

Aug. 13, 1980 
10 r  n= 3 

a n = 4 



-p — i  r"-i 1 — t P-°- 



10 



Oct. 9, 1980 

 n = 2 
□ n = 4 



-i r 



r 



20 




Dae. 13, 1980 




10 


■1 10=35 






30 






1 r l i i i 
Jan. 15, 1981 


1 1 


20 


- JL ,„•:!? 




10 


- 





60 


#1 






i ' i ' I i i 


' I 


SO 


~ 






40 


_ 


April 29, 1981 




30 


- 


 n ; 3 
, on: 163 




20 


- 






10 


- 









^ <"i 








1 i r  - i  i 




20 40 60 80 100 






Length (cm) 





FIGURE 9. — Seasonal size distributions of Gadus morhua at station 
B in the Bay of Fundy and station A in Passamaquoddy Bay, 1980 
and 1981. 



young cod (0+, 1 + , <17 cm) has been previously 
reported in the western North Atlantic (Schroeder 
1930) but is not well documented. On the other hand, 
this occurrence of young cod in the North Sea is well 
known (Daan 1978). During winter, juvenile cod were 
abundant at station A or in colder winters at station B 
(Fig. 8, 1980 and 1981). Both juvenile and adult cod 
were more abundant at station A during our study 
than during 1965 (20-70/tow, Tyler (1971); 1976-81, 
50-400/tow). 

Haddock, Melanogrammus aeglefinus, were never 
abundant during our study. Adults were captured 
only at the hard-bottom station A during summer 
(Fig. 8) and juvenile haddock (1+) were occasionally 
captured at the same site in winter. Catches of had- 
dock declined from a maximum of 25/tow to <5/tow 
during the study period (Fig. 8). However, up to 260 
haddock/tow were caught at station A during 1965 
(Tyler 197 1). Decline in abundance after 1965 might 
be the cause for the collapse of the Gulf of Maine had- 
dock stock in 1970 (Hare 1977; Clark et al. 1982). 

Only juvenile pollock, Pollachius virens, were cap- 
tured during the study. Pollock of the annual year 
class (0+) were either rare or extremely abundant at 



beach sites (100+/seine haul) in a given year, 
depending perhaps, on the size of the annual year 
class. Pollock dominated beach catches during early 
summer but disappeared from this region by Sep- 
tember (Fig. 5). In years when 0+ pollock were abun- 
dant along the beach in summer, members of the 
same year class were also abundant the following 
winter at station A (1976-77, 1981) and, in summers 
of low abundance on the beach, they were correspond- 
ingly rare offshore in winter (1977-78; Fig. 8). Large 
numbers of pollock larvae were present in the 
plankton during March 1979 (Scott 1980), and we 
again encountered large number of 0+ juveniles at 
station A in the winter of 1979-80. Present findings 
suggest there may have been three large year classes 
produced during our study period, 1976, 1979, and 
1981. 

Adult white, Urophycis tenuis, and red, U. chuss, 
hakes were common summer components at offshore 
stations A and B (Markle et al. 1982). Juvenile white 
hake (<15 cm) were a summer component at beach 
stations (Fig. 5), but were rarely captured thereafter 
and only then at offshore sites in winter. Also in 1965 
few small hake were captured after December (Tyler 
1971). Apparently hake leave Passamaquoddy Bay 
in winter (Markle et al. 1982). In the present study, 
the one time hake were observed during winter was at 
station B in the Bay of Fundy (Fig. 8). 

The fourbeard rockling, Enchelyopus cimbrius, was 
a regular component at station B in the Bay of Fundy 
and occasional in summer at station A (Fig. 8). The 
mesh size of our gear was just small enough to cap- 
ture large individuals of this species, and it was prob- 
ably more abundant than indicated. Larval rockling 
were a rare summer component of inshore sites 
(Table 4). Battle (1930) and Tyler (1971) both con- 
sidered rockling a summer occasional in Passama- 
quoddy Bay, occurring there during spawning 
migration. Tyler's catch rate at station A (2-3/tow) 
was similar to ours at that site. Larger catch rates at 
station B (10-50/tow) may be due to rocklings pref- 
erence for soft-bottom habitat (Bigelow and 
Schroeder 1939). 

Silver hake, Merluccius bilinearis, was often the 
most abundant gadoid found at offshore stations dur- 
ing summer, and juveniles were a regular component 
at station B year round (Fig. 10). Large numbers of 
adult silver hake were present during fall (Fig. 10) in 
company with other migratory summer occasionals, 
including American shad, Alosa sapidissima; spiny 
dogfish, Squalus acanthias; and butterfish, Porono- 
tus triacanthus . All these fishes may carry out coun- 
terclockwise spring to fall migrations around the Bay 
of Fundy similar to the shad (Dadswell et al. 1983). 



130 



MACDONALD ET AL.: FISH ASSEMBLAGES AND THEIR SEASONAL MOVEMENTS 



The Atlantic tomcod, Microgadus tomcod, was a summer (Fig. 5) and in estuaries in early winter 
regular component of the inshore assemblage and (Table 3). 

was particularly abundant at beach sites during early Clupeids and osmerids made up a major portion of 



Table 4.— Catch of fishes at intertidal seining station 3 (Brandy Cove) during period May 1976-November 1977. Fish captured during 

three 5-min seine hauls (100 X 15 m) (|j] = juvenile; [1| = larvae). 



Species 



1976 



1977 



15/05 14/06 13/07 18/08 15/09 10/10 08/12 15/02 20/03 10/04 30/05 29/06 15/07 18/09 10/10 17/11 



fla/a radiata [j] 

R. ennacea [j] 

Alosa pseudoharengus 

Clupea harengus 

Salmo salar [j] 

Osmerus mordax 

Fundulus heteroclftus 

Gasterosteus aculeatus 

G wheatlandi 

Enchelyopus cimbnus [I] 

Gadus morhua [j] 

Microgadus tomcod 

Pollachius virens [|| 

Urophycis tenuis [)) 

Scomber scombrus 

Pho/is gunnel/us 

Cyctopterus lumpus 

Myoxocephalus aeneus 

M scorpius [|] 

M octodecemspinosus [j] 

Hemitnpterus amencanus 

Pseudopleuronectes amencanus 

Liopsetta putnami 

Memdia menidia 



25 



— 5 



51 
1 



1 
115 



2 3 — 

132 15 12 

1 2 — 

- - -observed never captured - 
— 1 2 — — 



3 

1 

10 



2 — 



2 — — 



— — 15 



3 1 



4 1 








1 
1 


I 


1 


2 


— 


— 


2 


I 


15 


4 
10 


— 


— 


4 




- 


3 


- 


6 


4 


5 


26 


32 
3 


5 

1 


27 
3 
2 


3 




8 
2 


— 


3 


2 


— 


— 


3 


6 


1 1 


— 


— 


- 


4 


1 
2 

1 
12 


2 
1 

1 


3 
2 


1 
3 


2 


1 


- 


22 


8 


3 
2 


1 


1 
2 





— 


2 


— 


— 


2 


4 



300 
200 
100 

.2 o 

I 150 



■o 100 



1000 ^.eso 



Clup»a harengus 

harengus 



~^*J — I — l^^n 



L 



i-A +C means 

3-B means 
-A + C no catch 
-B no catch 



-P-P-r-i'M-P-^T* ' i i | i^ I i i*t*P^*i»* \ j * i^ 



600r> g600 



< 



50 



Merlucclus bilinear!* 



, , | ,lif , a- a ^A-^+ 




I I I I ' M I I l*f 



1200 700 

I I 



t-| " i  "  M < A rt l[*4- 



40 



20 



Squalus acanthlas 



 i   



J AODFAJ A0DFAJ AOD|FAJ AOD|FAJ A0D|FAJ 
1976 1977 1978 1979 1980 1981 

Winter 



FIGURE 10.— Seasonal occurrence and abundance of pelagic fishes and dogfish at offshore stations in the Bay of Fundy 

and Passamaquoddy Bay, 1976-81. 



131 



FISHERY BULLETIN: VOL. 82, NO. 1 



the fishes caught at inshore sites (Table 2). At beach 
station 8, alewives, Alosa pseudoharengus; Atlantic 
herring, Clupea harengus harengus; and American 
smelt, Osmerus mordax; appeared in mid-July and 
increased in abundance during August (Fig. 5). Her- 
ring were abundant in estuaries during summer and 
were replaced there by smelt in winter (Table 3). 
Large American smelt were present at offshore sites 
in Passamaquoddy Bay in mid-summer as observed 
by Tyler (1971). During most winters, juvenile Atlan- 
tic herring (10-20 cm) were abundant at offshore 
sites, particularly inside Passamaquoddy Bay at 
intermediate depths (station C; Fig. 10). Catches 
were variable, possibly because of schooling 
behavior (Brawn 1960). Tagging experiments indi- 
cate herring move from inshore during summer to 
deeper water in winter (McKenzie and Tibbo 
1961). 
Six species of sculpin (Table 2) were commonly 
encountered at offshore station of which two — 
longhorn sculpin, Myoxocephalus octodecemspino- 
sus, and sea raven, Hemitripterus americanus — were 
abundant, regular components (Fig. 1 1). Juveniles of 



most species were common at beach sites in summer 
(Table 4) and at station A in winter (Table 2). 
Increases in abundance of longhorn sculpins at sta- 
tion B during winter were observed (Fig. 1 1) and may 
be the result of migration out of Passamaquoddy 
Bay. Two small species, Arctic hookear sculpin, 
Artediellus uncinatus, and mailed sculpin, Triglops 
murrayi, were winter occasionals at station A. They 
were perhaps more abundant than catch rates 
indicated (2-5/tow) because their maximum size 
range was at the lower limit of catchability for our 
trawl. 

The blennioid-like fishes were represented by 
seven species (Tables 2, 3) of which ocean pout, 
Macrozoarces americanus, was regular at offshore 
stations in Passamaquoddy Bay (Fig. 11), and rock 
gunnel, Pholis gunnellus, was a regular component at 
beach sites (Table 4). Ocean pout abundance in 
Passamaquoddy Bay was generally highest in early 
summer and declined thereafter (Tyler 1971; Fig. 
11). Abundance of ocean pout usually increased at 
station B in late summer and fall, suggesting move- 
ment from Passamaquoddy Bay to the Bay of Fundy. 



200f : 



150 



100 



o 50 



a. 



u 

c 

CO 

■o 50 

a 

< 



O^T 

100 r 



500 
I 



-i — i — i — ►" 



i i P" A 



360 
I 



u 



n f 



Myoxocephalus octodecemspinosus 

430 

h-A + C means 
z=i-B means 

* -A + C no catch 

* -B no catch 



ul 



,I»I|A + f\- 



•^4 i*l ^ ivU- 



j 



i fn i i 



200 

i 



Hemitripterus americanus 



A-t-i-jlV v P < Pi i  a M *i , i . ^ n i^i > i ^ T  i [ P. i i .'i*4 «i AM i j « P i 



r* P Mi 1M-1M 



-P^r-r-rV 



50- 

J A o d| f a j 



JAODFAJA 
1976 1977 



Ma i , " 



O D| F 



Macrozoarces americanus 



4-lM, 



^ffl 



A J A '6 'd| 'f' A J A 



1978 



1979 



ODJFAJ A O D| F A J 



1980 



1981 



Winter 



FIGURE ll.— Seasonal occurrence and abundance of sculpins and ocean pout at offshore stations in the Bay of Fundy and Passama- 
quoddy Bay, 1976-81. 



132 



MACDONALD ET AL.: FISH ASSEMBLAGES AND THEIR SEASONAL MOVEMENTS 



It may be a response to avoid warm temperatures 
(Olsen and Merriman 1946). Movement of ocean 
pout is generally thought to cover only short dis- 
tances (Orach-Maza 1975; Sheehy et al. 1977). 

Other blennioids occurred infrequently at station A 
(Table 2). Selectivity of our shrimp trawl may have 
been a factor in these low catches. One species, 
radiated shanny, Ulvaria subbifurcata, which was 
thought to be rare in Passamaquoddy Bay (Leim and 
Scott 1966), was often captured (5/tow) at station A 
during winter. Scuba searches during summer 
revealed radiated shanny were abundant inshore, 
under rocks in 6-9 m of water (Dadswell and Melvin, 
pers. obs.). 

Five species of skate were captured during the 
study (Table 2): Two species, thorny skate, Raja 
radiata, and smooth skate, R. senta, were common, 
regular components of the offshore site in the Bay of 
Fundy; two little skate, R. erinacea, and winter skate, 
R. ocellata, were regular components of station A in 
Passamaquoddy Bay; and one species, the barndoor 
skate, R. laeuis, was encountered occasionally at sta- 
tion B. The species cooccurrences of skates and their 
habitat selection are as described by McEachran and 
Musick (1975). Some seasonal movement into 
Passamaquoddy Bay was exhibited. Abundance of 
smooth and thorny skates at station A increased dur- 
ing summer and declined after late fall. Juveniles of 
thorny, little, and winter skates were often captured 
at beach sites during summer (Table 3). 

Several smaller fishes were captured at inshore 
sites only, but again this may be an artifact of sam- 
pling gear. Threespine stickleback, Gasterosteus 
aculeatus, was a regular component at most beach 
sites (Table 4). Other sticklebacks were more or less 
confined to estuarine areas (Table 3). Mummichog, 
Fundulus heteroclitus, and Atlantic silversides, 
Menidia menidia, occurred mainly in estuaries during 
summer but were part of the winter community at 
beach sites (Table 4). 



Assemblages and Diversity 

Species assemblages in the study area varied 
according to site and season. If juveniles and adults 
of some dominant species are considered as separate 
taxonomic units (Table 2), calculated coefficients of 
community show similarity between similar habitat 
types (e.g., soft bottom) at a given season, and be- 
tween the summer assemblage of one habitat and the 
winter assemblage of the next seaward habitat 
(Table 5). In general, movement of assemblages was 
from inshore in summer to offshore in winter with 
some return movement in spring (Fig. 12). Some 
species, however, exhibited a partial reverse of this 
pattern (Atlantic tomcod, ocean pout). 

Specific groupings of fish were segregated among 
the available habitats according to season. The "es- 
tuarine" assemblage was dominated by warmwater, 
euryhaline species, including sticklebacks, Atlantic 
silversides, mummichogs, and juvenile clupeids. 
Most of this group moved to adjacent, inshroe marine 
habitat in winter (Tables 3,4), but Atlantic tomcod 
and American smelt moved in the reverse direction to 
form a winter estuarine group (Table 3). 

The summer "beach" assemblage consisted of 
regulars such as threespine stickleback and rock gun- 
nel and a summer component including juvenile 
gadids, juvenile sculpins, flounders, and juvenile 
alosids. Juvenile gadids (pollock, white hake, and 
Atlantic tomcod) were most abundant in early sum- 
mer but were replaced by steadily increasing num- 
bers of clupeids in late summer (Fig. 5). Numerous 
other postlarval and juvenile fishes, including four- 
beard rockling and lumpfish, Cyclopterus lumpus, 
appeared in the beach zone during the summer 
(Table 3). In late fall, most of this assemblage left the 
beaches and occupied offshore sites in Passama- 
quoddy Bay. Atlantic herring concentrated at the 
soft-bottom station C and the gadids, sculpins, and 
winter flounder (juveniles) at the hard-bottom sta- 
tion A. Threespine stickleback and rock gunnel 



Table 5. — Coefficients of community among seasonal fish assemblages in the lower Bay of Fundy. 













Se 


award 












Estuarine 


Estuarine 


Beach 


Beach 


C 


C 


A 


A 


B 


B 




winter 


summer 


winter 


summer 


winter 


summer 


winter 


summer 


winter 


summer 


Estuarine winter 


— 


10 


200 


70 


00 


12.5 


00 


2.0 


0.0 


00 


Estuarine summer 


— 


— 


50.0 


12.5 


20.0 


0.0 


0.0 


00 


00 


0.0 


Beach winter 


— 


— 


— 


66 


14 3 


3 8 


42 


0.0 


0.0 


0.0 


Beach summer 


— 


— 


— 


— 


6 6 


33.3 


36.1 


17 3 


21 2 


0.0 


C winter 


— 


— 


— 


— 


— 


12.5 


4 2 


5 7 


6.6 


00 


C summer 


— 


— 


— 


— 


— 


— 


48 


400 


40.0 


47.0 


A winter 


— 


— 


— 


— 


— 


— 


— 


20.9 


43.0 


26.3 


A summer 


— 


— 


— 


— 


— 


— 


— 


— 


36.4 


42 8 


B winter 


— 


— 


— 


— 


— 


— 


— 


— 


— 


258 



133 



FISHERY BULLETIN: VOL. 82, NO. 1 



Estuarine 



Station A 



Winter Community 

Pollock (juvenile) 

Cod (juvenile) 

Haddock (juvenile) 

Winter flounder (juvenile) 

Herring (adult) 

Regular Community 

Sea raven 
Little skate 
Longhorn sculpin 
Ocean pout 

Summer Community 

Cod (adult) 

Haddock (adult) 

Winter flounder (adult) 

Thorny skate 

Silver hake 

White hake 

Fourbeard rockling 





11 



Gulf of Maine- 
Scotian Shelf 



Winter Community 

Witch (adult) 
Cod (adult) 
Haddock (adult) 
Silver hake 
Dogfish 
White hake 




Beach 



Winter Community 

Silversides 

Mummichog 

Regular Community 

3-spine stickleback 

Tomcod 

Rock gunnel 

Summer Community 

Pollock (juvenile) 
Cod (juvenile) 
White hake (juvenile) 
Winter flounder 

(juvenile, adult) 
Herring (juvenile) 
Sea raven (juvenile) 




Winter Community 

Tomcod 
Smelt 

Summer Community 

Herring (juvenile) 
Sticklebacks 
Mummichog 
Silversides 
Smooth flounder 
American eel 



Station C 




Winter Community 

Herring (juvenile, adult) 

Summer Community 

Plaice 
Silver hake 
Winter flounder 
Ocean pout 



Station B 



Winter Community 

Winter flounder (adult) 

Witch (juvenile) 

Longhorn sculpin 

Herring 

Atlantic sturgeon 

Regular Community 

Plaice 
Sea raven 
Thorny skate 
Smooth skate 
Silver hake 
Fourbeard rockling 

Summer Community 

White hake 
Witch (adult) 
Dogfish 
Ocean pout 
American shad 



FIGURE 12. — Communities of fishes occurring at each site divided into summer component (SC), winter component 
(WC), and regular component (RC). Arrows indicate direction of seasonal movement. 



remained at beach sites over winter and were joined 
by Atlantic silversides and mummichog to form a 
winter assemblage (Table 4). 

During summer an "offshore, hard-bottom" 
assemblage consisting of adult gadids (Atlantic cod, 
haddock, white and red hake), adult flounders (win- 
ter yellowtail), ocean pout, adult sculpins, and skates 
assembled inside Passamaquoddy Bay. Sea raven, 
longhorn sculpin, ocean pout, and little skate 
remained at this site over winter and were joined by 
juvenile fishes from the beach zone. The other 



species apparently move to offshore sites in the Bay 
of Fundy and/or to the Scotian Shelf (McCracken 
1959; Wise 1962; Edwards 1965; Kulka and Stobo 
1981). 

The "offshore, soft-bottom" assemblage consisted 
of American plaice, witch flounder, white hake, four- 
beard rockling, and skates as described by Bigelow 
and Schroeder (1939). This group at station B was 
the most stable assemblage studied and had the 
largest regular component. Conversely, similar 
assemblages which occurred at the shallower, soft- 



134 



MACDONALD ET AL.: FISH ASSEMBLAGES AND THEIR SEASONAL MOVEMENTS 



bottom station C were the most seasonally dynamic 
(Fig. 12). Adult witch flounder and most hakes left 
station B in winter for grounds further offshore in the 
Gulf of Maine (Powles and Kohler 1970; Kulka and 
Stobo 1 98 1), and this site was occupied by adult win- 
ter flounder and longhorn sculpin, perhaps from 
inside Passamaquoddy Bay or other adjacent 
inshore sites (McCracken 1963). 

Superimposed on the two offshore, essentially 
benthic fish assemblages was a seasonal semipelagic 
component. In summer, silver hake was the 
numerically dominant species. During fall, diversity 
increased with the arrival of spiny dogfish, butterfish, 
and American shad. In winter, Atlantic herring 
numerically dominated the pelagic component at all 
offshore sites (Fig. 12). 

Diversity, expressed simply as number of species 
captured, varied appreciably at beach sites during 
the year. Diversity was 2-5 species in winter-spring, 
9-13 species in summer, and 4-6 species in fall-winter 
(Fig. 13). Total number of species captured at 
inshore sites was 35, compared with 51 species cap- 
tured at offshore sites. 

Diveristy of assemblages at deep offshore sites 
(80+ m) was more stable on an annual basis because 
of the seasonal influx and departure of species from 
and to adjacent habitats (Fig. 14). Species number 
varied between 7 and 1 7 fishes at station B and 7 and 



20 fishes at station A, fluctuating about a mean of 1 2/ 
sampling trip. During 1965, Tyler (1971) observed a 
higher mean diversity of 17 species/ trip at station A 
with a maximum occurrence of 24. The difference 
between his observations and ours may be accounted 
for partially by the decline in haddock abundance 



o 

5 



14 



12 - 



10 - 



8- 



Q. 
CO 

~ 6 



2 - 







A 
l\ 

/ 1 

- 1 \ 

/ 1 

: A 


"*\ Station 


' i \ 


Station 8 


— A 




i II 


1 1 1 1 



J J A 
Month 



FIGURE 13.— Monthly diversity of fishes at intertidal stations 3 and 
8 in Passamaquoddy Bay. Species/month for station 3 is mean of 
1976 and 1977 samples. 



20 



16 



12 



2 



a 

CO 



Station A \ 



Station B 




' ' 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 

F A J A O D| F A J A O D| F A J A 
1976 1977 1978 



Figure 14.— Seasonal diversity of 
fishes at station A (Passamaquoddy 
Bay) and station B (Bay of Fundy). Ver- 
tical bars represent the range among 
replicated collections. 




01 — I — 1—1 — I — I — 1 — I — I I 1—1 ,_l I I 1 L_l I I I i i i 1 I I I I I l_l 

F A J A O D] F A J A O D| F A J A 
1979 1980 1981 



135 



FISHERY BULLETIN: VOL. 82, NO. 1 



since 1965 and the recent absence of American 
plaice from this site, and partially by his use of a 0.6 
cm cod end liner, which would have retained small, 
occasional species more often than our 2.5 cm cod 
end. 

Highest diversities occurred during winter at sta- 
tion B and during summer at station A (Fig. 14) as a 
result of seasonal exchange between these sites and 
the arrival of periodics. The highest diversities record- 
ed during the study period occurred at station A dur- 
ing the fall, coinciding with maximum annual 
temperatures (Fig. 2). Diversity at station C, the mid- 
depth site, decreased from 13 species in May 1978 to 
4 species in May 1980, perhaps in response to a 
general decline in lower Bay of Fundy temperatures 
during the study period (Fig. 3). 

GENERAL DISCUSSION 

Most authors have related the occurrence and dis- 
tribution of adult benthic fishes in the North Atlantic 
to substrate type and temperature (Edwards 1965; 
Colton 1972; McEachran and Musick 1975; Scott 
1976) and have shown that there is a marked seasonal 
variation (Lux and Nichy 1971; Jeffries and Johnson 
1974). Our findings agree and suggest yearly dif- 
ferences at the same site for a given time may be 
influenced mainly by annual ocean climate pertuba- 
tion. Species occurrence and abundance appeared to 
change in response to seemingly small changes in 
temperature. Jeffries and Johnson (1974) reported a 
similar observation concerning winter flounder 
abundance over a 7-yr period in Narragansett Bay. 
Pelagic and semipelagic species (Atlantic herring, 
silver hake) demonstrated little or no substrate pref- 
erence. Occurrence was apparently related to annual 
migratory behavior. 

Seasonal movements of the various species was 
largely from an inshore, shallow-water locality in 
summer to an offshore, deepwater locality in winter 
with a reverse movement occurring in spring. Cause 
of this movement may have a large physiological com- 
ponent related to temperature effects on the 
osmoregulation of marine fishes (Potts and Parry 
1964). In the southern part of their range, fish such as 
winter flounder migrate onshore in winter (Bigelow 
and Schroeder 1953) in response to availability of 
preferred temperature but never encounter the low 
temperatures found at northern latitudes. Atlantic 
tomcod, a species known to produce an antifreeze in 
its blood (Fletcher et al. 1982), was one of the few 
fishes exhibiting onshore migration to lower 
salinities during winter in this area. For many species 

(pollock, Atlantic herring, white hake), migration 



from inshore habitat to offshore is unidirectional for 
the individual, since each year the beach community 
consists of the new 0+ year class. For other species 
(winter flounder, juvenile sculpins, radiated shanny), 
the return inshore is an annual occurrence, triggered 
perhaps as much by resource availability and pre- 
dator avoidance as by physiology. 

Tyler (1971) concluded that in Passamaquoddy 
Bay movements of large fish independent of the 
small individuals of a species were not evident for 
fishes other than hake, but we found obvious dif- 
ferences in size-class distributions and abundance 
between summer and winter populations of winter 
flounder, witch flounder, Atlantic cod, and pollock at 
offshore sites and a complete lack of most fish 
inshore. This suggests marked segregation between 
juveniles (at least 0+ age group) and adults for these 
species. The use of shallow water habitat as nursery 
area by fishes of commercial important in the Cana- 
dian North Atlantic has received little attention. In 
Europe, this fact has been amply demonstrated for 
many fish species, including Atlantic cod and pollock 
(Zijlstra 1972; Daan 1978; Burd 1978; Rauck and 
Zijlstra 1978). The use of beach habitat as nursery by 
these fishes makes them susceptible to coastal pollu- 
tion impacts and puts their adult fisheries at risk to 
coastal degradation and development, 

Decline in haddock abundance in Passamaquoddy 
Bay since 1965 coincides with increased numbers of 
Atlantic cod. However, previous studies indicate lit- 
tle interaction between these two species (Tyler 1972; 
Jones 1978). Catches in 1965 (Tyler 1971) coincided 
with the largest haddock abundance on record 
(Clark et al. 1981). Fishermen in Passamaquoddy 
Bay may only catch haddock consistently during 
years preceded by large recruitment on Georges 
Bank, the Scotian Shelf, and the Gulf of Maine. 

In the Bay of Fundy region, fish assemblages are 
segregated according to habitat and, although fish 
movement is influenced by seasonal climatic regime, 
assemblages appear cohesive through time. In sum- 
mer, fishes assembled and exploited the available 
resources as members of 1) estuarine, 2) beach, 3) 
offshore, hard-bottom, 4) offshore, soft-bottom, and 
5) migratory-pelagic assemblages. With winter, 
movement of species and/or age groups resulted in 
different seasonal assemblages in each habitat, but 
major groupings remained essentially intact and 
replaced each other seaward. The reverse movement 
occurred in spring. A large portion of benthic and 
pelagic components occurring at the offshore, hard- 
bottom habitat were migratory. In contrast, the 
offshore, soft-bottom assemblage was more senden- 
tary. Smaller seasonal variation in the water tem- 



136 



MACDONALD ET AL.: FISH ASSEMBLAGES AND THEIR SEASONAL MOVEMENTS 



perature at the Bay of Fundy, soft-bottom site, and 
the greater seasonal stability of invertebrate food 
resource production in this type of habitat ( Wildish 
and Dadswell in press) may also be important. The 
dynamic nature of the hard-bottom community, par- 
ticularly among commercially valuable species, 
emphasizes the need for well-designed, seasonal 
sampling programs in order to properly assess the 
occurrence of species and abundance offish stocks in 
a local area. Long-term changes are apparent from 
annual assessment data (Brown et al. 1973), but 
higher resolution surveys at "type" localities are 
needed to properly determine causative factors, 
whether physical or biological. 

ACKNOWLEDGMENTS 

We thank Captain Tom Allen and Floyd Johnson, 
crew of the Pandalus for their help. Bill McMullon 
and Frank Cunningham prepared the figures and 
Brenda Fawkes and Jeanine Hurley typed the 
manuscript. J. S. Scott, W. B. Scott, D. Markle, and 
D. J. Scarratt reviewed the manuscript. Work by J. S. 
Macdonald was in partial fulfillment for Ph.D. 
requirements supported by NSERC Grants to R. H. 
Green. This work was supported in part by a grant 
from the Canadian National Sportsman's Fund (4- 
R88) to D. Methven. 

LITERATURE CITED 

Battle, H. I. 

1930. Spawning periodicity and embryonic death rate of 
Enchelyopus cimbrius (L.) in Passamaquoddy Bay. Con- 
trib. Can. Biol. Fish., New Ser. 5:361-380. 

BlGELOW, H. B., AND W. C. SCHROEDER. 

1939. Notes on the fauna above mud bottoms in deep water in 
the Gulf of Maine. Biol. Bull. (Woods Hole) 76:305- 
324. 
1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish. 
Bull. 53, 57 p. 
Brawn, V. M. 

1960. Seasonal and diurnal vertical distribution of herring 
(Clupea harengus L.) in Passamaquoddy Bay, N.B. J. 
Fish. Res. Board Can. 17:699-711. 
Brown, B. E„ J. A. Brennan, E. G. Heyerdahl, M. D. Grosslein, 

AND R. C. HENNEMUTH. 

1973. An evaluation of the effect of fishing on the total finfish 

biomass in ICNAF Sub-area 5 and statistical area 6. Int. 

Comm. Northw. Atl. Fish. Res. Doc. 73/8, 32 p. 
Burd, A. C. 

1978. Long-term changes in North Sea herring stocks. In G. 

Hempel (editor), North Sea fish stocks - recent changes 

and their causes, p. 137-153. Rapp. P. — V. Reun. Cons. 

Perm. Int. explor. Mer 172. 
Clark, S. H., W. J. Overholtz, and R. C. Hennemuth. 

1982. Reveiw and assessment of the Georges Bank and Gulf 

of Maine haddock fishery. J. Northw. Atl. Fish. Sci. 3:1- 

27. 



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. 
Cushing, D. H. 

1980. The decline of the herring stocks and the gadoid out- 
burst. J. Cons. Int. Explor. Mer 39:70-81. 
Daan, N. 

1978. Changes in cod stock and cod fisheries in the North 
Sea. In G. Hempel (editor), North Sea fish stocks - recent 
changes and their causes, p. 39-57. Rapp. P.— V. Reun. 
Cons. Perm. Int. Explor. Mer 172. 

1980. A review of replacement of depleted stocks by other 
species and the mechanisms underlying such re- 
placement. In A. Saville (editor), The assessment and 
management of pelagic fish stocks, p. 405-421. Rapp. P.- 
V. Reun. Cons. Perm. Int. Explor. Mer 177. 

Dadswell, M. J., G. D. Melvin, and P. J. Williams. 

1983. Effect of turbidity on the temporal and spatial utiliza- 
tion of the inner Bay of Fundy by American shad (Alosa 
sapidissima) (Pisces: Clupeidae) and its relationship to 
local fisheries. Can. J. Fish. Aquat. Sci. 40 (Suppl. 1):322- 
330. 

Dow, R. L. 

1964. A comparison among selected marine species of an 
assocation between sea water temperature and relative 
abundance. Extr. J. Cons., Cons. Int. Explor. Mer 
28:425-431. 

Edwards, R. L. 

1965. Relation of temperature to fish abundance and dis- 
tribution in the southern New England area. Int. Comm. 
Northw. Atl. Fish., Spec. Publ. 6:95-110. 

Fletcher, G. L., C. L. Hew, and S. B. Joshi. 

1982. Isolation and characterization of antifreeze glyco- 
proteins from the frostfish, Microgadus tomcod. Can. J. 
Zool. 60:348-355. 

HACUNDA, J. S. 

198 1. Trophic relationships among demersal fishes in a coast- 
al area of the Gulf of Maine. Fish. Bull., U.S. 79:775- 
788. 

Hare, G. M. 

1977. Atlas of the major Atlantic coast fish and invertebrate 
resources adjacent to the Canada-United States boun- 
dary areas. Can. Fish. Mar. Serv. Tech. Rep. 681, 97 p. 

Jaccard, P. 

1932. Die statistisch-florestische Methode als Grundlage der 
Pflanzen-soziologie. Handb. Biol. Arbeitsmed. 5:165- 
202. 

Jeffries, H. P., and W. C. Johnson. 

1974. Seasonal distributions of bottom fishes in the 
Narragansett Bay area: Seven-year variations in the abun- 
dance of winter flounder {Pscudopleuronectes 
americanus). J. Fish. Res. Board Can. 31:1057-1066. 

Jones, R. 

1978. Competition and co-existance with particular 
reference to gadoid fish species. In G. Hempel (editor), 
North Sea fish stocks - recent changes and their causes, p. 
292-300. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. 
Mer 172. 

Keast, A. 

1970. Food, specializations and bioenergetic interrelations in 
the fish faunas of some small Ontario waterways. In J. H. 
Steel (editor), Marine food chains, p. 377-411. Univ. 
Calif. Press. 
Kontkanen, P. 

1957. On the delimitation of communities in research on 



137 



FISHERY BULLETIN: VOL. 82, NO. 1 



animal biocoenotics. Cold Spring Harbour Symp. Quant. 
Biol. 22:373-378. 
KULKA, D. W., AND W. T. STOBO. 

1981. Winter distributions and feeding ofm a ckerel on the 
Scotian Shelf and outer Georges Bank with reference to 
the winter distribution of other finfish species. Can. 
Tech. Rep. Fish. Aquat. Sci. 1038, 38 p. 

Leim, A. H., and W. B. Scott. 

1966. Fishes of the Atlantic coast of Canada. Fish. Res. 
Board Can., Bull. 155, 485 p. 
Lux, F. E., and F. E. NlCHY. 

1971. Number and lengths, by season, of fishes caught with 
an otter trawl near Woods Hole, Massachusetts, Septem- 
ber 1961 to December 1962. U.S. Dep. Commer., Natl. 
Mar. Fish. Ser., Spec. Sci. Rep. Fish. 622, 15 p. 
Markle, D. F. 

1975. Young witch flounder, Glyptoeephalus cynoglossus , on 
the slope off Virginia. J. Fish. Res. Board Can. 32:1447- 
1450. 
Markle, D. F., D. A. Methven, and L. J. Coates-Markle. 

1982. Aspects of spatial and temporal cooccurrence in the life 
history stages of the sibling hakes, Urophycis chuss 
(Walbaum 1792) and Urophycis tenuis (Mitchill 1815) 
(Pisces: Gadidae). Can. J. Zool. 60:2057-2078. 

McCracken, F. D. 

1959. Studies of haddock in the Passamaquoddy Bay 
region. J. Fish. Res. Board Can. 17:175-180. 

1963. Seasonal movements of the winter flounder, 
Pseudopleuronectes americanus (Walbaum), on the Atlan- 
tic coast. J. Fish. Res. Board Can. 20:551-586. 

McEachran, J. D., and J. A. Musick. 

1973. Characters for distinguishing between immature 
specimens of the sibling species, Raja erinacea and Raja 
evellata (Pisces: Rajidae). Copeia 1973:238-250. 
1975. Distribution and relative abundance of seven species 
of skates (Pisces: Rajidae) which occur between Nova 
Scotia and Cape Hatteras. Fish. Bull., U.S. 73:110-136. 
McKenzie, R. A., and S. N. Tibbo. 

1961. Herring movements in the Bay of Fundy and Gulf of 
Maine, 1957 and 1958. J. Fish. Res. Board Can. 18:221- 
252. 
Moore, R. H. 

1977. Variations in the diversity of summer estuarine fish 
populations in Aransas Bay, Texas, 1966-1973. 
Estuarine Coastal Mar. Sci. 6:495-501. 
Musick, J. A. 

1973. A meristic and morphometric comparison of the hakes, 
Urophycis chuss and U. tenuis (Pisces, Gadidae). Fish. 
Bull., U.S. 71:479-488. 
Nl.I-H. 

1982. Meristic variation in beaked redfishes, Sebastes men- 
tella and S. fasciatus, in the Northwest Atlantic. Can. J. 
Fish. Aquat. Sci. 39:1664-1685. 
Olsen, Y. H., and D. Merriman. 

1946. Studies on the marine resources of southern New 
England. IV. The biology and economic importance of the 
ocean pout, Macrooarces americanus (Block and 
Schneider). Bull. Bingham Oceanogr. Collect. Yale Univ. 
9(4):1-184. 
Orach Meza, F. L. 

1975. Distribution and abundance of the ocean pout, Mac- 
rozoarces americanus (Block and Schneider, 1801) in the 
Western North Atlantic ocean. M.S. Thesis, Univ. Rhode 
Island, Kingston, 108 p. 
Potts, W. T. W., and G. Parry. 

1964. Osmotic and ionic regulation in animals. Pergammon, 



N.Y., 423 p. 
Powles, P. M., and A. C. Kohler. 

1 970. Depth distributions of various stages of witch flounder 
(Glyptoeephalus cynoglossus) off Nova Scotia and in the 
Gulf of St. Lawrence. J. Fish. Res. Board Can. 27:2053- 
2062. 

QuiNN, N. J. 

1980. Analysis of temporal changes in fish assemblages in 
Serpentine Creek, Queensland. Environ. Biol. Fish. 
5:117-133. 
RAUCK, G., AND J. J. ZlJLSTRA. 

1978. On the nursery-aspects of the Waddensea for some 
commercial fish species and possible long-term 
changes. In G. Hempel (editor), North Sea fish stocks - 
recent changes and their causes, p. 266-275. Rapp. 
P.-V. Reun. Cons. Perm. Int. Explor. Mer 172. 

Richards, S. W. 

1963. The demersal fish population of Long Island 
Sound. Bull. Bingham Oceanogr. Collect,, Yale Univ. 
18:1-101. 
Saila, S. B. 

1961. A study of winter flounder movements. Limnol. 
Oceanogr. 6:292-298. 
Schroeder, W. C. 

1930. Migrations and other phases in the life history of cod off 
southern New England. Bull. U.S. Bur. Fish. 46:1-136. 
SCOTT, J. S. 

1976. Summer distribution of groundfish on the Scotian 
Shelf, 1970-74. Can. Fish. Mar. Serv. Tech. Rep. 635, 
13 p. 

1980. Occurrence of pollock, Pollachius virens, and sand 
lance, Ammodytes sp., larvae in the Bay of Fundy. J. 
Northw. Atl. Fish. Sci. 1:45-48. 

Sheehy, P. J., S. Y. K. Sheriouda, A. J. Alton, S. B. Saila, and S. 
M. Constanttnides. 

1977. The ocean pout: An example of underexploited 
fisheries resource development. Mar. Fish. Rev. 39 (6):5- 
15. 

Sherman, K., C. Jones, L. Sullivan, W. Smith, P. Berrien, and L. 
Ejsymont. 

1981. Congruent shifts in sand eel abundance in western and 
eastern North Atlantic ecosystems. Nature (Lond.) 
291:486-489. 

Sissenwtne, M. P., E. B. Brown, J. E. Palmer, and R. J. Esig. 

1982. Empirical examination of population interactions for 
the fishery resources off the northeastern USA. In M. C. 
Mercer (editor), Multispecies approaches to fisheries 
management advice, p. 82-94. Can. Spec. Publ. Fish. 
Aquat. Sci. 59. 

Steiner, R. G. 

1976. Neritic reef fish communities - some problems and a 
possible solution. In C. A. Simenstad and S. J. Lipovsky 
(editors), Fish food habit studies. First Pacific Northw. 
Tech. Workshop, 193 p. 

Sutcliffe, W. H., Jr., K. Drinkwater, andB. S. Muir. 

1977. Correlations offish catch and environmental factors in 
the Gulf of Maine. J. Fish. Res. Board Can. 34:19-30. 

Tyler, A. V. 

1971. Periodic and resident components in communities of 
Atlantic fishes. J. Fish. Res. Board Can. 28:935-946. 

1972. Food resource division among northern, marine, 
demersal fishes. J. Fish. Res. Board Can. 29:997-1003. 

Van Guelpen, L., and C. C. Davis. 

1979. Seasonal movements of the winter flounder, 
Pseudopleuronectes americanus, in two contrasting 



138 



MACDONALD ET AL.: FISH ASSEMBLAGES AND THEIR SEASONAL MOVEMENTS 

inshore locations in Newfoundland. Trans. Am. Fish. Soc. Bay of Fundy. Proc. N.S. Acad. Sci. 

108:26-37. Wise, J. P. 

WARFEL, H. E., AND D. MERRIMAN. 1962. Cod groups in the New England area. U.S. Fish Wildl. 

1944. Studies on the marine resources of southern New Serv., Fish. Bull. 63:189-203. 

England. I. An analysis of the fish population of the shore ZlJLSTRA, J. J 

zone. Bull. Bingham Oceanogr. Collect., Yale Univ. 1972. On the importance of the Wadden Sea as a nursery area 

9(2): 1-91. in relation to the conservation of the southern North Sea 

Wildlish, D. J., AND M. J. Dadswell. fishery resources. Symp. Zool. Soc. (Lond.) 29:233- 

In press. Sublittoral Gammaridae of soft sediments in the 258. 



139 



THE DETECTION AND DISTRIBUTION OF 

LARVAL ARCTO-NORWEGIAN COD, GADUS MORHUA, 

FOOD ORGANISMS BY AN IN SITU PARTICLE COUNTER 



S. TlLSETH AND B. ELLERTSEN 1 



ABSTRACT 

An in situ particle counter system was developed to count measure food particles in numbers per liter within 
the size range 150-600 /urn, the sizes of copepod nauplii captured by first feeding cod larvae. Patches of 
particles/nauplii of 50- 1 00 per liter were found in the spawning and larval first feeding area. Different sizes of 
copepod nauplii showed diel vertical migration, and this influenced the formation of patches. Mixing of the 
water column by wind forces created a homogeneous vertical distribution of particles. Gut content analysis of 
cod larvae during these hydrographical conditions indicated reduced accessibility of food organisms to 
larvae. 



During the last few years fisheries scientists have 
done a great deal of laboratory work on the behavior 
of fish larvae and their energy requirements for 
growth and survival (Hunter 1972; Laurence 1974; 
Lasker and Zweifel 1978; Houde 1978; Werner and 
Blaxter 1980). A review of these data (Hunter 1981) 
shows that differences exist between the required 
density of prey particles for first feeding larvae to 
survive and the densities found in the sea. Since 
pelagic fish larvae are successful in their environ- 
ment, it is recognized that there must be patches of 
suitable concentrations of food organisms for first 
feeding larvae (Lasker and Zweifel 1978). This has 
been demonstrated for the northern anchovy, 
Engraulis mordax, in laboratory experiments by 
Hunter and Thomas (1974) and in a series of field 
investigations by Lasker (1978). Houde and Schek- 
ter ( 1 978) have shown increased survival of larval bay 
anchovy, Anchoa mitchilli, and sea bream, 
Archosargus rhomboidalis, when exposed to sim- 
ulated food patches in a laboratory experiment. 

This work has been stimulated by Hjort's (1914) 
hypothesis which simply stated that larval mortality 
rates may be due to variable feeding conditions at a 
critical stage, which in turn causes variations in year- 
class strength. It has been difficult to test this simple 
hypothesis in field surveys because of the inade- 
quacy of the sampling gear used (May 1974). To 
obtain a better understanding of the relationship be- 
tween estimates of food densities required by fish 
larvae in the laboratory and densities found in the 



'Institute of Marine Research, Directorate of Fisheries, 5011 
Bergen-Nordnes, Norway. 



open sea, samples should be taken which are relevant 
to larval searching behavior. This would require an 
enormous number of plankton samples. It would be 
time-consuming to obtain these samples with con- 
ventional plankton gear. Furthermore, water move- 
ment and dispersion would make it difficult to obtain 
time and space relationships for studying the forma- 
tion and dynamics of plankton patches (Steele 1978). 
One way of studying these relationships is by using in 
situ instruments (Boyd 1973; Pugh 1978; Tungate 
and Reynolds 1980). 

In this study an instrument designed to count and 
measure particles in situ in the size range of food 
organisms most frequently captured by cod larvae 
was used. Investigations were made on the spawning 
and first feeding grounds of the Arcto-Norwegian 
cod, Gadus morhua Linnaeus, during two successive 
years (1980-81). During the first survey, inves- 
tigations were made in a sheltered fjord where cod 
larvae are known to appear in high numbers 
(Ellertsen et al. 1977) and where the current system 
has been described (Furnes and Sundby 1981). The 
objective was to find and study the formation of mi- 
crozooplankton patches and to study larval cod feed- 
ing under different environmental conditions with 
regard to food density, water turbulence, etc. In the 
following year, the main first feeding area, an open 
ocean bay, was surveyed in order to find and study 
the vertical and horizontal distribution of micro- 
zooplankton patches in this exposed area. 

The present study is part of a project, started in 
1975, dealing with growth, mortality, and drift of cod 
larvae in the Lofoten area (Ellertsen et al. 1976). 



Manuscript accepted June 1983. 

FISHERY BULLETIN: VOL. 82. NO. 1, 1984. 



141 



FISHERY BULLETIN: VOL. 82, NO. 1 



MATERIALS AND METHODS 

The Particle Counter 

The in situ particle counter system was built and 
described by Mohus (1981), Eriksen (1981), and 
Eriksen and Mohus (1981). It is presented 
schematically in Figure 1. The system is based on a 
Hiac PC-320 Particle Counter 2 which works on the 
principle of light blockage. The sensor (E-2500, 
dynamic range 80-2500 ^.m) is installed in a 
pressure-proof box together with a depth detector. A 
pump is connected to the sensor, and the sensor and 
pump are mounted to a rig which is lowered into the 
sea by winch. Seawater is pumped through a 60 cm 
long by 2.5 cm diameter hose through the sensor 
orifice (3 mm), at a flow rate of 6.15 1/min. Particles 
are counted by the Hiac PC-320 Particle Counter 
and depth is monitored by the depth detector unit. 
The "Micro -count" datalogger unit contains an 



Reference to trade names does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 



input-output interface to accomodate incoming data, 
a large internal data storage area, operator com- 
munication via a small CRT display, a keyboard, and 
a microprocessor with program to control the system. 
The microcomputer samples data from the Hiac PC- 
320 Particle Counter and the data sample time can 
be selected from 1 to 99 s. Finally, a Silent 733 ter- 
minal is connected to the microcomputer. This ter- 
minal contains a full text keyboard and a page printer 
used for initial operator communication and printout 
of data tables. Two cassette tape stations are 
included in the terminal. 

The system operates from the surface to 50 m 
depth, and the registration of particles is presented 
on the TV monitor as the sensors are lowered into the 
sea. The vertical distribution of particles can be pre- 
sented on the monitor at 1, 2 , or 5 m depth, depend- 
ing on the selected depth intervals. Data are, 
however, printed out in 1 m depth intervals from the 
surface to 50 m depth as concentration of particles 
per liter in six different size groups (150-600 ju.m) on 
the Silent 733 terminal immediately after the sam- 
ples have been made. An in situ particle profile is 



I! TEST 
11 BOX 



Cr - 



i 



DEPTH 
DETECTOR 



i 

V 

I 

I 



HIAC PC-320 

PARTICLE 

COUNTER 



" ^ JL ^ - 

SENSORS 



INPUT-OUTPUT 



MICRO 
PROCE- 
SSOR 



DATA 

STOR- 
AGE 



COMMUNICATION 



Figure 1.— The particle counter system. 




 



l_ 



SILENT-733 



CASSETTE 
TAPE 




KEYBOARD-PRINTER 



MICRO-COUNT 



142 



TILSETH and ELLERTSEN: FOOD ORGANISMS OF LARVAL COD 



defined in the present paper as the concentration 
of particles within the size range of 150-600 /xm 
from the surface to 50 m depth in 1 m depth 
intervals. 

An object found in the Hiac sensor was measured so 
that the largest projected area was converted to a cir- 
cle of the same area. By calibration, the object was 
given a length similar to the diameter of this circle. 

The contours of Artemio nauplii were drawn by 
using a microscope drawing tube. Their areas were 
estimated by planimeter and converted to areas of 
circles and their diameters calculated. Their size dis- 
tribution was then divided into four 50 jum length 
groups of 200 to 400 /xm. Four of the Hiac Particle 
Counter channels were set according to the sensor 
calibration diagram to the corresponding size 
groups. 

The instrument system was tested and calibrated in 
the laboratory by comparing microscope and Hiac 
measurements of the size-frequency distribution of a 
sample of laboratory hatched Artemia nauplii. Tests 
were also made at sea when the research vessel was 
anchored. The in situ instrument data were com- 
pared with plankton pump samples taken simul- 
taneously. These samples were taken by a submer- 
sible electric pump (Flygt 2051, 250 1/min) which 
pumped samples on deck through a 50 m long by 5 cm 
diameter hose. Samples were taken at 0, 2.5, 5, 7.5, 
10, 12.5, 15, 20, 25, 30, and 40 m depths. This is 
defined as a zooplankton pump profile. Seawater was 
collected in calibrated tanks (23.7 1), and zoo- 
plankton were filtered through 90 itm mesh plankton 
nets. Zooplankton were identified and counted by 
microscope, the whole sample (23.7 1) was counted. 
Results of the samples from these 11 depths were 
statistically compared with the in situ counts from 
corresponding depths by paired Mests. 

Field Investigations 

The main objectives of field investigations were to 
use the in situ instrument system to find particle 
patches and to identify larval cod food organisms and 
study their vertical distribution. Observations were 
made in the Lofoten area (Fig. 2). The effect of wind 
driven turbulence on the distribution of particles and 
the consequences on larval cod feeding incidence 
were studied in the Austnesfjord (Fig. 3), which is in 
the main spawning area of the Arcto-Norwegian cod. 
Stations and sections in the Austnesfjord are shown 
in Figure 3. A section is a transect with a series of 
stations. Austnesfjord was chosen because cod lar- 
vae are known to appear in high numbers (Ellertsen 
et al. 1977), and the dynamics of the current system 



are known (Fumes and Sundby 1981). During the 
1980 cruise, a Wolfe wind recorder was placed on 
land in the fjord to continuously measure wind 
velocity and direction. 

In 1981, observations were also made in the main 
first feeding area, an open ocean bay (Fig. 2), for cod 
larvae. The objectives were to find these food parti- 
cle patches for cod larvae and to investigate the 
extent and densities of these patches in this 
exposed area. 

Distribution of cod larvae in the first feeding areas 
was studied from the Juday net (80 cm, 375 jum 
mesh) samples taken in vertical hauls from 30 to m. 
In the Austnesfjord, three stations were taken on 
eight sections (Fig. 3). The vertical distribution of 
cod larvae in the Austnesfjord was investigated only 
when the ship was anchored. A total of 42 samples 
were taken by a submersible electric pump (Flygt 
B2125, 3.4 mVmin) at 5, 10, 15, 20, 25, 30, and 35 m 
depths every 3 h from 1600 h 13 May to 1000 h 14 
May 1980. Fifteen cubic meters of seawater was sam- 
pled at each depth. Seawater was pumped through a 
40 m long by 15 cm diameter hose and filtered 
through a Juday net (40 cm, 180 /xm mesh) into a 
large tank on deck. Cod larvae were preserved in 4% 
Formalin in 10%o seawater solution. Gut contents of 




FIGURE 2.— Map of the Lofoten area with stations and sections 21 
April-8 May 1 98 1. The figures on the stations refer to number of cod 
larvae/m 2 surface. 



143 



FISHERY BULLETIN: VOL. 82, NO. 1 




TABLE 1. — Size frequency distribution of Artemia 
nauplii measured by the Hiac Particle Counter (n = 
1542) and by microscope (n = 45). 



FIGURE 3. — Map of the Austnesfjord with stations. • Juday net 
and particle/zooplankton stations, position of the 24 h station -k, 
and the Wolfe wind recorder i£r. 



about 20 larvae from each depth were examined by 
dissecting the larval gut under the microscope. 

During 24-h stations in situ particle profiles, CTD 
profiles, and zooplankton pump profiles were made 
simultaneously every 2 h. On sections, zooplankton 
pump profiles were made on every second station. 



RESULTS 
In Situ Instrument Tests 

Results of the comparison between microscope and 
particle counter measurements is presented in Table 

144 



Size 


M 
Part 


o. of Anemia 


nai 


pi, 


counted by 


Iflm) 


cle counter 






Microscope 


200-249 




101 






5 


250-299 




416 






14 


300-349 




848 






23 


350-399 




1 77 






3 



1. A chi-square test for independence in the 4X2 
table (3 df) showed no significant difference (P < 
0.05) between the two methods of measuringArterata 
nauplii. 

Paired tests between microscope and in situ parti- 
cle counts were done on data from two different 24-h 
stations in the Austnesfjord (Figs. 4, 5). Plankton 
pump samples were taken from 11 different depths 
on each profile, and the mean counts from these 
depths were tested against the mean in situ counts 
from the same depths. A comparison was also made 
between the mean of all plankton pump counts from 
each profile, and the mean of all in situ counts from 
the corresponding profile. 

During the first 24-h station, 19 vertical profiles 
were made. No significant differences (P < 0.05) was 
found when the mean counts (n = 19) from each of 1 1 
different depths were compared, nor when the mean 
counts from the different profiles were compared. 
The same statistical test was made on data from 14 
vertical profiles on the second 24-h station. There 
had been an increase in the variability of mi- 
crozooplankton both horizontally and vertically dur- 
ing this 24-h station (Fig. 5A, B). No significant 
differences (P < 0.05) was found between the mean 
in situ counts and the mean plankton pump counts 
when the different profiles were tested. We found, 
however, a significant difference (P < 0.05) when the 
mean counts from corresponding depths were tested. 
This difference was found between in situ and 
plankton pump counts both from 30 and 40 m 
depths. No significant difference (P < 0.05) was 
found between counts from 0, 0.5, 7.5, 10, 12.5, 15, 
20, and 25 m depths. This difference may have 
resulted from samples having been taken at different 
depths. The in situ instrument was equipped with a 
depth detector, but the depth of the submersible 
pump was controlled only by the meter wheel on 
the winch. 

Distribution of Particle/Nauplii 
in the Fjord 

The vertical distribution of particles/nauplii for a 
24-h station made during 22-24 April 1981 in the 



TILSETH and ELLERTSEN: FOOD ORGANISMS OF LARVAL COD 

A 



21 



22-24 Apr. 81 Austnesf jorden 
01 05 09 13 17 21 01 




FIGURE 4. — Isopleth diagrams of the particle concen- 
trations (per liter) (A), and nauplii (per liter) (B), center 
station, section 5 in Austnesfjord, 22-24 April 1981. 



21 



22 -24 Apr. 81 Austnesf jorden 

01 05 09 13 17 21 01 05 09 H 




Austnesfjord is presented in Figure 4A and B. The 
maximum observed particle concentration was a 
small patch of 50 particles/1 at about 15 m (Fig. 4A). 
A patch of 40 nauplii/1 at the same depth was iden- 
tified from pump samples (Fig. 4B). The particle/ 
nauplii isolines in the upper 20 m show a tendency of 
ascending towards the surface at midnight, indicat- 
ing their diel vertical migration. This observation was 
repeated on another 24-h station made 6 d later at the 
same position (Fig. 5A, B). Particle concentration 
had increased markedly during this period; more 
than 50 particles/1 were found at 25-35 m depth on 
every profile. A very dense surface patch was found 
at midnight with more than 500 particles/1. Figure 5B 
shows a similar distribution of nauplii during the 
same 24-h station. Since there was no wind in the 
fjord and consequently little or no vertical turbulence, 
the hydrographic conditions during this 24-h station 
were perfect for this type of observation. This is 
shown in Figure 6 where the hydrographic conditions 
is presented by the temperature distribution in the 
upper 60 m. 



Figure 7AandB presents the particle (1 50-600 jiim) 
distribution from to 40 m depth through a section of 
the Austnesfjord made at night on 27-28 April 1981 
from 2130 to 0420 h. There was little or no wind in the 
fjord when the section was made. Patches of more 
than 100 particles/1 were found in the surface water 
of the outer parts of the fjord. A particle minimum 
layer (<10/1) was observed at 10 m in the middle of 
the fjord. In the bottom of the fjord three patches of 
more than 50 particles/1 were found at different 
depths. Figure 7B shows the naupliar distribution on 
the same section. Highest concentrations (> 100/1) 
were observed in the bottom of the fjord, at inter- 
mediate depths and in the surface water of the outer 
parts of the fjord. 

The same section made through the fjord the next 
day from 0950 to 1610 h (Fig. 8A, B) showed that the 
particle/nauplii distribution in the fjord had changed 
completely. A particle/nauplii minimum layer (< 10/ 
1) was found from the surface down to about 20 m 
through most of the fjord length. The surface patches 
in the outer parts of the fjord had disappeared. Only 



145 



FISHERY BULLETIN: VOL. 82, NO. 1 



28 -29 Apr. 81 Austnesfjorden 

07 09 11 13 15 17 19 21 



01 03 05 07 09 H 




B 



a 
» 



28 - 29 Apr. 81 
11 13 15 



Austnesfjorden 
17 19 21 



23 



01 



03 05 07 09 H 




FIGURE 5.— Isopleth diagrams of the particle concentrations (per liter) (A), and nauplii (per liter) (B), cen- 
ter station, section 5 in Austnesfjord, 28-29 April 1981. 



07 09 



28 -29 Apr.81 
11 13 15 



Austnesfjorden 
17 19 21 23 



01 03 05 07 09 H 




146 



FIGURE 6. — Isopleth diagram of the temperature distribution, middle station, section 5 in Austnesfjord, 

28-29 April 1981. 



TILSETH and ELLERTSEN: FOOD ORGANISMS OF LARVAL COD 

Austnesfjorden Hella 27 - 28 Apr. 81 
A 0420 



2 n. miles 



21 30 H 




Austnesfjorden Hella 27 -28 Apr. 81 



2 n. mile s 



21 JU H 




FIGURE 7.— Particle (A) and nauplii (B) distributions (per liter) in the upper 40 m of Austnesfjord, 27-28 April 1981, at 21.30 to 0420 h. 

(Particle size range 150-600 (im, nauplii all sizes.) 



one patch with >50 particles/nauplii per 1 was ob- 
served between 20 and 40 m at the bottom of the 
fjord. 

Effect of wind driven turbulence on vertically 
migrating particles is presented in Figure 9A, B, and 
C. The figure presents data collected continuously 
from 9 to 15 May 1980, on wind velocity and direc- 
tion, temperature, and particle distribution in the 
water column. Due to technical problems, only par- 
ticles within the size range 300-500 ju.m were 
measured by the particle counter in 1980. From 9 to 
12 May the wind was blowing downfjord with varying 
velocity. On 1 2 May the wind changed direction 180° 
and blew upfjord with a velocity of 5- 1 m/s (Fig. 9A). 
Unfortunately, observations of temperature and par- 
ticle distribution were not made from 10 to 11 May. 
However, one 24-h station was made on 9 May during 
the period when the wind was blowing downfjord. At 
this time, the upper 10 m of the water column showed 
tendencies of mixing, and colder intermediate water 



masses were observed from 15 to 55 m above the 
transition layer. Within the cold intermediate water 
masses a particle maximum layer was found (Fig. 
9C). It is believed that the wind was blowing the sur- 
face water downfjord and this was compensated for 
by intermediate water masses moving in the opposite 
direction. On 9 May we observed a patch of particle- 
rich intermediate water moving in from the outer part 
of the fjord. The particle isolines in the upper 10 m 
followed the isotherms (Fig. 9B, C). When the wind 
direction reversed and increased in velocity on 12 
May (Fig. 9A), the fjord became more exposed to the 
wind force and the wave action from the open ocean 
outside the fjord. Under this condition the current 
system will reverse (Furnes and Sundby 1981). The 
surface water became completely mixed within about 
24 h (Fig 9B), and no particle diel vertical migration 
was observed during this condition (Fig. 9C). The 
particle concentration decreased and became almost 
homogeneous from the surface to 40 m. 



147 



FISHERY BULLETIN: VOL. 82, NO. 1 



Austnesfjorden -Holla 29 Apr. 81 



a 
a 



2 n. miles 




1610 H 



B 0950 



Austnesfjorden Holla 29 Apr. d1 



2 n. miles 



1610 H 



a 

Q 10H 



20 



30 J 



40 




FIGURE 8.— Particle (A) and nauplii (B) distributions (per liter) in the upper 40 m of Austnesfjord, 29 April 1981, at 0950 to 

1610 h. 



Distribution of Cod Larvae 

The highest concentration of cod larvae (140-290 
larvae/m 2 ) was observed in the middle of May at the 
bottom of the Austnesfjord both in 1980 and 1981 
(Fig. 10). This has also been observed on previous 
cruises (Ellertsen et al. 1977). The research vessel 
was therefore anchored at the middle station on sec- 
tion 5, where 24-h stations were made. 

In 1 98 1 , the study of the distribution of cod larvae in 
the exposed open ocean bay of Vesteralsfjorden 
showed that larvae were only found on the innermost 
stations with a maximum of 4 larvae/m 2 (Fig. 2), e.g., 
only two cod larvae in vertical Juday net hauls from 
30 m depth. 

Gut contents of 738 cod larvae were examined from 
39 pump samples. Fewer than 10 larvae were found 
in pump samples from 30 and 35 m depths from the 
01-02 h pump profile and from 35 m depth from the 



04-05 h pump profile. These larvae have not been 
included in the analysis (Fig. 11B). A total of 1,204 
prey organisms were found, out of which 96.5% were 
identified as copepod nauplii. Only 1.7% of the prey 
organisms could not be identified. About 0.5% of the 
larval cod gut content was bivalve veliger larvae, 
copepod eggs, and phytoplankton (Peridinium sp.), 
and 1.3% was identified as copepod fecal pellets. The 
size distribution of the main prey organisms (e.g., 
copepod nauplii) ranged from 140 to 520 jiim with a 
mean size of 224 /xm (all measurements as 
carapace length). 

Gut content analysis of cod larvae is presented in 
Figure 1 IB as feeding incidence (percent larvae with 
gut content) and larval feeding ratio (number of prey 
organisms per larval gut). The feeding incidence 
varied between 73 and 100% in samples from the 
three pump profiles taken before midnight. In 6 1 % of 
these samples the feeding incidence was as high as 



148 



TILSETH and ELLERTSEN: FOOD ORGANISMS OF LARVAL COD 




80- 



a. 

q 50 



>150 



20- 



40 



r~ft~^~ Particle density 
\C- 150 nos. per liter 

' 100 



40 



fe\ 40 30 l^^ x 20 7 rO C/ 20 






60 



FIGURE 9. — Wind velocity (length of vector, see m/s scale) and direction from the abcissa (A), 
isopleth diagrams of temperature (B), and particle concentration (300-500 yum) distribution (C), 
at the middle station on section 5 in Austnesfjord, 9-15 May 1980. 



90-100% .The larval feeding ratio was >1 prey/larval 
gut in all samples taken before midnight. In 71% of 
these samples the feeding ratio was >2 prey/larval 
gut and in 14% of the samples >3 prey/larval gut. In 
samples taken after midnight, however, the feeding 
incidence varied between 4 and 92%. The lowest 
level was found in pump samples from 25 m depth 
from the 01-02 h profile. In 38% of the samples taken 
after midnight the feeding incidence was <50%. Only 
in the last pump profile made at 09-10 h the larval 
feeding incidence was more than 50% in all samples. 
The feeding ratio was < 1 prey/larval gut in all sam- 
ples from 01-02 h profile, and <1 prey/larval gut in 
61% of all samples taken after midnight. A feeding 
ratio level <1 prey/larval gut was not observed in 
samples taken before midnight. The highest feeding 
ratio observed in samples taken after midnight was 



1.65 prey/larval gut from the 25 m depth samples 
taken from the 09-10 h pump profile. 

Distribution of Particles/Nauplii 
in Open Ocean Waters 

The main first feeding area of the Arcto-Norwegian 
cod is thought to be the waters outside the Lofoten 
islands and in the open ocean bay of the Vesterals- 
fjord (unpubl. data). Figure 12A and B shows the 
particle and nauplii distributions in the northeast 
section in the Vesteralsfjord. Plankton pump samples 
were only taken at every second station on the sec- 
tion. The figure shows a similar distribution pattern. 
However, due to the more frequent samples taken by 
the particle counter, a more accurate distribution 
picture of the particles on the section was achieved. 



149 



FISHERY BULLETIN: VOL. 82, NO. 1 



LARVAE /M^ 



80 • 
60 
40 
20 H 







SURFACE 




29 APRIL 1980 
1000M 



7 6 



3 2 1 



LARVAE /M< 



STATIONS 



140- 
120- 

1 DO 
80 
60 
iO 
20 



SURFACE 



5 MAY 1980 
100OM 



7 6 



LARVAE/M 2 



4 3 2 1 

STATIONS 



IX 




K 



300- 


SURFACE 

• 






12 


MAY 1980 




280 


\ 








1000M 






\ 








1 ' 




100 


\ 

• 












80- 


\ 


• 










60 


\y 

• 


\ 










4 n 






\ 








20 

n 










• 


• 



7 6 5 4 3 2 1 

STATIONS 



IX 



LARVAE/M 2 
SURFACE 



24 APRIL 1981 

1000M 




LARVAE/M 2 
SURFACE 



120 
100 

80 

60 H 

40 
20 





LARVAE /M 2 

SURFACE 
HO- 

120 - 

100 
8 
60H 
40 

20 H 




4 2 

STATIONS 



29 APRIL 1981 



1000M 




7 6 5 4 3 2 1 
STATIONS 



IX 



8 MAY 1981 
1000M 




8 7 6 5 4 3 2 1 

STATIONS 



IX 



FIGURE 10.— The average number of cod larvae/m 2 surface on sections 1-8 and stations X and IX in the Austnesfjord, April-May 1980 

and 1981. 



Sections were also made at four locations in the 
open water off the Lofoten islands. On three of these 
sections (Eggum, Myrland, and Fuglehuk), patches 
with high particle concentrations (>50/l) were ob- 
served about 1 1 km (8 n mi) off shore. All sections had 
low particle concentrations (10-30/1) in the surround- 
ing water masses (Figs. 13-15). The similarity of the 
positions of these three patches suggests that they 
are components of the same water mass with higher 



particle concentrations than the surrounding water 
masses. On the Skiva section (Fig. 16A-D) the parti- 
cle distribution patterns were more complicated. 
The section was surveyed during daytime and two 
patches were observed, one at about 5-10 m (>100 
particles/1) and another 20-25 m (>50 particles/1). 
Particle concentration decreased further offshore. 
The same section was surveyed at night (Fig. 16C), 
and two surface patches were found. 



150 



TILSETH and ELLERTSEN: FOOD ORGANISMS OK LARVAL COD 



Q. 
LU 

o 





16-17 


HOURS 


5 


10 




0-«. 




10- 




"""^-° 




°c 




20- 






30- 


./" 





19-20 HOURS 

5 10 15 



22-23 HOURS 01-02 HOURS 0^-05 HOURS 09-10 HOURS 



/ 



10 15 




\ 

O 

/ 



10 



10 



a 

\ 

O 

J 
\ 



\ 



JOLarvae/rrr 



10 



Q- 20- 

LU 
Q 



3 



50 100 
"I 2 



50 100 



v 



? 


4 


1 

\ 


\ 

• 



3 a 



50 100 



J L. 



\ 



j a 



\ 



50 ip o q_ 

1 2 



\\ 
/ 

/I 



50 ip o g_ 

1 2 



50 100 




4 



I 



/ 

V 



»/oFI 

Naupl/larv 



K 



FIGURE 1 1 .—Distribution of first feeding cod larvae (per m 3 ) (A), and the larval feeding incidence (% larvae with gut content) V and lar- 
val feeding ratio (nauplii/larval gut) O (B), during the 24 h sampling station, 13-14 May 1980, at middle station, section 5 in 
Austnesfjord. 



A Vesteralsf jorden 30 Apr 1 May 81 

20 00 ,2n.miles 



5«H 

Q. 

a 



20 



30 



0213 h 




B Vesteralsf jorden 30 Apr. - 1 May 

2000 



40- 1 



. 2 n.miles 



02«H 




Figure 12. -Particle (A) and nauplii (B) distributions (per liter) in the upper 4 m 
on the section in Vesteralsfjord, 30 April- 1 May 1981. 



151 



FISHERY BULLETIN: VOL. 82, NO. 1 



£ 

£ 
a. 

01 

Q 
10 



20 



1033 



Eggum 26Apr.81 



2 n. miles 



1535 H 



30 ^ 



40 




10 



30 



Eggum 26 Apr. 81 



2 n. miles 



15 40 H 



g-IOH 
Q 



20 



30- 



40 J 




FIGURE 13.— Particle (A) and nauplii (B) distributions (per liter) in the upper 40 m on the 

Eggum section, 26 April 1981. 



DISCUSSION 

Food particles found in the alimentary tract of larval 
cod consist, with few exceptions, of copepod nauplii 
in the size range of 140-520 tun. This observation did 
not differ significantly from that of Ellertsen et al. 
(1977), who found the size variation to be within 140- 
600 /Am. The in situ instrument was set to detect par- 
ticles in this size range. Investigations have shown 
that in May copepod nauplii outnumber all other par- 
ticles in this size range in the Lofoten area (Ellertsen 
et al. 1977; Wiborg 1948a, b). The main objective 
when designing this instrument was to obtain a quick, 
reliable impression of naupliar distributions without 
laborious, time-consuming countings by microscope. 
The tests performed to compare the in situ instru- 
ment system and the plankton pump samples 
showed good agreement between the two methods. 
The critical food concentrations for first feeding cod 
larvae are not precisely known. They are thought to 
be on the order of 40-200 nauplii/1 based on studies 
of swimming activity, larval search volume, and 
oxygen requirements of first feeding cod larvae 



(Solberg and Tilseth 1984). Patches of particles/ 
nauplii with the required densities for first feeding 
cod larvae to survive were found in the spawning and 
first feeding area by these methods. 

