Full text of "Fishery bulletin"

See other formats

Fishery Bulletin

'ATES O^ ^

Vol. 86, No. 1

Marine Biological Laboratory '
LIBRARY

JUL 6 1000

Woods Hole, m^IS^^'^^^^QS

oplank-

1
25

LOEB, VALERIE J., and OMAR ROJAS. Interannual variation of ichthyopl
ton composition and abundance relations off northern Chile, 1964-85

VETTER, E. F. Estimation of natural mortality in fish stocks: A review ...

NOTARBARTOLO-DI-SCIARA, GUISEPPE. Natural history of the rays on the
genus Mobula in the Gulf of California 45

WILLIAMS, AUSTIN B. Notes on decapod and euphausiid crustaceans, continen-
tal margin, western Atlantic, Georges Bank to western Florida, USA 67

WARLEN, STANLEY M. Age and growth of larval gulf menhaden, Brevoortia
patronus , in the northern Gulf of Mexico 77

O'BRIEN, LORETTA, and RALPH K. MAYO. Sources of variation in catch per
unit effort of yellowtail flounder, Limanda ferruginea (Storer), harvested off the
coast of New England 91

McEACHRON, LAWRENCE W., JEFF F. DOERZBACHER, GARY C. MATLOCK,
ALBERT W. GREEN, and GARY E. SAUL. Reducing the bycatch in a commer-
cial trotline fishery 109

BRUCE, B. D. Larval development of blue grenadier, Macruronus novaezelandiae
(Hector), in Tasmanian waters 119

COWAN, JAMES H., JR., and RICHARD F. SHAW. The distribution abundance,
and transport of larval sciaenids collected during winter and early spring from the
continental shelf waters off west Louisiana 129

HAMNER, WILLIAM M., GREGORY S. STONE, and BRYAN S. OBST. Behavior
of southern right whales, Eubalaena australis, feeding on the Antarctic krill,
Euphausia superba 143

Notes

PENSON, JOHN B., JR., ERNEST 0. TETTY, and WADE L. GRIFFIN. An econo-
metric analysis of net investment in Gulf shrimp fishing vessels 151

SHIRLEY, SUSAN M., and THOMAS C. SHIRLEY. Appendage injury in Dunge-
ness crabs. Cancer magister, in southeastern Alaska 156

(Continued on back cover)

Seattle, Washington

U.S. DEPARTMENT OF COMMERCE
C. William Verity, Jr., Secretary

NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION
William E. Evans, Under Secretary for Oceans and Atmosphere

NATIONAL MARINE FISHERIES SERVICE

Fishery Bulletin

The Fishery Bulletin carries original research reports and technical notes on investigations in fishery science, engineering, and
economics. The Bulletin of the United States Fish Commission was begun in 1881; it became the Bulletin of the Bureau of Fisheries
in 1904 and the Fishery Bulletin of the Fish and Wildlife Service in 1941. Separates were issued as documents through volume 46;
the last document was No. 1103. Beginning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate
appeared as a numbered bulletin. A new system began in 1963 with volume 63 in which papers are bound together in a single issue
of the bulletin instead of being issued individually. Beginning with volume 70, number 1, January 1972, the Fishery Bulletin became
a periodical, issued quarterly. In this form, it is available by subscription from the Superintendent of Documents, U.S. Government
Printing Office, Washington, DC 20402. It is also available free in limited numbers to libraries, research institutions. State and
Federal agencies, and in exchange for other scientific publications.

SCIENTIFIC EDITOR, Fishery Bulletin

Dr. Andrew E. Dizon

Southwest Fisheries Center La Jolla Laboratory

National Marine Fisheries Service, NOAA

P.O. Box 271

La Jolla, CA 92038

Editorial Committee

Dr. Jay Barlow

National Marine Fisheries Service

Dr. William H. Bayliff

Inter-American Tropical Tuna Commission

Dr. George W. Boehlert

National Marine Fisheries Service

Dr. Robert C. Francis
University of Washington

Dr. James R. Kitchell
University of Wisconsin

Dr. William J. Richards
National Marine Fisheries Service

Dr. Bruce B. CoUette

National Marine Fisheries Service

Dr. Tim D. Smith

National Marine Fisheries Service

Mary S. Fukuyama, Managing Editor

The Fishery Bulletin (ISSN 0090-0656) is published quarterly by the Scientific Publications GfTice, National Marine Fisheries Service,
NOAA, 7600 Sand Point Way NE, BIN C15700, Seattle, WA 98115. Second class postage is paid in Seattle, Wash., and additional
offices. POSTMASTER send address changes for subscriptions to Fishery Bulletin, Superintendent of Documents, U.S. Government
Printing Office, Washington, DC 20402.

Although the contents have not been copyrighted and may be reprinted entirely, reference to source is appreciated.

The Secretary of Commerce has determined that the publication of this periodical is necessary in the transaction of the public business
required by law of this Department. Use of funds for printing this periodical has been approved by the Director of the Office of
Management and Budget.

Fishery Bulletin

■ Marine Biolomlil^Z^
CONTENTS / LIBRARY

JUL 6 ma

Vol. 86, No. 1 I January 1988

LOEB, VALERIE J., and OMAR ROJAS. Interannual variation oPrc'Sfhj^^k^k-Mass.
ton composition and abundance relations off northern Chile, 1964-85

VETTER, E. F. Estimation of natural mortality in fish stocks: A review 25

NOTARBARTOLO-DI-SCIARA, GUISEPPE. Natural history of the rays on the

genus Mobula in the Gulf of California 45

WILLIAMS, AUSTIN B. Notes on decapod and euphausiid crustaceans, continen-
tal margin, western Atlantic, Georges Bank to western Florida, USA 67

WARLEN, STANLEY M. Age and growth of larval gulf menhaden, Brevoortia
patronus , in the northern Gulf of Mexico 77

O'BRIEN, LORETTA, and RALPH K. MAYO. Sources of variation in catch per
unit effort of yellowtail flounder, Limanda ferruginea (Storer), harvested off the
coast of New England 91

McEACHRON, LAWRENCE W., JEFF F. DOERZBACHER, GARY C. MATLOCK,
ALBERT W. GREEN, and GARY E. SAUL. Reducing the bycatch in a commer-
cial trotline fishery 109

BRUCE, B. D. Larval development of blue grenadier, Macruronus nouaezelandiae

(Hector), in Tasmanian waters 119

HAMNER, WILLIAM M., GREGORY S. STONE, and BRYAN S. OBST. Behavior
of southern right whales, Eubalaena australis, feeding on the Antarctic krill,
Euphausia superba 143

Notes

PENSON, JOHN B., JR., ERNEST O. TETTY, and WADE L. GRIFFIN. An econo-
metric analysis of net investment in Gulf shrimp fishing vessels 151

SHIRLEY, SUSAN M., and THOMAS C. SHIRLEY. Appendage injury in Dunge-

ness crabs, Cancer magister, in southeastern Alaska 156

iContinued on next page)

Seattle, Washington

1988

For sale by the Superintendent of Documents, U.S. Government Printing Office, Washing-
ton DC 20402— Subscription price per year: $16.00 domestic and $20.00 foreign. Cost per
single issue: $9.00 domestic and $11.25 foreign.

Contents — Continued

CURRNES, KENNETH P., CARL B. SCHRECK, and HIRAM W. LI. Reexamina-
tion of the use of otolith nuclear dimensions to identify juvenile anadromous and
nonanadromous rainbow trout, Salmo gairdneri 160

BUTLER, JOHN L., and DARLENE PICKETT. Age-specific vulnerability of
Pacific sardine, Sardinops sagax, larvae to predation by northern anchovy, En-
graulis mordax 163

EPIFANIO, CHARLES E., DAVID GOSHORN, and TIMOTHY E. TARGETT.

Induction of spawning in the weakfish, Cynoscion regalis 168

The National Marine Fisheries Service (NMFS) does not approve, recommend or
endorse any proprietary product or proprietary material mentioned in this publi-
cation. No reference shall be made to NMFS, or to this publication furnished by
NMFS, in any advertising or sales promotion which would indicate or imply that
NMFS approves, recommends or endorses any proprietary product or proprietary
material mentioned herein, or which has as its purpose an intent to cause directly
or indirectly the advertised product to be used or purchased because of this NMFS
publication.

AWARDS

The Publications Advisory Committee of the National Marine
Fisheries Service has announced the best bublications authored
by the NMFS scientists and published in the Fishery Bulletin for
1986 and the Marine Fisheries Review for 1985. only effective
and interpretive articles which significantly contribute to the
understanding and knowledge of NMFS mission-related studies
are eligible, and the following papers were judged as the best
In meeting this requirement.

Fishery Bulletin 1986 — Starvation-induced mortalitv of young
sea-caught jack mackerel, Trachurussymmetrlcus, determined
with histological and morphological methods, by Gail M.
Theilacker. Fish. Bull., U.S. 84: 1-17. Gail M. Theilacker is with the
Southwest Fisheries Center La Jolla Laboratory National Marine
Fisheries Service, NOAA, La Jolla, CA.

Marine Fisheries Review 1985 — Biology of the red sea urchin,
Strongylocentrotus franclscanus, and its fishery in California, by
Susumu Kato and Stephen c. schroeter. Mar. Fish. Rev 47(3):i-20.
Susumu Kato is with the Southwest Fisheries center Tiburon
Laboratory National Marine Fisheries Service, NOAA, Tiburon, CA,
and Stephen C. schroeter is with the Department of Biological
Sciences, university of California, university Park, CA.

INTERANNUAL VARIATION OF ICHTHYOPLANKTON COMPOSITION
AND ABUNDANCE RELATIONS OFF NORTHERN CHILE, 1964-83

VaLKRIK J LOEB' AND OMAR ROJAS^

ABSTRACT

Larval fishes were collected off of northern Chile during winter (July-September) ichthyoplankton
surveys undertaken in 1964-70, 1972-73, and 1983. The 19-year timespan included a wide variety
of hydrographic conditions in the Humboldt Current area (cold years. El Nino events, and intervening
transition years); it also included the decline and collapse of the anchoveta fisheries and increases of
sardine, mackerel, and jack mackerel stocks off northern Chile and Peru. The ichthyoplankton data
are examined in relation to ambient hydrographic conditions as well as to possible chronological
changes in environmental conditions which led to the increased Chilean sardine stocks and anchoveta
fishery collapse.

More coherent patterns come from considerations of larval fish species composition in 1964-69 and
1970-73 data sets than from years of "similar" hydrographic conditions. A marked shift in relative
abundances of nonfished mesopelagic species in 1969-70 is associated with changes within long-term
physical data bases from Chile and Peru suggesting a large-scale environmental change. Sardine
stock growth began with successful larval survival of 1968-69 and later year classes. Anchoveta stock
decline began in 1972 probably due to poor larval survival. Affiliation of anchoveta and coastal
species larval abundance implies that they are similarly influenced by coastal processes. An atmos-
pherically driven oceanic circulation change beginning in the late 1960's and possibly involving
onshore presence of subtropical and or oceanic waters and altered coastal processes may have been
responsible for the changes in the northern Chilean fish assemblages.

The Humboldt Current region, like the other
major eastern boundary current systems (Califor-
nia, Benguela, and Canary Currents) is domi-
nated by pelagic schooling fish stocks including
anchoveta (Engraulis), sardine iSardinops),
hake iMerluccius), mackerel (Scomber), jack
mackerel (Trachurus), and bonita iSarda) (Par-
rish et al. 1983). These fish stocks, like those in
the other eastern boundary current areas, exhibit
extreme population fiuctuations. Most notable in
the past 30 years are the collapses of Peruvian
and Chilean anchoveta stocks in the mid-1970's
and their succession by sardine and, to a lesser
extent, mackerel and jack mackerel stocks (San-
tander and Flores 1983; Serra 1983).

Hydrographic complexity and variability are
characteristic of eastern boundary current sys-
tems. Included in the Humboldt Current region
are equatorial, subequatorial, subantarctic, and
antarctic oceanic water masses; northward flow-
ing currents and opposing countercurrents; and
wind driven, seasonally variable coastal up-
welling (Wyrtki 1967). Additionally, the region is

iMoss Landing Marine Laboratories, P.O. Box 450, Moss
Landing, CA 95039.

2Instituto de Fomento Pesquero, Avenida Pedro de Valdivia
2633, Casilla 1287, Santiago, Chile.

subject to 1) large seasonal and longer period
fluctuations in advection of water masses of
markedly different properties and 2) large in-
terannual differences in the timing and intensity
of seasonal upwelling processes (Bakun 1987;
Bernal et al. 1983; Parrish et al. 1983; Robles et
al. 1976). The clearest and generally considered
most important of the nonseasonal processes in-
fluencing the biology of the current system is the
El Nino phenomenon (Bernal et al. 1983; Guillen
1983). El Nino events off Peru and Chile are
marked by large-scale atmospherically driven
southward and coastward advection of warm,
high-salinity equatorial and subequatorial sur-
face waters, weakening of coastal upwelling (or
upwelling of warm nutrient-poor waters), and
weakening of subsequent phytoplankton blooms.
These El Nino or warm-water periods are vari-
able in their intensity and duration (Guillen
1983; Santander and Flores 1983). In contrast to
these periods are more "normal" cold-water
events resulting from atmospherically driven in-
tensification of northward flowing cold, low-
salinity subantarctic waters and seasonal up-
welling of cold, nutrient-rich water. Major El
Nino events occurred in 1891, 1925-26, 1940-41,
1957-58, 1965, 1972-73, 1976, and 1982-83;

Manu.script accepted September 1987.
FISHERY BULLETIN: VOL. 86, NO. 1, 1988.

FISHERY HULLKTIN: VOL 8(i, NO 1

major cold events over the past 20 years occurred
in 1964, 1967-68, 1970-71, and 1974-75 (Guillen
1983).

The decline and ultimate collapse of the an-
choveta fisheries of Peru began in 1970 and was
finalized by the intense 1972-73 El Nino; the
northern Chilean stock decline started in 1972
and was finalized by 1977. Factors facilitating
these declines are generally believed to include
overfishing and the devastating effects of the El
Nino on anchoveta spawning behavior and in-
tensity as well as on subsequent recruitment.
Competition and/or predation pressure result-
ing from increasing abundances and distributions
of sardine and mackerel have also been hy-
pothesized (Santander and Flores 1983; Serra
1983).

Because of the great socioeconomic value of the
dominant pelagic fish species of the Peru-Chile
ecosystem, their population fluctuations have re-
ceived a great deal of attention over the past 20
years. However, coincidental changes in the com-
position, abundance, or spawning intensities of
other commercially less important and non-
harvested species have not been examined. Infor-
mation on the changes of these unfished species
in relation to hydrographic conditions and fluctu-
ations of the dominant pelagic fish stocks provide
additional insight into the ecology of the Hum-
boldt Current and may elucidate possible causes
for the dramatic changes which occurred during
the 1970's.

In the present work we examine the abundance
and composition of total ichthyoplankton assem-
blages collected off of northern Chile (lat. 18°-
24°S) during 1964-73 and 1983 in relation to am-
bient hydrographic conditions. "Normal" cold
water as well as warm-water and El Nino events
occurred during the 19-yr sampling span. We also
examine our results with respect to possible
chronological change in environmental condi-
tions which led to the 1977 anchoveta fishery col-
lapse off northern Chile. Our results may be ap-
plicable for interpreting coincidental changes
in the Peruvian ecosystem and may also be
broadly applicable for studies of similar changes
in the other eastern boundary current ecosys-
tems.

METHODS

Samples were collected during 1964-73 and
1983 ichthyoplankton surveys conducted by the
Instituto de Fomento Pesquero. The area most

intensively surveyed was a narrow coastal strip
extending between Arica and Antofagasta (lat.
18°-24°S, long. 70°-72°W; Fig. 1). This area in-
cludes one of two major anchoveta (Engraulis rin-
gens ) spawning grounds off Chile and the pri-
mary sardine iSardinops sagax) spawning area
off Chile prior to 1973 (Fig. 2A, B). All samples
used for interannual comparisons were collected
during late July-September following peak win-
ter anchoveta and sardine spawning periods. Be-
tween 21 and 87 samples from the 18°-24°S area
were analyzed for each of 11 cruises (Table 1). In
one case data from two cruises (August and Sep-
tember 1968) were pooled to provide adequate
coverage. Sampling was done annually from 1964
to 1970 and in 1972 and 1973. There was a 10-yr
hiatus before regular sampling was resumed in
1983.

The 1964-73 samples were collected with
Hensen nets (0.28 m^ mouth opening; 300 ixm
mesh). Prior to 1973 the vertical net hauls were
50-0 m; in 1973 haul depth was increased to 100
m. The 1983 100-0 m vertical hauls were made
with WP2 nets (0.25 m^ mouth opening;
UNESCO 1968) of 300 ^JLm mesh. Samples were
preserved using buffered 5% formalin solution.
Sea surface temperature and salinity data were
collected at most sampling stations for all but two
winter cruises; these data are lacking for 1970
and salinity data are minimal for 1967.

All fish eggs and larvae were removed from
samples, and invertebrate zooplankton biomass
was measured. Wet weight displacement volume
was measured for 1964-73 samples; in 1983 the
Yashnov (1959) technique modified by Robertson
(1970) was used. A calculated correction factor
of 1.44 (±3.34) was applied to the 1983 biomass
values to permit comparison with the earlier
data.

All fish larvae were identified to lowest taxon
possible and counted. We herein treat the larvae
of six commercially important species (anchoveta
[Engraulis ringens], Pacific sardine [Sardinops
sagax], jack mackerel [Trachurus murphyi; also
known as T. symmetricus in U.S.A.], chub mack-
erel [Scomber japonicus]. South Pacific men-
haden [Ethmidium maculatum], and hake [Mer-
luccius gayi]) separately from the other 35
identified taxa. These six species are referred to
as the "PL" (larvae of pelagic schooling species).
The other larval taxa considered together are the
"OL". The PL and OL categories are treated sepa-
rately because abundances of the PL (especially of
anchoveta and sardine) mask abundance rela-

LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE

90° 75° 60°

45<

z
<

u
o

<
a.

- 30'

45'

60'

Figure 1. — Ichthyoplankton study area between Arica and Antofagasta, Chile (18°-24°S), 1964-83.

tions of the OL. The OL are further separated into
myctophid, "other mesopelagic" and "coastal"
fractions.

Ichthyoplankton abundances are expressed as
numbers per 10 m^ sea surface. Flow meters were
not used with the vertical Hensen net hauls;
numbers of larvae caught in each tow are multi-

plied by 30 to provide numbers per 10 m^ esti-
mates. A flow meter used with WP2 nets provided
water volume measurements and more accurate
abundance estimates. Based on these measure-
ments the conversion factor applied to Hensen net
catches appears to be reasonable: water volumes
filtered by 100-0 m vertical WP2 net hauls aver-

76*

74°

"I T"

COQUIMBO

VALPARAISO

FALC AHUANO

VALDIVIA

PTO MONTT

_ai aJl_

24'

26 °

28 °

30°

34'

36'

38'

- 40'

42<

44'

FISHKKY lUlI.I.KTIN: VOl. Hfi. NO 1

72° 70°

I 8 °

- 20 °

22 o

-24

-28

- 30

- 32

- 34

- 36

-38

- 40'

- 42

Figure 2. — Spawning areas of (A) anchoveta iEngraiilis ringens) and (Bi sardine tSardinops sagax) off Chile based on egg

abundances during July-September ichthyoplankton surveys, 1964-73.

LOEB and ROJAS: ICHTHYOPLANKTON COiMPOSITION AND ABUNDANCE

Table 1. — Cruises yielding samples used for examination of ichthyoplankton abundance and composition variations
oft northern Chile Only samples from 1 8 -24 S are used for interannual comparisons. Data from two 1 968 cruises are
combined; data from cruise 71(4)69CD are used for analysis of sampling depth-related catch differ-
ences. N = number of samples used in ichthyoplankton analyses Tow types: H = Hensen net; WP2 = WP2 net;
V = vertical.

Year

Cruise

Dates

Location

Tow

Type

Depth
(m)

1964

06(3)64GE

16 08-23 09

18 20',

, 23 38'S

70 ir.

, 71 50'W

0-50

1965

13(3)65CD

15 08-09 09

18 20'

, 23 50'S

70 00',

, 72 08'W

0-50

1966

25{3)66CD

21 08-31 09

18 28'

, 23 52'S

7016'

, 72 16'W

0-50

1967

37(3)67CD

17 08-1009

18 25',

, 23 45'S

70 05'

, 71 38'W

0-50

1968

47(3)68NO

25 08

18 28'

, 23 OO'S

70 05'

, 70 58'W

0-50

49(3)68NO

29 09

18=26'

, 23 01'S

70 11'

, 70 59'W

0-50

1969

70(3)69NO

23 08-25 08

18^29'

, 23 01'S

70 09'

, 71 10'W

0-50

71(4)69CD
71(4)69CD

03 12-17 12
03 12-17 12

28 29'
28 29'

, 38 OO'S
, 38 OO'S

71 22'
71 22'

, 73 55'W
, 73 55'W

43
39

H
H

V
V

0-100
0-50

1970

86(3)70NO

25 09-26 09

19°27'

, 21 58'S

70 14'

, 71 03'W

0-50

1972

109(3)72NO

04 09-15 09

18 '29'

, 22^58 'S

70 '10'

, 71 25'W

0-50

1973

130(3)73CP

28 07-08 08

18 17'

, 23 OS'S

7005'

, 72 20'W

0-100

1983

277(3)83CP

07 08-15 09

18°33'

. 23 48'S

70 09'

, 71 38'W

WP2

0-100

aged 28.3 m'^ yielding a raw count to numbers per
10 m'- conversion factor of 35; this 179f increase in
conversion factor is associated with a 12*7^ de-
crease in mouth opening of WP2 vs. Hensen nets.
Larval fish diversity is expressed as total num-
bers of taxa per sampling period and mean num-
bers of taxa per tow.

Various parametric and nonparametric tests
were used for statistical analyses. Differences in
mean abundances are tested with 2-tailed Z tests
and Mann Whitney U tests (Dixon and Massey
1969; Conover 19711. Similarity of abundance
ranks within data sets are tested with Kendall's
concordance iW) test (Tate and Clelland 1969)
and Spearman's rho (p) correlation test (Conover
1971). Significant values resulting from these
tests are indicated, but due to multiple testing
these values should be used only as indicators of
the relative strengths of relationships. Percent
similarity indices (PSI; Whittaker 1975) are used
for comparisons of species percentage composi-
tion. Because PSIs are strongly influenced by
abundant species, we apply these tests to the OL
fraction as well as to total larvae. We define as
"high" all PSI values >80, as "moderate" PSI's
>65 and <80, and as "low" values <65.

SAMPLING CONSIDERATIONS

Sampling Depth Differences

The 100 m sampling depths in 1973 and 1983
potentially effect direct comparisons of abun-
dance estimates and species composition in these
vs. earlier data sets owing to individual species'
depth distributions. Evaluation of depth-related
sampling differences is possible through a com-
parison of data obtained from coincidental 50-0
m and 100-0 m Hensen net samples taken during
December 1969 at 28°-38"S (Table 1). These data
indicate that abundance estimates and percent-
age contribution of the PL to total larvae were
higher in 50 m tows while those of the OL were
higher in 100 m tows (Table 2). Four of five PL
species caught were more abundant and frequent
in the 50 m samples. The greatest differences
were for Trachurus murphyi, which was 5x more
abundant and 3.6 x more frequent, and Merluc-
cius gayi , which was present only in 50 m sam-
ples. Abundance estimates were higher for 11 of
16 OL taxa in 100 m tows although, with a few
exceptions, catch frequencies were similar. The
greatest difference was for Triphoturus mexi-

FISHERY BULLETIN: VOL 86, NO. 1

Table 2. — Catch comparisons of 50-0 m and 100-0 m Hensen net tows taken at lat. 28 -38 S, long. 71 -74'W during
3-17 December 1969. Abundances expressed as mean and standard deviations of numbers per 10 m2, F = percent
frequency of occurrence in samples Depth-related catch differences are shown as ratios of 50:100 m abundance
estimates, species percentage contribution to total identified larvae, PL and OL percentage contribution to total larvae, and
taxonomic diversity (mean numbers per tow and total numbers of taxa). N = number of samples. PL = larvae of pelagic
species; OL = other larval taxa.

50-

-0 m

100-

■0 m

= 43)

(/V =

39)

Taxon

(S)

Ratio

Engraulis ringens

34.2

(122.8)

23.2

13.42

22.3

69.4)

15.4

7.23

1.53

Clupea bentincki

6.3

( 28 2)

9.3

2.47

1.5

6.7)

5.1

0.49

4.20

Ethmidium maculatum

0.7

( 4.6)

2.3

0.27

0.8

4.8)

2.6

0.26

0.88

Merluccius gayi

1.4

( 6.4)

0.55

Trachurus murphyi

4.2

( 14.0)

9.3

1.65

0.8

4.8)

2.6

0.26

5.25

Total PL

46.8

(131.1)

34.9

1836

25.4

71.0)

23 1

8.24

1.84

Bathylagus nlgngenys

0.7

( 4.6)

2.3

0.27

0.8

4.8)

2.6

0.26

0.88

Vinciguerria lucetia

10.5

( 33.9)

14.0

4.12

16.9

75.6)

7.7

5.48

0.62

Diogenlchthys spp.

31.4

(107.0)

20.9

12.32

37.7

79.9)

28.2

12.22

0.83

Hygophum bruuni

121.4

(238.2)

58.1

47.64

166.2

291.0)

53.8

53.86

0.73

Protomyctophum sp.

3.1

9.2)

10.3

1.00

Diaphus sp.

( 18.6)

18.6

3.02

9.2

24.0)

17.9

2.98

0.84

Lampanyctus parvicauda

7.7

( 18.6)

163

3.02

6.9

21.2)

15.4

2.24

1.12

Lampanyctus sp.

1.4

{ 6.4)

4.6

0.55

1.5

6.7)

5.1

0.49

0.93

Triphoturus mexicanus

7.7

( 21.8)

14.0

3.02

18.5

57.4)

20.5

5.99

0.42

Scopelosaurus sp.

0.8

4.8)

2.6

0.26

Normanichthys crockeri

15.3

( 38.9)

18.6

6.00

20.0

66.5)

10.3

6.48

0.76

Sebastes sp.

1.4

( 9.1)

2.3

0.55

0.8

4.8)

2.6

0.26

1.75

Blennild A

0.7

( 4.6)

2.3

0.27

Blenniid D

0.7

( 4.6)

2.3

0.27

0.8

4.8)

2.6

0.26

0.88

Bothid

0.7

( 4.6)

2.3

0.27

Unid. 2

0.7

( 4.6)

2.3

0.27

Small damaged myctophids

4.8

9.2

Other unidentified

13.2

7.0

Total OL

226.0

(316.0)

86.0

81.59

299.4

405.9)

82.0

91.78

0.84

Total larvae

272.8

(340.7)

324.8

400.0)

0.84

No. taxa/tow

2.5

( 2.0)

2.2

2.0)

1.13

Total no. taxa

0.94

canus which was 2.4 x more abundant and 1.5 x
more frequent in 100 m tows. Because of the large
catch variability, none of the species abundance
differences nor the abundance differences of the
PL, OL, and total larvae are significant (Z tests,
P's all >0.10). Additionally, species abundances
within all positive tows from the two sampling
depths are not significantly different (Mann
Whitney U tests, P > 0.10 in all cases).

The overall species composition of 50 and 100 m
tows was similar. Despite greater proportions of
PL in 50 m tows, the PSI value from comparisons
of total species lists was high (87.7). Species per-
centage contribution within the OL fraction of the
two tow types was also quite similar (94.5). Spe-
cies abundance rankings within the two total lar-
val data sets are significantly correlated
(p = +0.80; P < 0.01 ). Species diversity estimates
(total numbers of taxa and mean numbers of taxa
per tow) are also similar.

From these comparisons it is apparent that the

PL predominantly occur within the upper 50 m.
Similar shallow (e.g., <50 m ) distributions have
been described for dominant PL species off of Peru
(anchoveta, sardine, and hake; Sameoto 1982).
The generally lower 100 m abundance estimates
of these species is puzzling, but suggests possi-
bly shorter sampling time and/or less efficient
sampling within the upper 50 m of these tows.
Higher catch frequency and abundance of T.
mexicanus in 100 m tows suggest that large pro-
portions (e.g., 30-60%) of these larvae are at 50-
100 m.

As a result of these catch differences we suggest
caution in making direct numerical comparisons
between the 1973 and 1983 vs. earlier data sets.
Although the overall compositions and abun-
dance relations should not be markedly altered,
some accommodation should be allowed for the
percentage contributions and across-year abun-
dance ranks of PL species (especially Trachurus)
and Triphoturus mexicanus.

LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE

Net Type Differences

There are no data available for a direct evalua-
tion of catch differences between the vertical
Hensen and WP2 net tows. However, comparisons
of data from the 1983 WP2 net hauls (Table 3) do
not indicate that this net is more or less efficient
than the Hensen net. Mean abundances and spe-
cies diversity (total numbers of taxa and mean
numbers of taxa per tow) are within the ranges of
values from Hensen net tows.

Day-Night Catchi Considerations

Day and night sample data have been combined
for each cruise. Day samples (0600 -1800 hours)
outnumbered (generally 55-64%) night samples
during all cruises. In nine cases there were non-
significant day-night larval catch differences (Z
tests, P >0.05) and overall similarity in night:
day catch ratios (0.74-1.8:1; mode = 1.2:1). One
cruise had a significantly higher night vs. day
catch (2.5 X; P < 0.01); this cruise and one with

Table 3. — Abundance estimates and diversity of larval fishes and abundance estimates of zooplankton collected in July-September
samples off northern Chile (18 -24-S), 1964-83^ Larval fish abundances as mean numbers per 10 m^. PL = larvae of commercially
important pelagic species; OL = other larval taxa (myctophid, other mesopelagic, and other categories). Number of larvae is total raw count
of identified larvae for each sampling period. Larval fish diversity expressed as total number of identified taxa and mean number of taxa
per tow. Zooplankton abundance is mean displacement volume (cc/10 m2). N.A. = data not available.

Mean abundance for

Taxon

1964

1965

1966

1967

1968

1969

1970

1972

1973

1983

Engraulis ringens

302.0

36.7

3,478.3

224.7

72.3

181.7

620.0

1,634.1

1,816.4

—

Sardinops sagax

5.3

5.1

12.5

4.6

21.1

24.9

1.4

327.2

52.1

594.9

Ethmidium maculatum

—

2.6

—

0.7

—

Trachurus murphyi

2.6

7.1

10.8

3.1

0.4

6.8

5.7

100.3

2.1

—

Scomber japonicus

—

0.4

—

25.9

—

Merlucaus gayi

—

0.7

2.6

Total PL

309.9

49.3

3,501.6

232.4

93.8

216.0

627.1

2,088.2

1,871.3

597.5

Bathylagus nigrigenys

4.0

5.1

6.7

9.7

6.2

0.9

7.2

56.6

35.0

31.5

Vinciguema lucetia

3.1

0.4

2.9

6.1

3.1

2.6

21.4

54.8

18.6

3.7

Stemoptyx diaphana

—

0.7

1.0

Chaultodus sp.

—

0.3

—

Stomias spp.

—

1.9

Lestidiops pacificum

—

0.3

0.7

—

Melamphaes sp.

—

0.5

0.4

—

0.3

2.1

—

Beryciform

—

1.4

—

Other mesopelagics

7.1

5.5

9.6

16.3

9.7

3.5

30.0

113.0

57.1

38.1

Diogenichthys spp.

6.2

18.9

14.6

30.5

7.3

14.6

21.4

59.3

39.3

114.2

Hygophum bruuni

—

2.8

1.2

3.6

—

0.9

Hygophum atratum

—

6.7

Metelectrona ventralis

—

■ —

0.9

Myctophum nitidulum

0.4

—

0.4

—

0.4

0.9

—

4.8

4.3

4.6

Diaphus sp.

—

0.4

—

1.4

0.3

—

4.5

Lampanyctus parvicauda

41.0

32.0

65.8

63.0

10.8

38.6

5.7

29.7

12.1

15.5

Lampanyctus spp.

—

0.8

—

1.1

—

0.7

—

Triphoturus mexicanus

31.8

15.8

20.8

29.0

4.6

31.7

11.4

19.7

23.6

33.6

fvlyctophids

79.4

69.9

104.0

126.1

24.2

85.8

39.9

113.8

80.0

180.9

Normanichtys crockeri

17.2

1.5

281.2

5.6

2.3

26.6

10.0

2.4

31.4

—

Sebastes sp.

0.4

—

2.5

1.0

0.8

—

0.7

—

Gadiform D

—

0.8

—

Macrourid A

—

0.5

—

0.7

—

Macrourid C

—

0.5

—

Blenniid A

0.4

—

0.8

—

1.2

—

1.4

—

Blenniid B

—

0.4

—

Blenniid C

—

1.5

—

0.3

—

Blenniid D

5.3

0.4

16.7

7.1

4.2

2.6

—

2.8

8.6

1.0

Gobiesocid A

0.4

—

0.8

—

1.0

0.7

—

Gobiesocid B

—

0.4

—

Unid. 1

—

0.8

—

5.7

8.6

Unid. 2

—

0.4

—

0.3

12.1

—

Unid. 3

—

0.4

1.7

—

0.3

0.7

—

Unid 4

—

0.3

0.7

—

Ophidiid

—

1.0

0.8

—

Hippoglossina sp.

—

0.3

—

Other larvae

23.7

1.9

303.6

17.2

10.9

30.9

10.0

7.7

62.7

9.6

Total OL

110.2

77.3

417.2

159.6

44.8

120.2

79.9

234.5

199.8

228.4

FISHERY BULLETIN: VOL. 86, NO 1

Table 3. — Continued.