The results presented in this paper show some of 
the dynamics in the formation and distribution in 
time and space of microzooplankton patches. The 
vertical distribution and density of nauplii changes 
due to the diel vertical migration of these organisms 
(Figs. 5, 6). 

The concentration of particles/nauplii in a patch 
was dependent on the hydrographic situation and on 
the distribution and concentration of micro- 
zooplankton in the water column (Figs. 5, 6). Conse- 
quently the vertical distribution of particles and 
nauplii will be dependent on factors such as hydro- 
graphic conditions and time of day when the ob- 
servations are made. 

Increased wind force caused mixing of the surface 
layers and led to a homogeneous vertical particle dis- 
tribution. No surface patch was observed at night 
during windy conditions, and the mean particle con- 
centration in the water column dropped steadily dur- 



152 



TILSETH and ELLERTSEN: FOOD ORGANISMS OF LARVAL COD 

A Myrland 26 Apr. 81 

2126 



Q. 
41 

° 10 



20 



30- 



40- 1 



2 n. miles 



17 14 H 




Myrland 26 Apr. 81 





21°° 




2 n. miles 


17 50 H 


• 0- 










Q. 


2 ? >20 






^~""— 10 


Q 10- 


] 










J / 
20 -^ 








20- 


V 








 


A 


10 


<10 




30- 


<A 








jn 


\ , 









FIGURE 14.— Particle (A) and nauplii (B) distributions (per liter) in the upper 40 m on the 

Myrland section, 25 April 1981. 



2215 



Fuglehuk 26 -27 Apr.81 



2 n. miles 



0239h 



E 

e 
Q 

20 



30- 



40-I 




FIGURE 15.— The particle distribution (per liter) in the upper 40 m on the Fuglehuk section 

26-27 April 1981. 



ing the observation period (Fig. 10). This indicates 
that wind forces have caused increased water tur- 
bulence, and that these forces have exceeded the 
naupliar swimming rate. Mixing of surface layers and 
reduction in particle concentration occurred a few 
hours before midnight 13-14 May, and the water 



column became completely mixed down to a depth of 
16 m (see Figure 10). Cod larvae were sampled both 
before and after this condition occurred (see Figure 
11). Larval gut content analysis from these samples 
showed a reduction both in feeding incidence and 
feeding ratio in samples taken the first few hours 



153 



A i io 



Skiva 27 Apr.81 



2 n.miles 



FISHERY BULLETIN: VOL. 82, NO. 1 



0430 h 




p Q9 45 Skiva 2 7 Apr. 81 



2 n.miles 



050° H 



u 


>10 


z 10 ^ 










> 10 












c 




-20^p 






10- 








10 








<10 










E 

t 20 ~ 

a. 

Q 

30 

40- 


<10 







Skiva 29 -30 Apr. 81 
C 22*7 



_i i_ 



2 n.miles 



0410H 



10- 



Q. 

o 



20- 



30 



4a 




q 22 47 Skiva 29-30Apr. 81 





2 n. miles 



40 J 



041°H 




FIGURE 16.— The particle and nauplii distribution (per liter) in the upper 40 m on the Skiva section 27 April 1981 (A, B) and the particle 

and nauplii distribution 29-30 April 1981 (C, D). 



154 



TILSETH and ELLERTSEN: FOOD ORGANISMS OF LARVAL COD 



after this hydrographic condition had occurred. Dur- 
ing the following hours the larval feeding incidence 
increased again, most rapidly in larvae sampled at 
15-30 m, indicating that food particle concentration 
did not become critical. (Note that the particle con- 
centration in Figure IOC only represents particles 
within 300-500 /Am size range.) However, the feeding 
ratio did not increase significantly, indicating a more 
difficult accessibility of food particles to the larvae. 
Similar observations were made by Lasker (1975, 
1978), where stability of the water column in the 
upper 30 m was necessary for food organisms to 
aggregate in concentrations high enough to exceed 
the threshold for feeding stimulus of first feeding 
northern anchovy larvae. This observed reduced 
feeding in cod larvae cannot be explained by a diel 
feeding rhythm. Cod larvae are visual feeders; the 
light intensity threshold for feeding is 0.1 lx 
(Ellertsen et al. 1980). The light intensity in the 
upper 40 m does not drop below this level in Lofoten 
in May, and cod larvae are found with newly captured 
nauplii in the gut at all hours (Gj0saeter and 
Tilseth 1981). 

The number of cod larvae found in the main first 
feeding area was too small to do a comparison on lar- 
val feeding conditions. However, patches with 
particle/nauplii concentrations of more than 50/1 
were observed on every section made in this area. 
Sizes of these patches were, on the other hand, small 
compared with the volume of water surveyed. The 
life span of these patches is probably very short 
because of the influence of biological and physical 
factors, especially when the upper 50 m of the water 
column is unstable. This is the normal situation in the 
Lofoten area in May (Furnes and Sundby 1981). 
Therefore, prey organism patches with concen- 
trations above the critical level for first feeding cod 
larvae would probably be broken down, due to 
increased water turbulence when the wind forces 
increase. A series of storms during the larval cod first 
feeding period could thereby have serious effects on 
larval feeding conditions and consequently on sur- 
vival and recruitment. 



ACKNOWLEDGMENTS 

The in situ instrument system was developed in 
collaboration with The Foundation of Scientific and 
Industrial Research at the Norwegian Institute of 
Technology. We thank I. Mohus, B. Holand, and I. O. 
Eriksen who were responsible for this work. We also 
thank 0. Ulltang at the Institute of Marine Research 
for his advice and help on statistics. 



LITERATURE CITED 

Boyd, C. N. 

1973. Small scale spatial patterns of marine zooplankton 
examined by an electronic in situ zooplankton detecting 
device. Neth. J. Sea Res. 7:103-11 1. 
Ellertsen, B., E. Moksness, P. Solemdal, T. Str0mme, S. 
Tilseth, and V. 0iestad. 

1976. The influence of light and food density on the feeding 
success in larvae of cod (Gadus morhua L.); field and 
laboratory observations. ICES, C. M. 1976/F:34, 31 
p. [Processed.] 

Ellertsen, B., E. Moksness, P. Solemdal, T. Str0mme, S. 
Tilseth, T.Westgard, E. Moksness, and V. 0iestad. 

1977. Vertical distribution and feeding of cod larvae in rela- 
tion to occurrence and size of prey organisms. ICES, C. 
M. 1977/L:33, 31 p. [Processed.) 

Ellertsen, B., P. Solemdal, T. Str0mme, S. Tilseth, T. 
Westgard, E. Moksness, and V. 0iestad. 

1980. Some biological aspects of cod larvae (Gadus morhua 
L.). Fiskeridir. Skr. Ser. Havunders. 17:29-47. 

Eriksen, J. O. 

1981. "Micro-count". Particle datalogger. Program man- 
ual. Sintefrep. STF 48 F 81019, 203 p. [Processed.] 

Eriksen, J. O., and I. Mohus. 

1981. "Micro-count". Particle datalogger. User's man- 
ual. Sintefrep. STF 48 F 81017, 54 p. [Processed.] 
Furnes, G.K., and S. Sundby. 

1981. Upwelling and wind induced circulation in Vestfjor- 
den. In R. Saetre and M. Mork (editors), Proceedings 
from the Norwegian Coastal Current Symposium, Geilo, 
Norway, 9-12 Sept. 1980, Vol. I, p. 152-177. Univ. 
Bergen, Norway. 
GJ0SAETER, H., AND S. TlLSETH. 

1981. Primary growth increments in otoliths of cod larvae 
(Gadus morhua L.) of the Arcto-Norwegian cod stock. 
Fiskeridir. Skr. Ser. Havunders. 17:287-295. 
H.JORT, J. 

1914. Fluctuations in the great fisheries of northern Europe 
viewed in the light of biological research. Rapp. P.— V. 
Reun. Cons. Perm. Int. Explor. Mer 20:1-228. 
HOUDE, E. D. 

1978. Critical food concentrations for larvae of three species 
of subtropical marine fishes. Bull. Mar. Sci. 28:395- 
411. 

HOUDE, E. D., AND R. C. SCHEKTER. 

1978. Simulated food patches and survival of larval bay 
anchovy, Anchoa mitchilli, and sea bream, Archosargus 
rhomboidalis. Fish. Bull., U.S. 76:483-487. 
Hunter, J. R. 

1972. Swimming and feeding behavior of larval anchovy 
Engraulis mordax. Fish. Bull., U.S. 70:821-838. 

1981. Feeding ecology and predation of marine fish lar- 
vae. In R. Lasker (editor), Marine fish larvae, morphol- 
ogy, ecology, and relation to fisheries, p. 33-77. Univ. 
Wash. Press, Seattle. 
Hunter, J. R., and G. L. Thomas. 

1974. Effect of prey distribution and density on the searching 
and feeding behaviour of larval anchovy Engraulis mordax 
Girard. In J. H. S. Blaxter (editor), The early life history of 
fish, p. 559-574. Springer- Verlag, Berl. 

Lasker, R. 

1975. Field criteria for survival of anchovy larvae: The rela- 
tion between inshore chlorophyll maximum layers and 
successful first feeding. Fish. Bull., U.S. 73:453-462. 

1978. The relation between oceanographic conditions and 

155 



FISHERY BULLETIN: VOL. 82, NO. 1 



larval anchovy food in the California Current: identifica- 
tion of factors contrihuting to recruitment failure. Rapp. 
P.-V. Reun. Cons. Int. Explor. Mer 173:212-230. 

LASKER, R.. AND J. R. ZWEIFEL. 

1978. Growth and survival of first-feeding Northern anchovy 
larvae {Engraulis mordax) in patches containing different 
proportions of large and small prey. In J. H. Steele 
(editor). Spatial patterns in plankton communities, p. 329- 
354. Plenum Press, N.Y. 

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. 

May, R. C. 

1974. Larval mortality in marine fishes and I he critical period 
concept. //; J. H. S. Blaxter (editor). The early life history 
offish, p. 3-19. Springer-Verlag, Berl. 
Moms, I. 

1981. "Micro-count". Particle datalogger. Equipment man- 
ual. Sintef rep. STF 48 F 81018, 90 p. [Processed.] 
PUGH, P. R. 

1978. The application of particle counting to an understand- 
ing of the small-scale distribution of plankton. In J. H. 
Steele (editor), Spatial patterns in plankton communities, 
p. 111-129. Plenum Press, N.Y. 



SOLBERT, T., AND S. TlLSETH. 

1984. Growth, energy consumption and prey density 
requirements in first feeding larvae of cod ((Indus morhua 
L.). In E. Dahl, D. S. Danielssen, E. Moksness, and P. 
Solemdal (editors). The propagation of cod Gadus morhua 
L., p. 145-166. F10devigen Rapp. Ser. 1. 
Steele, J. H. 

1978. Some comments on plankton patches. In J. H. Steele 
(editor). Spatial patterns in plankton communities, p. 1- 
20. Plenum Press, N.Y. 
TUNGATE, D. S., AND E. REYNOLDS. 

1980. The MAFF on-line counting system. Fish. Res. Tech. 
Rep., MAFF Direct. Fish. Res., Lowestoft, (58), 11 p. 
Werner, R. G., and J. H. S. Blaxter. 

1980. Growth and survival of larval herring (Clupea harengus) 
in relation to prey density. Can. J. Fish. Aquat. Sci. 
37:1063-1069. 
WlBORG, K. F. 

1948a. Experiments with the Clarke-Bumpus plankton sam- 
pler and with a plankton pump in the Lofoten area in 
Northern Norway. Fiskeridir. Skr. Ser. Havunders. 
9(2):l-32. 

1948b. Investigations on cod larvae in the coastal waters of 
Northern Norway. Fiskeridir. Skr. Ser. Havunders 
9(3):l-27. 



156 



EFFECTS OF SIZE AND TIME OF RELEASE ON 

SEAWARD MIGRATION OF SPRING 

CHINOOK SALMON, ONCORHYNCHUS TSHAWYTSCHA 

R. D. Ewing,' C. E. Hart, 2 C. A. Fustish, 3 and 
Greg Concannon' 

ABSTRACT 

Juvenile spring chinook salmon, Oncorhynchus tshawytscha, from Round Butte Hatchery on the Deschutes 
River, Oregon, were released monthly into a 3.7 km fish ladder. Fish released into the ladder from February 
to May migrated through the ladder in mid-May in both 1977 and 1978. Fish released after mid-May 
migrated through the ladder within 2 weeks after release. The extent of migration decreased progressively in 
fish released after 15 June. The migration was presumably photoperiod dependent, although temperature 
may have acted both as a releasing factor for migration and as a stimulus for growth. In the fish ladder, size of 
the fish remained constant over a 3-week migration period, suggesting that larger fish migrated before 
smaller fish. After a migration of 213 km, fish captured at the Dalles Dam had very large apparent growth 
rates, suggesting that larger fish were faster migrants. 



Maximum survival of juvenile salmonids after 
release from hatcheries is dependent upon their 
rapid migration to the sea (Raymond 1979). Delays in 
this seaward migration may subject the juveniles to 
starvation and stress which rapidly deplete their 
numbers (Miller 1952, 1958). Residual hatchery 
juveniles in a river often have an impact on wild 
stocks of fish through piscivory (Sholes and Hallock 
1979) and competition for food (Chapman 1966). 
Rapid migration of hatchery juveniles ensures max- 
imum survival to adulthood with minimal interaction 
with wild stocks. 

Timing and duration of the physiological conditions 
which result in migratory behavior are still relatively 
unknown. Timing of seaward migration in juvenile 
salmonids depends upon a number of environmental 
factors, including photoperiod (Wagner 1974), tem- 
perature (Solomon 1978), water flow (Mains and 
Smith 1964), and fish size (Wagner 1974). The 
interrelationships between these are not well 
understood, but the available data suggest that these 
relationships may be complex. Hoar (1958) and 
Baggerman (1960) have postulated that these 
environmental factors act as "releasers" which, in 
conjunction with a physiological readiness to 
migrate, trigger overt migrational behavior. 



'Corvallis Fish Research Laboratory, Oregon State University, 
Corvallis, OR 97331. 

department of Zoology, Oregon State University, Corvallis, OR 
97331. 

'Oregon Department of Fish and Wildlife, Research and Develop- 
ment Section, Corvallis, OR 97331. 



Manuscript accepted August 198:!. 

FISHERY BCLLETIN: VOL. 82. NO. 1. 1984. 



In most river systems, the relative influence of such 
factors is estimated by extensive sampling programs 
which use multivariate analysis of the data. Control 
of environmental variables in such a system is not 
possible. Furthermore, the size of many river sys- 
tems prevents an unbiased sampling of juveniles dur- 
ing migration. It is difficult, therefore, to obtain 
reliable estimates of the size of fish at migration, the 
timing of migration, and the influence of the environ- 
ment on that timing. 

In the present study, an unused fish ladder provided 
a relatively constant environment for migration of 
juvenile spring chinook salmon, Oncorhynchus 
tshawytscha, over a 3.7 km distance. Serial releases 
of hatchery-reared juveniles into this system permit- 
ted an investigation of the timing of seaward migra- 
tion, the duration of the migration tendency of the 
juveniles, and the relationship of several environ- 
mental variables to seaward migration. 

METHODS 
Study Area 

The study area included the lower 175 km of the 
Deschutes River, Oreg., and the lower Columbia 
River from its confluence with the Deschutes River to 
the Dalles Dam (Fig. 1). 

Rearing Conditions 

Progeny from spring chinook salmon spawned at 

157 



FISHERY BULLETIN: VOL. 82, NO. 1 




PELTON REGULATION DAM< 

P€LTOn\ 
PELTON DAM<^~ - LADOER 

c^., k .^ D .,-r-r,- ^... jUrOUND BUTTE HATCHERY 
ROUND BUTTE DAM^SS 

-V-I80 

FIGURE 1.— Map of the lower 175 km of the Deschutes River and its 
confluence with the Columbia River. Numbers refer to kilometers 
from the mouth of the Deschutes River. 



Round Butte Hatchery (river km 175 from the 
Columbia River) in 1976 and 1977 were used for 
experiments in 1977 and 1978, respectively. Eggs 
from 1976 brood fish were incubated in Heath 4 
incubators in 10 U C spring water, and the resulting fry 
were reared in raceways using the same water source. 
Eggs from 1977 brood fish were divided into two 
groups. One group was reared under conditions as 
described above and referred to as "fast-reared". 
The second group of eggs was incubated in Heath 
incubators in spring water chilled to 5°-6°C. The 
resulting fry were transferred to raceways and reared 
in 7°-8°C tail-race water from Round Butte Dam. 
After 2 mo, the group was transferred to 10°C spring 
water and reared there until release. This group was 
referred to as "slow-reared" and was released in 
March 1979 as yearlings. 



■"Reference to trade names does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 



In May and June, production lots of fast-reared 
spring chinook juveniles were released into the 
Deschutes River below Pelton Regulation Dam. At 
this time, experimental groups were transferred to 
oval fiber glass ponds supplied with 10 U C spring wa- 
ter at 9.5 1/s. In May 1977, 5,600 fast-reared spring 
chinook juveniles (average fork length 10.0 cm) were 
transferred to a fiber glass pond and reared there 
through June 1978. In late March 1978, 2,500 fast- 
reared fish (average fork length 8.5 cm) were 
transferred to a fiber glass pond and reared there 
through August. 

All fish were reared under a natural photoperiod 
and fed to repletion daily with Oregon Moist Pellet. 

Seaward Migration 

Migratory behavior of the spring chinook salmon 
was assessed by the release and recapture of 
hatchery-reared juveniles from two groups. Migra- 
tion tendency of the experimental groups was 
assessed by monthly release of about 200 fish into 
the upper end of Pelton ladder during 1977 and 1978 
(Fig. 1). The ladder is 3.7 km long and is constructed 
with concrete walls and bottom except for a 1.1 km 
central section which is a natural stream channel. It is 
supplied with water from Lake Simtustus (directly 
above Pelton Dam) at a constant flow rate of 1,130 
1/s. Maximum depth of the ladder is 2. 1 m. The ladder 
is closed by revolving screens at both the upper and 
lower ends. A trap located at the lower end of the lad- 
der was used to capture migrants. Temperature of 
the water at the lower end of the ladder was measured 
by a thermograph placed near the trap. 

Fish from the various experimental groups were 
identified upon recapture in the trap at the lower end 
of the ladder by unique combinations of polystyrene 
dye (Phinney et al. 1967) and fin clips. The trap was 
checked 5 d a week during May and June and 2 d a 
week during the remainder of the year. Fish captured 
in the trap were considered migrants while those 
remaining in the ladder following the date of peak 
recapture were assumed to be residuals. Fork lengths 
and marks of each migrant were recorded upon cap- 
ture. In January 1978, the ladder was drained and all 
residual fish from the 1977 studies were removed 
before the 1978 releases. 

The second group of hatchery-reared fish used for 
assessment of migration were production lots of fast- 
reared juvenile chinook released into the Deschutes 
River immediately below Pelton Regulation Dam 
(river km 161). These fish were marked with coded 
wire tags (Jefferts et al. 1963). In 1977, 62,000 fast- 
reared fish were released on 2 May and 73,000 fast- 



158 



EWING ET AL.: EFFECTS OF SIZE AND RELEASE TIME ON SALMON 



reared fish were released on 3 June. These fish 
averaged 9.7 cm and 11.2 cm FL, respectively. On 31 
May 1978, 121,000 fast-reared fish, which had been 
graded according to fork length, were released in two 
groups of 95,000 and 26,000 fish to test the effects of 
size on migration and survival to adulthood. These 
fish averaged 10.9 and 11.8 cm FL, respectively. 
Downstream movement in both years was monitored 
in the Columbia River at the Dalles Dam (52 km 
downstream from the mouth of the Deschutes River) 
by gatewell sampling conducted by the National 
Marine Fisheries Service and the Oregon Depart- 
ment of Fish and Wildlife. Sampling was conducted 
5 d a week throughout May and June. Juveniles 
originating at Round Butte Hatchery were identified 
by analysis of coded wire tags. 

Apparent Growth Rates 

Apparent growth rates in Pelton ladder and in the 
Deschutes River were calculated from the size of the 
juveniles released into the ladder or the river and the 
size and time at which they were recaptured. Actual 
growth rates could not be measured, because selec- 
tive mortality of small fish or migration of larger ones 
could not be estimated. Differences in fork lengths 
were tested for significance at the 95% confidence 
level using Student's t test. 

RESULTS 

Timing of Migration 

Maximum migration of chinook salmon juveniles 
released in February and March into Pelton ladder 
occurred between mid-May and the first of June in 
both 1977 and 1978. There was little migration in 



these groups before or after this 4-wk period (Tables 
1, 2). Fish released in April showed two peaks in 
migration. A large percentage of the fish moved 
through the ladder within 2 wk after release, while a 
second peak of migration occurred during the last 2 
wk of May. Fish released in early May also had a large 
percent migration within 2 wk after release, but the 
greatest percent migration occurred during the first 2 
wk in June. When chinook salmon juveniles were 
released from June to November, most of the fish 
moved through the ladder within 7 d after release. 
The maximum percent migration within 7 d after 
release occurred in fish released in early June 1977 
(Fig. 2) and in mid-June 1978 (Fig. 3). Fish released 
in August and at later times had reduced migration 
and had a higher tendency to become residual 
(Tables 1, 2). Migration of slow-reared fish released 
into Pelton ladder from May to August 1978 was less 
than half that of fast-reared fish released at the same 
time (Fig. 3B). 

Daily migrations of two groups released in February 
and March 1978 were compared with those from 8 
May to 8 June. Movement of both groups was coin- 
cidental throughout this period (Fig. 4), suggesting 
that environmental factors such as temperature 
influenced migration tendency. Temperatures in the 
ladder varied seasonally due to solar warming (Fig. 
5). Maximum temperatures of 17°C were attained in 
August 1977 and in July and August 1978. Tem- 
peratures in both years exceeded 13 °C by June, sug- 
gesting a possible temperature threshold for 
migration. While the relationship between migration 
and temperature was very poor (correlation coeffi- 
cient, R 2 = 0.074), there may have been a tendency 
for peaks in seaward migration to occur 1-2 d after 
transient increases in temperature (Fig. 4). 



IOO 

z 
o 

5 80-1 

a: 

o 

5 60 



S 40 



20 



IOO 



RELEASE M 
DATE 8 



M 
31 



A 
12 



IOO 

H 



200 



M 
2 



M 
II 



J 
14 



J 
12 



A 
9 




14 



N 
16 



X LENGTH 7 3 8 5 91 97 IOI 112 12 134 14 9 168 18 19 1 

FIGURE 2. — Percentage seaward migration within 7 d following 
release for each group of fast-reared spring chinook salmon released 
into Pelton ladder in 1977. Above each bar is the number of fish 
released. Lengths are means of samples of 30 fish taken from the 
population at the time of release. 



IOO 



80 



60 



40- 



20 



I, OOP 



RELEASE 
DATE 



F 
14 



M 
15 



A 
15 



M 
15 



J 
15 



J 

14 



A 

15 



XLENGTH 63 80 89 99 116 129 149 



M 
15 



n 



j 

15 



J 
14 



83 98 114 



A 
15 

127 



FIGURE 3. — Percentage seaward migration within 7 d following 
release for each group of spring chinook salmon released into Pelton 
ladder in 1978. A) Fast-reared chinook salmon. B) Slow-reared 
chinook salmon. Above each bar is the number of fish released. 
Lengths are means of samples of 30 fish taken from the population at 
the time of release. 



159 



FISHERY BULLETIN: VOL. 82, NO. 1 
Table 1.— Percentage downstream migration for fast-reared spring chinook salmon released into the Pelton ladder in 1977. 



Release date: 


8 Mar. 


31 Mar 


12 Apr, 


2 May 


11 May 


3 June 


14 June 


12 July 


9 Aug 


9 Sept. 


15 Oct. 


16 Nov. 


Capture X length (cm): 


7.2 


8 5 


9.1 


9 7 


10 2 


11.2 


120 


13.4 


14 9 


16.8 


18.0 


19.1 


dates n: 


200 


99 


194 


100 


200 


100 


198 


198 


200 


199 


200 


175 


3/1-3/15 


3 5 
























3/16-3/31 


1.0 
























4 1-415 


10 


1.0 


17 5 




















4/16-4/30 








5 




















5/1-5/15 


1 5 


1.0 


3 


S 'i 


370 
















5/16-5/31 


34.5 


420 


27 


19.0 


5 5 
















6/1-6/15 


8.0 


16 


180 


37.0 


355 


78 


7.0 












6/16-6/30 


0.0 


1.0 


0.5 


6.0 


3.0 


40 


35 












7/1-7/15 


5 


0.0 


0.0 


1)0 


5 


1 


20 


40.0 










7/16-7/31 


5 








1.0 


0.0 


0.0 


o 5 


18.5 










8/1-8/15 


00 





00 





o o 


0.0 





05 


260 








8/16-8/31 


00 


0.0 


5 








1 


1.0 


5 


7 5 








9/1-9/15 


0.0 


0.0 


0.0 


1.0 


o o 


1.0 


3 5 


b 


2.0 


29 






9/16-9/30 








o o 


0.0 


0.0 








00 





1 


3 






10/1-10/15 


0.0 


0.0 


i) 5 














o 


0.0 


0.0 


00 






10/16-10/31 


0.0 








0.0 


0.0 








0.0 


0.0 


0.0 


0.0 


7.5 




11/1-11/15 


00 





0.0 


0.0 











0.0 


0.0 


0.0 


00 


0.0 




11/16-11/30 


0.0 


0.0 








oo 











00 


05 


0.5 


2 5 


50 


12/1-12 15 


0.0 





0.0 


00 


00 








0.0 


0.0 


00 


1.5 


1.5 


0.5 


12/16-12 31 


00 





on 





I 1 


(III 


0.0 


0.0 


0.0 


00 


00 


00 


1/1-1/15 


0.0 











00 





0.0 





0.0 


2 


1.5 


1 


Total percentage 


























migration 


505 


61 


62 


73.0 


81.5 


85.0 


490 


60.0 


37 


36 


13.0 


6 5 


Percent 


























residuals 





0.0 





2 





00 





2.0 


14.5 


31.5 


43 5 


55 


Total percentage 


























recovered 


50.5 


61 


62 


750 


81.5 


850 


49.0 


620 


51 5 


67 5 


56.5 


61.5 



TABLE 2. — Percentage downstream migration over semimonthly intervals for fast-reared spring 
chinook salmon released into the Pelton ladder in 1978. 





R, 


lease daie 


14 Feb. 


15 Mar, 


1 5 Apr. 


15 May 


1 5 June 


14 July 


1 5 Aug. 


Capture 


.X 


ength (cm): 


6 3 


80 


89 


99 


116 


12.9 


14.9 


dates 




n: 


1,000 


199 


198 


192 


96 


192 


200 


2/15-2/28 






1 














3/1-3/15 





















3/16-3/31 









3 












4/1-4/15 






















4/16-4/30 






0.1 


1 


37 










5/1-5/15 






17 3 


80 


2.5 










5/16-5/31 






648 


62 8 


328 


41 7 








6/1-6/15 






3 8 


8.5 


12 


33 8 








6/1 6-6/30 






1.7 





1.0 


3.1 


77 






7/1-7/15 






5 











00 






7/16-7/31 






1 3 


1.0 


0.0 





2.0 


55.0 




8/1-8/15 






Oil 


0.0 


0.0 


0.0 


o 


5 




8/16-8/31 






7 





0.0 


5 


1.0 


5 


54 5 


9/1-9/15 






2.2 


..I \, 





2 5 


1 


2 5 


50 


Total percen 


m 


gration 


92 5 


84.8 


85 3 


81 6 


81 


58 5 


59 5 



Recovery of Released Fish 

In 1978, the greatest recovery of fish liberated into 
Pelton ladder (92.5%) was from the large group of 
1,000 fish released on 14 February (Table 2). From 
81.0 to 85.3 f /(: of the fish released from 15 March 
through 15 June were recovered. Only 58.5 and 
59.5 f 7f of the fish released on 14 July and 15 August, 
respectively, were recovered in the trap as migrants. 
Presumably the remainder were residuals in the 
ladder. 

In 1 97 7, recovery of both migrants and nonmigrants 
from all groups was lower than in 1978 (Table 1), 



although the extent of migration of fish released near 
the time of maximum migration tendency on 11 May 
and 3 June was 8 1.5 and 85 c /'c, respectively, similar to 
that observed for most release groups in 1978. Few 
residual chinook salmon from releases before August 
1977 were found when the ladder was drained in 
January 1978. Nonmigrant fish were recaptured in 
increasing numbers from releases from 12 July on. 

Size and Growth Relationships 

Growth rates of juvenile chinook salmon reared at 
Round Butte Hatchery were 0.046 and 0.058 cm/d 



160 



EWING ET AL.: EFFECTS OF SIZE AND RELEASE TIME ON SALMON 

r l6 



TABLE 3. — Apparent growth rates of juvenile chinook salmon re- 
leased into Pelton ladder, 1977 and 1978. 



■Z. 16-, 

o 








-j 




Release 


Av 


srage reca 


pture 


Apparent growth 


i= '«" 


A 2 






m 
m 




date 




date 




rate (cm/d) 


or 12-1 


1977 


3/8 




5/24 




0078 


- 10 J 


. . t /  V | 


v // \ 




3) 

• 5j 




4/12 
5/2 




6/1 
5/28 




0081 
0.124 




y \ /v\Js\ / ; 




33 

12 m 


Hatchery 


5/11 




6/4 




0.120 
0.048 


V b- 


\ / A i V < 


i*f V 




o 


1978 


2/14 




5/27 




0034 


< "- 








O 




3/15 




5/20 




0071 


Q 2 - 


vj-' ' 










4/15 




5/26 




097 






-*~»— • -^ti 


-10 
3 


Hatchery 


5/15 




6/7 




097 




8 10 12 14 16 18 20 22 


24 26 28 30 


13 5 7 

JUN 


046 


MAY 















I978 



Fic.l'RE 4. — Daily percent seaward migration from 8 May to 8 June 
for groups released 14 February (solid circles) and 15 March (open 
circles). Temperature (triangles) is the average daily temperature. 



20 



or 

or 

UJ 



o 

S 

UJ 

< 

or 

UJ 




4 - 



J I L 



JAN MAR MAY JUL SEP NOV 

FEB APR JUN AUG OCT DEC 



FIGURE 5. — Average monthly temperature in Pelton ladder 
in 1977 (solid circles) and 1978 (open circles). 



TABLE 4. — Fork lengths of juvenile chinook salmon at time of release 
into Pelton ladder and at time of recapture, 1977 and 1978. Values 
are means ± standard errors. Number of samples is given in 
parentheses. 



Date 


Mean fork 1 


Bngth 


Date 


Mean fork I 


sngth 


of 


at release 


of 


at recapture 


release 


(cm) 




recapture 


(cm) 




1978: 












2/14 


6.5±0 1 


60) 


5/15-6/9 


13 4+0.1 


245) 


3/15 


8 2±0.1 


30) 


5/16-5/24 


13.0+.0 1 


81) 


4/15 


9.0±0.1 


100) 


5/22-5/31 


13.0+0 2 


42) 


5/15 


10.1 ±0.1 


29) 


6/5-6/9 


12 3±0 1 


54) 


6/15 


1 1 3±0 1 


30) 


6/16 


1 1.9±0 1 


30) 


7/14 


13 7±0 .1 


30) 


7/17 


13 2±0.2 


30) 


8/15 


15 1±03 


30) 


8/16 


15.5+0 4 


9) 


9/15 


17.1±0.3 


28) 


9/18 


17.2+.0.2 


30) 


1977 












3/8 


7.5±0 1 


30) 


5/24 


13 5±0 1 


25) 


3/31 


8.5±0.1 


30) 


5/24-5/28 


13.5±0.1 


26) 


4/12 


9 4+0 1 


30) 


6,1 


13.4±0 1 


26) 


5/2 


9 7±0 1 


30) 


5/28 


12 9±0 1 


7) 


5/11 


9.9±0 .1 


29) 


5/12-6/4 


11 7±0 1 


61) 


6/3 


1 1.2±0 1 


30) 


6/3 


11.4±0 1 


25) 


6/14 


12.0+.0.1 


30) 


6/17 


12 3±0 .1 


30) 


7/12 


1 3 5±0 1 


60) 


7/15 


1 3 9±0 1 


28) 


8/9 


15.1±02 


88) 


8/15 


16 1+02 


20) 


9/9 


16 7±0 2 


88) 


9/9-9/13 


16.8±0 2 


42) 


10/15 


17 5±0.5 


30) 


10/17 


18 4+0 4 


15) 



for fast-reared fish in 1977 and 1978, respectively. 
Slow-reared fish in 1978 grew at 0.043 cm/d. 
Apparent growth rates offish placed in Pelton ladder 
varied from 0.034 to 0.124 cm/d (Table 3). These 
apparent growth rates increased in later introduc- 
tions, reflecting the increasing water temperature of 
the ladder (Fig. 5). 

There was no evidence for differences in migration 
timing by fish of different sizes. Fork lengths of fish 
recaptured in the trap within a few days of release 
were usually not significantly different (P > 0.05) 
from those of fish at release (Table 4). However, fish 
recaptured from the large group of juveniles released 
on 1 4 February 1978 were similar over a 3-wk period 
(Table 5), suggesting that faster growing fish were 
migrating more rapidly that slower growing fish. 

Apparent growth rates of marked spring chinook 
juveniles released below Pelton Regulation Dam in 
1977 were calculated from fork lengths of recaptured 
fish at the Dalles Dam, after a migration distance of 



TABLE 5.— Mean fork lengths of juvenile 
spring chinook salmon recovered in 1978 
after release into Pelton ladder on 14 Feb- 
ruary 1978. Values are means ± standard 
errors. Number of samples is given in 
parentheses. 



Date of 
recovery 



Fork length (cm) 



2/17 

5/15 

5/16 

5/18 

5/22 

5/24 

5/30 

6/1 

6/5 

6/9 



67+01 


(30) 


12.8±0.1 


(30) 


13 2±0.1 


(30) 


13 1±0.1 


(30) 


13.4±0.1 


(30) 


13.2±0.1 


(30) 


13.9±0.1 


(30) 


13 5±0.2 


(19) 


13.6±0.1 


(21) 


11.2±0.1 


(25) 



213 km (Table 6). This apparent growth rate is nearly 
twice that of fish reared at Round Butte Hatchery. 



181 



FISHERY BULLETIN: VOL. 82. NO. 1 



TABLE 6.— Fork lengths and apparent growth rates of juvenile 
spring chinook salmon recaptured at the Dalles Dam after release 
into the Deschutes River. 1977. Fork lengths are means ± standard 
errors for the number of samples shown in parentheses. 





Apparent growth rate 


Recapture date Fork length (cm) 


(cm/d) 


2 May release (9 7±0.1 cm fork length) 




5/27 11.6±0 2 (12) 


0075 


6/3 12.2±0 1 (19) 


0078 


6 7 12 2±0 1 (22) 


0.077 


6/8 12.2±0.1 (21) 


0076 


3 June release (1 1 2±0 1 cm fork length) 




6 7 11.9±0.1 (23) 


0.168 


6/8 11.8±0.1 (30) 


120 



DISCUSSION 

Determination of the migratory characteristics of 
juvenile chinook salmon during smolting has been 
complicated by the variety of migratory behaviors 
displayed by the juveniles. Some fry migrate from 
tributaries shortly after emergence from the gravel 
(Reimers 1973; Ewing et al. 1980), but there is little 
evidence that the fry move into the estuary at that 
time (Schluchter and Lichatowich 1977). In some 
stocks, a general movement of fish through the river 
occurs during the fall of the first year (Reimers 1973) 
with a majority of the fish entering the ocean during 
the fall of the first year (Reimers 1973; Schluchter 
and Lichatowich 1977; Buckman and Ewing 1982). 
In other stocks, seaward movement occurs primarily 
in the following spring when the fish are more than 1 
yr old (Mains and Smith 1964; Diamond and Pribble 
1978: Raymond 1979). Krcma and Raleigh (1970) 
reported migration of juvenile chinook salmon into 
Brownlee Reservoir (Snake River, Idaho) in fall and 
spring for 2 consecutive years. The migration pattern 
seems to depend upon stock, size, and rearing con- 
ditions and may be highly variable. It is therefore 
important in the culture of various stocks of juvenile 
chinook salmon to determine the timing of maximum 
migration tendency. 

In the present study, the major migration of fish 
released early into Pelton ladder occurred in mid- 
May. Fish from the same brood released into the 
Deschutes River at about this time were found to 
migrate 213 km to the Dalles Dam within 7 d, sug- 
gesting that the migrational behavior was seaward 
directed (Hart et al. 1981). It is difficult to confirm in 
the Deschutes River that the release of fish into 
Pelton ladder 1 mo before the time of maximal migra- 
tion tends to increase the time during which the fish 
will migrate. Release of the fish 1 mo later than the 
time of maximal migration tends to decrease the time 
for migration. It is important to note that it is not 
necessary to release the fish early to insure that all 



migrate to sea. Releases late in the migration period 
were recovered to the same extent as those released 
earlier. Migration tendency seems to be retained for 
some time, even though the fish are not permitted to 
begin migration. These results suggest that late re- 
leases hasten the seaward migration, thus removing 
the populations of hatchery fish quickly from the 
river system and affording maximum protection to 
the wild stocks. 

Those groups released later than July were recap- 
tured in the trap in decreasing numbers (Tables 1,2). 
In 1977, nonmigrant fish were recaptured in increas- 
ing numbers from releases after 12 July (Table 1). 
This result indicates that the decrease in numbers of 
fish recaptured at the trap was due to decreased 
migration tendency and not due to increased mor- 
talities at the higher water temperatures. 

A major advantage of utilizing a closed system such 
as the Pelton ladder for studies of migration was that 
fish populations and flows could be effectively con- 
trolled. Variables which remained uncontrolled in- 
cluded photoperiod, lunar periodicity, temperature, 
and food supply. Of these, photoperiod seems the 
most important in stimulating seaward migration. 
Previous studies utilizing a closed system for study- 
ing seaward migration of steelhead trout, Salmo 
gairdnvri, (Zaugg and Wagner 1973; Wagner 1974) 
and coho salmon, Ocorhynchus kisutch, (Lorz and 
McPherson 1976) also concluded that photoperiod 
was an important factor affecting the timing of sea- 
ward migration. 

Lunar phase has been suggested to affect the onset 
of migration, based on the correlation between peaks 
in plasma thyroxine levels and lunar phase (Grau et 
al. 1981). Assuming maximal migration occurred on 
22 May in both 1977 and 1978, this date correspond- 
ed to the time of a new moon in 1977 and that of a full 
moon in 1978. These brief data do not support the 
hypothesis that the migration is influenced by the 
lunar phase. 

Temperature may have had a dual influence on 
migration. Temperature has been suggested as a 
releasing factor for salmon migration (Hoar 1958; 
Baggerman 1960), but we were unable to show a 
statistical relationship between daily migration and 
average daily temperature (Fig. 4). 

Temperature also serves to increase growth rates in 
salmonids in the presence of abundant food supplies. 
Wagner (1974) suggested that a critical size was 
required in steelhead if migration were to take place. 
The importance of size on migration of spring 
chinook salmon can be seen by comparing the extent 
of migration of the slow- and fast-reared fish in 1978 
(Fig 3). The slow-reared fish may have failed to mi- 



162 



EWING ET AL.: EFFECTS OF SIZE AND RELEASE TIME ON SALMON 



grate because they did not reach a critical size and/or 
growth rate by the appropriate photoperiod. Migra- 
tion from Pelton ladder seemed to occur as fish 
reached a particular size, since during a 3-wk period 
of migration, there was no difference in average 
fork length of the fish recaptured (Table 5). From 
estimated growth rates (Table 3), fish at the end of 
the migration period might be expected to be nearly 2 
cm larger than those at the beginning. This influence 
of size on migration could be best demonstrated in 
fish recaptured at the Dalles Dam after a migration 
distance of 213 km. Apparent growth rates were 
much higher than that of fish reared at Round Butte 
Hatchery, suggesting that a selection for larger fish 
occurs during the long migration distance. 

A major concern in utilizing a closed system for 
studying seaward migration is the importance of 
aggressive behavior by resident fish toward newly 
introduced fish. Chapman (1962) found that aggres- 
sive behavior of resident fish may be partly respon- 
sible for emigration of fish introduced into the 
system. Aggressive behavior may have caused the 
rapid movement immediately following release for 
the March and April release groups in both 1977 and 
1978. Further movement of these fish was not ob- 
served until May. Alternatively, migration in these 
fish immediately after release may have been due to 
disorientation of the fish upon release and a passive 
drifting downstream with the current. Fish released 
earliest into Pelton ladder migrated first in both 1977 
and 1978. 

The importance of determining appropriate times 
for hatchery releases of spring chinook salmon in 
order to obtain maximum seaward migration is de- 
monstrated by the short time during which maximum 
migration occurred (Tables 1, 2). In both 1977 and 
1978 peak migration occurred within a period of a few 
weeks. Releases made on either side of this time 
period exhibited decreased migratory activity. The 
use of model systems, such as the Pelton ladder, to 
determine when peak migration occurs can benefit 
hatchery programs by suggesting sizes and times for 
release of salmonids which maximize seaward mi- 
gration. 

ACKNOWLEDGMENTS 

We thank the members of the Deschutes River 
Salmon Study, Round Butte Hatchery personnel, 
and biologists of Portland General Electric Company 
for their help and cooperation throughout this study. 
We acknowledge the special assistance of Ray Hill, 
Jerome Diamond, Zeke Madden, Garet Soules, and 
Richard Aho. This study was supported by a grant 



from Portland General Electric Company to the 
Oregon Department of Fish and Wildlife. 

LITERATURE CITED 

Baggerman. B. 

1960. Factors in the diadromous migrations of fish. Symp. 
Zool. Soc. Lond. 1:33-58. 
BUCKMAN, M., AND R. D. EWING. 

1982. Relationship between size and time of entry into the 
sea and gill (Na+K)-ATPase activity for juvenile spring 
chinook salmon. Trans. Am. Fish. Soc. 111:681-687. 
Chapman, D. W. 

1962. Aggressive behavior in juvenile coho salmon as a cause 
of emigration. J. Fish. Res. Board Can. 19:1047-1080. 

1966. Food and space as regulators of salmonid populations 
in streams. Am. Nat. 100:345-357. 
Diamond, J., and H. J. Pribble. 

1978. A review of factors affecting seaward migration and 
survival of juvenile salmon in the Columbia River and 
ocean. Oreg. Dep. Fish Wildl. Inf. Rep. Ser., Fish. 78-7. 
Ewing, R. D.. C. A. Fustish, S. L. Johnson, and H. J. Pribble. 

1980. Seaward migration of juvenile chinook salmon without 
elevated gill (Na+K)-ATPase activities. Trans. Am. Fish. 
Soc. 109:349-356. 

Grau, E. G., W. W. Dickhoff, R. S. Nishioka, H. A. Bern, and 
L. C. Folmar. 

1981. Lunar phasing of the thyroxine surge preparatory to 
seaward migration of salmonid fish. Science (Wash., D.C.) 
211:607-609. 

Hart, C. E., G. Concannon, C. A. Fustish, and R. D. Ewing. 

198 1 . Seaward migration and gill (Na+K)-ATPase activity of 
spring chinook salmon in an artificial stream. Trans. Am. 
Fish. Soc. 110:44-50. 
Hoar, W. S. 

1958. The analysis of behaviour of fish. In P. A. Larkin 
(editor), The investigation of fish-power problems. Univ. 
Br. Columbia, Inst. Fish., p. 99-111. 
Jefferts, K. B„ P. K. Bergman, and H. F. Fiscus. 

1963. A coded wire identification system for macro- 
organisms. Nature (Lond.) 198:460-462. 

Krcma, R. F., and R. F. Raleigh. 

1970. Migration of juvenile salmon and trout into Brownlee 
Reservoir, 1962-65. U.S. Fish Wildl. Serv., Fish. Bull. 
68:203-217. 

Lorz, H. W., and B. P. McPherson. 

1976. Effects of copper or zinc in fresh water on the adapta- 
tion to sea water and ATPase activity, and the effects of 
copper on migratory disposition of coho salmon 
(Oncorhynchus kisutch). J. Fish. Res. Board Can. 
33:2023-2030. 

Mains, E. M., and J. M. Smith. 

1964. The distribution, size, time and current preferences of 
seaward migrant chinook salmon in the Columbia and 
Snake Rivers. Wash. Dep. Fish., Res. Pap. 2:5-43. 

Miller, R. B. 

1952. Survival of hatchery-reared cutthroat trout in an 

Alberta stream. Trans. Am. Fish. Soc. 81:35-42. 
1955. The role of competition in the mortality of hatchery- 
trout. J. Fish. Res. Board Can. 15:27-45. 
Phinney, D. E., D. M. Miller, and M. L. Dahlberg. 

1967. Mass-marking young salmonids with fluorescent 
pigments. Trans. Am. Fish. Soc. 96:157-162. 
Raymond, H. L. 

1979. Effects of dams and impoundments on migrations of 



163 



FISHERY BULLETIN: VOL. 82. NO. 1 



juvenile chinook salmon and steelhead from the Snake 
River, 1966 to 1975. Trans. Am. Fish. Soc. 108:505- 
529. 
Reimers, P. E. 

197:!. The length of residence of juvenile fall chinook salmon 
in Sixes River, Oregon. Res. Rep. Fish Comm. Oreg. 
4(2):3-43. 

SCHLUCHTER, M. D., AM) J. A. LlCHATOWICH. 

1977. Juvenile life histories of Rogue River spring chinook 
salmon (Oncorhynchus tshawytscha Walbaum) as deter- 
mined by scale analysis. Oreg. Dep. Fish Wildl., Inf. Rep. 
Ser., Fish. 77-5. 

Sholes, W. h.. and R. J. Hallock. 

1979. A evaluation of rearing fall-run chinook salmon, 
Oncorhynchus tshawytscha, to yearlings at Feather River 



Hatchery, with a comparison of returns from hatchery and 
downstream releases. Calif. Fish Game 65:239-255. 
Solomon, D. J. 

1978. Some observations on salmon smolt migration in an 
chalkstream. J. Fish Biol. 12:571-574. 

Wagner, H. H. 

1974. Photoperiod and temperature regulation of smoltingin 
steelhead trout {Salmn gairdneri). Can. J. Zool. 42:219- 
234. 
Zaugg, W. S., and H. H. Wagner. 

1973. Gill ATPase activity related to parr-smolt transforma- 
tion and migration in steelhead trout (Salmo gairdneri): 
Influence of photoperiod and temperature. Comp. 
Biochem. Phvsiol. 45B:955-965. 



164 



INTERACTIVE EFFECTS OF AGE AND ENVIRONMENTAL 

MODIFIERS ON THE PRODUCTION OF DAILY GROWTH 

INCREMENTS IN OTOLITHS OF 

PLAINFIN MIDSHIPMAN, PORICHTHYS NOTATUS 

Steven E. Campana' 
ABSTRACT 

Plainfin midshipman, Porichthys notatus, were reared in the laboratory under three environmental regimes to 
determine the influence of certain variables upon otolith growth increment formation. Both larval and 
juvenile midshipman were used to test diel cycles and constant conditions of light and temperature. Daily 
growth increments were formed upon hatch unless a diel photoperiod was absent. However, under constant 
light, an endogenous circadian rhythm became evident aftera 2-3 week acclimation period, resulting in daily 
increment production. With increasing age, the influence of light as a zeitgeber decreased, while daily 
increments became more prominent in all environments. Temperature fluctuation affected increment 
appearance, but did not entrain increment deposition. 



Daily growth increments in the otoliths of fishes have 
been observed in a large number of species (Pannella 
1971; Brothers et al. 1976; Taubert and Coble 1977; 
Wilson and Larkin 1980). These concentrically 
formed increments may be counted or measured to 
provide a chronological record of past fish growth. 
Information on hatching date/age (Ralston 1976; 
Struhsaker and Uchiyama 1976), daily growth rates 
(Methot 1981), and timing of life history transitions 
(Pannella 1980; Brothers and McFarland 1981) has 
been derived from the examination of otolith micro- 
structure. Such data are difficult to obtain from larval 
and juvenile fishes by other means. 

Daily increments are produced through a diel 
periodicity in the deposition of calcium carbonate on 
the otolith (Mugiya et al. 1981). However, there is 
some controversy as to the zeitgeber behind the daily 
cycle of deposition, if indeed one exists. In a series of 
experiments upon larval Lepomis, Taubert and 
Coble (1977) determined that a 24-h light-dark cycle 
was necessary to entrain an endogenous rhythm of 
increment production. Reversal of the light-dark 
cycle reversed the daily sequence of increment for- 
mation in larval Tilapia (Tanaka et al. 1981). 
However, 36-h "days" and constant light conditions 
had no effect on daily increment production in 
juvenile starry flounders, Platichthys stellatus (Cam- 
pana and Neilson 1982). Similarly, constant light or 



institute of Animal Resource Ecology, University of British 
< Columbia, Vancouver, British Columbia, Canada V6T 1W5; pre- 
sent address: Marine Fish Division, Bedford Institute of Ocean- 
ography, P.O. Box 1006, Dartmouth, Nova Scotia, Canada B2Y 

4A2. 



Manuscript accepted .July 198.'!. 

FISHERY BULLETIN: VOL. 82. NO. 1. 1984. 



dark conditions did not inhibit the formation of daily 
increments in young chinook salmon, Oncorhynchus 
tshawytscha (Neilson and Geen 1982). The con- 
tradictory results of the above studies suggest that 
photoperiod effects on increment production may 
vary with age or species of fish. 

Other environmental variables may influence the 
daily rhythm of otolith deposition. Diel temperature 
fluctuation has been implicated as a factor in daily 
increment production of temperate stream fishes 
(Brothers 1981), although this suggestion has not 
been supported by other studies (Campana and 
Neilson 1982; Neilson and Geen 1982). Feeding fre- 
quency may also influence otolith increment produc- 
tion; fish given multiple daily feedings have been 
reported to produce nondaily increments (Pannella 
1980; Neilson and Geen 1982), although recent 
studies suggest that feeding effects are limited 
(Tanaka etal. 1981 ; Marshall and Parker 1982; Cam- 
pana 1983). 

Confidence in the reliability of otolith microstruc- 
ture examination requires knowledge of those factors 
that may influence otolith increment production. 
Conflicting results in the literature suggest that age 
influences the response of daily increment produc- 
tion to environmental variables such as photoperiod 
and temperature. This study was undertaken to test 
that hypothesis. Plainfin midshipman, Porichthys 
notatus, were reared from the egg stage under various 
light and temperature regimes; constant conditions 
and diel cycles of each variable were tested. The 
effect of the regimes on otolith microstructure was 
noted for both newly hatched and juvenile fish. 

165 



FISHERY BULLETIN: VOL. 82, NO. 1 



Juveniles were then subdivided and transferred to 
different regimes, allowing an examination of the 
interactive influence of greater age and novel 
environment on increment production. 

MATERIALS AND METHODS 

Fertilized Porichthys eggs were collected inter- 
tidally from White Rock, British Columbia, on 9 and 
22 June 1982. Yolk-sac larvae remain attached to the 
rock upon which the eggs were originally deposited 
(Arora 1948), necessitating the collection of both 
rocks and egg masses. Upon return to the laboratory, 
eight separate egg masses (50-250 ova each) were 
isolated in individual saltwater aquaria and main- 
tained under a diel photoperiod and a temperature of 
13°C. Small amounts of methylene blue, strep- 
tomycin sulphate, and penicillin G were used to con- 
trol bacterial and fungal infection. Embryo 
development varied both among and within egg 
masses, but the difference appeared to be <2-3 d. 

On 1 July, egg masses were exposed to an 
experimental environment. Environmental regimes 
were selected to provide a diel periodicity of either 
photoperiod or temperature. A third regime main- 
tained constant conditions of both variables. In this 
manner, the influence of both factors on increment 
formation could be determined for newly hatched lar- 
vae. Daily increment production in the constant 
environment would suggest the presence of an 
endogenous circadian rhythm. Regimes were as 
follows: 

14L:10D at a constant temperature of 19°C 

(14L:10D/CT) 
24L with 14 h at 21 C C and 10 h at 19 C (24L/ 

14TV10T,) 
24L at a constant temperature of 19°C (24L/ 

CT) 

Duplicate aquaria, each containing an egg mass (or 2 
small masses, if at similar developmental stages), 
were kept in light-proof, temperature-controlled 
cubicles under each of the above environments. All 
lighting was fluorescent (30 jiiEs/m 2 /s). Temperature 
fluctuations were timer-controlled and conducted 
parallel to the light cycle. New temperatures were 
reached VA h after initiation. Mean temperatures 
approximated those of the egg collection site; diel 
temperature fluctuations were present at the site, 
but were not recorded. Aquarium water was changed 
at 7-10 d intervals. Hatching date varied among and 
within egg masses, beginning between 7 and 1 1 July. 
Release from the rock (before completion of yolk-sac 



resorption) was more variable, and occurred between 
23 July and 9 August. Live adult Artemia were first 
provided as food on 30 July and were consumed by 
both released and attached larvae. Thereafter, 
Artemia were maintained in all aquaria at all times, 
with the exception of two 3-d periods when food was 
not available. Food abundance did not differ among 
the aquaria. Observations of feeding fish indicated 
that the accessibility of Artemia did not limit 
growth. 

By 10 August, all fish were about 32-d old 
(posthatch) and had become juveniles (i.e., had 
assumed the appearance of an adult). To test the 
effect of an altered photoperiod or temperature cycle 
on juveniles, one tank from each of the environmental 
regimes was subdivided (Fig. 1). About 25 fish were 
transferred from one aquarium ("cohort") to each of 
the remaining environments, while leaving 25 fish in 
the original environment as a control. Sagittae were 
removed from up to 25 of the excess fish to determine 
the effect of the original environment on newly 
hatched larvae. In order to remove any intercohort 
variability of hatching dates, only one of the two avail- 
able cohorts from each environment was subdivided 
and sampled. However, low numbers of 1 4L: 1 0D/CT 
fish necessitated the transfer of an entire cohort. 

For processing, the sagittae were brushed free of 
tissue and glued sulcus-side up with instant glue on a 
standard microscope slide. Sagittae were ground and 
polished with metallurgical lapping film (grit size 30 



AUG 10 
JULY 1- AUG 10 TRANSFER 



14L 10D/CT 



24L/ 14T, 10T 



24L/CT 



AUG 10- SEPT 10 



24L/14T, 10T 2 141 10D/CT 24L/CT 



Fk;i RE 1. — Summary of experimental environmental regimes of 
plainfin midshipman through time. Fish transferred to new environ- 
ments on 1(1 August came from the same egg mass as that sampled 
on 10 August. 



166 



CAMP ANA: DAILY GROWTH INCREMENTS IN OTOLITHS OF PLAINFIN MIDSHIPMAN 



jum to 0.3 jixm) until the growth increments in the 
region of maximal growth were most visible. I defined 
a growth increment as a bipartite structure, consist- 
ing of a narrow opaque band and an adjacent broad 
translucent region. Growth increments between the 
otolith periphery and the hatch check were counted 
at least twice through a compound microscope at a 
magnification of 400X. Duplicate counts of an otolith 
never differed by more than 10 f A. The use of a hand 
counter eliminated the possibility of a count converg- 
ing on an expected value. There was little doubt con- 
cerning the nature of the hatch check; its radius 
matched that of radii of otoliths removed at hatch. 
Growth increments in 14L:10D/CT fish sampled 10 
September were counted as above. However, a 
second series of counts was made from the hatch 
check to the prominent 10 August check; the second 
data set served as a substitute for the actual sampling 
of 14L:10D/CT fish on 10 August. 

Increment counts were made from both the left- and 
right-hand side sagittae. Since the two sides did not 
differ systematically under any of the environments 
(paired /-test, P > 0.05), the means were used in all 
data analyses. 

Increment widths were measured from photo- 
graphs with a micrometer. Expected increment 
widths were calculated from radial measurements 
(central nucleus to rostral tip) of otoliths from all 
environments and a variety of sampling dates (N = 
10 per date). Values for mean increase in radial 
otolith growth per day were then compared to ob- 
served values. 

Since individual otoliths often displayed erratic but 
discernable width trends through time, a measure of 
the similarity of the widths of two adjacent daily 
increments was calculated: 



IR, 



w,- w;._, 

(W, + W^/2 



where IR, is the index of increment width regularity 
for day/, and W, is the increment width for day (.Such 
an index gives low values when adjacent increments 
are similar in width, despite any trends in the data. 
Index values were calculated for a range of ages in 
otoliths from a given environment. 

RESULTS 

Porichthys larvae and juveniles survived and grew 
under all laboratory environments. Survival ex- 
ceeded 95 '/< after hatch. Fish sampled about 1 mo 
after hatch (10 August) did not differ significantly in 
standard length (ANOVA, P > 0.05). By the end of 



the study, only those fish maintained in the 24L/ 
1 4T, : 1 0T 2 environment were significantly smaller in 
length (Scheffe's test P < 0.01); the difference was 
apparently due to unintentional overcrowding from 
the date of transfer. 

Hatching was initiated simultaneously in two of the 
three initial environments, but started 4 d later in the 
24L/CT aquaria. The delay did not appear to be due 
to the artificial environment, since embryo develop- 
ment among the 24L/CT egg masses lagged behind 
that of the others at the time of collection. In the 
aquarium, about 959? of the viable ova hatched 
within 4 d of hatch initiation. Intratank hatch-date 
variance would be expected to affect the variance of 
increment counts. However, the 17-d range of larval 
release dates (from the rock) was not reflected in the 
otolith microstructure. 

Unground sagittae derived from both pre- and 
posthatch fish were extremely lobulated in structure. 
The origin of the numerous lobes was 5-10 
"peripheral" nuclei, from which the majority of the 
growth increments emanated. A central nucleus also 
had growth increments associated with it, although 
these were incorporated into the peripheral incre- 
ments within 10-20 d/increments. A prominent hatch 
check occurred within 5-10 major increments of the 
central nucleus. The most prominent check of the 
older otoliths was that associated with the sub- 
division/transfer date of 10 August. 

Many growth increments were visible in the 
polished otoliths sampled after hatch. When plotted 
as a function of time, total increment counts were 
significantly greater than those expected of daily pro- 
duction (P < 0.05) (Fig. 2). Diel light and tempera- 
ture cycles both produced an increment: age slope of 
about 3.0, suggesting that numerous subdaily 
increments were being counted with any daily 
increments present. Increment clarity, prominence, 
and width varied substantially within an otolith. 
However, most increments could be assigned to one 
of two "levels" — visually prominent/relatively wide 
and visually faint/relatively narrow. To determine if 
the first level consisted primarily of daily increments, 
the expected width of a daily increment was 
calculated. 

23 July 30 July 9 Aug. 10 Sept. 
Mean otolith 

radius (jam): 270 430 620 875 

Daily increments on the order of 12-23 and 5-8 fim 
wide would be expected in the first and second month 
posthatch, respectively. These expected increment 
widths were similar to those observed in the first 
"level" of growth increments. 



167 



FISHERY BULLETIN: VOL. 82, NO. 1 



200-1 




• =14L10D/CT 
a = 24L/14Ti:10T 2 



40 
AGE (DAYS) 



50 



70 



FIGURE 2.— Total otolith increment count as a function of age for 
plainfin midshipman from two cyclic experimental environments. A 
straight line has been fitted to the data, although the relationship is 
probably curvilinear. N = 5 for each data point. 



Criteria for distinguishing daily from subdaily 
increments have been reported previously (Taubert 
and Coble 1977; Campana and Neilson 1982; 
Marshall and Parker 1982). Nevertheless, no objec- 
tive criteria have yet been defined which can be 
applied to all otoliths. In this study, I have used visual 
prominence and increment width as guides for dif- 
ferentiating daily and subdaily increments. In- 
crements assigned as daily were 1) of similar visual 
prominence (contrast) to adjacent daily increments 
(±30%), 2) of similar increment width to adjacent 
daily increments (±50%), 3) not merged with adja- 
cent daily increments in the nearest radial groove of 
the sagitta. Some increments met only some of the 
criteria and were subjectively assigned as daily or 
subdaily. The observed widths of daily increments, 
as classified above, were similar to those expected on 
the basis of otolith growth calculations (see 
previous paragraph). 

Diel Light Cycle 

Otoliths offish reared under a diel photoperiod and 
constant temperature ( 1 4L: 1 OD/CT) produced clear 
daily growth increments from the time of hatch. 
Regression of major increment number against elapsed 
time produced a slope not significantly different 
from 1.0 (P> 0.05); a slope of 1.0 would indicate that 
one increment was formed every day. 



Increment width varied with location on the otolith 
and fish age (Fig. 3). Subdaily increments were com- 
mon at all ages, numbering up to 5 between adjacent 
daily increments. They were most abundant in the 
first month after hatch. The distinction between 
daily and subdaily increments was generally clear; 
however, increments produced 5-20 d after hatch 
were the most irregular on the otolith, and were 
sometimes difficult to interpret. Subdaily incre- 
ments tended to be prominent in this region, so that 
distinction was a matter of degree (Fig. 4A). 



5 
I 

H 
Q 

i 



5 

cr 
U 

z 



z 
< 



16 



12 



4- 



24L/14T, 10T 2 



24L/CT 




14L 10D/CT 



i 

10 



i i 

20 30 

AGE (DAYS) 



— r~ 
40 



- 1 
50 



Figure 3. -Daily increment width as a function of age for otolith 
samples of plainfin midshipman from each of the three experimental 
environments. At a given age, mean widths do not differ significant- 
ly among environments, with the exception of values at age 40 d (P 
< 0.05). 



Fish transferred to a constant environment (24 L/ 
CT) as juveniles produced posttransfer increments 
that were very different from those produced prior 
to transfer. Posttransfer increments were visually 
faint and, in some cases, virtually invisible (Fig. 5A). 
Subdaily increments were also present. Transfer to a 
constant environment was not associated with a 
recognizable lag period during which increments 
gradually shifted their appearance. Increments pro- 
duced within 1-2 d of transfer were virtually nonexis- 
tent. Nevertheless, posttransfer increments were 
daily in nature, as indicated by increment counts 
similar to those expected of daily increment produc- 
tion (Table 1). Daily increments gradually became 
more prominent after about 15 d posttransfer, their 
visual contrast improving until the end of the 
experiment. 



168 



CAMP ANA: DAILY GROWTH INCREMENTS IN OTOLITHS OF PLArNFIN MIDSHIPMAN 

TABLE 1 .—Growth increment counts in otoliths of plainfin midshipman, Porichthys notatus, in rela- 
tion to elapsed time for various experimental environments. Fish were transferred to new environ- 
ments (or kept in the original environment as a control) on 10 August, 





10 


Aug. samples 




Environ- 




10 Sept. samples 




Fnviron- 


Days after 


No. major 




Days aft€ 


ir No. major 




ment 1 


hatch 


increments 


SE 


ment 2 


hatch 


increments 


SE 


14L10D/CT 


34 


'34.3 


057 


14L10D/CT 


65 


66.7 


0.80 


14L10D/CT 


— 


— 


— 


24I7CT 


65 


65.1 


1.21 


24L/14T p :10T 2 


34 


41.1 


1.29 


241VCT 


65 


71 2 


0.70 


2417CT 


30 


49.1 


1.33 


24L7CT 


61 


76.9 


1.04 


2417CT 


— 


— 


— 


14L10D/CT 


61 


72 7 


1.10 


24UCT 


— 


— 


— 


24L'14T,:10T 2 


61 


69.3 


0.92 



'This value was derived from 14L: 10D/CT otoliths sampled 10 September; counts were made from the hatch check to the 
prominent subdivision/transfer check 



Diel Temperature Cycle 

Fish hatched under a 24L/14T 1 :10T 2 regime 
deposited growth increments that differed in many 
respects from those produced under a cyclic 
photoperiod (14L:10D/CT). Increments produced 
up to 8 d medial and distal of the peripheral nuclei 
were characterized by a high incidence of prominent 
subdaily increments (Fig. 4B), more so than was the 
case under a cyclic photoperiod. Daily/subdaily 
similarities are reflected in the data of 10 August 
(Table 1), where the observed major increment count 
was significantly different from that expected of daily 
increments (P < 0.05). The high increment count 
indicates that some subdaily increments were promi- 
nent enough to be classified as daily. 

Increments produced in the 15-20 d before transfer 
were generally distinct and regular in appearance. 
Increment width and the incidence of subdaily 
increments were similar to those observed in the cor- 
responding region of the cyclic photoperiod otoliths 
(Fig. 3). However, the appearance of the major 
increments was unusual in that the opaque portion of 
each increment was relatively broad and sharply 
delineated (Fig. 6). 

Fish maintained in the 24L/14T,:10T 2 environ- 
ment after 10 August were overcrowded and did not 
grow well. As a result, posttransfer otolith growth was 
limited, increments were very narrow, and reliable 
counts could not be made. However, increment 
counts of representative otoliths suggested that daily 
increments were deposited after the transfer date. 

Juvenile fish transferred from the fluctuating tem- 
perature regime to a constant environment (24L/CT) 
produced posttransfer increments similar to those of 
fish transferred from 14L:10D/CT to 24L/CT (Fig. 
5B). The difference between August and September 
increment counts corresponds to that expected of 
daily increment deposition (P > 0.05) (Table 1). The 
first five posttransfer increments were faint and vir- 
tually nonexistent; subsequent increments became 



more distinct and regular with time. Opaque regions 
within each increment never became as broad and 
discrete as was observed prior to transfer. 

Constant Environment 

Otoliths of fish hatched under constant conditions 
(24L/CT) initially resembled those of the other two 
environments (with respect to the first 5-8 
increments). The subsequent region resembled that 
of 24L/14T,: 10T 2 fish in that subdaily increments 
were prominent (Fig. 4C). Although the difference 
was not significant (Scheffe's test, P = 0.07), incre- 
ment widths tended to be more irregular than those 
of 1 4L: 1 OD/CT fish of similar age (Fig. 7) . The confu- 
sion of daily and subdaily increments in the early lar- 
val region resulted in a high variance and a mean 
increment count that was significantly higher than 
would be expected of daily increments (P < 0.05) 
(Table 1). After age 10-25 d, daily increments de- 
creased in width (Fig. 3) and became more regular in 
width (Fig. 7) and appearance, although subdaily 
increments were still present. Increments with 
broad, discrete opaque portions were not observed in 
the 24L/CT fish, as they were in the fluctuating tem- 
perature regime. For an unknown reason, otolith 
growth (but not fish growth) under a 24L/CT regime 
significantly exceeded that observed under 
14L:10D/CT(P<0.05). 

Fish remaining in a constant environment after the 
10 August transfer date continued to produce dis- 
tinct increments, although daily and subdaily 
increments were occasionally difficult to differen- 
tiate. Increment width was significantly more 
irregular than in the posttransfer region of 1 4L: 1 0D/ 
CT fish (<-test, P < 0.05) (Fig. 7). Major increments 
in the posttransfer region were daily; the regression 
of increment number against elapsed time resulted in 
a slope not significantly different from unity (P > 
0.05). 

Posttransfer increments of fish hatched and reared 
under constant conditions were prominent, although 



169 



FISHERY BULLETIN: VOL. 82, NO. 1 




1 lilt 




I III I 




I 




111 



I I 




FlGl KK 4. — Growth increments on the polished sagittae of larval plainfin midshipman. Subdaily increments are visible between some of the 
indicated daily increments. Daily increments became more clear with age, but were most prominent/consistent in width in (A). Bar = 'AO fim. 



170 



CAMPANA: DAILY GROWTH INCREMENTS IN OTOLITHS OF PLAINFEM MIDSHIPMAN 




PN = peripheral nucleus. (A) Hatched under a diel light cycle; (B) hatched under a diel temperature cycle; (C) hatched under a constant 
environment. 



irregular in width (Fig. 5C). In contrast, increments of 
fish transferred to the constant environment as 
juveniles were visually faint, becoming more promi- 
nent after 2-3 wk. Juveniles transferred from a cons- 
tant environment to a cyclic regime deposited 
similar-appearing increments before and after 
transfer. However, posttransfer increments tended 
to be more regular in width than in constant environ- 
ment fish; the change generally became apparent 2-4 
d after transfer. Visual contrast of daily increments 
may have increased in the fluctuating temperature 
regime, but the change was not consistent among all 
otoliths. No such change was evident among the post- 
transfer increments of fish shifted from 24L/CT to 
14L:10D/CT, although the incidence of subdaily 
increments appeared to decrease. Fish transferred 
from the constant environment to either of the cyclic 
regimes produced daily increments after transfer; 
high increment counts (Table 1) were derived from 
the irregular, pretransfer region of the otolith. 

DISCUSSION 

Daily growth increments were deposited on the 
otoliths of plainfin midshipman under a variety of 
environmental conditions. My results indicated that 



light, temperature, age, and an endogenous circadian 
rhythm may all influence the production and/or 
appearance of daily and subdaily increments. 
However, some of the variables tested interacted to a 
large degree, and their influence on increment pro- 
duction was subject to alteration through time. 

A cyclic light regime influenced increment produc- 
tion in larval fish more than any other variable tested. 
Under a natural photoperiod, daily increments were 
produced from the time of hatch. In contrast, con- 
stant light conditions disrupted the production of 
posthatch increments, resulting in a high incidence of 
prominent nondaily increments (> 1 increment/24 h) 
and irregular increment widths. My observations are 
consistent with those of Taubert and Coble (1977), 
who observed numerous, nondaily increments in lar- 
val Tilapia hatched under constant light conditions. 
Those authors concluded that light acted as a 
zeitgeber for an endogenous rhythm and that without 
a cyclic photoperiod, daily increment production was 
not possible. My results only partially support their 
conclusion. Photoperiod entrained daily increment 
production in newly hatched midshipman. However, 
in the absence of cyclic light or temperature stimuli, 
an endogenous circadian rhythm of increment 
deposition became apparent after an acclimation 



171 



FISHERY BULLETIN: VOL. 82, NO. 1 



\ 



A 



1 * i | , 



T ' ' '  







B 




Kiel kk 5. — Growth increments in sagittae of plainfin midshipman produced before and after transfer to a constant 
environment. Fish hatched under 24L/CT produced clearer daily increments than those transferred from a different 



172 



CAMPANA: DAILY GROWTH INCREMENTS IN OTOLITHS OF PLAINKIN MIDSHIPMAN 




'* 







environment. Daily increments are indicated, as is direction of sagittal growth (arrow). T = transfer check. Bar= 30ju.ni. 
(A) 14L:10D/CTto24L/CT; (B) 24L/14T,:10T, to 24L/CT; (C) 24L/CT to 24L/CT. 



period of 2-4 wk. Therefore, photoperiod acted as a 
zeitgeber for an endogenous rhythm during the early 
larval stages, but became unnecessary with increas- 
ing age. The nondaily increments produced after 
hatch in this study (and that of Taubert and Coble 
1977), probably comprised both daily and subdaily 
increments. The combination resulted in the deposi- 
tion of more than 1 increment/24 h. 

If a constant photoperiod was present at hatch, an 
endogenous rhythm of increment deposition became 
apparent after an acclimation period. Acclimation 
also occurred when older fish were transferred from a 
natural light cycle to constant light conditions. 
However, the pattern of increment production during 
acclimation differed at the two ages (Table 2). The 
larval fish acclimation period may be analogous to 
that of newborn rats transferred from a diel 
photoperiod to constant conditions. An arhythmic 
activity pattern continues for almost 2 wk in rats 
before an endogenous circadian rhythm becomes 
apparent (Davis 1981). 

The length of the acclimation period could not be 
determined with accuracy. A shift in increment 
appearance after transfer from a constant to a cyclic 
environment generally occurred in 2-5 d. The reverse 
transfer resulted in almost nonexistent increments 



Table 2. — Age effects on growth increment production in otoliths of 
plainfin midshipman, Porichthys notatus, reared under three artifi- 
cial environments. 



Larvae 



Juveniles 



Light important as zeitgeber 
Daily & subdaily increments 
similar during acclimation to 24L 

Long acclimation to 24L 
Immature circadian rhythm 



Light not important as zeitgeber. 

Faint daily increments, but subdaily 
increments dissimilar during acclima- 
tion to 24L 

Short acclimation to 24L 

Mature circadian rhythm 



for a period of 5 d, but the visual contrast of the 
growth patterns improved over the subsequent 10- 
15 d. Therefore, the critical stage of the adaptation 
process appears to have been completed in 2-5 d. 
This result is consistent with that of Tanaka et al. 
( 1981), who observed a 6-d transitory period of incre- 
ment formation when a 24-h light-dark cycle was sud- 
denly reversed. 

Age-related changes in endogenous circadian 
rhythms have not been examined in fishes. Mam- 
malian studies indicate that endogenous rhythms 
often appear after birth; once present, cycle 
amplitude tends to increase with time until the 
rhythm is "mature" (Davis 1981). Porichthys larvae 
hatched under constant light appear to fit this pat- 
tern. Daily and subdaily increments were not easily 



173 



^.HERY BULLETIN: VOL. 82, NO. 1 



X 






Fic.i'RE 6. — Daily growth increments produced on the sagittae of plainfin midshipman after 15-25 d of rearing under a diel temperature cycle. 
The increments were visually prominent and sharply delineated relative to those produced under other environmental regimes. Bar = 20 
/i in. 



O 5i 



O 4- 



cc 




< 




_l 




D 




a 


3 


DC 




I 




t- 




n 






2 


3 




L^ 




o 




« 




LU 


O I 


c 




2 





o 



24L/14T, 10T ; 



24L CT 




14L 10D CT 



10 



20 



i 

30 



40 



50 



AGE (DAYS 



FlGi RE 7. Index of daily increment width regularity as a function of 
age for otolith samples of plainfin midshipman from each of the three 
experimental environments. Bars represent ±1 SE. 



differentiated at first, suggesting that the circadian 
deposition rhythm was not yet mature. Maturation 
apparently occurred by days 10-20. Early larval 
increments were only indistinct temporarily in the 
14L:10D/CT fish, suggesting that the cyclic 
photoperiod entrained the maturing rhythm fairly 
quickly. In addition, very young animals may be more 
responsive to a diel light cycle, due to age-related 
characters of the rhythm cycle (Sacher and Duffy 
1978). For instance, the metabolic rate of newly 
hatched rats is very sensitive to changes in light level, 
while older rats are less affected. In this study, larval 
fish exposed to a constant environment took longer to 
produce daily increments than did juvenile fish, sug- 
gesting an analogy with the rat study. Similar age- 
related results were reported by Gibson et al. (1978) 
in an ontogenetic study of flatfish activity cycles. A 
constant photoperiod eliminated a diel activity cycle 
in larval plaice (Pleuronectes platessa), but had no 
such effect on juveniles of the same species. 

Increasing age of midshipman was correlated with 
decreasing increment width and fewer subdaily 



174 



CAMPANA: DAILY GROWTH INCREMENTS IN OTOLITHS OF PLAINFIN MIDSHIPMAN 



increments in all environments. However, foremost 
among the age-associated effects (Table 2) was the 
prominence of daily increments in juveniles relative 
to larvae. Distinction between daily and subdaily 
increments was seldom difficult in juveniles (outside 
of the acclimation period) unlike the situation in lar- 
val otoliths. If this age-related difference in daily 
increment formation is universal, daily increment 
counts in larvae may be unreliable relative to slightly 
older fish. This suggestion has serious implications 
for the application of growth increments in aging lar- 
val fish. Similarly, the absence of definitive criteria 
for differentiating daily and subdaily increments 
could cause problems in aging field-collected fish. 
Subdaily increments can be numerous and confusing 
in some species (Campana, unpubl. data). 

The demonstration of and age-related rhythm and 
the existence of an acclimation period may have 
resolved some of the conflicting results in the litera- 
ture concerning the zeitgeber effect of light. In a pre- 
vious study, a constant light regime did not influence 
the production of daily increments in juvenile starry 
flounders (Campana and Neilson 1982). The floun- 
ders were about 8 mo-old, suggesting that the 
necessary acclimation period would be short. In addi- 
tion. the fish were exposed to the experimental 
environment for 2 wk prior to tetracycline injection 
(marking the start of the experiment); it is probable 
that acclimation occurred during this period, resulting 
in clear daily increment production by the time the 
experiment began. An analogous explanation may 
explain the results of another study, where chinook 
salmon eggs, reared in darkness, produced daily 
increments after hatch (Neilson and Geen 1982). The 
embryos were held in total darkness for 50 d before 
hatch, suggesting that their endogenous circadian 
rhythm had time to acclimate before hatch. 

A fluctuating temperature regime did not entrain 
increment production under constant light con- 
ditions. Fish reared in this environment produced 
more increments than would be expected of daily 
production, similar to those of 24L/CT fish. The 
variance of larval increment counts was similar to 
that produced under a constant environment, both of 
which were significantly larger than the 1 4L: 1 OD/CT 
variance (Bartlett's test,P< 0.01). Once acclimation 
occurred, daily increments were produced through 
an apparently endogenous periodicity, and not 
through temperature entrainment of an internal 
clock. However, the formation of a broad, optically 
dense, sharply delineated opaque zone in postac- 
climation daily increments indicates that tempera- 
ture fluctuation did affect increment production. The 
opaque portion of a daily increment consists of 



calcium carbonate and a proteinaceous matrix, with 
the latter component predominating (Brothers 
1981;Mugiyaetal. 1981). Falling temperatures, such 
as would occur at night, may have increased the pro- 
portion of protein deposited in the opaque region, 
resulting in an increment that had increased visual 
contrast. Accentuation of contrast renders in- 
crements visually prominent, and could easily be 
interpreted as an entraining mechanism. Diel tem- 
perature fluctuations noticeably accentuated incre- 
ment contrast in young chinook salmon otoliths ( J. D. 
Neilson 2 ). A correlation of increasing protein deposi- 
tion with decreasing temperature suggests that the 
broad opaque zone formed during the low tempera- 
ture, 1 0-h, experimental "night", overlaid the opaque 
zone formed under circadian control through a 3-h 
period (Mugiya et al. 1981). If temperature does 
exert a "masking" effect (Enright 1981), a low 
temperature-induced opaque zone would appear 
independently of any endogenous circadian rhythm 
of deposition. Therefore, multiple daily oscillations 
in temperature could conceivably produce a distinct 
increment after each cycle, in addition to the daily 
increment formed under endogenous control. In 
some situations, the masking effect of temperature 
fluctuations may be substantial, obscuring most of 
the increments formed through the action of an 
endogenous rhythm of deposition (E. B. Brothers 3 ). 
This hypothesis is consistent with studies that 
demonstrated that temperature cycles do not entrain 
daily increment production (Campana and Neilson 
1982; Neilson and Geen 1982), but can influence 
increment formation (Brothers 1981). 

My results suggest that a diel light cycle entrains an 
endogenous circadian rhythm of increment deposi- 
tion. Increasing age mitigated the zeitgeber effect of 
photoperiod, while temperature fluctuation influ- 
enced increment appearance, rather than perio- 
dicity. In other studies, the incidence of subdaily 
increments was correlated with feeding periodicity 
(Neilson and Geen 1982; Campana 1983). The fact 
that so many variables may affect increment deposi- 
tion suggests that the environment does not 
influence the rhythm of otolith deposition directly, 
but acts through some penultimate process. 
Metabolic rate is susceptible to environmental 
influence, as well as being subject to an endogenous 
circadian rhythm (Matty 1978) that changes with age 
(Davis 1981). However, metabolic rate is in turn 



-J. D. Neilson, Marine Fish Division. Biological Station, St. 
Andrews, New Brunswick, Canada EOG 2X0, pers. comraun. Jan- 
uary 1983. 

'K. H. Brothers, Division of Biological Sciences, Cornell Univer- 
sity, Ithaca, XV 1 1850, pers. comraun. May 198 3 



175 



FISHERY BULLETIN: VOL. 82. NO. 1 



regulated by endocrine levels, and it may be the 
environmental modulation ofendocrine rhythms that 
ultimately controls increment periodicity on the 
otolith (Menaker and Binkley 1981). Endocrine se- 
cretion often follows a circadian pattern (Simpson 
1978) and, in mammals at least, is closely linked to 
the circadian pacemaker itself (Menaker and Binkley 
1981). Hormones regulate many aspects of meta- 
bolism and growth, including skeletal calcification 
(Simpson 1978). Therefore, it seems reasonable to 
postulate that those factors that entrain and/or mod- 
erate the circadian rhythm of endocrine secretion will 
have a subsequent effect on increment deposition in 
the otolith. 

ACKNOWLEDGMENTS 

Jim McNutt of Ayerst Laboratories kindly donated 
the antibiotics used in this study. I appreciate the 
assistance and technical innovation of Doug Begle in 
the field and laboratory. John D. Neilson and Nor- 
man J. Wilimovsky provided many helpful comments 
on an earlier draft of the manuscript. This study was 
supported by a grant from the Natural Sciences and 
Engineering Research Council of Canada to Norman 
J. Wilimovsky. 

LITERATURE CITED 

Arora, H. L 

1948. Observations on the habits and early life history of the 
batrachoid fish Porichthys notatus Girard. Copeia 
1948:89-93. 
Brothers, E. B. 

1981. What can otolith microstructure tell us about daily and 
subdaily events in the early life history of fish? Rapp. P.- 
V. Reun. Cons. Int. Explor. Mer 178:393-394. 
Brothers, E. B., C. P. Mathews, and R. Lasker. 

1976. Daily growth increments in otoliths from larval and 
adult fishes. Fish. Bull., U.S. 74:1-8. 
Brothers, E. B., and W. N. McFarland. 

1981. Correlations between otolith microstructure, growth, 
and life history transitions in newly recruited French grunts 
\Haemulon flavoUneatum (Desmarest), Haemulidae]. Rapp. 
P-V Reun. Cons. Int. Explor. Mer 178:369-374. 

Campana, S. E. 

1983. Feeding periodicity and the production of daily growth 
increments in otoliths of steelhead trout (Salmo gairdneri) 
and starry flounder (Platichthys stellatus). Can. J. Zool. 
61:1591-1597. 

Campana, S. E., and J. D. Neilson. 

1982. Daily growth increments in otoliths of starry flounder 
{Platichthys stellatus) and the influence of some environ- 
mental variables in their production. Can. J. Fish. Aquat. 
Sci. 39:937-942. 

Davis, F. C. 

1981. Ontogeny of circadian rhythms. In J. Aschoff (editor). 
Handbook of behavioral neurobiology, Vol. 4, p. 257- 
274. Plenum Press, N.Y. 



ENRIGHT, J. T. 

1981. Methodology. In J. Aschoff (editorl. Handbook of be- 
havioral neurobiology. Vol. 4, p. 11-19. Plenum Press, N.Y. 

Gibson, R. N., J. H. S. Blaxter, and S. J. de Groot. 

1978. Developmental changes in the activity rhythms of the 
plaice (Pleuronectes platessa L.). In J. E. Thorpe (editor). 
Rhythmic activity of fishes, p. 169-186. Acad. Press, 
N.Y. 

Marshall, S. L., and S. S. Parker. 

1982. Pattern identification in the microstructure of sockeye 
salmon {Oncorhynchus nerka) otoliths. Can. J. Fish. 
Aquat. Sci. 39:542-547. 

Matty, A. J. 

1978. Pineal and some pituitary hormone rhythms in fish. In 
J. E. Thorpe (editor), Rhythmic activity of fishes, p. 21- 
30. Acad. Press, N.Y. 
Menaker, M., and S. Binkley. 

1981. Neural and endocrine control of circadian rhythms in 
the vertebrates. In J. Aschoff (editor). Handbook of 
behavioral neurobiology, Vol. 4, p. 243-256. Plenum 
Press, N.Y. 
Methot, R. D„ Jr. 

1981. Spatial covariation of daily growth rates of larval 

northern anchovy, Engraulis mordax, and northern 

lampfish, Stenobrachius leucopsarus. Rapp. P.-V. Reun. 

Cons. Int. Explor. Mer 178:424-431. 

Mugiya, Y., N. Watabe, J. Yamada, J. M. Dean, D. G. 

Dunkelberger, and M. Shimuzu. 

1981. Diurnal rhythm in otolith formation in the gold fish, 
Carassius auratus. Comp. Biochem. Physiol. 68A:659- 
662. 

Neilson, J. D., and G. H. Geen. 

1982. Otoliths of chinook salmon (Oncorhynchus tsha- 
wytscha): daily growth increments and factors influencing 
their production. Can. J. Fish. Aquat. Sci. 39:1340-1347. 

PANNELLA, G. 

1971. Fish otoliths: daily growth layers and periodical pat- 
terns. Science (Wash., D.C.) 173:1124-1127. 

1980. Growth patterns in fish sagittae. In D. C. Rhoads and 
R. A. Lutz (editors). Skeletal growth of aquatic organisms: 
Biological records of environmental change, p. 519- 
560. Plenum Press, N.Y. 

Ralston, S. 

1976. Age determination of a tropical reef butterflyfish utiliz- 
ing daily growth rings of otoliths. Fish. Bull., U.S. 74:990- 
994. 
Sacher, G. A., and P. H. Duffy. 

1978. Age changes in rhythms of energy metabolism, activity 
and body temperature in Mus and Peromyscus. In H. V. 
Samis, Jr., and S. Capobianco (editors). Aging and biologi- 
cal rhythms, p. 105-124. Plenum Press, N.Y. 
Simpson, T. H. 

1978. An interpretation of some endocrine rhythms in 
fish. //; J. E. Thorpe (editor). Rhythmic activity of fishes, 
p. 55-68. Acad. Press, N.Y. 
Struhsaker, P., and J. H. Uchiyama. 

1976. Age and growth of the nehu, Stolephorus purpureus 
(Pisces: Engraulidae), from the Hawaiian islands as 
indicated by daily growth increments of sagittae. Fish. 
Bull., U.S. 74:9-17. 

Tanaka, K., Y. Mugiya, and J. Yamada. 

198 1. Effects of photoperiod and feeding on daily growth pat- 
terns in otoliths of juvenile Tilapia nilotica. Fish. Bull., 
U.S. 79:459-466. 

Taubert, B. D., and D. W. Coble. 

1977. Daily rings in otoliths of three species of Lepomis and 



176 



CAMPANA: DAILY GROWTH INCREMENTS IN OTOLITHS OF PLAINFIN MIDSHIPMAN 

Tilapia mossambica. J. Fish. Res. Board Can. 34:332- 
340. 
Wilson, K. H., and P. A. Larkin. 

1980. Daily growth increments in the otoliths of juvenile sock- 
eye salmon [Oncorhynchus nerka). Can. J. Fish. Aquat. 
Sci. 37:1495-1498. 



177 



ASPECTS OF THE LIFE HISTORY AND FISHERY OF 
THE WHITE CROAKER, GENYONEMUS LINEATUS (SCI AENIDAE), 

OFF CALIFORNIA 

Milton S. Love, 1 Gerald E. McGowen, 2 William Westphal ' 
Robert J. Lavenberg, 2 and Linda Martin' 

ABSTRACT 

White croaker. Genyonemus lineatus (Ayres), was a dominant species off southern California in nearshore, 
sandy substratum waters, and comprised 29.7% of all fish taxa taken in otter trawl hauls. Juveniles occurred 
in waters < 27 m and the mean length of all individuals increased with depth. The maximum depth of capture 
was 183 m. 

White croaker live to 1 2 years, exhibiting rapid growth which is essentially constant throughout the species' 
life. Females grew at a slightly faster rate than males. Von Bertalanffy age-length parameters for females 
wereL„ = 60.7,fc = 0.04,( =-7.6,andformalesL oc = 59.2,& = 0.03,r o =-8.7.Afterlyear,morethan50% 
of the individuals are mature, but others delay maturity for 4 years. Larger females had longer spawning 
seasons than did smaller individuals. Although spawning occurred throughout the year, principal spawning 
occurred between November and April, with a February-March peak. White croaker are batch spawners; 
females spawned 18-24 times a season. Batch fecundities ranged from 800 to 37,200 eggs. White croaker 
reproduction off Monterey differed significantly from that off southern California. Large-scale spawning 
occurred from at least July through February, and continued throughout the year. Colder water off Monterey 
may have allowed for extended spawning activity. 

White croaker larvae were a significant constituent of the southern California ichthyoplankton fauna, 
second in abundance to northern anchovy, Engraulis mordax, in waters <36 m deep. Data from ichthyoplankton 
surveys indicated two spawning centers, one located from Redondo Beach to Laguna Beach and a smaller 
one centered about Ventura. Highest larval densities were found near the substratum in 15-22 m of water. 
White croaker is an important part of the skiff sportfishery and the basis of a growing commercial gill net 
fishery. Size frequencies of white croaker taken in both fisheries indicated that few juveniles were 
captured. 



Fishes of the family Sciaenidae (drums) are a major 
constituent of the fauna of the eastern temperate 
Pacific coast off California (Skogsberg 1939; Frey 
1971). Eight species have been recorded off Califor- 
nia, primarily in inshore waters. With the exception of 
the shortfin corvma,Cynoscionparvipinnis, and black 
croaker, Cheilotrema saturnum, all six of the other 
species known from off California (white seabass, 
Atractoscion nobilis; white croaker, Genyonemus 
lineatus; California corbina, Mentieirrhus undulatus; 
spotfin croaker, Roncador stearnsii; queenfish, 
Seriphus politus; yellowfin croaker, Umbrina ron- 
cador) are of sport or commercial importance. 

The white croaker is an abundant species that 
associates with soft (primarily sand) substrata in the 
coastal zone. White croaker are small (reaching 



'Vantuna Research Group, Department of Biology, Occidental 
College, Los Angeles, CA 90041. 

'Natural History Museum of Los Angeles County, 900 Exposition 
Blvd., Los Angeles, CA 90007. 

'Moss Landing Marine Laboratory, P.O. Box 223, Moss Landing, 
CA 95039. 



lengths of 41.4 cm total length, Miller and Lea 1972) 
and active fishes that range from the surf zone to 
depths of 183 m between Vancouver Island, British 
Columbia, Canada, south to Magdalena Bay, Baja 
California, Mexico. Within this geographic range, 
they are most abundant between San Francisco Bay 
and northern Baja California. White croaker are 
omnivores, feeding on a variety of benthic and 
epibenthic forms (crustaceans, clams, polychaetes, 
and small fishes, particularly the northern anchovy, 
Engraulis mordax (Phillips et al. 1972; Morejohn et 
al. 1978; Ware 1979)). 

White croaker are the mainstay of pier and small 
boat sportfish catches in both southern (Pinkas et al. 
1968; Wine and Hoban 1976) and central California 
(Miller and Gotshall 1965). In addition, commercial 
catches have increased in recent years to 200,000 kg/ 
yr. 4 Despite this, G. lineatus is a much maligned 
species, as it is small and adept at bait-stealing. More- 



Manuscript accepted June 1983. 

FISHERY BULLETIN: VOL. 82, NO. 1, 1984. 



4 M. Oliphant, California Department of Fish and Game, Long 
Beach, CA 90802, pers commun. July 1981. 

179 



FISHERY BULLETIN: VOL. 82, NO. 1 



over, there is a firmly held belief that white croaker 
are unusually wormy. In fact, the frequency of 
occurrence of nematodes (larval Anisakis and 
Phocanema) in white croaker muscle is lower than 
that for at least some other important sport and com- 
mercial species such as California halibut, 
Paralichthys californicus, and chilipepper rockfish, 
Sebastes goodei (Dailey et al. 1981). 

Because white croaker are abundant around sewage 
outfalls and tolerant of degraded environments, 
much of the recent research on this species has been 
pollution-centered. Several published works deal 
with pesticide levels (Castle and Woods 1972; 
MacGregor 1972; Stout and Beezhold 1981) and 
pollution-implicated diseases and abnormalities 
(Russell and Kotin 1957; Mearns 1974, 1979; 
Mearns and Sherwood 1977; Sherwood 1978). Five 
small-scale studies have been conducted on its life 
history (Issacson 1964, 1967; Goldberg 1976; More- 
john et al. 1978; Ware 1979) 

This contribution represents a summation of 
unpublished white croaker data obtained from three 
sources: a life history and fishery study by Love, 
ichthyoplankton work by McGowen and Lavenberg, 
and a trawling survey by Westphal. 

METHODS 

Collection of Juveniles and Adults 

Samples were collected monthly (3-6 per month) 
from October 1978 to February 1981 with a 7.6 m or 
4.9 m headrope otter trawl in 15-65 m of water be- 
tween Palos Verdes and Huntington Beach, Calif. 
Reduced numbers of white croaker also were collect- 
ed monthly from April 1979 to September 1981 in 
Monterey Bay with a 4.9 m otter trawl in 10-60 m of 
water or were purchased from local fishermen. All of 
these specimens were frozen for later dissection. 
After thawing, all fish were measured (total length, 
fork length, standard length), weighed, sexed, and 
the gonads were weighed. 

Collection of Depth Preference Data 
for Adults and Juveniles 

Information on white croaker depth preference was 
based on data from a trawling program aboard the 
RV Vantuna . Trawling was conducted at a speed of 2- 
3 kn for 20 min with a 7.6 m (occasionally 4.9 m) otter 
trawl having a net of 0.6 cm stretch mesh. From Sep- 
tember 1972 through December 1980, 18 stations 
(Fig. 1) were sporadically sampled at 10 depths, 
although most of the trawling effort was performed at 



depths between 59 and 91 m. After shipboard sort- 
ing, fishes were measured (board standard length) 
and discarded. All lengths were converted to total 
length (TL) using conversion factors based on 
measurements of 100 white croaker (Table 1). 

TABLE 1. — Conversion factors between standard (SL), fork (FL), 
and total (TL) lengths (cm), based on measurements of 100 white 
croaker from southern California. 



SL = 0.442 + 79 TL 
= 0.379+ 82 FL 



FL = 



0.088 + 96TL 
0.849+ 1 14 SL 



TL=0892 + 1.19 SL 
= 0.023+ 1.04 FL 



Techniques for Aging Juveniles and 

Adults 

Sagitta were removed from each side of the head, 
and the otoliths were cleaned, air dried, and stored in 
vials. Because whole croaker otoliths are difficult to 
age, they were sectioned on a Buehler Isomer 5 low 
speed saw, Otoliths were placed on wood blocks and 
completely embedded in clear epoxy (Ciba 825 hard- 
ener and Ciba 6010 resin). Each block with its otolith 
was emplaced on the saw and a dorsal-ventral 0.05 
cm wafer was cut through the otolith, using two 
diamond-edge blades separated by a stainless steel 
shim. Wafers were stored in water for a few days to 
soften the epoxy (which was removed), then the 
wafers were placed in a black-bottomed water glass 
filled with water and read under a dissecting micro- 
scope at a magnification of 10X. All otoliths were 
read twice, about 4 mo apart, by Love. When 
readings did not agree, the otoliths were read again. 
The value of two coincident readings was accepted as 
the best estimate of age. Fifteen percent of all 
otoliths were unreadable due to a lack of recogniz- 
able annuli. 

Procedures for Determining 

the Timing of Maturation and 

Reproduction 

We estimated length at first maturity by classifying 
gonads as immature or mature based on the tech- 
niques of Bagenal and Braum (1971). Smallermature 
fish and fish just entering their first mature season 
become reproductive later in the spawning season. 
Hence we estimated length at first maturity during 
the peak spawning period of January, February, and 
March. To ascertain spawning season duration and 
its relation to body size, we sampled at least 150 
females/mo in 1 cm size intervals throughout the 



'Reference to trade names does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 



180 



LOVE ET AL.: LIFE HISTORY AND FISHERY OF WHITE CROAKER 



Fliti'RE 1.— Location of white croaker 
sampling sites. 




^C^>> 



100km 



year. A gonadosomatic index (gonad weight)/(total 
body weight) X 100 was computed from frozen 
specimens to quantify changes in gonad size with 
season. Ovaries for use in fecundity studies were 
fixed in modified Gilson's fluid (Bagenal and Braum 
1971) for 4-8 mo. We measured fixed egg diameters 
from 11 individuals, all of which contained some hy- 
drated eggs. Batch fecundity was estimated by the 
gravimetric method of Bagenal and Braum (1971). 
The time between spawning events per female was 
computed by estimating the percent of females with 
hydrated eggs on any given night during the spawning 
season. 
We computed condition factor 1UU (W u W) , 

L 3 
where W = body weight in grams, GW — gonad 
weight in grams, and L = total length in centime- 
ters — of mature southern California and Monterey 
croaker. Condition factor was computed using body 



weight with gonad weight subtracted to mini- 
mize the effects of seasonal changes in gonad size. 

Larval Sampling 

Ichthyoplankton data presented here were collect- 
ed monthly between August 1979 and July 1980 
along 20 sites within the Southern California Bight 
aboard the RV Seawatch (Table 2). Stations were 
established at 8 and 22 m along each transect (with 
exception of Palos Verdes and Laguna Beach where 
15 m was substituted for 8 m). Additional stations at 
15 and 36 m depths were maintained at three sites 
(Ormond Beach, Redondo Beach, San Onofre). Oblique 
bongo tows from the bottom to the surface were made 
at all stations. A 70 cm diameter bongo net sampler 
(McGowan and Brown 1966), equipped with wheels 
to prevent damage when the sampler encountered 
the bottom, was lowered to the bottom with canvas 



181 



FISHERY BULLETIN: VOL. 82, NO. 1 



TABLE 2.— Southern California ichthyoplankton collection 
sites, August 1979-July 1980. Location abbreviations used 
in Figures 13-15 are in parentheses. 



Collection sites 


Lat. N 


Long. W 


Coho Bay (80) 


34 26' 


120 c 26' 


Refugio to El Capitan. 8 m (DR) 


34°27' 


120' 02'- 
120°05' 


North of Refugio. 22 m 


34°27' 


1 20 r 06' 


Santa Barbara to Goleta Pt. (8. 1 5) 


34°25' 


119 44'- 
119°51' 


Pt. Gorda to Rincon Pt |RN) 


3422'- 
34 23' 


119°28' 


Ventura (83) 


34° 1 6' 


119°17' 


Ormond Beach (OB) 


34°07' 


1 1 9 1 0' 


Arroyo Sequit (85) 


34°03' 


118°57' 


Mahbu Beach (MU| 


34°02' 


118°41' 


Playa del Rey (87) 


33°57' 


118°27' 


Redondo Beach (RB) 






Redondo Breakwater, 8, 1 5, and 22 m 


33°51' 


118°24' 


Hermosa Pier. 36 m 


33°52' 


118°25' 


Palos Verdes (PV) 


33°43' 


118°25' 


Huntington Harbor (88) 


33°4T 


11 8° 04' 


Balboa (BA) 


33° 36' 


1 1 7°54' 


Aliso Creek (Laguna Beach) (90) 


33°31' 


1 1 7°46' 


San Onofre (SO) 


33°21' 


1 1 7°33' 


Santa Margarita River (91 ) 


33° 15' 


1 1 7°28' 


Agua Hedionda (Carlsbad) (CD) 


33°08' 


117 23' 


San Dieguito River (Del Mar) (93) 


32° 58' 


117 16' 


Mission Beach (MB) 


32°48' 


117"16' 


San Diego (95) 


32°38' 


1 1 7°09' 



doors over the mouth openings. The canvas doors 
were removed by a cable messenger, allowing the 
nets to fish. Immediately thereafter the sampler was 
retrieved at a constant rate of about 10 m/min (0.17 
m/s); a wire angle of 51 ± 5° was maintained. The 
ship's speed (0.95 ± 0.03 m/s) plus the retrieval rate 
brought the net speed to about 1.12 m/s. 

In addition, stratified (surface, midwater, bottom) 
tows were made at each of the four stations on tran- 
sects at Ormond Beach, Redondo Beach, and San 
Onofre. Horizontal midwater tows were made with 
the previously described bongo sampler towed at a 
rate of 1.06 ± 0.06 m/s. For these tows the sampler 
was lowered to a depth about half-way between the 
surface and the bottom, opened via cable messenger, 
fished, closed via cable messenger, and retrieved. 
Surface samples were taken with a manta sampler 
(Brown 1979) towed at a rate of 1 .07 ± 0.06 m/s. This 
net had a rectangular opening (88 X 16 cm). Bottom 
collections were taken using an auriga net 6 with a 200 
X 50 cm mouth. The auriga net fished a zone 2 mwide 
by 0.5 m deep, about 0.25 m above the substratum, 
and was fished at a rate of 1.07 ± 0.46 m/s. All nets 
were equipped with 335 ju mesh. A General Oceanics 
flowmeter was mounted in the mouth of each net. The 
field program is described in greater detail by Laven- 
berg and McGowen. 7 



Additional data from a 4-yr study off Redondo 
Beach were derived from monthly surface tows made 
from January 1974 to February 1977, using meter 
nets with 335 /i mesh. A TSK flowmeter was mount- 
ed in the mouth of each net. This field program is de- 
scribed in greater detail by McGowen. 8 

Fishery 

Although white croaker are usually the most impor- 
tant species in the private vessel sportfishery, no 
size-frequency data were available. For this reason, 
4,941 croaker taken by anglers aboard skiffs and 
other small private vessels were measured during the 
period June 1980 to July 1981, between Oxnard and 
Dana Point. From September 1980 through August 
1981, 1,748 white croaker were taken off southern 
California by commercial gill net vessels and were 
measured. 

RESULTS 
Depth Preference 

Our trawling study indicated that white croaker pre- 
ferred nearshore habitats and their abundance 
declined in deeper waters. Ranking first of all species 
taken, white croaker was the dominant species at the 
shallowest (18-27 m) stations (Table 3), and com- 
posed 29.7% by number of the total catch and 
appeared in 68% of the trawls. At the 59-73 m 
stations, white croaker catches had declined to 3.3% 
of total catch, frequency of occurrence 20.7%, and at 
the 91-109 m station, the species made up 1.2% of 
total catch, frequency of occurrence 14.0%. At 
stations between 165 and 183 m, white croaker com- 
prised 0.6% of the total catch, with a frequency of 
occurrence of 1.7%. On the basis that no individuals 
were captured at greater depths, we accept 183 m as 
their maximum depth. 

Though white croaker was supplanted as the domi- 
nant species at deeper stations, it remained an 
important community component to depths of 109 
m. Two other species, the California tonguefish, 
Symphurus atricauda, and the Pacific sanddab, 
Citharichthys sordidus, were among the 10 most 
abundant species throughout these depths. Pacific 



''Mitchell, C. T. Auriga: A wheeled epibenthic plankton sampler 
for rocky bottoms. Unpubl. rep., 12 p. Marine Biological Con- 
sultants Inc., 947 Newhall Street, Costa Mesa, CA 92627. 
'Lavenberg, R. •!.. and (1. E. McGowen. Coastal ichthyoplankton 



of the Southern California Right: temporal and spatial distribution 
(Augusl 1979-July 1980). Manuscr. in prep. Los Angeles County 
Museum of Natural History, 900 Exposition Blvd., Los Angeles, 

CA 9 7. 

8 McGowen, G. E. 1978. Effects of thermal effluent from 
Southern California Edison's Redondo Beach steam generating 
plant on the warm temperate fish fauna of King Harbor 
Marina. SCE Research and Development Series: 78-RD-47, 65 
P. 



182 



LOVE ET AL.: LIFE HISTORY AND FISHERY OF WHITE CROAKER 



Table 3. — The 1 most abundant fish species taken by otter trawls 
in three depth intervals off Southern California, 1972-80. 





Total no 


% total 


% frequency 




taken 


no. 


occurence 


Depth interval. 18-27 m 








Number of collections. 109 








Total no. of fish, 14.313 








Total Species. 80 








Genyonemus /meatus 


4.252 


297 


679 


Cithanchthys stigmaeus 


2.221 


15 5 


63.3 


Symphurus atncauda 


2.031 


14.2 


60.6 


Senphus politus 


1.341 


94 


44.0 


Phanerodon furcatus 


595 


4.2 


59.6 


Engraulis mordax 


591 


4.1 


22.0 


Pleuromchthys verticalis 


476 


33 


62.4 


Hyperprosopon argenteurn 


395 


2.8 


33.0 


Cithanchthys sordidus 


301 


2.1 


12.8 


Synodus lucioceps 


206 


1.4 


38.5 


Depth interval. 59-73 m 








Number of collections. 82 








Total no. of fish, 1 3,337 








Total species. 62 








Cithanchthys sordidus 


3.196 


24.0 


72.0 


Microstomus pacificus 


2,769 


20.8 


65.9 


Sebastes dalh 


1.565 


11.3 


65.9 


Sebastes saxicola 


867 


6.5 


29.3 


Ponchthys notatus 


786 


5.9 


59.8 


Sebastes jordani 


694 


52 


17.1 


Symphurus atncauda 


512 


3.8 


51.2 


Scorpaena guttata 


506 


38 


63.4 


Genyonemus /meatus 


436 


3.3 


207 


Icelmus quadnseriatus 


297 


2.2 


25.6 


Depth interval. 91-109 m 








Number of collections, 1 72 








Total no. of fish, 35,488 








Total species, 77 








Microstomus pacificus 


12.386 


34.9 


762 


Cithanchthys sordidus 


9.655 


27.2 


73.8 


Sebastes saxicola 


4,262 


12.0 


65.1 


Ponchthys notatus 


1,688 


4.8 


63.4 


Glyptocephalus zachirus 


1,249 


3.5 


30.2 


Scorpaena guttata 


875 


2.5 


44.2 


Sebastes jordani 


802 


2.3 


21.5 


Genyonemus tmeatus 


441 


1.2 


14.0 


Symphurus atncauda 


377 


1.1 


24.4 


Zaniotepis frenata 


299 


0.8 


250 



where L, 
k 



sanddab dominated in waters between 59 and 109 m, 
declining in numbers both inshore and offshore. 
California tonguefish exhibited an abundance pat- 
tern like white croaker, with numbers peaking in 
inshore waters and declining with greater depth. 

Most juvenile white croaker (50% mature by 15 cm) 
were limited to the inshore (18-27 m) stations (Fig. 2). 
Larger individuals inhabited greater depths. In fact, 
the mean size of white croaker was successively 
larger as depth increased (ANOVA, F = 284.2, P < 
0.001). 

Age and Growth 

Lengths at ages were estimated by direct observa- 
tion of otolith annuli and through the von Bertalanffy 
growth curve model 

L t = L x [1 - exp -k {t-t )\ 



= length at time t 
= theoretical maximum length 
= constant expressing the rate of ap- 
proach to L x 
= theoretical age at which L, = 

to the direct observation age-length 



was fitted 
data. 

We transformed male and female growth equations 
to linear form (Allen 1976) and compared these by 
analysis of variance. Females were found to grow 
significantly faster than males (F = 16.8, P < 0.05), 
hence we separated growth data by sex (Table 4). 



TABLE 4.— Parameters of the von Bertalanffy equation for 
white croaker off southern Calfornia. 



Sex 


L m 


SE 


* 


SE 


to 


SE 


Female 
Male 


60.72 
59.17 


0.23 

0.29 


0037 
0033 


0.02 
0.03 


-7.54 
-866 


1.1 

1.3 



The oldest male and female white croaker we 
examined were 12 yr old (Fig. 3). Females grew 
slightly faster than males and reached a greater size. 
Females from age 1 (at which over 50% of the fish 
were mature) outgrew males. 

White croaker grew at a fairly constant rate 
throughout their lives, exemplified in their very low/? 
values. No asymptote was reached within the observed 
12-yr life span. Thus, the maximum predicted 
lengths were longer than both published (41.4 cm 
TL, Miller and Lea 1972) and unpublished (44.2 cm 9 ) 
records, although the r values for the von Bertalanffy 
equations were high (0.84 for both sexes). 

Length - Weight Relationships 

A total of 58 1 males and 665 females from southern 
California and a total of 94 males and 161 females 
from Monterey Bay were weighed and measured. 
The relationships between total length and weight fit 
the relationship W = aL h , where W = weight in 
grams, L = total length in centimeters, and a and b 
are constants, with values determined using log 10 
transformation and fitting the values to a straight line 
by least squares (Figs. 4, 5). In southern California, 
males tended to be heavier at a given length than 
females (analysis of variance, F — 10.18, P < 0.01), 
whereas off Monterey no significant difference was 
found (analysis of variance, F = 0.67, P > 0.4). To 
test whether this difference was an artifact caused by 
seasonal and gender-related factors, we subtracted 



'R. N. Lea, California Department of Fish and Game, 2201 Garden 
Road, Monterey, CA 93940, pers. commun. May 1982. 



183 



50 - 



50 - 



UJ 

< 

I- 

I 
£2 

Li. 

LL 
O 

en 

CO 

D 
z 



50 



300 



200 



100 



SAMPLES=1 
N=60 
X=26.1 

DEPTH 165-183 m 



SAMPLES=19 
N=308 
X=23.9 

DEPTH 91-109 m 



SAMPLES=13 
N=286 
X=17.3 

DEPTH 59-73 m 



SAMPLES=69 

N= 3,764 
X=16.2 

DEPTH 18-27 m 



6-6.9 8-8.9 



FISHERY BULLETIN: VOL. 82, NO. 1 




10- 


12- 


14- 


16- 


18- 


20 


22- 


24 


26 


28- 


30- 


10.9 


12.9 


14.9 


16.9 


18.9 


20.9 


22.9 


24.9 


26.9 


28.9 


3 30.0 



32- 



32.9 



TOTAL LENGTH INTERVALS (cm) 



FIGURE 2.— Length intervals of white croaker taken by otter trawl off southern California. 



gonad weight from body weight, generated the 
length-weight relationships for each sex and tested 
these between sexes. Again, differences between 
sexes existed in southern California (ANOVA, F = 
1 1.13, P < 0.01), but not in Monterey Bay (ANOVA, 
F= 1.33, P> 0.05). 

Condition Factor 

Both male and female southern California white 
croaker displayed differences in condition between 
peak spawning and resting seasons (Table 5). In both 
sexes, fish were more robust during the resting 
season, perhaps because energy normally utilized for 
somatic maintenance and growth was shifted to egg 
and sperm production and spawning behavior. Over 
all seasons, whereas southern California females 
were more robust than males (Table 5), no such sex- 



TaBLE 5. — Condition factor (K) of white croaker from southern 
California 1978-81 and Monterey Bay. Calif., 1979-81. 





N 


K 


SD 


F 


P 


Southern California 












Males 












Jan -Mar 


264 


0.34 


0.53 


1 17.4 


<0.001 


May-Aug. 


91 


0.98 


0.32 






Females 












Jan-Mar 


280 


0.46 


0.56 


24,4 


<0.001 


May-Aug 


76 


0.80 


0.49 






Sexes Combined 












Jan-Mar 


544 


0.40 


0.55 


118.3 


<0.001 


May-Aug. 


167 


0.90 


41 






All Seasons 












Males 


535 


0.71 


056 


4 5 


<0.05 


Females 


617 


078 


0.54 






Monterey — All seasons 










Males 


80 


1 03 


0.09 


1 29 


>0.2 


Females 


142 


1 02 


0.10 






Southern California 


and Monterey 










Males 












Monterey 


80 


1 03 


0.09 


26.54 


<0.001 


S Calif 


535 


0.71 


0.56 






Females 












Monterey 


142 


1.02 


0.10 


27.83 


<0.001 


S. Calif. 


617 


0.78 


0.54 







184 



LOVE ET AL.: LIFE HISTORY AND FISHERY OF WHITE CROAKER 



30 



25 



5 
U 



a 



H 

O 



20 



15 - 



10 



FEMALES 



MALES 




10 



11 



12 



AGE (YEARS) 



FIGURE 3. — Von Bertalanffy growth curves of female and male white croaker. Also 
included are mean lengths at ages (females-circles, males-triangles) computed from 
direct observation of otolith annuli. Based on 332 females and 250 males taken off 
southern California, 1977-81. 



ual dimorphism was observed off Monterey. Both 
males and females off Monterey were more robust 
than their southern California counterparts (Table 
5). 

Maturation and Reproduction 

Although a few white croaker matured before 1 yr 
(12.9-13.4 cm TL), over 50% of the males were 
mature by 14 cm TL and over 50% of the females by 
about 15 cm TL, which equals an age of 1 yr (Fig. 6). 
All fish were mature by 19 cm TL (3-4 yr). 

Larger females (greater than about 1 7 cm TL and 1 - 
2+ yr) spawned earlier in the year and continued to 
spawn later than smaller and younger individuals 



(Table 6). The smallest spawning females may spawn 
for 3-4 mo whereas larger individuals may spawn for 
as long as 7 mo. 

Off Long Beach, white croaker spawned primarily 
from November through April, with January through 
March the peak months, based on the occurrence of 
hydrated eggs within ovaries. A few individuals (> 18 
cm TL) spawned from May through October. Ovaries 
increased in size in the fall and peaked in January, 
when they averaged 4.67c of body weight (maximum 
11.8%, minimum 0.8%). Thereafter, ovarian size 
declined in summer to a minimum of about 1.0% of 
body weight (maximum 1.3%, minimum 0.07%) and 
remained constant through August (Fig. 7). 

Similarly, testes were small during summer months 



185 



FISHERY BULLETIN: VOL. 82, NO. 1 



350 H 



300 



250 



200 



o 

I- 
I 
(J 

UJ 



150 



100 



FEMALES 

W = .0109 L 30239 

R = .9836 







• • 2 A 






•2 4/ 3 






• 335»**2 22 






•2234/222. 






3.33*534 • 






... 5>24*«3 






 •••3J794.82.2. 






• 4>5276«'2« 






2' 2059424 
•3 5WS25525 




2. 


4 6<2962«2« 




345; 


>Jf>272 2 




• 29 70< 


) 352. 


2 


.5^5533* 



50 



22.&ZB82.2 
• 24 5/TC • • 
 4*^2»« 
222 



_L 



_L 



_L 



11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 

TOTAL LENGTH (cm) 

FIGURE 4. — Length-weight relationship based on 665 female white croaker sampled off southern California, 1978-81. 



TABLE 6. —The percent per month of female white croaker from southern California (1978-81) that were 

sexually mature. 



Mean total 
length (cm) 










Percen 


sexually 


mature 












Sept. 


Oct. 


Nov. 


Dec 


Jan. 


Feb. 


Mar. 


Apr. 


May 


June 


July 


Aug 


13.0 











2 


16 


15 


6 

















140 











1 1 


26 


26 


8 

















15 











21 


73 


72 


15 

















16 











18 


88 


88 


27 

















17.0 








2 


20 


91 


90 


35 


tr 1 


1 











18.0 








6 


21 


96 


94 


61 


tr 


tr 











19.0 








7 


21 


100 


1 ( 10 


83 


48 


tr 











20.0 





tr 


7 


23 


100 


100 


82 


52 


tr 





tr 


tr 


21 








5 


31 


100 


100 


94 


51 


2 


tr 








22.0 


tr 





6 


32 


99 


99 


93 


58 


1 











23.0 








7 


48 


100 


100 


95 


60 


tr 








tr 


24.0 


tr 





6 


47 


100 


100 


93 


58 


2 











25.0 








6 


47 


100 


100 


99 


57 


2 











26.0 





tr 


6 


46 


100 


100 


98 


59 


1 





tr 






'Trace <1%. 



186 



LOVE ET AL.: LIFE HISTORY AND FISHERY OF WHITE CROAKER 



300 - 



250 



200 



O 

\- 
I 

o 

LLI 



150 



100 



50 



MALES 

W = .0111l3-0114 

R = .9750 



2 / 2» 













•2 »2 > 

2 y/ 


r • * 












• 4 5^ • • 
. «2>«2.. 












2 
23. 
..2 36» 

3.22 3, 


. ».*<.32. 
232/222. .. 
4%*.24 •• 
<H3 •• 


. 






• 


• • 
•33 


3233Z/33 
... 33»4/<f.2 
.223S77«22 
...694^83. 
2332635243. •• 
•633/22.2 • 

68^.3. 


. . 




2 


2 


• 
■4 


2..23* 1 

>*32 

«5623« 


252* 







S43» 



J_ 



_1_ 



_L 



_L 



_L 



_L 



J_ 



X 



12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 

TOTAL LENGTH (cm) 

FIGURE 5.— Length-weight relationship based on 581 male white croaker sampled off southern California, 1978-81. 



(Fig. 8), averaging 0.39c of body weight (maximum 
0.99c, minimum 0.057c), and increased in the fall to a 
January peak averaging 2.69c of body weight (max- 
imum 7.7%, minimum 0.4%). 

In contrast, white croaker off Monterey Bay 
spawned over a longer period and may have winter 
and summer spawning peaks. Ovarian weights were 
highest in January and September (averaging about 
6.5 and 7.0% of body weight, respectively) and lowest 
in May (1.3% of body weight). Ovaries never shrank 
to the minimum sizes typical of individuals in the 
southern California population during summer 
months. Testes grew to a much larger maximum size 
(4.6% vs. 2.6%) off Monterey. Northern white 
croaker spawned nearly every month, though spring 
spawning might have been limited. In limited sam- 



pling during the following year, 10 the second 
(January) peak was not evident, and thus may not be 
an annual event. 

Batch fecundities ranged from an estimated 800 
eggs in a 15.5 cm female to about 37,200 in a 26 cm 
female (Fig. 9). During the spawning period about 
19% of all mature female white croaker sampled con- 
tained hydrated eggs, implying that a female 
spawned about once every 5 d. Females of ages 1 and 
2 (13-18 cm) have a spawning season of 3 mo (Table 
6) and spawn about 18 times per season, whereas 
older fish ( 1 9 cm and larger) spawn over a period of 4 
mo, about 24 times/season. 



10 T. Keating, Moss Landing Marine Laboratory, P.O. Box 223, 
Moss Landing, CA 95039, pers. commun. January 1982. 



187 




14 15 16 17 18 
TOTAL LENGTH (cm) 

FIGURE 6.— Length-maturity relationship 
in 995 female and 941 male white croaker 
collected off southern California, 1978-81. 



FISHERY BULLETIN: VOL. 82, NO. 1 

Larvae 

Data from our ichthyoplankton surveys showed that 
white croaker spawning occurs every month of the 
year (Fig. 10). However, a distinct seasonal spawning 
period can be deduced from findings that few larvae 
were collected from June through November, 
whereas high densities were encountered from 
January through April with a strong peak in March. 
Results of our study in King Harbor, Redondo Beach 
(Fig. 11), confirm the peak densities of white croaker 
larvae in January, February, and March. 

White croaker larvae constitute an important com- 
ponent of the neritie ichthyoplankton fauna of the 
Southern California Bight, ranking second in overall 
abundance behind the northern anchovy, Engraulis 
mordax. On a per transect basis (Fig. 12), white 
croaker larvae ranked first in abundance at all tran- 
sects between Palos Verdes 1 ' and Laguna Beach and 



n Genyonemus and Engraulis were virtually tied for first place at 
Redondo Beach with 40.1% and 40.39;, respectively. 




MONTEREY 




SOUTHERN 
CALIFORNIA 



1 



I 



i 



1 



1 



1 



_L 



JAN 



FEB MAR APR MAY JUNE JULY AUG SEPT OCT NOV DEC 



Figure 7. — Seasonal changes in the gonosomatic index (GSI-gonad weight as a percentage 
of total body weight) of female white croaker (based on 720 southern California and 223 
Monterey individuals). Vertical lines indicate 95',' confidence intervals of the mean. 



188 



LOVE ET AL.: LIFE HISTORY AND FISHERY OF WHITE CROAKER 



5 - 



MONTEREY 



< 

5 



4 - 



3 - 



2 - 




SOUTHERN 
CALIFORNIA 



JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT NOV DEC 

FIGURE 8. — Seasonal changes in the gonosomatic index of male white croaker (based on 631 
southern California and 1 14 Monterey individuals). Vertical lines indicate 95'* confidence 
intervals of the mean. 



40 



«" 36 - 



b 
o 

2 32 



z 

D 24 



1 20 

u 

2 16 



1? 

a 

4 



F = .000093 L 
R = 0.80 



6.08 




14 15 16 17 



19 20 21 £?. 

TOTAL LENGTH 



25 26 



Figure 9.— Batch fecundity— total length relationship for 44 white 
croaker collected off southern California during February and 
March 1979-81. 



second at the remaining transects, except Mission 
Beach where it ranked third behind Engraulis and an 
unidentified goby. In the King Harbor study, white 
croaker larvae ranked either fourth or fifth depend- 
ing on the year and the stations sampled. 

Larval density data (number of individuals per unit 
volume of water) indicate two spawning centers be- 
tween Point Conception and the U.S. -Mexican bor- 
der (Fig. 13): The larger one extends north and south 
of the Palos Verdes Peninsula, from Redondo Beach 



7000 



6000 



o 
o 

o 



5000 



Q- 
UJ 

< 
> 

DC 

< 



O 
a. 



4000 



3000 



3 
Z 



2000 



1000 — 




AUG SEP OCT NOV DEC JAN FEB MAR APR MAYJUN JUL 
1979 1980 

FIGURE 10.— Mean density of white croaker larvae collected in the 
oblique bongo tows per month between August 1979 and July 
1980. 

to Laguna Beach, whereas the smaller one is further 
north around Ventura. That area from San Onofre 
south to the international border was striking for its 
low densities of white croaker larvae. Along this sec- 



189 



FISHERY BULLETIN: VOL. 82, NO. 1 



M 

E 
o 
O 



550 - 



500 - 



450 - 



400- 



350- 



HI 

a. 300 

HI 
< 

> 

< 



250 - 



O 

jjj 200- 

D 

150 - 



100 - 



50 - 




1974 



MIJIJIAIS 
1977 




FIGURE 1 1 . — Mean densities of white croaker larvae in the vicinity of King Harbor, Redondo Beach, Calif., between January 1974 and February 

1978. 



tion of the coast, white croaker larvae accounted for 
11.7% of the larval fishes collected in oblique tows 
versus 43.6% from Laguna Beach to Redondo Beach 
and 17.9% from Playa del Rey to Point Conception 
(Fig. 14). 

Our data indicate that highest densities of white 
croaker larvae occur near the bottom (Fig. 15). In the 
coastal zone, between the 15 and 36 m isobaths, rela- 
tive densities indicate little variation through the 
water column, being 1.5-3.5% with surface waters, 
55.0-58.0% in the bottom waters, and 40.0-42.5% in 
middepth waters (Fig. 16). Relative densities in the 
surface waters at the shallow 8 m stations 
dramatically increased to 17.5% with a correspond- 
ing decrease in both bottom and middepth waters. 

White croaker larval densities peaked at stations 
located at 1 5 and 22m depths (Fig. 15). The densities 



declined sharply at the deeper (36 m) and shallower 
stations (8 m). The only exception to this trend was in 
surface water where densities steadily decreased in 
an offshore direction. 

Only 1 5 of our 20 transects had stations at 8 and 22 
m isobaths. Data in Figure 15 suggest that an abun- 
dance estimate based on the 8 and 22 m stations may 
approximate one based on the 15 and 36 m stations. 
If so, an estimate based on either of those station 
pairs should approximate one based on all four. We 
examined this at the three transects (OB, RB, SO), 
where data for all four stations were available. We 
tested the data from each transect for each of the 12 
mo of the sampling program using the sign test 
(Dixon and Massey 1957). The estimated number of 
white croaker larvae per 1 ,000 m 3 based on the 8 and 
22 m stations was compared with the estimate based 



190 



LOVE ET AL.: LIFE HISTORY AND FISHERY OF WHITE CROAKER 



on the 8, 15, 22, and 36 m stations; no statistically 
significant difference was found (N = 26; P > 0.05). 
The similarities of the overall estimates based on 
these two station groupings are shown in Figure 13. 

On the basis of our 8 and 22 m stations we have 
extrapolated density estimates to 36 m. Estimates 
were made for the truncated Palos Verdes and 
Laguna Beach transects as well, which are likely to be 
upwardly biased as they are based on two high den- 
sity stations (15 and 22 m). Data in Figure 13 show 
that these two transects are not high density ones; in 
fact, Palos Verdes is low for that section of the coast. 
The Laguna Beach transect is lower than the next two 
transects to the north. We included the Laguna 
Beach transect in the portion of the Southern Califor- 
nia Bight where white croaker larvae are in high abun- 
dance, on the basis that the density would still be 
higher than the portion of the coast from San Onofre 
to San Diego, based on just the 8 and 22 m stations, 
even if the 8 m station contributed no larvae. 

We estimated, from oblique bongo tows taken at the 



8 and 22 m stations (15 and 22 m stations at Palos 
Verdes and Laguna Beach), the average density of 
white croaker larvae between August 1979 and July 
1980 to have been 740/1,000 m-\ 2,203/1,000 m\ 
and 4 1 1/1,000 m 3 for the regions between Point Con- 
ception and Playa del Rey, Redondo Beach and 
Laguna Beach, and San Onofre and the international 
border, respectively. On the basis that there is no 
significant difference between estimates based on 
the 8 and 22 m stations and one based on the 8, 15, 
22, and 36 m stations, we use the 8 and 22 m density 
estimates to project the average number of white 
croaker larvae to the 36 m isobath. 

It has been estimated (Lavenberg and McGowen 
footnote 7) that about 3 1 km 3 of water are located in a 
band along the coast between Point Conception and 
the U.S. -Mexican international border and extend- 
ing seaward to the 36 m isobath. Of this, 15.6 km 3 
(50.6%) is located in the region between Point Con- 
ception and Playa del Rey, 7.9 km 3 (25.9%) between 
Redondo Beach and Laguna Beach, and 7.2 km 3 



■j-i 
\- 

u 

LU 

w 

2 

< 



80 

DR 
81.5 

RN 
83 
OB 

85 

MU 

87 
RB 
PV 

88 
BA 
90 
SO 
91 
CD 

93 

MB 
95 



SANTA 

BARBARA 



3 2 

RANKINGS 




SAN 

DIEGO 



FIGURE 12. — Rank abundance of white croaker larvae collected in 
oblique bongo tows taken along 20 transects in the Southern Califor- 
nia Bight between August 1979 and July 1980. See Table 2 for sta- 
tion abbreviation definitions. 



DR 

81.5 



RN 



88 

BA 

90 

SO 
91 



93 
MB 
95 



Based on 2 Stations 
Based on 4 Stations 




r 



T 




SAN 
DfEGO 



500 



1000 1500 2000 2500 3000 3500 
NUMBER OF LARVAE PER 1000 m 3 



FIGURE 13. — Mean densities of white croaker larvae along 20 tran- 
sects in the Southern California Bight between August 1979 and 
July 1980. See Table 2 for station abbreviation definitions. 



191 



FISHERY BULLETIN: VOL. 82, NO. 1 



BO 

DR 
81 5 

KN 

B3 
OB 



h- 

u 

I/) 

z 
< 



MU 
87 
RB 
PV 

88 
BA 

90 

SO 
91 



93 
MB 
95 



10 



— I - 

20 



— 1 1 — 

30 40 

PERCENT 




SAM 
DtEGO 



< 

> 

< 



3 



3300 - 



3000 



2700 



2400 



2100 



1800 



1500 



1200 



900 



600 



300 - 



• = MANTA 

□ = MID-DEPTH 

 = BENTHIC 
A= OBLIQUE 




"I 1 I - 

08 15 22 

BOTTOM DEPTH (m) 



36 



60 



FIGURE 15.— Mean density of white croaker larvae collected with 
each of four different tow types along four isobaths — 8, 15, 22, and 
36 m— in the Southern California Bight between August 1979 and 
July 1980. 



FIGURE 14. — The percentage contributed by white croaker to the 
total number of larvae collected along each of 20 transects in the 
Southern California Bight between August 1979 and July 1980. 
See Table 2 for station abbreviation definitions. 



(23.57c) between San Onofre and the international 
border. Based on these values plus the density 
estimates, we project the average number of white 
croaker larvae in each of the three areas during this 
period to have been 1.15 X 10 10 , 1.75 X 10 10 , and 2.97 
X 10 9 , respectively. Thus, about 55% of the white 
croaker spawned in the area between Redondo 
Beach and Laguna Beach, 36% between Playa del 
Rey and Point Conception, and about 9% between 
San Onofre and the border. 

Fishery 

Most of the white croaker retained by sportfisher- 
men were adults (Fig. 17), being 21-25 cm and 5-7 yr 



Figure 16. — Mean percentage of white croaker larvae collected 
near the surface, near the bottom and in between along each of four 
isobaths — 8, 15, 22, and 36 m — in the Southern California Bight be- 
tween August 1979 and July 1980. 



6IJ 



50 - 



O 



4(1 



30 



M 



10 



| NEUSTON 

| | MID-DEPTH 

I BENTHIC 



4 



08 15 22 

BOTTOM DEPTH (m) 



q 



36 



192 



LOVE ET AL.: LIFE HISTORY AND FISHERY OF WHITE CROAKER 



900 



800 



700 



600 



O 500 
CC 

111 

co 

E 

Z 

400 



300 



200 - 



100 



100% 



!=□_ 



10 12 14 16 18 20 22 24 

TOTAL LENGTH (cm) 



M 



28 



30 32 



FIGURE 17. — Lengths of white croaker retained by skiff sportfishermen off southern 
California, 1980-81, with length at IOO'a maturity noted. 



old. Small fish were only occasionally hooked, and 
rarely retained. 

Within the Southern California Bight, about 10 
vessels fished white croaker full time. Two areas, 
Long Beach south to Dana Point and Oxnard to 
Santa Barbara, were fished most heavily, which cor- 
responded to the sites of peak white croaker larvae 
concentrations reported here. 

This is a gill net fishery, and an informal agreement 
among fishermen sets the net mesh at 7.0 cm (2.75 in) 
stretch. Nets are 1.3 km (0.8 mi) long and are set on the 
bottom in depths of 5.5-37 m (3-20 fathoms). Mean 
catches of white croaker are 270-400 kg (600-900 lb) 
per set with maximum catches of 680-770 kg (1,500- 
1,700 lb). Largest catches occurred in January and 
February, during spawning season, when white 
croaker aggregated in large numbers. The prices for 
1982 to fishermen were 13-18«/kg (30-40*/lb). Most 
fish taken during our study were 26-29 cm long 



(Fig. 18) and 8-10 yr old. We found no immature 
fish. 

DISCUSSION 
Depth Preference 

Though most species of Sciaenidae prefer inshore 
waters, white croaker are distributed over a wider 
depth range than other northeastern Pacific species. 
Queenfish was the fourth most abundant species 
taken in our survey at the shallowest station (Table 
3); its abundance declined rapidly with depth. 
Though it was present in deeper water, it contributed 
<0.1% of the fishes taken at 59-73 m. The white 
seabass is common within the 30 m contour (though 
they are taken as deep as 90 m during winter months). 
Umbrina roncador, Roncador stearnsi, and Men- 
ticirrhus undulatus prefer sandy beaches and bays to 



193 



FISHERY BULLETIN: VOL. 82, NO. 1 



500 



400 



< 

I 300 



m 

3 
Z 



200 



100 



FIGURE 18.— Lengths of white croaker retained by com- 
mercial gill net fishermen off southern California, 1980- 
81. 



100% 



18 20 22 24 26 28 

TOTAL LENGTH (cml 



30 



32 



34 



depths of perhaps 9 m (Skogsberg 1939), whereas 
Cheilotrema saturnum are common over reefs to 
perhaps 15 m (occasionally to 45 m) (Limbaugh 
1961). 

Most eastern Pacific drums are limited to the warmer 
waters south of Point Conception (Miller and Lea 
1972) or, like the queenfish and white seabass, are 
rare north of the Point. Conversely, white croaker are 
abundant north to San Francisco. Temperature pre- 
ference experiments 12 indicate that juvenile white 
croaker have wide metabolic thermal optima (11°- 
17 C, based on routine oxygen consumption rates) 
that may account for their wide depth and 
latitudinal ranges. 

Though white croaker are most abundant over sandy, 
featureless substrata, they are occasionally found in 
large numbers in kelp beds. This is particularly the 
case in beds anchored on sand, such as those off San 
Onofre and Santa Barbara. Similarly, though they 
spend most of their time near the bottom, we have 
noted schools in midwater, 20-40 m or more above 
the substrata. We have also seen white croaker at the 
surface, chasing anchovy schools. 

Maturation and Reproduction 

We computed the length-maturity relationship 
using standard length to compare our results with 



1! Hose, J. E., and W.H.Hunt. 1981. Physiological responses of 
juvenile marine fish to temperature. Occidental College Annual 
Report submitted to Southern California Edison, 17 p. 



those of Issacson (1967). We found 50% of the males 
mature by 12.0 cm SL and 50% of females by 13.0 cm 
SL, both at 1 yr. This was in sharp contrast to Issac- 
son's statement that "The white croaker matures 
between 147 and 164 mm standard length at an age of 
3 to 4 years." Why such a disparity should exist is 
unclear. 

White croaker is the only southern California drum 
that spawns in the winter. Winter spawning is 
unusual even among tropically derived temperate 
species off California. All species in the families 
Blenniidae, Carangidae, Labridae, Pomacentridae, 
Scombridae, and Sphyraenidae are either summer 
spawners or spring and summer spawners with a 
summer spawning peak. An exception are the rock- 
fishes (Scorpaenidae), the vast majority of which 
spawn in winter and/or spring. 

The more or less continuous (or perhaps dual- 
peaked) spawning season seen in white croaker in 
Monterey Bay is an interesting phenomenon. Most 
California marine fishes have restricted spawning 
seasons. If spawning does continue for extended 
periods (as in the bocaccio, Sebastes paucispinis), 
there is usually only one peak spawning period. An 
exception is the northern anchovy, Engraulis mor- 
dax, that may spawn year-round and which exhibits a 
major peak in late winter-early spring and a minor 
one in early fall. 

Fishes of the northeastern Pacific tend to have a 
longer spawning season in the southern part of their 
range, as favorable conditions are usually more re- 



194 



LOVE ET AL.: LIFE HISTORY AND FISHERY OF WHITE CROAKER 



stricted in northern waters (Westrheim 1975). 
However, on examination, the water temperatures in 
Monterey Bay more closely approximate optimal 
white croaker spawning conditions than those off 
southern California. The peak spawning periods, 
based on gonosomatic indices and ichthyoplankton 
surveys, in southern California occur between 
January and March, when mean surface tem- 
peratures decrease to 13°-14°C (U.S. Department of 
Commerce 1956). Off Monterey, the mean tem- 
peratures of the warmest months are 13°-14°C 
(June-October), whereas the other months are 1°- 
3°C cooler. Thus white croaker encounter tem- 
peratures conducive to spawning for more months off 
Monterey than off southern California. 

White croaker reproductive behavior is in some re- 
spects the opposite of the cooccuring queenfish. White 
croaker spawn almost entirely during late winter and 
early spring (peak February-March), but our 
ichthyoplankton survey gives a March-April peak, 
whereas queenfish are spring and summer spawners 
(peak April-May, DeMartini and Fountain 1981). 
Most egg hydration in white croaker takes place dur- 
ing the night, with spawning occurring from just 
before dawn to midmorning. Queenfish spawn be- 
tween late afternoon and evening. We have not ascer- 
tained the extent that habitat partitioning has played 
in this separation. Off Monterey, where queenfish are 
rare, white croaker spawn virtually year round. As 
discussed before, this is perhaps a reflection of a 
more favorable temperature regime. It would be 
instructive to know if in the absence of queenfish, egg 
hydration and spawning time are simlar to those off 
southern California. 

Larvae 

Data from both gonosomatic indices and 
ichthyoplankton surveys show white croaker spawn 
year-round in southern California waters. However, 
peak spawning clearly is in the winter and spring. Our 
data, combined with Watson's (1982), indicate that 
peak densities of white croaker larvae were in either 
January, February, or March from 1974 through 
1980. This is out of phase with other southern 
California sciaenids, all of which spawn primarily in 
the spring and summer (Lavenberg and McGowen 
footnote 7). 

White croaker larvae are an important component 
of the southern California neritic ichthyoplankton 
fauna. Along the three sections of the Southern 
California Bight, defined and studied during this 
investigation, white croaker larvae contributed 11.7, 
43.6, and 17.9% of the total larvae from south to 



north. Highest densities were found at stations 
located in 15-22 m depths (Fig. 15). The decreasing 
densities, as one moves shoreward of the 15 m 
isobath, apparently continues into the enclosed bays 
and estuaries of southern California. McGowen 
(1981) did not collect any white croaker larvae in 
south San Diego Bay during a 13-mo study. Larval 
white croaker ranked sixth, contributing 0.6% of the 
larvae collected in Newport Bay during an 18-mo 
study by White (1977). The percentage reported by 
White may have a bias toward lower values because 
the period of peak spawning was sampled only once 
during the 18 mo. However, even a doubling of 
White's percentages does not make white croaker 
larvae dominant members of the Newport Bay ich- 
thyoplankton assemblage. Leithiser (1981) reported 
white croaker to contribute 1.9% of the total catch of 
larval fishes in Anaheim Bay during a 12-mo 
study. 

King Harbor is typical of the estuarine-enclosed bay 
habitat rather than that of the open coast and is 
dominated by blennies, clinids, gobies, and 
engraulids (McGowen footnote 8). White croaker lar- 
vae ranked either fourth or fifth in the King Harbor 
study, depending on the year and the stations 
sampled. 

Densities of white croaker larvae also decreased 
between the 22 and 36 m isobaths (Fig. 15). This 
indication that white croaker larvae are not common 
in offshore waters is supported by CalCOFI data. 
The highest any sciaenid ranked in these collections 
between 1955 and 1958 was 18th, contributing 
0.30% of the total larval catch (Ahlstrom 1965). 

This pattern of white croaker larvae being dis- 
tributed in a narrow band along the coast, between 
the 15 and 22 m isobaths, is similar to the pattern 
reported by Watson (1982) and Barnett et al. 13 off 
San Onofre. They designated white croaker larvae as 
having an inner nearshore epibenthic pattern. Bar- 
nett et al. (footnote 13) indicated highest densities on 
the bottom, shoreward of the 22 m isobath, and the 
second highest densities in the water column be- 
tween the 12 and 22 m isobaths and on the bottom 
between the 22 and 45 m isobaths. The major dis- 
crepancy between their data and ours is the higher 
epibenthic densities that they report shoreward of 
the 12 m isobath and seaward of the 22 m isobath. 
This discrepancy may be partially explained by dif- 



"Barnett, A. M., A. E. Jahn, P. E. Sertic, and W. Wat- 
son. 1980. Long term spatial patterns of ichthyoplankton off San 
Onofre and their relationship to the position of the SONGS cooling 
system. A study submitted to the Marine Review Committee of the 
California Coastal Commission, July 22, 1980, Unpubl. rep., 32 
p. Marine Ecological Consultants of Southern California, 533 
Stevens Ave., Suite D-57, Solana Beach, CA 92075. 



195 



FISHERY BULLETIN: VOL. 82, NO. 1 



ferences in sampling strategy. They sampled within 
blocks defined by depth contours whereas we sampled 
at specific isobaths. Thus, part of their block D (be- 
tween the 22 and 45 m isobaths) is located at a depth 
where we found high densities (22 m) and part of it 
where we found low densities (36 m). All of their 
block B (between 9 and 12 m) is located at depths 
where we did not sample. Their block A (between 6 
and 9 m) is located in a zone where our data suggest 
lower densities. 