Mean abundance for

Taxon

1964

1965

1966

1967

1968

1969

1970

1972

1973

1983

Larval fish abundance and di

iversity

Total ID PL

309.9

493

3,501 6

2324

938

216.0

627 1

2,0882

1,871 3

597.5

Total ID OL

110.2

77,3

417.2

159 6

448

120.2

799

234 5

1998

2286

Total ID larvae

420.1

126.6

3,9188

3920

138.6

336.2

707.0

2,322.7

2,071.1

826 1

Unid missing OL

10.2

4.4

170

220

3.4

11.9

7.2

104.5

57.2

76.6

Total larvae

430.3

131.0

3.9358

4140

142.0

348.1

714.2

2,427.2

2,1283

9027

Number of larvae

953

321

9,406

771

360

650

495

6,738

2,900

893

Number of taxa

Number of taxa tow

1.4

1.9

5.3

4 1

3.7

Number of samples

Zooplankton abundance

333.9

2422

406.5

370.8

279.2

141.0

N.A

301 7

100.6

168.7

Number of samples

124

110

a 1.8:1 night:day catch ratio were represented
by fairly equal day {55-5T/( ) and night sam-
ples.

TAXONOMIC PROBLEMS

The 576 samples used for interannual compari-
sons yielded a total of 41 taxa including 19 spe-
cies, 7 genera, and 11 higher taxa (Table 4). The
PL and most mesopelagic forms were identified to
species. During several cruises there were large
proportions of small Diogenichthys spp. (Myc-
tophidae) larvae which could not be identified to
species. As a consequence, data on the two spe-
cies, D. atlanticus and D. laternatiis, were lumped
to permit reasonable between-year taxonomic
composition comparisons. In all but one cruise, D.
laternatiis dominated (77-1007f ) the identifiable
Diogenichthys larvae. Total within-year Dio-
genichthys spp. abundances were multiplied by
proportions of identified D. laternatus and D. at-
lanticus larvae to provide between-year abun-
dance rankings for each species.

Species identifications of coastal forms are lim-
ited by inadequate taxonomic information and by
the presence of generally early larval develop-
mental stages in samples. These larvae are pri-
marily classified at familial and ordinal levels.
Because the classifications include few multispe-
cies groupings and those were numerically rare
the taxonomic limitations offer no severe analyti-
cal problems.

Largest taxonomic problems occurred in
cruises when large numbers of small unidentifi-
able larvae were caught (e.g., 1972 and 1983;
Table 3). Additionally, most cruises had "miss-
ing" larvae (e.g., "other larvae" enumerated when
the samples were first processed but not ac-
counted for during later species identification

work). With the exception of 1983 the uniden-
tified and missing larvae made up <5% of the
total larval abundance for each sampling pe-
riod.

RESULTS

Overall Ichthyoplankton Composition

The 576 July-September samples used for in-
terannual comparisons yielded a total of 23,487
identified larvae. These larvae were dominated
(85.17r) by PL species (Table 4). Overall domi-
nants were anchoveta (Engraulis ringens; 74.37c)
and sardine (Sardinops sagax; 9.3%). The other
PL species were relatively rare: Trachurus mur-
phyi contributed 1.2% and Scomber japonicus,
Merluccius gayi , and Ethmidium maculatum to-
gether formed 0.3% of the total. The larval abun-
dances of these species off of Chile are strongly
influenced by sampling time and location. Mer-
luccius gayi occurs primarily to the south of the
study area (24°-43°S) and Scomber japonicus and
Trachurus murphyi have later summer (Novem-
ber-February) spawning peaks.

The OL were dominated by mesopelagic fishes
(18 taxa, 10.6% of total larvae). Myctophids were
most abundant (8.0%) primarily because of the
large numbers of Diogenichthys spp., Lampanyc-
tus parvicouda , and Triphoturus mexicanus,
which together made up 7.7% of the total. One
bathylagid (Bathylagus nigrigenys) and one
gonostomatid (Vinciguerria lucetia ) were also rel-
atively abundant (together 2.5%). Coastal fish
larvae (14 taxa) made up 4.2% of the total; a
scorpaeniform (Normanichthys crockeri; 3.4%)
and blenniid (Blenniid D; 0.4%) dominated this
group.

Eighteen taxa were relatively frequent (e.g., in

LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE

>50% of cruises) and/or abundant across the 10
sampling periods (Table 4). These taxa (four PL
species, seven myctophid taxa, two other meso-
pelagic species, and five coastal forms) made up
99.69; of the identified larvae; they also con-
tributed 97-1009r of the identified larvae (91.2-
99.6% of total larvae) and included the top 9-12
ranked taxa within each sampling period (Table
5).

Table 4— Ichthyoplankton species collected in July-September
samples oft northiern Chile (18 -24 S, 70 -72 W), 1964-83. Rela-
tive abundance (ROA) and percentage composition based on
summed cruise mean abundances (no. 10 m2) of all identified
forms. Frequency is number of total 10 sampling periods when
taxon was caught. Categories are PL (commercially important
pelagic species) and other taxonomic components (M = myc-
tophids: OM = other mesopelagic taxa; C = coastal forms).

Fre-

Cate-

Taxon

ROA

quency

gory

Engraulis ringens

74.29

Sardinops sagax

9.33

Normanichthys crockerl

3.36

Diogenichthys latema-

tus + D. atlanticus

2.90

Lampanyctus parvicauda

2.79

Tnphoturus mexicanus

1.97

Bathylagus nigngenys

1.44

Trachurus murphyi

1.23

Vinciguerna lucetia

1.04

Blenniid D

0.43

Scomber japonicus

023

Myctophum nttidulum

0 14

Unid. 1

0.13

Unid. 2

0.11

Hygophum bruuni

0.075

Diaphus sp.

0.062

Hygophum atratum

0.060

Sebastes sp.

0050

Blenniid A

0.034

Ethmidium macu latum

20.5

0030

Merluccius gayi

20.5

0.030

Unid. 3

0028

Melamphaes sp

235

0.026

Gobiesocid A

23.5

0026

Lampanyctus spp.

0024

Sternoptyx diaphana

0,020

Stomias spp.

0.017

Blenniid C

285

0016

Ophidiid

285

0016

Beryciform

0.012

Macrourid A

0011

Lestidiops pacificum

32.5

0.009

Unid. 4

32.5

0.009

Metelectrona ventralis

0.008

Gadiform D

0.007

Macround C

0,004

Blenniid B

37.5

0.0036

Gobiesocid B

37.5

0.0036

Chaultodus sp.

39.5

0.0030

Htppoglossina sp.

39.5

0.0030

Total number of identified larvae:

23,487

Total number of samples:

576

Total number of taxa:

INTERANNUAL VARIATIONS IN
ABUNDANCE AND COMPOSITION

The ichthyoplankton demonstrated extreme in-
terannual variations in abundance and composi-
tion (Tables 3, 5; Fig. 3A, B). Most obvious are the
1) total larval and PL abundance peaks of 1966,
1972, and 1973 and 2) shift from anchoveta to
sardine dominance between 1973 and 1983. The
maximum total larval abundance values in 1966,

1972, and 1973 were 2.4-30 x higher than those
of other years; the PL had 3-71 x higher abun-
dances during these vs. other years. Interannual
abundance variations during the seven years of
lower abundance were also large for total larvae
(to 6.9 X ) and the PL (to 12.7 x ). Anchoveta domi-
nated the ichthyoplankton through 1973 (29.0-
88.8% ) and was directly responsible for the ex-
treme abundance variations; anchoveta were
absent from the 1983 samples, and sardine larvae
(72.0%) contributed to the moderately high total
larval and PL abundances (Table 5).

The OL fraction had less extreme abundance
variations than the PL: maximum 1966, 1972,

1973, and 1983 mean abundance values ranged
from 1.4 to 9x those of the other years; mean
values within the six years of lower abundance
varied to 3.8 x. Unlike the PL, OL abundance
fluctuations were not attributed to any one taxo-
nomic component (Fig. 3B): the 1966 peak was
largely due to coastal taxa (73% of OL); myc-
tophids and other mesopelagic taxa equally domi-
nated the 1972 peak (48%); myctophids domi-
nated (79%) in 1983; and all three components
were relatively abundant (29-40%) in 1973. Myc-
tophids were generally the dominant component
(50-90% ) during the six years of lower OL abun-
dance. Abundance fluctuations (as range of mean
abundance values) across the 10 years were less
extreme for myctophids (8.1 x ) than for the other
mesopelagic taxa (32.3 x) and coastal forms
(160x).

In contrast to the ichthyoplankton, July-Sep-
tember zooplankton biomass values were rela-
tively constant between years and exhibited only
a 4x range in values (Table 3).

Despite large between-year variability in rela-
tive proportions of the PL and OL, there is a sig-
nificant agreement of their ranked mean abun-
dances across the 10 years (p = +0.81, P < 0.01).
There is also a general agreement of the ranked
mean abundances of PL and the three OL compo-
nents across the 10 years (W = 0.44; P = 0.05).
These categories were generally more abundant

FISHERY BULLETIN: VOL. 86, NO 1

Table 5. — Comparisons of relative abundances of dominant larval fish taxa collected off of northern Chile (18-
24°S, 70°-72 W) during July-September sampling periods, 1964-83. Relative abundances within each year are
presented as (A) percentage contribution to total identified larvae and (B) ranked abundance. Taxa are listed in
order of total summed 10-yr mean abundances

Taxon

1964

1965

1966

1967

1968

1969

1970

1972

1973

1983

A. Percentage contribution to total identified

larvae

Engraulis nngens

71.89

28.99

88.76

57.32

52 20

54.05

87.69

70 35

87.70

—

Sardinops sagax

1.26

4.03

0.32

1.17

15 23

7.41

0.20

14.09

2.52

72.01

Normanichthys crocken

409

1.18

718

1.43

1.66

7.91

1.41

0.10

1.52

—

Diogenichthys spp.

1.48

14.93

0.37

7.78

5.27

4.34

303

2.55

1.90

13.82

Lampanyctus pan/icauda

9.76

25.28

1.68

16.07

7.80

11.48

0.81

1.28

058

1 88

Tnphoturus mexicanus

7.57

12.48

0.53

7.40

332

9.43

1 61

085

1.14

4.07

Bathylagus nigrigenys

095

4.03

0.17

2.47

4.48

027

1.02

244

1.69

381

Trachurus murphyi

0.62

5.61

0.28

0.79

0.29

2.02

0.81

4.32

0.10

—

Vinciguerna lucetia

0.74

0.32

0.07

1.56

2.24

0.77

3.03

2.36

0.90

0.45

Blenniid D

1.26

0.32

0.43

1.81

3.03

0.77

—

0.12

0.42

012

Scomber japonicus

—

0.32

—

1.12

—

Myctophum nitidulum

0.10

—

0.01

—

0.29

0.27

—

0.21

0.56

Unid. 1

—

0.56

—

0.27

0.96

Unid. 2

—

0.01

—

• —

—

0.01

0.57

—

Hygophum bruuni

—

2.21

0.03

0.92

—

0.11

Diaphus sp.

—

0.32

0.01

—

0.20

0.01

—

0.54

Hygophum atratum

—

0.81

Sebastes sp.

0.10

—

0.06

0.26

0.58

—

0.03

—

Other taxa

0.19

—

0.09

1.02

3.03

1.28

0.20

0.19

0.43

0.78

B. Ranked within-year abundance

Engraulis nngens

—

Sardinops sagax

6.5

Normanichthys crocken

—

Diogenichthys spp.

2.5

Lampanyctus parvicauda

7.5

8.5

Triphoturus mexicanus

Bathylagus nigrigenys

6.5

12.5

Trachurus murphyi

7.5

13.5

—

Vinciguerria lucetia

11.5

2.5

Blenniid D

6.5

11.5

—

13.5

Scomber japonicus

—

11.5

—

Myctophum nitidulum

12.5

—

18.5

—

12.5

—

Unid. 1

—

12.5

—

Unid. 2

—

18.5

—

■ —

—

20.5

8.5

—

Hygophum bruuni

—

15.5

Diaphus sp.

—

11.5

18.5

—

20.5

—

Hygophum atratum

—

Sebastes sp.

12.5

—

13.5

—

in 1966, 1972, and 1973 and relatively rare in
1965 and 1968. The ranked abundance patterns of
each of the components differ from one another
(e.g., all pairwise correlation coefficients
[p = —0.21 to +0.61] are nonsignificant). Larval
diversity is strongly correlated with total larval
abundance (p = +0.88, P < 0.01).

There are no significant correlations between
abundance ranks of invertebrate zooplankton
biomass and total larvae (p = +0.27) or any of the
larval categories (p = -0.03 to +0.22; P > 0.05 in
all cases).

Species Abundance Variations
and Relations

The top 10 ranking larval fish taxa were caught

during at least 9 of the 10 sampling periods (Table
4). All of these taxa exhibited large interannual
abundance fluctuations (Table 3). Most marked
were the abundance changes of anchoveta, sar-
dine, and coastal species Normanichthys crockeri.
This latter species (rank 3 in overall abundance)
was frequently abundant prior to 1983; like an-
choveta it was absent from 1983 samples. Among
the 10 taxa only Triphoturus mexicanus had
<10x changes in mean abundance values; <20x
changes occurred for Lampanyctus parvicauda
(11.5X) and Diogenichthys spp. (18.4x); all other
taxa had >20x mean abundance changes over
the 10 years. The abundance fluctuations of these
10 taxa are primarily responsible for the interan-
nual abundance and composition variations
(Table 5; Fig. 3A, B).

LOEB and ROJAS: ICHTHYOPLANKTON COMI'OSITION AND AHUNDANCE

A. TOTAL ICHTHYOPLANKTON

E
O

t :1 Anchoveta
Other PL
I I Sordine
■ OL

OL CATEGORIES

II Other mesopeiogic toxo
M yctophids
Coastal toxa

*j>\ij

Figure 3. — Mean abundance (numbers per 10 m2) of (A) total ichthyoplankton and major PL components and
(Bl total OL and major OL components collected off northern Chile (18'-24°S) during July-September sampling
periods, 1964-83.

Between-year comparisons of the species per-
centage compositions of total larvae give a wide
range of PSI values (3.3-95.0; Table 6A) which
primarily reflect similarity in percentages of an-
choveta; 51% of these values are moderate to high
(e.g., >65). Highest values (91.5-95.0) come from
comparisons between 1966, 1970, and 1973 when
anchoveta contributed >87% of the larvae. High

values (80.7-82.5) also result from comparisons
between 1964, 1967, and 1969 and result from
moderate anchoveta abundance (54.0-71.9%) and
relatively similar proportions of other taxa. Low-
est values (3.3-31.0) result from comparisons of
1983 vs. all other years and reflect the absence of
anchoveta larvae in 1983 samples. The 1973 and
1983 PSI values are little affected (e.g., <2.3) by

FISHERY BULLETIN: VOL. 86, NO 1

Table 6. — Between-year percent similarity index (PS!) values from compari-
sons of (A) total larvae and (B) \he OL (ottier larval taxa) fraction collected
during July-September sampling periods off northern Cfiile (18-24 S), 1964-
83.

PS! values for

1965

1966

1967

1968

1969

1970

1972

1973

1983

A. Total larvae

1964

52.44

79.94

8222

71.25

80.70

79.70

77.89

80.30

10.30

1965

33.95

67.41

55.54

62.38

38.16

45.32

38.74

28.47

1966

62.69

57.83

65.08

91.55

73.65

91 88

3.32

1967

77.96

8247

67.75

68.21

66.82

18.05

1968

79.39

62.81

76.39

63.68

30.95

1969

62.95

69.67

63.50

18.80

1970

79.06

94.95

7.31

1972

79 42

22.02

1973

8.95

B. OL fraction

1964

69.88

47.01

75.76

57.56

85.16

45.99

36.08

51.29

33.15

1965

29 22

88.07

60 36

68.42

55.43

54.30

47.13

54.39

1966

34.96

3678

50.09

30 53

26 15

37 50

18.48

1967

70.21

71.01

53.92

52.55

55.68

49.12

1968

58.46

54.85

61.57

67 92

51.92

1969

48.98

39.22

51 92

37.18

1970

74.35

68.40

60.21

1972

66 23

58.74

1973

58.61

accommodations for possible sampling depth-
catch differences of Trachurus gayi and Triphotu-
rus mexicanus.