Our trawling data also support this narrow band as 
important for the young stages of white croaker. 
Almost all of the juvenile white croaker taken during 
our study were collected at stations located between 
the 18 and 27 m isobaths (Fig. 2). 

In summary, these data suggest that adult white 
croaker migrate shoreward (larger adults were taken 
at deeper depths; Fig. 2) and spawn in a narrow band 
along the coast. This band has its shoreward boun- 
dary located between the 8 and 12 m isobaths, and its 
seaward boundary located between the 22 and 36 m 
isobaths. Furthermore, the pelagic stages remain pri- 
marily within this band. At the end of the pelagic 
phase young white croaker move into 3-6 m and take 
up residence near the bottom. As these juvenile fish 
mature, they migrate to deeper waters (Fig. 2). 

Based on this hypothesis, we believe that a realistic 
evaluation of the spawning activities of the white 
croaker can be based on data collected from the 
shore to the 36 m isobath. We have done this and 
found that about 9% of the spawning by white croaker 
occurred along the coast from San Onofre to the 
international border, about 55% from Laguna Beach 
to Redondo Beach, and around 36% from Playa del 
Rey to Point Conception. If this represents the typi- 
cal annual pattern, the portion of the Southern 
California Bight from Laguna Beach to at least Point 
Conception is important for white croaker, especially 
the region around the Palos Verdes Peninsula from 
Redondo Beach to Laguna Beach. However, that 
portion of the bight from San Onofre to the border is 
relatively insignificant. The only remaining coastal 
zone in the U.S. portion of the Southern California 
Bight is around the Channel Islands. We have not 
investigated the coastal zones of these islands and 
cannot appraise their significance to the spawning 
activities of white croaker in the Southern 
California Bight. 

Fishery 

Historically, the commercial white croaker fishery 
has been minor, rarely exceeding 1 million lb/yr (Frey 
1971). Most fish were caught and landed in the Long 



Beach-San Pedro region and Monterey Bay. 
Southern California accounted for about two-thirds 
of the catch and Monterey one-third, although during 
World War II, Monterey produced over one-half the 
total catch. Until recently, white croaker were taken 
commercially by otter trawl, round haul net, mul- 
tifilament gill net, and hook and line. However, in the 
past few years, significant changes have occurred in 
the fishery. Gill nets, particularly monofilament nets, 
have almost entirely supplanted other methods. 

The ubiquity of white croaker along the southern 
California mainland makes this species accessible to 
small boat sportfishermen. The ease with which it 
may be taken, using minimum skill or equipment, 
ensures that this species will be caught in consider- 
able numbers. We commonly found two fishermen 
with at least 50 or more white croaker after a half 
day's effort. Though traditionally scorned by many, 
we found that the species is popular with a number of 
ethnic groups. 

The Monterey fishery has been revived in the past 
2-3 yr by newly arrived Vietnamese fishermen. 14 
White croaker are fished throughout Monterey Bay, 
over the entire year, in 12-24 m (40-80 ft), 
occasionally to 37 m (120 ft) with 1.6-2.4 km (1-1.5 
mi) long monofilament gill nets [6.3 cm (2.5 in) 
stretch mesh]. Nets are tended daily, and 450-900 kg 
(1,000-2,000 lb) catches are common with maximum 
catches to 1,800 kg (4,000 lb). Depending on catch 
size and fish condition, payment to fishermen ranges 
from 6 to 22C/kg (15 to 50C/lb). These white croaker 
are sold principally within central California (par- 
ticularly the San Francisco area), although a small 
amount is shipped to southern California. Demand is 
increasing, particularly among various Asian com- 
munities. 15 

SUMMARY 

In this study, white croaker was the most abundant 
species in nearshore (18-27 m) otter trawl collections 
in southern California. This species dwelled prin- 
cipally in shallow water and juveniles were restricted 
to the shallower (<27 m) parts of the species depth 
range. Living to 12 yr, white croaker grew at a nearly 



l4 D.J. Miller, California Department of Fish and Game, 2201 Gar- 
den Road, Monterey, CA 93940, and T. Keating, Moss Landing 
Marine Laboratory, P.O. Box 233, Moss Landing, CA 95039, pers. 
commun. August 1981. 

"Though most white croaker are retailed fresh, there is reason to 
believe that a potential market exists for them as surimi (fish cakes). 
A fish cake plant existed in Ventura during 1979, processing 3,000- 
4,000 lb (1,360-1,800 kg) of white croaker per day. All cakes were 
sold to the Asian community in Los Angeles. Demand for the pro- 
duct was very strong and the plant closed for reasons unrelated 
to profitability. 



196 



LOVE ET AL.: LIFE HISTORY AND FISHERY OF WHITE CROAKER 



constant rate throughout the species' life. A majority 
of both males and females matured at about 1 yr and 
all were mature by 4 yr. We noted a difference in 
spawning season between southern and central 
California. Off southern California, significant 
spawning occurred between November and April, 
while central California individuals spawned all year, 
with large-scale activity occurring from July through 
February. Our ichthyoplankton survey indicated that 
two spawning centers occurred off southern 
California — one located from Redondo Beach to 
Long Beach and the other centered about Ventura. 
White croaker larvae, which were second in abun- 
dance to northern anchovy in nearshore waters, were 
found in greatest abundance near the substratum in 
15-22 m of water. The abundance of white croaker 
and its ease of capture make it a major sportfish in the 
skiff fishery and a growing component of the com- 
mercial gill net fishery. Our study indicates that the 
vast majority of fishes taken in both fisheries were 
adults. 

ACKNOWLEDGMENTS 

We thank J. Stephens for his continual support of 
our work. J. Palmer, T. Sciarrotta, and J. Stock of the 
Southern California Edison Company and G. Brewer 
of the University of Southern California assisted in 
project design and logistical support. 

L. McCluskey helped estimate batch fecundities, 
and L. Natanson and E. Taylor conducted the small 
vessel creel census. T. Keating supplied numerous 
Monterey specimens, and J. Balesteri supplied data 
on the southern California commercial operation. 

Majority of the larval identifications were made by 
D. Carlson, D. Chandler, D. Eto, R. Feeney, S. Good- 
man, N. Singleton, D. Winkler, and R. Woodsum of 
the University of Southern California and the Natural 
History Museum of Los Angeles County. E. Gray and 
L. Games of the Southern California Edison Com- 
pany and the Natural History Museum of Los 
Angeles County, respectively, assisted with data 
reduction and computer programming. We also 
thank M. Butler (illustrations) and R. Meier 
(photography) of the Los Angeles County Natural 
History Museum. Lastly, we thank the many people 
who assisted in the sorting and collecting of samples, 
especially the crews of RV Vantuna and RV 
Seawatch. 

LITERATURE CITED 

AHLSTROM, E. H. 

1965. Kinds and abundances of fishes in the California 



Current region based on egg and larval surveys. Calif. 

Coop Oceanic Fish. Invest. Rep. 10:31-52. 
Allen, R. L. 

1976. Method for comparing fish growth curves. N.Z.J. Mar. 

Freshw. Res. 10:687-692. 
BAGENAL, T. B.. AND E. Braum. 

1971. Eggs and early life history. In W. E. Ricker (editor), 
Methods for assessment offish production in fresh waters, 
p. 166-198. (Int. Biol. Programme) Handb. 3., Blackwell 
Sci. Publ., Oxf. and Edinb. 

Brown, D. M 

1979. The manta net: quantitative neuston sampler. Inst. 
Mar. Res. Mar. Tech. Handb. IMR TR 64, 15 p. 
Castle, W. T., and L. A. Woods, Jr. 

1972. DDT residues in white croakers. Calif. Fish Game 
58:198-203. 

Dailey, M. D., L. A. Jensen, and B. W. Hill. 

1981. Larval anisakine roundworms of marine fishes from 
southern and central California, with comments on public 
health significance. Calif. Fish Game 67:240-245. 
DeMartini, E. E., and R. K. Fountain. 

1981. Ovarian cycling frequency and batch fecundity in the 
queenfish, Seriphus politus: attributes representative of 
serial spawning fishes. Fish. Bull., U.S. 79:547-560. 
Dixon, W. J., and F. J. Massey, Jr. 

1957. Introduction to statistical analysis. 2d ed. McGraw- 
Hill Book Co., Inc., N.Y., 488 p. 
FREY, H. W. 

1971. California's living marine resources and their 
utilization. Calif. Dep. Fish Game, 148 p. 

Goldberg, S. R. 

1976. Seasonal spawning cycles of the sciaenid fishes 
Genyonemus lineatus and Seriphus politus. Fish. Bull., 
U.S. 74:983-984. 

ISSACSON, P. A. 

1964. Length-weight relationship of the white croaker. 

Trans. Am. Fish. Soc. 93:302-303. 
1967. Notes on the biology of the white croaker, Genyonemus 
lineatus (Ayres). Trans. Kentucky Acad. Sci. 28:73-76. 
Leithiser, R. M. 

1981. Distribution and seasonal abundance of larval fishes in 
a pristine southern California salt marsh. Rapp. P.-V. 
Reun. Cons. Int. Explor. Mer 178:174-175. 
LlMBAUGH, C. 

1961. Life-history and ecologic notes on the black 
croaker. Calif. Fish Game 47:163-174. 
MacGregor, J. S. 

1972. Pesticide research at the fishery-oceanography cen- 
ter. Calif. Coop. Oceanic Fish. Invest. 16:103-106. 

McGowan, J. A., and D. M. Brown. 

1966. A new opening-closing paired zooplankton net. 
SIO Ref. 66-23, 56 p. 
McGowen, G. E. 

1981. Composition, distribution, and seasonality of 
ichthyoplankton populations near an electricity generat- 
ing station in south San Diego Bay, California. Rapp. 
P.-V. Reun. Cons. Int. Explor. Mer 178:112-114. 
MEARNS, A. J. 

1974. Southern California's inshore demersal fishes: diver- 
sity, distribution, and disease as reponses to environmen- 
tal quality. CalCOFI (Calif. Coop. Oceanic Fish. Invest.) 
Rep. 17:140-148. 
1979. Responses of coastal fishes and invertebrates to waste- 
water discharges. Prog. Water Technol. 4:19-32. 
Mearns, A. J., and M. J. Sherwood. 

1977. Distribution of neoplasms and other diseases in marine 



197 



FISHERY BULLETIN: VOL. 82, NO. 1 



fishes relative to the discharge of waste water. Ann. N.Y. 
Acad. Sci. 298:210-224. 
Miller, D. J., and D. Gotshall. 

1965. Ocean sportfish catch and effort from Oregon to Point 
Arguello, California. Calif. Dep. Fish Game, Fish Bull. 
130, 135 p. 
Miller, D. J., and R. N. Lea. 

1972. Guide to the coastal marine fishes of California. Calif. 
Dep. Fish Game, Fish Bull. 157, 235 p. 
Morejohn, G. V., J. T. Harvey, and L. T. Krasnow. 

1978. The importance oiLoligo opalescens in the food web of 
marine vertebrates in Monterey Bay, California. In C. W. 
Recksiek and H. W. Frey (editors). Biological, ocean- 
ographic, and acoustic aspects of the market squid, Loligo 
opalescens Berry, p. 67-98. Calif. Dep. Fish Game, Fish 
Bull. 196. 
Phillips, L. C, C. Terry, and J. S. Stephens. 

1972. Status of the white croaker (Genyonemus lineatus) in 
the San Pedro Bay region. South Calif. Coastal Water 
Proj. Tech. Rep. 109, 49 p. 
Pinkas, L., M. S. Oliphant, and C. W. Haugen. 

1968. Southern California marine sportfishing survey: 
private boats, 1964; shoreline, 1965-66. Calif. Dep. Fish 
Game, Fish Bull. 143, 42 p. 

Russell, F., and P. Kotin. 

1957. Squamous papilloma in the white croaker. Nat. Can. 
Inst. J. 18:857-861. 
Sherwood, M. J. 

1978. The fin erosion syndrome. South. Calif. Coastal Water 
Res. Proj., Annu. Rep., p 203-221. 
Skogsberg, T. 

1939. The fishes of the family Sciaenidae (croakers) of 
California. Calif. Dep. Fish Game, Fish Bull. 54, 62 p. 



Stout, V. F., and F. L. Beezhold. 

1981. Chlorinated hydrocarbon levels in fishes and shell- 
fishes of the northeastern Pacific Ocean, including the 
Hawaiian Islands. Mar. Fish. Rev. 43(1):1-12. 

U.S. Department of Commerce. 

1956. Surface and water temperatures at tide stations, 
Pacific coast, North and South America and Pacific Ocean 
islands. U.S. Coast Geod. Surv. Spec. Publ. 280, 74 p. 

Ware, R. R. 

1979. The food habits of the white croaker Genyonemus 
lineatus and an infaunal analysis near areas of waste dis- 
charge in outer Los Angeles Harbor. M.A. Thesis, 
California State Univ., Long Beach, 113 p. 

Watson, W. 

1982. Development of eggs and larvae of the white croaker, 
Genyonemus lineatus Ayres (Pisces: Sciaenidae), off the 
southern California coast. Fish. Bull., U.S. 80:403-417. 

Westerheim, S. J. 

1975. Reproduction, maturation, and identification of larvae 
of some Sebastes (Scorpaenidae) species in the northeast 
Pacific Ocean. J. Fish. Res. Board Can. 32:2399-2411. 

White, W. S. 

1977. Taxonomic composition, abundance, distribution and 
seasonality of fish eggs and larvae in Newport Bay, 
California. M.A. Thesis, California State Univ., Fuller- 
ton, 107 p. 
Wine, V., and T. Hoban. 

1976. Southern California independent sportfishing survey 
annual report, July 1, 1975-June 30, 1976. Calif. Dep. 
Fish Game, 109 p. 



198 



FEEDING HABITS OF BLACKSMITH, CHROMIS PUNCTIPINNIS, 
ASSOCIATED WITH A THERMAL OUTFALL 



Pamela A. Morris 1 



ABSTRACT 

The availability and use of food by blacksmith, Chromispunctipinnis, were examined at a thermal outfall and 
a control site in King Harbor, California. Stomach analysis showed that blacksmith from the outfall area con- 
sumed a significantly greater amount of food, consist ing of larger prey items, than control fish. Movements of 
water created by the outflow may provide dietary benefits by reducing zooplankton predator avoidance and 
by entraining and entrapping organisms not normally planktonic. This dietary enrichment may result in 
attraction of blacksmith to the King Harbor outfall. 



An increased demand for energy resulting in growth 
of coastal power plant activity has created concern 
for the effects of heated effluents upon the fish com- 
munity (Miller 1977; Stephens 1978, 2 1980 3 ; 
Stephens and Palmer 1979 4 ). Few studies have 
examined the factors attracting fish to outfall areas. 
White et al. (1977) found less diversity and lower 
abundance of fish at an outfall station, while Kelso 
(1976) and Minns et al. (1978) reported a clustering 
offish in the vicinity of thermal outfalls. Underwater 
observations suggest that fish are attracted to ther- 
mal outfalls to feed. Kelso (1976) found that fish in 
proximity to a thermal discharge exhibited a complex 
swimming behavior that could represent feeding 
activity. Moreover, this behavior continued when 
unheated effluent was discharged. 

The blacksmith, Chromis punctipinnis (family 
Pomacentridae), an abundant planktivorous tem- 
perate reef inhabitant, has been regularly observed 
feeding at the thermal outfall of a steam electrical 
generating station in King Harbor, Redondo Beach, 
Calif. Recent studies on the effects of thermal effluents 
upon blacksmith have concentrated on behavioral 



'VANTUNA Research Group, Department of Biology, Occidental 
College, Los Angeles, CA 90041. 

'Stephens, J. S., Jr. 1978. Effects of thermal effluent from 
Southern California Edison's Redondo Beach steam generating 
plant on the warm temperate fish fauna of King Harbor Marina. Fish 
and laboratory study reports for Phase III. VANTUNA Research 
Group, Department of Biology, Occidental College, Los Angeles, CA 
90041. 

'Stephens, J. S., Jr. 1980. Effects of thermal effluent from 
Southern California Edison's Redondo Beach steam generating 
plant on the warm temperate fish fauna of King Harbor Marina. Fish 
and laboratory study reports for 1977-1978. VANTUNA Research 
Group, Department of Biology-, Occidental College, Los Angeles, CA 
90041. 

4 Stephens, J. S., Jr., and J. B. Palmer. 1979. Can coastal power 
stations be designed to offset impacts by habitat enrichment? Gen. 
Tech. Rep. RM-65, p. 446-450. Paper presented at Mitigation 
Symposium, U.S. Department of Agriculture, Fort Collins, Colo. 



responses to intermittent chlorination (Hose and 
Stoffel 1980; Hose et al. in press). The objective of 
this study was to examine the feeding habits of black- 
smith and determine whether the discharge was 
attracting them through dietary enrichment. 

MATERIALS AND METHODS 

This study was conducted at King Harbor, Redondo 
Beach, Calif., at the southern end of Santa Monica 
Bay, just north of the Palos Verdes Peninsula (Fig. 1, 
lat. 33°51'N, long. 188°24'W) (Terry and Stephens 
1976; Stephens and Zerba 1981). Situated just 
offshore is the head of the Redondo Submarine 
Canyon, a source of cold upwelling water for the har- 
bor. In contrast, thermal effluent from Units 7 and 8 
of Southern California Edison's Redondo Beach 
steam electrical generating plant is discharged just 
inside the harbor mouth. 

The thermal outfall study site consists of a vertical 
conduit, 4 m in diameter, out of which the effluent is 
pumped. The circular outlet is level with the sub- 
strate at a depth of 7 m. Effluent is discharged at a 
rate of 1.78 X 10 6 1/min during peak operation. 

A control site was chosen about 500 m from the dis- 
charge. This area, referred to as the Point, is located 
at the tip of the breakwater that partially encloses the 
harbor. This site has been surveyed by Stephens and 
Zerba (1981) who note that blacksmith are an abun- 
dant resident species. 

A form of presence/absence monitoring was used as 
an indicator offish abundance at the discharge. Mean 
estimates (0-25, 26-50, 51-75, 76-100, or >100) were 
made by two scuba divers swimming a circular tran- 
sect around the discharge. The position of fish was 
recorded: in the plume (the column of water directly 
over the discharge), in the outer plume (the area of 



Manuscript accepted June 1983. 

FISHERY BULLETIN: VOL. 82. NO. 1. 1984. 



199 



FISHERY BULLETIN: VOL. 82. NO. 1 



Santa Monica 
Bay 



Redondo Beach 




Los 
Angel es 



100 500 

I 1 1 1 1 1 I 

meters 



FIGURE 1.— Study area at King Harbor, Redondo Beach, Calif. 

water immediately surrounding the plume), or at the 
base (the substrate surrounding the discharge). 

The abundance of fishes at the Point has been 
documented since 1974 (Stephens and Zerba 1981), 
and work continued at this area during the same time 
period the discharge was examined. Two divers 
equipped with slates and depth gauges swam in one 
direction along the rock breakwater at a fixed depth 
for 5 min, counting all fish seen 1.5 m above and 
below them and within sight to either side. Transects 
were run at depths of 1.5, 4.5, 7.5, and 10.5 m, with 
replicates at each depth. 

In order to determine the nature of the feeding 
habits of blacksmith at the discharge versus those 
feeding at the Point, utilization of food items based 
on stomach analysis was examined for each area. 
General availability of food was estimated by sam- 
pling plankton at both sites. 

Stomach analysis closely followed methods em- 
ployed by Ellison et al. (1979). Fish were collected 
from each study site by scuba divers using pole 
spears. During fish collection, a temperature profile 
was taken using a temperature probe coupled to a 
telethermometer (Yellow Springs Instruments Co., 
Model 43 ID 5 ). After capture the fish were placed on 



ice. The body wall was cut open and the stomach 
injected with a 207c Formalin solution. The fish were 
then preserved in a 1 0% Formalin solution for at least 
48 h, rinsed in running water for 2 h, and placed in 
70% isopropyl alcohol. 

Within 2 wk from date of capture, fish stomachs 
were removed and placed in vials of 70% isopropyl 
alcohol. At this time the standard length, wet weight, 
and sex of each fish were noted. Each stomach was 
then blotted dry (with special care taken to remove 
the internal fluid) and weighed, food items dissected 
out, and the empty stomach weighed again. Stomach 
fullness was estimated using a scale from (empty) to 
5 (full). 

Individual prey items were separated into the 
lowest identifiable taxa and counted, and the percent 
of the total volume estimated. In most cases, only 
whole organisms or whole organism indicators were 
counted. In prey items which were not eaten whole 
(i.e., algae and ectoprocts), only the percent volume 
was estimated. 

In 1979-80, 73 fish were collected at the discharge 
area from 13 sampling days during a 15-mo period. 
Four sampling days were in the afternoon (1430- 
1830 h) and 10 were in the morning (0830-1100 h). A 
total of 35 blacksmith were collected from the Point 
area before noon (1000-1130 h). 

During the study period, 28 plankton samples from 
the discharge plume and 13 plankton samples from 
the Point were collected. The mean rank order abun- 
dance of prey items from each site was determined 
for comparison with blacksmith stomach contents. 

Observations comparing different prey items from 
two locations were tested using contingency table 
analysis, the G-test (Crow 1982), and Kendall's coef- 
ficient of rank correlation. When only one variable 
(fish weight, stomach fullness etc.) was tested be- 
tween two locations, a two-sample f-test was used, 
assuming separate variances. Values of the Index of 
Relative Importance (IRI) were calculated for con- 
sumed prey from the sum of the percent number and 
the percent volume, multiplied by the frequency of 
occurrence (Foe) (Pinkas et al. 1971). 

Dietary overlap between blaeksmth from the Point 
and discharge was examined using the formula of 
Schoener (1970): 



a 



1-0.5 ( Z \Px r Py, 



'Reference to trade names does not imply endoresement by the 
National Marine Fisheries Service, NOAA. 



where n is the number of food categories, x, is the 
average percentage of estimated volume that food 
category ;' contributed to species at location x, andy, 
is the average percentage of estimated volume that 
food category i contributed to species at location y. 



200 



MORRIS: FEEDING HABITS OF BLACKSMITH 



An estimate of mean prey size was obtained by 
dividing the total number of prey consumed into the 
stomach weight for each fish collected. 

RESULTS 

Thirty species of fish were identified from the area 
surrounding the discharge. Blacksmith were the 
most abundant and frequently occurring fish (mean 
estimate of abundance per transect >100 in- 
dividuals, rank of the mean number per transect = 1 , 
and frequency of occurrence per transect = 92.3). 
Large schools arrived in the morning and began feed- 
ing in the plume and outer plume. When feeding in 
the outer plume, blacksmith would orient themselves 
toward the plume, surrounding it, and feed on the 
organisms that settled out of the rising effluent. 
When in the plume, blacksmith were in constant 
motion, being tossed about by the irregular flow, but 
it was apparent from mouth action that these fish 
were also feeding on suspended food items. 

The mean abundance per transect of blacksmith at 
the Point for the quarterly sampling days in 1979 and 
1980 was 148.4. They ranked first in mean number 
per transect, with a mean frequency of occurrence of 
86.2, and used the breakwater as their primary noc- 
turnal sheltering site. 

There were no significant differences in either fish 
length or fish weight, but there were significant dif- 
ferences in stomach weight and stomach fullness be- 
tween the two collection sites (Table 1). Fish 
collected from the discharge had a greater amount of 
food in their stomachs (an increase of 138%). 

Stomach fullness was not influenced by collection 
time. The stomach weight and stomach fullness were 



TABLE 1. — Comparison of blacksmith, Chromis punctipinnis, collect- 
ed from the discharge (thermal outfall) and the Point (Control Site), 
King Harbor, Calif. 





Discharge 


Point 




n = 73 


n = 35 


Fish weight (g| 


Mean = 175.6 g 


Mean = 168.3 g 




SD= 38.4 


SD = 44 7 




r = 0819 






P = 416 




Fish length (SL mm) 


Mean = 1 72.8 mm 


Mean = 1 72.2 mm 




SD= 12 9 


SD= 14,3 




t = 569 






P = 0.571 




Stomach weight (g) 


Mean = 1 10 g 


Mean = 0.30 g 




SD = 0.53 


SD = 0.25 




t= 9 726 






'P<0.001 




Stomach fullness (0-5) 


Mean = 3.89 


Mean = 1.63 




SD= 1.06 


SD= 1 .09 




f = 10.175 






P<0.001 





'Note: The statistical package (SPSS) used was unable to compute P values lower 
than 0.001 . Values below this number are represented as P < 0.001 . 



not significantly different between morning and 
afternoon collections (£-test:£ = 1. 359, P= 0.181 and 
t = 1.471, P = 0.147, respectively). Consequently, 
the data collected from the discharge samples were 
combined. 

The mean prey abundance, percent number, per- 
cent volume, frequency of occurrence, and the 
calculated IRI value of the 30 most abundant prey 
items from each location are given in Table 2. A con- 
tingency table analysis of the mean abundance 
indicates thatthere was a significant difference in the 
stomach contents between the two locations (G = 
570.6, P <0.001, df = 17). The 10 most abundant 
prey from each site (eliminating the smaller values) 
are significantly different (G = 56 1 . 1 , P < 0.00 1 , df = 
12). A comparison of the 10 highest IRI values from 
each site are not significantly correlated (Kendall's 
tau, t = 0.1868, P = 0.324, n = 14). A pictorial rep- 
resentation of the IRI values is given in Figures 2 
and 3. 

A comparison of the mean prey weight from each 
sampling site revealed that blacksmith from the dis 
charge ate larger prey than blacksmith from the Point 
(discharge mean prey weight = 3.22 mg, SD = 4.01, 
Point = 0.82 mg, SD = 0.81, t = 4.439, P <0.001). 

Temperatures from the discharge plume and base 
were compared with surface and bottom tem- 
peratures at the Point. The mean plume temperature 
(26.3°C, SD = 3.3, n = 15) was significantly greater 
(t-test: t = 5.69, P <0.001) than the mean surface 
temperature from the Point (20.8°C, SD = 2.5, n = 
30). Similarly, the mean base temperature (18.2°C, 
SD= 2.4, n = 17) was significantly greater (t = 4.12, 
P < 0.001) than the mean bottom temperature from 
the Point (15.2° C, SD = 2.4, n = 30). 

The rank of the 10 most abundantly consumed prey 
items was compared with the rank of the 10 most 
abundant plankton items for both the discharge and 
Point. There was no significant correlation for either 
study site (discharge t = 0.01 10, P = 0.956, n = 14; 
Point t= 0.2051, P= 0.329, n = 13). 

Between-site comparisons of the mean abundance 
of six abundantly consumed prey items from both 
stomach contents and plankton samples (Table 3) 
show that two prey items, gammarids and 
Polyophthalmus pictus, had a significantly higher 
usage and availability at the discharge than the Point, 
and that Calanus sp. and mysids had a higher usage 
at the discharge but were not significantly more avail- 
able. There was no significant difference in the usage 
or availability of Oikopleura sp. between the Point 
and discharge (although blacksmith from the Point 
tended to eat a greater amount). 

The diets of blacksmith at the discharge and Point 



201 



FISHERY BULLETIN: VOL. 82. NO. 1 



Table 2. — The 30 most abundant food items consumed by blacksmith, Chromis punctipinnis , at the dis- 
charge (thermal outfall) and the Point (control site), King Harbor, Calif. Foc= frequency of occurrence; 
IRI = index of relative importance. 









Point 








Discharge 








 


% 


% 






X 


% 


% 








no 


no. 


vol. 


Foe. 


IRI 


no. 


no 


vol. 


Foe. 


IRI 


Oikopleura 


43049 


77.5 


41.3 


77.1 


9,159 5 


29063 


33.7 


108 


87 7 


3.902.7 


Ac am a 


5903 


10.6 


3.2 


71.4 


985.3 


46.38 


5.4 


2.1 


67,1 


503.3 


Calanoids. misc. 


26.1 1 


*1 


3.2 


71.4 


564 1 


14.86 


1.7 


1.4 


75 3 


2334 


Polychaeta. misc. 


8.03 


14 


1 6 


57 1 


171.2 


2.53 


0.3 


1.2 


50 7 


76 .1 


Corycaeus 


6.51 


1.2 


1.4 


45 7 


118 8 


3.68 


0.4 


0,5 


42.5 


38.3 


Calanus 


529 


9 


5.7 


62 9 


415.1 


298 36 


34.6 


110 


76.7 


3.497.5 


Chaetognath 


4 51 


0.8 


3 4 


57.1 


239 8 


4,14 


0.5 


0.5 


34.2 


27.4 


Labidocera 


2 60 


5 


:> 2 


48.6 


131 2 


2.12 


0.2 


0.6 


52 1 


41.7 


Brachyuran zoea 


1 89 


03 


9 


22.9 


27 5 


8 14 


09 


1.2 


46.6 


97 9 


Gammandae 


1.83 


3 


1 1 


42.9 


60 1 


1 11.33 


12 9 


25.3 


91.8 


3,506.8 


Pagurid zoea 


1.63 


3 


h 


25.7 


23.1 


3 27 


0.4 


0.5 


43.8 


394 


Cladocera 


1.60 


0.3 


4 


28.6 


200 


0.49 


0.1 


0,1 


20,5 


5.7 


Hhincalanus 


1 .17 


2 


06 


34.3 


27.4 


2.27 


3 


0,6 


41.1 


30.9 


Euphausids 


097 


2 


04 


25.7 


154 


082 


0.1 


I 


26.0 


5 2 


Tortanus 


077 


0.1 


1.2 


286 


37 2 


1 52 


0.2 


0,4 


32.9 


17 2 


Cypns larvae 


0.54 


1 


02 


22 9 


69 


1.10 


1 


0.2 


37,0 


11.1 


Fish eggs 


0.49 


0.1 


0.2 


28.6 


8 6 


0.53 


1 


0.1 


24,7 


4 9 


Cirnpide exoskel. 


0.46 


(I I 


0.1 


143 


2.9 


0.42 


1.) 1 


0.6 


27.4 


10 


Polyophthalmus pictus 


0.34 


0.1 


1 


2.9 


3 


25.70 


3.0 


6 8 


28.8 


282.2 


Gastropoda 


0.34 


0.1 


2 


22.9 


6 9 


0.37 


1 


1 


205 


4 1 


Fish larvae 


0.31 


0.1 


0.3 


17 1 


6.8 


3 29 


0.4 


1 4 


35.6 


64.1 


Mysids 


0.31 


1 


02 


200 


6.0 


3601 


4.2 


7 7 


80.8 


961.5 


Ophelndae 


0.14 


0.1 


0.1 


86 


1.7 


0.90 


1 


03 


30.1 


12.0 


Decapoda. misc. 


0.06 


0.1 


2 


5.7 


1.1 


0.55 


1 


9 


30.1 


30 1 


Caprellidae 


003 


1 


1 


2 9 


3 


1.90 


2 


1.0 


466 


55 9 


Porcellanid zoea 


003 


0.1 


1 


2 9 


3 


0.89 


1 


0.2 


32.9 


9 9 


Pelecypoda 


n 














0.60 


1 


0.4 


24,7 


12 4 


Anemone 

















3.29 


0.4 


0,5 


15.1 


13.6 


Ecto-Entoprocta 


— 


— 


0.1 


2 9 


0.3 


— 


— 


1 5 


8.6 


12 9 


Unidentified, misc. 


— 


— 


11 2 


35.3 


395 4 


— 


— 


69 


42.9 


296.0 



TABLE 3. — Usage and availability of selected prey items from the 
Point (control site) and discharge (thermal outfall), King Harbor, 
Calif. 





In stomachs 1 


In plar 


ikton 2 


Prey items 


Discharge 


Point 


Discharge 


Point 


Polyophthalmus pictus 








Mean 


25 70 


0.34 


30.59 





SD 


67.49 


2.03 


86 52 







t= 3.207 


P = 0.002 


3p<0.001 




Acartia 










Mean 


46.38 


5903 


181,987.13 


167,487.59 


SD 


112.02 


130.95 


323.297.81 


133,525.13 




t = 0.492 


P= 0.625 


r = 0.031 


P> 0.840 


Calanus 










Mean 


29936 


5 29 


364.41 


721.00 


SD 


753.29 


11.96 


717.27 


1,260.84 




t = 3.323 


P= 0.001 


t = -0.959 


^ = 0.353 


Mysidacea 










Mean 


3601 


031 


943.28 


306 38 


SD 


75.38 


0.72 


3.562.00 


568.29 




t = 4.046 


P < 0.001 


r = 0981 


P= 0.333 


Grammandae 










Mean 


111.33 


1.83 


6.291.81 


472.92 


SD 


174.51 


4.52 


10.784.44 


845.73 




f = 5.357 


P < 001 


t = 3.029 


P = 0.005 


Oikopleura 










Mean 


290.63 


43049 


6.82981 


4,582,08 


SD 


471.00 


557.59 


19.821.55 


9,906 22 




t = -1.281 


P= 205 


r = 505 


P= 0616 



'Mean number of prey consumed per fish. 
2 Mean number per 100 m 3 of water sampled. 

3 Note: The statistical package (SPSS) used was unable to compute P values low- 
er than 0.001 . Values below this number are represented as P <0.001 , 



did not overlap (a = 0.522, with a value >0.60 con- 
sidered significant, Zaret and Rand 1971). 

DISCUSSION 

Blacksmith were a numerically dominant species at 
both study sites. The daytime abundance of black- 
smith was similar at the discharge and the Point. 
Blacksmith may travel to the discharge from the 
breakwater and other nearby jetties during the day, 
since they do not seek shelter around the discharge at 
night. Such diel migrations of blacksmith between 
the Units 7 and 8 intake of Southern California 
Edison's Redondo Beach Station and the nocturnal 
rocky shelters at the Point have been previously 
observed. 6 

The feeding habits of blacksmith were significantly 
different between the Point and discharge (Figures 2 
and 3 best illustrate this difference). At the Point, 
Oikopleura and calanoid copepods (primarily Acar- 
tia) were the most heavily utilized organisms. At the 
discharge, blacksmith consumed larger organisms, 
gammarids, calanoid copepods of the genus Calanus, 



6 M. Helvey, VANTUNA Research Group, Occidental College, Los 
Angeles, CA 90041, pers. commun. 1980. 



202 



MORRIS: FEEDING HABITS OF BLACKSMITH 



50- 

01 

n 40 

E 

c 30 

>> 

n 20 

10 



0) 10 
E 

> 

>. 30- 

a 

^ 40 



CALANOIDS 
(1 Calanus) 












OIKOPLEURA 






? i 








i 




GAMMARIDS 




POLYCHAETE 


Other 

CRUSTACEANS 


MISC 


75 


43 


88 


92 




88 





92 



FREQUENCY OCCURRENCE 

FIGURE 2.— Graphic representation of the Index of Relative Importance of prey items consumed hy blacksmith, Chromis 
punctipinnis, at the discharge (thermal outfall) in King Harbor, Calif. 



-D 

E 

3 
C 

>< 



E 

3 

o 

> 

>> 
n 






80 












70 






60 












50 












40 












30 






OIKOPLEURA 






20 




GAMRDS 




P0LYCHAETE CRUSTACEAN 




10 



10 


" CALANOIDS 


i 

O 1 

i 


(1°Acartia) 


43 




57 ' 71 


MISC 




35 


20 


83 




30 












40 














77 




50 













P 40 - 



FREQUENCY OCCURRENCE 

FIGURE 3. — Graphic representation of the Index of Relative Importance of prey items consumed by blacksmith, 
Chromis punctipinnis, at the Point (control site) in King Harbor, Calif. 



large polychaetes, other crustaceans, as well as 
Oikopleura. At both sites blacksmith were selective 
in their planktonic feeding, consuming the largest 
prey items available. Brooks (1968) stated that there 



is selection for larger zooplankters, with smaller ones 
eaten as the larger ones become scarce. At the Point, 
Oikopleura was the largest prey item found in abun- 
dance, while at the discharge other larger food items 



203 



FISHERY BULLETIN: VOL. 82, NO. 1 



were common along with Oikopleura (gammarids, 
Polyophthalmus pictus, and mysids). The amount of 
dietary overlap between the two locations was not 
considered significant. 

Although more abundant at the Point, a significant- 
ly greater amount of Calanus sp. was eaten by black- 
smith at the discharge than at the Point. A possible 
explanation for the high usage of Calanus at the dis- 
charge could be the increased susceptibility of 
zooplankton to predation as a result of turbulent out- 
flow. Entrained Calanus are more accessible to 
planktivorous fishes, since the mortality rate of 
copepods passing through a power plant may reach 
70% (Carpenter et al. 1974). Dead or damaged 
copepods would appear as viable prey upon dis- 
charge from the plant and could be easily consumed. 
Increased mortality from turbulence has also been 
shown for other zooplankters (Gregg and Ber- 
gersen 1980). 

There is evidence that alterations in plankton dis- 
tributions at outfall areas are the result of upward 
vertical displacement of deep-water organisms. 
Evans (1981) noted that deeper living zooplankton 
are carried vertically upward to the turbulent waters 
over the discharge jets. Although analysis of 
plankton sampled did not prove the existence of such 
currents, in a previous study at King Harbor dye 
injections were carried to the plume from bottom 
water 20 m away from the discharge. 7 

Large gammarids, polychaetes, and juvenile 
anemones, all of which were common in stomachs of 
blacksmith from the discharge, are not normal con- 
stituents of King Harbor plankton. The force of the 
swirling effluent is strong enough to detach and 
entrap these organisms from their normal habitat 
inside and around the discharge pipe. Once 
entrapped in the plume, these large invertebrates are 
accessible to the planktivorous blacksmith. 

Zooplankton avoid predation through escape 
movements upon detection of suction currents 
created by predatory fish (Dreeneretal. 1978; Kettle 
and O'Brien 1978). Once entrained in the effluent 
plume, the ability of zooplankton to detect these 
currents becomes impaired (Evans 1981). As a 
result, fish frequenting the plume have the potential 
for feeding on a high concentration of zooplankton 
with limited predator avoidance. The greater 
stomach weight and stomach fullness of blacksmith 
feeding at the discharge support this theory. 

Results from other studies examining the feeding 



7 Kinnetic Laboratories, Inc. 1981. Hydrodynamic characteris- 
tics of offshore intake structures. Field verification studies. Kin- 
netic Labs., Inc., P.O. Box 1040, 1 Potrero St., Santa Cruz, CA 
95061. 



habits of blacksmith appear to be similar to those 
found at the Point. The food items consumed by 
blacksmth at Santa Catalina Island are (listed in de- 
creasing abundance) Oikopleura, calanoid and 
cyclopoid copepods, fish eggs, cladocerans, and 
other crustaceans (Hobson and Chess 1976). At 
Naples Reef, off Santa Barbara, Calif., Bray (1981) 
found the diet of blacksmith to consist of larvaceans 
(Oikopleura), copepods, cladocerans, chaetognaths, 
decapods, and polychaetes. In the two above- 
mentioned studies and from the Point, blacksmith 
consumed at least twice as many Oikopleura as any of 
the other food items, while at the discharge, Calanus 
was the most abundantly consumed prey and gam- 
marids comprised the greatest volume of prey eaten 
(Table 2). When Calanus, gammarids, mysids, and 
the polychaete Polyophthalmus pictus are removed 
from the analysis of the 10 most abundant prey con- 
sumed, no significant difference was observed be- 
tween the two locations (G — 9.4, n.s. atP= 0.05, df = 
7). 

It has long been recognized that blacksmith forage 
on plankton in areas where currents are present 
(Limbaugh 1955, 1964; Feder et al. 1974; Ebeling 
and Bray 1976; Hobson and Chess 1976; Bray 1981). 
The tropical species of damselfish (family Pomacen- 
tridae) also prefer feeding in areas where currents are 
strong (Hobson and Chess 1978). Blacksmith have 
been shown to prefer incoming currents (Limbaugh 
1955, 1964; Ebeling and Bray 1976; Bray 1981), and 
Limbaugh believed they materially affected the 
amount of plankton entering the kelp beds. In Bray's 
(1981) study, stomach fullness was greater in fish at 
the incurrent end of the reef than in fish at the 
excurrent end. 

Areas of strong currents are rich in zooplankters 
(Hobson and Chess 1978) as is the discharge which 
receives both entrained and entrapped organisms. 
Although the discharge releases warm water, the 
current created by the outflow is the major attract- 
ant. Blacksmith, a species which prefers warm water 
(mean preferred temperature = 14°-15°C), are found 
in 26°-32°C discharge plume water, above their 
upper temperature avoidance limit of 23°-25°C 
(Shrode et al. 1982). In the presence of food, black- 
smith will disregard their normal avoidance limits for 
chlorine, intermittently present in most power plant 
effluents (Hose and Stoffel 1980). 

It can be concluded that the outflowing effluent and 
its related phenomena attract blacksmith to the dis- 
charge. This theory is further supported by 
documentation of similar attraction and rheotropic 
behavior by blacksmith at an offshore water intake 
structure (Helvey and Dorn 1981). 



204 



MORRIS: FEEDING HABITS OF BLACKSMITH 

ACKNOWLEDGMENTS 

For their help in the collection of field data, I wish to 
thank K. Zerba, D. Terry, K. Shriner, C. Rand, T 
Wong, B. Johnson, J. Hough, and L. McCluskey. I 
also wish to thank M. Love, J. E. Hose, G. Martin, and R. 
N. Bray for their help in reviewing this manuscript. My 
utmost appreciation goes to J. Stephens, Jr., for his 
support and guidance throughout this project and to S. 
Warschaw for her generous help in preparing the 
final copy. 

I wish to acknowledge the support of J. Palmer and 
Southern California Edison Research and Develop- 
ment Project No. C0650901 to J. Stephens, Jr. 

LITERATURE CITED 

Bray, R. N. 

1981. Influence of water currents and zooplankton densities 
on daily foraging movements of blacksmith, Chromis 
punctipinnis, a planktivorous reef fish. Fish. Bull., U.S. 
78:829-841. 

Brooks, J. L. 

1968. The effects of prey size selection by lake plank- 
tivores. Syst. Zool. 17:272-291. 
Campbell, R. C. 

1974. Statistics for Biologists. 2d. ed. Cambridge Univ. 
Press, 385 p. 
Carpenter, E. J., B. B. Peck, and S. J. Anderson. 

1974. Survival of copepods passing through a nuclear power 
station on northeastern Long Island Sound, USA. Mar. 
Biol. (Berl.) 24:49-55. 
Crow, M. E. 

1982. Some statistical techniques for analyzing the stomach 
contents of fish. //; G. M. Cailliet and C. A. Simenstad 
(editors), Fish food habits studies, p. 8-15. Proceedings 
of the Third Pacific Workshop. Washington Sea Grant 
Publ., Univ. Washington. 

Dreener, R. W., J. R. Strkkler, and W. J. O'Brien. 

1978. Capture probability: the role of zooplankter escape in 
the selective feeding of planktivorous fish. J. Fish. Res. 
Board Can. 35:1370-1373. 

Ebeling, A. W., and R. N. Bray. 

1976. Day versus night activity of reef fishes in a kelp forest 
off Santa Barbara, California. Fish. Bull., U.S. 74:703- 
717. 
Ellison, J. P., C. Terry, and J. S. Stephens, Jr. 

1979. Food resource utilization among five species of 
embiotocids at King Harbor, California, with preliminary 
estimates of caloric intake. Mar. Biol. (Berl.) 52:161- 
169. 

Evans, M. S. 

1981. Distribution of zooplankton populations within and 
adjacent to a thermal plume. Can. J. Fish. Aquat. Sci. 
38:441-448. 
Feder, H., C. H. Turner, and C. Limbaugh. 

1974. Observations on fishes associated with kelp beds in 
Southern California. Calif. Dep. Fish Game, Fish Bull. 
160, 144 p. 

Gregg, R. E., and E. P. Bergersen. 

1980. Mysis relicta: Effects of turbidity and turbulence on 
short-term survival. Trans. Am. Fish. Soc. 109:207-212. 



Hklvey, M., AND P. Dorn. 

1981. The fish population associated with an offshore water 
intake structure. Bull. South. Calif. Acad. Sci. 80:23-31. 

Hohson, E. S., and J. R. Chess. 

1976. Trophic interactions among fishes and zooplankters 

near shore at Santa Catalina Island, California. Fish. 

Bull., U.S. 74:567-598. 
1978. Trophic relationships among fishes and plankton in the 

lagoon at Enewetak Atoll, Marshall Islands. Fish. Bull., 

U.S. 76:133-153. 
Hose, J. E., AND R. J. STOFFEL. 

1980. Avoidance response of juvenile Chromis punctipinnis 
to chlorinated seawater. Bull. Environ. Contam. Toxicol. 
25:929-935. 

Hose, J. E., R. J. Stoffel, and K. E. Zerba. 

In press. Behavioral responses of selected marine fishes to 
chlorinated seawater. Mar. Environ. Res. 
Kelso, J. R. M. 

1976. Movement of yellow perch (Perca flavescens) and white 
sucker (Catostomus commersoni) in a nearshore great 
lakes habitat subject to a thermal discharge. J. Fish. Res. 
Board Can. 33:42-53. 

Kettel, D., and W. J. O'Brien. 

1978. Vulnerability of arctic zooplankton species to preda- 
tion by small lake trout (Saluelinus namaycush). J. Fish. 
Res. Board Can. 35:1495-1500. 
Limbaugh, C. 

1955. Fish life in the kelp beds and the effects of kelp harvest- 
ing. Univ. Calif. Inst. Mar. Resour. Ref. 55-9, 158 p. 
1964. Notes on the life history of two Californian pomacen- 
trids: Garibaldis, Hypsypops rubicunda (Girard), and 
blacksmiths, Chromis punctipinnis (Cooper). Pac. Sci. 
18:41-50. 
Miller, S. 

1977. The impact of thermal effluents on fish. Environ. Biol. 
Fish. 1:219-222. 

Minns, C. K., J. R. M. Kelso, and W. Hyatt. 

1978. Spatial distribution of nearshore fish in the vicinity of 
two thermal generating stations, Nanticoke and Douglas 
Point, on the Great Lakes. J. Fish. Res. Board Can. 
35:885-892. 

Pinkas, L., M. S. Oliphant, and I. L. K. Iverson. 

1971. Food habits of Albacore, Bluefin Tuna, and Bonito in 

California waters. Calif. Dep. Fish Game, Fish Bull. 152, 

105 p. 
Schoener, T. W. 

1970. Nonsynchronous spatial overlap of lizards in patchy 
habitats. Ecology 51:408-418. 

Shrode, J. B., K. E. Zerba, and J. S. Stephens, Jr. 

1982. Ecological significance of temperature tolerance and 
preference of some inshore California fishes. Trans. Am. 
Fish. Soc. 111:45-51. 

Stephens, J. S., Jr., and K. E. Zerba. 

1981. Factors affecting fish diversity on a temperate 
reef. Environ. Biol. Fish. 6:111-121. 

Terry, C. B., and J. S. Stephens, Jr. 

1976. A study of the orientation of selected embiotocid fishes 
to depth and shifting seasonal vertical temperature 
gradients. Bull. South. Calif. Acad. Sci. 75:170-183. 

White, J. W., W. S. Woolcott, and W. L. Kirk. 

1977. A study of the fish community in the vicinity of a ther- 
mal discharge in the James River, Virginia. Chesapeake 
Sci. 18:161-171. 

Zaret, T. M., and A. S. Rand. 

1971. Competition in tropical stream fishes: support for the 
competitive exclusion principle. Ecology 52:336-342. 



205 



CALIBRATION OF DENTAL LAYERS IN SEVEN CAPTIVE 

HAWAIIAN SPINNER DOLPHINS, STENELLA LONGIROSTRIS, 

BASED ON TETRACYCLINE LABELING 



Albert C. Myrick, Jr., 1 Edward W. Shallenberger, 2 
Ingrid Rang, 2 and David B. Mac-Kay' 



ABSTRACT 



To calibrate dentinal and cemental growth layer groups (GLGs) with real time, a study was conducted on the 
teeth from seven captive Hawaiian spinner dolphins that had been treated clinically with tetracycline (TCL) 
at numerous times over multiple years at Sea Life Park, Hawaii. To monitor layer accumulation as it occurred 
for 1 year, we gave single injections to three animals every 3 months and pulled a tooth from each every 6 months. 
By comparing dental-layer patterns between TCL labels that had been introduced at 6-month and 1-year 
intervals, annual patterns were distinguished. In the dentine, a thin, light layer (the first being the neonatal 
line) was formed about every 6 months. Each annual GLG contained 13 lunar monthly layers (LMLs). Using 
LML or light-layer counts, age, month, and year of birth were estimated for each of the seven specimens. All 
seven deposited nearly the same dentinal GLG thickness in the same year of life. Estimates of birth months 
indicated that five of the animals were born in late summer or early autumn and two were born in spring. Com- 
parisons of dentinal labels with clinical records for a captive-born animal showed that TCL given to its 
mother was imparted via milk to the nursing calf. Time calibration of cemental GLGs showed that usually one 
cemental GLG was deposited annually, but in some cases a GLG was formed every second year or twice 
a vear. 



The technique of "reading" layers or growth layer 
groups (GLGs, terminology of Perrin and Myrick 
1980) in teeth, developed to determine ages for pin- 
nipeds in the early 1950's by Scheffer (1950) and 
Laws ( 1952), is now used routinely in dolphin studies 
(see reviews by Klevezal' and Kleinenberg 1967; 
Jonsgard 1969; Scheffer and Myrick 1980). Early 
work on dolphins (e.g., Nishiwaki and Yagi 1953; 
Sergeant 1959), showing a correlation between 
apparent age and number of GLGs led to the working 
assumption that GLG-deposition cycles are con- 
stant, each GLG usually, but not always, interpreted 
as representing 1 yr. Critical analysis of this assump- 
tion has been impaired by a lack of suitable 
material. 

Three approaches have been used in efforts to 
calibrate dental GLGs with time and to determine 
their deposition rate: 1) In vivo labeling of tooth 
layers, 2) multiple extractions of teeth over time, and 
3) examination of teeth from animals of known age. 
Nishiwaki and Yagi (1953) labeled the layered den- 
tine in four wild-caught striped dolphins, Stenella 
coeruleoalbo, by intramuscular injection of lead ace- 
tate paste. None of the four survived long enough for 
the labels to provide useful data. 



'Southwest Fisheries Center, National Marine Fisheries Service, 
NOAA, La Jolla, CA 92038. 
2 Sea Life Park, Makapuu Point, Waimanalo, HI 96795. 
'Kaneohe Veterinary Clinic, Kaneohe, HI 96744. 



Nielsen (1972) treated a young wild-caught harbor 
porpoise, Phocoena phocoena, with tetracycline 
(TCL) three times over a 370-d period. Three 
fluorescent labels were found in thin sections of its 
teeth examined in ultraviolet (UV) light "... but the 
uniform [unlayered] dentine made it impossible to 
determine the number of growth-layers formed per 
year" (Nielsen 1972:72). 

Best (1976) administered oral doses of TCL hy- 
drochloride, "Mysteclin-V", on each day over an 8-d 
period to each of three wild-caught dusky dolphins, 
Lagenorhynchus obscurus. Labels were detected in 
teeth of two of the three specimens after their deaths. 
In one specimen, dentine accumulated for 703 d be- 
tween treatment and death averaged 200 jum/yr and 
0.56 jum/d. In the other (older) specimen, the average 
deposition rate in dentine between the treatment 
label and the pulp-cavity wall was 77 jiim/yr and 0.21 
ftm/d. Best concluded that the thickness of GLGs 
decreases significantly with age in dusky dolphins. 

Gurevich et al. (1980) successfully introduced a 
single TCL label into the teeth of three of four wild- 
caught adult common dolphins, Delphinus delphis. 
The three labeled animals died 328, 354, and 441 d, 
respectively, after the date of treatment. By estimat- 
ing the dentinal pattern laid down in about 1 yr, the 
investigators characterized an annual GLG. They 
estimated the ages of the animals by assuming that 
the GLGs in the unlabeled regions of the teeth rep- 



Manuscript accepted August 1983. 
FISHERY BULLETIN: VOL. 82, No. 1, 1984. 



207 



FISHERY BULLETIN: VOL. 82, NO. 1 



resented the same amount of time as the single GLG 
interpreted from the labeled region of each tooth. 

A study by Hui (1978) included two tooth extrac- 
tions made 2.5 yr apart from a captive male bot- 
tlenose dolphin, Tursiops truncatus (No. 10,"Kona"). 
Comparisons of longitudinal thin sections of the two 
teeth led Hui to conclude that "... almost three den- 
tin layers [GLGs] had been deposited during the 
intervening period. . . " (p. 1 1). Other than indicating 
GLG boundaries in figures of the two thin sections 
(his fig. 3), Hui did not describe the GLGs or their 
components. 

Three published studies (Sergeant 1959; Sergeant 
et al. 1973; Hui 1978) have attempted to 
demonstrate time content in GLGs using teeth of 
known-age, i.e., captive-born dolphins. All three had 
access to only a small number of specimens, all of 
Tursiops truncatus. Apparently, the investigators 
knew the ages of the specimens before defining and 
counting dentinal GLGs in the teeth, and no 
assurance was provided that the GLGs counted cor- 
responded to annual periods between birth and 
death. Hui's study demonstrated that GLGs may be 
defined in such a manner as to verify the age that is 
already known for a specimen (Myrick 1980a). The 
incorrect age data (3.3 yr) provided to Hui for one of 
two "known-age" specimens studied by him (Hui 
1978) led to his subsequent division of its dentinal 
layering pattern into three GLGs and a small fraction 
(Hui 4 ). The original clinical records for the specimen 
(Hui's No. 29, LACM 54698) show, however, that the 
dolphin was born on 28 August 1965 and died on 8 
August 1969, at nearly 4 yr of age. 

Used independently, teeth of known-age animals, 
single-labeled teeth, or teeth extracted on two dates 
do not provide reliable means by which to determine 
tissue accumulation rates fully or to define GLGs 
with precision. Each method yields only two dates 
bracketing a segment of layered tissue into which the 
known elapsed time is divided. Myrick (1980b) de- 
scribed approaches that combine the use of two or 
more labels and two or more tooth extractions over an 
extended period to monitor rates and calibrate 
GLGs. The present paper is an account of such a 
study which used TCL-labeled teeth from seven cap- 
tive Hawaiian spinner dolphins, Stenella longi- 
rostris. 

MATERIALS AND METHODS 

The study consisted of two phases. The first was a 



retrospective examination of TCL labels in the 
dolphins' dental tissues produced incidentally by 
clinical treatments administered during their cap- 
tivity at Sea Life Park, Hawaii. Teeth were used from 
four frozen carcasses (Nos. WFP 606, 669, 670, and 
67 1 5 ), including one specimen of known age, and 
three live animals (Nos. ACM 103, 104, and 106) 
from which teeth were extracted in early 1980. 

The second phase was a 1-yr monitoring of tissue- 
accumulation rates in teeth of three live animals. 
Each animal was given intramuscular injections of 
TCL at about 3 -mo intervals and underwent three 
tooth extractions during the monitored period. 

To restrain the dolphins during injections and 
extractions, an elevated rigid litter was placed near 
the edge of the dolphin holding tank in which the 
water level had been lowered to a depth of 0.5 m. The 
sloped tank bottom inclined the litter at an angle of 
20° relative to the water surface. Each dolphin in turn 
was guided on its belly onto the litter until the front 
half of its body was above the water surface. In this 
position the dolphin could be held firmly with little 
apparent discomfort to the animal. 

The procedure used to extract teeth was adapted 
for the spinners from the method described by 
Ridgway et al. (1975) for bottlenose dolphins. The 
dolphin's mouth was held open by moistened rolled 
toweling placed around the upper and lower jaws. 
Carbocaine 6 (5-10 cc) was injected into the right or 
left interalveolar nerve immediately behind the 
anterior border of the mandibular foramen. After 
allowing about 10 min for the anesthetic to take 
effect, a tooth was removed from the middle of the 
corresponding mandibular tooth row using an 
elevator and an extractor. The vacated alveolus was 
packed with cotton soaked with a ferric solution to 
control bleeding and promote healing. 

Liquamycin 100, a form of TCL, was injected into 
the dorsal musculature between the dorsal fin and 
the blow hole. To reduce the possibility of local 
inflammation of the tissue — a problem known to 
result from concentrations of TCL — each dose (25 
mg/kg body weight) was distributed along the dor- 
sum at three separate sites. 

Untreated (cut or ground) thin sections and 
decalcified and haematoxylin-stained (D/S) thin sec- 
tions are the two most widely used preparations for 
dolphin teeth in age determination studies (see 
Perrin and Myrick 1980: 21 ff.). D/S sections pro- 
duce simpler, more uniform GLG patterns, but de- 



4 Clifford Hui, Naval Ocean Systems Center, San Diego, Calif., pers. 
commun. 1981. 



'Skeletons are in the synoptic collection at Southwest Fisheries 
Center, NMFS, La Jolla, Calif. 

^Reference to trade names does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 



208 



MYRICK ET AL.: DENTAL LAYERS IN HAWAIIAN SPINNER DOLPHINS 



calcification removes TCL labels (Nielsen 1972). We 
prepared the spinner teeth using both methods. 

Untreated, mid-longitudinal thin sections, 100 [xm 
thick, were prepared by hand grinding and polishing 
teeth using 240 and 600 grit AI2O3 on a glass plate. 
Other teeth were decalcified in RDO 7 for 6-8 h, 
rinsed, and cut with a microtome in longitudinal 
plane to produce 30 /xm thick sections that were 
stained in Mayer's haematoxylin for 15-30 min. 
Untreated and D/S preparations were mounted on 
slides in Permount or glycerin gel and covered with 
coverslips. 

To determine the pattern components of GLGs, the 
D/S and untreated thin sections were examined in 
plain transmitted light 39X and 150X with a Zeiss 
photomicroscope. TCL labels were viewed at the 
same magnifications with UV reflected light using a 
Zeiss fluorescent vertical illuminator with a filter- 
reflector No. 44-75-05 combination attached to the 
same instrument. 

Retrospective Calibration of 
Dentinal GLGs 

Dates and durations of treatment, date of birth (for 
one specimen) or capture, and dates of death (for four 
carcasses) were taken from clinical records main- 
tained for each dolphin during its captive life at Sea 
Life Park (for summaries see Myrick et al. in press). 
Data for each specimen were transcribed onto a cali- 
bration chart as the chronological series of event blocks, 
the relative width of a given block corresponding to 
the length of a given period of treatment. 

In each thin section showing distinct fluorescent 
labels under UV light, label thicknesses and 
interlabel distances were measured. Label- 
measurement data for each dolphin were entered on 
its chart as a series of blocks below the event blocks, 
with spacing and thickness scaled to the correspond- 
ing measurements. The treatment and label blocks 
were compared for spacing and thickness to identify 
the date each label was introduced. Connecting lines 
were drawn from the beginning and the end of each 
matched pair of blocks (Fig. 1C). 

A UV photograph of each thin section was used to 
identify and letter key labels that enclosed 6- or 12- 
mo segments of dentine. Labels and structural 
landmarks in the UV photograph were traced with a 
china marker on an overlay of transparent plastic. 
Using the landmarks, the tracing was lined up on the 
corresponding plain-light photograph onto which the 



1 A commercial rapid decalcifying agent available through Dupage 
Kinetic Laboratories, Inc., Plainfield, 111. 



labels were reproduced to delineate layering pat- 
terns within the time segments. Each marked 
photograph was then inspected for repeating layer 
components to define GLGs and their subunits in the 
untreated thin section. GLGs defined in the labeled 
dentine of each thin section were used as a basis for 
identifying similar GLGs in the unlabeled regions of 
the dentine and permitted a complete series of GLG- 
thickness measurements and an estimate of dentinal 
age in years to be made for each animal. 

Dentinal GLGs in dolphin teeth are most easily dis- 
cerned in the region of the "shoulder", i.e., along a 
transect from near the base of the neonatal line (the 
first layer of the postnatal dentine), downward and 
inward at about a 30°-40° angle to the margin of the 
pulp cavity (for examples see Perrin and Myrick 
1980: fig. 2; Hui 1978: figs. 1, 2, 3). For consistency, 
measurements of GLG and label thickness, taken 
perpendicular to the long axis of the teeth of the 
Hawaiian spinner dolphins, were made along tran- 
sects at a similar position and angle (Figs. 1A, B). 
However, a GLG or label may vary in thickness in 
localized regions of the dentine and may not be the 
same on both sides of a tooth because of tooth asym- 
metry. For these reasons, measurements were made 
on the most symmetrical side of a tooth and in regions 
where GLGs and labels were clearest and least vari- 
able in thickness; departing slightly from a uniform 
angle of transect. GLGs in the dentine of the corres- 
ponding D/S thin sections were defined and counted 
with the aid of GLG-thickness measurements 
obtained from the untreated section. 

Retrospective Calibration of 
Cemental GLGs 

Because fewer labels were observed in the cemen- 
tum than in the dentine of the same untreated thin 
section, it was assumed that those visible represent- 
ed condensed forms of only the brightest, thickest, or 
closely spaced groups of dentinal labels. This has 
been verified in bottlenose dolphins (Myrick 1980b) 
and recently in the present sample of Hawaiian spin- 
ners by observations that bright dentinal labels at the 
tooth base are continuous with cemental labels. 
Hence, cemental labels were lettered to correspond 
to the brightest dentinal labels, and the cemental 
layers between labels were calibrated using the time 
segments represented between the dentinal labels. 

The annual GLG pattern was defined as precisely as 
possible using the calibrated segments of the tissue, 
and the cemental GLG definition was tested by com- 
paring the dentinal GLG count with the cemental 
GLG count in untreated thin sections. In D/S thin 



209 



FISHERY BULLETIN: VOL. 82, NO. 1 



Enamel 



Enamel 



Prenatal dentine 



Neonatal line 




Approximate region 
and angle of GLG and 
label measurements 



Cementum 



ABC P Pulp cavity 
Tetracycline margin 
labels , 



654 3 2 1 

i i 

Dentinal 
GLGs 



VIEW IN UV LIGHT 



VIEW IN PLAIN LIGHT 



B 



Periods of treatment 



Dentinal labels 



ig78 197g 



1980 

.i i i i i i i i i i I 




Death 



Pulp cavity 



FIGURE 1 .—Line drawing of hypothetical dolphin tooth in thin section showing appearance of TCL labels, A, B, C, D, 
under ultraviolet light (1A, left-hand side) and dentinal growth layer group (GLG) layering patterns under plain 
transmitted light (IB, right-hand side), and standard positions in tooth where label and GLG thickness are measured. 
1 C illustrates method of identifying labels in tooth section with TCL treatment dates by comparing relative thickness 
and spacing of labels with treatment periods. 



210 



MYRICK ET AL.: DENTAL LAYERS IN HAWAIIAN SPINNER DOLPHINS 



sections, cemental GLGs were defined indirectly 
by comparing them with the pattern and number 
of cemental GLGs determined in untreated 
sections. 

Direct Monitoring 

Calculation of depositional rates and calibration 
and definition of GLGs in dentine and cementum 
were achieved by comparing tooth specimens con- 
taining successively introduced labels and/or 
additional tissue accumulated over the 1-yr period of 
monitoring. To make determinations, for cases in 
which labels were not distinct or not successfully pro- 
duced, the additional tissue was measured from 
structural landmarks or labels in the extracted series 
of thin sections. 

RESULTS 

Dentinal labels. — The untreated thin sections for 
all seven specimens contained multiple labels. Most 
attempts to match labels with treatments were suc- 
cessful (Figs. 2-6). However, in four specimens more 
labels occurred than could be accounted for from 
clinical records. In the only captive-born specimen, 
WFP 670, numerous TCL labels were observed (Fig. 
7A, B), but only three were found to have been caused 
by intentional therapeutic treatments (Fig. 7D, 
labels C, F, and G). Labels A and B apparently were a 
result of TCL impaired to the then-calf through the 
milk of its mother, who was treated with the drug for 
two periods while the calf nursed. The other labels 
appear to have resulted from frequent ingestion of 
stolen TCL-dosed smelt intended for other dolphins 
being treated at various times while sharing a com- 
mon tank with this animal. 

No treatment was recorded for label A found in the 
dentine of dolphin carcass WFP 669 (Fig. 4A, C) and 
live dolphin ACM 104 (Fig. 6A, C). Judging from the 
relative positions of the "A" labels to the other labels 
for which matches were found with recorded treat- 
ments, "A" labels were introduced into both 
specimens at or about their respective dates of cap- 
ture. It is a fairly common practice in commercial 
aquaria to give medication (often tetracycline) to 
newly captured dolphins recovering from stress of 
capture and adjusting to the captive environment 8 . 

Labels B and G in the dentine of dolphin carcass 
WFP 671 could not be identified from clinical 
records (Fig. 5A, C), although the numerous other 



"William A. Walker, Los Angeles County Museum of Natural His- 
tory, Los Angeles, Calif., pers. commun. 1982. 



labels match well in relative thickness and spacing 
with the treatment dates for this specimen. 

In teeth of live dolphin ACM 103 the labels were 
indistinct. The presence of TCL, introduced 
clinically during three periods of treatment over 2 yr 
and experimentally at 3-mo intervals in 1980, was 
indicated only by several areas of hazy fluorescence 
in the dentine near the pulp cavity. 

Dentinal GLG pattern. — The use of plastic overlays 
of key labels enclosing 6-mo or 1 -yr segments of den- 
tine on plain-light photographs of the dentine for 
each specimen permitted repeated calibrations of 
the annual dentinal layering pattern for six of the 
seven specimens (the seventh specimen, ACM 103, 
had no discrete labels). In untreated thin sections, a 
dentinal GLG contained four major components 
deposited in the following sequence: 1) A thin, light 
(GLG-boundary) layer, 2) a thicker dark layer, 3) 
another thin, light (mid-GLG) layer, and 4) a second 
thick, dark layer (Figs. 3A, 4B, 5B, 6B). 

In addition to the four components, many of the 
earliest deposited GLGs had an infrastructure com- 
posed of finer alternating dark and light layers. 
Counts made at 150X under low transmitted light 
showed that each of these annual GLGs contained 13 
pairs of fine layers (Figs. 3A, 4B, 6D, 7C). Where 
layers were sufficiently distinct to be counted be- 
tween labels (e.g., between label B and M, Fig. 4A, B), 
counts indicated that each pair |"LML," (lunar 
monthly layer) Myrick 1980b] represented about 1 
lunar month. The full complement of LMLs was vis- 
ible throughout the dentine in the captive-born 
specimen, WFP 670, i.e., 13 LMLs in each of the first 
three complete GLGs and 9 in the incomplete fourth 
GLG (Fig. 7C). In specimen ACM 103, 13 LMLs were 
observed in the first 12 of the 14.5 GLGs present (Fig. 
8). But in other specimens, LMLs were clear enough 
to be counted only in the first five or six GLGs. 

In D/S thin sections, the annual GLG pattern con- 
sisted of two lightly stained and two darkly stained 
layers. The thin, light, GLG-boundary layers and 
mid-GLG layers in untreated thin sections corres- 
ponded to the lightly stained layers in D/S thin- 
sections (Fig. 9A, B). LMLs were indistinct in almost 
all GLGs in D/S preparations. 

Age-specific GLG thickness.— Table 1, showing 
dentinal GLG thickness measurements made from 
the most symmetrical side of the tooth of each of the 
seven dolphins, indicates that for each animal a GLG 
of a specific thickness was produced that appears 
to be related to the year of life in which the GLG 
was formed, i.e., an age-specific GLG thickness. 



211 



FISHERY BULLETIN: VOL. 82, NO. 1 





<c 




e 


Tt 


3. 


tv. 


o 


CD 


o 


T— 



CO _ 




o 



o •£ M 

-C "S -= 

g — — 
co 

c 



S 



-a £ 



3 
O 



ft 

X 

<v 
cu 



7 ° 



G 

0) 

-a 

c 

-3 
c 
o 
ft 



£ 5 






to 7 



_co - E 

C T3 

■S T3 

O o 

-C 3 

to -g 

o 2 

00 -w 

OS S 



1^ 

CO CO 



c 

'Is 

o 

cc 
u 

CO 

O 



T3 

cu 

CJ 
CO 

t< 



£5? 

■s s 

O ^H 

en ^ 

<D 

CM _ 

c -a 
•- o 



o a 



o 



e 

IS 






C 
P 



o 

§e2 



co a» 



D3 



^ o 



c 



— T3 

CO C 
«J CO 



> t> 






o 

ft 



5 s 

cu ■•-> 

** CM 

— Oi 

"3 "- 1 



k £ 



-a cu 

O TO 

O .> 

t- ^_> 

CU _TO 

I E 

H o 



o i_, 



_c 


-r 


CO 


cu 


— 


Cu 


-a 


Eh 


o 


U 



P £ 5 



212 



MYRICK ET AL.: DENTAL LAYERS IN HAWAIIAN SPINNER DOLPHINS 




100 pm 



I I > 



A B 



* 



* 




Treatment 



Labels 



1976 1977 

i i i I i i n li i_i i i 



1979 



1980 

I "in ii Hi 




lllllllllll 



T"c"dTT g\ IIIIIIIIIMI 

b l d t (■ b-\p C of second too(h 



FIGURE 3. — Labeled tooth taken from dolphin carcass WFP 606. A. Untreated thin section in plain light showing 
about eight annual GLGs in dentine (separated by arrows). GLGs divided approximately in half by thin, light mid- 
GLG layers (heavy dark marks). GLGs 6, 7, and 8 were interpreted from positions of tetracycline labels (lettered). 
Finer dark layers represent lunar monthly layers ( 1 50X). B. Dentine labels in UV light ( 1 50X) . C. Chart showing 
dates labels were introduced. 



213 



FISHERY BULLETIN: VOL. 82, NO. 1 





O) - 



IN _ 



O) _ 




■- <k 2 

« c .S 

CO ;I2 -^ 

0) ^ C 

■§ = | 



.2 2 
> -J 



c 

o 



CQ (D 

.S 

o is 



a. « 

* a 



> 



>,-2 

to O 

-§■ >> 

ft ^ 



£ o 

a a 

° a; 



co 
<u 



o 

he 
C 

o 

J3 

-C en 
w o 

°& 

to C 

cfl 0) 

c E 



o 
| E 

bo »> 






o> 



a> -C i3 

ft — <*- 

. CD O 

.2 c 

O js 



c 



> 

CO 



ffl CO 

0> — 






05 

tr 
to 

PL, 



CO 



a 



§1E 



b5 



■n 

■> § 

en O 

co o 

CO ^ 

CO &c 

.s >> 

o K 

-a 



T3 

5 >= 
p, 

T3 



cc 



0> Cut 



a> 

T3 

u 
o 

.o 

CO 

o 
o 

"3 

3 

c 
c 

CO 

a 

'% 
o 

-C 
to 

< 

CO 



ft CO 

ft 



, ° 

CO >> 

en -O 

O T3 

J .2 

o 5 

C o> 

"^ XI 

CO — 

CU tO 

£ "co 



214 



MYRICK ET AL.: DENTAL LAYERS IN HAWAIIAN SPINNER DOLPHINS 




B 



H 



h 
200pm 




1974 

I I i I l l l l 



Treatments 



Labels 



1975 



1976 1977 

I I I I i I I n I ii i I 



1979 1980 

l 1 1 I I I ii l 1 1 i I I i i 




A^B C 
No record 



E F ^G H 

No record 



FIGURE 5.— Labeled tooth of dolphin carcass WFP 671. A. Untreated thin section in UV light showing TCL labels 
(39X). B. Thin section in plain light showing almost eight complete GLGs as interpreted from labels. Light GLG boun- 
dary layers appear to have been deposited in or about March (39X). C. Chart showing match between labels and treat- 
ments. 



TABLE 1. — Mean age-specific thicknesses (fim) of completed dentinal growth layer groups (GLGs) in 
teeth of seven Hawaiian spinner dolphins, Stenella longirostris. Values are averages of at least three 
measurements per specimen, taken perpendicular to the long axis of the tooth in a stairstep fashion 
downward and inward from the base of the neonatal line to the pulp-cavity wall. 

Specimen GLG number 

no 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 

WFP 670 240 240 180 — — — — — — — — — — — — 

WFP669 1 240 240 172 5 150 1275 1275 92.5 85 65 57.5 — — — — — 

WFP 671 240 240 165 140 120 95 80 ______ __ 

ACM 104 240 230 170 140 130 90 90 70 70 60 55 — — — — 

ACM 103 240 240 180 160 110 80 80 70 60 60 60 65 55 40 40 

ACM 106 240 240 1 50 1 30 1 1 90 90 60 60 60 55 — — — — 

WFP 606 240 240 1 80 1 50 1 20 90 90 — — — — — — — — 

/V 7776666444311 11 

* — 2386 1710 1450 1196 95 4 87 713 613 58 1 55 — — — — 

SD 3.8 110 10 5 8 4 16 5 5 6 10 3 2 5 2.4 5.0 — — — — 

SE 1.4 4 2 4 3 34 6 7 2.3 5.2 1.3 12 2.9 — — — — 

'Mean values ol measurements in untreated and D/S (decalcified and haematoxylin-stained) sections 



215 



FISHERY BULLETIN: VOL. 82, NO. 1 




O 1976 1977 

INlMll.lllllllll 



Treatments 



Labels 



1978 



1979 



1980 




PC of 3rd Tooth 



100 pm 






I'll 




FIGURE 6.— Tooth of live dolphin ACM 104 extracted 2 February 1981. A. Untreated thin section in UV light showing 
location of TCL labels (150X). B. Same section as in 6A in plain light showing position of key labels bracketing last 4 yr of 
deposition. Light GLG boundary layers appear to have been deposited in or about August. C. Chart showing match of 
labels and treatments. D. Thin section showing 1 1 complete annual GLGs (separated by dark marks) as interpreted 
from labels (39X). 



216 



MYRICK ET AL.: DENTAL LAYERS IN HAWAIIAN SPINNER DOLPHINS 




200|jm 




N 



100pm 




D 

Treatments 



1975 1976 

I i I i I I i I II i I ll I I I I 



1977 



1978 



labels 




I Ml I M I I I I I L UU J I I I I I I I m^J I I III 



1979 

l i I I I I I 




Treatments to otners 
in same tank 



\ ABC 

v Neonatal line 



D E 



FG\. 



PC 



FIGURE 7.— Teeth from dolphin carcass WFP 670 (captive-born animal). A. Untreated thin section showing TCL labels in den- 
tine. Labels A-B apparently represent TCL imparted to this animal through its mother's milk (UV, 39X). N =neonatal line; PC = 
pulp cavity margin. B. Portion as shown in 1A showing numerous labels from TCL-dosed smelt stolen from other dolphins 
occupying the same tank. Labels F and G represent direct treatments administered shortly before death (UV, 150X). C. Thin- 
sectioned tooth showing three entire and one partial GLGs (indicated by heavy dark marks) in the postnatal dentine as inter- 
preted from TCL labels. LMLs are indicated by fine dark markers (plain transmitted light, 150>^. D. Chart showing dates of 
direct and presumed incidental introduction of TCL and corresponding labels identified in the dentine by relative label position 
and thickness. 



217 



FISHERY BULLETIN: VOL. 82, NO. 1 







200jum 



B 





/ 



/*>..J 






/K'&~ 



^ 



>• 



ft 



50/im 



Figure 8.— Untreated tooth of live dolphin ACM 103 extracted 25 January 1980. A. About 
14'/2GLGs indicated (heavy dark marks) (39X). B. GLGs 8-14% showing thin, light boundary layers 
with dark margins. Thirteen LMLs indicated in each GLG 8-12 are particularly well developed 
(150X). 



Comparisons of age-specific GLG thickness among 
the specimens suggest that the animals deposited a 
GLG of similar thickness in the same year of life. In 
the first and second year, 240 /xm thick GLGs were 
deposited. In the third, fourth, fifth, sixth, and 
seventh years, thickness of GLGs averaged 171, 145, 

218 



119, 95, and 87 ju.m respectively. From the 8th to the 
1 1th year, GLGs were between 71 and 55 jum thick. 
The data in Table 1 represent averages of at least 
three measurements per GLG per specimen. 

Cemental labels.— Relatively few TCL labels were 



MYRICK ET AL.: DENTAL LAYERS IN HAWAIIAN SPINNER DOLPHINS 




B 




FIGURE 9.— Comparison of GLG patterns in teeth 
from dolphin carcass WFP 606 prepared by two 
methods: A. Untreated thin section (39X). B. De- 
calcified and stained thin section (39X). 



Pi 



76 



r 



found in the cementum compared with those in the 
dentine of the specimens. In the captive-born 
specimen, WFP 670, with about 25 dentinal labels, 
the cementum contained only three labels. In 
specimen WFP 669, only four cemental labels were 
observed (Fig. 10A) compared with 30 dentinal 
labels (Fig. 4A). The cementum in the other 
specimens had either zero or 1 label, despite the 
numerous dentinal labels observed for each. 

Cemental GLG pattern. — In untreated thin sec- 
tions, a cemental GLG consisted of a dark layer and a 
light layer (Fig. 10B). In D/S sections it was com- 
posed of a dark-stained layer, corresponding to the 
dark layer in untreated sections, and a lightly stained 
layer (Fig. 1 1). In both types of preparations, the dark 



layers contained larger concentrations of ceraen- 
tocytes than did the light layers. 

Calibration of cemental GLGs. — Calibrations of 
cemental GLGs with those in the dentine were car- 
ried out using the assumption that cementum is a less 
sensitive recording structure than dentine (Klevezal' 
1980) and that labels occurring in the cementum cor- 
responded only to the brightest and thickest labels or 
label groups in the dentine. Thus, for example, the 
four labels detected in the cementum of specimen 
WFP 669 (Fig. 10A) were flagged with the same let- 
ters used to identify multiple label concentrations in 
the dentine (Fig. 4A). 

In some cases, such as in WFP 669, plastic overlays 
were used to determine that a cemental GLG rep- 



219 



FISHERY BULLETIN: VOL. 82, NO. 1 




B 




FIGURE 10.— Tooth cementum of dolphin carcass WFP 669. A. TCL labels interpreted as corresponding to 
lettered dentinal labels (150X). B. Positions of TCL labels (arrows) in layered cement. About 10 GLGs are 
indicated (150X). 



resented the same amount of time as a dentinal GLG, 
i.e., 1 yr (e.g., Fig. 10B). In other cases, where labels 
were absent or where only one label occurred, cali- 
bration of cemental GLGs with dentinal GLGs was 
made indirectly by comparing GLG counts from both 
tissues. This method usually demonstrated a one-to- 
one relationship of GLGs in dentine and cementum, 
but in a few regions of the cementum of the captive- 
born specimen, WFP 670, there were twice as many 
GLGs as in the dentine (Fig. 11), indicating that a 
GLG may have been deposited twice a year in the 



cementum. In expanded regions of the cementum in 
another specimen (ACM 104; see Table 2), the 
cemental count was equal to the dentinal count; but 
in thinner regions the cemental count was only half 
that of the dentine, suggesting a cemental GLG being 
deposited every 2 yr in some cases. 

Direct monitoring. — The results of examinations of 
thin sections of the series of three teeth from each of 
three live animals, taken at the beginning, at mid- 
point, and at the end of a 1-yr monitored period are 



220 



MYRICK ET AL.: DENTAL LAYERS IN HAWAIIAN SPINNER DOLPHINS 







FIGURE 11.— Cementum of dolphin carcass WFP 670 tooth in de- 
calcified and stained thin section. The number of dark layers is eight, 
about double the age in years of this captive-born specimen (150X, 
plain transmitted light). 



presented in Table 2. Although distinct labels were 
not always successfully introduced, dentinal and 
cemental GLGs continued to be accumulated at a 
uniform rate of one per year. A comparison of 
accumulated dentine and labels in the first two 
extracted teeth of specimen ACM 106 (Fig. 12) 
showed two experimental treatments and one 
(unscheduled) clinical treatment accounted for in the 
second tooth (Fig. 2A-C). Specimen ACM 103, in 
which premonitor labels were indistinct, showed no 
experimental labels but clearly showed continued 
accumulation of dentine, the third extracted tooth 
having added about one GLG over the 1-yr period. 
No experimentally introduced labels were observed 
in the (less sensitive) cementum in any teeth of the 
three animals, but cemental deposition of about one 
complete GLG occurred in each animal for the period. 

Seasons of birth. — By determining the dates of key 
dentinal labels introduced at or near the thin, light 
component layers of GLGs and by noting the approx- 
imate time of formulation of component layers in the 
teeth extracted during the monitor period, it was 
found that GLG-boundary and mid-GLG layers were 
formed at about 6 mo intervals. In five specimens, 
GLG-boundary layers were deposited in or about 
August and the mid-GLG layers were deposited in or 
about March. In the two other specimens the timing 
was reversed, i.e., GLG-boundary layers formed in 
March, mid-GLG layers in or near August, Proceed- 
ing on the assumption that the timing of layer forma- 
tion (determined from the labeled or monitored 



Table 2.— Results of examinations of teeth extracted from three live Hawaiian spinner dolphins, 
Stenella longirostris , over a 1-yr period monitoring accumulation of layers and labels. GLGs = 
growth layer groups. 



Specimen 
no. and tooth 


Date of tooth 


Date label 


Dentine 




Cementum 




extraction 


introduced 


Additional labels 


No. GLGs 


Additional labels No. GLGs 


ACM 103 














First 


25 Jan. 1980 


25 Jan 1980 
30 Apr 1980 


— 


14 5 


" 


14.5 


Second 


30 Julv 1980 


30 July 1980 
30 Nov 1980 


indistinct 


150 


None 


150 


Third 


2 Feb. 1981 


— 


indistinct 


15 5 


None 


15.5 


ACM 106 














First 


19 Mar. 1980 


19 Mar. 1980 
11-28 Apr 1980' 
5 June 1980 




103 




10 


Second 


30 July 1980 


30 July 1980 
30 Nov. 1 980 


3 


10.7 


None 


10+ 


Third 


2 Feb 1981 


— 


indistinct 


112 


None 


11 


ACM 104 














First 


25 Jan 1980 


25 Jan 1980 
30 Apr 1980 


— 


< 2 > 




( 2 ) 


Second 


30 July 1980 


30 July 1980 
30 Nov. 1980 


2 


10.7 


None 


3 5/10 


Third 


2 Feb. 1981 


— 


4 


11.3 


None 


3 6/11 



Unscheduled medical treatment 1 8 d in duration. 
Not examined because of poor preparation of section. 
3 Cementum showed a number of GLGs equal to that of the dentine as well as half that of the dentine. 



221 



FISHERY BULLETIN: VOL. 82. NO. 1 




222 



MYRICK ET AL.: DENTAL LAYERS IN HAWAIIAN SPINNER DOLPHINS 



regions of the dentine) was uniform throughout all of 
the dentine for a given specimen, GLG-boundary and 
mid-GLG layers were counted in reverse order of 
deposition up to the first boundary layer, the neona- 
tal line, to estimate month and year of birth. 

Table 3 summarizes the month- and year-of-birth 
estimates made from boundary-layer counts in six 
specimens and birth dates taken from park records of 
two captive-born specimens, WFP 670 and the calf of 
ACM 104. Six were born in late summer/early 
autumn and two in March. 

TABLE 3.— Estimated birthdates of eight cap- 
tive Hawaiian spinner dolphins, Stenella long- 
imstris.. 



Specimen 


Month and year 


no. 


of birth 


ACM 103 


August 1964 


ACM 106 


August 1969 


WFP 669 


August 1969 


ACM 104 


September 1969 


WFP 606 


March 1972 


WFP 671 


March 1973 


WFP 670' 


8 September 1975 


Calf of ACM 104 2 


21 July 1977 



'Born in captivity. 

2 Born in captivity, survived 3 d 



DISCUSSION 
Age-Specific GLG Thickness 

Dentinal GLG thickness appears to be age-specific 
for the Hawaiian spinner dolphin teeth examined. 
There was little variability from tooth to tooth or 
from animal to animal in the sequence of GLG thick- 
ness through the 11th GLG, despite deposition of a 
specific GLG in some specimens while still in the wild 
and in other specimens during their captive lives. 
This suggests that, to some extent at least, the 
amount of dentine deposited by animals at a given 
age may be predetermined and that animals of a 
given stock, species, or higher common phylogenetic 
affinity may follow the same or similar pattern of age- 
specific GLG deposition unaffected by environ- 
ment. 

Used in conjunction with the GLG component-layer 
pattern, the regularity in thickness of age-specific 
GLGs may be useful as an aid in locating GLG boun- 
daries and counting GLGs in teeth of wild Hawaiian 
spinner dolphins and dolphins of related species in 
which GLG thickness and component-layer patterns 
are found to be similar. When measurements are 
taken at standard positions in the teeth of such 
dolphins, one may make fairly rapid age estimates 
without having to examine each GLG in detail (see 
Myrick et al. 1983). 



Lunar Monthly Layers (LMLs) 

Laws ( 1 962) was the first to suggest that the system 
of fine layers within dentinal GLGs of pinniped teeth 
corresponded to lunar monthly cycles. Putative 
LMLs have been reported in dentine of dugongs 
(Kasuya and Nishiwaki 1978; Marsh 1980), in den- 
tine of beaked whales ("short cycles," Kasuya 1977; 
"accessory layers," Perrin and Myrick 1980:3, 5), in 
fossil dolphin teeth (Myrick 1979), and in the man- 
dibular bone (Myrick 1980b) and dentine of modern 
dolphins (Myrick 1980b; Hohn 1980a, b). Hui (1978) 
reported finding no relationship between the fine 
layers that he counted in a tooth from a known-age 
bottlenose dolphin and its age in lunar months; but 
with no prior knowledge of its age, Myrick (1980b) 
made dentinal LML counts in the same specimen 
that closely agreed with its known age. 

The present study has furnished verification that 
LMLs are deposited with lunar-monthly regularity in 
the animals studied. In the 3.7-yr-old captive-born 
spinner dolphin (WFP 670), 13 LMLs were counted 
in each of the three complete annual dentinal GLGs 
and 9 were counted in the partial fourth GLG. Where 
LMLs were visible between TCL labels in the den- 
tine in this and other specimens, they were found to 
correspond consistently in number to the time in 
months represented between labeling dates. 

Where LMLs could be seen clearly, no departure 
from the 13 LML/GLG pattern was detected in the 
teeth used in the present study. Variability has been 
reported in studies of other marine mammals. Marsh 
(1980:197) found only "about 12 [LMLs] per GLG" 
in the dentine of the deciduous incisor of a dugong. 
Ten to 15 LMLs/GLG were observed in dugong tusks 
by Kasuya and Nishiwaki (1978). Kasuya (1977) 
found between 11 and 13.4 LMLs ("short cycles")/ 
GLG in teeth of Baird's beaked whales, Berardius 
bairdii. Hohn ( 1 980b) counted 10-13 LMLs/dentinal 
GLG in Atlantic bottlenose dolphin teeth. Pre- 
sumably, LML variability will be found to occur also 
in Hawaiian spinner dolphins when larger samples 
are examined. 

Relationship of Cemental GLGs to 
Dentinal GLGs 

None of the teeth of the studied specimens had 
reached the stage of pulp-cavity occlusion or dentinal 
irregularity that necessitated age estimation solely 
from cemental GLG counts (Kasuya 1976; Myrick et 
al. 1983). Although the pulp cavities were small in 
some specimens and some later-administered TCL 

223 



FISHERY BULLETIN: VOL. 82, NO. 1 



failed to produce distinct labels, none showed 
evidence of cessation of dentine deposition. 

All cemental GLG counts corresponded in number 
to dentinal annual GLG counts except in the case of 
specimen WFP 670, where some regions of the 
cementum showed double the number, and in 
specimen ACM 104, where in some places the 
cemental count was half that of the dentine. The find- 
ing that in some cases cemental GLGs may form at 
half or twice the rate of dentinal GLG deposition 
points up the problem of using cemental GLGs to 
estimate ages without reference to the dentine 
(Myricket al. 1983). 

Evidence for an Internal Clock 

In the dentine of the animals studied, a thin GLG 
boundary layer, beginning with the neonatal line, was 
formed in the month of birth and on anniversaries of 
the month of birth. Mid-GLG layers were formed 
about 6 mo after formation of boundary layers. 
Where LMLs could be calibrated, one was found to 
form about every (lunar) month with high uniformity 
in relative spacing. Such a cycle of deposition is 
indicative of an internal clock, or clocks. The pattern 
commences at birth and apparently is reset with solar 
and/or lunar regularity without perceptible altera- 
tion by fluctuation in the dolphins' natural or captive 
environment or in calendric season of birth. That it 
may not be a totally free-running system, i.e., not 
without external cues, is suggested by the precisely 
synchorized deposition of the fine and coarse pat- 
terns of the dentine repeated over many years. 

Age at Sexual Maturity 

Perrin et al. (1977) indicated that sexual maturity 
may be reached in females of Stenella longirostris at 
an average 5.5 yr (range of 5-9 yr) and the average 
period of gestation may be about 1 1 mo. From the 
study of dentinal GLGs and TCL labels of specimen 
ACM 104, it was possible to determine that this 
animal was about 8-yr-old when she gave birth to her 
calf. Assuming an 11 -mo gestation, we estimate that 
she would have been 7-yr-old when she conceived. It 
is not known whether the pregnancy resulted from 
fertilization at her first or subsequent ovulations. 
ACM 104 remains alive. This precludes examination 
of her ovaries for ovulation scars. 

Reproductive Seasonality 

Based on the birth records of specimen WFP 670 
and the calf of ACM 104 and the deductions made 



from dentinal layers, six animals were born in late 
summer/early fall, and two were born in March. Since 
all animals in the study represented the same popula- 
tion off Kona, Hawaii, the early-spring and early-fall 
birth patterns might indicate a corresponding two- 
cycle pattern of reproductive peaks for the wild pop- 
ulation generally. Such a seasonal pattern has been 
suggested by Norris and Dohl (1980, fig. 16), but 
Wells (in press), who has studied the population in 
considerable detail, concluded that the breeding 
season occurs from spring to fall, with most births in 
the fall. Our sample was too small to verify Wells' 
findings. 

Tetracycline Exposure to the Calf 
Through the Milk 

The first two labels found in the dentine of 
specimen WFP 670, the captive-born animal, were 
interpreted as having been introduced through milk 
received by the calf while the mother was being 
treated with TCL. This recommends a possible prac- 
tical application in indirectly treating newborns in ill 
health. Excessive handling of such animals fre- 
quently results in a worsening of their condition, 
making the treatment more dangerous than the 
malady. (Nursing calves not on solid food cannot be 
treated with TCL-dosed fish and must be force-fed 
or injected with drugs.) Separating the young calf 
from its mother may produce additional com- 
plications. 9 If treatments for the calf could be 
administered through the milk by treating the mother 
with TCL-dosed food, it seems likely that most of the 
problem could be minimized. The question invites 
further study. 

ACKNOWLEDGMENTS 

We thank D. G. Chapman, R. L. Brownell, Jr., D. B. 
Siniff, A. Wild, W. F. Perrin, F. Hester, A. Dizon, and 
J. Barlow for their critical reviews of the manuscript. 
K. Raymond and R. Allen prepared the figures. M. 
DeWitt typed the manuscript. 

LITERATURE CITED 

Best, P. B 

1976. Tetracycline marking and the rate of growth layer for- 
mation in the teeth of a dolphin (Lagenorhynchus 
obscurus). S. Afr. J. Sci. 72:216-218. 
Gurevich, V. S., B. S. Stewart, and L. H. Cornell. 

1980. The use of tetracycline in age determination of com- 



9 L. H. Cornell, Sea World, Inc., San Diego, Calif., pers. commun. 
1980. 



224 



MYRICK ET AL.: DENTAL LAYERS IN HAWAIIAN SPINNER DOLPHINS 



mon dolphins, Delphinus delphis . In W.F.PerrinandA. C. 
Myrick, Jr. (editors). Age determination of toothed whales 
and sirenians, p. 165-169. Rep. Int. Whaling Comm., 
Spec. Issue 3. 
HOHN, A. 

1 980a. Age determination and age related factors in the teeth 
of western North Atlantic bottlenose dolphins. Sci. Rep. 
Whales Res. Inst. Tokyo 32:39-66. 

1980b. Analysis of growth layers in teeth of Tursiops trun- 
catus, using light microscopy, microradiography, and 
SEM. In W. F. Perrin and A. C. Myrick, Jr. (editors), Age 
determination of toothed whales and sirenians, p. 155- 
160. Rep. Int. Whaling Comm. Spec. Issue 3. 
HUI, C. A. 

1978. Reliability of using dentin layers for age determination 
in Tursiops truncatus. U.S. Marine Mammal Comm. Rep. 
No. MMC-77/09 (Nat. Tech. Inf. Serv. PB288444), 25 
P. o 

JONSGARD, A 

1969. Age determination of marine mammals. In H. T. 

Anderson (editor), The biology of marine mammals, p. 1- 

30. Acad. Press, N.Y. 
Kasuya, T. 

1976. Reconsideration of life history parameters of the spot- 
ted and striped dolphins based on cemental layers. Sci. 
Rep. Whales Res. Inst. 28:73-106. 

1977. Age determination and growth of Baird's beaked whale 
with a comment on the fetal growth rate. Sci. Rep. Whales 
Res. Inst. Tokyo 29:1-20. 

Kasuya, T., and M. Nishiwaki. 

1978. On the age characteristics and anatomy of the tusk of 
Dugong dugon. Sci. Rep. Whales Res. Inst. Tokyo 30:301- 
311. 

Klevezal, G. A. 

1980. Layers in the hard tissues of mammals as a record of 

growth rhythms of individuals. In W. F. Perrin and A. C. 

Myrick, Jr. (editors) , Age determination of toothed whales 

and sirenians, p. 89-94. Rep. Int. Whaling Comm. Spec. 

Issue 3. 
Klevezal', G. A., and S. E. Kleinenberg. 

1967. Opredelenie vozrasta mlekopitayushchikh po sloityn 

strukturam zubov i kosti (Age determination of mammals 

by layered structure in teeth and bone). Izdatel'stvo 

Nauka, Moscow, 144 p. 
Laws, R. M. 

1952. A new method of age determination in mammals with 

special reference to the elephant seal (Mirounga leonina 

Linn.). Falkland Is. Dep. Surv. Sci. Rep. 2, 11 p. 
1962. Age determination of pinnipeds with special reference 

to growth layers in the teeth. Saugetier. Mitt. 27(3):129- 

146. 
Marsh, H. 

1980. Age determination of the dugong {Dugong dugon 

(Muller)) in northern Australia. In W. F. Perrin and A. C. 

Myrick, Jr. (editors), Age determination of toothed whales 

and sirenians, p. 181-201. Rep. Int. Whaling Comm. 

Spec. Issue 3. 
Myrick, A. C, Jr. 

1979. Variation, taphonomy, and adaptation of the Rhabdos- 
teidae (=Eurhinodelphidae) (Odontoceti, mammalia) 
from the Calvert formation of Maryland and Vir- 
ginia. Ph.D. Thesis, Univ. California Los Ang., 41 1 p. 



1 980a. Some approaches to calibration of age in odontocetes 
using layered hard tissues. In W. F. Perrin and A. C. 
Myrick, Jr. (editors) , Age determination of toothed whales 
and sirenians, p. 95-97. Rep. Int. Whaling Comm. Spec. 
Issue 3. 
1980b. Examination of layered tissues of odontocetes for age 
determination using polarized light microscopy. In W. F. 
Perrin and A. C. Myrick, Jr. (editors), Age determination 
of toothed whales and sirenians, p. 105-112. Rep. Int. 
Whaling Comm. Spec. Issue 3. 
Myrick, A. C, Jr., A. A. Hohn, P. A. Sloan, M. Kimura, and D. 
D. Stanley. 

1983. Estimating age of spotted and spinner dolphins 
(Stenella attenuate and Stenella longirostris) from 
teeth. NOAA-TM-NMFS-SWFC-30, 17 p. 
Myrick, A. C, Jr., E. W. Shallenberger, and I. Kang. 

In press. Records used in the calibration of dental layers in 
seven captive Hawaiian spinner dolphins, Stenella 
longirostris. NOAA-TM-NMFS-SWFC. 
Nielsen, H. G. 

1972. Age determination of the harbour porpoise Phocoena 
phocoena (L.) (Cetacea). Vidensk. Medd. Dan. Naturhist. 
Foren. 135:61-84. 

Nishiwaki, M., andT. Yagi. 

1953. On the age and the growth of teeth in a dolphin, (Pro- 
delphinus caeruleo-albus). (I). Sci. Rep. Whales Res. Inst. 
Tokyo 8:133-146. 
NORRIS, K. S., AND T. P. DOHL. 

1980. Behavior of the Hawaiian spinner dolphin, Stenella lon- 
girostris. Fish. Bull., U.S. 77:821-849. 
Perrin, W. F., D. B. Holts, and R. B. Miller. 

1977. Growth and reproduction of the eastern spinner 
dolphin, a geographical form of Stenella longirostris in the 
eastern tropical Pacific. Fish. Bull, U.S. 75:725-750. 
Perrin, W. F., and A. C. Myrick, Jr. (editors). 

1980. Age determination of toothed whales and sireni- 
ans. Rep. Int. Whaling Comm. Spec. Issue 3, 229 p. 
Ridgway, S. H., R. F. Green, and J. C. Sweeny. 

1975. Mandibular anesthesia and tooth extraction in the bot- 
tlenose dolphin. J. Wildl. Dis. 11:415-418. 
SCHEFFER, V. B. 

1950. Growth layers on the teeth of Pinnipedia as an indica- 
tion of age. Science (Wash., D.C.) 112:309-311. 

SCHEFFER, V. B., AND A. C. MYRICK, JR. 

1980. A review of studies to 1 970 of growth layers in the teeth 

of marine mammals. In W. F. Perrin and A. C. Myrick, Jr. 

(editors), Age determination of toothed whales and 

sirenians, p. 51-63. Rep. Int. Whaling Comm. Spec. 

Issue 3. 
Sergeant, D. E. 

1959. Age determination in odontocete whales from dentinal 

growth layers. Nor. Hvalfangst-tidende 48:273-288. 
Sergeant, D. E., D. K. Caldwell, and M. C. Caldwell. 

1973. Age, growth, and maturity of bottlenosed dolphin (Tur- 
siops truncatus) from northeast Florida. J. Fish. Res. 
Board Can. 30:1009-1011. 

Wells, R. S. 

In press. Reproductive seasonality and social behavior of 
Hawaiian spinner dolphins, Stenella longirostris. In W. F. 
Perrin and D. DeMaster (editors), Cetacean reproduc- 
tion. Rep. Int. Whaling Comm. Spec. Issue 6. 



225 



REPRODUCTION OF THE BANDED DRUM, LARIMUS FASCIATUS, 

IN NORTH CAROLINA 1 



Steve W. Ross 



ABSTRACT 

The reproductive biology of Larimus fasciatus was examined in coastal North Carolina from September 1975 
through September 1976. Spawning occurred in nearshore waters from April through Sept ember with a peak 
in August. Maturity in fameles was reached by the first year between 120 and 130 mm SL. Generally the 
larger, older fish matured earlier and also continued spawning later in the season than the younger ones. 
Fecundity ranged from 12,750 to 320,819 ova with first spawners preducted to have between 31 ,088 and 
65,038 eggs. Fecundity was best predicted by ovary weights during August. Sex ratios generally favored 
more females. As fish grew the sex ratio changed from predominately males to predominately females. 



The banded drum, Larimus fasciatus Holbrook, 
occurs from Massachusetts to southeastern Florida 
and along the northern Gulf of Mexico from the 
Florida west coast to Mexico. Unlike other drums it 
appears to be largely restricted to nearshore coastal 
waters at all sizes and is rarely collected in estuaries 
or from the outer continental shelf (Gunter 1938; 
Dahlberg 1972; Chao 1978: Powles 1980). Larimus 
fasciatus is a small sciaenid reported by Holbrook 
(1860) to reach 305 mm TL (total length), but it 
seldom grows larger than 220 mm (Chao 1978). Its 
small size, low abundance, and lack of status as a food 
or game fish afford this species little commercial or 
recreational value, although it was reported as a com- 
ponent of the North Carolina (Wolff 1972) and Gulf 
of Mexico (Gutherz et al. 1975) industrial fisheries. 

Published data on life history aspects of L. fasciatus 
are largely lacking. Hildebrand and Cable (1934) 
reported limited information on spawning, growth, 
and juvenile descriptions of North Carolina 
specimens, and Powles (1980) presented data on lar- 
val description, spawning seasons, and areas in the 
South Atlantic Bight. Feeding habits were briefly 
examined by Welsh and Breder (1923) and Chao and 
Musick (1977). Standard and Chittenden (in press) 
have studied banded drum life history off of Texas. 

This study describes the following aspects of L. fas- 
ciatus life history in North Carolina: 1) spawning 
seasonality, 2) age and size at maturity, 3) fecundity, 
and 4) sex ratios. 



METHODS 

Most banded drum were collected in the ocean near 
the mouth of the Cape Fear River, N.C., about 4-6 km 
off Oak Island in depths of 4-14 m (Fig. 1). Bottom 
topography was uniform with sediments of fine sand 
and mud. Hydrographic conditions were heavily 
influenced by discharge from the Cape Fear River 
(Ross 1978). 

This area was sampled weekly from September 
1975 through September 1976, except only monthly 
samples were made during January, June, July, and 
August. Each sample consisted of repetitive (4-12) 
30-min trawls with a 12.4 m semiballoon otter trawl 
of 3.85 cm stretched mesh during daylight hours. 

Additional specimens were collected from Septem- 
ber 1975 through September 1976 during twice month- 
ly, daylight sampling between Beaufort Inlet and 
Cape Lookout, N.C. (Fig. 1), except that there was 
no sampling in December 1975 and only monthly 
sampling in January and February 1976. Repetitive 
trawls were made in this area in a depth range of 9-12 
m over a flat, sand bottom using the aforementioned 
gear and tow times. Specimens were also collected 
near Cape Hatteras (9-17 m depth) in November and 
December 1975 and April 1976 by the North Car- 
olina Division of Marine Fisheries (Fig. 1). 

Larimus fasciatus were preserved in the field in 1 0% 
Formalin 3 and later stored in 40% isopropanol. Total 
length (TL) and standard length (SL) were measured 
to the nearest mm. Body weights (BW) were deter- 
mined to the nearest 0.1 g, and gonads >0.01 g were 



'Adapted from part of a thesis submitted to the Zoology Depart- 
ment, University of North Carolina, in partial fulfillment of the 
requirements for the MA degree. 

2 North Carolina Division of Marine Fisheries, P.O. Box 769. 
Morehead City, NC 28557. 



'Reference to trade names does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 



Manuscript accepted August 1983. 
FISHERY BULLETIN: VOL. 82. NO. 1, 1984. 



227 



FISHERY BULLETIN: VOL. 82, NO. 1 




•i a n U iS rt *y Ca P e 

inlet . r 

Lookout 



©-Larimus fasciatus 
collection sites 



34 



Cape Fear 



FIGURE 1. — Collection sites for Larimus fasciatus in North Carolina. 



blotted dry and weighed (gonad weight (GW)) to the 
nearest 0.01 g. Gonad indices (GI) were calculated 
as follows: 

GI = GW (both) X 100/(BW - GW) 

and were used to determine spawning seasons and 
maturity. 

Fecundity was determined for both maturing 
gonads by relating the number of eggs in a subsample 
to the whole gonad. Each subsample (weighed to the 
nearest 0.001 g), removed from the middle and both 
ends of each alcohol-preserved gonad, represented 
roughly 5% of the total gonad weight. All eggs 
(excluding those <0.01 mm in diameter and atretic 
eggs) in the subsample were counted, and the modal 
ovum diameter was measured to the nearest 0.05 
mm. Total fecundity used in the analysis equaled the 
number of eggs in both gonads combined. 



RESULTS 
Spawning 

Larimus fasciatus spawned from April through Sep- 
tember with peak activity in August as indicated by 
female gonad indices (n = 126, Fig. 2). Male gonad 
indices (n = 53) somewhat mirrored the female pat- 
tern, but the spawning cycle was not clearly illus- 
trated because the testes composed a small per- 
centage of the body weight at any maturity stage in all 
months (Fig. 2). Some running ripe males were ob- 
served in the field from June through August. Since 
the mean gonad index was still high in September 
(Fig. 2), spawning may have continued after Septem- 
ber, although I have no collections to substantiate 
this. 

The large size range of juveniles and the collection 
of young-of-the-year <40 mm SL in all months ex- 



228 



ROSS: REPRODUCTION OF BANDED DRUM 



UJ 
Q 



4.5 
4. OH 
3.5 
3.0 



§2.5 

z 

O 2.0 
C3 

-Z- 1.5 
< 

UJ 

5 1.0 

0.5 




7 




- P" 
N 



F M A 

MONTH 



M 



FIGURE 2.— Monthly mean gonad index of male and female banded drum from October 
1975 to September 1976 in North Carolina, including sample size and ±1 standard 
error of the mean. 



cept December 1975 and January, April, May, and 
June 1976 (Fig. 3) support an extended late spring 
through early fall spawning season. Major young-of- 
the-year (1976 year class) recruitment, evidently 
from Spring spawning, first appeared in July 1976 
and continued through September 1976. Young-of- 
the-year from the 1975 year class were evident from 
September 1975 through November 1975 and ap- 
peared again in February 1976 (Fig. 3). This young- 
of-the-year recruitment over a long period with a lack 
of bimodal length frequencies indicated sustained 
spawning effort. Other collections in and near the 
lower Cape Fear River of Larimus fasciatus <40 mm 
SL in January, February, April, June, July, Septem- 
ber, November, and December also indicated ex- 
tended spawning (K. A. MacPherson 4 ). 

The majority of the reproductively active adults 
were collected near the Cape Lookout area (Fig. 1), 
especially during August and September where bot- 
tom water temperature averaged 27° (August) to 
20°C (September). A high percentage (48.9-100%) of 
the total number of females collected in the Cape 
Lookout area exhibited maturing or ripe gonads 
while corresponding percentages from Cape Fear 



4 K. A. MacPherson, biologist, Carolina Power and Light Company, 
Brunswick Biological Laboratory, P.O. Box 10429, Southport, N.C. 
28461, pers. commun. 1977. 



FIGURE 3. — Length frequencies of Larimus fasciatus collected in 
North Carolina from September 1975 through September 1976. 



SEP 1975 
n=118 




25 55 85 115 145 175 

STANDARD LENGTH(mm) 



229 



FISHERY BULLETIN: VOL. 82, NO. 1 



were low (0-8.1%) (Table 1). Although sampling 
effort in the Cape Fear area was half of that near Cape 
Lookout from June through August, more female 
banded drum were collected near Cape Fear; 
however, the percent of females with large gonads 
was much greater in the Cape Lookout area (Table 1 ) . 
Cape Fear area sampling effort doubled over that 
near Cape Lookout in September and yielded many 
more female banded drum, but only 0.7'r were re- 
productively active compared with 48.9% in the 
Cape Lookout area (Table 1). Irregular sampling 
from the Cape Hatteras area (Fig. 1) yielded matur- 
ing or ripe L. fasciatus only during April when 82.4% 
of the females collected had gonad indices between 
1.7 and 6.1 (Table 1). Bottom water temperature in 
this area was 17°C. 

Ovum diameter is often an indication of sexual 
maturity (Higham and Nicholson 1964), and the 
relationship between egg size (OD) and gonad index 
(GI) for banded drum {n = 90) was 

OD - 0.34 + 0.1 1 (In GI), r = 0.77 

(Fig. 4). This relationship is an objective, quantita- 
tive way to determine degree of maturity (Yuen 1955; 



Schaeferand Orange 1956) and was used to differen- 
tiate maturing from immature female banded drum. 
The point on the graph (Fig. 4) where gonad index 
began to increase more rapidly than egg size was used 
as the boundary between immature and maturing 
gonads and occurred around a gonad index of 1.0 and 
an ovum diameter of 0.35 mm. Mean ova diameters 
peaked from July through September at 0.48 mm 
(Table 2), which also coincided with the highest 
gonad indices. 

Maturity 

Female banded drum reached sexual maturity be- 
tween 120 and 130 mm SL (n = 112). All fish <120 
mm SL were immature (GI <1.0) and 97% of those 
>130 mm were mature, with 607t between 120 and 
130 mm reaching maturity (Table 3). During the 
spawning season, females between 120 and 130 mm 
indicated increased gonad activity. Females smaller 
than 120 mm displayed no seasonal gonad activity, 
while only three fish >130 mm were not maturing 
during the spawning season (Fig. 5). Only the larger 
adults >150 mm matured and spawned early (April), 
and generally a higher proportion of the older 



Table 1.— 


Percent of female Larimus fasciatus with gonad indices >1.0 and sample size 
collection area during the spawning months of 1976. 


TV) from each 


Area 


April 


May 




June 


July 


Aug. 


Sept. 


Total 


Cape Fear 
Cape Lookout 
Cape Hatteras 
Total 


(274) 
(1) 
82.4 (17) 
4.8 (292) 


0(219) 
0(219) 




(9) 
53.8(26) 

40.0(35) 


2.8 (36) 
100 (12) 

27 .1 (48) 


8.1 (11 1) 
75.0(28) 

21.6(139) 


0.7 (153) 
48.9(45) 

11 6(198) 


1 4(802) 
61 6(112) 
82.4 (17) 
10.1 (931) 



0.6 



5 



£ 
E 

<x 



4 



•S 0.3 

< 

Q 

I 0.2 

> 

o 

o i 



••• • ••« 



OD = 0.34 + 0.11ILnGII 

r = 0.77 

n=90 



3 4 

GONAD INDEX 



FIGURE 4. — Relationship between famale gonad index and ova diameters of North 

Carolina Larimus fasciatus. 



230 



ROSS: REPRODUCTION OF BANDED DRUM 



TABLE 2. — Mean monthly ova diameters of Larimus 
fasciatus from March through September 1976. 





M 


•■in 


ova diameter 






Month 






(mm) 


Sa 


mple size 


March 






0.01 




1 


April 






0.41 




7 


June 






0.46 




16 


July 






0.48 




13 


August 






0.48 




32 


September 






0.48 




21 



females continued spawning later (September) (Fig. 
5). Most of the smallest reproductively active 
females (between 120 and 130mmSL) matured from 
June to August (Fig. 5). 

Using age-length relationships of Ross (1978), 
Larimus fasciatus reached maturity shortly after 
turning 1-yr-old. They continued spawning through- 
out life until age 3, which was the maximum age 
encountered. 

Fecundity 



TABLE 3. — Number and percentage of mature and 
immature female banded drum by 10 mm size cate- 
gories off North Carolina, April-September 1976. 
Maturity was judged by gonad index (GI) value. 



Standard 


Immature 


Mature 


Percent 


length (mm) 


GI < 1 .0 


GI > 1.0 


mature 


<90 


1 





00 


90-99 








0.0 


1 00- 1 09 


4 





0.0 


110-119 


4 





0.0 


120-129 


6 


9 


60.0 


130-139 


1 


12 


92.3 


140-149 


1 


22 


95.7 


150-159 





17 


100.0 


160-169 


1 


26 


963 


170-179 





7 


100.0 


180-189 





1 


100.0 


Total 


18 


94 





130 mm SL and predicted fecundity in this size range 
is 31,088-65,038 ova. Body weight (BW) minus the 
gonad weight (GW) was regressed onto fecundity 
yielding the equation: 



Number of ova increased with increasing fish size, 
ranging from 12,750 ova in a 118 mm SL female to 
320,819 in a 179 mm female. The relationship be- 
tween fecundity (F) and SL for 86 females was linear 
and expressed by the equation: 

F = -376,312 + 3,395 (SL), r = 0.76 
(Fig. 6). Length at first spawning is between 120 and 



F = -52,741 + 1,887 (BW), r = 0.76, n = 85. 

Gonad weight varies seasonally and is closely related 
to fecundity; therefore, eliminating it from body 
weight reduced the possibility of autocorrelation. 
Even without the gonad weight, body weight varies 
seasonally and to some extent daily as a function of 
diet; therefore, body weight is not the best predictor 
of fecundity. The fecundity to ovary weight (0 W) 



7- 



6- 



5- 



X 

LU 

Q 



4- 



Q 

I 3 

O 

o 



1- 



6 7 63 4 ^ *** 



88 
8 



4 9 



9 



8 8 



7 6 

4 8 8 

7 4 
6 9 9 9 8 8 8 

'788 8 8 

9 6 8 4 

9 9 8 6 «8> 

9 6 9 

89 9 99 9 9 8 9 4 « 

6 ^ ' 



6 9 7 
7 

6 6 

6 
6 6 6 

7 8 



4 4 
4 



— r~ 
90 



100 110 120 130 140 150 160 170 180 

STANDARD LENGTH(mm) 



FIGURE 5. — Relationship between famale gonad index and standard length by month 
for banded drum during January (l)-September (9) 1976 (n = 124). 



231 



FISHERY BULLETIN: VOL. 82, NO. 1 



35- 












F = 


-376,312 


+ 3395ISL] 




r = 


0.76 




• 


30- 


n : 


86 




* • / 


25- 










^^ 








• • / • 


"<* 








• / 


O 








• / 


--20- 








/ 


X 








• / 


' — 








• y 


> 








• • * • 


t 15- 








• / • •• 










• •• • • • • 


Q 








• • /• • 


2 








• Am* 

• • •* • 


z> 








• / • 


Q 10- 






• 
• 


/• •• 

• • ••• • • 

/•• • • 


LLI 










U_ 






* 4 


• 
• 


5 - 




t '• 


/ • 

• * 

• 
• 


• 

• 


o- 


1 






1 1 T— 



100 125 150 175 200 

STANDARD LENGTH(mm) 

FIGURE 6. — Relationship of fecundity to standard length for banded 
drum collected in North Carolina from April through September 
1976. 



relationship was expressed by 

F = 15,490 + 28,024 (OW), r = 0.94, n = 85 

and had a much higher correlation coefficient than 
either the length or body weight regressions. To 
minimize monthly variation (Morse 1980) the most 



accurate prediction of fecundity was derived from 
ovary weights only from the peak spawning month, 
August, expressed by 

F = 18,532 + 28,181 (OW), r = 0.97, n = 31 



(Fig. 7). 



Sex Ratios 



Sex was determined for 2,729 banded drum and the 
overall ratio of males to females varied significantly 
from 1:1 in favor of females (Table 4). This non- 
homogeneity of total sex ratios could not be account- 
ed for by any consistent pattern of seasonal ratio 
differences. The two largest size groups exhibited sex 
ratios significantly in favor of females. The disparity 
between sexes in the size range 100-139 mm SL was 
accounted for during winter, spring, and summer, 
while that in the fish > 140 mm SL was accounted for 
during fall and winter (Table 4). Contingency table 
analysis indicated strong dependency between sex 
and size group (x 2 = 17.84, df — 3,P < 0.001), even 
though differences in the smallest two size groups 
were nonsignificant (Table 4). As fish grew, the 
population shifted from more males to more females. 
There were more total females than males in all 
seasons except summer; however, the differences 
were only significant in the fall. The fall divergence 
from a 1:1 ratio was explained by differences in the 
60-99 mm and >140 mm SL size groups (Table 4). 





33- 




30- 




27- 


o 


24- 
21- 


X 


18- 


> 




Q 


15- 
12 


D 
O 


9- 


LU 
Ll_ 


6 




3- 




0- 



F = 18,532+ 28,181I0WI 
r =0.97 
n = 31 



-i 1 1 1 1 I r- 

2 3 4 5 6 7 8 

OVARY WEIGHT (g) 



10 



11 



FIGURE 7. — Relationship of fecundity to overy weight during August 1976 for North Car- 
olina banded drum. 



232 



ROSS: REPRODUCTION OF BANDED DRUM 



Table 4. — Larimus fasciatus male/female sex ratios by season and size group from 
North Carolina (September 1975-September 1976) with chi-square values from testing 
a 1:1 ratio. 







Season 










Size group 


Fall 


Winter 


Spring 


Summer 




(mm SL) 


(Sept. -Nov.) 


(Dec. -Feb.) 


(Mar. -May) 


(June-Aug.) 


Total 


df 


X 1 


<59 


151/144 


22/27 


104/100 


65/34 


342/305 


3 


5.23 


60-99 


77/103 


15/12 


492/479 


65/64 


649/658 


3 


2.14 


100-139 


71/67 


19/34 


87/123 


48/74 


225/298 


3 


8.04' 


>140 


21/64 


14/23 


19/22 


49/50 


103/149 


3 


8.47' 


Total 


320/268 


70/96 


702/724 


227/222 


1.319/1.410 






df 


3 


3 


3 


3 








X 1 


928* 


3.64 


7.28 


763 


10.38* 







'P < 05 



DISCUSSION 

The prolonged April-September spawning season 
of L. fasciatus in this study is supported by the few 
published references to its reproduction. From 
analysis of larval occurrence in North Carolina, 
Hildebrand and Cable (1934) proposed a May 
through October spawning season. Powles (1980) 
reported a May to October spawning in the South 
Atlantic Bight also based on larval collections. Gun- 
ter (1938) suggested April spawning for banded 
drum in Louisiana. Standard and Chittenden (in 
press) found two spawning peaks forL. fasciatus off 
Texas, a minor one in the spring (April-June) and the 
major one in the fall (September-November). They 
did not find significant evidence of spawning in July 
or August. 

My data suggested a prolonged spawning effort in 
North Carolina beginning as early as April, peaking in 
August, and possibly continuing after September. 
This major departure from Standard and Chitten- 
den's (in press) biomodel spawning was supported by 
1) a steady increase in gonad indices with a single 
August peak, 2) a single peak mode of ova diameters 
of 0.48 mm from July through September, 3) con- 
tinuous recruitment of young-of-the-year through 
the summer and fall months, and 4) the collection of 
larvae in all months except March (Powles 1980; K. 
A. MacPherson footnote 4). Although it is fairly cer- 
tain that spawning begins in April, at least for larger 
fish, I did not determine if spawning continued into 
October because samples of adults were lacking. 
Although the September gonad index declined, 
young-of-the-year recruitment in North Carolina in 
February and larval collections in November, 
December, January, and February (K. A. MacPher- 
son footnote 4) indicated that spawning may last at 
least through October. Protracted spawning is also 
characteristic of many other Sciaenidae (Welsh and 
Breder 1923; Thomas 1971; Merriner 1976; 
Warlen 1980). 



Maturation at an early age is typical in sciaenids 
(Schaefer 1965; Meriner 1976; Shlossman and Chit- 
tenden 1981) and in short-lived fishes in general 
which tend toward r strategy life histories (Adams 
1980). Since L. fasciatus is a short-lived sciaenid, 
rarely completing a fourth year, the small size (120 
mm SL) at first maturity, attained shortly after reach- 
ing 1 yr of age, is not surprising (Ross 1978). Larimus 
fasciatus off of Texas apparently live only 2 yr and 
consequently mature earlier (80 mm TL) than North 
Carolina individuals (Standard and Chittenden in 
press) . In addition to short life and early maturation, r 
strategists' traits are rapid growth, high fecundity 
(even at early ages), small maximum size, high mor- 
tality, and low maximum age (Adams 1980), all of 
which are related to emphasizing reproductive pro- 
ductivity. Banded drum have all of these characteris- 
tics as indicated in this study and by Ross (1978) and 
Standard and Chittenden (in press). 

As banded drum became older their growth rate 
slows (Ross 1978; Standard and Chittenden in 
press), as is typical of most fishes, and they can 
devote relatively more energy toward reproductive 
activity than at earlier ages. Only the largest fe- 
males (>150 mm) appeared to spawn as early as 
April and continue spawning into September. 
Although the phenomenon of older fish having a lon- 
ger spawning season has not been reported in United 
States east or gulf coasts sciaenids, it does occur in 
other fishes (Quast 1968; Grimes and Huntsman 
1980). 

Larimus fasciatus spawns as far north as Cape Hat- 
teras. Although larvae have been collected off 
Chesapeake Bay (Berrien et al. 1978), there are no 
records of reproductively active adults north of Cape 
Hatteras and this species is rare north of Chesapeake 
Bay (Hildebrand and Schroeder 1928; Johnson 
1978); therefore, Cape Hatteras is probably the 
northern limit of banded drum reproduction. 
Larimus fasciatus in spawning condition were most 
often collected in the nearshore waters between 



233 



FISHERY BULLETIN: VOL. 82, NO. 1 



Beaufort Inlet and Cape Lookout, larval dis- 
tributions have not clarified the preferred spawning 
depth range, since larvae have been collected over a 
wide range of the continental shelf (Berrien et al. 
1978; Powles 1980); there is, however, some ten- 
dency toward increased abundance over the inner 
shelf (Powles 1980). Miller et al. (in press) suggested 
that onshore transport by currents into estuarine 
nurseries of offshore spawned larvae is most favor- 
able during the winter off North Carolina south of 
Cape Hatteras. Several winter spawners with 
estuarine dependent young spawn along the outer 
continental shelf {Leiostomus xanthurus, Dawson 
1958; Mugil cephalus, Anderson 1958; Breuoortia 
tyrannus, Nelson et al. 1977; Micropogonias 
undulatus, Warlen 1980); thus, the young could take 
advantage of the inshore directed currents. A cor- 
ollary to this theory indicates that summer spawners 
should reproduce near shore or in the estuary if lar- 
vae are to be retained in the more productive shallow 
waters because net current movement is offshore 
(Miller et al. in press). In addition to L. fasciatus, 
other fishes also spawn in nearshore or estuarine 
waters south of Cape Hatteras during the summer 
(Cynoscion regalis, Merriner 1976; C. nebulosus, 
Mahood 1975; Stellifer lanceolatus and Bairdiella 
chrysoura, Powles 1980). 

ACKNOWLEDGMENTS 

I am especially grateful to Sheryan P. Epperly for 
her time and statistical advice. I also thank D. R. 
Colby, G. W. Link, K. A. MacPherson, and F. C. 
Rohde for advice and field assistance. J. W. Gillikin, 
B. F. Holland, S. G. Keefe, J. B. Sullivan, and J. 
Vaughn made special efforts to collect specimens for 
this study. C. S. Manooch III and C. B. Grimes con- 
tributed much through review of a preliminary 
manuscript. I wish to acknowledge my thesis commit- 
tee, F. J. Schwartz, E. A. McMahan, and A. F. 
Chestnut, for their support. Major support for this 
project was through a grant from Carolina Power and 
Light Company. 

LITERATURE CITED 



Adams, P. R. 

1980. Life history patterns in marine fishes and their conse- 
quences for fisheries management. Fish. Bull., U.S. 78:1- 
12. 

Anderson, W. W. 

1958. Larval development, growth, and spawning of striped 
mullet (Mu?>il cephalus) along the South Atlantic coast of 
the United States. U.S. Fish Wildl. Serv., Fish. Bull. 

58:501-519. 



Berrien, P. L., M. P. Fahay, A. W. Kendall, Jr., and W. G. 
Smith. 

1978. Ichthyoplankton from the RV Dolphin survey of con- 
tinental shelf waters between Martha's Vineyard, 
Massachusetts and Cape Lookout, North Carolina, 1965- 
66. U.S. Dep. Commer., NOAA, Sandy Hook Lab. Tech. 
Serv. Rep. 15, 152 p. 
CHAO, L. N. 

1978. Family: Sciaenidae.Lon>m/.s fasciatus, Holbrook, 
1860. In W. Fischer (editor), FAO species identification 
sheets for fishery purposes, western central Atlantic (fish- 
ing area 31), Vol. 4, unpaged. FAO, Rome. 
CHAO, L. N., AND J. A. MUSICK. 

1977. Life history, feeding habits, and functional morphology 
of juvenile sciaenid fishes in the York River estuary, 
Virginia. Fish. Bull., U.S. 75:657-702. 

Dahlberg, M. D. 

1972. An ecological study of Georgia coastal fishes. Fish. 
Bull., U.S. 70:323-353. 
Dawson, C. E. 

1958. Study of the biology and life history of the spot, Leio- 
stomus xanthrus Lacepede, with special reference to 
Smith Carolina. Contrib. Bears Bluff Lab, 28, 48 p. 
Grimes, C. B., and G. R. Huntsman. 

1980. Reproductive biology of the vermilion snapper, Rhnm- 
boplites aurorubens, from North Carolina and South Car- 
olina. Fish. Bull., U.S. 78:137-146. 
GUNTER, G. 

1938. Seasonal variations in abundance of certain estuarine 
and marine fishes in Louisiana, with particular reference 
to life histories. Ecol. Monogr. 8:313-346. 
Gutherz, E. J., G. M. Russell, A. F. Serra, and B. A. Rohr. 

1975. Synopsis of the northern Gulf of Mexico industrial and 
foodfish industries. Mar. Fish. Rev. 37(7): 1-1 1. 
Higham, J. R., and W. R. Nicholson. 

1964. Sexual maturation and spawning of Atlantic men- 
haden. U.S. Fish Wildl. Serv., Fish. Bull. 63:255-271. 
Hildebrand, S. F„ and L. E. Cable. 

1934. Reproduction and development of whitings or kingfish, 
drums, spot, croaker, and weakfishes or seatrouts, family 
Sciaenidae, of the Atlantic coast of the United 
States. Bull. [U.S.| Bur. Fish. 48:41-117. 
Hildebrand, S. F., and W. C. Schroeder. 

1928. Fishes of Chesapeake Bay. Bull. U.S. Bur. Fish. 43(1), 
366 p. 
Holbrook,.]. E. 

1860. Ichthyology of South Carolina. 2d ed. Russell and 
Jones, Charleston, S.C., 205 p. 
Johnson, G. D. 

1978. Development of fishes of the mid-Atlantic Bight: an 
atlas of egg, larval and juvenile states, Vol IV. Carangidae 
through Ephippidae. U.S. Fish Wildl Serv. Biol. Serv. 
Program FWS/OBS-78/12, 314 p. 

Mahood, R. K. 

1975. Spotted seatrout in coastal waters of Georgia. Proc. 
29th Annu. Conf. Southeast. Assoc. Game Fish Comm., p. 
195-207. 

Merriner, J. V. 

1976. Aspects of the reproductive biology of the weakfish, 
Cynoscion regalis (Sciaenidae), in North Carolina. Fish. 
Bull., U.S. 75:18-26. 

Miller, J. M., J. P. Reed, and L. T. Pietrafesa. 

In press. Patterns, mechanisms, and approaches to the study 
of migrations of estuarine dependent fish larvae and 
juveniles. In J. D. McCleave (editor), Mechanisms of 
migration in fishes. NATO Adv. Res. Inst, 13-17 Dec. 



234 



ROSS: REPRODUCTION OF BANDKI) DRl M 



1982. Acquafredda, Italy. 

Morse. W. W. 

1980. Maturity, spawning, and fecundity of Atlantic croaker, 
Micropogonias undulatus, occurring north of Cape Hat- 
teras. North Carolina. Fish. Hull., U.S. 78:190-1 Of,. 

Nelson, W. R.. M. C. Ingham, and W. E. Schaaf. 

1977. Larval transport and year-class strength of Atlantic 
menhaden. Brevoortia tyrannus. Fish. Hull., U.S. 75:23- 
41. 

POWLES, H. 

1980. Descriptions of larval silver perch, Bairdiella 
chrysoura, banded drum, Larimus fasciatus, and star 
drum. Stellifer lanceolatus (Sciaenidae). Fish. Hull., U.S. 
78:119-136. 

Ql AST, J. C. 

1968. Observations on the food and biology of the kelp bass, 
Paralobrax clathratus with notes on its sportfishery at San 
Diego. California. In W. J. North and C. I.. Hubbs (com- 
pilers and editors), Utilization of kelp-bed resources in 
southern California, p. 81-108. Calif. Fish Game, Fish 
Bull. 139. 

Ross. S. W. 

1978. The life history of the banded drum, Larimus fasciatus, 
in North Carolina waters. M.A. Thesis, Univ. North Car- 
olina, Chapel Hill, 143 p. 

Schaefer. m. b.. and C. .J. Oranok. 

1956. Studies of the sexual development and spawning of 
yellowfin tuna (Neothunnus macropterus) and skipjack 
{Katsuwanus pelamis) in three areas of the eastern Pacific 
Ocean, by examination of gonads. [In Engl, and 
Span.] Inter-Am. Trop. tuna Comm., Bull. 1:281-349. 



Schaefer, R. H. 

1965. Age and growth of the northern kingfish in New York 
waters. N.Y. Fish. Game J. 12:191-216. 
Sholssman, P. A., and M. E. Chittenden, Jr. 

1981. Reproduction, movements, and population dynamics 
of the sand seatrout, Cynoscion arenarius. Fish. Bull.. 
U.S. 79:649-669 
Standard, G. W., and M. E. Chittenden, Jr. 

In press. Reproduction, movements, and population dy- 
namics of the banded drum,Larimus fasciatus, in the Gulf 
of Mexico. Fish. Bull., IS 
Thomas, D. L. 

1971. The early life history and ecology of six species of drum 
(Sciaenidae) in the lower Delaware River, a brackish tidal 
estuary. Ichthyol. Assoc. Bull. 3, Part III, 247 p. 

WARLEN, S. M. 

1980. Age and growth of larvae and spawning time of Atlantic 
croaker in North Carolina. Proc. Annu. Conf. Southeast. 
Assoc. Fish Wildl. Agencies 34:204-214. 
Welsh, W. W., andC. M. Breder. 

L923. Contributions to life histories of Sciaenidae of the east- 
ern United States coast. Hull. U.S. Bur. Fish. 39:141- 
201. 
Wolff, M. 

1972. A study of North Carolina scrap fishery. N.C. Div. 
Comm. Sport Fish., Dep. Nat. Econ. Resour., Spec. Sci. 
Rep. 20, 29 p. 

Yuen, H. S. H. 

1955. Maturity and fecundity of bigeye tuna in the 
Pacific. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 150, 
30 p. 



235 



NOTES 



MARKING GROWTH INCREMENTS IN 

OTOLITHS OF LARVAL AND JUVENILE 

FISH BY IMMERSION IN TETRACYCLINE 

TO EXAMINE THE RATE OF 

INCREMENT FORMATION 

Age determination of fishes by counting daily growth 
increments in their otoliths is becoming a widely used 
technique in growth and population studies. Daily 
formation of otolith increments was first reported by 
Pannella (1971) for three species of temperate fish. 
Since then a number of workers, using three basic 
techniques for confirming the periodicity of incre- 
ment formation, have reported the presence of daily 
increments in larval or adult otoliths of at least 15 
species of marine and freshwater fishes. Laboratory 
rearing from eggs to larvae of known age was used to 
confirm daily increments by brothers et al. (1976), 
Taubert and Coble (1977), Barkman (1978), Tanaka 
etal. (1981), and Laroche et al. (1982). The change in 
the mean number of increments over time in fish cap- 
tured in the wild and held in captivity was used to 
validate daily increments by Struhsaker and 
Uchiyama (1976), Wilson and Larkin (1980), and 
Uchiyama and Struhsaker (1981). The third method 
makes use of chemical agents to mark the growing 
margin of calcified structures in order to examine their 
rate of growth (Harris 1960). Tetracycline is one of 
the best chemical markers because it is relatively 
nontoxic and produces a fluorescent mark which is 
easily viewed in ultraviolet light (Harris 1 960; Weber 
and Ridgway 1962). It has been administered to fish 
by feeding (Choate 1964; Weber and Ridgway 1967; 
Trojnar 1973; Odense and Logan 1974) and by injec- 
tion (Kobayashi et al. 1964 and others below). Tet- 
racycline has been used in two studies to determine 
the rate of increment formation in otoliths. Wild and 
Foreman (1980) injected the drug into large juveniles 
and adult skipjack tuna, Kotsuwonus pelamis, and 
yellowfin tuna, Thunnus albacares, in a mark- 
recapture program in the tropical eastern Pacific. 
They found that otoliths of yellowfin tuna of 40-110 
cm FL showed daily average increment formation, 
but that skipjack tuna of 42-64 cm FL showed <1 
increment/d. Campana and Neilson (1982) injected 
tetracycline into juvenile starry flounders, Platich- 
thys stellatus, and found that daily increments were 
subsequently produced in both field and laboratory 
conditions. These authors briefly mentioned obtain- 
ing similar marking results by immersion, but did not 
detail their procedure. 



This paper presents a technique for marking otolith 
increments by immersing larval and juvenile fish in a 
solution of tetracycline in seawater, and reports the 
rate of increment formation under laboratory con- 
ditions for two species from the Great Barrier Reef, 
Australia: Hypoatherina tropicalis (Altherinidae) and 
Spratelloides dellicatulus (Dussumeriidae). 

Materials and Methods 

The experiments were conducted between July 
1980 and February 1982 at One Tree Island Field 
Station and Lizard Island Research Station, during a 
field study of the population dynamics of these 
species. 

Achromycin (a brand of tetracycline HC1 1 ) was used 
in all experiments. The concentration that would 
mark the otoliths but not kill the fish was determined 
by testing three concentrations (400 mg/1, 250 mg/1, 
and 40 mg/1) using//, tropicalis from 12.8 to 23.0 mm 
SL. The otoliths of survivors were compared with 
untreated specimens to assess the effectiveness of 
the mark. 

The appropriate concentration, 250 mg/1, was then 
used in a series of similar experiments to determine 
the rate of increment formation (Table 1). The 
experiment number (I-IV) designates a group offish 
collected at the same time. In each experiment, fish 
were killed at two different times, designated as A or 
B, to compare the number of increments in fish held 
for different time periods. In experiment IV, the 
treatment times also differed, but in all other 
experiments the treatment time was the same for 
both groups A and B. 

Both species are small (adults <7 cm SL), mid- 
water, reef-associated, schooling fishes which do not 
undergo a marked metamorphosis between larval 
and juvenile stages (pers. observ.). Both attain their 
full complement of fin elements and begin to form 
scales and adult pigmentation at a standard length of 
17-19 mm. Following the convention of Ahlstrom 
(1968), I consider this to be the size at which larvae 
become juveniles. Hypoatherina tropicalis used in the 
rate-determination experiments ranged from 12.8 to 
27.2 mm SL, with 10 of 21 fish classed as larvae 
(<17.0 mm SL). Spratelloides delicatulus ranged 
from 15.5 to 22.9 mm SL, with 2 of 29 being larvae 
(Table 1). 



'Manufactured by Lederle Labs, a division of Cyanamid Australia 
Pty. Ltd. References to trade names do not imply endorsement by 
the National Marine Fisheries Service, NOAA. 



FISHERY Bl'LLKTIN: VOL. 82, NO. 1, 1984. 



237 



The fish were collected at night with a light and a dip 
net, and placed in 25 1 aquaria without aeration or 
running seawater as soon as possible after collection. 
The aquaria were located outdoors under an awning, 
and therefore were exposed to the ambient diel light 
cycle, but not to direct sunlight. The fish were 
allowed to acclimatize for 12-24 h before treatment. 
Usually there was mortality during this period, but 
the proportion was not determined. All dead fish 
were removed prior to treatment. 

The fish were exposed to 250 mg tetracycline/1 
seawater for 12 h from sunset to sunrise, except in 
experiment IVB when the immersion period was 
from sunrise to sunset (Table 1). After an immersion 
period, the aquarium was flushed with 90% water 
changes until no visible color remained. The 
tetracycline-seawater solution is yellow until ex- 
posed to sunlight for more than ~3 h, when it turns 
pink, due to oxidative photolysis. Following the treat- 
ment, fish were maintained in clean seawater for 2-6 
d by feeding either fresh wild plankton > 125 mm 
diameter once a day (experiment I) or Artemia salina 
nauplii 3-4 times/d (all subsequent experiments). 
Artemia nauplii were more convenient for frequent 
feedings than fresh wild plankton. Ninety percent of 
the water in each aquarium was changed each morn- 
ing by siphoning, to minimize handling the fish. Tank 
water temperatures were measured over the diel 
cycle during February 1982 (summer) at One Tree 
Island. The temperature ranged from 25°C at 0630 h 
to 30°C at 1800 h. Replacement water, added at 0700 
h from the surface of the lagoon, measured 27°C. 



Larvae were killed at the end of each experiment by 
placing them into 707c ethanol. Fish were subsequent- 
ly measured to the nearest 0. 1 mm SL. Their otoliths 
(both sagittae and lapilli) were removed and mounted 
whole on glass slides without coverslips, using 
Protexx. 

The following terms are used in this report for the 
concentric rings seen in otoliths. A growth zone is a 
wide ring which appears light or hyaline under 
transmitted light. A discontinuous zone is the nar- 
rower ring between two growth zones, often called the 
opaque zone because it appears dark under trans- 
mitted light. A growth increment, or simply an incre- 
ment, is a growth zone plus a discontinuous zone. 

Otoliths were examined at 250-l,000X magnifica- 
tion with a combination of incident ultraviolet light to 
reveal the fluorescent tetracycline-marked rings, and 
polarized transmitted light to count the rings. The 
fluorescence microscope used ultraviolet light from a 
50W mercury lamp. Excitation wavelength was 
limited by a band pass filter (450-490 nm) and a long 
pass suppression filter (515 nm). 

In most cases, one sagitta from each fish was read, 
although occasionally the lapillus was used if its rings 
were clearer. The area to be counted was selected by 
scanning the margin of each otolith to find the place 
where the greatest number of distinct rings could be 
seen between the innermost fluorescent increment 
and the edge. A datum was considered valid only if 
identical counts were obtained in at least two out of 
three blind readings. No other otoliths were con- 
sidered in the analysis. Of 21 H. tropicalis otoliths 



Table 1. 


—Summary of tetracyeline-marking experiments to determine the rate of increment formation in H. tropicalis and S. delicatulus. 


Experiment 


No. of fish with various 
Standard length Predicted no deviations from the 
(mm) Treatment Date and of discontin- predicted number 
N Mean (range) period time of killing uous zones . 



Hypoathenna tropicalis 
IA 



IB 
MA 
IIB 

IMA 

Total 



2( 1 ) 

4 

6( 2 1) 

6 

3 

21| 2 1) 



Spratelloides delicatulus 
IIIA 6 



1MB 
IVA 
IVB 
Total 



5( 2 2) 
9 

9( 2 1) 
29( 2 3) 



14.0 (13.6-14.4) 
13.7 (12.8-14.7) 
20.5 (16.2-27.2) 
18.9(16.8-20.7) 

16.1 (15.4-17.2) 

17.5 (15.5-19 .1) 
17 9(17.6-18.2) 

19 9(18.8-22.8) 

20 5 (17.9-22 9) 



2130, 8 July to 

0830. 9 July 1980 
2130, 8 July to 

0830. 9 July 1980 
1830. 31 Oct. to 

0630. 1 Nov. 1980 
1830. 31 Oct. to 

0630, 1 Nov 1980 
2000, 6 Nov to 

0700. 7 Nov 1981 



2000. 6 Nov to 
0700. 7 Nov. 1981 

2000, 6 Nov. to 
0700. 7 Nov. 1981 

1800. 31 Jan to 
0630. 1 Feb. 1982 

0600 to 1 800 
31 Jun 1982 



0830, 12 July 
1730, 14 July 
0730, 6 Nov. 
0600, 7 Nov. 
0545, 12 Nov. 

0545, 12 Nov. 
1800, 9 Nov. 
1 800. 6 Feb. 
0715, 6 Feb 



2+1 

5 

4+1 

5+1 

4+1 

4+1 
2 
5 
4+1 



1 

2 
5 
6 
3 

17 



5 3 
2 6 

7 12 



'Otoliths of two treated fish were destroyed by poor preservation. 