When the PL are excluded, comparisons be-
tween the OL taxa yield generally lower PSI val-
ues than those of the total larvae; only 2¥7( of the
18.5-88.1 values are moderate to high (Table 6B).
Moderate to high values (69.9-88.1) come from
comparisons between 1964, 1965, 1967, and 1969,
and in part result from similar proportions of
Lampanyctus parvicauda (32.1-41.4%) and Tri-
photurus mexicanus (18.2—28.99^) during those
years. Moderate values (66.2-74.4) also come
from comparisons of 1967 vs. 1968 and 1969 (sim-
ilar proportions of L. parvicauda, T. mexicanus,
and Diogenichthys [12.2-39.5%!); 1970 vs. 1972
(similar proportions of Vinciguerria lucetia and
Diogenichthys [23.4-26.8%!); and 1972 vs. 1973
(similar proportions of Bathylagus nigrigenys, V.
lucetia, and Diogenichthys [9.3-25.3%!). Lowest
PSI values (<30) result from comparisons of 1966
vs. 1965, 1970, 1972, and 1983, and are due
largely to extreme dominance by Normanichthys
crockeri (67.4% of OL) in 1966. Recalculations to
accommodate for possible depth-related increased
catches of T. mexicanus in most cases decrease
1973 and 1983 PSI values (e.g., by 2.6-7.2) and in
two cases (1968 and 1970 vs. 1973) change the
value characterization from moderate (67.9 and
68.4) to low (63.8 and 64.5). With one exception

(1972 vs. 1973, PSI = 67.0)
justed values are low.

all of the other ad-

Species Across-Year Ranked
Abundance Patterns

Individual species across-year abundance rank-
ings demonstrate a variety of patterns. Three pat-
terns are shared by nine of the more frequently
occurring taxa (Table 7). These involve 1) a
group formed by anchoveta and three coastal
forms; 2) a group formed by one myctophid and
two other mesopelagic species; and 3) a species
pair consisting of sardine and a myctophid. An-
other species pair (two myctophids) can be formed
if the 1973 and 1983 abundances of T. mexicanus
are adjusted.

Group I includes anchoveta, Normanichthys
crockeri, Blenniid D, and Sebastes sp. (Table 7).
There is a significant concordance among these
species as to years of highest (1966 and 1973) and
lowest (1965 and 1983) abundance (W = 0.69,
P < 0.01). The abundance rankings of anchoveta
and A'^. crockeri (p = +0.79) and of Blenniid D and
Sebastes (p = +0.88) are significantly correlated
(P < 0.01). None of the correlations between spe-
cies of the two pairs are significant due to differ-
ences in 1967-68 vs. 1970-72 relative abun-
dances.

The three Group II species, Diogenichthys later-

LOEB and RO.I.XS ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE

notus. Bathylagus nigrigenya , and Vinciguerna
lucetia, have a concordance of higher abundances
in 1972, 1973, and 1983 vs. other years (W = 0.81,
P<0.01). The abundance rankings of B. ni-
grigenys are strongly correlated with those of D.
laternatus and V. lucetia (p = +0.82, P < 0.01 in
both cases); the correlation between D. laternatus
and V . lucetia is not significant.

The abundance rankings of sardine and Mycto-
phum nitidulum (Pair I species) are significantly
correlated (p = +0.96, P<0.01); both species
were rare or absent in 1970 and most abundant in
1972, 1973, and 1983.

With adjustment to 1973 and 1983 abundances
of T. mexicanus, its abundance rankings are
strongly correlated with those of Lampanyctus
parvicauda (Pair II species; p = +0.84, P < 0.01).
Highest relative abundances of both species were
in 1964, 1966, 1967, and 1969.

Three relatively frequent species iTrachurus
murphyi, Diogenichthys atlanticus , and Diaphus

sp.) do not conform to any of the above patterns.
If the 1973 Trachurus abundance is adjusted to
accommodate for possible undersampling, its
abundance pattern is similar to that of anchoveta
(p = +0.68, P = 0.05) but not to any of the other
Group I species (p = +0.03 to +0.46).

VARIATIONS IN ABUNDANCE AND

COMPOSITION RELATIVE TO

HYDROGRAPHIC CONDITIONS

Ichthyoplankton abundances in the 10 years
sampled show no consistent patterns relative to
warm water-cold water events (Table 8). High PL
and OL abundances occurred during strong El
Nino events (1972 and 1983) and during cold or
transition years immediately following El Nihos
(1966 and 1973). Lowest abundances of both frac-
tions were associated with the 1965 El Nino and
warm 1968. Neither the PL nor OL have signifi-
cant correlations with ranked (high to low) July-

Table 7. — Larval fish taxa grouped according to similar across-year (1964-83) ranked abundance patterns.
Significant agreement of group rankings indicated by Kendall's concordance {W) values. Correlations between
species pair rankings indicated by Spearman's rhio (p) values. Significant values at P £ 0.05 are indicated, but
note use of multiple testing. " = abundance ranks adjusted to accommodate for possible sampling deptti
related catch differences (Table 2).

Abundance rank foi

Species

1964

196£

. 1966

1967

1968

1969

1970

1972

1973

1983

Group 1

Engraulis nngens
Normanichthys crocken
Blenniid D

5
4
4

9
9
9

1
1
1

6
6
3

8
8
5

7
3

7
6

2
2
2

Sebastes sp.

E nngens-N- crocken
Blenniid D-Sebastes sp,
E nngens- Blenniid D
E- nngens-Sebastes sp.
N. croc/(en -Blenniid D

W =
P =
P =
P =
P =
P =

0.69 (P
+ 0.79 (P
+ 0.88 (P
+0.59
+0.44
+ 0.61

o o o

V II V

N. crocken-Sebastes sp.

P =

+ 0.48

Group II

Diogenichthys laternatus
Bathylagus nigngenys
Vinciguerna lucetia

10
9
6.5

7.5

4
4
4

9
7
6.5

7.5
10
9

6
5
2

1
1

3
2
3

1
3
5

B nigngenys-D. laternatus
B nigrigenys-V. lucetia
D. laternatus-V lucetia

W =
P =
P =
P =

0.81 (P
+ 0.82 (P
+ 0.82 (P
+ 0.52

<0.01)
<0.01)
< 0.01)

Pair 1

Sardinops sagax
Myctophum nitidulum

7
6

8
9

6
6

9
9

5
6

4
4

10
9

3
3

1
2

S sagax-M. nitidulum

P =

+ 0.96 (P<0.01)

Pair II

Lanpanyctus parvicauda
Tnphoturus mexicanus ' '

5
6

1
4

9
10

4
2

10
8

6
5

8
9

7
7

L. parvicauda-T. mexicanus"

P =

+ 0.84 (P

<0.01)

FISHERY BULLETIN VOL H6, NO 1

Table 8. — Range, mean, standard deviation and ranked (high to low) values of sea
ichthyoplankton sampling periods, 1964-83. N = number of observations. N.A. ^ data not
(1976), Bernal et al. (1982), and Kelly and Blanco (1983).

1964

1965

1966

1967

1968

Temperature (°C)
N

128

Range

13.5-17.9

13.6-18.1

13.7-17.1

13.3-16.0

14.3-18.6

15.7

16.5

15.3

14.8

16.5

(S)

(1.0)

(0.8)

(0.9)

Rank

3.5

Salinity (%o)
N

124

Range

34.51-34.99

34.80-35.40

34.53-35.03

34.74-34.82

34.69-35.23

34.74

35.06

34.83

34.78

34.92

(S)

(0.12)

(0.11)

(0.04)

(0.12)

Rank

Hydrographic
condition:

Cold

El Nino

Transition

Cold

Warm

September mean temperature and salinity val-
ues; larval diversity (mean number of taxa/tow)
also shows no correlation with these values
(Table 9). Within the PL, anchoveta were most
abundant during years immediately following El
Nihos (cold 1966, transition 1973), the 1972 El
Nirio, and cold 1970; lowest abundances were dur-
ing the 1965 and 1983 El Nihos (Table 8). There
are no significant correlations between ranked
anchoveta abundances and ranked values of tem-
perature or salinity (Table 10). Sardine larvae
were most abundant during and after the 1972 El
Niiio; prior to this moderate abundances and rela-
tively large percentage contributions to the
ichthyoplankton occurred only during the warm
1968-69 period (Tables 5, 7). Lowest sardine
abundances were during cold years 1964, 1967,
and 1970 and the 1965 El Nino. Despite low abun-
dances during the 1965 El Nino, there is a signif-
icant positive correlation between ranked sardine
abundance and temperature (p = +0.69, P
< 0.05). Ranked larval anchoveta and sardine
abundances are not correlated (p = -0.07).

The only apparent warm-cold year abundance
pattern among the OL categories is that of the
coastal taxa; this group had lowest abundances
during the 1965, 1972, and 1983 El Nihos and
highest abundances in subsequent 1966 and 1973
transition years. The ranked abundance pattern
of this category has negative correlations
(P < 0.05) with ranked temperature (p = -0.69)
and salinity (p = -0.68) values (Table 9). Both
the myctophid and other mesopelagic categories
appear to have abundance patterns unrelated to
warm-cold hydrographic conditions (Table 9).

Table 9. — Correlations of across-year abundance ranks of
zooplankton and ichthyoplankton categories with ranked (high to
low) mean temperature and salinity values from nine July-Septem-
ber sampling periods off northern Chile, 1964-83. Correlations
based on Spearman's rho tests. Significant values at P s 0.05 are
indicated, but note use of multiple testing. PL = larvae of pelagic
species; OL = other larval taxa.

Temperature

Salinity

(^■C)

(%o)

Zooplankton

-0.48

-0.37

-0.10

-0.17

-0.02

+ 0.05

Larval diversity:

(mean no. taxa/tow)

+ 0.12

0.00

Myctophids

+ 0.02

+0.12

Other mesopelagic

taxa

+ 0.28

+ 0.12

Coastal taxa

-0.69 P<0.05

-0.68 P = 0.05

Zooplankton biomass values show negative but
nonsignificant correlations with temperature and
salinity (Table 9). This is in agreement with the
time-series analysis results of Bernal et al. (1983)
which demonstrated no consistent relations of
zooplankton biomass with cold- or warm-water
events.

Species Groups and
Hydrographic Conditions

The species groups formed by similarity of
between-year abundance ranks demonstrate both
positive and negative correlations with cold- and
warm-year conditions (Table 10). Group I and
Pair II and their member species have negative
correlations with ranked temperature and
salinity values indicating a tendency for higher

LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE

surface temperature (°C) and salinity (%o) values during July-September
available Hydrographic "condition" descriptions for these penods are from Robles et al.

1969

1970

1972

1973

1983

Temperature ("C)
N

N A

110

Range

14.6-17.5

15.5-18.5

14.3-17.4

15.1-19.0

16.0

17.2

15.8

17.2

(S)

(1.0)

(0.7)

(0.9)

Rank

1.5

Salinity (%o)
N

N.A.

108

Range

34.60-35.19

34.75-35.96

34.60-35.13

34 81-35.39

34.87

35.05

34.82

35.09

(S)

(0.16)

(0.18)

(0.13)

(0.19)

Rank

Hydrographic
condition:

Warm

Cold

El Nirio

Transition

El Nino

Table 10. — Correlations of across-year abundance ranks of larval
fish species groups and member species with ranked (high to low)
mean temperature and salinity values from nine July-September
sampling periods, 1964-83. Correlations based on Spearman's
rho tests. Group correlations based on ranks of summed within-
year ranks of member species. Significant values at P < 0.05 are
indicated but note use of multiple testing. " = abundance ranks
adjusted to accommodate for apparent large sampling depth-
related catch differences.

Temperature

Salinity

Species

(C)

(%o)

Group 1

-0.78

P < 0.05

-0.78 P<0.05

Engraulis nngens

-0.52

-0.62

Normanichthys crocken

-0.69

P < 0.05

-0.72 P<0.05

Blenniid D

-0.74

P < 0.05

-0.77 P<0.05

Sebastes sp.

-0.67

-0.55

Group II

+ 0.27

+ 0.20

Diogenichthys laternatus

+ 0.45

+ 0.50

Bathylagus nigngenys

+ 0.27

+ 0.20

Vinciguerna lucetia

+ 0.12

-0.14

Pair 1

+ 0.70

P < 0.05

+ 0,49

Sardinops sagax

+ 0.69

P < 0.05

+ 0.50

Myctophum nitidulum

+ 0.61

+ 0.35

Pair II

-0.62

-0.57

Lampanyctus parvicauda

-0.68

P = 0.05

-0.48

Tnphoturus mexicanus"

-0.50

-0.53

abundances during colder, lower salinity periods.
The rankings of Group I (based on ranks of
summed within-year member species ranks) are
significantly correlated (P < 0.05) with both
temperature and salinity (p = -0.78 in both
cases). Within this group the rankings of
Normanichthys crockeri and Blenniid D are cor-
related (P < 0.05) with temperature and salinity
(p = -0.69 to -0.77); the correlation of Sebastes
with temperature is also relatively strong
(p = -0.67). Within Species Pair II, Lampanyctus
parvicauda abundance has a relatively strong

negative correlation with temperature (p =
-0.68).

Group II and Pair I and their member species
have positive correlations of ranked abundance
with temperature and (with one exception) salin-
ity values (Table 10) suggesting a tendency for
higher abundances during warmer, higher salin-
ity conditions. These correlations are all non-
significant and generally weak for Group II and
its member species (Bathylagus nigrigenys , Vin-
ciguerria lucetia, and Diogenichthys laternatus).
Pair I has a positive correlation (P < 0.05) with
temperature (p = +0.70) primarily due to sardine
abundance ranks.

Species Percentage Composition
Relative to Hydrographic Conditions

Ichthyoplankton percentage composition shows
no striking warm year vs. cold year related
patterns. Total ichthyoplankton composition
comparisons between years of "similar" hydro-
graphic conditions do not give overall higher PSI
values than do comparisons between years of dif-
ferent conditions (Table llA). PSI values (ranges,
means, and proportions of high and moderate val-
ues) from comparisons of cold, transition, and
warm years are similar. However, highest values
(91.6-94.9) come from cold vs. transition year
(1970 vs. 1966 and 1973) and between-transition
year (1966 vs. 1973) comparisons. Additionally,
intercomparisons of the transition, warm, and
El Niilo years give relatively lower values than
do cold-year comparisons. Comparisons between
El Nino years give generally PSI low values.

FISHERY BULLETIN: VOL. 86, NO. 1

Table 11. — Within- and between-hydrographic period ichthyoplankton composition
comparisons presented as range, mean, and standard errors of percent similarity index
(PSI) values and numbers (N) out of total comparisons hiaving moderate to high (e.g.,
-65) values. A. Total larvae. B. OL (other larval taxa) fraction.

PSI values for

Cold

Transition

Warm

El Nino

A. Total larvae

Cold years

Range

67 8-822

62.7-94.9

62.8-82.5

7.3-79.1

(1964, 1967, 1970)

76.6

79.4

73.0

46.5

(SE)

( 4.5)

( 5.2)

( 3.6)

( 9 6)

5/6

4 6

Transition years

Range

57.8-65.1

3.3-79.4

(1966, 1973)

91.9

625

39.7

(SE)

( 1-6)

(12.9)

1/1

1/4

2/6

Warm years

Range

188-76.4

(1968, 1969)

(SE)

79.4

52.3

( 9.2)

1/1

2/6

El Nmo years

Range

220-45.3

(1965, 1972, 1983)

(SE)

31.9

( 7.0)

B. OL

Cold years

Range

46 0-758

30 5-68 3

47.0-85,2

332-88 1

(1964, 1967, 1970)

58.6

48.0

64.6

57.6

(SE)

( 8,9)

( 5.7)

( 5.4)

( 5.9)

1,3

1/6

3/6

3/9

Transition years

Range

36.8-67.9

18 5-66.2

(1966, 1973)

37.5

51.7

41.0

(SE)

( 6.4)

( 7.8)

0/1

1 4

1 6

Warm years

Range

37,2-68,4

(1968, 1969)

(SE)

58.5

53,1
( 5.2)

0/1

1/6

El Nino years

Range

54.3-58.7

(1965, 1972. 1983)

(SE)

55.8

( 1.5)

The OL percentage composition similarly does
not demonstrate clear hydrographically related
patterns (Table IIB). As with the total larvae,
comparisons of cold vs. other years yield most of
the moderate to high PSI values. Highest values
(88.1 and 85.2) come from comparisons of cold vs.
El Nino (1965 vs. 1967) and cold vs. warm (1964
vs. 1969) years. Comparisons within and between
transition, warm and El Nino years give primar-
ily low values. Recalculation of PSIs to accommo-
date for Triphoturus mexicanus lowers mean val-
ues for comparisons with 1973 and 1983 by only
0.4-2.3 and does not affect the overall results.

Chronological Considerations of
Species Composition

When the total larval and OL PSI data are con-
sidered in terms of chronological rather than hy-

drographic periods, various patterns become ap-
parent (Table 12). For the total ichthyoplankton,
comparisons within the 1964-69 data set and be-
tween this and the 1970-73 data set give similar
means, ranges, and proportions of moderate to
high values. In contrast, comparisons within the
1970-73 data set provide more similar values and
a significantly higher mean value than results
from comparisons within the 1964-69 set (Z test,
P < 0.01). This suggests that, despite the varied
hydrographic conditions represented during the
1970-73 period, conditions were favorable for a
repeated fairly similar anchoveta-dominated
ichthyoplankton assemblage during July-
September months.

Chronologically grouped comparisons of the OL
fraction provide somewhat different patterns
from those of the total ichthyoplankton (Table
12B) and indicate a marked change in species

LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE

proportions between 1964-69 and later years.
Comparisons within the 1964-69 OL data set
yield all of the high and most of the moderate PSI
values. The mean PSI value from these compari-
sons is significantly higher (P < 0.05) than that
from 1964-69 vs. 1970-73 comparisons. As with
the total larvae, the 1970-73 PSI values are sim-
ilar and moderately high and the 1983 compari-
son values are relatively low compared with other

Table 12. — Within- and between-time period ichthyoplank-
ton composition comparisons presented as ranges,
means, and standard errors of percent similarity index
(PSI) values and numbers (A/) out of total comparisons
having moderate to high (e.g.. -65) values. A. Total larvae.
B. OL {other larval taxa) fraction.

PSI values for

1964-69

1970-73

1983

A. Total larvae

1964-69

Range

33 9-825

38 2-91.9

3 3-31.0

67.4

67.7

18.2

(SE)

{ 3.6)

( 3.6)

( 4.3)

9 15

11 18

1970-73

Range

79.1-95.0

7.3-22.0

84.4

12.8

(SE)

( 5.2)

( 4.7)

3/3

B. OL

1964-69

Range

29.2-88.1

26.2-67.9

18.5-54.4

60.2

48.4

40.7

(SE)

( 4.6)

( 2.5)

( 5.6)

715

1 18

1970-73

Range

66.2-74.4

58.6-60.2

69.7

592

(SE)

( 2.4)

( 0.5)

0/3

years. Accommodation for T. mexicanus abun-
dance reduces slightly (1.1-3.1) the mean values
from comparisons with 1970-73 and 1983 data
sets and strengthens the significance of difference
(P<0.01) between 1964-69 and 1964-69 vs.
1970-73 mean values.

The difference in OL PSI values between 1964-
69 and 1970-83 is, to a great extent, due to abun-
dance shifts of Group II and Pair II species. The
abundance ranks of all three Group II species
(Table 7) indicate significantly higher abun-
dances during 1970-83 than in earlier years
(Man Whitney U tests, all P's < 0.05). Addition-
ally, the averaged abundance estimates from
these four years are significantly higher than
from earlier years (Z tests; P < 0.01 for Bathyla-
gus nigrigenys and Vinciguerria lucetia,P < 0.05
for Diogenichthys spp.). Together B. nigrigenys ,
V. lucetia, and Diogenichthys spp. (primarily D.
laternatus ) contributed 46.5-72.8^7^ of the OL col-
lected during 1970-83 compared with 5.8-37.1%
during 1964-69 (Table 13).

In contrast. Pair II species Lampanyctus parvi-
cauda and Triphoturus mexicanus were rela-
tively less abundant during 1970-83 than in pre-
vious years. These two species contributed
21-66% of the OL during 1964-69 compared with
<22% during 1970-83 (Table 13). With adjust-
ments to T. mexicanus abundance these species
proportions in 1973 and 1983 decrease to 11.8%
and 13.2%, and those of the Group II species in-
crease to 49.9% and 71.4%, respectively. The rela-
tive abundance decrease of Pair II was primarily

Table 13.— Percentage contribution by dominant OL (other larval taxa) species collected
sampling penods, 1 964-83. Species arranged according to group affiliations based on across
patterns.

during July-September
■year ranked abundance

Percentage contribution for

1964

1965

1966

1967

1968

1969

1970

1972

1973

1983

Group 1
Normanichthys crocken
Sebasles sp.
Blenniid D

1561
0.36
4.81

1.94
0.52

67.40
060
4.00

3.51
0.63
4.45

5.15
1.79
940

22.13
2.16

12.52

1.02
1.19

15.72
0.35
4.31

0.44

Group total

20.78

2.46

72.00

8.59

16.34

24.29

12.52

2.21

20.38

0.44

Group II
Bathylagus nigrigenys
Vinciguerria lucetia
Diogenichthys spp.

3.63
2.81
5.63

6.60

0.52

24.45

1.61
0.70
350

608

3.82

19.11

13.87

6.94

16.33

0.75

2.16

12.15

9.01
26.78
26.78

24.14
23.37
25.29

17.53

9.31

19.68

13.78

1.62

49.96

Group total

12.07

31.57

5.81

2901

37.14

15.06

62.57

72.80

46.52

65.36

Pair 1
Myctophum nitidulum

0.36

—

0.10

—

0.89

0.75

—

2.05

2.15

2.01

Pair II
Lampanyctus parvicauda
Triphoturus mexicanus

37.21
28.86

41.40
20.44

15.77
4.99

3947
18 17

24.16
10.29

32.11
26.37

7.13
14.27

12.67
8.40

6.06
11.82

6.78
14.70

Pair total

66.07

61.84

20,76

57.64

34.45

58 48

21.40

21.07

17.88

21.48

FISHERY BULLETIN: VOL. 86. NO 1

due to the increased numbers of Group II species
and decreased numbers of L. parvicauda . The
1970-83 averaged abundance of L. parvicauda is
significantly lower (Z test; P < 0.01) than that of
1964-69. Triphoturus mexicanus averaged abun-
dance (both adjusted and unadjusted values)
is similar (P > 0.05) between the two time peri-
ods.

DISCUSSION

The northern Chilean ichthyoplankton data set
is obviously weakened by lack of information
from the 1974-82 period; this missing informa-
tion is critical for an appreciation of the temporal
extent and relative constancy of the apparent
ichthyoplankton composition change in 1970-73
vs. earlier years. This data set also suffers from
limited seasonal coverage which prohibits exami-
nation of between-year variations in spawning
time and intensity as a cause of interannual
abundance fluctuations and apparent composi-
tion change. However, the existing data set does
provide coherent coverage over varied hydro-
graphic conditions between 1964 and 1973 and is
sufficient to test for correlations with short-term
(e.g., year to year) fluctuations in hydrographic
conditions.

The large interannual changes in abundance
and composition of the northern Chilean ichthy-
oplankton can to a certain extent be related to
interannual changes of hydrographic conditions
in the Humboldt Current. This has been demon-
strated through correlations of ranked tempera-
ture and salinity values and abundances of
coastal species, sardine, and Lampanyctus parvi-
cauda (Tables 9, 10). The temperature and salin-
ity values used in these correlation tests
represent ambient conditions during the July-
September spawning period and therefore
possibly reflect only conditions affecting egg and
early larval (e.g., to stages capable of substantial
net avoidance) survival. These values do not nec-
essarily reflect longer term conditions affecting
abundance, distributions, and fecundities of adult
populations or later larval survival and recruit-
ment. However, there is a generally good corre-
spondence between these values and reported
longer term hydrographic conditions in the Hum-
boldt Current over the 19-yr timespan (e.g.,
Table 8; Robles et al. 1976; Bernal et al. 1983;
Guillen 1983; Bakun 1987).

Despite significant correlations between abun-
dances of some ichthyoplankton components and

temperature and salinity values, there is no ap-
parent consistency of total larval or OL species
percentage compositions during years of
"similar" hydrographic conditions (Table 11).
More coherent patterns emerge from consider-
ations of the 1964-69 and 1970-73 data sets
(Table 12). This chronological separation is also
supported by the ranked abundance patterns of
the various species groups and pairs {Table 7).

Among the least confusing across-year abun-
dance patterns demonstrated by the ichthy-
oplankton are 1) generally greater abundance
of Group II species after 1969, 2) greatest abun-
dance of Pair I species after 1970, and 3) pre-
dominantly higher abundances of one of the Pair
II species prior to 1970 (Table 7). Associated with
the Group II and Pair II abundance patterns are
large shifts in their relative proportions
(Table 13).

The shift from relatively large percentage con-
tributions by Lampanyctus parvicauda and
Triphoturus mexicanus to larger proportions of
Diogenichthys spp., Bathylagus nigrigenys , and
Vinciguerria lucetia after 1969 is notable. The
abundances of these mesopelagic species, unlike
those of anchoveta and sardine, are not directly
influenced by man's fishing activities and so may
be interpreted as indicators of environmental
change. Furthermore, the timing of these species
absolute and relative abundance changes pre-
ceded by several years the dramatic changes in
anchoveta and sardine stocks off of northern
Chile (Fig. 4) and so cannot be directly related to
biological consequences of change in the domi-
nant pelagic schooling fish stocks.

Although fragmentary, there is evidence for a
change in zooplankton biomass values off north-
ern Chile (18°-24°S) occurring in 1969 (Fig. 5)
which, like OL percentage composition, suggests
a possible environmental change. Time series
analysis of quarterly zooplankton biomass values
during 1964-73 indicate generally lower biomass
during 1969-73 relative to the 1964-68 period.
As with total larval abundance (Table 9), these
zooplankton biomass variations do not appear to
be related to warm year-cold year events (Bernal
et al. 1983).

The changes in OL composition and zooplank-
ton biomass suggest that there was subtle but
large-scale (low-frequency) environmental tran-
sition occurring in the 1969—70 period. Various
indications of environmental change occurring
about this time are present in long-term physical
data bases from Chile and Peru. Predominantly

LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE

3,0

1955

83 84

Figure 4.— Total catch and catch of dominant species taken in the northern Chilean pelagic fishery, 1955-84.

negative sea surface atmospheric pressure
anomalies occurred off of Arica from 1960 to 1972
with a strong negative anomaly occurring in
1969; predominantly positive anomalies occurred
there after 1972 (Fig. 6A). A similar but less ex-
treme change from negative or neutral anomalies
to predominantly positive anomalies occurred off
Iquique (20"S) in 1970 (Fig. 6B; Kelly and Blanco
1983).

Off Peru (5°-15°S) the wind driven turbulent
mixing index of surface waters shows a general
increase during and after the 1972 El Nino event
(Fig. 7A; Bakun 1987). A probable result of this
increased turbulence is an increase in standard
deviations associated with monthly temperature
values; standard deviations above the 30-yr mean
generally persisted throughout the year from
1972 to 1984 (Fig. 7B) and suggest increased
physical variability and heterogeneity in this
later period. Comparable data sets from northern
Chile are not available to determine if these lat-

ter two features were also characteristic of the
Chilean area.

How these observed atmospherically related
changes could be related to changes in the marine
environment off northern Chile is uncertain. It is
possible that the observed changes in atmos-
pheric pressure off Arica and Iquique have associ-
ated changes in advection of water mass and fau-
nal sources. Bernal et al. (1983) discussed El Nino
related changes in water mass distribution off
Chile in 1972 and 1973 relative to cold-year 1967.
These changes involved southerly extensions of
oceanic subtropical and equatorial subsurface
waters, strengthening of the spring-summer ther-
mocline, and cessation of coastal upwelling.
These authors did not examine water mass distri-
butions in the 1968-70 period. However, lowered
zooplankton biomass starting in 1969 and the OL
composition change starting around 1970 suggest
that the hydrographic conditions attributed to the
1972 El Nino may have been an intensification of

ZOOPLANKTON BIOMASS

FISHERY BULLETIN: VOl- 86, NO 1

2 0

1 .0

o
■o

c
o

I 0 -

- 2 0

—

— 1

1 1

1966

1967

1966

969

1970

1971

972

1973

Year

Figure 5. — Time series estimates of quarterly zooplankton biomass values from northern Chile
(18°-24°S), 1964-73, standardized accordmg to the long term standard deviation. W = warm
years; C = cold years. From Bernal et al. 1983.

SEA SURFACE
ATMOSPHERIC PRESSURE ANOMALIES

3 -

3 -
2 -

1 -

/^ 1

L-A

\ ''

1 -

J ^

/•M

-J

V \H

I960 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82

Year

Figure 6. — Sea surface atmospheric pressure anomalies off of (A) Arica (18'Sl and (B) Iquique (20''Sl, 1960-82. From

Kelly and Blanco 1983.

LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE

TURBULENT MIXING INDEX

350 -

300 -

250 -

200 -

65 70

Year
STANDARD DEVIATIONS

4 0-
3 5
3 0
2 5
2.0

I 5

1.0

Std. dev

12- mo- running meon of std. dev.

.— Long - Itrtn meon of tid. dev.

Yeor

Figure 7.— (A) Surface layer turbulent mixing index values and (B) standard deviations and 12-mo running mean of standard
deviations associated with mean monthly sea surface temperature values off of Peru (5°-15°S), 1953-84. Dotted lines represent
long-term mean values of each index. From Bakun 1987.

conditions initiated during the warm 1968-69 pe-
riod; these could possibly have persisted and in-
tensified again during the 1976 and 1983 El Nino
events.

It is also likely that the 1970 change in OL
composition is related to onshore advection of
northern or oceanic water masses and associated
faunal assemblages, but this cannot be confirmed.
All involved species are relatively abundant in
coastal Peruvian and south eastern tropical
Pacific waters (Ahlstrom 1971, 1972; de Castillo
1979; Santander and de Castillo 1979) and in
more southern coastal Chilean waters (Table 2),
but their wintertime relative abundances in these
areas have not been documented. It is also possi-
ble that the changes in species composition are
related to locally lowered zooplankton abundance
(e.g., that the Group II species are relatively more
successful than Lampanyctus parvicauda and
Triphoturus mexicanus during periods of lowered
secondary productivity levels). Alternatively, the

observed change could be due to altered seasonal
spawning activity which is not treated in the
present study. However, whatever the cause,
there is evidence for an environmental change
in the study area, and this may be also implicated
in changes occurring within the PL ichthyoplank-
ton fraction.

Increased abundances of sardine and Mycto-
phum nitidulum (Pair I species) during and after
1972 may be further evidence for a changed
marine environment off of northern Chile. Addi-
tionally, the significant correlation between lar-
val sardine abundance and temperature
(Table 10) suggests that elevated temperatures
may have been important for increased spawning
activity and/or increased success of hatching and
early larval survival. Given this observation one
may speculate that the increased sardine catches
after 1973 (Fig. 4) are related to increased fre-
quency of warm-water events in the 1964-84 pe-
riod relative to earlier years. Increased sardine

FISHERY BULLETIN: VOL. 86, NO. 1

catches in 1973 resulted from apparently ex-
tremely good survival and recruitment of individ-
uals spawned during the warm 1968-69 period
(Serra"^). Similarly, good survival of the large
1972 El Nino spawn could explain the huge catch
increases in 1976 and later years (Fig. 4). How-
ever, mean biomass estimates of age groups con-
tributing to the Chilean fisheries catch from 1974
to 1981 (Serra 1983) indicate increasing contribu-
tions after the 1967 year class with marked in-
creases beginning with the 1970 year class; this
suggests that factors other than temperature
(e.g., environmental change starting in 1969-70)
may also be responsible for increased larval sur-
vival and recruitment. A possible cause is in-
creased nearshore influence of equatorial and
subtropical waters (Santander and Flores 1983).
Because the sardine abundance increase was ini-
tiated prior to the 1972 anchoveta decline off
northern Chile (Fig. 4), it is difficult to implicate
reduced anchoveta competition as the cause of the
early sardine population growth in this area.

The grouping of anchoveta with three coastal
species (Group I), and significant correlation of
anchoveta and Normanichthys crockeri larval
abundances are extremely interesting and imply
that the spawning intensity and/or early stage
survival of these four species are influenced in
similar ways by interannual changes off northern
Chile. Unfortunately, little is known about the
natural histories or population abundances of the
coastal species. Because of the group composition,
it is logical to suspect that coastal processes are
important factors influencing their larval abun-
dance. The significant negative correlation of the
group as a whole, and of two of the coastal species,
with ranked temperature and salinity values
(Table 10), suggests that coastal upwelling and/or
increased coastal influence by subantarctic
waters, and theoretically enhanced food supplies,
are important factors.

Given the present data set and information
from recent publications, a case can be made for a
low-frequency environmental change influencing
the abundances of anchoveta and sardine larvae
as well as the larvae of coastal and mesopelagic
species during the 1964-84 period. The Chilean
OL composition suggests an environmental
change (e.g., an atmospherically related oceanic
circulation change) starting with the 1968-69
warm-water event. This coincided with apparent

successful survival of sardine larvae and
markedly increased recruitment by 1968 and
later year classes despite varied warm water-cold
water events between 1968 and 1973. Physiologi-
cal anomalies of Peruvian anchoveta stocks in
1971 suggest that these fishes may have experi-
enced environmental change at that time. Unusu-
ally low proportions (e.g., 40% vs. typically 90%)
of potential spawning-sized fish were sexually
mature during the 1971 spawning season and fat
content of the 1971-72 catch was anomolously
high, indicating unusually low transfer of body
fat to gonadal products (Sharp 1980). Starting
with the 1972 El Nino was 1) an obvious in-
creased incidence of penetration of subtropical
surface waters toward the Peruvian coast, 2) co-
incidental onshore and southward expansion of
sardine spawning activity off both Peru and
Chile, 3) southward expansion of Peruvian an-
choveta spawning activity into new spawning
areas between 14°S and 18°S (e.g., to northern
Chile), and 4) a succession of years of poor an-
choveta larval survival off Peru and Chile (San-
tander and Flores 1983; Serra 1983). Environ-
mental conditions favorable for growth of sardine
populations, as well as of mackerel and jack
mackerel populations, off both Chile and Peru
have persisted since the early to mid-1970's (San-
tander and Flores 1983; Serra 1983).