2 Number of fish discarded because of inconsistency between otolith readings. 



238 



examined, 1 (4.8%) was discarded. Of 29 S. 
delicatulus, 3 (10.3%) were discarded (Table 1). 

Results and Discussion 
Marking Technique 

In the experiment to determine an effective 
tetracycline-marking concentration, all fish (n = 17) 
in 400 mg/1 died during the 12-h immersion period. 
Of 10 fish treated with 250 mg/1, 1 died during 
treatment, and 1 died during the subsequent holding 
period. Of 10 fish treated witlv50 mg/1, 1 died dur- 
ing treatment. 

Otoliths of untreated specimens showed faint 
fluorescence around the edge and occasionally along 
cracks and surface irregularities (Fig. 1A); this is a 
naturally occurring autofluorescence (Campana and 
Neilson 1982). Otoliths of fish in 50 mg/1 were indis- 
tinguishable from those of untreated specimens. 
Otoliths of fish in 250 mg/1 showed a strong fluores- 
cent band medial to the edge, in addition to the weak 
fluorescence at the edge (Fig. IB, C). This strong 
band consisted of two growth zones and one discon- 
tinuous zone (Fig. 2). 

It is not known how long it takes for tetracycline to 
be incorporated into the growing otoliths when 
administered by immersion. Campana and Neilson 
(1982) reported that after injection, 50% of fish 
showed fluorescent otoliths after 10 h, and 100% 
after 24 h. If one assumes similar or slightly longer 
incorporation times in the present study, then the 
inner fluorescent growth zone was probably formed 
the day after the immersion period. The subsequent 
discontinuous zone and growth zone were formed 
while there was residual tetracycline in the water or 
fish. Another possible explanation is that the 
appearance of fluorescence in two growth zones is an 
artifact of viewing whole otoliths. 

The results of this experiment indicate that immer- 
sion in a concentration of 250 mg Achromycin/1 of 
seawater for 12 h is adequate to mark one or more 
growth increments in//, tropicalis and S. delicatulus 
larvae and juveniles. The overall mortality rate in 
experiments I, II, and III (total n = 37), was 5.4% 
during treatment and 2.7% during the holding phase. 

To determine whether fluorescent marking would 
occur if the tetracycline immersion period was during 
daylight hours, an experiment was conducted using 
S. delicatulus from 17.9 to 22.9 mm SL (experiment 
IV). The fish were collected and divided between two 
tanks. One tank received tetracycline from 1800 h to 
0630 h, the other from 0600 to 1800 h. Mortality due 
to treatment was not monitored. After 6 d, the fish 






FIGURE 1. — Flourescence photomicrographs of sagittae of larval 
Hypaatherina tropicalis. A. Untreated otoliths, showing autofluores- 
cence around the edge (10.1 mm SL). B. Tetracycline-marked oto- 
lith, showing fluorescent band medial to the edge (16.2 mm SL). C. 
Marked otolith under higher magnificaton (17.6 mm SL). 



were killed and examined. The fluorescent bands 
medial to the edges were similar in width and inten- 
sity to those in previous experiments, and showed no 
difference between the two treatments. This 
indicates that tetracycline is incorporated into grow- 
ing otoliths and produces fluorescent increments 
equally well during the day and night, regardless of 
whether the solution is yellow or has oxidized to 
pink. 



239 





FIGURE 2.— Tetracycline-marked otolith from H. tropicalis (17.6 
mm SL), photographed with a combination of fluorescent and trans- 
mitted polarized light. Arrows indicate the fluorescent band pro- 
duced by the marking technique. This individual is from experiment 
IIB, and shows six discontinuous zones between the innermost 
fluorescent growth zone and the edge. The edge appears to be a 
growth zone. 



In summary, tetracycline can be administered by 
three techniques: feeding, injection, and immersion. 
Feeding has apparently not been used in otolith 
studies. The immersion technique presented here 
has advantages over injection in some situations. It 
can be used on fish which are too small or fragile for 
inj ection. The fluorescent mark obtained is relatively 
narrow, covering only two increments, compared 
with the wider mark resulting from injection 
(Kobayashi et al. 1964; Campana and Neilson 1982). 
Therefore, it is distinguishable from edge autofluores- 
cence after a shorter period of time, and allows finer 
resolution of increment formation, which may be use- 
ful in some experimental situations. Also, immersion 
requires minimum equipment, facilities, and han- 
dling of fish. 

Rate of Increment Formation 

In interpreting the results of my experiments, the 
number of discontinuous zones between the inner- 
most fluorescent growth zone and the edge was com- 
pared with the number predicted if one dis- 
continuous zone formed every day from ca. 0700 to 
1000 h. Tanaka et al. (1981) found that growth zones 
in juvenile Tilapia nilotica held under various 
photoperiods started forming a few hours after 
lights-on, continued through the dark period, and 
stopped or slowed down about the time of the follow- 
ing lights-on. The discontinuous zone was formed in 
the few hours after lights-on. Mugiya et al. (1981) 
demonstrated that the deposition of calcium in 
goldfish, Carassius auratus, slowed down or stopped 



at sunrise and resumed in 3 h. Since otoliths are made 
of a matrix of organic fibers, which are calcified in the 
growth zones and not calcified in the discontinuous 
zones (Panella 1980; Watabe et al. 1982), the find- 
ings of Mugiya et al. (1981) support Tanaka et al. 
(1981). Whether this rhythm of increment formation 
is found in most fish remains to be investigated. 

The results for all experiments are presented in 
Table 1. For fish that were killed between 0545 and 
0830 h, the predicted number includes an additional 
discontinuous zone that should have been forming at 
the time of death, although this ring was probably not 
always sufficiently formed to be counted. In these 
cases, an otolith was considered to show daily incre- 
ment formation even if the number of discontinuous 
zones was one less than predicted. 

One growth increment was formed each day in 85% 
of H. tropicalis (n — 20); the rest had one more than 
the predicted number of increments. In S. 
delicatulus, 46% (n = 26) showed daily formation of 
growth increments; 27% showed one less, and 27% 
showed one more, than expected if increments form 
daily. Thus, the variability in rate of increment for- 
mation was greater in S. delicatulus than in H. 
tropicalis, but the average rate for S. delicatulus was 
still 1 increment/d. 

This apparent difference in the rate of increment 
formation between species may be partially due to a 
difference between larvae and juveniles. Almost all 
(93%) of the S. delicatulus treated were juveniles, but 
only about half (52%) of the H. tropicalis were 
juveniles. However, no conclusion can be drawn from 
these data because the experiments were not 
designed to examine this factor, and the numbers are 
too small to compare larvae with juveniles. 

It is possible that tetracycline may affect the rate of 
increment formation. Some workers have reported 
that tetracycline inhibits mineralization in scales and 
bone (Harris 1960; Kobayashi et al. 1964), although 
others note neither growth promotion, retardation, 
nor structural weakness in bone as a result of tet- 
racycline administration (Weber and Ridgway 1967). 
The possibility that the tetracycline treatment inter- 
feres with growth of otoliths or fish was not con- 
sidered in this study, but should be examined before 
further use is made of this technique. 

In conclusion, the rate of increment formation has 
been examined in only a small number of species 
under a limited range of conditions. Recent evidence 
suggests that increment formation may be affected in 
some species by temperature, food availability and 
feeding frequency, photoperiod, and developmental 
stage (Taubert and Coble 1976; Brothers 1978; Pan- 
nella 1980; Wild and Foreman 1980; Geffen 1982; 



240 



Lough et al. 1982; Neilson and Geen 1982). It is 
therefore desirable to examine the rate of increment 
formation under various conditions before using 
otoliths for age determination (Brothers 1979). The 
technique presented here is a tool for studying incre- 
ment formation in otoliths of young fish under 
laboratory and possibly field conditions. It can be 
used for reef and nearshore benthic species which 
can be captured while larvae or juveniles and kept in 
containers or enclosures. 

Acknowledgments 

I want to thank Jeffrey M. Leis, Keith A. McGuin- 
ness, Richard Methot, Jr., Peter F. Sale, and two 
anonymous reviewers for their helpful comments on 
early drafts of the manuscript. I am grateful to J. M. 
Leis for the initial suggestion which led to this work, 
and to Peter Clarke for assistance with fluo- 
rescence microscopy. 

Literature Cited 

Ahlstrom, E. H. 

1968. Reviews and Comments: Development of fishes of the 
Chesapeake Bay region, an atlas of egg, larval, and 
juvenile stages, Part 1. Copeia 1968:648-651. 
Barkman, R. C. 

1978. The use of otolith growth rings to age young Atlantic 
silversides, Menidia menidia. Trans. Am. Fish. Soc. 
107:790-792. 
Brothers, E. B. 

1978. Exogenous factors and the formation of daily and sub- 
daily growth increments in fish otoliths. Am. Zool. 
18:631. 

1979. Age and growth studies on tropical fishes. In S. B. 
Saila and P. M. Roedel (editors), Stock assessment for 
tropical small-scale fisheries, p. 1 19-136. Proceedings of 
international workshop held Sept. 19-21, 1979, at Univer- 
sity of Rhode Island. (Copies available from Agency for 
International Development, AID-DIHF/ARDA, 7222 - 
47th St., Chevey Chase, MD 20815.) 

Brothers, E. B., C. P. Mathews, and R. Lasker. 

1976. Daily growth increments in otoliths from larval and 
adult fishes. Fish. Bull., U.S. 74:1-8. 
Campana, S. E., and J. D. Neilson. 

1982. Daily growth increments in otoliths of starry flounder 
(Platichthys stellatus) and the influence of some environ- 
mental variables in their production. Can. J. Fish. Aquat. 
Sci. 39:937-942. 
Choate, J. 

1964. Use of tetracycline drugs to mark advanced fry and 
fingerling brook trout (Salvelinus fontinalis). Trans. Am. 
Fish. Soc. 93:309-311. 
Geffen, A. J. 

1982. Otolith ring deposition in relation to growth rate in her- 
ring (Clupea harengus) and turbot (Scophthalmus max- 
imus) larvae. Mar. Biol. (Berl.) 71:317-326. 
Harris, W. H. 

1960. A microscopic method of determining rates of bone 
growth. Nature (Lond.) 188:1038-1039. 



KOBAYASHI, S., R. YUKI, T. FURUI, AND T. KOSUGIYAMA. 

1964. Calcification in fish and shell-fish - I. Tetracycline 
labelling patterns on scale, centrum and otolith in young 
goldfish. Bull. Jpn. Soc. Sci. Fish. 30:6-13. 
Laroche, J. L., S. L. Richardson, and A. A. Rosenberg. 

1982. Age and growth of a pleuronectid, Parophrys vetulus, 
during the pelagic larval period in Oregon coastal 
waters. Fish. Bull., U.S. 80:93-104. 
Lough, R. G., M. Pennington, G. R. Bolz, and A. A. Rosenberg. 
1982. Age and growth of larval Atlantic herring, Clupea 
harengus L., in the Gulf of Maine-Georges Bank region 
based on otolith growth increments. Fish. Bull., U.S. 
80:187-199. 
Mugiya, Y., N. Watabe, J. Yamada, J. M. Dean, D. G. 
Dunkelberger, and M. Shimizu. 

1981. Diurnal rhythm in otolith formation in the goldfish, 
CaraxsiiLs auratus. Comp. Biochem. Physiol. 68A:659-662. 
Neilson, J. D., and G. H. Geen. 

1982 Otoliths of chinook salmon {Oncorhynchus 
tshawytscha): daily growth increments and factors 
influencing their production. Can. J. Fish. Aquat. Sci. 
39:1340-1347. 
Odense, P. H., and V. H. Logan. 

1974. Marking Atlantic salmon (Salmo salar) with oxytet- 
racycline. J. Fish. Res. Board Can. 31:348-350. 
Pannella, G. 

1971. Fish otoliths: daily growth layers and periodical pat- 
terns. Science (Wash., D.C.) 173:1124-1127. 

1980. Growth patterns in fish sagittae. In D. C. Rhoads and 
R. A. Lutz (editors), Skeletal growth of aquatic organisms. 
Biological records of environmental change, p. 519- 
560. Plenum Press, N.Y. 

Stri hsaker, P., and J. H. Uchiyama. 

1976. Age and growth of the nehu, Stolephorus purpureus 
(Pisces: Engraulidae), from the Hawaiian Islands as 
indicated by daily growth increments of sagittae. Fish 
Bull., U.S. 74:9-17. 

Tanaka, K., Y. Mugiya, and J. Yamada. 

1981. E ff ects of photoperiod and feeding on daily growth pat- 
terns in otoliths of juvenile tilapia nilotiea. Fish. Bull., 
U.S. 79:459-466. 

Taubert, B. D., and D. W. Coble. 

1977. Daily rings in otoliths of three species of Lepomis and 
Tilapia mossambica. J. Fish. Res. Board Can. 34:332-340. 

Trojnar, J. R. 

1973. Marking rainbow trout fry with tetracycline. Prog. 
Fish-Cult. 35:52-54. 
Uchiyama, J. H, and P. Struhsaker. 

1981. Age and growth of skipjack tuna, Katsuwonus pelamis, 
and yellowfin tuna, Thunnus albacares, as indicated by 
daily growth increments of sagittae. Fish. Bull., U.S. 
79:151-162. 

Watabe, N., K. Tanaka, J. Yamada, and J. Dean. 

1982. Scanning electron microscope observations of the 
organic matrix in the otolith of the teleost fish Fundulus 
heteroclitus (Linnaeus) and tilapia nilotiea (Linnaeus). J. 
Exp. Mar. Biol. Ecol. 58:127-134. 

Weber, D., and G. J. Ridgway. 

1962. The deposition of tetracycline drugs in bones and 
scales offish and its possible use for marking. Prog. Fish- 
Cult. 24:150-155. 
1967. Marking Pacific salmon with tetracycline anti- 
biotics. J. Fish. Res. Board Can. 24:849-865. 
Wild, A., and T. J. Foreman. 

1980. The relationship between otolith increments and time 
for yellowfin and skipjack tuna marked with tet- 



241 



racycline. [In Engl, and Span.] Inter.-Am. Trop. Tuna 
Comm. Bull. 17:509-560. 
Wilson, K. H., and P. A. Larkin. 

1980. Daily growth rings in the otoliths of juvenile sockeye 
salmon (Onchorhynchus nerka). Can. J. Fish. Aquat. Sci. 
37:1495-1498. 



P. D. SCHMITT 



School of Biological Sciences, University of Sydney 

Sydney, N.S.W. 2006, Australia 

Present address: 

Southwest Fisheries Center La Jolla Laboratory 

National Marine Fisheries Service, NOAA 

P.O. Box 277, La Jolla, CA 92038 



TAG-RECAPTURE VALIDATION OF 

MOLT AND EGG EXTRUSION PREDICTIONS 

BASED UPON PLEOPOD EXAMINATION IN 

THE AMERICAN LOBSTER, 

HOMARUS AMERICANUS 

Techniques for molt prediction based upon epider- 
mal and setal development in pleopods (Aiken 1973) 
and for egg extrusion prediction based upon pleopod 
cement gland development (Waddy and Aiken 1980; 
Aiken and Waddy 1982) provide opportunities for 
more comprehensive studies of growth and re- 
productive potential in natural American lobster, 
Homarus americanus, populations than have pre- 
viously been possible. These laboratory-developed 
techniques have only recently been applied to field 
samples from a number of areas of Atlantic Canada 
(Robinson 1979; Campbell and Robinson 1983; 
Ennis 1984). Although the methodologies are fairly 
straightforward and may be applied in field studies 
quite readily, in practice the investigator will some- 
times be faced with specimens for which predictions 
can only be made with some degree of uncertainty. A 
study of Newfoundland lobsters using these tech- 
niques has included the tagging of animals from 
which pleopods were obtained. This paper presents 
results from observations on recaptured lobsters 
which validate the predictions that were made at the 
time of tagging that molting or egg extrusion would or 
would not occur during the current molting/spawning 
period. 

Materials and Methods 

Pleopods were obtained from American lobsters 
(ranging from 33 mm to 130 mm CL (carapace 
length)) caught in traps and by scuba divers near 
Arnold's Cove, Placentia Bay, Newfoundland, be- 
tween 24 June and 17 July 1981. These were 



examined for molt and cement gland stages accord- 
ing to the methodologies of Aiken (1973), Waddy and 
Aiken (1980), and Aiken and Waddy (1982). 

It is clear from Aiken (1973) that one can predict 
with considerable confidence that lobsters with 
pleopod stages 3.0 and higher just prior to or early in 
the molting season will molt that year. It is also clear, 
however, that for animals with pleopod stages 1.0-2.5 
one cannot predict with confidence that molting will 
or will not occur. Molt prediction for these stages is 
not reliable because of development plateaus that 
occur during D (i.e., molt stages 1.0-2.5). However, 
most such plateaus occur at stages 1.5-2.0, and a 
lobster will rarely remain at stage 2.5 for more than 2 
wk. Once an animal has passed beyond stage 2.5, 
there will be no further plateaus, and proecdysis will 
proceed at a rate that is regulated by temperature 
(Aiken 1973). Aiken (1980) also stated that at stage 
2.5, the epidermis in the general integument begins 
to show signs of activity, indicating imminent transi- 
tion from indecisive D„ into the irreversible premolt 
development of D,. Considering that animals with 
stage 2.5 pleopods should molt in 48-52 d at 10°C 
(Aiken 1973) plus the fact that at Arnold's Cove the 
July-August temperatures on the lobster grounds 
average in excess of 10°C (mean daily temperatures 
from 24 June to 31 August averaged 12. 1°C in 1981), 
it appeared more likely that lobsters with stage 2.5 
pleopods during the 24 June-17 July sampling at 
Arnold's Cove would molt. As a working hypothesis, 
it was decided to predict that lobsters with pleopod 
stages 2.5 and higher would molt during the 1981 
molting season at Arnold's Cove and that those with 
pleopod stages 0-2.0 would not molt. 

Cement glands were initially staged according to 
the classification scheme of Waddy and Aiken 
(1980). These stages were subsequently converted to 
their more recent scheme (Aiken and Waddy 1982). 
It is clear from these papers that for lobsters with 
stage or stage 1 cement glands just prior to or early 
in the spawning season one can confidently predict 
that egg extrusion will not occur that year, whereas 
for those with stage 2 or higher cement glands one can 
confidently predict that egg extrusion will occur. 

During the sampling at Arnold's Cove, 356 of the 
lobsters from which pleopods were removed for molt 
and cement gland staging were tagged with 
"sphyrion" tags, which are designed to remain 
attached through ecdysis (Scarratt and Elson 1965), 
and released within a few minutes of being taken from 
the water very close to where they were captured. 
Observations on 171 of these lobsters recaptured 
subsequent to the molting/spawning period (mainly 
during the 1982 fishing season, 20 April-30 June) 



242 



FISHERY lU'LLETIN: VOL. 82, NO. 1, 1984 



provide a basis for validating the molt or egg extru- 
sion predictions. 



TABLE 1. — Summary of molt predictions and subsequent valida- 
tions for American lobsters sampled and tagged at Arnold's Cove, 
Newfoundland, 24 June-17 July, 1981. 



Results 



Molt Predictions 



Four of the 11 males (36.4%) and 11 of the 27 
females (40.7%) with pleopod stages 0-2.0 molted 
instead of not molting as was predicted (Table 1). 
Even some with pleopod stage molted. Of the 16 
females which did not molt, 1 4 extruded eggs, and the 
2 females which did not extrude eggs had stage 1 
cement glands, indicating that egg extrusion would 
not occur. Six out of 21 males (28.6%) with pleopod 
stages 2.5 and 3.0 did not molt, whereas all with 
pleopod stages >3.5 and all females with pleopod 
stages >2.5 did molt (Table 1). Overall, 78.4% of the 
predictions which could be validated were correct. 
There was greater success with predicting that molt- 
ing would occur (89.8% correct predictions) than 
with predicting it would not (60.5% correct predic- 
tions). There was no pleopod stage at and below 
which none molted; however, at stage 3.5 and higher 
all molted. 

Validations of molt prediction are available for 
males ranging in size from 73 to 104 mm CL. Except 
for one animal at 99 mm, it was only for animals 
smaller than 8 1 mm that any of the predictions were 
incorrect. The size range for which validations are 
available for females is limited (75-82 mm CL). 

Egg Extrusion Predictions 

All of 17 females with either stage or stage 1 
cement glands did not extrude eggs, and all of 7 with 
stage 3 cement glands did extrude eggs as predicted. 
However, 2 out of 9 with stage 2 cement glands, which 
were predicted would lay eggs, did not do so (Table 
2). Overall, 93.9% of the predictions which could be 
validated were correct. The 2 females which failed to 
extrude eggs as predicted, molted, despite having 
molt stage pleopods. 



Discussion 

There have long been problems associated with 
growth rate and functional maturity determinations 
in American lobsters. Reliable data on annual pro- 
portions molting (or molt frequency) and proportions 
laying eggs in relation to size are difficult to obtain. 
Both these parameters are essential in assessing the 
impact of size limit and/or exploitation rate changes 







Number 


of molt predictions/va 


lidat 


ions 










Ma 


les 








rem3les 


1 




Pleopod 




Cor- 




Cor- 






Cor- 




Cor- 


stage 


Yes 


rect 


No 


rect 


Yes 




rect 


No 


rect 



















14 


9 


1.0 






1 


1 








2 


2 


1.5 






8 


5 








8 


4 


2.0 






2 


1 








3 


1 


2.5 


7 


4 






2 




2 






3.0 


14 


11 
















3.5 


13 


13 






1 




1 






4.0 


11 


1 1 






3 




3 






4.5 


1 


1 
















5.0 


2 


2 
















5.5 


5 


5 

















'This table does not include 69 females which were ovigerous with old eggs at the 
time of sampling/tagging, all of which subsequently molted 



Table 2. — Summary of egg extrusion predic- 
tions and subsequent validations for female 
American lobsters sampled and tagged at 
Arnold's Cove, Newfoundland, 24 June- 1 7 July, 
1981. Sixty-nine (69) females which were 
ovigerous with old eggs at the time of sampling/ 
tagging, all of which subsequently molted, are 
not included in the table. 









Number of egg 


extrusion 




Cement g 


land 




predictions/va 


lidat 


ions 




stage 


Yes 


Correct 




No 


Correct 













8 


8 


1 










9 


9 


2 




9 


7 








3 




7 


7 









in a lobster fishery on yield per recruit and reproduc- 
tive potential. Such assessments are important to 
proper lobster fishery management. 

The techniques used here to predict molting and 
egg extrusion provide new approaches to the study of 
lobster growth and maturity that have only recently 
been used in studies of lobster populations. Results 
of this validation study, however, clearly indicate that 
caution has to be used in their application. 

In the case of molt prediction it appears that the 
time of sampling in relation to the molting period is 
critical. The ideal situation would be a very short 
annual molting period with sampling just prior to the 
start of molting when all animals going to molt would 
have well-developed (stage 3 or higher) pleopods. 
American lobsters reach the northern limit of their 
range in Newfoundland waters, and it is probably 
here that their annual molting period is the shortest. 
In the Arnold's Cove area, molting starts early in July 
and is virtually completed by early September. In the 
present study, 5 out of 1 4 lobsters (all females, Table 
1), sampled and tagged between 24 June and 17 July 
1981 and had stage pleopods (for which it was pre- 



243 



dieted that molting would not occur that year), had 
molted when recaptured prior to the molting period 
the following year. For these animals premolt 
development must have occurred very rapidly during 
the 1 98 1 molting period. This indicates that periodic 
sampling throughout the molting period along with a 
validation study are required in order to use these 
molt prediction techniques as a basis for estimating 
annual proportions molting in a lobster population. 
The overall success rate with predicting egg extru- 
sion was much greater than with molt prediction 
(94% cf. 78%). The small number of incorrect predic- 
tions may have resulted from loss of eggs rather than 
failure of the animals to extrude. One of 6 ovigerous 
females with newly laid eggs that were tagged during 
the 24 June-1 7 July sampling period had molted and 
was nonovigerous when recaptured. While egg extru- 
sion prediction based upon the cement gland staging 
technique provides a reliable basis for estimating 
annual proportions laying eggs in a lobster popula- 
tion, it is clear that such estimates should be ad- 
justed, using the kind of information that can be 
obtained from a validation study before being used in 
an assessment of reproductive potential in a 
population. 



Ennis, G. P. 

1984. Comparison of physiological and functional size- 
maturity relationships in two Newfoundland populations 
of lobsters Homarus americanus. Fish. Bull, U.S. 82: 
244-249. 

Robinson, D. G. 

1979. Consideration of the lobster (Homarus americanus) 
recruitment overfishing hypothesis; with special 
reference to the Canso Causeway. In F. D. McCracken 
(editor), Canso marine environment workshop, Part 3 of 4 
Parts, Fishery impacts, p. 77-99. Fish. Mar. Serv. Tech. 
Rep. 834. 

SCARRATT, D. J., AND P. F. ELSON. 

1965. Preliminary trials of a tag for salmon and lobsters. J. 
Fish. Res. Board Can. 22:421-423. 
Waddy, S. L., andD. E. Aiken. 

1980. Determining size at maturity and predicting egg extru- 
sion from cement gland development in Homarus 
americanus. CAFSAC (Can. Atl. Fish. Sci. Advis. 
Comm.) Res. Doc. 80/43, 9 p. 



G. P. ENNIS 



Department of Fisheries and Oceans 

Fisheries Research Branch 

P.O. Box 5667 

St. John's. Newfoundland, Canada A 1C 5X1 



Acknowledgments 

I am grateful to S. L. Waddy for her courtesy and 
cooperation in teaching P. W. Collins the pleopod 
and cement gland staging techniques. I am indebted 
to Collins who, in addition to examining all the 
pleopods in this study, participated in the field work 
involved in obtaining samples, tagging, and recovery 
of tagged animals and provided the data summaries. 
Assistance with field work was provided by G. Dawe 
and D. G. Badcock to whom I am also very 
grateful. 

Literature Cited 

AlKEN, D. E. 

1973. Proecdysis, setal development, and molt prediction in 
the American lobster (Homarus americanus). J. Fish. 
Res. Board Can. 30:1337-1344. 
1980. Molting and growth. In J. S. Cobb and B. F. Phillips 
(editors), The biology and management of lobsters. Vol. I, 
Physiology and behavior, p. 91-163. Acad. Press, N.Y. 
Aiken, D. E., and S. L. Waddy. 

1982. Cement gland development, ovary maturation, and 
reproductive cycles in the American lobster Homarus 
americanus. J. Crust. Biol. 2:315-327. 

Campbell, A., and D. G. Robinson. 

1983. Reproductive potential of three American lobster 
(Homarus americanus) stocks in the Canadian Mari- 
times. Can. J. Fish. Aquat. Sci. 40:1958-1967. 



COMPARISON OF PHYSIOLOGICAL AND 

FUNCTIONAL SIZE-MATURITY 

RELATIONSHIPS IN TWO 

NEWFOUNDLAND 

POPULATIONS OF LOBSTERS 

HOMARUS AMERICANUS 

Lobster (genus Homarus) fisheries are characterized 
by excessive exploitation rates and small, minimum 
legal sizes in relation to sizes at maturity 
(Anonymous 1977, 1979). Under such conditions, 
widespread recruitment overfishing is a distinct 
possibility and in eastern Canada appears to be the 
cause of stock collapses in certain areas (Robinson 
1979). Stock-recruitment relationships as such are 
poorly known for the genus Homarus; however, since 
current levels of landings are well below historical 
levels in most fisheries, it is reasonable to assume 
that, within the limits of habitat carrying capacity, 
increased egg production will result in increased re- 
cruitment. It is clear that increasing the minimum 
legal size and/or reducing exploitation rates will 
result in increased egg production within a lobster 
stock; however, detailed knowledge of size-fecundity 
and size-maturity relationships is required to pro- 
perly assess the impact of changes in fishery 
regulatory measures on annual egg production within 
a given stock. 



244 



FISHERY Bl'LLKTIN: VOL. 82. NO. 1. 1984. 



Size-maturity relationships, based mainly on obser- 
vations of ovary color and ova size in nonovigerous 
females for five Newfoundland lobster populations, 
indicate 100% maturity (physiological) for sizes at 
which tagging results show that substantially < 100% 
of the nonovigerous females lay eggs in a given 
spawning season (Ennis 1980). Resorption of the 
mature ovary near the expected time of oviposition is 
a common phenomenon in//, americanus (Aiken and 
Waddy 1980a) and presumably is the main reason for 
failure on the part of physiologically mature females 
to express their maturity by extruding eggs. Clearly, 
it is an "expressed" or functional size-maturity 
relationship that is required to assess the impact of 
size limit and/or exploitation rate changes in a fishery 
on annual egg production. Using the pleopod cement 
gland staging technique described by Aiken and 
Waddy (1982) as a basis for predicting egg extrusion, 
such a relationship was derived for two Newfound- 
land populations. These are compared with physio- 
logical size-maturity relationships for the same 
populations. 

Materials and Methods 




52° 

-I — 



52° 



50° 



48° 



$2° 



FIGURE 1. — Map of Newfoundland showing location of Arnold's 
Cove and Comfort Cove. 



Pleopods were obtained from 172 nonovigerous 
female lobsters caught between 24 June and 17 July 
1981 and 77 caught between 14 and 18 June 1982 
near Arnold's Cove, Placentia Bay, and 246 caught 
between 1 and 7 July 1982 at Comfort Cove, Notre 
Dame Bay, Newfoundland, (Fig. 1) using traps and by 
scuba diving. Sizes ranged from 40 to 111 mm CL 
(carapace length) at Arnold's Cove and from 58 to 
113 mm at Comfort Cove. Pleopods were examined 
for molt stage according to the method of Aiken 
(1973) and for cement gland development according 
to the method of Aiken and Waddy (1982) to deter- 
mine whether molting or egg extrusion would occur 
during the current molting/spawning period. In this 
study it was predicted that females with cement 
glands in stages and 1 would not extrude eggs dur- 
ing the current spawning period whereas those with 
stage 2 or higher cement glands would (see Aiken 
and Waddy 1982 for descriptions of cement gland 
stages). A validation study (Ennis 1983) has 
demonstrated that egg extrusion prediction based on 
cement gland staging is quite reliable. Of the predic- 
tions that could be validated, 947c were correct. The 
only incorrect predictions were for females with stage 
2 cement glands of which 2 out of 9 (22%) failed to 
extrude eggs. Accordingly, in the data analyzed here 
the number of animals with stage 2 cement glands in 
each size group was reduced by 22% to obtain a more 
accurate estimate of the number that would actually 



extrude eggs. Where 22% of the number was < 0.5, 
nothing was subtracted. 

The two Arnold's Cove samples were combined. 
For each area the numbers examined and numbers 
functionally mature (i.e., going to extrude eggs during 
the current season) were grouped by 1 mm CL and 
subjected to probit analysis. Although good statisti- 
cal fits were obtained (P values >0.9), the fitted 
curves did not approximate the data very well at the 
upper and lower ends. Proportions from the same 
data were analyzed using the logistic equation 



Y = 



1 +e b+cX 



(1) 



An SAS 1 program, which performs this analysis by 
means of a nonlinear regression procedure using the 
Marquardt method, was used. Curves were obtained 
with substantially improved visual fits to the data. 

Previously published size-maturity relationships 
for Arnold's Cove and Comfort Cove lobsters (Ennis 
1980) were based mainly on detailed examination of 
the gonads of nonovigerous females, but ovigerous 
females in the samples were included as mature 
animals. For this paper the ovigerous specimens 
were excluded from these samples and the data 



*SAS User's Guide: Statistics, 1982 ed. SAS Institute Inc., Cary, 
N.C., 584 p. 



245 



reanalyzed using the above equation. The size 
maturity relationships thus derived are a more 
accurate reflection of the proportions of non- 
ovigerous females whose gonads are developing for 
extrusion during the upcoming spawning season (i.e., 
physiologically mature). 

Results 

The smallest female lobsters with cement glands in 
stage 2 (or higher), indicating that egg extrusion 
would occur during the current spawning period, 
were 73 mm CL at Arnold's Cove and 7 1 mm at Com- 
fort Cove (Tables 1, 2). All smaller animals had stage 
or 1 cement glands, indicating that egg extrusion 
would not occur. The largest female lobsters with 
cement glands in stage or 1 were 96 mm CL at 
Arnold's Cove and 88 mm at Comfort Cove. All larger 
animals had stage 2 (or higher) cement glands. 

Functional and physiological size-maturity re- 
lationships were derived for each area and plotted 
together (Figs. 2, 3). Sizes at 50% functionally mature 
female lobsters from the relationships were 81 mm 
CL at Arnold's Cove and 80 mm at Comfort Cove. 
These compare with sizes at 50% physiologically 



mature female lobsters of 74 mm and 76 mm for 
Arnold's Cove and Comfort Cove, respectively. 

Observations taken from the data indicate that at 
Arnold's Cove the shift in physiological maturity 
from none to all occurred over a 9 mm CL size range 
(71-80 mm) compared with a 25 mm size range (72- 
97 mm) for functional maturity. The equivalent size 
ranges for Comfort Cove lobsters were 22 mm CL 
(64-86 mm) for physiological maturity and 23 mm 
(70-93 mm) for functional maturity. Examination of 
the fitted curves shows considerable disparity be- 
tween proportions of physiologically mature and 
functionally mature lobsters at given sizes over much 
of the size range in each area. In order to quantify this 
disparity, points on the curves were treated as num- 
bers (out of 100) rather than percentages and the dif- 
ference determined between the two curves at any 
given size. The greatest disparities were for 73 mm 
CL lobsters at Arnold's Cove (Fig. 2) and for 70 mm 
lobsters at Comfort Cove (Fig. 3) where this com- 
parison of the curves indicates that 60% and 41%, 
respectively, of the physiologically mature animals 
fail to extrude eggs. This percentage decreases with 
increasing size in each area. To derive an estimate of 
this percentage for the population as a whole, the 



Table 1 . - Cement gland stages for female lobsters 
caught at Arnold's Cove, Newfoundland, 24 June - 
17 July 1981 and 14-18 June 1982. 



Carapace length 




Cement 


gland si 


age 






(mm) 





1 


2 


3 


4 


Total 


40-69 


31 










31 


70 


2 


1 








3 


72 


I 










1 


73 


3 


1 


1 


1 




6 


74 


2 










:• 


75 


2 










2 


/)_, 


2 


3 






2 


7 


77 


5 




3 


3 


4 


15 


78 


3 


4 


3 


3 


2 


15 


79 


3 


6 


7 


6 


9 


31 


80 


3 


3 


B 


4 


6 


24 


81 


2 


6 


9 


4 


1 


22 


82 


1 


1 




1 


1 


4 


83 


4 


2 


5 


1 


1 


13 


84 




1 


1 


1 




3 


85 


2 


2 


3 


4 


1 


12 


86 




2 


1 


3 


1 


7 


87 




1 


1 


1 


3 


6 


88 


1 




9 




1 


1 1 


89 


1 


1 


3 


1 




6 


90 






2 


1 




3 


91 




1 


3 






4 


92 






I 






1 


93 






1 


1 




2 


94 






1 






1 


95 








2 


1 


3 


96 


1 




1 


2 




4 


97 






1 


1 




2 


98 






2 


2 




4 


100 








1 




1 


102 








2 




2 


107 






1 






1 


109-111 








1 


1 


2 



TABLE 2. — Cement gland stages for female lobsters 
caught at Comfort Cove, Newfoundland, 1-7 July 
1982. 



Carapace length 




Cement 


gland 


stage 




(mm) 





1 


2 


3 4 


Total 


58-69 


7 


1 






8 


70 


2 


1 






3 


71 


1 


1 


1 




3 


72 


3 


1 






4 


73 


2 




1 




3 


74 


1 


2 


1 




4 


75 






5 




5 


76 


1 


2 






3 


77 






2 


1 


3 


78 


1 


2 


1 




4 


79 


1 


1 


7 




9 


80 




1 


7 




8 


81 




2 


7 




9 


82 


3 


2 


8 




13 


83 


1 


4 


16 


4 


25 


84 




4 


15 




19 


85 


1 


2 


15 


1 


19 


86 




4 


4 


4 


12 


87 


1 


2 


8 


6 


17 


88 




1 


13 


1 


15 


89 






8 


1 


9 


90 






5 


5 


10 


91 






4 


1 


5 


92 






3 


2 


5 


93 






2 


4 


6 


94 






1 


2 1 


4 


95 






1 


1 


2 


96 








4 


4 


97 








1 


1 


98 






2 




2 


100 






1 


2 


3 


101-1 13 








7 2 


9 



246 



100 


i 






I 


I 1 


1 


I T 1 1 1 

X^_X-*-*-X — X-» — * — X 


I 1 
— x x-x 




90 














/ x / 




- 


80 














/ / 




- 


70 






PHY 
1  


SI0L0GICAL MATURITY 
10404 


\ 


/ / x 

/ / x 

IX X 
IX I 




- 




14-0943 

l + e 


-0I896X 


£ 60 

>— 

3 








N = 167 






x / 

/ " 

X 




" 


-50 














X XX 




- 


O 














x / x 






"" 40 


















- 


30 












X 

X 


/ \ FUNCTIONAL MATURITY 
' N 0-9694 




- 




15 9542 -0-1983 X 
l + e 


20 














N = 230 




- 


10 


















- 





— 1 


=r »Hr«- 




"fx XX 

1 


XXX X 

1 1 


X XX 

1 


i i i i i 








i i 





50 55 60 65 70 75 80 85 90 95 100 105 110 115 

CARAPACE LENGTH (mm) 



FlGl'RK 2. — Physiological and functional size-maturity relationships for female lobsters at Ar- 
nold's Cove, Newfoundland. Functional maturity data only are provided. 



i 1 r 

X D M X » » — « - «   X « « 



100 



90 



80 



70 - 



60 



50 



40 - 



30 - 



20 



10 



PHYSIOLOGICAL MATURITY 
10254 

16-0316 -0-21 10 X 
l + e 

N = 250 




FUNCTIONAL MATURITY 
. 0-9801 

' 15-9997 -0-2019 X 

l + e 

N=246 



XXXXXXXX X 

I I 1 



50 55 



65 70 75 80 85 90 95 100 105 110 

CARAPACE LENGTH (mm) 



FIGURE 3. — Physiological and functional size-maturity relationships for female lobsters at Com- 
fort Cove, Newfoundland. Functional maturity data only are provided. 



247 



above procedure was followed for those sizes be- 
tween the largest with 100% functionally immature 
and the smallest with 100% functionally mature 
(from the data) and the numbers added. The result- 
ing estimates were 25% at Arnold's Cove and 20% at 
Comfort Cove. 

Discussion 

This study has demonstrated that, failure on the 
part of physiologically mature female lobsters to 
"express" their maturity by extruding eggs is quite 
common in the wild. Resorption of the mature ovary 
near the expected time of extrusion appears to be the 
main reason. Resorption occurs when the molting 
and reproductive cycles conflict (Aiken and Waddy 
1976, 1980a, b). These cycles are normally syn- 
chronized by temperature and photoperiod regimes 
so that conflict between them is minimized. 
However, final ovary maturation is disrupted, if it 
coincides with middle to late premolt, and the ovary 
is resorbed prior to the impending molt. Not only 
would this ensure the conservation of energy, but it 
might also serve to resynchronize the molt and re- 
productive cycles (Aiken and Waddy 1980b). 

Nonfertilization may also be a cause of resorption. 
In Jasus lalandii, for example, oviposition will not 
occur in unfertilized females (Heydorn 1969). While 
oviposition will occur in H. americanus even if the 
female has not successfully mated (Aiken and Waddy 
1980a), it is not clear if this is the rule or the excep- 
tion. Physiologically mature H. americanus females 
which are unfertilized (i.e., empty seminal recep- 
tacles) occur in the wild (Krouse 1973; Ennis 1980). 
In sampling from January to June 1973 at St. Chads, 
Bonavista Bay, on the northeast coast of Newfound- 
land, Ennis (1980) found 6 (11.5%) of 52 
physiologically mature females to be unfertilized. At 
Arnold's Cove in August and September 1981, 98 of 
100 females >79 mm CL were fertilized as deter- 
mined by the presence of spermatophores in seminal 
receptacles. While nonfertilization may be a con- 
tributing factor in some areas, it does not appear to 
be a major cause of ovary resorption in wild H. 
americanus. 

A validation study (Ennis 1983) has demonstrated 
that the cement gland staging technique enables a 
reliable prediction of whether a female lobster will 
extrude eggs during the upcoming spawning season. 
However, caution has to be exercised in applying a 
functional size-maturity relationship based on these 
predictions because there is substantial loss of eggs 
subsequent to spawning. For example, 2 of 15 
females with well-developed (stages 3 and 4) cement 



glands, indicating extrusion to be imminent, and 1 of 
6 females with newly laid eggs (all tagged during the 
24 June to 17 July 1981 sampling period at Arnold's 
Cove) had molted and were nonovigerous when 
recaptured prior to the 1982 molting/spawning 
period. 

There is also substantial loss of eggs other than 
through molting. Some of this loss may be the result 
of eggs not being fertilized. Unfertilized eggs do not 
attach securely and may be lost soon after oviposi- 
tion, but in some cases a fair number will remain 
attached for several months (Aiken and Waddy 
1980a, b). However, it is common for fertizlied eggs 
to be lost as well (Aiken and Waddy 1980a , b). Nor- 
mal attrition of properly attached (fertilized) eggs 
over the 9-12 mo incubation period has been 
estimated at around 36% (Perkins 1971); however, 
some females lose up to 100% of their eggs. The six 
ovigerous females referred to above (i.e., tagged dur- 
ing 24 June to 17 July 1981 at Arnold's Cove) had 
apparently normal clutches of eggs when tagged, but, 
of the five that had eggs when recaptured, four had 
normal clutches and one had < 200 eggs remaining. A 
normal clutch for this particular animal, which was 79 
mm CL, would have been about 10,000 eggs (Ennis 
1981). Similar observations were made on animals 
tagged between 1 and 14 August 1981 at Arnold's 
Cove. Of six females with newly laid, normal-sized 
clutches of eggs, one had just a few hundred eggs 
remaining when recaptured. Another female, which 
had well-developed (stage 4) cement glands, had no 
eggs but had pleopods covered with cement when 
recaptured, indicating that eggs had been extruded 
and subsequently lost (Templeman 1940). 

These observations demonstrate that there is sub- 
stantial loss of eggs subsequent to extrusion over and 
above that attributed to normal attrition. This loss of 
eggs should be taken into account in any assessment 
of the impact of changes in fishery regulatory 
measures on reproductive potential (i.e., annual egg 
production) in a population. 

Acknowledgments 

I am grateful to P.W. Collins who was responsible 
for collecting the samples and examining the 
pleopods for molt stage and cement gland develop- 
ment and to G. Dawe and D. G. Badcock who assisted 
in the collection of the samples. 

Literature Cited 

Aiken, D. E. 

1973. Proecdysis, setal development, and molt prediction in 



248 



the American lobster (Homarus americanus). J. Fish. 
Res. Board Can. 30:1337-1344. 

AIKEN, D. E., AND S. L. WADDY. 

1976. Controlling growth and reproduction in the American 
lobster. In J. W. Avault, Jr. (editor), Proceedings of the 
7th Annual Meeting World Mariculture Society, p. 415- 
430. Louisiana St. Univ. Press, Baton Rouge. 

1980a. Reproductive biology. In J. S.Cobb and B. F.Phillips 
(editors), The biology and management of lobsters. Vol. I, 
Physiology and behavior, p. 215-276. Acad. Press, N.Y. 

1980b. Maturity and reproduction in the American 
lobster. In V. C. Anthony and J. F. Caddy (editors), Pro- 
ceedings of the Canada-U.S. Workshop on Status of 
Assessment Science for N.W. Atlantic Lobster {Homarus 
americanus) Stocks, St. Andrews, N.B., Oct. 24-26, 1978, 
p. 59-71. Can. Tech. Rep. Fish. Aquat. Sci. 932, St. 
Andrews, Can. 

1982. Cement gland development, ovary maturation, and 
reproductive cycles in the American lobster Homarus 
americanus. J. Crust. Biol. 2:315-327. 

Anonymous. 

1977. Report of the working group onHomarus stocks. ICES 
CM. 1977/K:11, 19 p. 

1979. Report of the Homarus working group. ICES CM. 
1979/K:8, 49 p. 

ENNIS, G. P. 

1980. Size-maturity relationships and related observations 
in Newfoundland populations of the lobster (Homarus 
americanus). Can. J. Fish. Aquat. Sci. 37:945-956. 

1981. Fecundity of the American lobster, Homarus 
americanus, in Newfoundland waters. Fish. Bull., U.S. 
79:796-800. 

1983. Tag-recapture validation of molt and egg extrusion 
predictions based upon pleopod examination in the 
American lobster, Homarus americanus. Fish. Bull., 
U.S. 

Heydorn, A. E. F. 

1969. The rock lobster of the South African west coast Jasus 
lalandii (H. Milne-Edwards). 2. Population studies, 
behaviour, reproduction, moulting, growth and mi- 
gration. S. Afr. Div. Sea Fish. Invest. Rep. 7:1-52. 
KROUSE, J. S. 

1973. Maturity, sex ratio, and size composition of the natural 
population of American lobsters, Homarus americanus, 
along the Maine coast. Fish. Bull., U.S. 71:165-173. 
Perkins, H. C. 

1971. Egg loss during incubation from offshore northern 
lobsters (Decapoda: Homaridae). Fish. Bull., U.S. 69:451- 
453. 
Robinson, D. G. 

1979. Consideration of the lobster (Homarus americanus) 
recruitment overfishing hypothesis; with special 
reference to the Canso Causeway. In F. D. McCracken 
(editor), Canso Marine Environment Workshop. Part 3 of 
4 parts. Fishery impacts, p. 77-99 Fish. Mar. Serv. Tech. 
Rep. 843. 
TEMPLEMAN, W. 

1940. The washing of berried lobsters and the enforcement of 
berried lobster laws. Newfoundland Dep. Nat. Resour. 
Res. Bull. (Fish.) 10,21 p. 

G. P. ENNIS 

Fisheries Reasearch Branch 

Department of Fisheries and Oceans 

P.O. Box 5667 

St. John's, Newfoundland, Canada AlC 5X1 



CONVERSIONS BETWEEN TOTAL, FORK, 

AND STANDARD LENGTHS IN 35 SPECIES 

OF SEBASTES FROM CALIFORNIA 

In recent years, the rockfishes (Scorpaenidae: Sebas- 
tes) of the northeastern Pacific Ocean have been 
investigated extensively. With many institutions 
studying diverse aspects of their biology and 
fisheries, a lack of standardized methods has ham- 
pered attempts to synthesize the data. A particular 
problem has been the reporting of different length 
measurements. To provide the means to convert one 
of these length measurements to another, we report 
here the linear regression statistics necessary for 
conversions in 35 species of Sebastes. 

Specimens were collected from fishery catches be- 
tween Cape Blanco, Oreg., and San Diego, Calif., dur- 
ing 1977-82. The sample included five fish for each 
centimeter of body length throughout the size range 
of each species. Measurements were taken on a 
meter board in millimeters on frozen, then thawed, 
carcasses. Standard length was measured from the 
anterior tip of the upper jaw to the posterior end of 
the vertebral column (Hubbs and Lagler 1970:25); 
fork length was measured from the anterior tip of the 
longest jaw to the median point of the caudal fin; and 
the total length was measured from the most anterior 
tip of the longest j aw to the most posterior part of the 
tail when the caudal rays are squeezed together (Holt 
1959:71). Linear regressions were run on all com- 
binations of the measurements of length. Outliers 
(±3.0 standard deviations) from the line were noted 
by the computer program, then checked for data 
entry error and corrected when possible. If a data 
entry error was not found, an outlier was assumed to 
result from measurement error and the observation 
was deleted. 

Statistics reported for each species arey-intercept 
(a), slope (/?), standard error of estimate (S vx ), cor- 
relation coefficient (r), range in length, and the sam- 
ple size used in the regression (n) (Tables 1-3). 
Estimates of a imply impossible values for the 
dependent variable when the independent variable is 
zero. The impossible results could be caused by ran- 
dom error in estimation of a or nonlinearity for values 
less than those observed. The high values of r and 
examination of scattergrams indicate that the length 
relationships are linear over the observed range of 
values. The standard precaution of limiting the 
application of these regressions to the ranges of 
observed values is advised. To calculate the total 
length (TL) of S. alutus, given a standard length (SL) 
of 250 mm, the regression values from Table 1, total 
length on standard length, are used so that 



FISHERY BULLETIN: VOL. 82. NO. 1. 1984. 



249 



Table 1. — Results of linear regressions of standard length versus total length for Sebastes. 

Measurements are in millimeters. 



Species of 
Sebastes 


n 


r 


Star 
len 


dard 
gth 


Total length 
standard len 


on 
gth 


Standard length 
on total length 


Mm 


Max 


a 


P 


Vx 


a 


P 


Vx 


alutus 


49 


0995 


232 


361 


1454 1 


249 


3.746 


2.056 


0.792 


2.984 


aunculatus 


1 16 


1 000 


72 


426 


-1423 1 


240 


3.787 


1.369 


0806 


3054 


aurora 


43 


0.991 


164 


324 


0.098 1 


220 


4 709 


4.398 


0.806 


3.827 


babcocki 


74 


999 


185 


532 


6.478 1 


196 


4 833 


-4.614 


0.834 


4035 


cam at us 


105 


0999 


75 


292 


3.676 1 


201 


2.206 


-2.866 


0832 


1 836 


caunnus 


113 


0997 


1 1 1 


443 


3 873 1 


209 


5 769 


-0.653 


0820 


4.568 


chlorostictus 


107 


0999 


107 


382 


5316 1 


202 


3 636 


-3.931 


0830 


3.023 


chrysomelas 


60 


0998 


77 


316 


1 007 1 


211 


3 161 


-0.123 


0.822 


2 605 


constellatus 


105 


0999 


148 


365 


4497 1 


175 


3 119 


-3.204 


0849 


2.651 


cramen 


102 


999 


102 


394 


-0.304 1 


266 


4.153 


0.737 


0.788 


3 278 


diploproa 


80 


0999 


87 


308 


1 286 1 


242 


2 718 


-0.740 


804 


2 188 


elongatus 


108 


0998 


107 


317 


15.238 1 


165 


3 543 


-12.144 


0855 


3036 


entomelas 


105 


0998 


194 


435 


9.496 1 


211 


5.679 


-6.296 


0822 


4.679 


fl avid us 


193 


0.997 


191 


453 


0.468 1 


247 


5.700 


1 379 


0.798 


4.558 


goodei 


99 


1 000 


101 


449 


4.199 1 


224 


2 870 


-3.196 


0816 


2.344 


hopkinsi 


71 


0993 


99 


251 


3059 1 


200 


4 788 


-0.195 


0.822 


3 964 


lordani 


145 


0998 


77 


260 


4 610 1 


216 


2 903 


-3.128 


0819 


2 382 


levis 


31 


1 000 


190 


717 


-4.500 1 


248 


4 907 


3.813 


0.801 


3.932 


mahger 


42 


0996 


174 


397 


1463 1 


220 


5 639 


1.120 


0.813 


4.604 


melanops 


138 


0999 


74 


495 


7 724 1 


221 


5.193 


-5 596 


0817 


4.247 


me/anostomus 


8/ 


0.994 


207 


421 


-0954 1 


244 


6897 


4 780 


0.794 


5 508 


mimatus 


109 


0.994 


237 


550 


9629 1 


229 


9.765 


-3.095 


0804 


7 900 


mystmus 


163 


0998 


102 


387 


2.930 1 


238 


5694 


-1.192 


0.804 


4,588 


nebulosus 


69 


0995 


213 


366 


4.294 1 


196 


3.962 


-0.731 


0828 


3296 


ovahs 


83 


0.997 


181 


375 


0550 1 


225 


4.374 


1.329 


0.81 1 


3 558 


paucispims 


163 


0999 


103 


649 


-5.035 1 


262 


7.550 


4882 


0.790 


5.974 


p/nn/ger 


136 


997 


196 


565 


11476 1 


239 


8002 


-7 447 


0803 


6.443 


rosaceus 


83 


0.996 


132 


263 


3.917 1 


199 


2 867 


-1 794 


828 


2 383 


rosenblatti 


104 


999 


132 


428 


9.567 1 


182 


3.653 


-7.374 


0844 


3086 


ruberrtmus 


118 


0996 


203 


565 


5.856 1 


202 


9465 


-1 71 7 


0826 


7.843 


rufus 


26 


0999 


152 


447 


12.946 1 


1 77 


5 963 


-10.316 


0.848 


5.061 


saxtcola 


68 


0999 


109 


288 


3.226 1 


242 


2 456 


-2.252 


0.804 


1.976 


semicmctus 


31 


0.979 


101 


147 


8179 1 


170 


3.617 


-1.752 


0.820 


3027 


serranoides 


129 


0995 


190 


441 


8.292 1 


209 


7 277 


-3542 


0.819 


5988 


witsom 


48 


999 


71 


126 


0572 1 


234 


1 071 


-0.231 


0808 


868 



Table 2. — Results of linear regressions of standard length versus fork length for Sebastes. 

Measurements are in millimeters. 



Species of 
Sebastes 


n 


r 


Star 

lei 


dard 
gth 


Fork 1 
stand 


ength on 
ard length 




Standard length 
on fork length 


Mm 


Max 


a 


P 


Vx 


a 


P 


Vx 


alutus 


48 


0996 


232 


361 


-0281 1 


195 


3024 


2.492 


0.831 


2 521 


aunculatus 


114 


999 


72 


426 


-0.369 1 


228 


4.126 


0575 


0.813 


3 358 


aurora 


44 


993 


164 


324 


-3 046 1 


201 


4 237 


6237 


0.821 


3 502 


babcocki 


76 


0999 


185 


532 


9034 1 


153 


5 190 


-6 860 


865 


4.496 


carnatus 


104 


0999 


75 


292 


4.601 1 


194 


2 425 


-3 613 


836 


2.030 


caunnus 


1 17 


996 


11 1 


448 


5896 1 


187 


6 764 


-2 272 


836 


5 674 


chlorostictus 


107 


0999 


107 


382 


5 289 1 


171 


3.719 


-3 987 


852 


3 173 


chrysomelas 


58 


0997 


77 


226 


1.137 1 


209 


3 209 


-0009 


822 


2.647 


constellatus 


107 


0999 


148 


365 


3 883 1 


.152 


2 964 


-2.774 


0.866 


2 571 


cramen 


103 


0999 


102 


394 


1 390 1 


205 


4.282 


-0.565 


0.828 


3 550 


diploproa 


82 


0999 


87 


308 


2 092 1 


181 


2.627 


-1 460 


0845 


2 223 


elongatus 


116 


0998 


107 


317 


14 186 1 


116 


3469 


-11 724 


892 


3 102 


entomelas 


106 


0997 


194 


435 


1 6 964 1 


124 


5602 


-13 326 


0885 


4 970 


tl avid us 


198 


0998 


191 


453 


-0 918 1 


213 


5367 


2 363 


0.820 


4.412 


goodei 


99 


1.000 


101 


449 


1 515 1 


159 


3 085 


-0988 


0862 


2 660 


hopkinsi 


72 


994 


99 


251 


3.011 1 


153 


4 465 


-0.372 


0856 


3 847 


jordani 


154 


0998 


77 


260 


5.645 1 


124 


2.519 


-4 418 


0887 


2 238 


levis 


34 


999 


190 


717 


033 1 


1 77 


8 446 


0688 


848 


7.169 


maliger 


41 


0997 


1 74 


397 


11835 1 


173 


4867 


-8202 


0848 


4.138 


melanops 


135 


0999 


74 


495 


7 149 1 


197 


5.042 


-5 247 


0834 


4 209 


me/anostomus 


86 


0.994 


207 


421 


-0.828 1 


201 


6.853 


4.912 


0822 


5670 


mimatus 


106 


0.994 


237 


550 


16 442 1 


168 


9.200 


-9445 


0.847 


7836 


mystmus 


164 


0998 


102 


387 


352 1 


192 


4.975 


644 


0836 


4.166 


nebulosus 


71 


993 


213 


366 


6934 1 


181 


4623 


-1.852 


0835 


3 888 


ovahs 


83 


0996 


181 


375 


-3 554 1 


187 


4 677 


5 130 


836 


3925 


paucispims 


162 


0999 


103 


649 


-4 082 1 


209 


6819 


4 183 


826 


5636 


pmniger 


138 


0998 


196 


565 


12880 1 


164 


7,440 


-9 326 


855 


6378 


rosaceus 


83 


0997 


132 


263 


1399 1 


187 


2.730 


0.225 


837 


2 293 


rosenblatti 


104 


0999 


132 


428 


9 938 1 


147 


3.347 


-8.023 


0870 


2.915 


ruberrtmus 


1 18 


996 


203 


565 


6 665 1 


181 


9028 


-2 664 


0841 


7.620 



250 



TABLE 2.— Continued 









Stand 


ard 




Fork length 


on 


Standard length 


Species of 
Se hastes 






leng 


rh 


standard ten 


gth 


on 


fork length 


n 


r 


Mm 


Max 


f> 


fi 


Vx 


a 


P Vx 


rufus 


26 


0999 


152 


447 


14 246 


1.112 


4.416 


-12 392 


898 3 969 


saxicola 


77 


0.999 


109 


288 


3 234 


1 200 


2.511 


-2 315 


0831 2.090 


semicmclus 


31 


0.978 


101 


147 


6486 


1.128 


3 562 


-0.343 


0849 3.091 


serranoides 


126 


0995 


190 


441 


4 422 


1.184 


6.779 


-0672 


0837 5.700 


wilsoni 


53 


0999 


71 


126 


0671 


1 203 


0884 


-0.372 


0.830 0.734 



Table 3. — Results of linear regressions of fork length versus total length for Sebastes. 
Measurements are in millimeters. 









Fork 


Total 


ength on 




Fork I 


ength on 


Species of 
Sebastes 






length 


fork length 




total length 




n 


r 


Mm 


Max 


a 


P 


Vx 


a 


P 


Vx 


alutus 


48 


0999 


278 


430 


-0.003 1 


050 


1 483 


1.321 


0.949 


1.272 


aunculatus 


113 


1 000 


90 


529 


-0.586 1 


007 


1 637 


0634 


0.993 


1.626 


aurora 


43 


0998 


198 


388 


2 293 1 


019 


2 349 


-0.917 


0.977 


2 300 


babcocki 


72 


1.000 


222 


635 


-1.146 1 


032 


2.392 


1.336 


0968 


2.316 


carnatus 


101 


1.000 


92 


351 


-0.759 1 


005 


0.510 


0.768 


0.995 


0.507 


caurinus 


107 


0999 


135 


538 


0629 1 


010 


3.022 


0005 


0988 


2 988 


chlorostictus 


106 


1 000 


127 


449 


-0 723 1 


028 


1.905 


0858 


0.972 


1.852 


chrysomelas ' 






















constellatus 


104 


1.000 


174 


422 


-0 134 1 


023 


1.504 


0.301 


0.977 


1.470 


era men 


99 


1000 


124 


480 


-1.700 1 


051 


2 


002 


1 756 


0.951 


1.904 


diploproa 


80 


1 000 


106 


364 


-0558 1 


049 


1 


704 


0669 


0.953 


1 625 


elongatus 


102 


1.000 


129 


360 


-0 552 1 


047 


1 


449 


0.701 


0954 


1.383 


entomelas 


100 


0999 


231 


496 


-6 845 1 


072 


3 


251 


6954 


0.931 


3.029 


it avid us 


191 


1.000 


226 


551 


2.358 1 


025 


2 


439 


-1 906 


0.974 


2.377 


goodei 


96 


1 000 


122 


527 


2 468 1 


057 


2 


647 


-2.096 


0.945 


2.503 


hopkinsi 


70 


0999 


115 


292 


0002 1 


041 


1 917 


0.428 


0959 


1.840 


jordani 


140 


0999 


89 


296 


-1.872 1 


086 


1 885 


2.036 


0.920 


1.735 


levis 


34 


1 000 


228 


855 


-3.335 1 


055 


4452 


3369 


0.947 


4.219 


matiger 


40 


0999 


215 


480 


-8.696 1 


034 


2 782 


9075 


0965 


2 687 


melanops 


132 


1.000 


91 


599 


1595 1 


017 


2.099 


-1.421 


0983 


2 063 


melanostornus 


82 


0.999 


247 


519 


-0.635 1 


036 


2.181 


1.065 


964 


2.103 


mmtatus 


103 


0999 


293 


654 


-7 857 1 


054 


4638 


8 665 


0946 


4394 


mystmus 


158 


1 000 


122 


463 


2 495 1 


039 


2.329 


-2 164 


0962 


2.241 


nebulosus 


71 


1 000 


256 


498 


0.854 1 


001 


1.423 


-0.487 


0998 


1.420 


ova lis 


78 


0999 


225 


438 


3.914 1 


033 


1.996 


-3.311 


0967 


1.931 


paucispinis 


157 


1.000 


123 


781 


-0.870 1 


045 


2.273 


0.930 


0.956 


2.174 


pinniger 


132 


1 000 


235 


586 


-4.107 1 


070 


2.822 


4.108 


0934 


2 638 


rosaceus 


79 


0999 


158 


316 


1.409 1 


015 


1.173 


-1.085 


0984 


1.155 


rosenblatti 


103 


1.000 


155 


497 


-0453 1 


030 


2.026 


0692 


0970 


1 966 


rubernmus 


118 


1.000 


243 


680 


-0 758 1 


018 


3.640 


1.296 


0.981 


3.573 


rufus 


24 


1.000 


182 


517 


-2.197 1 


057 


1 659 


2.135 


0946 


1.569 


saxicola 


69 


0999 


136 


347 


-0.669 1 


038 


2009 


0.921 


0963 


1.935 


semicmctus 


29 


0998 


119 


174 


-0 422 1 


050 


1 178 


1.010 


949 


1.120 


serranoides 


125 


0999 


222 


518 


1419 1 


029 


2.623 


-0862 


0.971 


2 548 


wilsoni 


45 


1.000 


86 


151 


-1.141 1 


035 


0.560 


1.182 


966 


0.541 



1 No regression was run because total length and fork length are equal. 



TL = a + p (SL) 

TL= 1.454 + (1.249) (250) 

TL= 313.7 mm. 



HlJBBS, C. L.. AND K. F. Lagler. 

1970. Fishes of the Great Lakes region. Univ. Michigan 
Press, Ann Arbor, 213 p. 



Literature Cited 

Holt, S. J. 

1959. Report of the international training center on the 
methodology and technqiues of research on mackerel 
(Rastrelliger). FAO/ETAP Rep. 1095, 129 p. 



Tina Echeyerria 
William H. Lenarz 



Southwest Fisheries Center Tiburon Laboratory- 
National Marine Fisheries Service, NOAA 
3150 Paradise Drive 
Tiburon, CA 94920 



251 



NOTICES 

NOAA Technical Reports NMFS published during first 6 months of 1983 

Circular 

448. Synopsis of biological data on the grunts Haemulon aurolineatum and H. 
plumieri (Pisces: Haemulidae). By George H. Darcy. February 1983, iv + 
37 p., 33 figs., 26 tables. 

449. Synopsis of biological data on the pigfish, Orthopristis chrysoptera (Pisces: 
Haemulidae). By George H. Darcy. March 1983, iv + 23 p., 22 figs., 15 
tables. 

450. The utility of developmental osteology in taxonomic and systematic 
studies of teleost larvae: A review. By JeanR.Dunn. June 1983, iii+ 19p., 
7 figs., 5 tables. 

Special Scientific Report — Fisheries 

761. Sea level variations at Monterey, California. By Dale Emil Bretschneider 
and Douglas R. McLain. January 1983, iii + 50 p., 16 figs., 3 tables, 
App. A, B. 

762. Abundance of pelagic resources off California, 1963-78, as measured by an 
airborne fish monitoring program. By James L. Squire, Jr. February 1983, 
v + 75 p., 65 figs., 4 tables. 

763. Climatology of surface heat fluxes over the California Current region. By 
Craig S. Nelson and David M. Husby. February 1983, iii + 155 p., 21 figs., 
1 table, App. I, II, III. 

764. Demersal fishes and invertebrates trawled in the northeastern Chukchi 
and western Beaufort Seas, 1976-77. By Kathryn J. Frost and Lloyd F. 
Lowry. February 1983, iii + 22 p., 4 figs., 6 tables, App. A. 

765. Distribution and abundance of larvae of king crab, Paralithodes 
camtschatica, and pandalid shrimp in the Kachemak Bay area, Alaska, 1972 
and 1976. By Evan Haynes. April 1983, iii + 64 p., 29 figs., 1 table, 
App. 

766. An atlas of the distribution and abundance of dominant benthic inverte- 
brates in the New York Bight apex with reviews of their life histories. By 
Janice V. Caracciolo and Frank W. Steimle, Jr. March 1983, v + 58 p., 69 
figs., 5 tables. 

767. A commercial sampling program for sandworms, Nereis virens Sars, and 
bloodworms, Glycera dibranchiata Ehlers, harvested along the Maine 
coast. By Edwin P. Creaser, Jr., David A. Clifford, Michael J. Hogan, and 
David B. Sampson. April 1983, iv + 56 p., 16 figs., 30 tables, App. A. 

768. Distribution and abundance of east coast bivalve mollusks based on speci- 
mens in the National Marine Fisheries Service Woods Hole collection. By 
Roger B.Theroux and Roland L.Wigley. June 1983, xvi+ 172 p., 121 figs., 
327 tables. 



Some NOAA publications are available by purchase from the Superintendent of Documents, U.S. Govern- 
ment Printing Office, Washington, DC 20402. Individual copies of NOAA Technical Reports (in limited 
numbers) are available free to Federal and State government agencies and may be obtained by writing to 
Publication Services Branch (E/Al 13), National Environmental Satellite, Data, and Information Service, 
NOAA, 3300 Whitehaven Street, N.W., Washington, DC 20235. 



Ctmtcnts — conlmtivti 

LOVE, MILTON S., GERALD E. McGOWEN, WILLIAM WESTPHAL, ROBERT J. 
LAVENBERG, and LINDA MARTIN. Aspects of the life history and fishery of the 
white croaker, Genyonemus lineatus (Sciaenidae), off California 179 

MORRIS, PAMELA A. Feeding habits of blacksmith, Chromis punctipinnis , associated 

with a thermal outfall 199 

MYRICK, ALBERT C., JR., EDWARD W. SHALLENBERGER, INGRID KANG, and 
DAVID B. MacKAY. Calibration of dental layers in seven captive Hawaiian spinner 
dolphins, Stenella longirostris, based on tetracycline labeling 207 

ROSS, STEVE W. Reproduction of the banded drum, Larimus fasciatus, in North 

Carolina 227 

Notes 

SCHMITT, P. D. Marking growth increments in otoliths of larval and juvenile fish by 

immersion in tetracycline to examine the rate of increment formation 237 

ENNIS, G. P. Tag-recapture validation of molt and egg extrusion predictions based upon 

pleopod examination in the American lobster, Homarus americanus 242 

ENNIS, G. P. Comparison of physiological and functional size-maturity relationships in 

two Newfoundland populations of lobsters Homarus americanus 244 

ECHEVERRIA, TINA, and WILLIAM H. LENARZ. Conversions between total, fork, 

and standard lengths in 35 species of Sebastes from California 249 



# GPO 693-007 



^°^o. 