The lack of Chilean ichthyoplankton data from
the 1974-82 period precludes evaluation of the
constancy of altered species composition during
that time. However, there are indications that
change is once more occurring off northern Chile.
Preliminary analysis of ichthyoplankton samples
collected between Arica and Antofagasta during
4-14 August 1985 indicates a clear dominance by
anchoveta larvae at a markedly higher mean
abundance level than encountered in the 1964-
83 samples; sardine and Trachurus larval abun-
dances are comparable to those in the 1973 sam-
ples (Table 14). The other species have not yet
been analyzed, but Normanichthys crockeri is

Table 14. — Mean abundance estimates and standard errors (num-
bers per 10 m2) and percent frequency of occurrence (F) of PL
taxa, OL and total larvae collected in 81 1 00-0 m WP2 net samples
off northiern Chile (18°-24°S) during 4-24 August 1985.

Species

(SE)

(F)

3Serra, R. Unpubl. manuscr. Subsecretaria de Pesca,
Teatinos 120, Piso 11, Of. 44, Santiago, Chile.

Engraulls ringens 5,535.3 (1,844.1) (90.4)

Sardinops sagax 63.8 ( 24.9) (26.6)

Trachurus murphyi 1.2 ( 0.9) (2.1)

Otfier species 233.7 ( 28.4) (92.5)

Total 5,834.0 (1,838.2)

LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE

once again noted to be among the abundant OL
taxa. It will be of great interest to see the recruit-
ment resulting from this 1985 anchoveta spawn.
If the 1985 anchoveta year class is relatively suc-
cessful during this period of continued large sar-
dine stocks, it will lend support to the idea that
the anchoveta-sardine population fluctuations
have been primarily regulated by low-frequency
hydrographic events. In a long-term context such
events may not be unusual to the Humboldt Cur-
rent area. Fish scales present in sediment records
from coastal Peruvian waters indicate that sar-
dine replacements of typically dominant an-
choveta stocks have occurred at infrequent inter-
vals over the past 11,000 years (De Vries and
Pearcy 1982). In light of this, the ecological
events of the past 20 years may be naturally oc-
curring, physically mediated, but probably fish-
eries enhanced, fluctuations in this hydrographi-
cally complex region. Because of our limited time
reference, we have not previously acknowledged
such fluctuations as being "normal".

ACKNOWLEDGMENTS

We extend our appreciation to Dick Parrish and
Andy Bakun, Pacific Fisheries Environmental
Group (SWFC/NMFS/NOAA), whose interest in
eastern boundary current comparison studies has
made this work possible. We also thank Rodolfo
Serra and Gary Sharp for their valuable com-
ments and discussions of the manuscript. Appre-
ciation is extended to numerous other people in-
cluding ship captains and crew and laboratory
assistants, especially Hernan Miles, who have
helped in sample collection and processing over
the 20-yr period.

This work was funded by NMFS/NOAA Solici-
tation WASC-84-00075.

LITERATURE CITED

Ahlstrom, E. H.

1971. Kinds and abundance of fish larvae in the eastern
tropical Pacific, based on collections made on
EASTROPAC I. Fish. Bull., U.S. 69:3-77.

1972. Kinds and abundance of fish larvae in the eastern
tropical Pacific on the second multivessel EASTROPAC
survey, and observations on the annual cycle of larval
abundance. Fish. Bull., U.S. 70:1153-1242.

Bakun, A.

1987. Monthly variability in the ocean habitat off Peru as
deduced from maritime observations, 1953-84. In D.
Pauly and I. Tsukayama (editors), the anchoveta and its
ecosystem, p. 46-74. International Center for Living
Aquatic Resources Management (ICLARM), Manila.

BERNAL, p. a., F. L. RoBLES, AND O. RoJAS.

1983. Variabilidad fisica y biologica en la region merid-
ional del sistema de corrientes Chile-Peru. FAO Fish.
Rep. 291:683-711.
CONOVER, W. J.

1971. Practical nonparametric statistics. John Wiley
and Sons, N.Y., 462 p.
DE Castillo, O. S.

1979. Distribucion y variacion estacional de larvas de
peces en la costa Peruana. Inf. Inst. Mar Peru 63:1-
32.

De Vries. T J . and W G. Pearcy.

1982. Fish debris in sediments of the upwelling zone off
central Peru: a late quarternary record. Deep-Sea Res.
28(1A):87-109.

DiXON. W J., AND F. M. Massey, Jr.

1969. Introduction to statistical analysis. McGraw-Hill,
N.Y., 638 p.

Guillen, O

1983. Condiciones oceanograficas y sus fluctuaciones el el
Pacifico sur oriental. FAO Fish. Rep. 291:607-658.

Kelly, R , and J. L. Blanco.

1983. Fluctuaciones ambientales y su relacion con la
abundancia de recursos pelagicos en la zona norte-centro
de Chile. Inst. Fom. Pesq. 830040, 22 p.
Parrish. R H , A Bakun. D M Husby, and C S Nelson.

1983. Comparative climatology of selected environmental
processes in relation to eastern boundary current pelagic
fish reproduction. FAO Fish. Rep. 291:731-777.
Robertson. A

1970. An improved apparatus for determining plankton
volume. Fish. Bull., S. Afr. 6:23-26.

ROBLES, F L , E ALARCON. AND A ULLOA.

1980. Water masses in the northern Chilean zone and
their variations in the cold period (1967) and warm
periods ( 1969, 1971-73). Proceedings of the workshop on
the phenomenon known as "En Nino", p. 83-174.
UNESCO.

Sameoto. D

1980. Distribution and abundance of six species of fish
larvae in Peruvian waters and their relationship with the
physical and biological environment. Bol. Inst. Mar
Peru Callao 5:164-170.

Santander, H , and O S de Castillo

1979. El ictioplancton de la costa Peruana. Bol. Inst.
Mar Peru Callao 4:69-112.

Santander. H . and R Flores

1983. Los desoves y distribucion larval de cuatro especies
pelagicas y sus relaciones con las variaciones del ambi-
ente marino frente al Peru. FAO Fish. Rep. 291:835-
867.
Serra. J R

1983. Changes in the abundance of pelagic resources
along the Chilean coast. FAO Fish. Rep. 291:255-284.
Sharp. G D

1980. Report of the workshop on effects of environmental
variation on survival of larval pelagic fishes. In G. D.
Sharp (editor). Workshop on the effects of environmental
variation on the survival of larval pelagic fishes, Lima,
Peru, April-May 1980. Workshop Report No. 28, p. 15-
59. Intergovernmental Oceanographic Commission,
UNESCO, Paris.

Tate. M W , and R C Clelland

1957. Nonparametric and shortcut statistics in the social,
biological and medical sciences. Interstate Printers and
Publishers, Danville, IL, 171 p.

FISIIKRY BUI.LKTIN VOL HH. NO 1

UNESCO Wyktki, K

1968. Zooplankton sampling. Monographs on oceano- 1967. Circulation and water masses in the eastern equa-

graphic methodology 2. Imprimeries Fopulaires, torial Pacific Ocean. Int. J. Oceanol. Limnol 1:1 17-147.

Geneva. 174 p. Yashnov, V A

WmiTAKKK. K H 1959. A new model of a volume meter for rapid and pre-

1975. Communities and ecosystems MacMillan Pub- cise plankton evaluation undt-r field conditions. Zool.

lishing Co., N.Y., 385 p. Zh. (Moscowl 38:1741-1744.

ESTIMATION OF NATURAL MORTALITY IN FISH STOCKS:

A REVIEW

E. F VetterI

ABSTRACT

The instantaneous rate of natural mortality (M) is an important but poorly quantified parameter m
most mathematical models of fish stock dynamics. This report reviews methods used commonly to
estimate M for fish stocks, sensitivity of some common fishery models to values chosen for M, and
evidence refuting the common assumption that a constant value can be an adequate approximation
of A/ within single stocks.

With the exception of simple surplus production
models (e.g., Schaefer 1954; Pella and Tomlinson
1969) all mathematical models offish stock dy-
namics include as a parameter the instantaneous
rate of natural mortality (M). The models do not
require explicitly any particular form for M; it
can be constant or can vary in any imaginable
form. But because natural mortality has proved
extremely difficult to measure directly, M is as-
sumed almost universally to be some constant
specific to whatever stock is being modeled. This
is particularly true for analyses of commercial
fish stocks, which often require estimates of M
only for the postrecruit ages. Decreases in natural
mortality with increasing age during egg and
postlarval stages are so dramatic compared to ap-
parent changes during postrecruitment ages (e.g.,
Gushing 1975) or compared to differences be-
tween different sexes, collection sites, seasons,
years, cohorts, or stocks within species, that vari-
ations in M during these later (postrecruitment)
ages are often assumed negligible.

Whether this assumption is in fact acceptable is
the subject of this report. The answer is no, it is
probably not acceptable in most cases. That an-
swer follows from the information presented in
Sections II through V, with the following conclu-
sions:

Section II: Current methods for estimating nat-
ural mortality: a review of methods used cur-
rently to estimate M in fish populations. All of
these methods have strong limitations or disad-
vantages.

^Southwest Fisheries Center La Jolla Laboratory, National
Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA
92038.

Manuscript accepted October 1987.
FISHERY BULLETIN: VOL. 86, NO. 1. 1988.

Section III: Sensitivity of fishery models to
choices for M: a review of the sensitivity of some
standard fishery models to different choices for
input value(s) of M. Existing studies show that
the models are sensitive and that sensitivity is
affected not only by the values chosen for M, but
also by interactions between M and the values
chosen for other parameters in the models.

Section IV: Evidence for nonconstant M; factors
influencing death rate: a review of factors as-
sumed or shown to affect M in fish stocks. Avail-
able information implies that many such factors
exist, acting alone or in concert.

Section V: Evidence for nonconstant M; vari-
ability within and between groups: a review of
existing quantitative evidence for the extent of
variability in M between but especially within
stocks. Because almost all fishery models focus on
single stocks, variability within stocks (as op-
posed to between stocks) is the most important
question. Some studies show strong differences
between mortality rates of various groups offish;
some do not. Those which do not have tended to
assume there would be none, and have often used
catch curve regression analysis to derive a single
estimate from data combined over many groups
(usually years) of data. The few studies from
which it is possible to determine ranges of esti-
mates show differences of at least 50 to 100% be-
tween minimum and maximum estimates for sin-
gle groups (e.g., stocks) offish.

The report's major conlcusions are that natural
mortality is far from constant for many fish
stocks, and that this variability is extensive
enough that it should not be ignored. Analyses of

FISHERY BULLETIN: VOL. 86, NO. 1

fish stock dynamics need much more rigorous es-
timates of within-stock variability (both trends
and variance) in M for exploited fish stocks.

II. CURRENT METHODS FOR
ESTIMATING NATURAL MORTALITY

Three methods are used currently or have been
proposed to estimate M in fish populations:
1) analysis of catch data, usually from commer-
cial fisheries but also from sampling programs
specifically conducted for stock assessment (this
includes mark-recapture studies), 2) correlations
of M with other life history parameters, and
3) estimation of deaths due to predation. I de-
scribe below each method in turn, listing both
advantages and disadvantages of each.

Catch-Analysis Methods

Methods for deriving estimates of natural mor-
tality from catch data are based on measuring
decreases in abundance, either relative or abso-
lute, in groups of fish during two or more succes-
sive periods of time. Groups may be distinguished
on the basis of any identifiable characteristic,
e.g., size (length or weight), age, sex, location and
time of capture, or some identifiable tag or mark.

The most common grouping is by age, for two
reasons. First, age has been considered histori-
cally the most important factor potentially affect-
ing estimates of mortality rate and subsequent
results from the most commonly used fishery
models (e.g., Heincke 1913; Baranov 1918). This
is probably because the methods were developed
for temperate water fisheries which tend to have
obvious annual reproductive cycles, so that indi-
vidual year classes are often relatively easy to
distinguish. Second, the earliest method of esti-
mating M (catch curve analysis, discussed below)
depends on determining the rate during succes-
sive ages.

Regardless of the grouping criterion, methods
for estimating M use generally one of two types of
data. The first type is simply subsamples of un-
marked catch. These subsamples contain fish se-
lected randomly and classified into groups on the
basis of size (length or weight). The second type is
mark-recaptures, in which previously marked in-
dividual fish can be identified and classified after
recapture into groups on the basis of this positive
identification. Estimates of mortality are usually
derived from samples of unmarked fish by analy-
sis of resulting "catch curves" (Ricker 1975). Be-

cause it has been used so frequently, catch curve
analysis is discussed below in some detail.

With marking it is possible to follow the history
of individual fish, so many different types of esti-
mation procedures exist for deriving estimates of
mortality from mark-recapture data (e.g., Ricker
1975; Jones 1979; Brownie et al. 1985). Because
so many variations are possible, marking experi-
ments are discussed only generally, stressing
the basic advantages and disadvantages of mark-
ing data relative to data from unmarked samples,
in deriving estimates of mortality from these
data.

Size-frequency distributions from unmarked
subsamples of catch (the first type of data) are
converted usually to age-frequency distributions,
on the basis of previously determined relation-
ships between age and length or age and weight.
Subsequent analyses concentrate on analyzing
this resulting curve of age-composition (e.g.,
Ricker 1975). Abundance usually decreases expo-
nentially with size (or age) in this type of sample.
Converting the abundances to their logarithmic
values often results in a relatively linear decrease
during most exploited ages (or sizes), after some
initial increase in vulnerability. Graphs of these
logged-frequency distributions are usually called
"catch curves", and their analysis, "catch curve
analysis". "Catch curve analysis" generally con-
sists of determining the best-fit straight line
through the decreasing portion of the logged-
frequency distribution, because if the decrease in
abundance is truly exponential, the slope of this
line through the log-transformed data is the in-
stantaneous rate of decrease in abundance (e.g.,
Ricker 1975).

There are two basic t3T)es of catch curves, dis-
tinguished on the basis of when the data were
collected and how many groups are represented in
the curves. The first, horizontal catch curves, in-
cludes data from several groups (e.g., size or age
classes) collected at a single point in time (or com-
bined from two or more points in time). Thus,
horizontal catch curves reflect "ancient history".
The individuals contributing to the frequency dis-
tribution were not originally all members of the
same group. To use this type of catch curve, one
must assume that for each successive age, risk of
mortality has been historically the same for all
individuals achieving that age. If this has not
been the case, the catch curves may show various
types of curvature in the descending leg, but ab-
sence of curvature is no guarantee that the rates
have in fact been constant.

VETTER: NATURAL MORTALITY IN FISH STOCKS

Thus, horizontal curves are subject to the ex-
tremely restrictive assumption that the groups
from which the data were collected must be in
steady state relative to each other, i.e., their rela-
tive abundances must be constant through time.
If this is true, then a graph of data collected at a
single point in time, which may include, for ex-
ample, individuals from 5 consecutive year
classes displayed as frequencies at 5 consecutive
ages, will look the same as the 5 graphs of data
that will result from collecting samples during 5
consecutive years (ages) from each of the 5 year
classes. If these conditions are not met, simple
linear fitting to determine a single estimate for
mortality will be inappropriate.

The second type of catch curve, longitudinal,
includes data collected from a single identifiable
group over a protracted period of time. Most often,
this will be a single cohort of fish such as single
year class, sampled during successive years. Lon-
gitudinal curves are not subject to the assumption
of steady state, but do share with horizontal catch
curves several other severe disadvantages. These
include 1) groups must be adequately identifi-
able; 2) groups must be closed to migration, so
that changes in abundance are due only to fishing
or natural mortality, or if migration does occur, it
must occur in proportion to the age distributions
in the local groups; 3) samples must represent
adequately the true composition of the groups in
nature; 4) rate(s) of mortality must be relatively
constant between groups over time, so that the
log-transformed frequency distributions are truly
linear (e.g., Jensen 1984); 5) compensatory rela-
tionships between stock levels and natural mor-
tality, or fishing mortality and natural mortality,
must not be present.

Methods for estimating M, which assume to
greater or lesser degrees that the conditions listed
above are met, have been described repeatedly.
The methods tend to fall into two categories.
Methods in the first category estimate M from
catch records of unexploited or lightly exploited
groups of fish. In these groups, F equals or ap-
proximates zero. Therefore, the observed rate of
decrease (Z ) equals or approximates M, because Z
equals the sum of F and M (e.g., Heincke 1913;
Baranov 1918; Ricker 1947; Beverton and Holt
1957; Robson and Chapman 1961; Pauly 1982;
Munro 1982; and among others).

Methods in the second category estimate M by
determining Z at various levels of fishing effort,
then using the observed relationship between Z
and effort to predict, via regression analysis or

manipulation of various ratios, the value of Z
at zero effort (e.g., Silliman 1943; Beverton and
Holt 1957; Paloheimo 1961; Lander 1962; Chap-
man and Murphy 1965; Paulik and Robson 1969;
Gulland 1983; Butler and MacDonald 1979;
Fournier and Archibald 1982; Caddy 1984; and
others).

These methods are most appropriate for analyz-
ing catches of unmarked fish. Accurate results
depend strongly on accurate measures of catch
per unit effort (CPUE) and constant catchability
(q) because if these conditions (in addition to
those listed above) are not met, observed relation-
ships between abundances in different sample
groups may not reflect true differences between
groups in situ.

Marked fish present fewer problems. Advan-
tages include 1) concentration on measuring rela-
tive rather than absolute differences between
abundances of different groups, 2) immigration
need not be considered, as entire original groups
are known to carry marks, and 3) with suffi-
ciently large samples, it becomes possible to test
for differences in mortality rate between different
groups (e.g., between ages, between sexes, or be-
tween sampling sites), rather than having to as-
sume that such effects are negligible.

Reviews and descriptions of various mark-
recapture methods appear in papers by Seber
(1973), Ricker (1975), Jones (1979), and Brownie
et al. (1985). Some of the newer types of marking
analyses can solve many of the most vexing prob-
lems associated with traditional catch curve anal-
ysis (e.g.. Reed and Davies 1980; Hochbaum and
Walters 1984; Burnham and Andersen 1984;
Burnham et al. 1984; Brownie et al. 1985).

Several disadvantages unique to marking oper-
ations counteract these advantages, however,
even with the newer methods. These disadvan-
tages include various types of mark-induced
effects on mortality rates, behavior, and vulnera-
bility to capture, as well as mark loss, unrepre-
sentative mixing of marked fish with their origi-
nal groups prior to recapture (e.g., Ricker 1975),
and especially in commercial fisheries, under-
reporting or incorrect reporting of recaptures.

Both analysis of catch curves from unmarked
fish and analysis of mark-recapture data have the
advantage of requiring only catch (and usually
effort) data, and these data can generally be col-
lected by sampling catches from commercial fish-
eries. However, in addition to problems specific to
each method, they have in common one or more
other major disadvantages: 1) inability to distin-

FISHERY BULLETIN; VOL 86, NO 1

guish between losses (or gains) from migration or
recruitment versus losses due to fishing or natu-
ral causes, 2) imprecision in the estimates of M
obtained (e.g., Beverton and Holt 1957; Taylor
1958; Bishop 1959; Paloheimo and Dickie 1966
Ricker 1975, 1977; Doubleday 1976; Pauly 1980
Larkin and Gazey 1982; Paloheimo 1980, 1982
Myers and Doyle 1983; Roff 1984), 3) sensitivity
to size-specific mortality affecting the estimated
age-structure of the group (Ricker 1969), 4) errors
in estimates of age, such that abundances-at-age
derived from age-length conversions are unrepre-
sentative, 5) where analyses are conducted on
data combined over two or more cohorts, the un-
likely condition that mortality rates were in fact
similar for all cohorts, and 6) problems inherent
in the analyses themselves (e.g.. Barlow 1984).
Disadvantages 1, 4, and 5 may not apply to
marked fish. Disadvantage 5 does not apply to
single cohorts. But collections from marked
groups and single cohorts are still vulnerable to
the other problems.

Further, although in principle it would be pos-
sible to estimate M for different ages, times, or
places, most commonly in practice a single,
fishery-wide constant M is estimated by pooling
data from throughout the fishery. By implication,
the analyst is assuming that the exploited stock
was more or less in steady-state over all times and
areas of catch so that M was relatively constant
while that data set was collected and while (his-
torically) the observed age-distributions were
being created. In fact, substantial evidence exists
that M is not constant, either within a single
stock over time (age) or between stocks of a given

species in different areas (Sections IV and V).

A final disadvantage is that catch-curve analy-
ses are fundamentally unmechanistic, generated
simply by charting changes in abundance. Catch-
curve analyses cannot predict the effect of
changes in factors that control M; thus there is
little hope of predicting M in the future should
conditions change.

Life History Methods

A second approach to estimating the instanta-
neous rate of natural mortality in fish stocks is
based on the observation that M often correlates
strongly with life history parameters, such as
growth rate, age at sexual maturity, costs of re-
production, and maximum age (Table 1).

Typically in such studies, analytical formulas
are derived from theoretical relationships be-
tween the various parameters (e.g., Beverton
1964), or empirical formulas are derived from re-
gression of M against one or more of the parame-
ters (e.g., Hoenig 1983). These models have two
significant advantages: 1) they require minimal
amounts of data, and 2) they are useful in demon-
strating broad trends across species and in devel-
oping ecological theory. But because they produce
only a single and often very imprecise estimate of
M for any given group offish, they are not partic-
ularly effective for generating precise estimates
of natural mortality or for determining the exis-
tence or extent of trends and variability in M for
given stocks. They will also be no better than the
methods used to estimate the values of M used in
the regressions.

Table 1. — Studies relating instantaneous rate of natural mortality to life history traits in fish.

Traits

Species

Source

^^max. ^'f' ~''-inf' nietabolic rate,

Various

Beverton and Holt 1959

reproduction

^max. !<• '-inf. "'-asm. f'Shing

clupeids, engraulids

Beverton 1963

^^V,n,

general

Ursin 1967

'max' 'maxbiomass, '*•

general

Alverson and Carney 1975

growth rate

young fish

Ware 1975

'-asm. gonad size, condition factor

gadoids

Jones and Johnston 1977

'max

general

Blinov 1977

gonad body weight index.

Ms/w, r^3„ /.,„

general

Gunderson 1980

tV|n(, L|n), k. water temperature

175 stocks

Pauly 1980

energy cost of reproduction

general

Myers and Doyle 1983

max

various

Hoenig 1983

weight

various

Peterson and Wroblewski 1983

'<. '-inf. '-asm

various

Roff 1986

^Maximum age,

2Von Bertalanffy growth parameter.

3Maximum length.

''Length at age of sexual maturity.

^Maximum weight.

6Age at occurrence of cohort's max biomass.

''Age of sexual maturity.

VETTER: NATURAL MORTALITY IN FISH STOCKS

Predation Methods

A third class of estimators extends single spe-
cies cohort analysis to a multispecies assemblage
incorporating the major predators and alterna-
tive prey of the stock in question. Single species
cohort analysis is used to estimate population
abundances and annual values for the instanta-
neous rate of fishing mortality (F) for single
groups, usually year classes, of fish (e.g., Pope
1972; Ricker 1975; Gulland 1983). The multispe-
cies extension simply combines cohort analyses
for several species (e.g., Anderson and Ursin
1977). The methods all generate estimates of M as
the sum of some constant rate of nonpredatory,
nonfishing mortality plus the total estimated flux
of prey (stock) to each of the major predators. This
feeding flux to predators is estimated by first
using cohort analyses to reconstruct population
sizes of the various groups of predator and prey,
then combining these population sizes with ob-
served growth rates for the predators and with
estimated preferences for various prey. Thus it
becomes possible to estimate the predatory com-
ponent of M.

Versions of the method have been described by
Anderson and Ursin (1977), Majkowski (1981),
and Pope and Knights (1982). Applications in a
marine system (North Sea) have been described
by Anderson and Ursin (1977), in an ecosystem
context, by Laevastu et al. (1982) and in lake
systems by Forney (1977) and Stein et al. (1981).

The predation method has been developed pri-
marily from analyses of marine systems, espe-
cially the North Sea, and much of the literature
exists only as "mimeos" or notes associated with
ICES (International Council for the Exploration
of the Seas) activities. The most readily available
discussion of this approach appeared in Mercer
(1982), which includes a critical review and dis-
cussion by Ursin (1982) of the various methods.
Several other discussions appear in Pauly and
Murphy's (1982) volume of collected papers from
a symposium on theory and management of trop-
ical fisheries. Most of these papers specifically
address tropical multispecies systems, but the
concepts are broadly applicable. References to
other, often less accessible, works can be found in
these two general references.

The predation method is elegant in concept but
often difficult to apply. Studies by Forney (1977)
and Stein (1981) had the distinct advantages of
limited species numbers in a small system, and
direct quantification of stomach contents. Yet

even in lake systems, the sampling problems of
estimating Z, population abundances, and so
forth, remain often as intractable as in large
marine systems. The two greatest problems are 1)
the difficulty in defining vulnerability and prefer-
ence functions for the various prey stocks (e.g.,
Ursin 1982) and 2) the need to include cohort
analyses of all the major interacting species, some
or many of which may not be available commer-
cially (and for which therefore data will be
scarce).

Despite these problems the approach can cer-
tainly generate, for stocks that suffer heavy
predatory mortality from other fished stocks,
more realistic estimates of M than approaches
that simply generate a globally fixed and invari-
ant M. More importantly (and in contrast to the
age-frequency or life history methods) the preda-
tion method has the advantage of being mecha-
nistic. Predation-related causes and conse-
quences of age, size, site, stock, geographic, or
time trends in M can be investigated via pertur-
bation and sensitivity analysis in computer simu-
lation studies or, alternatively, investigated
through analysis of existing catch data. It be-
comes possible (not necessarily feasible) to inves-
tigate the implications of varying age or abun-
dance structures of interacting fishery resources.

Thus the predation approach has considerable
conceptual appeal for fairly simple systems in
which 1 ) predation is the major force controlling
prey abundance, 2) predators have few alterna-
tive prey, 3) the possibility can be ignored that
predators prefer moribund prey which were about
to die anyway, and 4) all major species of predator
and prey are sought commercially so that data on
abundances and feeding preferences are or can be
made available.

Unfortunately, the number of systems satisfy-
ing these requirements appears to be fairly small,
and of course where predation is a relatively
small fraction of M, the multispecies predation
method will be particularly ineffective.

III. SENSITIVITY OF FISHERY MODELS
TO CHOICES FOR M

Although catch-analysis, life history, and pre-
dation methods all exist currently for estimating
M in fish stocks, in practice the only method used
extensively is the first — direct estimation of M
from analysis of catch structure. Thus the discus-
sion below of model sensitivity to M is based on
this type of estimate. The conclusions reached are

FISHERY BULLETIN: VOL. 86, NO. 1

not specific to this one method. Model sensitivity
to a given derived value of M will be the same,
regardless of the method used to derive the value.

General Patterns

Sensitivity analyses of M in fishery models
have evolved through two phases. Earlier studies
noted the influence of M on estimates of maxi-
mum yield (Fmax) or maximum yield per recruit
{(Y/R)jnax\ and on F^ax (the fishing pressure re-
quired to produce maximum yield) in Beverton-
Holt yield models (Beverton and Holt 1957; Hen-
nemuth 1961; Francis 1974; Parks 1977; Bartoo
and Coan 1979; Bulgakova and Efimov 1982).
More recently, as cohort analyses have become
more popular, more attention has been directed
toward assessing the influence of M on age-
specific estimates of stock sizes (A^, ) and fishing
mortalities (F, ) produced by these models (Mur-
phy 1965; Pope 1971; Ricker 1971; Agger et al.
1973; Doubleday 1976; Ulltang 1977; Doubleday
and Beacham 1982; Pope and Shepard 1982; Sims
1982a, 1982b, 1984). A few other studies have
investigated the effect of M on estimates of maxi-
mum sustainable yield (MSY) or total biomass
(Francis 1976; Deriso 1982; Beddington and
Cooke 1983; Tyler et al. 1985).

Most of these studies have used a single, invari-
ant value for M . Model sensitivity is then as-
sessed by comparing model results using some
"best" estimate of M, to results using one (or
rarely, more) pair(s) of M values some arbitrary
percentage above and below the best estimate.
Only a few studies exist of the effects of noncon-
stant M, where M varies in different groups of
fish within a given stock. These include Beverton
and Holt's (1957) example of density -dependent
M in plaice, and several investigations of age-
specific M (Parks 1977; Ulltang 1977; Bartoo and
Coan 1979; Sandland 1982; Bulgakova and Efi-
mov 1982; Caddy 1984; Tyler et al. 1985).

No study to date has specifically addressed the
problems of estimating values of M for a full fish-
ery analysis, leading from cohort analyses (using
M to estimate F, , Ni , and recruitment R ) to esti-
mates of yield or yield-per- recruit using the same
M(s) and R subsequently in the Beverton-Holt
formulas. Also, no study to date has addressed the
possibility and consequences of differing patterns
of variability in M , although it has been sug-
gested in one case (Ulltang 1977) that random
variations will be unimportant if the rate is con-
stant (on average) over the fished ages.

In general, the earlier analyses with yield mod-
els assuming a constant M show that higher esti-
mates of M lead to 1) lower estimates of y^ax oi"
iY/R)jnax (because fewer survive to be caught),

2) higher estimates of Fj^ax 'yo^J must fish a bit
harder to catch a given amount of those left), and

3) lower estimates of age at first capture it^.; be-
cause it pays to catch them before they die, rather
than waiting for them to grow bigger but less
abundant).

Including density-dependence tends to exag-
gerate these trends, at least for plaice in the
North Sea (Beverton and Holt 1957). Including
age-structured M in yield models also affects the
estimates, but not necessarily in a straightfor-
ward manner. As described below in the section
on numeric results, change in model output for a
given change in M depends not just on the values
chosen for M, but also on those chosen for the
other parameters. M is not an independent
parameter in these models.

Analyses with cohort or virtual population
models which assume a constant value for M
show that in general the effect of increasing M is
to increase estimates of N, (because the higher M
is, the more fish died in addition to those being
caught) and to decrease estimates of F,. The data
show only Z , which is the sum of M and F, . As-
suming Z has been constant, a decrease in F,
requires an increase in M. If Z has been variable,
the lower F, may be explained on the basis of
higher A^,, a smaller proportion of which (F,)
would account for the observed catch.

The actual effect, particularly on estimates of
A^,, is not necessarily that simple. As with yield
models, a given change in M does not always
produce the same change in model output. The
result depends also on values chosen for other
parameters; M is not an independent param-
eter.

In cohort analysis the results (estimates of A'^,
and F, ) are particularly sensitive to the relative
sizes of F and M (i.e., to the exploitation ratio
E = F/(F + M)). The effect of assuming an incor-
rect value (or series of values) of M tends to build
up as the analysis proceeds backward in time.
This is because with every time step backward
the catch (C) is inflated by the factor M in order
to estimate at that time the size of the entire
stock, not just the size of the catch. That is

A^, =A^, + i + C,(F, +M)/F,
where F, satisfies the catch equation

(1)

VETTER: NATURAL MORTALITY IN FISH STOCKS

C, =N,,i (FJF, + A/)(e'

(2)

If M is large relative to F (i.e., the exploitation
ratio is low), then errors in A^, can increase pro-
gressively and become quite large at the younger
ages (e.g., Agger et al. 1973; Murphy 1965; Ull-
tang 1977; Sims 1982a, 1982b, 1984).

Numeric Results

Although general responses of various models
can be determined simply by inspection of the
analytic models themselves, the quantitative
change to expect in the result (output) for a quan-
tified change in M (input) is not always immedi-
ately obvious. This is because M tends to occur
more than once in various formulas. For example,
M appears in both the numerator and denomina-
tor in the solution to the Beverton-Holt yield
equation (Ricker 1975).

Y = FA^oe'-^'-WxdAM + F)

-3e'-*'"V(M +F + k)

+ 3e(-2*^V(M +F + 2k)

e'-s^'-'/CM +F ^3k)).

(3)

So, rather than derive analytical expressions
(e.g., Sims 1984), I resort below to a simpler ap-
proach. Sensitivity of fishery models to changes of
given magnitude in M is assessed by comparing
percent change reported in model response (out-
put) to percent change in M (input). In cases for

vector (age or density-dependent) M, I have
merely described the shape of the M -vector. For
these different vectors, I report the percent
change in the result due to switching from a vec-
tor of one shape to a vector of another shape.

Yield Models

At least four studies (Beverton and Holt 1957;
Hennemuth 1961; Francis 1974; Bartoo and Coan
1979) have shown that errors in estimates of
M propagate into roughly equal errors in esti-
mates of (y//? )maxj but with sign reversed (Table
2). For example, a 10% overestimate in M will
lead to approximately 10% underestimate of (Y/
R 'max- An equally important result is that the
actual magnitude of the effect induced depends
strongly not just on the error in M, but on the
values chosen for the other parameters in the
model.

In another study, Beddington and Cooke (1983)
used the Beverton-Hol formulation to investi-
gate the influence of M (constant; 0.1 to 0.8
year"M, t^ (0 to 4 years), and K (the von Berta-
lanffy growth parameter; 0.1 to 0.5 year"M on
MSY (maximum sustainable yield), expressing
the result as "MSY as a % of Bq," where Bq is the
initial or recruited biomass. Higher percentages
indicate that more of the original biomass is
being taken at MSY. Increasing M by a factor of
8 (0.1 to 0.8 year-i) increased MSY/Bq by a factor
of about 4 to 8, depending on the particular values
of tf. and K. Again, errors in M produced roughly
the same relative error in the result; and again
the actual effect of any given change in M de-

Table 2— Sensitivity of estimated maximum yield per recruit ((V/Rj^ax) 'o changes in instantaneous
rate of natural mortality (M) and other input conditions. Sensitivity of (>^'W)max ^^d of changes in Ware
expressed as percentage difference from nominal responses at nominal (best-guess) M. Symbols
are: t^ = age-at-first-capture, F = instantaneous rate of fishing mortality, M = nominal value for
M. Frances (1974) used an age-structured simulation model. All other citations used standard yield-
per-recruit analyses.