Sr 4T£S 0< * 



Fishery Bulletin 



"\ 



r 



Vol. 82, No. 2 



April 1984 



ROPES, JOHN W., STEVEN A. MURAWSKI, and FREDRIC M. SERCHUK. 
Size, age, sexual maturity, and sex ratio in ocean quahogs, Arctica islandica Linne, 
off Long Island, New York 253 

BRODEUR, RICHARD D., and WILLIAM G. PEARCY. Food habits and dietary 
overlap of some shelf rockfishes (genus Sebastes) from the northeastern Pacific 
Ocean 269 

COLVOCORESSES, J. A., and J. A. MUSICK. Species associations and community 
composition of Middle Atlantic Bight continental shelf demersal fishes 295 

HAYNES, EVAN B. Early zoeal stages of Placetron wosnessenskii and Rhinolith- 
odes wosnessenskii (Decapoda, Anomura, Lithodidae) and review of lithodid larvae 
of the northern North Pacific Ocean 315 

ZIMMERMAN, ROGER J., THOMAS J. MINELLO, and GILBERT ZAMORA, Jr. 
Selection of vegetated habitat by brown shrimp, Penaeus aztecus, in a Galveston 
Bay salt marsh 325 

STANDARD, GARY W, and MARK E. CHITTENDEN, Jr. Reproduction, move- 
ments, and population dynamics of the banded drum, Larirnus fasciatus, in the 
Gulf of Mexico 337 

TETTEY, ERNEST, CHRISTOPHER PARDY, WADE GRIFFIN, and A. NELSON 
SWARTZ. Implications of investing under different economic conditions on the 
profitability of Gulf of Mexico shrimp vessels operating out of Texas 365 

BUCK, JOHN D. Quantitative and qualitative bacteriology of elasmobranch fish 
from the Gulf of Mexico 375 

BOTTON, MARK L., and HAROLD H. HASKIN. Distribution and feeding of the 
horseshoe crab, Limulus polyphemus, on the continental shelf off New Jersey .... 383 

PEARCY, W, T NISHIYAMA, T. FUJII, and K. MASUDA. Diel variations in the 
feeding habits of Pacific salmon caught in gill nets during a 24-hour period in the 
Gulf of Alaska 391 

RUGGERONE, GREGORY T, and DONALD E. ROGERS. Arctic char predation 
on sockeye salmon smolts at Little Togiak River, Alaska 401 



Notes 



MAULE, ALEC G., and HOWARD F. HORTON. Feeding ecology of walleye, 
Stizostedion vitreum vitreum, in the mid-Columbia River, with emphasis on the 
interactions between walleye and juvenile anadromous fishes 411 

(Continued on back coven 



Seattle, Washington 



U.S. DEPARTMENT OF COMMERCE 

Malcolm Baldnge, Secretary 

NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION 

John V. Byrne, Administrator 

NATIONAL MARINE FISHERIES SERVICE 

William G. Gordon. Assistant Administrator 



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 Stales Fish Commission was begun in 188), it became the Bulletin of the Bureau of Fisheries 
in 19dl and the Fishery Bulletin of the Fish and Wildlife Service in 1941. Separates were issued as documents through volume 46; the 
last document was No. 1 103. 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 ,. 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 exchari nl ific publications 



SCIENTIFIC EDITORS. Fishery Bulletin 



Dr. Carl J. Sindermann 

Northeast Fisheries Center Sandy Hook Laboratory 

National Marine Fisheries Service, NOAA 

Highlands, NJ 07732 



Dr. William J. Richards 

Southeast Fisheries Center Miami Laboratory 

National Marine Fisheries Service, NOAA 

Miami, FL 33149-1099 



Editorial Committee 



Dr. Bruce B. Collette 

National Marine Fisheries Service 

Dr. Edward D. Houde 
Chesapeake Biological Laboratory 

Dr. Merton C. Ingham 

National Marine Fisheries Service 

Dr. Reuben Lasker 

National Marine Fisheries Service 



Dr. Donald C. Malins 

National Marine Fisheries Service 

Dr. Jerome J. Pella 

National Marine Fisheries Service 

Dr. Jay C. Quast 

National Marine Fisheries Service 

Dr. Sally L. Richardson 

Gulf Coast Research Laboratory 



Mary S. Fukuyama, Managing Editor 



■\ Bulletin - published quarterly by the Scientific Publications Office, National Marine 

iinl Way NE, BIN C15700, Seattle, WA 98115. Second class postage is paid at Seattle. Wash., and 
additional offices. POSTMASTER send addre - for subscriptions to Superintendent of Docui S Government 

Printing Office. Washington, DC 20 10 

Although the contents have nm been copyrighted and may be reprinted entirely, reference to source is appreciated. 

The - mmerce has determined that the publication of this periodical is necessary in the transaction of the public 

iiss required by law of this Department. I is for printing of this periodical has been approved by the Director o( the 

Office of Management and Budget through 1 April 19 



Fishery Bulletin 

CONTENTS 

Vol. 82, No. 2 April 1984 

ROPES, JOHN W., STEVEN A. MURAWSKI, and FREDRIC M. SERCHUK. 
Size, age, sexual maturity, and sex ratio in ocean quahogs, Arctica islandica Linne, 
off Long Island, New York 253 

BRODEUR, RICHARD D., and WILLIAM G. PEARCY. Food habits and dietary 
overlap of some shelf rockfishes (genus Sebastes) from the northeastern Pacific 
Ocean 269 

COLVOCORESSES, J. A., and J. A. MUSICK. Species associations and community 
composition of Middle Atlantic Bight continental shelf demersal fishes 295 

HAYNES, EVAN B. Early zoeal stages of Placetron wosnessenskii and Rhmolith- 
odes wosnessenskii (Decapoda, Anomura, Lithodidae) and review of lithodid larvae 
of the northern North Pacific Ocean 315 

ZIMMERMAN, ROGER J., THOMAS J. MINELLO, and GILBERT ZAMORA, Jr. 
Selection of vegetated habitat by brown shrimp, Penaeus aztecus, in a Galveston 
Bay salt marsh 325 

STANDARD, GARY W., and MARK E. CHITTENDEN, Jr. Reproduction, move- 
ments, and population dynamics of the banded drum, Larimus fasciatus, in the 
Gulf of Mexico 337 

TETTEY, ERNEST, CHRISTOPHER PARDY, WADE GRIFFIN, and A. NELSON 
SWARTZ. Implications of investing under different economic conditions on the 
profitability of Gulf of Mexico shrimp vessels operating out of Texas 365 

BUCK, JOHN D. Quantitative and qualitative bacteriology of elasmobranch fish 
from the Gulf of Mexico 375 

BOTTON, MARK L., and HAROLD H. HASKIN. Distribution and feeding of the 
horseshoe crab, Limulus polyphemus , on the continental shelf off New Jersey .... 383 

PEARCY, W, T NISHIYAMA^ T FUJII, and K. MASUDA. Diel variations in the 
feeding habits of Pacific salmon caught in gill nets during a 24-hour period in the 
Gulf of Alaska 391 

RUGGERONE, GREGORY T, and DONALD E. ROGERS. Arctic char predation 
on sockeye salmon smolts at Little Togiak River, Alaska 401 



Notes 



MAULE, ALEC G., and HOWARD F. HORTON. Feeding ecology of walleye, 
Stizostedion vitreum uitreum, in the mid-Columbia River, with emphasis on the 
interactions between walleye and juvenile anadromous fishes 411 



(Continued on next page) 



Seattle, Washington 
1984 



Uborate; 






APR 1 7 1985 



For sale by the Superintendent of Documents, U.S. Government Printing Office, Washing- 
ton DC 20402 — Subscription price per year: $21.00 domestic and $26.25 foreign. Cost per 
single issue: $6.50 domestic and $8.15 foreign. 



Contents — continued 

ROCKETTE, MARK D., GARY W. £¥A*ffiARD, and MARK E. CHITTENDEN, Jr. 
Bathymetric distribution, spawning periodicity, sex ratios, and size compositions 
of the mantis shrimp, Squilla empusa, in the northwestern Gulf of Mexico 418 

CROWE, BARBARA J. Distribution, length-weight relationship, and length- 
frequency data of southern kingfish, Menticirrhus americanus , in Mississippi .... 427 

RADTKE, RICHARD. Scanning electron microscope evidence for yearly growth 
zones in giant bluefin tuna, Thunnus thynnus, otoliths from daily increments . . . 434 

PAYNE, P. MICHAEL, and DAVID C. SCHNEIDER. Yearly changes in abun- 
dance of harbor seals, Phoca vitulina, at a winter haul-out site in Massachusetts . 440 

GOLDBERG, STEPHEN R., VICTOR HUGO ALARCON, and JUERGEN ALHEIT 
Postovulatory follicle histology of the Pacific sardine, Sardinops sagax from 
Peru 443 

SHIMEK, RONALD L., DAVID FYFE, LEAH RAMSEY, ANNE BERGEY, JOEL 
ELLIOTT, and STEWART GUY. A note on spawning of the Pacific market squid, 
Loligo opalescens (Berry, 1911), in the Barkley Sound region, Vancouver Island, 
Canada 445 

EPPERLY SHERYAN P., and WALTER R. NELSON. Arithmetic versus expo- 
nential calculation of mean biomass 446 



Notice 
NOAA Technical Reports NMFS published during last 6 months of 1983. 



The National Marine Fisheries Service (NMFS) does not approve, recommend or 
endorse any proprietary product or proprietary material mentioned in this pub- 
lication. No reference shall be made to NMFS, or to this publication furnished by 
NMFS, in any advertising or sales promotion which would indicate or imply that 
NMFS approves, recommends or endorses any proprietary product or proprietary 
material mentioned herein, or which has as its purpose an intent to cause directly 
or indirectly the advertised product to be used or purchased because of this NMFS 
publication. 



SIZE, AGE, SEXUAL MATURITY, AND SEX RATIO IN 

OCEAN QUAHOGS, ARCTICA ISLANDICA LINNE, 

OFF LONG ISLAND, NEW YORK 

John W. Ropes, Steven A. Murawski, and Fredric M. Serchuk 1 

ABSTRACT 

Ocean quahogs, A rctica islandica. were collected off Long Island. New York, in 1978 for a determination 
of sexuality and gonadal condition. A microscopic examination of histologically prepared tissues of 133 
clams. 19-60 mm in shell length, revealed that 36 were in an undifferentiated condition and could not be 
sexed. Sexual differentiation was evident in 97 clams; of the latter, 69 were in two types of intermediate 
development: those with sparse (20) and moderate (49) tubule development. Only 28 clams were fully 
mature. Age and growth were assessed from acetate peels of shell cross sections. Determinations of sex 
of these, and of specimens 57-103 mm in shell length collected from the same area in 1980. indicated 
that the smallest and youngest ocean quahogs were predominantly male, but the largest and oldest 
were predominantly female. 



Ocean quahogs, Arctica islandica, like most other 
bivalves, lack external characteristics for a de- 
termination of sex, maturation, and gonadal con- 
dition. Sex determination has been made for other 
bivalves, such as the surf clam, Spisula solidis- 
sima (Ropes 1979a), from microscopic examina- 
tions of gametogenesis in histological prepara- 
tions of gonadal tissues. Similar examinations 
were lacking for ocean quahogs. The resource has 
become an important fishery within the past 
half-decade (Ropes 1979b; Serchuk and Murawski 
1980 2 ). 

In most bivalves that have been studied, sexual 
maturity occurs at a young age and small size, but 
species differences have been observed (Altman 
and Dittmer 1972). Thompson et al. (1980a, b) 
found that the ocean quahog is a slow growing, 
long-lived species which exhibits considerable 
variability in maturation with respect to size and 
age. The latter conclusion was based on examina- 
tions of 39 specimens, 87% of which were 40 mm or 
longer in shell length. The samples were collected 
in April-May, 3-4 mo before the spawning period 
reported for this species by Loosanoff (1953). It 
seemed reasonable to assume that mature, older 
quahogs in the sample would produce large num- 



■Northeast Fisheries Center Woods Hole Laboratory. National 
Marine Fisheries Service, NOAA, Woods Hole, MA 02543. 

2 Serchuk, F M.. and S. A. Murawski. 1980. Evaluation and 
status of ocean quahogs, Arctica islandica (Linnaeus) popula- 
tions off the Middle Atlantic coast of the United States. Woods 
Hole Lab. Ref. Doc. 80-32. 4 p. Northeast Fisheries Center 
Woods Hole Laboratorv, National Marine Fisheries Service, 
NOAA, Woods Hole. MA 02543. 



Manuscript accepted September 1983. 
FISHERY BULLETIN: VOL. 82. NO. 2. 1984. 



bers of sex cells, but it was not possible to deter- 
mine whether most of the undifferentiated gonads 
in the sample would do likewise. Their contribu- 
tion to the reproductive potential of the species 
was an enigma, and our knowledge of maturation 
was incomplete. 

In late July and early August 1978, the National 
Marine Fisheries Service marked large numbers 
of ocean quahogs at a location near a site sampled 
in the study of sexual maturity reported by 
Thompson et al. (1980b). This was an opportunity 
to collect specimens for a reexamination of 
gonadal condition at about the time of maximum 
ripeness, as Loosanoff (1953) had reported finding 
many ocean quahogs in the partial spawning con- 
dition in mid-August. The time of collection, then, 
seemed favorable for obtaining sexually mature 
quahogs with fully developed, ripe gonads that 
could be clearly separated from immature 
quahogs with undifferentiated sex cells in the 
gonads. 

METHODS 

A commercial clam dredge vessel, MV Diane 
Maria, was chartered for the marking project dur- 
ing 25 July-5 August 1978. The hydraulic clam 
dredge had a 100-in (2.54 m)-wide knife and was 
modified by lining the inner surfaces with 1/2-in 
(12.7 mm) square-mesh hardware cloth to retain 
small clams. Sample tows were of 4-5 min duration 
and usually resulted in a dredge filled with clams, 
shells, and bottom substrata. 

253 

*1 



FISHERY BULLETIN: VOL 82. NO. 2 



The sample site was 48 km SSE of Shinnecock 
Inlet, Long Island, N.Y., at lat. 40°21'N, long. 
72°24'W, and 53 m deep. This location contained 
high densities and a wide size range of ocean 
quahogs and had a low probability of being dis- 
turbed by the fishery: criteria important for suc- 
cess in the marking experiment (Murawski et al. 
1982). The wide size range of ocean quahogs found 
at and near the site included more small individu- 
als for a study of maturity than elsewhere in the 
Middle Atlantic Bight. 

Small quahogs ( ^ 65 mm shell length) were 
sorted from the catch during the marking opera- 
tion, and the soft bodies were immediately re- 
moved from the shells for preservation in Bouin's 
fixative; shells were saved and coded for reference 
to corresponding tissues. Slides of the gonadal tis- 
sues were prepared for microscopic examination 
using standard histological techniques. The clam 
bodies were cut dorsoventrally through the mid- 
section, and the anterior and posterior pieces of 
each clam were embedded to produce two sections 
for examination. The 6 /xm thick sections were 
stained in Harris' hematoxylin and eosin. Recog- 
nition of gametogenic stages was based on previ- 
ous studies of bivalve reproduction by Loosanoff 
( 1953 ); Ropes and Stickney ( 1965 ); Ropes ( 1968a, b; 
1971; 1979a ); Thompson et al. ( 1980b ); Jones ( 1981 1; 
and Mann (1982). 

The shells were processed for observation of 
internal age/growth lines in acetate peels by 
methods similar to those reported in Thompson et 
al. 1 1980a, b) and reported more fully by Ropes 
(1982) 3 . A radial section was made from the umbo 
to ventral margin of left valves, since these contain 
a single prominent tooth that Thompson et al. 
( 1980a, b) found had growth lines corresponding in 
number to those in the valve. Proper orientation of 
the valve for sectioning to retain the umbonal por- 
tion and broadest tooth surface in the anterior 
portion of the valve was a critical procedural step. 
The sections were made on a low-speed saw and by 
a 10.2 cm diameter by 0.03 cm thick diamond wa- 
tering blade. The cut edges were hand polished on 
wetable carborundum paper (240, 400, and 600 
grits) to remove saw marks, polished to a high 
luster on a vibrating lap machine charged with 
aluminum oxide, then etched in a \ c /< HC1 solution 
for one min. Peels were produced by flooding the 



;) Ropes, J. W. 1982. Procedures for preparing acetate peels 
of embedded valves of Arctica islandica for ageing. Woods Hole 
Lab. Ref. Doc. 82-18, 8 p. Northeast Fisheries (inter Woods 
Hole Laboratory, National Marine Fisheries Service. NOAA, 
Woods Hole, MA 02543 



etched surfaces with acetone and applying 0.127 
mm thick acetate film. After a 15-min drying 
period, the film was peeled off and sandwiched 
between glass slides. Peel images were enlarged 
on a microprojector to 40 x . Age/growth lines were 
counted and the exit location of each at the exter- 
nal edge was marked on the peel for a comparison 
with the external bands by placing the anterior 
valve portion on the peel image. This procedure 
clearly demonstrated correspondence between the 
number and location of internal lines and external 
bands. It also delimited sequential increments be- 
tween external bands for measurement to the 
nearest 0.1 mm with calipers. 

Periodic age/growth phenomena in the shells of 
ocean quahogs have been called "bands" for incre- 
ments of darker periostracum deposits on the ex- 
ternal shell surfaces and "lines" for those accreted 
in the shells. The latter have been identified as 
prismatic microstructures that demark bound- 
aries of growth increments (Ropes et al. in press); 
the external pigmented bands varied in intensity 
and width (from to -2 mm). A slight concentric 
depression often outlined the shell shape in the 
bands and corresponded to the location of internal 
lines. This and the method of marking the acetate 
peel aided in measuring increments of growth. 

After completing the study of the gonadal tis- 
sues of small ocean quahogs, it was evident that 
the sex ratio of larger clams from the same area 
should be examined. Therefore, squashes of 
thawed gonadal tissues from 199 marked clams 
57-103 mm shell length recaptured in August 1980 
were examined microscopically at the laboratory 
for determination of sex. 

RESULTS 

Observations of Age 

The shells and acetate peels of 137 clams were 
examined. Bands on the external shell surfaces 
were not equally distinct for all clams in the sam- 
ple. The bands were widely separated for small 
clams, but crowded together at the ventral margin 
for large clams. A few shells had poorly defined 
bands, but lines in the peels aided in locating 
them. Age annuli formed during the earliest on- 
togeny of ocean quahogs are difficult to detect on 
the valve surface and must be carefully exposed in 
the sectioned shell. A quahog 20.0 mm in shell 
length had three barely detectable bands on the 
surface of its valves; the two most recent annuli in 
peels of the valve and hinge tooth were most obvi- 



254 



ROPES ET AL : SIZE. AGE. AND SEX OF OCEAN QUAHOGS OFF LONG ISLAND. NY 



ous and the first was confounded by a secondary 
incomplete line that had formed slightly later 
(Fig. la, b). The formation of secondary lines is not 
typical at a young age. The formation of a complete 
line is, however, important in detecting annul i. 

Three clams had shell abnormalities related to 
an injury An ocean quahog with six bands had a 
slight depression at the anterior end of the left 
valve that was not detected as unusual growth 
lines or increments in the peel of either the valve 
or tooth; the right valve showed no evidence of an 
injury (Fig. 2a, b). Another quahog had a deep 
indentation, and part of the ventral margin was 
missing in the left valve before band six had been 
formed, but the right valve showed a slight inden- 
tation and darkening as evidence of an injury ( Fig. 
3a, b). The peel of the left valve showed age lines 
before and immediately after the site of the injury 
(Fig. 3c). The sixth annuli in the hinge tooth was 
very prominent (Fig. 3d). The valve of a quahog 
with seven bands had definite surface indenta- 
tions associated with annuli, and the hinge tooth 
showed regularly spaced growth increments (Fig. 



4a, b, c). An injury was not clearly evident. The 
annuli in peels of all these clams were easily re- 
lated to bands on the valve surface for mea- 
surements of growth. 

For 9 clams (47.5-60.4 mm long), all annuli in 
the peels were counted, but only some bands were 
measured because those near the ventral margin 
were too crowded and poorly defined. 

The shells of 3 clams (39.7-64.0 mm long) pro- 
duced a confused pattern of lines in the ventral 
third of the peels and extensive ridging and poorly 
defined bands on the external valve surfaces (Fig. 
5a, b, c). It was not evident that these clams had 
been injured, but they were omitted in analyses, 
since growth appeared to be aberrant. In all, 134 
clams, 18.7-60.4 mm long and averaging 38.9 mm 
(S.D. ± 8.65), were used. 



5mm 

i i i i 



5 mm . 




FIGURE 1. — <ai Right valve of a 3-yr-old ocean quahog, 
Arctica islandica, 20.0 mm shell length, (b) Photomi- 
crograph of the acetate peel image of the hinge tooth 
showing three annuli. 



i/. 



~A 









^* 




\ 



, 500>im 

I ' i i i 



FIGURE 2. — (a) Right valve of a 6-yr-old ocean quahog, Arctica islan- 
dica. 31.1 mm shell length, ibi Photomicrograph of the acetate peel 
image of the hinge tooth showing six annuli. 



255 



FISHERY BULLETIN: VOL. 82, NO. 2 








•■^ 



5mm 

i i i i 





,500 um, 

1 i i r i I 

FIGURE 3. — iai Right valve of a 6-yr-old ocean quahog, Arctica islandica, 33.1 mm shell length, lb) Sectioned anterior portion of the 
left valve showing injury, ic) Three serial photomicrographs of the acetate peel image. Arrows point to annuli formed before and after 
the injury. (d> Photomicrograph of the acetate peel image of the hinge tooth showing six annuli. 



Size measurements at age of the clams are 
shown in Figure 6. The mean shell length, one 
standard deviation from the mean, and range are 
given for clams 3-8 yr old. The bands on the shells 
and lines in the peels indicated rapid growth 
through age 8. From age 3 and a mean size of 23.4 
mm, the clams increased about 5 mm in shell 
length each year to age 8 and a mean size of 46.1 
mm. Thereafter, growth seemed to decrease in 
rate. The bands were well separated to age 13. The 



bands at the ventral margin of 14-yr-old and older 
clams were too indistinct for accurate mea- 
surements, but the growth lines in peels were 
clearly separated and easily counted. The oldest 
14- to 18-yr-old specimens may have been the 
smallest and slowest growing individuals in their 
year classes, but mean lengths were not progres- 
sively smaller than means for clams 9-13 yr old. 
Thus, a significant bias was not clearly indicated 
in the selection of older specimens. 



256 



ROPES ET AL.: SIZE, AGE, AND SEX OF OCEAN QUAHOGS OFF LONG ISLAND, NY. 





£. 
E. 
10. 









fclD 


c 








. > 




1^ O 




*t\ — 




KjHkV ^ 




299ft -c 








y^t. .« 




\# ° 




y* -*-* 




tC &> 




V t>£ 




&Jy " c 




. A  ~ 




^. ' ^ 




^ * 

^ "^ 






0) 


CO 


> 


£ 


d> 


"~" 


-C 


"3 


4-3 


cu 


c 


a. 




0/ 


■-; 


4-3 
CO 
*3 


3 
C 


cu 


c 


CO 


CO 


CU 


o 

4J 


X 

-l-> 


bo 


c~ 


a 


o 


*2 


J3 


c 


a. 

CO 


— 


bd 


y. 


o 

in 


S 


CJ 


o 




;- 


s 


CO 


o 




+3> 


-C 


o 


^ 


X 


5 


0- 


— 


CD 


,Q 


-- 








CO 




> 




= 


So 


- 


c 




<D 


-a 




cu 


_-. 


C 


"3 


c 


JS 


— 


[C 


o 




cu 


e 


a: 


£ 




cc 


4-> 


iri 


— 
o 


cc 


CO 


.J 


-C 


J^ 


— 




CO 


-^-, 








£ 


;i 


Q 


o 






./; 


o 






a 


E 


w 


c 


*^ 


— 




o 


i- 


~ 


^ 


a. 






SI 


-, 


c 


CU 


CO 


CO 


3 




cr 


~ 


c 


c 


CO 


-3- 


cu 


. 


o 


CJ 


o 




~ 








o 








Sh 


3 


>, 


C 


t^- 


C 


cfl 


CO 


c*-. 


c 


o 


cu 


cu 
_> 


> 

cu 

CO 


"cO 




> 




-*-> 




_c 




M 





w 

OS 
D 

o 



257 



FISHERY BULLETIN: VOL. 82, NO. 2 




£_ 
in 




4 









J3 — 




H ^ 




-— 4-3 




« S 




. o 




■$ & 




1 * 




*■ o> 




0> ""T 




m a. 




a £ 




'3 o 




01 - 




-E — ' 




■w -a 




<*-< c 




° 5 








— ai 




13 "° 




0) >-. 




o-t; 




1/ o 




■y o 




<s a. 




-*-> . 




4) V- 




o o 




CO OJ 




0) c 




-B g 




4_> N 




"3 =° 




-B s 




b- 2 




ca -B 




i» to 




M M 




O *J 




- Cu 




.5 ■* 




r- O 




o ■- 




o w 




-B C 




Oh -5 




_ bo 




X! b 




_ CS 




E 


e 




a,- 


5 s "cO 

.3 Sh 


o : 


"? c 


OJ 

in 


-5 > 

^ 0; 




'« — 




C *J 




~ +J 




*o CO 








^ O 




< z 




 CO 




bo > 









quah 
ioned 




B t3 




cO a; 




1; X 




8 jj 




_C +-> 




f <*- 




be 




Js be 




— CS 








01 B 




x. ■- 




en — * 




_ 4> 




£ 




% 




*^ 




-f g 




S CO 




S^ 




0) -4-> 




> <~ 




lo ° 




> CO 




*> -s 




-B S 1 




.5° 2 




OS gi 




— *■« 




CO u 








1 £ 




'. 




lO -w 









W J= 




« a 




P 




O 



258 



ROPES ET AL.: SIZE. AGE, AND SEX OF OCEAN Ql AH( >GS OFF LONG ISLAND, N.Y. 



5 






6.' 



5.: 



4C 



30 



20- 



10- 



:■■' 



2 



1 



10 



1 3 



29 



27 



W 26 
14 



; 9 



N = 134 

X =3 89 
SO = 8.65 



MEAN 




RANGE 



8 10 12 

AGE IN YEARS 



14 



16 



FIGURE 6. — Observed shell lengths at age of ocean quahogs, 
Arctica islandica, off Long Island, N.Y., late July-early August 
1978. 



Observations of Gonadal Condition 

Gametogenesis 

Gametogenesis in pelecypod molluscs exhibits 
similar basic characteristics. Each reproductive 
cycle begins with the production of the smallest, 
earliest cells at the basement membrane of folli- 
cles or alveoli. These infiltrate the lumina during 
maturation. Spermiogenesis through meiotic di- 
visions is completed within male gonadal alveoli; 
oogenesis undergoes mitotic division of the 
oogonia and growth of the primary oocytes within 
the female gonadal alveoli. Oocytes may reach 
metaphase of the first meiotic division in the ducts 
of spawning females, but are blocked from com- 
pleting maturation until after spawning and 
sperm penetration (Raven 1958). Most pelecypods 
expel the ripe cells into the surrounding environ- 
mental water where fertilization and larval devel- 
opment occur. A few pelecypods, and most notably 
female oysters of the genus Ostrea, are exceptions, 
since the eggs are held in the inhalent cavity dur- 
ing fertilization and initial developmental stages 



(Yonge 1960). A reproductive cycle corresponds to 
the initiation and completion of gametogenic 
stages and spawning. Single annual cycles have 
been described for many pelecypods, including the 
ocean quahog, although biannual and continuous 
cycles have been discribed for others (Sastry 1979). 
In some species, such as the ocean quahog, succes- 
sive reproductive cycles begin at or soon after 
spawning; in others, activation of a cycle is de- 
layed and the gonads are considered to be in a 
quiescent or resting stage (Sastry 1979). The latter 
condition frustrates determination of sex, since 
secondary sexual characteristics are generally 
lacking in most pelecypods. 

Spermiogenesis 

Spermatogonia about 5.5 /xm in diameter are 
the initial germinal cells produced by male Arc- 
tica islandica during a mitotic phase of sper- 
miogenesis. Successive meiotic stages follow and 
include primary and secondary spermatocytes 
( ~ 3.7 and 4.0 fj.m in diameter, respectively ), sper- 
matids (-2.2 /u.m), and flagellated spermatozoa. 
The respective cells proliferate into the lumina of 
alveoli. Sperm have conical heads —4.8 ^.m long. 

Oogenesis 

Oogonia are the initial germinal cells produced 
by female Arctica islandica during oogenesis. 
These are embedded in the basement membrane 
and are comprised of cytoplasm and a conspicu- 
ous nucleus or germinal vesicle with a basophilic 
nucleolus surrounded by a network of loose chro- 
matin. The distinction between oogonia, sper- 
matogonia, and other cells in the basement mem- 
brane is not obvious. Primary oocytes begin 
protruding from the basement membrane into the 
lumina of alveoli and retain an attachment with it 
during the growth stage. The large spherical, ve- 
sicular nucleus of primary oocytes is surrounded 
by a coarse cytoplasm containing granules of the 
golgi apparatus and acidophilic granules of pro- 
teid yolk (Raven 1958; Kennedy and Battle 1964). 
The nucleolus differentiates into an amphinu- 
cleolus with maturation. Mature oocytes appear 
free in the lumina of alveoli and are often of ir- 
regular shape and have a distinct vitelline mem- 
brane. Measurements of the diameter of 50 clearly 
spherical oocytes that were sectioned through the 
nucleus and amphinucleolus ranged from 49.4 to 
65.0 /ttm and averaged 56.6 fxm. 

Thirty-six gonadal tissues were in an undif- 



259 



FISHERY BULLETIN: VOL. 82, NO 2 



ferentiated condition (Table 1, Fig. 7a, b). Gonadal 
tubules were of small diamater, few in number, 
and surrounded by an abundant loose vesicular 
connective tissue. Gonia embedded in the germi- 
nal epithelium lacked definite cellular structures 
for sex determinations. The lumina of tubules 
were empty. 

Sex determinations were possible for 97 
quahogs, but in most (69) the gonads appeared to 
be in an intermediate stage and not fully devel- 
oped. These latter tissues were separated into two 
categories: Those with either sparse or moderate 
tubule development. 

Differentiated gonads with sparse tubule devel- 
opment were characterized by a limited number of 
gametogenic cells, as well as a limited number of 
tubules. The 16 male tissues examined were pro- 
ducing a few sperm; the 4 female tissues examined 
were producing a few oocytes. Abundant loose ve- 
sicular connective tissue occurred between the 
widely spaced gonadal tubules. In males, sper- 



matogenic cells at the germinal epithelium were 
about one layer thick, but were absent in portions 
of the epithelium (Fig. 8a, b). Some sperm were in 
close contact with the spermatogenic cells and a 
few were scattered in the lumina of tubules. In 
females, the few small oocytes occurred at the 
germinal epithelium, none were in the tubule 
lumina, and all were in an early developmental 
stage (Fig. 8c, d). 

For differentiated gonads with moderate tubule 
development, 39 males examined were producing 
sperm, while 10 females examined were producing 
oocytes. The gonadal tubules were more numerous 
than in gonads of sparse tubule condition, and 
some exhibited an expanded alveolar condition. 
Loose vesicular connective tissue clearly sepa- 
rated the tubules. In males, several layers of sper- 
matocytes proliferated from the germinal 
epithelium with some sperm forming a fringe ex- 
tending toward the empty lumina; however, por- 
tions of the germinal epithelium in some tubules 



TABLE 1. — Gonadal condition relative to age. sex. and size of three categories of ocean 
quahogs. An tua islandica — sexually immature, intermediate, and mature — off Long 
Island. N.Y.. late July-early August 197s. M = male; F = female. 









No 


clams (%] 


i 










Immature 




Intermediate 












Tubule development 




















(undiffer- 
entiated) 


Spai 


se 


Moderate 


MatL 


ire 


Total 




M 


F 


M 


F 


M 


F 


no. 


Age 


















(yO 


















2 


K08) 














1 


3 


4(3 0) 


2(1.5) 




2(1.5) 








8 


4 


7(5.3) 


5(3.7) 




2(1.5) 








14 


5 


11(8.2) 


4(3.0) 


1(0 8) 


9(6.7) 




1(0.8) 




26 


6 


9(6.7) 


3(22) 


(1 5) 


10(7.5) 




2(1.5) 


1(0 8) 


27 


7 


3(2 2) 


2(1 5) 


1(0 8) 


12(9 0) 


9(6.7) 


2(1.5) 




29 


8 


1(08) 






3(22) 


1(0.8) 


5(3.7) 




10 


9 












1(0.8) 




1 


10 








1(0 8) 




2(1.5) 




3 


11 












1(08) 


1(0 8) 


2 


12 












1(0 8) 


2(1 5) 


3 


13 












1(0.8) 




1 


14 












1(0 8) 


1(0.8) 


2 


15 


















16 














4(3.0) 


4 


17 


















18 












2(15) 




2 


Age range 


2-8 


3-7 


5-7 


3-10 


7-8 


5-18 


6-16 


2-18 


Mean 


5.03 


4 63 


6 00 


6.08 


7 .10 


9 79 


13.22 


6 71 


Shell 


















length 


















(mm) 


















20 


2 




















2 


20-29 


8 


4 





3 











15 


30-39 


16 


9 


3 


18 


2 


2 





50 


40-49 


10 


3 


1 


18 


8 


12 


1 


53 


50-59 

















5 


6 


11 


-59 




















2 


2 


Length range 


19-46 


21-44 


36-42 


20-48 


39-45 


36-58 


41-60 


19-60 


Mean 


34.4 


33.8 


384 


37.2 


41.8 


47.1 


55.0 


39.0 


Total no. 


36(27.1) 


16(12 0) 


4(3.0) 


39(29.3) 


10(7.5) 


19(14.3) 


9(6.8) 


'133 



'The tissues of a 21.1 mm. 3-yr-old clam were too poorly prepared for examination 



260 



ROPES ET AL.: SIZE, AGE, AND SEX OF OCEAN QUAHOGS OFF LONG ISLAND, NY. 






i* c? 









IOO>im 






• t 



**& • lie 



if • 



*  * o 



.■?--«. aft  j» . ,, - 



S . 



**. 



lOjJrn- 



FIGURE 7. — (a) Undifferentiated gonadal tissue section from a 5-yr-old ocean quahog, Arctica islandica, 37.2 mm shell length. 

(b) Enlargement of a gonadal tubule from the same clam. 



I 






\ 



s 

X 












>:« lift ■: 



*•*'*' 




*"" 




►• » 








ctf\*. 








> 










if.* • 

< 








» 






J 
II 






• 


/ 






i 




9tf 


» 








•*. 



L_l 
lOO^m 






>»«i : v? 









f 









lOjjm 



FIGURE 8. — (a) Differentiated gonadal tissue section in the sparse condition from a 3-yr-old male ocean quahog, Arctica islandica, 
21.0 mm shell length, (b) Enlargement of spermiogenesis in a portion of a gonadal tubule, (c) Differentiated gonadal tissue section in 
the sparse condition from a 5-yr-old, 37.5 mm shell length, ocean quahog. (d) Enlargement of oogenesis in a gonadal tubule. 



261 



FISHERY BULLETIN: VOL. 82, NO. 2 



again lacked obvious spermatogenic cells (Fig. 9a, 
b). Oocytes in females were at the same stage of 
development as seen for females with sparse 
gonadal tubules, but more were growing from the 
germinal epithelium and some portions of the 
germinal epithelium lacked obvious oogenic cells 
(Fig. 9c, d). 

The sexually mature condition was found in 19 
males and 9 females. In these quahogs the tubules 
were greatly expanded and filled the gonadal area; 
little connective tissue occurred between adjacent 
tubules. Developmental stages similar to those de- 
scribed for other bivalves by Ropes and Stickney 
( 1965) were recognized. Two males and one female 
were in an early gonadal condition. Sper- 
miogenesis and oogenesis had cellular charac- 
teristics as in gonads of moderate tubule develop- 
ment, but the tubules were more numerous and 



crowded together. Six males were in a late gonadal 
condition. Primary and secondary spermatocytes 
and spermatids were proliferating from the ger- 
minal epithelium, filling about half of the tubules 
and sperm crowded into the lumina. No females 
were found in the late gonadal condition, but 11 
males and 2 females were in an advanced late 
stage. In males, spermatocytes and spermatids 
proliferated from the germinal epithelium and 
sperm predominated in the lumina of the tubules 
(Fig. 10a, b). In females, oocytes crowded into the 
lumina of tubules and a few seemed to be attached 
to the germinal epithelium. No ripe males and 
only six ripe females with numerous ripe oocytes 
crowding into the tubules were found (Fig. 10c, d). 
The potential for developing large numbers of 
germinal cells was most evident and indicative of 
full sexual maturity in all of these quahogs. 



a 



,< 



, 



v . 



% 



: S 



 



m 



V 



\ *>. 



V 



I 







/"** 



■AS'4 






3< 









.•SO*- 'ft** 



SI 






*? 



'■ 



V 
t. 



* 









lOO^m* 







d 






&-?§£# U 







u 



FIGURE 9. — (a) Differentiated gonadal tissue section in the moderate condition from a 7-yr-old ocean quahog, Arctico islandica, 42.9 
mm shell length, (bi Enlargement of spermiogenesis in a portion of a gonadal tubule, ic) Differentiated gonadal tissue section in the 
moderate condition from an 8-yr-old female ocean quahog, 43.3 mm shell length, (d ) Enlargement of oogenesis in a portion of a gonadal 
tubule. 



262 



ROPES ET AL.: SIZE. AGE. AND SEX OF OCEAN QUAHOGS OFF LONG ISLAND, NY. 



I 



" 









m 



 






ft 









ft 



7 



* 



i 
I 



> 




FIGURE 10. — i a i Differentiated gonadal tissue section in the mature condition from an 18-yr-old male ocean quahog, Arctica islandica , 
57.8 mm shell length, (b) Enlargement of spermiogenesis in a portion of a gonadal tubule, (c) Differentiated gonadal tissue section in 
the mature condition from a 16-yr-old female quahog, 59.8 mm shell length, (d) Enlargement of ripe oocytes in a tubule. 



Gonadal Condition vs. 
Size and Age 

In an analysis of gonadal condition relative to 
age and size, quahogs in the undifferentiated, 
immature condition ranged from 2 to 8 yr old, 
averaged 5.0 yr old, and were from 19 to 46 mm 
long and averaged 34.4 mm (Table 1). This condi- 
tion was found in 27% of the gonads in the sample. 

For the three types of differentiated gonads, 
quahogs with sparse tubule development com- 
prised 15% of the sample. Males ranged from 3 to 7 
yr old, averaged 4.6 yr old, and were from 21 to 44 
mm long and averaged 33.8 mm; females ranged 
from 5 to 7 yr old, averaged 6.0 yr, and were from 
36 to 42 mm long and averaged 38.4 mm. This 
category contained the smallest and youngest 
female in the sample: 38 mm long and 5 yr old. 

Quahogs with moderate tubule development 



comprised 37% of the sample. Males ranged from 3 
to 10 yr old, averaged 6.1 yr, and were from 20 to 48 
mm long and averaged 37.2 mm; females ranged 
from 7 to 8 yr old, averaged 7.1 yr, and were from 39 
to 45 mm long and averaged 41.8 mm. This cate- 
gory contained the smallest and youngest male in 
the sample, which was 20 mm long and 3 yr old 
(Fig. la, b). 

Sexually mature quahogs comprised 21% of the 
sample. Males ranged from 5 to 18 yr old, averaged 
9.8 yr, and were from 36 to 58 mm long and aver- 
aged 47.1 mm; females ranged from 6 to 16 yr old, 
averaged 13.2 yr, and were from 41 to 60 mm long 
and averaged 55.0 mm. The smallest mature 
quahog found was a male 36 mm long and 6 yr old, 
although a 5-yr-old, 41 mm long male was also 
mature; the smallest and youngest mature female 
found was 41 mm long and 6 yr old. 

None of the gonads contained germinal cells 

263 



FISHERY BULLETIN: VOL. 82, NO. 2 



suggestive of ambisexuality. This is consistent 
with the conclusion of Loosanoff (1953) that the 
sexes are separate. The sex ratio, however, was 
particularly imbalanced in favor of males. In the 
69 quahogs considered less than fully mature, 55 
were males and 14 were females, while in the 28 
sexually mature specimens, 19 were males and 9 
were females; the observed ratios were 4:1 and 2:1, 
respectively. The data were subjected to goodness 
of fit tests under the hypothesis of a 1:1 ratio 
between the sexes; results indicated highly signif- 
icant (P<0.01) and significant (P<0.05) dif- 
ferences, respectively. 

Microscopic examinations of gonadal tissue 
squashes of the 199 clams collected in 1980 re- 
vealed an overall sex ratio of 96 males and 103 
females. These results were not significantly dif- 
ferent from parity (1 male:1.07 female), but by 
separating the data into 10 mm size groups, a 
significant difference (P 0.05) in favor of males 
was indicated in the size group 80-89.9 mm, and a 
highly signficant difference (P < 0.01) in favor of 
females was indicated in the 100-110 mm size 
group (Table 2). 

Figure 11 shows the combined observations of 
clam size and sex obtained from the 1978 and 1980 
samples. In these samples, males tended to 
decrease in occurrence relative to females with 
increasing shell size. 

TABLE 2. — Occurrence of male and 
female ocean quahogs, Arctica islan- 
dica, within 10 mm size groups off 
Long Island, N.Y., August 1980. 



too 



Size group 
(mm) 




Number 


Males 


Females 


50-59 


4 





60-69 


44 


32 


70-79 


12 


21 


80-89 


16 


5 


90-99 


19 


33 


100-109 


1 


12 


Total 


96 


103 



DISCUSSION 

The time of sampling, sample size, and capture 
of small quahogs provided a basis for detection of 
the differentiated and sexually mature stage at 
younger ages and smaller sizes as compared with 
the study of Thompson et al. (1980b). In the pres- 
ent study, 5- and 6-yr-old quahogs 41 and 36 mm 
long, respectively, were considered sexually ma- 
ture; the youngest mature quahog reported by 
Thompson et al. (1980b) was a 42 mm male 11 yr 
old. The intermediate gonadal condition was 



80 - 



-J 60 

1 

k 

0. 



.' 



N= 170 d*o* 126 $9 
1 :074 




J_ 



_L 



_L 



L 



< 20-29 40-49 60-69 

SIZE CROUPS 



80-89 



100-109 



FIGURE 11. — Sex of ocean quahogs. A rctica islandica, relative to 
shell length (mm) in collections off Long Island, N.Y., 1978 and 
19S0 



found to occur at lower ages and smaller sizes than 
by Thompson et al. (1980b), and slightly smaller 
sizes were found for sexually mature quahogs. 
Variability in attainment of sexual maturity at 
age/size was observed in both studies. 

The onset of sexual maturity at young ages has 
been reported for several bivalves. The bay scal- 
lop, A rgopecten irradians, attains maturity at 1 yr; 
the hard clam, Mercenaria mercenaria, soft clam, 
Mya arenaria, and blue mussel, Mytilus edulis, 
matures at 1-2 yr ( Altman and Dittmer 1972). Surf 
clams, Spisula solidissima, from an inshore 
habitat showed precocious sexuality in a few post- 
larvae or juveniles; they spawned at 1 yr, but 
reached full maturity at 2 yr (Ropes 1979a). Sea 
scallops, Placopecten magellanieus, spawned at 
about 1.5-2 yr after forming the first growth ring 
(Naidu 1970). In apposition to more mature 
gonadal conditions, some scallops in his collec- 
tions were considered undifferentiated and dif- 
ferentiated male and female immature specimens. 
Lucas (1966) observed precocious sexuality in a 
scallop (Chlamys varia) and two clams 
(Glycymeris glycymeris and Venus striatula) from 
waters off France. The development of the repro- 
ductive potential during the early life history of 
these several bivalves seems consistent with esti- 
mates of their life span, which are as short as 2 yr 
for the bay scallop and as long as 30 yr for the surf 
clam (Belding 1906; Ropes 1979a). In contrast, the 
present study revealed that ocean quahogs attain 
maturity at 5-10 yr of age, and Thompson et al. 



264 



ROPES ET AL.: SIZE. AGE, AND SEX OF OCEAN QUAHOGS OFF LONG ISLAND. NY. 



(1980a) reported a longevity of about 150 yr. They 
found that growth was vigorous at old age and that 
there were no obvious indications of reproductive 
senility. A small abyssal nuculoid bivalve, Tin- 
daria call isti for mis, studied by Turekian et al. 
(1975) seems most exceptional with regard to age 
and size at sexual maturity They found a longev- 
ity of about 100 yr for a large specimen (8.4 mm 
shell length) by radiometric techniques and 
counts of shell growth bands, but gonadal devel- 
opment was not recognized until the clams were 
about 4 mm long and 50-60 yr old. The attainment 
of sexual maturity about midway in the life span of 
Tindaria sets it apart from other species that re- 
produce at a younger age. Nevertheless, all have 
the potential to reproduce for many years. Repro- 
duction during a long life span of a species may be 
an evolutionary strategy in response to uncertain 
larval and juvenile survival (Krebs 1972). Repro- 
duction during a particularly long life span is most 
obvious for Arctica islandica. 

For the 69 gonads containing sexually differen- 
tiated germinal cells and sparse-to-moderate 
tubule development, some morphologically ripe 
sperm were present. In contrast, oogenesis never 
exceeded an early developmental state. Jones 
(1981), Loosanoff (1953), von Oertzen (1972), and 
Mann (1982) reported that mature ocean quahogs 
spawn each year. Thus, the sperm may be 
spawned, but the fate of the oocytes remains an 
enigma. In American oysters, Crassostrea vir- 
ginica, germinal cells remaining in the gonads 
after spawning are reabsorbed (Galtsoff 1964), but 
viable, nearly ripe, or ripe germinal cells may be 
retained by hard clams throughout the fall, 
winter, and into the following spring (Loosanoff 
and Davis 1951). Thus, bivalves appear to differ 
greatly in this respect. No conclusion can be drawn 
relative to retention of germinal cells after spawn- 
ing for ocean quahogs which were intermediate 
between the immature and mature condition in 
the absence of collected data. 

Gonadal development in 28 mature clams 
suggested that many (46% ) were approaching 
ripeness or were ripe (21% ). Later development 
probably resulted in a spawning which was begun 
in late August-September. This seems reasonable 
based on observations by Mann (1982) of the re- 
productive cycle of Arctica islandica from sample 
locations in Block Island Sound. At the beginning 
of his study in September 1978, most (69% ) were in 
the partially spent or spent condition and spawn- 
ing was indicated until mid-November. An exact 
correspondence of the time and duration of spawn- 



ing may be a hazardous assumption, since the two 
sample sites are about 110 km apart and some of 
the samples taken by Mann (1982) were at shallow 
depths (36 m). 

A disparity in the initiation of gametogenesis 
was observed between the sexes. Male ocean 
quahogs began producing germinal cells at a 
smaller size and younger age than females. This 
suggests that females require a longer period of 
development and growth. The later development 
of female sexuality is a probable explanation for 
the highly significant difference obtained in tests 
of the sex ratio of quahogs in the intermediate 
gonadal condition. The significant difference ob- 
served for fully mature quahogs may be due to the 
small number in the sample (Dixon and Massey 
1957), but Jones (1981) observed a similar dispar- 
ity (P = 0.008) for quahogs > 75 mm from offshore 
New Jersey. In his collections 184 were males and 
136 were females, a ratio of 1:0.74. Mann (1982) 
examined ocean quahogs that were mostly 80-100 
mm long and found 185 males and 169 females, a 
ratio of 1:0.91. These observations suggest that 
spatial variation may occur in the sex ratio of 
ocean quahog populations, but that males are 
more numerous than females. 

Pelseneer (1926) investigated the sex ratio of 
several mollusc species, including bivalves. He 
found more females among the older individuals of 
some populations and the converse among younger 
individuals. Coe ( 1936) recognized the existence of 
such disparities in molluscs and proposed the fol- 
lowing hypotheses as possible explanations: 1) 
That males have a shorter longevity than females, 
because of a differential mortality rate or less re- 
sistance to unfavorable environmental conditions; 
2) that the development of alternative sexual con- 
ditions is environmentally determined; and 3) 
that sex change may occur. Loosanoff (1953), von 
Oertzen ( 1972 ) , Thompson et al .( 1980b ) , and Jones 
(1981) all considered the species to be strictly 
dioecious, as did Mann (1982), although he found 
two hermaphrodites. These are anomalous, "acci- 
dental functional hermaphrodites" by the ter- 
minology of Coe (1943). Although Sastry (1979) 
hypothesized that a failure in the genetic sex- 
differentiating mechanism may produce some 
hermaphrodites, he found no evidence of a 
phenotypic or genetic basis for sex determination 
in pelecypods. 

It is unlikely that ocean quahogs are protandric. 
This condition in a typically hermaphroditic 
species is characterized by the development of 
male organs or maturation of their products before 



265 



FISHERY BULLETIN: VOL. 82, NO. 2 



the appearance of corresponding female products. 
In Ostrea lurida, for example, spermatogonia are 
proliferated first throughout the follicles, but be- 
fore the sperm mature oogonia have developed into 
numerous oocytes in the same follicles and the 
gonad has a definite intersexual character (Coe 
1932). More than 90 c /c of the young oysters exhibit 
the bisexual condition and no strictly male or 
female specimens occur. Old oysters in the female 
phase retain sperm balls and spermatogonia, and 
those in the male phase retain large and small 
oogonia. The two anomalous ocean quahogs found 
by Mann (1982) were examples of bilateral her- 
maphroditism, i.e., the germinal cells for each sex 
were in separate follicles. None of the inves- 
tigators of the reproductive cycle in ocean quahogs 
suggested finding ambisexual conditions 
(Loosanoff 1953; von Oertzen 1972; Jones 1981; 
Mann 1982 ). Thus, the characteristic germinal cell 
development for protandry is lacking in ocean 
quahogs. 

Sex reversal in some molluscs has been linked to 
castration from parasites invading the gonads, but 
evidence of causality was considered inconclusive 
by Noble and Noble (1961) and Malek and Cheng 
( 1974 ). Except for the occurrence of the commensal 
nemertean, Malacobdella grossa, in ocean 
quahogs (Gibson 1967; Jones 1979), parasites in 
the species have not been reported (Ropes and 
Lang 1975 ) 4 . The causality of hermaphroditism in 
ocean quahogs, then, remains uncertain and evi- 
dence is unavailable that sex may be environmen- 
tally determined. 

The hypothesis that female ocean quahogs 
may live longer than males has some support from 
determinations of the sex of specimens recovered 
from the marking site in August 1980. Based on 
predicted ages of ocean quahogs at the marking 
site reported by Murawski et al. ( 1982 ), the largest 
and oldest notched ocean quahogs were predomi- 
nantly female. Since this may be atypical for the 
extensive population of ocean quahogs inhabiting 
the Middle Atlantic Bight, samples from other lo- 
cations are being examined to determine possible 
spatial variations. 

ACKNOWLEDGMENTS 

We gratefully acknowledge the technical assis- 



4 Ropes, J. W, and H. S. Lang. 1975. An annotated bibliog- 
raphy of the ocean quahog, Arctica islandica (Lin- 
naeus). Xeroxed manuscr., 67 p. Northeast Fisheries Center 
Woods Hole Laboratory, National Marine Fisheries Service, 
NOAA, Woods Hole, MA 02543. 



tance of Dorothy W. Howard and Cecelia S. Smith 
of the Northeast Fisheries Center Oxford Labora- 
tory, National Marine Fisheries Service, NOAA, 
Oxford, Md., for histological preparations of 
gonadal tissues; and Taina Honkalehto, Frances 
Lefcort, and Miranda Olshansky, student trainees 
from Smith College, Northampton, Mass., for as- 
sistance in preparing acetate peels of the shells of 
ocean quahogs. 

LITERATURE CITED 

ALTMAN, p. l. and d. s. Dittmer. 

1972. Biological data book. 2d ed. Fed. Am. Soc. Exp. 
Biol., p. 154-155. 
BELDING, D. L. 

1906. Preliminary report upon the shellfisheries of Mas- 
sachusetts. Comm. Fish. Game, Mass., 15 p. 
COE, W. R. 

1932. Development of the gonads and the sequence of 

the sexual phases in the California oyster (Ostrea 

lurida i. Bull. Scripps. Inst. Oceanogr., Tech. Ser. 

3:119-144. 

1936. Sex ratios and sex changes in mollusks. Mem. Mus. 

Hist. Nat. Belg. 3:69-76. 
1943. Sexual differentiation in mollusks. I. 
Pelecypods. Q. Rev. Biol. 18:154-164. 
DIXON, W. J.. AND F J. MASSEY, JR. 

1957. Introduction to statistical analysis. 2d 
ed. McGraw-Hill. N.Y.. 488 p. 
GALTSOFF, P. S. 

1964. The American Oyster, Crassostrea virginica Gme- 
lin. U.S. Fish Wildl. Serv, Fish. Bull. 64, 480 p. 
GIBSON, R. 

1967. Occurrence of the entocommensal rhynchocoelan, 
Malacobdella grossa , in the oval piddock, Zirfaea cnspata , 
on the Yorkshire coast. J. Mar. Biol. Assoc. U.K. 47:301- 
317. 
JONES, D. S. 

1979. The nemertean, Malacobdella grossa, in the ocean 

quahog, A rcti lea islandica (Bivalvia). Nautilus 93:29-30. 

1981. Reproductive cycles of the Atlantic surf clam Spisula 

solidissima , and the ocean quahog Arctica islandica off 

New Jersey. J. Shellfish Res. 1:23-32. 

KENNEDY, A. V, AND H. I. BATTLE. 

1964. Cyclic changes in the gonad of the American oyster, 
Crassostrea virginica (Gmelin). Can. J. Zool. 42:305-321. 
KREBS, C. J. 

1972. Ecology; the experimental analysis of distribution 
and abundance. Harper and Row, Publ., N.Y., 694 p. 
LOOSANOFF, V L. 

1953. Reproductive cycle in Cyprina islandica. Biol. Bull. 
(Woods Hole) 104:146-155. 
LOOSANOFF, V. L., AND H. C. DAVIS. 

1951. Delaying spawning of lamellibranchs by low tem- 
perature. Sears Found., J. Mar. Res. 10:197-202. 
LUCAS, A. 

1966. Manifestation precoce de la sexualite chez quelques 
mollasques bivalves. Estratlo Lav. Soc. Malacol. Ital. 
3:153-158. 

Malek, E. a., and C. Cheng. 

1974. Medical and economic malacology. Acad. Press, 
NY, 398 p. 



266 



ROPES ET AL : SIZE, AGE, AND SEX OF OCEAN QUAHOGS OFF LONG ISLAND, NY. 



Mann, R. 

1982. The seasonal cycle of gonadal development in Arctica 
islandica from the Southern New England shelf. Fish. 
Bull.. U.S. 80:315-326. 
MURAWSKI, S. A., J. W. ROPES, AND F. M. SERCHUK. 

1982. Growth of the ocean quahog. A rctica islandica, in the 
Middle Atlantic Bight. Fish. Bull., U.S. 80:21-34. 
NAIDU, K. S. 

1970. Reproduction and breeding cycle of the giant scallop 
Placopecten magellanicus (Gmelini, in Port au Port Bay, 
Newfoundland. Can. J. Zool. 48:1003-1012. 

NOBLE. E. R.. AND G. A. NOBLE. 

1961. Parasitology, the Biology of Animal Parasites. Lea 
and Febiger, Phila., 767 p. 
OERTZEN, J. A. VON. 

1972. Cycles and rates of reproduction of six Baltic Sea 
bivalves of different zoogeographical origin. Mar. Biol. 
(Berl.) 14:143-149. 
PELSENEER, R 

1926. La proportion relative des sexes chez les animau et 
particulierement chez les mollusques. Mem. Acad. R. 
Belg., Classe Sci. 8:1-258. 
RAVEN, C. R 

1958. Morphogenesis: The analysis of molluscan devel- 
opment. Pergamon Press, N.Y., 311 p. 
ROPES, J. W. 

1968a. Reproductive cycle of the surf clam, Spisula solidis- 
sima, in offshore New Jersey. Biol. Bull. (Woods Hole) 
135:349-365. 
1968b. Hermaphroditism in the surf clam, Spisula solidis- 
sima. Proc. Natl. Shellfish. Assoc. 58:63-65. 

1971. Maryland's hard clam studied at Oxford labora- 
tory. Chesapeake Bay Affairs, Commer. Fish. News 
4(6):2-3. 



1979a. Shell length at sexual maturity of surf clams, 
Spisuhi solidissima, from an inshore habitat. Proc. Natl. 
Shellfish. Assoc. 69:85-91. 
1979b. Biology and distribution of surf clams (Spisula so- 
lidissima) and ocean quahogs (Artica islandica) off the 
Northeast Coast of the United States. Proc. Northeast 
Clam Ind.: Management for the future, p. 47-66. Univ. 
Mass. and Mass. Inst. Tech., Sea Grant Prog. SP-112. 
ROPES. J. W, AND A. P. STICKNEY. 

1965. Reproductive cycle of Mya arenaria in New En- 
gland. Biol. Bull. (Woods Hole) 128:315-327. 
ROPES, J. W, D. S. JONES, S. A. MURAWSKI, F M. SERCHUK, 
AND A. JEARLD, JR. 

1984. Documentation of annual growth lines in ocean 
quahogs, Arctica islandica Linne. Fish. Bull., U.S. 
82:1-19. 
SASTRY, A. N. 

1979. Pelecypoda (excluding Ostreidae). In A. C. Giese 
and J. S. Pearse (editors), Reproduction of marine inver- 
tebrates, Vol. V, p. 113-292. Acad. Press, N.Y. 
THOMPSON, I., D. S. JONES, AND D. DREIBELBIS. 

1980a. Annual internal growth banding and life history of 
the ocean quahog Arctica islandica (Mollusca: Bival- 
via). Mar. Biol. (Berl.) 57:25-34. 
THOMPSON, I., D. S. JONES, AND J. W ROPES. 

1980b. Advanced age of sexual maturity in the ocean 

quahog Arctica islandica (Mollusca: Bivalvia). Mar. 

Biol. (Berl.) 57:35-39. 

TUREKIAN, K. K., J. K. COCHRAN, D. R KHARKAR, R. M. CER- 

RATO, J. R. VAISNYS, H. L. SANDERS, J. F GRASSLE, AND J. A. 

ALLEN. 

1975. Slow growth rate of a deep-sea clam determined by 
228 Ra chronology. Proc. Natl. Acad. Sci. 72:2829-2832. 
YONGE, C. M. 

1960. Oysters. Collins, Lond, 209 p. 



267 



FOOD HABITS AND DIETARY OVERLAP OF 

SOME SHELF ROCKFISHES (GENUS SEBASTES) FROM 

THE NORTHEASTERN PACIFIC OCEAN 

Richard D. Brodeur and William G. Pearcy 1 

ABSTRACT 

Euphausiids were the major food of five co-occurring species of rockfishes I Sebastes spp. ) along the west 
coast of North America from Vancouver Island to northern California. Copepods, decapods, cephalo- 
pods, amphipods, fishes, and other pelagic prey were also consumed but were less important to the 
overall diet. Two species, S. flavidus and S. diploproa, were relatively euryphagous, utilizing a high 
number of prey taxa. The other species, S. pinniger, S. alutus, and S. crameri, had a more restricted 
diet comprised mostly of euphausiids. The numerical composition of prey in the diet of all species was 
similar due to the preponderance of the two dominant euphausiid species. Diet overlaps based on 
weight composition were high for S. pinniger, S. diploproa, and S. alutus but were moderate for most 
comparisons involving S. flavidus and S. crameri. 

The diets of S. flavidus and S. pinniger were examined in more detail to explain some of the vari- 
ability associated with their food habits. Both species exhibited peak feeding periods at the same time 
during the day. They consumed about the same mean size of prey, although S. flavidus consumed a 
wider size range of prey. Size of prey and dietary composition did not vary much with size offish. There 
were significant seasonal, geographical, and diel differences in food composition for both species, which 
may be a function of varying food availability. 



Factors that allow coexistence of a large number of 
morphologically similar species have been the 
focus of numerous studies and continued debate 
in the ecological literature. Competition and re- 
source partitioning have been reviewed in general 
by Schoener (1974), and for fishes by Helfman 
(1978). Potential competition for resources is 
thought to be most common in three aspects of the 
ecological niche in fish communities: habitat, 
food, and time of activity (Tyler 1972; Bray and 
Ebeling 1975; Ross 1977; Werner 1979; Larson 
1980; McPhersonl981). 

Rockfishes (Sebastes spp.) of the family Scor- 
paenidae are, a priori, interesting subjects for 
examining the various modes of resource parti- 
tioning. This genus is extremely speciose, with 
about 100 species reported from the North Pacific 
Ocean. At least 69 of these species are known to 
occur in the eastern North Pacific (Chen 1975). In 
addition to the large number of species, rockfishes 
also exhibit a high degree of overlap in their 
geographical distributions, with as many as 50 
species occurring in a narrow latitudinal band 
(lat. 34°-38°N) off central California (Chen 1971). 
Several of these congeners are morphologically 



'School of Oceanography, Oregon State University, Marine 
Science Center, Newport, OR 97365. 



similar and occupy similar habitats, so the poten- 
tial for resource overlap and competition is high 
(Larson 1980). 

Many of these species are abundant enough and 
of sufficient size to contribute substantially to 
commercial trawl landings in the northeastern 
Pacific (Alverson et al. 1964; Alton 1972; Gabriel 
and Tyler 1980; Gunderson and Sample 1980). De- 
spite their abundance in the northeastern Pacific, 
relatively few quantitative studies exist on rock- 
fish feeding habits. Most of the studies to date 
have dealt with shallow-water, neritic species 
often taken in recreational fisheries or accessible 
to in situ observations and sampling by scuba 
divers (Gotshall et al. 1965; Larson 1972; Hobson 
and Chess 1976; Love and Ebeling 1978). Descrip- 
tions of the diet of offshore species of Sebastes 
generally either lack taxonomic or quantitative 
detail (Phillips 1964) or encompass limited geo- 
graphical area or collection times (Pereyra et al. 
1969; Lorz et al. 1983). Skalkin (1964) and Somer- 
ton et al. (1978) 2 described food habits of rock- 
fishes from the Bering Sea and Gulf of Alaska, far 



Manuscript accepted October 1983. 
FISHERY BULLETIN: VOL. 82, NO. 2, 1984. 



2 Somerton, D., F. Funk, K. Mesmer, L. J. Bledsoe, and K. 
Thornburgh. 1978. A comparative study of the diets of Pacific 
ocean perch [Sebastes alutus) and walleye pollock [Theragra 
chalcogramma ) in the Gulf of Alaska. NORFISH Tech. Rep. 
NPB8, Wash. Sea Grant, 25 p. 

269 



FISHERY BULLETIN: VOL. 82, NO. 2 



north of our study area which extends from off 
northern California to off Vancouver Island, Brit- 
ish Columbia. 

This study represents the first attempt to exam- 
ine broad geographical and seasonal patterns in 
food utilization and overlap by several commer- 
cially important species of rockfish on the outer 
continental shelf. The species considered include 
the yellowtail rockfish, Sebastes flavidus; canary 
rockfish, S. pinniger; Pacific ocean perch, S. 
alutus; splitnose rockfish, S. diploproa; and the 
darkblotched rockfish, S. crameri, all important 
members of the demersal shelf rockfish complex 
(Gabriel and Tyler 1980). In addition, variability 
in the diet of two of these species, S. flavidus 
and S. pinniger, was examined for the purpose 
of determining the effects of factors such as sea- 
son, geographic area, time of capture, and pred- 
ator size. 

MATERIALS AND METHODS 

Sampling Methods 

The food utilization patterns of the five rockfish 
species were determined by examining stomach 
contents. Fishes were obtained by two different 
survey methods (hereafter referred to as the 
summer and seasonal surveys). As the collection 
methods differ, they will be discussed separately 
The laboratory methods are similar and will be 
presented together. 

Summer Survey Methods 

Collections for the summer survey were made 
during the National Marine Fisheries Service 
(NMFS) 1980 West Coast Survey which took place 
from 12 July to 28 September 1980. The purpose of 
this survey was to assess the distribution and 
abundance of commercially important rockfishes. 
The area encompassed by the survey included 
much of the continental shelf and inner slope 
(ranging in depth from 55 to 366 m) between 
Monterey, Calif., (lat. 36°48'N) and the northern 
end of Vancouver Island, British Columbia (lat. 
50° 00' N). Two commercial stern trawlers, the FV 
Mary Lou and the FV Pat San Marie, were 
utilized for the survey. A Nor'Eastern 3 high- 
opening bottom trawl with an estimated 13.4 m 
horizontal and an 8.8 m vertical mouth opening 



was used on both vessels. The main body was 
constructed of 127 mm stretched mesh with 89 mm 
mesh in the cod end. The cod end also contained 
a 32 mm mesh liner. Half-hour tows were made 
at random depth-stratified stations chosen by 
a method described in Gunderson and Sample 
(1980). 

The majority of the stomach samples used in 
this study were collected in August and Septem- 
ber from north of lat. 43° N (Table 1, Fig. 1). 
Complete station data are given in Brodeur (1983). 

Stomachs were removed at sea from a random 
subsample of the catch of the five target species 
(Table 1). Sebastes pinniger and S. flavidus were 
the primary target species, and stomachs of these 
species were collected first and the other species 
sampled as time allowed. Altogether, 480 stom- 
achs were collected during the survey, all from 
adult fish ( > 200 mm FL). Fork length (measured 
to the nearest millimeter) and sex were recorded 
for all fish sampled, and stomachs were then 
removed, individually wrapped and labeled, and 
preserved in a 10% Formalin-seawater mixture. 
The intestinal tracts of many of the fish were 
examined at sea but few contained any recogniz- 
able food and none were retained. Total elapsed 
time between bringing the fish on board and 
preserving the stomachs was < 1 h. The oral 
cavities of all fish were examined for signs of 
stomach eversion and regurgitation; any fish 
showing such signs were discarded. Individual 
fish weights were not recorded at sea but were 
later calculated using the length-weight relation- 
ships of Phillips (1964). 

TABLE 1. — Number of rockfish stomachs 
analyzed from the 1980 National Marine 
Fisheries Service summer survey. The 
approximate latitudinal ranges covered 
by each leg were I, lat. 37°-42°N; II, lat. 
43°-46°N; III, lat. 46°-50°N. 



Reference to trade names does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 



Leg 


Sampling 
dates 


Species 


Number 


1 


12-20 July 


S pinniger 
S flavidus 


9 
8 

17 


II 


4-29 Aug. 


S pinniger 
S. flavidus 


85 

127 






S. alutus 


54 






S. diploproa 


52 






S. crameri 


30 
348 


III 


4-28 Sept 


S pinniger 
S flavidus 


36 
50 






S. alutus 


19 






S diploproa 


10 
115 






Total number analyzed 480 



270 



BRODEUR and PEARCY: FOOD HABITS AND DIETARY OVERLAP OF SOME SEBASTES 



48 



FIGURE 1. — Location of sampling sta- 
tions from which stomach collections 
were taken. + sign denotes collections 
made during the National Marine Fish- 
eries Service's summer survey and the 
stippled area (inset) shows the sampling 
area on Heceta Bank of the Oregon De- 
partment of Fish and Wildlife's seasonal 
collections. All depth contours are in 
meters. 



4t> 



42" 




CALIFORNIA 

t 
l( 



EUREKAo^, A 
I 'I 



271 



FISHERY BULLETIN: VOL. 82, NO. 2 



Seasonal Survey Methods 

Stomachs for the seasonal study were collected 
during rockfish surveys conducted by the Oregon 
Department of Fish and Wildlife (ODFW) on 
Heceta Bank off the central coast of Oregon. These 
surveys obtained hydroacoustic and environmen- 
tal data along with the trawl catches. A total of 317 
stomach samples was collected during seven sur- 
veys conducted in 1980-81 (Table 2). All surveys 
used trawling gear similar to that used in the 
summer surveys. 

Locations of the tows were chosen on the basis of 
high concentrations offish found during acoustic 
surveys over the outside edge of Heceta Bank 
between lat. 44°20'N and 44°00'N between the 
128 m and 238 m bathymetric contours (inset, Fig. 
1). The duration of tows was variable but averaged 
< 1 h. No tows were attempted at night because of 
the lack of acoustical targets near the bottom at 
this time. Stomachs were collected as described 
earlier. 



TABLE 2. — Number of rockfish stomachs analyzed from the 
seasonal Oregon Department of Fish and Wildlife collections 
on Heceta Bank. All dates are in 1980 unless otherwise noted. 



Vessel 


Cruise 


Sampling dates 


Species 


Number 


Ronnie C 


1 


23-24 April 


S, pinniger 


42 


Bay Islander 


1 


17-18 June 


S pinniger 


24 


Queen Victoria 


1 


15-16 July 


S pinniger 
S flavidus 


47 
16 


Ronnie C 


II 


26-28 Sept. 


S. pinniger 
S. flavidus 


60 
23 


New Life 


1 


27 Oct. 


S. pinniger 
S. flavidus 


21 
2 


Ronnie C 


III 


17-18 Dec 


S pinniger 
S flavidus 


33 
25 


New Life 


II 


25 Jan. 1981 


S- pinniger 
S flavidus 


11 
13 






Total number analyzed 317 



Analysis of Stomach Contents 

The stomachs were opened and their contents 
transferred to 50% isopropyl alcohol in the labora- 
tory. Contents were examined using a variable 
power dissecting microscope. Individual stomach 
fullness was estimated according to a subjective 
rating ranging from (empty) to 5 (stomach fully 
distended with food). The condition of the contents 
was assigned a value from (well-digested, barely 
identifiable to phylum) to 4 (fresh). 

Prey were identified to the lowest possible taxon 
and enumerated. In stomachs containing many 
small prey, such as euphausiids, any large or rare 
prey items were removed first. The remaining 
contents were then subdivided by means of a 

272 



Folsom plankton splitter (McEwen et al. 1954), 
and the contents of one subsample were used 
to estimate the stomach contents of small prey. 
The digested state of the contents of many stom- 
achs made precise counts of some prey difficult. 
Some paired parts of prey animals (e.g., eyes of 
euphausiids, otoliths of teleosts) were more resis- 
tant to digestion and total counts of these parts 
were halved to yield minimum counts of prey in- 
gested. Total lengths or greatest dimensions of 
intact prey found in the stomach were measured to 
the nearest 0.1 mm for the total sample (or a sub- 
sample of at least 15 individuals) using a stage 
ruler or ocular micrometer. All prey were blotted 
dry with absorbent paper and wet weights of each 
taxon were recorded to the nearest milligram. 

Analysis of Food Habits 

The minimum number of stomach samples 
needed to adequately describe the diet of a species 
was determined for all five rockfish species, using 
a cumulative prey species curve. A subset of 
stomachs of a particular species was randomly 
chosen and the cumulative number of unique prey 
taxa were then plotted versus the number of stom- 
achs which produced these taxa. The point on the 
abscissa where the curve begins to level off is 
considered the minimum number of stomachs nec- 
essary to describe the diet of that species. An 
example of the cumulative prey curves for the first 
28 stomachs of each of the species in this study is 
shown in Figure 2. Although the curves assume 
different shapes, all approach an asymptote at 
sample sizes less than those analyzed. 

The contributions of the different prey items to 
the total diet of the rockfishes were expressed as 
percent frequency of occurrence, percent numeri- 
cal composition, and percent gravimetric composi- 
tion. Breadth and overlap were calculated for the 
five rockfishes from the summer surveys and for 
S. pinniger and S. flavidus from the seasonal 
surveys, using the pooled p ; 's (relative proportion 
of the total number or biomass of resource i used 
by each species) for the major taxa. These include 
all taxa identified to at least generic level that 
exceeded Q.\ c /< of the total weight or number of all 
identified foods. Resource breadth was computed 
for each species using the following formula: 



B = 



1 



-I Pi 



where B equals R (the total number of prey taxa 



BRODEUR and PEARCY: FOOD HABITS AND DIETARY OVERLAP OF SOME SEBASTES 

25- 



20- 



< 
x 

o 






< 



10- 



5- 



0- 



s 


pinniger 


s 


flavidus 


s 


olutus 


s 


crameri 


s 


diploproa 



.._./. 





/ 


1 




i 




i 




i 










T 



T 



T 



T 



8 12 16 20 

NUMBER OF POOLED STOMACHS 



24 



28 



FIGURE 2— Cumulative prey curves for the first 28 stomachs of each of the 5 rock- 
fish species. 



in a food spectrum) when all items are in equal 
proportion in the diet (Levins 1968). These values 
were normalized as B n = BIR, which ranges from 
(most uneven distribution) to 1 (totally even 
distribution among the prey present). This index 
assumes equal availabilities of the different prey 
to all predators. 

Several indices of dietary overlap have been 
proposed and tested with known distributions of 
prey organisms (see Cailliet and Barry 1979 4 ; 
Linton et al. 1981; Wallace 1981). The coefficient of 
overlap described by Colwell and Futuyma (1974; 
identical to Schoener's (1970) index but not ex- 
pressed as a percentage) was chosen as it was 
found to be realistic for a wide range of true over- 
laps (Linton et al. 1981). This coefficient is as 
follows: 



dh = 1.0 -0.5(2, 



Pij - phj I 



where p tJ and phj are the proportions of prey j 
found in the diets of species i and h respectively. 
This coefficient has a minimum of (no overlap 



4 Cailliet, G. M., and J. P. Barry. 1979. Comparison of food 
array overlap measures useful in fish feeding habits analysis. 
In S. J. Lipovsky and C. A. Simenstad (editors). Fish food habits 
studies, p. 67-79. Proc. 2d Pac. Northwest Tech. Workshop, 
Wash. Sea Grant. 



of prey) and a maximum of 1 (all items in equal 
proportions). 

Analysis of Diet Variations 

The sample sizes of S. pinniger and S. flauidus 
were sufficient to permit detailed analyses of their 
food habits, including seasonal, latitudinal, diel, 
and predator-size variations. 

The 368 specimens of S. pinniger and 264 of S. 
flavidus were grouped into 10 mm length catego- 
ries (Fig. 3). The distribution of S. pinniger 
lengths from the two surveys was similar and no 
significant differences in the means were found 
(Student's t-test; P > 0.05). Specimens of S. 
flavidus collected during the seasonal survey were 
significantly larger (P < 0.001) than those of 
the summer survey. Sebastes pinniger averaged 
about 40 mm larger than S. flavidus for both 
surveys combined. Corrections were made for this 
difference where appropriate in the analyses. 

To simplify the analysis of dietary variation 
in S. pinniger and S. flavidus, eight major types 
of prey were selected for comparison, based on 
their gravimetric importance or frequency of oc- 
currence. Numerical abundances were not used 
because of the great disparity in prey sizes en- 
countered and the problem of making counts on 



273 



FISHERY BULLETIN: VOL. 82, NO. 2 




350 4O0 450 500 550 

FORK LENGTH (mm) 



 i i r 

600 650 



x 



o 



3 




300 



350 



400 



450 



500 



550 



600 



FORK LENGTH (mm) 



FIGURE 3. — Size distributions of Sebastes pinniger and S. 
Ilavidus from summer (National Marine Fisheries Service) and 
seasonal (Oregon Department of Fish and Wildlife) surveys. 



incomplete animals. These prey categories include 
the two most important euphausiid species and 
other major taxonomic groups (Table 3). Other 
planktonic prey (e.g., copepods, chaetognaths, 
pteropods) were occasionally present in the diet of 
one or both species, but their contributions to the 
overall diets were minor. Cephalopods did not 



TABLE 3. — The major prey categories used in the analysis of 
diet variations and their respective size ranges found in the 
stomachs of S. pinniger and S. flavidus. 





Prey 




Category 


size range 

(mm) 


Inclusive taxa or life stages 


Euphausia pacifica 
Thysanoessa spinifera 
Total euphausuds 


8-26 
8-30 
8-30 


juvenile and adult stages 
juvenile and adult stages 
above two and other species, 


Decapods 


3-87 


unidentified euphausuds 
adult shrimp, crab zoea and 


Amphipods 

Cephalopods' 

Fishes 

Gelatinous zooplankton 


3-30 

18-150 + 
16-150 + 
10-22 


megalopae, shrimp mysis 
mostly hypernd but some gammand 
squid and octopods 
larvae, juvenile and adult stages 
ctenophores, thaliaceans, 

medusae, and siphonophores 



' Found in S flavidus stomachs only. 



occur in the diet of S. pinniger; thus only seven 
prey categories were used for this species. 

We analyzed four factors that may affect the 
diet of these two species: season, geographic area, 
time of day, and size of fish. Each factor was 
subdivided into four classes to elucidate the gen- 
eral trends within each factor. Stomach content 
data for all cruises were grouped into four sea- 
sons, based on major periods in the hydrographic 
regime on the continental shelf off Oregon (Huyer 
et al. 1975; Huyer 1977): spring (March-May), 
summer (June-August), fall (September- Novem- 
ber), and winter (December-February). The collec- 
tion stations for all cruises were divided into one 
of four latitudinally defined shelf regions: North- 
ern California-Southern Oregon (lat. 41° 00' to 
43°50'N), Heceta Bank-Central Oregon (lat. 
43° 50' to 45°00'N), Columbia Region (lat. 45°00' 
to 47°00'N), and Northern Washington-Vancou- 
ver (lat. 47° 00' to50°00'N). 

For the analysis of diel variation of feeding, the 
local mean sampling time was adjusted to account 
for latitudinal, longitudinal, and seasonal differ- 
ences in daylight. Each collection time was stan- 
dardized to an equinox day with 12 h between 
sunrise and sunset, based on solar table values. 
These adjusted collection times were assigned to 
one of four time periods: morning (0800-1200 h), 
early afternoon (1200-1600 h), late afternoon 
(1600-1800 h), and night (1800-0700 h). Only a 
small number of S. pinniger and S. flavidus were 
collected at night despite extensive nighttime 
trawling effort on several occasions during the 
summer survey. 

Since the length distributions of the two species 
were roughly normal (Fig. 3), dividing the length 
range into four equal size groups would result 
in disproportionately large sample sizes in the 
middle size ranges. On the other hand, setting the 
sample sizes of the four groups equal would result 
in narrow size ranges around the mode. As neither 
of these options seemed desirable, compromise 
groupings were chosen. For S. pinniger, we used 
the following size classes: <45 cm, 45-<50 cm, 
50- < 55 cm, and s 55 cm. Similar size classes were 
selected for S. flavidus but were offset 5 cm to 
reflect the smaller mean size of this species. 

To test whether significant within-factor varia- 
tion occurred in the diet of each species, contin- 
gency tables were constructed comparing the 
occurrence of food or a particular prey category 
versus the absence of food or that prey category. 
A variance test for homogeneity of binominally 
distributed data (Snedecor and Cochran 1967) was 



274 



BRODEUR and PEARCY: FOOD HABITS AND DIETARY OVERLAP OF SOME SEBASTES 



used for testing differences among the classes 
within each factor. Any comparisons which ex- 
ceeded the tabulated 0.05 \ 2 percentage caused a 
rejection of the null hypothesis of similar diets. 

RESULTS 

General Food Habits 

The results of the stomach content analysis are 
presented for both surveys and all five species in 
Tables 4 through 8. Each species will be discussed 
in detail in this section. 

Sebastes flavidus preyed on a diverse assem- 
blage of planktonic and micronektonic prey (Table 
4). Dominating the diet in terms of frequency of 
occurrence (F.O.), percent by number, and, to a 
lesser extent, percent by weight were euphausiids, 
principally Euphausia pacifica and Thysanoessa 
spimfera. Many species of hyperiid amphipods 
were represented in the diet, but these were not 
numerous and did not comprise a major portion of 
the food on a weight basis. Decapods and cephalo- 
pods were moderately important in stomachs 
examined from both surveys. Copepods and larval 
decapods occurred only in the stomachs from the 
summer survey, while gelatinous zooplankton 
were found only in the seasonal study, and were 
common during late fall and winter. Fish were an 
important component on a weight basis; they were 
mainly mesopelagic species and juvenile stages of 
predominantly benthic species, although many 
adult Pacific herring, Clupea harengus pallasi, 
and some smelts were also found. The mean 
number of taxa and mean number of myctophids 
per stomach were higher in fish from the seasonal 
than those from the summer survey. 

Sebastes pinniger had a much more limited diet 
both in number of prey species and major prey 
categories consumed than S. flavidus (Table 5). 
Euphausiids were again the dominant prey con- 
sumed with proportional abundances and weights 
exceeding 90% of the total in both surveys. Many 
stomachs were distended with adult euphausiids 
(>1,000 individuals). Hyperiid and gammarid 
amphipods were common but did not appear to be 
important components of the diet. Mesopelagic 
fishes, including myctophids and stomiatoids, con- 
tributed to the biomass consumed during the fall 
and winter months of the seasonal survey. There 
was a low number of taxa represented in each 
stomach, especially in the summer survey. 

Because of the advanced stage of digestion of 
most of the stomach contents (mean digestion 



score = 1.05), many taxa were not identified to 
species in the stomachs of S. alutus, although 
many major prey categories were represented 
(Table 6). Euphausiids were the principal prey 
by weight and number. Of the remaining prey 
species, amphipods were relatively common and 
numerous. The oceanic shrimp, Sergestes similis, 
appeared in a significant number of stomachs and 
may constitute an important prey item. Remains 
of fishes were found in only a few stomachs, a 
noteworthy difference compared with the other 
four species examined. 

Sebastes diploproa utilized a spectrum of prey 
items as wide as that of S. flavidus, but the 
smaller mean size of this species is reflected in 
generally smaller prey taken (Table 7). Euphau- 
siids were less important, and amphipods, cope- 
pods, and decapods were more important on a 
numerical and percentage occurrence basis than 
for the other species. Sergestes similis contributed 
heavily in all respects and was found in almost 
half the stomachs examined. The small hyperiid 
amphipod, Vibilia propinqua, was common and 
numerous but contributed little to the bulk of the 
diet. The mean number of prey found per stomach 
was second only to the seasonal number of S. 
flavidus. 

The diet of S. crameri was characterized by very 
few prey taxa, perhaps because only 30 stomachs 
were examined (Table 8). Of these, one-third of the 
stomachs were empty and only about one-third of 
the total biomass found in these stomachs was 
identifiable, resulting in very low mean fullness 
and digestion scores (1.03 and 1.05, respectively). 
This identifiable fraction was composed of equal 
numbers of euphausiids, amphipods, and cope- 
pods. Euphausiids contributed a greater share to 
the total biomass, however, and completely domi- 
nated the identifiable contents. Few prey taxa 
were found, overall, in the stomachs of S. crameri. 

Diet Breadth and Overlap 

In order to quantify the relative food resource 
used by the various species, niche breadth mea- 
sures were calculated for all species. The principal 
prey types (proportional biomasses exceeding 
1.07c of the total biomass), and niche breadth 
values (overall and normalized) are given in Table 
9 for all species analyzed from the summer sur- 
veys and for S. pinniger and S. flavidus collected 
during the seasonal surveys. 

Sebastes flavidus utilized the greatest number 
of prey types (R), had the widest niche breadth 



275 



FISHERY BULLETIN: VOL. 82, NO. 2 



TABLE 4. — Summary of yellowtail rockfish, Sebastes flavidus, stomach contents from the 
Oregon Department of Fish and Wildlife's seasonal and the National Marine Fisheries 
Service's summer samplings. F.O. = frequency of occurrence. 







Seasonal 




Summer 




FO 
(%) 


Number Weight (g) 


FO 
(%) 


Number Weight (g) 


Prey organism 


Mean % Mean % 


Mean % Mean % 



Euphausiacea 

Euphausia pacifica (juv.) 

Euphausia pacifica (adults) 

Thysanoessa spinifera (juv ) 

T. spinifera (adults) 

T longipes 

Thysanopoda acutifrons 

Euphausild unidentified 
Amphipoda 

Phronima sedentana 

Paraphronima gracilis 

Parathemisto pacifica 

Hyperia medusarum 

Hyperoche medusarum 

Streetsia challenger/ 

Vibilia propmqua 

Primno macropa 

Hyperndea unidentified 

Rhacotropis sp. 
Decapoda 

Sergestes similis 

Pandalus jordani 

Munida quadrispina (|uv) 

Pinnothendae megalopae 

Cancer sp megalopae 

Decapod mysis larvae 
Copepoda 

Calanus pacificus 

C marshallae 

Neocalanus sp. 

Euchirella sp 

Copepod unidentified 
Cephalopoda 

Abraliopsis felis 

Gonatus sp 

Loligo opalescens 

Japatella heathi 

Octopus sp. (juv.) 

Cephalopod unidentified 
Miscellaneous invertebrates 

Sagitta elegans 

bmacina helicma 

Alciopid polychaete 

Siphonophora 

Ctenophora 

Cnidana 
Osteichthyes 

Clupea harengus pallasi 

Thaleichthys pacificus 

Spirinchus starksi 

Stenobrachius leucopsarus 

Diaphus theta 

Tarletonbeania crenularis 

Symbolophorus californiensis 

Protomyctophum crocken 

Myctophidae unidentified 

Argyropelecus aculeatus 

Chauliodus macouni 

Nectoliparis pelagicus 

Lipandidae unidentified 

Stichaeidae unidentified (juv) 

Sebastes sp (juv) 

Glyptocephalus zachirus 

Lyopsetta exilis (juv) 

Psettichthys melanostictus (juv) 

Unidentified fish larvae 

Fish remains 
Unidentified animal remains 



367 24.5 6.4 0.34 2.3 — — — 

608 120 1 52.2 2 57 28.3 40 5 37 4 513 190 26.4 

— — — — — 6.0 11.5 2.3 43 9 

683 406 19.8 1.32 167 23 2 88 68 80 6.4 

— — 0.5 1.0 — 0.01 — 

1.3 1.0 — 0.12 — — — — — — 

49.4 56.3 19 9 108 9.9 16 7 61.1 34 7 2 60 14 9 

7.6 12 — 0.11 0.2 3.2 1.8 0.2 0.06 — 

1.3 10 — 001 — 1.1 10 — 0.01 — 

1.3 10 — 0.01 — 2.7 1.0 0.1 001 — 

2.5 1.0 — 0.01 — 2.7 1.2 0.1 01 — 

2.5 40 — 0.02 — 4.9 1.7 03 89 1.5 

38 1.3 — 0.03 — 05 1.0 — 0.04 — 

1.3 2.0 — 0.16 — 0.5 1.0 — 0.01 — 

3.8 1.0 — 0.02 — 0.5 2.0 — 02 — 

1.3 2.0 — 0.02 — — — — — 

1.3 1.0 — 0.01 — — — — — — 



7.6 


2.5 


0.1 


0.71 


1.0 


2.7 


2.8 


0.2 


0.75 


0.7 


1.3 


12.0 


0.1 


370 


0.9 


1.1 


1 


— 


5.19 


1.9 


3.8 


93 


0.2 


0.22 


02 


27 


5.6 


0.5 


.12 


0.1 


— 


— 


— 


— 


— 


0.5 


1.0 


— 


0.10 


— 












4.3 


1.9 


03 


02 


— 












1.6 


2.7 


0.1 


0.04 


— 


— 


— 


— 


— 





05 


1 





001 















1.6 


5.7 


0.3 


0.01 


— 












2.7 


44 


0.4 


0.01 


— 


— 


— 


— 


— 


— 


0.5 


10 


— 


0.02 


— 












22 


4.5 


0.3 


001 


— 


1.3 


1 


— 


0.93 


02 












1.3 


30 


— 


068 


0.2 


1.1 


10 


— 


057 


— 


38 


1.0 


— 


21.24 


149 


22 


1.5 


1 


2.26 


1.7 












1 6 


1.0 


— 


068 


0.3 


6.3 


26 


0.1 


071 


08 


6.5 


6.5 


0.3 


1.33 


2.1 


11.4 


1.2 


0.1 


1.82 


3.8 


22 


2.2 


0.2 


234 


1.7 



2.5 2.5 — 0.16 — 05 10 — 03 — 

1.3 1.0 — 0.01 — 5.4 1.5 0.3 0.04 — 

1.3 10 — 0.27 — — — — — — 

2.5 5.5 0.1 0.54 0.3 — — — — — 

1.3 8 — 1.56 0.3 — — — — — 

1.3 10 — 27 — — — — — 

— — — 3.8 1 6 1 14.17 184 
1.3 10 — 1.22 0.3 — — — — — 

2.5 2.0 — 1.65 0.7 5 10 — 0.28 — 

1.3 1.0 — 0.84 — 0.5 1.0 — 0.99 0.2 

2.5 1.0 — 1.97 0.9 0.5 2 — 9.96 18 

5.1 1.7 0.1 4.94 4.6 — — — — — 

1.3 10 — 0.07 — — — — — — 

1.3 10 — 0.14 — — — — — — 

11.4 1.3 0.1 1.37 2.9 0.5 1.0 — 0.71 0.1 

1.3 10 — 2.49 0.6 — — — — 

1.3 10 — 3 81 0.8 — — — — — 

— — 1.6 1.3 — 0.22 01 
1.3 10 — 0.17 — — — — — — 

1.3 20 — 0.44 0.1 1.1 10 — 026 01 

2.5 10 — 0.37 0.2 1.1 2.0 — 1.07 4 

1.3 10 — 0.88 0.2 — — — — — 

— — 1.1 1.5 — 0.29 01 

— — 0.5 1.0 — 0.06 — 

— — — — — 1.1 1.0 — 0.14 — 
152 — — 1.31 3.7 81 — 4.19 11.6 
30.4 — — 066 3.7 38 4 — 64 84 



276 



BRODEUR and PEARCY: FOOD HABITS AND DIETARY OVERLAP OF SOME SEBASTES 
TABLE 4.— Continued. 



Predator characteristics 




Number of stomachs examined 


79 


Number of empty stomachs 


4 


Mean weight per stomach 


5.192g ± 7 004 (SD) 


Mean total length 


478 35 mm r 29 06 (SD) 


Mean fullness score: 


2.87 


Mean digestion score 


290 


Mean no. prey taxa per fish 


281 



185 

38 

2.905 g ± 6.032 (SD) 

444 86 mm • 51 43 (SD) 

1.92 

1.95 

1.55 



TABLE 5. — Summary of canary rockfish, Sebastes pinniger, stomach contents from the 
Oregon Department of Fish and Wildlife's seasonal and the National Marine Fisheries 
Service's summer samplings. FO. = frequency of occurrence. 







Seasonal 








Summer 










Number 


Weight 


(q) 




Number 


Weight (g) 




FO 
(%) 










FO 










Prey organism 


Mean 


O 

o 


Mean 


°0 


(%) 


Mean 


% 


Mean 


% 


Euphausiacea 






















Euphausia pacitica (|uv ) 


21.0 


88 1 


12.1 


065 


48 


38 


125 


0.4 


0.09 


1 


E pacitica (adults) 


546 


141 3 


503 


3.21 


62.0 


408 


124 3 


51.6 


5 74 


435 


Thysanoessa spinilera 


22.3 


240 


3.5 


0.59 


4.6 


14 6 


64 7 


9.6 


740 


20.1 


Thysanopoda sp 


04 


34 


— 


003 


— 


— 


— 


— 


— 


— 


Euphausnd unidentified 


374 


139 6 


34 


1 54 


20 4 


26 1 


141 5 


37 7 


579 


28.1 


Mysidacea 






















Inusitatomysis sp 


— 


— 


— 


— 


— 


08 


2.0 


— 


0.02 


— 


Amphipoda 






















Parathemisto pacifica 


08 


2 


— 


01 


— 


46 


1.2 


— 


01 


— 


Hyperoche medusarum 


— 


— 


— 


— 


— 


08 


20 


— 


001 


— 


Phronima sedentana 


0.4 


1 


— 


0.03 














Streetsia challenger* 


04 


20 


— 


03 














Hypemdea unidentified 


— 


— 


— 


— 


— 


1.5 


1.0 


— 


001 


— 


Rhacotropis sp 


04 


4 


— 


005 


— 


1 5 


60 


1 


007 


— 


Atylus tndens 


04 


1 


— 


001 














Anonyx sp 


08 


1 5 


— 


18 














Lysianassidae unidentified 


— 


— 


— 


— 


— 


0.8 


1 


— 


0.02 


— 


Decapoda 






















Sergestes similis 


2.9 


1.7 


— 


0.09 


0.1 


1.5 


140 


0.2 


1.89 


0.6 


Pandalus /ordani 


04 


1 


— 


1 05 


1 


1.5 


1 


— 


1.55 


0.4 


Crangon sp 


— 


— 


— 


— 


— 


08 


1 


— 


0.03 


— 


Munida quadnspina (juv.) 


2.5 


50 


0.1 


006 














Chaetognatha 






















Sagitia elegans 


04 


6.0 


— 


0.07 














Osteichthyes 






















Stenobrachius leucopsarus 


0.8 


1.5 


— 


78 


0.2 


08 


10 


— 


0.59 


0.1 


Tarletonbeania crenulans 


0.4 


1.0 


— 


1 70 


0.3 












Myctophidae unidentified 


1.3 


1.3 


— 


1.47 


0.6 












Tactostoma macropus 


0.4 


1.0 


— 


1.73 


0.2 












Argyropelecus aculeatus 


0.4 


1.0 


— 


021 














Ammodytes hexapterus 












3.1 


4 5 


0.1 


76 


0.4 


Sebastes /ordani 


08 


1.0 


— 


1904 


56 












Fish remains 


84 


— 


— 


0.39 


1.2 


108 


— 


— 


300 


6.0 


Unidentified animal remains 


122 


— 


— 


003 


0.1 


42.3 


— 


— 


0.09 


0.7 


Predator characteristics 






















Number of stomachs examined 






238 










130 






Number of empty stomachs: 






39 










18 






Mean weight per stomach: 




2 828 


g ± 4.440 (SD) 






5.385 


g± 11 


.297 (SD 




Mean total length. 




191.45 mm ± 51.07 (SD) 




504 07 mm ± 50.34 (SD) 




Mean fullness score. 






202 










1.68 






Mean digestion score 






1.89 










1.55 






Mean no. prey taxa per fish: 






1.27 










1.00 







277 



FISHERY BULLETIN: VOL. 82, NO. 2 



TABLE 6. — Summary of Pacific ocean perch, Sebastes alutus, 
stomach contents from the National Marine Fisheries Service's 
summer sampling. F.O. = frequency of occurrence. 



TABLE 7. — Summary of splitnose rockfish, Sebastes diploproa, 
stomach contents from the National Marine Fisheries Service's 
summer sampling. F.O. = frequency of occurrence. 



Number of 
prey 



Weight of 
prey (g) 



Prey organism 


(%) 


Mean 


% 


Mean 


% 


Euphausiacea 












Euphausia pacifica 


52.1 


20.5 


62.4 


1 .12 


63.1 


Thysanoessa spinifera 


19.2 


7.1 


80 


0.47 


9.8 


Euphausnd unidentified 


20.6 


16.9 


20.3 


0.55 


12.2 


Amphipoda 












Phronima sedentaria 


2.7 


2.0 


0.3 


0.11 


03 


Paraphronima gracilis 


1.4 


1 


— 


0.02 


— 


Parathemisto pacifica 


6.8 


3.2 


1.3 


03 


0.2 


Vibilia propinqua 


1.4 


11.0 


09 


0.12 


0.2 


Pnmno macropa 


2.7 


10 


02 


0.02 


— 


Hyperndea unidentified 


6.8 


24 


1.0 


0.01 


— 


Cyphocans challenger/ 


2.7 


1.0 


0.2 


0.03 


0.1 


Copepoda 












Neocalanus plumchrus 


4.1 


1.3 


0.3 


001 


— 


Euchaeta sp. 


2.7 


3.0 


0.4 


0.01 


— 


Decapoda 












Sergesles similis 


20.6 


3 1 


3.7 


0.34 


7.5 


Pasiphaea pacifica 


1.4 


1 


— 


0.03 


— 


Decapod mysis larvae 


1 4 


1.0 


— 


0.01 


— 


Crustacea remains 


2.7 


— 


— 


0.19 


0.6 


Cephalopoda 












Loligo opalescens 


1.4 


1.0 


— 


0.53 


0.8 


Cephalopod unidentified 


6.8 


14 


0.5 


0.22 


16 


Osteichthyes remains 


5.5 


— 


— 


0.04 


0.1 


Predator characteristics 












Number of stomachs examined: 






73 






Number of empty stomachs: 






26 






Mean weight per stomach 




0923 


g ± 1 


954 (SD) 




Mean total length 




365 36 mm ± 60.01 (SD) 




Mean fullness score 






1.49 






Mean digestion score: 






1.05 






Mean no prey taxa per fish: 






1.68 







(B), and had the most even distribution among 
prey types (B n ) of all rockfish examined from the 
summer survey. Sebastes diploproa preyed on 
fewer taxa than S. flavidus but had moderately 
high overall and normalized food breadth values. 
Sebastes pinniger, S. crameri, and S. alutus uti- 
lized a similar number of distinct prey items and 
had similar breadth and evenness values with S. 
alutus having a more equitable distribution of 
prey than the other two. 

The seasonal results for the S. flavidus and S. 
pinniger were more divergent and represent the 
extreme values found among the species. Seven- 
teen principal prey types were important in the 
seasonal diet of S. flavidus, contributing toward 
a high B value. However, the dominance of a 
few species yielded a low evenness value for this 
species. Sebastes pinniger preyed on few taxa 
in fairly unequal proportions yielding fairly low 
niche breadth and evenness values. These low 
evenness values could be caused by the prepon- 
derance of euphausiids found in the guts of both 
species during the summer months. 

The individual overlap coefficients and the 
mean overlap for each species are presented for 



Number of 
prey 



Weight of 
prey (g) 



Prey organism 


(%) 


Mean 


% 


Mean 


% 


Euphausiacea 












Euphausia pacifica 


46.8 


26.5 


41.2 


1.53 


42.1 


Thysanoessa spinifera 


145 


29 


1.4 


0.16 


1.4 


Euphausnd remains 


290 


349 


33.7 


1 79 


306 


Amphipoda 












Parathemisto pacifica 


1 6 


1.0 


— 


001 


— 


Hyperoche medusarum 


32 


1.0 


— 


0.01 


— 


Paraphronima gracilis 


3.2 


1.0 


— 


002 


— 


Streetsia challenger/ 


1.6 


1.0 


— 


02 


— 


Vibilia propinqua 


32.3 


10.3 


111 


0.10 


19 


Pnmno macropa 


3.2 


10 


— 


0.01 


— 


Hyperndea unidentified 


97 


1.5 


0.4 


0.02 


1 


Cyphocans challenger/ 


4.8 


1.7 


03 


0.02 


— 


Lysianassidae unidentified 


1.6 


1.0 


— 


003 


— 


Gammandea unidentified 


1 6 


1.0 


— 


003 


— 


Isopoda unidentified 


1.6 


1.0 


— 


0.02 


— 


Copepoda 












Neocalanus cristatus 


6.5 


3.2 


0.7 


0.02 


0.1 


Euchaeta elongata 


48 


33 


0.5 


0.03 


1 


Euchirella sp 


3.2 


1.5 


0.2 


001 


— 


Candacia bipinnata 


4.8 


3.6 


0.6 


001 


— 


Metndia sp. 


32 


1.0 


0.1 


01 


— 


Decapoda 












Sergestes similis 


46.8 


44 


68 


60 


165 


Pasaphaea pacifica 


1.6 


1.0 


— 


59 


0.6 


Benthogenema burkenroadi 


1.6 


1.0 


— 


12 


0.1 


Munida quadnspina 


1.6 


60 


0.3 


0.12 


0.1 


Cancer sp. megalopae 


97 


1.3 


04 


002 


1 


Decapod mysis larvae 


1.6 


10 


— 


001 


— 


Mollusca 












Pteropoda unidentified 


1.6 


10 


— 


0.03 


— 


Gonatus sp. 


1.6 


1.0 


— 


07 


— 


Octopus sp. (juv.) 


1.6 


10 


— 


0.17 


0.2 


Osteichthyes 












Stenobrachius leucopsarus 


1.6 


1.0 


— 


036 


0.3 


Myctophidae unidentified 


6.5 


1.0 


0.2 


0.13 


0.5 


Tactostoma macropus 


1.6 


1.0 


— 


2.28 


22 


Lipandidae unidentified 


1.6 


1.0 


— 


0.15 


0.1 


Fish remains 


9.7 


— 


— 


0.38 


0.1 


Unidentified animal remains 


17.7 


— 


— 


0.03 


03 


Predator characteristics 












Number of stomachs examined: 






62 






Number of empty stomachs: 






15 






Mean weight per stomach: 




1 698 


g ± 3 449 (SD) 




Mean total length: 




264 82 mm ± 41 


.82 (SD) 




Mean fullness score: 






250 






Mean digestion score: 






1.25 






Mean no prey taxa per fish: 






2.48 







both the weight and numerical abundance of prey 
in Table 10 for the summer surveys. As overlap 
indices are affected by the level of taxonomic 
specificity at which the prey have been identified, 
no unbiased means for testing the significance of 
these values are available. We adopted the con- 
vention that overlap values from 0.00 to 0.29 are 
considered low, 0.30 to 0.60 considered medium, 
and those above 0.60 show highly similar diets 
(Langton 1982). 

The coefficients for numerical composition show 
high values for all possible combinations except 
those involving S. crameri. Very similar propor- 
tions of the major euphausiid prey groups resulted 
in an extremely high overlap value (0.93) between 



278 



BRODEUR and PEARCY: FOOD HABITS AND DIETARY OVERLAP OF SOME SEBASTES 



TABLE 8. — Summary of darkblotched rockfish, Sebastes cra- 
meri, stomach contents from the National Marine Fisheries 
Service's summer sampling. F.O. = frequency of occurrence. 







Numbe 


rof 


Weighl 


of 




FO 
(%) 


prey 


1 


prey ( 
Mean 


9) 


Prey organism 


Mean 


% 


% 


Euphausiacea 












Euphausia pacifica 


133 


80 


372 


042 


26.2 


Thysanoessa spimfera 


3.3 


1 


1.2 


0.06 


09 


Euphausnd remains 


33 


1 


12 


0.01 


02 


Amphipoda 












Parathemisto pacifica 


16.7 


44 


256 


01 


1.0 


Cyphocans challengeri 


6.7 


1.0 


2.3 


0.01 


0.3 


Lysianassidae unidentified 


3.3 


10 


1 2 


004 


08 


Copepoda 












Neocalanus cristatus 


3.3 


1 


1.2 


01 


02 


Euchaeta elongata 


100 


30 


105 


01 


0.5 


Copepod unidentified 


16 7 


30 


174 


001 


0.8 


Decapoda 












Sergestes similis 


33 


1 


1 2 


007 


11 


Osteichthyes 












Ammodytes hexaplerus 


3.3 


1 


1 2 


028 


4.3 


Unidentified animal remains 


53 3 


— 


— 


025 


625 


Predator characteristics 












Number of stomachs examined 






30 






Number of empty stomachs 






10 






Mean weight per stomach 




246 


g i 0.389 (SD) 




Mean total length 




330.36 mm ± 77.17 (SD) 




Mean fullness score 






1 03 






Mean digestion score 






1 05 






Mean no prey taxa per fish 






1.26 







pinniger, S. diploproa, and S. alutus are all rela- 
tively high (0.58, 0.56, and 0.61, respectively). 

Overlaps between S. pinniger and S. flavidus 
for the seasonal cruises are similar to the results of 
the summer surveys (dh = 0.80 by number; 0.46 
by weight). A possible explanation for the lower 
values may be changes in availability of both 
predator and prey (i.e., no S. flavidus stomachs 
were collected during spring and early summer 
when the euphausiid populations are generally 
the highest). The variability associated with the 
different cruises was examined by calculating the 
overlaps between these two species for the four 
seasonal cruises that contained at least 10 speci- 
mens of each species. The July cruise had the 
highest overlap of all on a weight basis (dh = 
0.88) and the September cruise had the lowest 
(dh = 0.32), while the December and January 
cruises had intermediate overlaps (dh = 0.52 
and 0.46), suggesting seasonal variations in prey 
availability for these species. 

For comparative purposes, the dietary composi- 



TABLE 9. — Principal prey types making up >1.09t of the diet and food breadths of the five 
species of Sebastes. R is the total number of distinct prey items identified to at least genus 
level and that make up 0.1' < ip, 0.0011 of the identified fraction of the total weight. These 
prey were used to calculate the overall diet breadth (B i and the evenness of distribution of 
the prey items in the diet iB n I. The seasonal values for S. flavidus and S. pinniger are given 
in parentheses. 





Sample 
size 


Principal prey types 
(Pi's > 0.01) 


Pooled species 


values 


Species 


R 


B 


B n 


S. flavidus 


185 

(79) 


Euphausia pacifica , Thysanoessa spinifera . 
hypernd amphipods. Sergestes similis. 
Loligo opalescens. myctophids. Clupea 
harengus pallasi 


12 
(17) 


3.64 
(377) 


0.303 
(0.222) 


S diploproa 


62 


E. pacifica. T spinifera, S similis. 
Vibilia propinqua 


8 


228 


0.285 


S pinniger 


130 
(238) 


E pacifica. T spinifera. Sebastes jordani 


8 

(6) 


1.86 

(133) 


0.232 
(0.222) 


S cramen 


30 


E pacifica. calanoid copepods, hypernd 
amphipods. Ammodytes hexapterus 


8 


1.80 


0.225 


S. alutus 


73 


E pacifica . T spinifera . S similis 


7 


1.73 


0.247 



S. pinniger and S. flavidus, although the diets 
are not similar for other prey items. 

Overlap on the basis of weight, which may be a 
better measure of the energy obtained from the 
various food items, indicates high overlap be- 
tween S. pinniger and S. diploproa and between 
S. alutus and S. pinniger, S. diploproa, and S. 
crameri. The rest of the values were <0.60, in- 
cluding S. pinniger with S. flavidus (dh = 0.48). 
The diet of S. flavidus overlaps the least with the 
other species (dh = 0.42) mainly due to its more 
piscivorous habits. The mean overlap values of S. 



tion of the five most important prey categories for 
each of the rockfish species is presented by percent 
number and percent weight in Figures 4 and 5. 
Both figures show the importance of euphausiids 
in all five species. The stomachs of S. crameri 
contained a more equitable distribution of num- 
bers of the major prey groups than the other 
species of rockfishes, with higher proportions of 
amphipods and copepods. Some of this difference 
may be ascribed to the smaller sample size. On a 
weight basis, S. flavidus was unique in that fishes 
and cephalopods were of greater importance in the 



279 



FISHERY BULLETIN: VOL. 82, NO. 2 



TABLE 10. — Overlap matrix for the five species of Sebastes. 
Only those prey that have proportional abundances exceeding 
V , were used in the analysis. Values above the rules are for 
proportional weight overlap and values below are for propor- 
tional abundance. The mean overlap for each species by weight 
and number is given in parentheses directly above and below 
the rules. 





S. pinniger 


S. flavidus 


S. diploproa 


S. crameri 


S alutus 




(0.58) 










S pinniger 


(0.71) 


0.48 
(042) 


0.72 


0.47 


066 

W 


S flavidus 


0.93 


(0.70) 


0.44 
(0.56) 


0.30 


046 E 
I 


S diploproa 


0.76 


0.78 


(0.65) 


0.48 
(0.48) 


0.63 G 

H 


S cramen 


040 


0.40 


0.42 


(0.41) 


0.69 T 
(0.61) 


S alutus 


074 


070 

N 


0.63 
UMBER 


0.42 


(0.62) 



diet of this species than any of the other rockfish. 
Decapods were of moderate importance to S. 
diploproa and, to a lesser extent, S. alutus. Fishes 
were an important food source by weight for all 
rockfishes but S. alutus. 

Seasonal Variation 

Differences in the diet of S. pinniger and S. 
flavidus are summarized in Table 11 for the four 
seasons. The spring cruise shows an extreme dom- 
inance of one prey item, Euphausia pacifica , in the 
diet of S. pinniger. This prey species was found in 
about three-quarters of the stomachs and made up 
almost all the prey biomass. Decapod shrimp and 
fishes were rarely found in the diet at this time. 
Euphausiids also dominated the diet in the sum- 



100-,-= 



90- 



80- 



70- 



m 

i 60. 



50- 



« 40-| 

r- 
z 

Ld 

g 30. 

Ld 
Q- 

20-1 



10- 







S . pinniger 
S . f lavidus 
S . diploproa 
S crameri 
S . alutus 




EUPHAUSIIDS DECAPODS AMPHIPODS COPEPODS 

PRINCIPAL PREY ITEMS 



FISHES 



FIGURE 4. — The proportions of the five major prey taxa found in the five rockfish species based on numerical composition. 



280 



BRODEUR and PEARCY: FOOD HABITS AND DIETARY OVERLAP OF SOME SEBASTES 

TABLE 11. — Variation in major prey taxa composition with season for Sebastes pinruger and S. flauidus. F.O. = frequency of occur- 
rence; % W = percent gravimetric composition; + = a prey category was present but made up O.V'r of the total weight. 



Species 
and 


No. of 
fish 

(% 
empty) 


Euphausia 


Thysanoessa 


Total 














Gelatinous 


pacifica 
FO % W 


spimfera 
F.O. % W 


euphausiids 
F.O. % W 


Decapods 
F.O. % W 


Amphipods 
FO %W 


Cephalopods 
F.O. % W 


Fishes 


zoopla 
F.O. 


ikton 


season 


FO. 


% W 


% W 


Sebastes pinniger 
































Spring' 


42 


73.8 


99.6 


— 


— 


738 


996 


2.4 


+ 


— — 


— — 


24 


04 


— 


— 


(Mar -May) 


(14.3) 






























Summer 


165 


61.2 


75.1 


79 


0.5 


68.5 


94.7 


36 


11 


2.4 + 


— — 


67 


42 


3.6 


+ 


(June-Aug.) 


(10.9) 






























Fall 


117 


35.9 


138 


32.5 


378 


55.6 


84.2 


9.4 


0.2 


120 0.1 


— — 


179 


15.1 


222 


0.3 


(Sept-Nov) 


(205) 






























Winter 


44 


500 


25.9 


477 


120 


81 8 


94.2 


— 


— 


6.8 0.4 


— — 


13.6 


5.5 


4.5 


+ 


(Dec. -Feb) 


(22.7) 






























Sebastes flavidus 
































Spring 2 

































(Mar -May) 
































Summer 


151 


523 


372 


23.2 


0.5 


58.3 


66.7 


9.9 


08 


132 0.1 


152 65 


139 


25.6 


3.3 


04 


(June-Aug.) 


(16.6) 






























Fall 


75 


34.7 


6.1 


427 


29.6 


54.7 


42.2 


93 


13.8 


26 7 8 


6.7 18 


28.0 


40.5 


10.7 


0.8 


(Sept -Nov) 


(22.7) 






























Winter 


38 


81.6 


195 


92.1 


15.3 


94 7 


46.7 


0.5 


0.7 


10.5 + 


289 306 


52.6 


154 


21 .1 


6.6 


(Dec -Feb.) 


(0.0) 































'All collections taken during one cruise All other seasons represent the means of at least two cruises spaced a minimum of 1 mo apart (Tables 1 and 2 give the 
exact dates and samples collected on each cruise). 
2 No stomachs of S flavidus were collected during this season. 



1 00 



90- 



80- 



70 -I 
x 
uj 60 



50- 



40- 



o 30 
cc 

Id 

20 -I 



10- 







lE 



5ta 



1 



jn 



^0 El 



r 



I 



EUPHAUSIIDS DECAPODS AMPHIPODS CEPHALOPODS FISHES 

FIGURE 5. — The proportions of the five major prey taxa found in the five rockfish species based on gravimetric composition. 



281 



FISHERY BULLETIN: VOL. 82, NO. 2 



mer but to a lesser degree. Thysanoessa spinifera 
appeared in the stomachs at this time, but E. 
pacifica continued to be the most important 
euphausiid consumed. Shrimp and fishes were 
slightly more important but together made up 
only a minor portion of the diet. A low percentage 
of empty stomachs occurred in the summer. 

The diet of S. pinniger in the fall showed 
substantial shifts in prey composition. Although 
the frequencies of occurrence were about equal for 
the two species of euphausiids, T. spinifera great- 
ly exceeded E. pacifica by weight. Decapods were 
common but were represented mainly by small 
shrimp (Sergestes similis) and juvenile pelagic 
crabs (Munida quadrispina), which contributed 
little on a weight basis. Amphipods and gelatinous 
zooplankton occurred frequently but were not 
important by weight. Fishes were important by 
occurrence and weight and consisted mostly of 
mesopelagic species and several adult Sebastes 
jordani which made a large contribution to the 
biomass consumed. 

Almost one-quarter of the fish collected in the 
winter had empty stomachs and contained much 
digested material. Euphausia pacifica and 71 
spinifera occurred in about the same number of 
stomachs, but E. pacifica contributed over twice 
as much of the total weight as T. spinifera. 
Subadult E. pacifica were very numerous at 
this time. The fishes consumed were mostly meso- 
pelagic species. 

Sebastes flavidus showed similar trends in 
food resource utilization among the three seasons 
from which collections were made (Table 11). 
Euphausiids, consisting mostly of E. pacifica, 



made up two-thirds of the diet by weight in the 
summer. Fishes were common and contributed 
heavily to the total biomass. Cephalopods were 
next in importance by either occurrence or weight. 
The diet in the fall showed the same shift in 
euphausiid species as was apparent for 8. pinni- 
ger, with T. spinifera the dominant species. Fishes 
were almost as important by weight as euphau- 
siids, but their weight total was mostly composed 
of adult clupeids. Cephalopods were least impor- 
tant in the fall months. 

Euphausiids represented about half the diet 
during the winter, but the remainder was shared 
mostly by cephalopods and fishes. Both species of 
euphausiids were commonly found, but E. pacifica 
(mostly subadults) were slightly more important 
in the overall diet. Cephalopods (mostly adult 
Loligo opalescens and juvenile copepods) did show 
a substantial increase in weight and occurrence 
during these months. Fishes were found in over 
half the stomachs but were mainly juveniles 
of relatively small myctophids. Gelatinous zoo- 
plankton were most common, and decapods were 
least common, during this season. In contrast to 
S. pinniger, all stomachs of this species contained 
some food and many stomachs were full during 
this season. 

Geographic Variation 

Several trends were evident when comparing 
the diet of S. pinniger between regions (Table 12). 
The two southernmost regions had similar diets 
dominated by E. pacifica with T. spinifera repre- 
senting only a minor portion of the diet. Meso- 



TABLE 12. — Variation in major prey taxa composition with geographic area for Sebastes pinniger and S. flavidus. F.O. = frequency 
of occurrence; % W = percent gravimetric composition; + = a prey category was present but made up < 0.1% of the total weight. 





No of 
fish 
(% 


Eupha 


usia 


Thysanoessa 


Total 
















Gelatinous 


Area 


pacif: 


<ca 


spinifera 


euphausiids 


Decapods 


Amphipods 


Cephe 


ilopods 


Fish 


es 


zooplankton 


taken 


empty) 


FO. 


% W 


F.O. 


%W 


F.O. 


% W 


F.O 


%W 


F.O. % W 


F.O. 


% W 


F.O. 


% W 


F.O. 


% W 


Sebastes pinniger 


































Southern 


51 


52.9 


63.1 


98 


0.6 


54.9 


93 1 


58 


0.8 


7.8 + 


— 


— 


5.8 


6.2 


58 


+ 


Oregon 


(13.7) 
































Heceta- 


281 


56.9 


67.1 


19.6 


4.4 


61.2 


91.3 


5.3 


0.8 


3.2 0.1 


— 


— 


11.7 


7.6 


64 


0.1 


Central 


(16.0) 
































Columbia 



































Region 1 


































Washington- 


36 


222 


4.6 


27.8 


57.7 


36.1 


92.8 


2.8 


+ 


22.2 0.1 


— 


— 


250 


6.7 


36 1 


0.4 


Vancouver 


(16.7) 
































Sebastes flavidus 


































Southern 


70 


58.6 


47.6 


30.0 


92 


65.7 


84.9 


10.0 


+ 


14.3 08 


24.3 


13.7 


86 


0.6 


5.7 


0.1 


Oregon 


(17.1) 
































Heceta- 


122 


61.5 


27.9 


49.2 


14.1 


70.5 


50.6 


11.4 


2.2 


18.0 0.3 


16.4 


16.9 


32.8 


27.6 


13.1 


2.3 


Central 


(11.5) 
































Columbia 


22 


27.3 


32 


9.1 


06 


36.4 


12.1 


18.2 


1.0 


4.6 + 


9.1 


0.3 


45.5 


86.5 


— 


— 


Region 


(13.6) 
































Washington - 


50 


28.0 


7.3 


34.0 


120 


48.0 


20.7 


12.0 


17.2 


16.0 0.6 


2.0 


1.0 


26.0 


60.5 


4.0 


+ 


Vancouver 


(26.0) 

































' No stomachs of S pinniger were collected from this region 

282 



BRODEUR and PEARCY: FOOD HABITS AND DIETARY OVERLAP OF SOME SEBASTES 



pelagic fishes and sergestid shrimps were common 
but generally contributed little to the diet on a 
weight basis. 

The northernmost region showed reduced occur- 
rences of euphausiids, overall, but they still com- 
posed a percent weight equivalent to the two 
southern areas. This could have resulted from a 
shift to T. spinifera, which is generally larger 
than E. pacifica, as the main euphausiid con- 
sumed. Decapods were of lesser importance, but 
gelatinous zooplankton were very common in the 
stomachs of fish from this region. This may 
explain the high abundances of hyperiid amphi- 
pods known to be associated with gelatinous 
zooplankton. Fishes were common, especially ju- 
venile Pacific sand lance, Ammodytes hexapterus, 
a prey species found only in the stomachs collected 
from this area. 

Sebastes flavidus showed a different pattern in 
food utilization. A general decrease in euphausiid 
abundance was observed with increasing latitude 
(Table 12). The euphausiids from the southern- 
most regions were mostly E. pacifica, although 
many were unidentified. The only other important 
prey groups in these southern regions were cepha- 
lopods (mostly Loligo opalescens) and relatively 
large fishes such as Pacific herring and mycto- 
phids iDiaphus theta ). All other prey groups were 
common but made little contribution to the diet. 

Specimens of S. flavidus collected in the north- 



ernmost regions consumed substantial amounts of 
fish (mainly clupeids and myctophids). Euphau- 
siids were relatively unimportant in these re- 
gions. As was the case with S. pinniger, T. 
spinifera was the dominant euphausiid eaten 
in the Washington-Vancouver region. Decapods, 
consisting mostly of Pandalus jordani, reached 
their highest proportion of the diet in the north- 
ernmost region. 

Diel Variation 

Both species showed variation in prey composi- 
tion with the diel period (Table 13). Sebastes 
pinniger contained high percentages of euphau- 
siids by weight during all four diel periods, with 
highest percentages occurring in the afternoon 
periods. Fishes, mostly non-mesopelagic species, 
were most important on a weight basis during 
morning and night when they occurred least fre- 
quently. Euphausiids were relatively more impor- 
tant by weight in the two afternoon periods. A 
high proportion of the fishes found in the stomachs 
during the afternoon periods were mesopelagic 
species. 

Sebastes flavidus exhibited the opposite trends 
in food consumption with respect to time of day. 
Euphausiids were found in the highest propor- 
tions by weight during the morning and night 
periods while proportions of fish were substan- 



TABLE 13. — Variation in major prey taxa composition with time of day for Sebastes pinniger and S. flavidus. F.O. = frequency of 
occurrence; 7c W = percent gravimetric composition; + = a prey category was present but made up < 0.195 of the total weight. 



Time 
of day 


No of 

fish 

(% 

empty) 


Euphausia 


Thysanoessa 


Total 
















Gelatinous 


pacifica 
F.O. % W 


spinifera 
FO % W 


euphausiids 
F.O. % W 


Decapods 
F.O. % W 


Amphipods 
F.O. % W 


Cepha 
F.O. 


ilopods 
% W 


Fishes 


zooplankton 


(h) 


F.O 


% W 


F.O % W 


Sebastes pinniger 
































Morning 


68 


25.0 


32.1 


19.1 


5.7 


51.5 


83 1 


7.3 


0.2 


13.2 0.1 


— 


— 


8.8 


15.9 


23.5 0.6 


(0800- 


(19.1) 






























1200) 
































Early aft 


128 


62.5 


51 2 


21.1 


205 


69.5 


95.9 


4.7 


0.5 


6.2 + 


— 


— 


14.1 


3.4 


7.8 + 


(1200- 


(10.2) 






























1600) 
































Late aft 


79 


50.6 


80.2 


16 


1.8 


709 


96.3 


3.8 


0.7 


— — 


— 


— 


13.9 


2.9 


5.1 + 


(1600- 


(17.7) 






























1800) 
































Night 


93 


527 


56.8 


20.4 


5.9 


69.9 


83.4 


4.3 


1.4 


4.3 02 


— 


— 


97 


14.8 


7.5 0.1 


(1800- 


(183) 






























0700) 
































Sebastes flavidus 
































Morning 


81 


432 


38 1 


358 


9 1 


58.0 


767 


25 


1 


3.7 + 


6.2 


5.3 


12.4 


17.1 


49 0.9 


(0800- 


(18.5) 






























1200) 
































Early aft. 


71 


54.9 


255 


28.2 


75 


57.7 


48.8 


14 1 


44 


19.7 02 


15.5 


76 


25.3 


37.4 


7.0 1.3 


(1200- 


(14.5) 






























1600) 
































Late aft 


57 


50.9 


164 


43.9 


18.5 


63.2 


46.2 


7.0 


1 


22.6 03 


21.0 


6.6 


33.3 


46.3 


7.0 0.3 


(1600- 


(17.7) 






























1800) 
































Night 


55 


60.0 


31.8 


43.6 


11.8 


69.1 


50.6 


18.2 


48 


16.4 02 


21.8 


24.9 


27.3 


15.8 


164 3.4 


(1800- 


(12.7) 






























0700) 

































283 



FISHERY BULLETIN: VOL. 82, NO. 2 



tially lower during these periods. Collections 
taken around late afternoon had equal amounts 
of fishes and euphausiids, while those taken at 
night had high occurrences and biomass of cepha- 
lopods (mostly Loligo opalescens) and gelatinous 
zooplankton. 

The mean fullness score, mean digestion score, 
mean weight ratio (equal to the weight of stomach 
contents divided by weight of fish), and the per- 
centage of empty stomachs were plotted for each 
adjusted collection time for both species. Sebastes 
pinniger had a distinct periodicity in its feeding 
cycle (Fig. 6). Peak periods of feeding intensity 
occurred midday and shortly after dusk. One col- 
lection (eight stomachs) taken at 0400 h had low 
values for fullness score and weight ratio but 
average values for digestion score and percentage 
of empty stomachs. The fullness and digestion 
scores (Fig. 6A) follow each other fairly well 
except that the midday digestion peak was several 



hours later than the fullness peak. A very distinct 
peak in the weight ratio at 1200 h and a smaller 
one shortly after dusk are evident (Fig. 6B). 

Sebastes flavidus also appears to show a diel 
periodicity in its feeding pattern (Fig. 7). The 
fullness and digestion scores track each other very 
closely and show distinct peaks of feeding inten- 
sity around noon and shortly after dusk, although 
the number of samples in the latter period was 
limited (Fig. 7A). The actual mean weight ratio 
showed similar trends, but the noon peak is 
somewhat obscured (Fig. 7B). The percentage of 
empty stomachs was highest in the morning and 
remained low through the remainder of the day 
unlike that found for S. pinniger. 

A high degree of variability in the mean weight 
ratio was found several times, especially during 
periods of peak feeding when both totally dis- 
tended and almost empty stomachs were often 
found together. The differences in the weight 



ui 

o 5 
co 



< 2 



(A) 



• • MEAN FULLNESS SCORE 

0---0 MEAN DIGESTION SCORE 



SAMPLE 
SIZE 




12 16 10 30 45 34 10 39 46 33 24 40 21 



SR 



ss 



10 



rr 

I 

Ld 
^ 4 

< 
Id 



(B) 



WT STOM CONT (g) / WT FISH (kg) 



■O PERCENT EMPTY STOMACHS 




100 



CO 

80 5 



-60 



40 



co 



2 



LlI 
20 O 

Ld 
0_ 



0400 0800 



1200 




1600 2000 2400 



SR 



SS 



TIME (hours) 



a. 
o 
o 
co 



< 

Ld 



• • MEAN FULLNESS SCORE 

0---0 MEAN DIGESTION SCORE 



10 



o 

< 
q: 

t- 
I 

Ld 



■z. 








SAMPLE SIZE 17 16 2028 22 12 14 23 46 II 43 10 2 



~r 



-t- 

SR 



SS 



• • WT STOM CONT. (g) / 

WT FISH (kg) 

0----0 PERCENT EMPTY 
STOMACHS 



(B) 




1 s /-Vr. 
/ 1 •, »i 



1 A > < 
^-r-6-cJ- 



-6-0- 



100 



CO 

X 

u 

< 

o 

(- 
co 



60 >- 



80 



40 



20 










0400 0800 1200 1600 2000 2400 



SR SS 

TIME (hours) 



FIGURE 6. — Feeding intensity indices for Sebastes pinniger at 
adjusted times of the day. See text for explanation of indices. 



FIGURE 7. — Feeding intensity indices for Sebastes flavidus at 
adjusted times of the day. See text for explanation of indices. 



284 



BRODEUR and PEARCY: FOOD HABITS AND DIETARY OVERLAP OF SOME SEBASTES 



ratios at the individual times were subjected to an 
analysis of covariance which compared the weight 
ratios adjusted for fish size (Jenkins and Green 
1977). Both S. pinniger (F (12 , 3 54> : 5.68, P < 
0.001) and S. flavidus (F( U ,262) = 6.51, P < 0.001) 
showed significant differences in the mean weight 
ratios over the times tested, implying that feeding 
varied during the diel period. No significant dif- 
ferences (P > 0.05) in stomach fullness were 
associated with size or sex of the predator for 
either species. 

Predator-Size Variation 

The proportion of empty S. pinniger stomachs 
as well as the percent frequency of occurrence and 
percent weight of prey taxa were remarkably 
invariant among the four predator size classes 
(Table 14). Only the largest size class (2:55+ cm) 
shows any substantial variation with a larger pro- 
portion by weight of fishes and a commensurate 
decrease in weight of euphausiids consumed. 
Much of this fish weight was contributed by a few 
individual fish of large relative size (mostly adult 
S. jordam); the frequency of occurrence of fishes 
is only slightly higher for this largest size class. 

Few obvious size-related trends were apparent 
for S. flavidus. The two smallest size classes 
consumed the largest proportion of euphausiids. 
Euphausia pacifica were less important for large 
fish. Decapods and cephalopods showed similar 
trends except that the frequencies of occurrence 
were highest for cephalopods but lowest for deca- 
pods in the largest size class. Fishes were consis- 
tent in their weight and occurrence proportions 



except that one size class (40-<45 cm) had much 
lower proportions than the others. Few trends 
were apparent for either amphipods or gelatinous 
zooplankton although both groups were commonly 
found. 

To determine if different sizes of rockfish se- 
lected different sizes of prey, all fish that con- 
tained measurable prey were grouped into 10 mm 
length intervals and the means and ranges of their 
prey were plotted against fish size (Fig. 8). Al- 
though some exceptions exist, the majority of the 
prey of S. pinniger are found within a narrow 
range of prey sizes, a range (15-27 mm) largely 
determined by adult euphausiids, the dominant 
prey category (Fig. 8). Fishes of the largest two 
size classes consumed larger prey on average, and 
their prey had the largest variation in size due to 
high numbers of both small and large prey con- 
sumed by these fish. No significant relationship 
was found between length of fish and either 
overall size of prey or size of euphausiid prey. 

Sebastes flavidus showed a much greater range 
in the sizes of prey consumed with the variation 
and range in prey length increasing with size 
of predator (Fig. 8). The mean size of prey eaten 
did not appreciably increase until the very largest 
size classes. Although the maximum prey size 
increases with fish size, the minimum size varies 
little throughout the length ranges examined. 
Again for this species, no relationship was found 
between fish length and overall or euphausiid 
prey lengths. 

The size distribution of prey is shown for both 
species in Figure 9. The prey-size spectrum of S. 
pinniger was distributed fairly normally with the 



TABLE 14. — Variation in major prey taxa composition with size of predator for Sebastes pinniger and S. flavidus. F.O. = frequency 
of occurrence; '/ W = percent gravimetric composition; + = a prey category was present but made up < O.l'X of the total weight. 



Size 
range 
(cm) 


No of 
fish 
(% 

empty) 


Euphausia 
pacifica 

FO. % W 


Thysanoessa 
spinifera 

FO. % W 


Total 
euphausiids 

F.O. % W 


Decapods 
F.O. % W 


Amphipods 
F.O. % W 


Cephalopods 
F.O. % W 


Fishes 


Gelatinous 
zooplankton 


F.O. 


% W 


F.O. % W 


Sebastes pinniger 
45 


64 


484 


43.2 


21 9 


9.7 


687 


91.4 


4.7 3.1 


6.2 0.3 






14.1 


48 


4.7 0.4 


45- 50 


(172) 
102 


51.8 


46.1 


176 


186 


676 


92.5 


4.9 0.9 


4.9 0.1 


_ 


_ 


13.7 


6.5 


5.9 0.1 


50- < 55 


(17.6) 

146 


47.3 


65.6 


21.2 


12.0 


61.1 


94.9 


5.5 0.2 


4 1 + 








11.0 


4.7 


8.9 0.1 


'55 


(14.4) 
56 


589 


49.3 


125 


7.6 


67.9 


83.4 


5.4 02 


7.1 + 


_ 





16.1 


163 


143 0.2 


Sebastes flavidus 


(14.3) 




























- 40 


35 


886 


44.7 


57.1 


11.7 


943 


61.3 


11.4 1 


14.3 0.1 


14.3 


2.8 


31.4 


34.6 


29 0.1 


40- 45 


(0.0) 

61 


459 


50.2 


29 5 


12.4 


525 


83.9 


13.1 1.2 


246 0.2 


13.1 


8.4 


9.8 


58 


49 04 


45-<50 


(22.9) 
126 


47.6 


22.9 


38.1 


13.7 


57.1 


46.8 


11.1 3.7 


12.0 0.1 


127 


17.4 


29.3 


29.3 


79 24 


^50 


(21.4) 

42 

(2.4) 


47.6 


25.8 


30.9 


83 


54.7 


51.9 


7.1 2.6 


238 0.3 


21.4 


136 


45.2 


303 


214 0.1 



285 



FISHERY BULLETIN: VOL. 82, NO. 2 



150-7 
125- 



E 

E 100 H 

UJ 
M 

OT 75H 

>- 

Q- 504 



25- 



0- 



400 



S. pinniger 



U?t 



)ii> 



f^ + i |t 



llll 



450 



500 



—I — i — i — •- 

550 



600 



FIGURE 8. — Mean (horizontal lines) ±95% confidence 
limits (boxes) and ranges (vertical lines) of prey sizes 
found for each 10 mm interval of Sebastes pinniger and 
S. flavidus. 



1507 



125- 



100- 



<" 75 
>- 

UJ 

cr 

Q- 50 



25- 



S flavidus 



it^nHt 






iiim 



«fi 



.mi 



} i i i I I « i I I i ' I i  I I I ' I I I I I I I i I t I I 

300 350 400 450 500 550 600 

PREDATOR LENGTH (mm) 



mode coinciding with the mean (x = 10.38 mm), 
although disjunct groups of small and large prey 
were found (Fig. 9). The prey-size spectrum of S. 
flavidus was slightly skewed toward the larger 
sized prey with the mean size (x = 18.44 mm) less 
than the mode. A smaller peak also appeared 
around 25 mm. No significant differences were 
found in the mean prey sizes utilized by the two 
species (Student's f-test, P > 0.05). 

Analysis of Variation 

The results of the chi-square analyses for S. 
pinniger showed that none of the factors analyzed 
had a significant effect on the occurrence of food in 
the stomachs (Table 15). At least one of the factors 
was related to the occurrences of all seven prey 
categories examined. Seasonal effects were the 
most significant (all P ^ 0.01) and were due to the 
higher occurrences of hyperiid amphipods, fishes, 
and gelatinous zooplankton in fall and winter. 
Area and time of capture showed both highly 
significant (P s 0.001) and insignificant effects 



depending on the prey category, but most compar- 
isons were significant at the 0.05 level. In none of 
the prey categories examined did the size of the 
predator have a significant effect on the relative 
proportions consumed. 

For S. flavidus, season of capture and size of 
predator affected the proportion of empty stom- 
achs found (Table 15). Again season had the most 
significant influence on prey occurrence and was 
significant in all eight prey categories. Highly 
significant differences were found in area of cap- 
ture and size of predator especially in the euphau- 
siid and fish categories. Differences in occurrence 
of prey with time of capture deviated from ex- 
pected the least of all the factors analyzed. 

DISCUSSION 

The five species of rockfishes examined rely 
heavily, if not exclusively, on pelagic macrozoo- 
plankton and micronekton. Although some ben- 
thic species appear in the prey lists (e.g., Lyopsetta 
exilis, Munida quadrispina, Psettichthys melan- 



286 



BRODEUR and PEARCY: FOOD HABITS AND DIETARY OVERLAP OF SOME SEBASTES 

TABLE 15. — Results of chi-square analyses testing for differences in the occurrence of food and specific prey categories within the 
various factors. All significances are with three degrees of freedom except where noted. 



Factor 
analyzed 



Occurrence 
of food 



Euphausia 
pacifica 



Thysanoessa 
spinifera 



Total 
euphausiids 



Decapods Amphipods Cephalopods Fishes 



• = P - 0.05. " = P • 0.01. •" = P  001 
1 Significance with two degrees of freedom. 



Gelatinous 
zooplankton 



Sebasfes pinniger 




















Season 


659 


23.09" - 


72.48— 


11 26" 


113 28 - " 


1676*" 





1587— 


39.96— 


Area' 


0.19 


8.61- 


5.22 


11 42— 


0.46 


22.44— 


— 


7.95* 


35.66— 


Time 


432 


28 18*" 


53.77— 


11.86" 


1.37 


13.05— 


— 


1.43 


1921 — 


Size 


0.67 


3.12 


324 


626 


0.04 


098 





1.26 


4.35 


Sebastes flavidus 




















Season' 


979" 


1332— 


51 33— 


10.02" 


30.27— 


11.15*" 


6.65 - 


21 43— 


11.67— 


Area 


5.76 


20.83— 


15.81 — 


1423— 


1.21 


2.76 


11 62" 


22.41 — 


6.43 


Time 


1 12 


3.21 


542 


238 


10.43* 


12.30" 


8.25* 


9.33* 


5.60 


Size 


1750" - 


13 78— 


6.35 


14.17— 


202 


11 69" 


0.94 


12.92" - 


886" 



25 



20 ■• 



< 

o 



(J 

rr 

UJ 

a. 



o 

t- 
u. 
O 



o 



15-' 



S pinniger 



n 




25 



20  



15 •• 



10 • 



5 •• 



S flavidus 



r^ 




o \ >   i ^  ■' '  l m    |   i i i i i i i i  i 1 1 1 | i i i i 

5 10 15 20 25 30 35 

PREY LENGTH INTERVAL (mm) 



FIGURE 9. — Prey size spectra in percent for Sebastes pinniger 
and S. flavidus. 



ostictus ), they were represented by postlarval 
or juvenile forms commonly found in the plank- 
ton. Several comparatively large nektonic fishes 
and cephalopods (e.g., Clupea harengus pallasi, 
Sebastes jordani, Loligo opalescens) were eaten, 
but their occurrences were relatively rare. Con- 
versely, the virtual absence of many common 
benthic and epibenthic organisms of appropriate 
size such as mysids, cumaceans, and gammaridean 
amphipods further implies that these fish do not 
normally feed on benthic animals. 



These findings concur with the limited number 
of previous studies dealing with food habits of off- 
shore rockfish. Phillips (1964) reported on the diet 
of all the species included here except S. alutus. 
Although little taxonomic detail and no quanti- 
tative data on prey consumption were given, 
euphausiids were listed as important forage items 
for all four species. Fishes were also important 
prey for several species, especially S. flavidus. 
Skalkin (1964), in a study of S. alutus in the 
Bering Sea, found mostly euphausiids and cope- 
pods in the stomachs, but also stated that a few 
nektobenthic species and "fragments" of benthic 
echinoderms were present. 

The food habits of S. flavidus have been de- 
scribed in several studies off Oregon and Wash- 
ington. Pereyra et al. (1969) found unusually high 
abundances and volumes of the mesopelagic fish, 
Stenobrachius leucopsarus, in S. flavidus stom- 
achs collected near Astoria Canyon and hypothe- 
sized that local hydrographic conditions may have 
aggregated these prey at high densities. Gunder- 
son et al. (1980) 5 reported that S. flavidus off the 
coast of Washington fed mostly on fishes, includ- 
ing some pleuronectid fishes possibly eaten near 
the bottom along with benthic polychaetes. Lorz et 
al. (1983) found euphausiids dominating the diet 
of S. flavidus off Washington and Queen Char- 
lotte Sound, with fishes of greater importance in 
the latter region. Another deepwater species, 
S. marinus, found in the North Atlantic Ocean, 
also fed chiefly on pelagic prey (Lambert 1960). 
Euphausiids, hyperiid amphipods, and copepods 
were the most abundant prey, but mesopelagic 
fishes were also found in large numbers. 

Among the species considered here, two diver- 
gent feeding patterns are apparent, assuming that 



5 Gunderson, D. R., G. L. Thomas, P. Cullenberg, D. M. Eggers, 
and R. Thome. 1980. Rockfish investigations off the Wash- 
ington coast. Ann. Rep., prep, for NMFS, Univ. Wash., 68 p. 



287 



FISHERY BULLETIN: VOL. 82, NO. 2 



the same prey items are equally available to all 
species. These can be seen most clearly in the 
divergence of the cumulative curves of the number 
of prey species (Fig. 2). Three species (S. pinniger, 
S. alutus, and S. crameri) tend to be steno- 
phagous, with very few prey items represented in 
large volumes of prey organisms. Euphausiids 
appear to be the most sought after or available 
prey, and other prey taxa occur in low numbers. 
These three species show similar low food breadth 
values. 

Sebastes flavidus and S. diploproa , on the other 
hand, have steadily rising prey curves that con- 
tinue to rise and approach an asymptote beyond 
the limits of the figure. These curves are charac- 
teristic of euryphagous predators which show high 
overall prey diversity as well as high within- 
stomach diversity. This high prey diversity can be 
seen in the greater food breadth values attained 
by these two species. Although euphausiids pre- 
dominate in these stomachs, high abundances of 
other prey, which may be preferred but have lower 
abundances and availabilities than euphausiids, 
also occur. 

The diet overlap measurements calculated here 
may be useful in comparing how similar the food 
habits of two species are but may be of limited 
use when interpreted in an ecological sense. The 
interaction of factors that affect or determine the 
diet of a particular species is complex and may 
include such factors as temporal and spatial dis- 
tribution of prey, behavioral adaptations of pred- 
ator and prey, prey detection capabilities, and 
feeding morphologies of predators (Hyatt 1979). 
Caution should be exercised when inferences are 
made about possible species interactions based on 
diet overlap measurements alone. Two species 
may have broadly overlapping diets in terms of 
prey composition but segregate with respect to 
prey sizes selected, time of feeding, or habitat 
utilization (Schoener 1974; Ross 1977; Werner 
1979; Macpherson 1981). 

Sebastes pinniger and S. flavidus are two of the 
most abundant rockfish species within the geo- 
graphical confines of this study. They inhabit 
similar depth ranges, latitudinal ranges, and 
show broadly overlapping areas of peak abun- 
dances according to trawl survey data (Alverson et 
al. 1964; Richardson and Laroche 1979; Gunderson 
and Sample 1980). Adams (1980) found that these 
two species had the highest positive association in 
trawl catches using presence-absence data of the 
seven abundant species he examined. Little is 
known, however, about their small-scale hori- 



zontal and vertical distribution. Although they 
may occupy similar bottom habitat, S. flavidus 
may be more pelagic (Alton 1972). 

Seasonal, geographical, and diel variations in 
the abundance and availability of the important 
prey of S. pinniger and S. flavidus could be a 
major cause of the variations in the diet of these 
species. These variations may be the result of 
intrinsic prey population fluctuations with sea- 
son, behavioral adaptations such as diel and 
ontogenetic vertical migration, or may stem from 
the prevailing oceanographic conditions either 
concentrating, dispersing, or transporting prey 
so that all prey are not equally available in the 
limited time and space frame of the individual 
predator. Current patterns alone are known to 
vary with season, depth, and geographic area 
(Huyer et al. 1975; Ingraham and Love 1978) and 
may affect the availability and concentration 
of prey. 

Quantitative estimates of the seasonal and 
areal distributions of the total prey spectrum 
consumed by these rockfishes are limited. Day 
(1971) sampled macrozooplankton and micronek- 
ton from the northern part of the range of this 
study (lat. 46° 45 '-50° 02' N) using a 0.9 m Isaacs- 
Kidd midwater trawl in the upper 150 m of the 
water column during the spring and fall. He found 
a peak in the biomass of catches at the outer edge 
of the continental shelf. Euph