Input conditions

% change % change in

in M (^/'^Jmax Species

Source

fc = constant (3.72)
F = variable
/W = 0.10

-1-50
-50

-20
-^30

fp = variable

F = constant (0.73)

M = 0.10

-1-50
-50

-60

-^50

F = constant (0.95)
M = 0.8

+ 20
-20

-21
+ 32

/W = 0.8

-HO
-10

-14
-M6

M = 0.60

+ 2b

-20

plaice

Beverton and Holt 1957

yellowfin tuna Hennemuth 1961
yellowfin tuna Francis 1974
yellov\rfin tuna Bartoo and Coan 1979

FISHERY BULLETIN: VOL. 86, NO. 1

pended on the values chosen for the other
parameters.

Pope and Garrod (1973) present another exam-
ple of sensitivity in MSY to values chosen for M.
They describe briefly the consequences of using
an incorrect constant for M of cod stocks when
estimating the F required to generate MSY
(Fmsy*- Underestimating M by 507r (assumed
M = 0.1 year^^; true M = 0.2 year"M leads to a
choice of Fmsy that is 67% too high. Overestimat-
ing M by 50% (assumed M = 0.3 year ^ true
M - 0.2 year"M underestimated Fmsy by 50%.

The simulations described above tested the ef-
fects of choosing alternative constant values for
M. Choosing a vector alternative can also have
significant effects; again, the magnitude of the
effect depends on the values chosen for other
parameters. Beverton and Holt (1957) showed
that incorporating density-dependence in M for
plaice decreased (y/i?)max by 12%, when holding
tc constant at 3.72 years and letting F vary. Con-
versely, holding F constant and letting t^, vary
decreased {Y/R )max by about 37%.

Age-dependent values for M were compared
with age-constant values by Bartoo and Coan
(1979), Bulgakova and Efimov (1982), and Tyler
et al. (1985). In their analysis of Atlantic yel-
lowfin tuna stocks, Bartoo and Coan found that
replacing an assumed constant M of 0.8 year"^
with an age-structured M increasing from 0.1
year~^ at age 0 to 1.2 year~^ at age 7, increased
(Y/R )max by 17% (from 6 to 7 kg).

Estimating total yield (7,) rather than {Y/R )
and estimating R as a function of constant versus
age-specific M in analysis of catch curves for rela-
tively unexploited stocks of Pacific ocean perch
and Oregon hake, Bulgakova and Efimov (1982)
found that replacing a constant (age-averaged) M
with age-variable M tended to increase estimated
Yi when fish recruited fairly late to the fishery,
but decreased Yf if the fish recruited early. This is
because of the interaction between the values as-
sumed for M (constant or age-variable) and the
value calculated for R from each type of mortality
curve.

Starting with a given value for recruitment at
age 6 years (from Efimov 1976), they calculated R
twice for ages 4 and 8 years — once with age-
averaged M and once with age-specific M. Be-
cause in this set of data the age-averaged M was
generally higher than the age-specific M at the
tested ages of recruitment (ages 4, 6, or 8 years),
back-calculations with age-averaged (i.e., con-
stant) M predicted fewer recruits than back-

calculations with age-specific M. With fewer re-
cruits and generally higher M , potential yield at
later ages obviously must drop. Differences in
predicted potential yield ranged from about
-30% at ^4 (age-specific estimate lower than age-
averaged estimate, when fish were assumed to
recruit to the fishery at age 4 years) to +15% at
^6 (age-specific estimate higher) and to +60% at

Tyler et al. (1985) tested (among other things)
the effects of ignoring "true" age-structure in M
and using instead a constant value in estimating
stock biomass using Deriso's (1980) delay-
difference model. They did the tests on catch data
generated by Walter's (1969) age-structured sim-
ulation model of cod, using three different (input)
age structures for M in Walter's model. After gen-
erating "catch data" from Walter's model, they
analyzed the simulated data set using Deriso's
model with constant M (= 0.5 year~^). The age
structures tested were 1) mortality increasing
and then decreasing with age (Walter's original
mortality vector spanning ages 3 to 12 years; age-
averaged M = 0.55 year~\ range = 0.33 to 0.70
year~^), 2) mortality increasing with age (ages 7
to 12 years; average M = 0.5 year~^, range 0.3 to
0.7 year"^) and 3) mortality decreasing with age
(ages 7 to 12 years; average M = 0.5 year"\
range 0.7 to 0.3 year M. In all three cases Deriso's
model with constant M misestimated the "true"
biomass generated by Walter's model (with age-
structured values for M). The differences were
relatively small, however: —13% for the increas-
ing and then decreasing series, +19% for the de-
creasing series, and +4% for the increasing
series. These differences were due to the differ-
ences in M, and not the differences in model
structure; generating and analyzing biomass
with the same constant M in both models led to a
discrepancy of only 0.5%.

By analogy to life history patterns in other
adult animals, M (after recruitment into most
fished stocks) is more likely to increase with age
than to cycle or decrease. By implication, the sim-
ulation results from the increasing series are
probably most realistic. If so, the effects of ignor-
ing age-structure in favor of using a constant M
may be relatively small (5 to 20%), at least for the
cod stock simulated in this study. But the results
obviously depend again not just on correctly
choosing the values for M, but on the values cho-
sen for the other parameters. In this case, Tyler et
al.'s (1985) results imply that age-structure in M
can be relatively unimportant, at least when the

VETTER: NATURAL MORTALITY IN FISH STOCKS

assumed constant is evenly bracketed by the
"true" age-structure in M .

Further simulations by Tyler et al. (1985) using
a wide range of constant values for M (0.4 to 1.4
year M and the growth rate parameter rho
(mean Ford growth coefficient for the fishable
stock; 0.46 to 1.6) showed that incorrect guesses of
M (and rho) could produce errors up to 1,000^^ in
estimated biomass. More realistic ranges for the
two parameters (0.4 to 0.8 year" ^ forM, 0.6 to 1:2
for rho), extending about 509r above and below
the "true" values for these parameters, induced
much lower error in biomass estimates (about the
same order of magnitude, 50 to 100% below and
above the "true" biomass). As before, changes
(errors) of a given amount in M (expressed as
fraction or percentage of the original value) ap-
pear to produce about the same amount of change
(expressed as percent of original value) in simple
estimates of yield, depending on the conditions of
other parameters in the model.

Chatwin (1958) compared estimates of Yj^ax
from lingcod populations. Rather than compare
constant and age-variable values for M, he as-
sumed several different values for an average
(constant) M in adults, but assumed that M in-
creased from the assumed average for adults to
higher values in both juveniles and senescent
fish. He reports no quantitative results but states,
as found above, that increasing the average M,
for a given F, considerably decreased l^max' that
decreasing M increased Yrna-x' snd that size at
first capture changed relatively little with those
changes in M.

These comparisons between age-structured
versus constant M, or between different constants
have demonstrated that effects on results can be
large for some combinations of parameters yet
small for others. Alternative choices drawn from
apparently realistic parameter values lead to rel-
atively small differences in estimates of M.
Specific amounts of change depend strongly not
only on the values chosen for M, but also on the
value of M relative to values chosen for the other
interacting parameters in the yield models. For
most choices of parameter values, sensitivity of
output is roughly equal to perturbation of input.

Cohort Analyses

Effects of interactions between changes in M
and values chosen for other parameters is even
more obvious in stock reconstruction analyses
(e.g., cohort analysis and virtual population anal-

ysis (VPA)). These analyses are used to "recon-
struct" estimates of stock abundance during pre-
vious years, based on catch data and assumptions
about the value(s) of M during those previous
years. Studies of sensitivity to M in Beverton-
Holt types of yield or biomass assessments were
usually empirical, based on analyses of catch data
from specific fisheries. Studies of sensitivity to M
in VPA and cohort analysis include both theoret-
ical and empirical studies; i.e., simulations using
totally contrived data sets (e.g.. Agger et al.
1973), analyses of specific data sets (e.g.. Pope
1971; Doubleday and Beacham 1982) and combi-
nations of analytical evaluations and analysis of
specific data sets (e.g., Doubleday 1976; Ulltang
1977; Sims 1982a, 1982b, 1984).

Simple analyses of sensitivity to M, in which M
is varied but all else is held constant, include
1) Pope's (1972) analysis of Atlantic yellowfin
tuna, in which he found that replacing constant
M with age-structured M (higher Ms for older
fish) produced lower estimates for fishing mortal-
ity {F, ) in the later ages, but had little effect on
estimates for the younger ages, and 2) Doubleday
and Beacham's (1982) statement that I07c error
in constant M translated into 9 to 149f error in
estimates of i? (at age 3) for cod in the Gulf of St.
Lawrence.

Somewhat more complicated analyses are pre-
sented by Ulltang ( 1977 ) and Sims ( 1982a, 1982b,
1984). Ulltang evaluated the effects on model pre-
dictions of F, and A'^, , of several types of variation
in M. These included no variation (uniformly con-
stant M), M constant within years but varying
randomly between years, M varying with age,
and M varying with season. Sims evaluated the
effects of choosing various constants for M on esti-
mates of A'^,, and derived an analytical expression
relating variance in M to expected variance in
estimates of abundance.

In Ulltang's simulations, increasing (decreas-
ing) a constant M by 50*^ (from 0.2) decreased
(increased) F by about 207c ("true" F's ranging
from 0.4 to 0.8). Creating a data set with M vary-
ing randomly from one year to the next, then an-
alyzing those data with an assumed constant M,
Ulltang (1977) found that the Z calculated from
the constant-M model was on average the same as
the "true" Z from the random-M model. He con-
cluded that random fluctuations in M will cancel
out during analysis and so can be ignored. Ull-
tang assessed the influence of age-dependent M
compared with constant M by generating a catch
curve with age-variable M (decreasing curvilin-

FISHERY BULLETIN: VOL 86, NO 1

early from 0.3 at age 1 to 0.1 at age 10, average
about 0.2) and F equal to 0.2, then analyzing the
catch with F equal to 0.2 or 0.6, and M equal
either to 0.1 or 0.2. Choice of M made little differ-
ence in estimates of stock size for the case of high
F (0.6), because most of the deaths were due to
(observed) fishing. When F was low (0.2), stock-
size estimates were much more sensitive to incor-
rect choices for M , because most of the deaths
were due in this case to M, which was unmea-
sured and therefore unobserved.

Ulltang (1977) simulated seasonal changes in
M by concentrating all deaths in either the first or
last quarter of a year. Estimated stock sizes (N, )
changed relatively little; with F = 1.2 and
M - 0.4, A'^, was a maximum of 10% higher if all
deaths occurred first quarter, 10% lower if all oc-
curred in the last quarter.

A serious problem with the conclusions reached
by Ulltang (1977) is also common to all the other
studies discussed above; they are based on rela-
tively few combinations of values for the various
parameters, and relatively few simulations. For
example, the conclusion that random errors in M
will tend to even out is intuitively attractive, pro-
vided the time scale of variation is short relative
to the generation time of the fish. In fact random
variation in M did even out in the two sets of
simulations he conducted. But the examples he
chose included only one set of ages (2 to 10 years),
with relatively high values of F (0.5 to 0.8 year"^)
compared to the values tested for M (0.1, 0.3
year"M. The gravity of consequences from choos-
ing an incorrect M depends very heavily on the
size of M relative to the size of F, i.e., on E. Had
he chosen different values for his simulations, he
might have reached very different conclusions.
This is probably the basis for the discrepancy be-
tween Ulltang's conclusion that seasonal effects
are minor, versus Sims' (1984) conclusion that
seasonal effects can be quite large, if M is high.

Sims (1984) attempted to overcome this prob-
lem (trying to draw general conclusions from the
results of simulations based on particular, or rel-
atively few, sets of parameters) by analytically
deriving formulas for relative error in stock-size
estimates, and then testing the formulas with
data from actual fisheries. He used this approach
twice: once to assess the effects of seasonality
(Sims 1982a) and once to consider in general the
effects of different choices (errors) for constant M
(Sims 1984). But his results (and equations for
error) show clearly that error in estimated stock
sizes depends on several parameters and that the

effects of one can be strongly dependent on the
values chosen for the others. Choosing a high M
(0.6 year~M and concentrating catch during the
first quarter of the year overestimated R by 20%;
concentrating catch during the last quarter
underestimated R by 23% (compared with the
10% error found by Ulltang).

Within the same analysis, reducing M by half
(to 0.3 year^ ^) reduced the error in R by half, but
the same reduction of error inR was also achieved
by leaving M high and reducing F . In assessing
specifically the effects of error in M on error in R ,
Sims (1984) showed very different effects on esti-
mates of/? in heavily fished versus lightly fished
cohorts of Atlantic bluefin tuna. Changing M by
50% led to changes in estimated R of 60 to 260%
in the lightly fished cohort, but only to relatively
smaller changes of 35 to 70% in the heavily fished
cohort. Again, the magnitude of the error in
model predictions depended not just on the mag-
nitude of M, but on its relationship to the other
parameters in the catch equation, particularly F.

Errors (expressed as percentage change in out-
put for a given change in input) in model output
in the simulations described above, all of which
tried to use apparently realistic values for model
parameters, rarely exceeded 50%, and were often
less than the error introduced into values chosen
for M . By implication, the effects of incorrectly
guessing M may be relatively unimportant if M is
relatively small (e.g., in this situation not more
than about 0.5 year"M and relatively invariant,
although the actual magnitude of effect due to
any given percentage change in M depends on the
values chosen for other parameters.

So, inaccurate estimates of M might be impor-
tant or they might not. It all depends on the mag-
nitude and variability of M within a given stock
(or group). Although untested, it seems likely
that estimates of M for groups in which M varies
little and is relatively low, are more likely to be
reasonably accurate than estimates of M from
groups in which M is large and variable. The fol-
lowing section reviews evidence that M does in
fact vary both within and between groups of fish,
and the succeeding section reviews evidence for
the magnitude of that variability in ostensibly
similar groups.

IV. FACTORS INFLUENCING DEATH
RATE

Despite the fact that in most fishery models, M
is assumed to be constant for all exploited ages in

VETTER; NATURAL MORTALITY IN FISH STOCKS

any given stock, abundant evidence exists to the
contrary. Natural mortality has been shown to
vary with age, density, disease, parasites, food
supply, predator abundance, water temperature,
fishing pressure, sex, and size. Evidence for rela-
tionships between these factors and M, and se-
lected references for each, are presented below.

Changes in mortality rate with age, within sin-
gle groups of fish, have been demonstrated and
discussed more frequently than changes with any
other factor. References include, among others,
Baranov (1918, plaice), Sette (1943, Atlantic
mackerel), Ricker (1945, 1947, lake fish; 1969,
1975, various species), Beverton and Holt (1959,
many species of marine fish), Beverton (1963, en-
graulids and clupeids), Boiko (1964, sturgeon).
Gushing (1975, plaice), Blinov (1977, fish in gen-
eral), Bulgakova and Efimov (1982, Oregon hake
and sea perch), Sandland (1982, fish in general).
Smith (1985, clupeoids), Roff (1986, fish in gen-
eral). Evidence for changes with senescence for
fish in general has been discussed or documented
by, among others, Woodhead (1979) and Craig
(1984).

Although specific patterns vary with species
(e.g., Woodhead 1979), in general M is extremely
high during egg and larval stages (e.g., 2 to 10%
per day in plaice and clupeoids (Cushing 1975;
Smith 1985)), falls precipitously during the juve-
nile period, becomes relatively stable during in-
termediate adult ages and increases again with
senescence. But even during these relatively sta-
ble mid-adult ages, changes in M with age can be
substantial, particularly in short-lived fish (e.g.,
Ricker 1947, stunted versus "normal" whitefish).

Changes in natural mortality rate with size
(rather than age) within single groups of fish
(usually stocks), have been discussed by Baranov
(1918, plaice), Ricker (1969, size-selective mortal-
ity in general). Ware (1975, larval fish), and
Peterson and Wroblewski (1984, many species).
Differences in natural mortality rate between
populations of the same species in different envi-
ronments, or even in different areas of a single
environment (e.g., a single lake) are documented
by Ricker (1947), Kennedy (1954), and Schupp
(1978). Year-to-year differences in natural mor-
tality rates of single stocks from a given area are
shown by Pope and Knights (1982, plaice) and by
Henderson et al. (1983, whitefish). Density-
dependent changes in M are discussed by Bever-
ton and Holt (1957), Cushing (1967), Tyler and
Gallucci (1980), Backiel and LeCren (1978), Jones
(1982), and others. Differences in M between

sexes have been documented by Beverton and
Holt (1957, plaice), Ricker (1947, rock bass), and
others. Changes in natural mortality rate related
to the cost of reproduction have been discussed
by Jones and Johnston (1977), Roff (1984), and
others.

Other factors that affect M either alone or in
combination with other factors include disease
and parasitism (reviewed by Lester 1984), starva-
tion (Hewitt et al. 1985; Theilacker 1986: larval
anchovy), physiological state (Smith 1985), and
fishing pressure (Ursin 1982; Munro 1982). Addi-
tional examples are cited by Beverton and Holt
(1957), Anderson and Ursin (1977), Sissenwine
(1984), and Hunter (1984).

Most of the factors listed above (e.g., age, size,
sex) are indirect influences on M . The most im-
portant factor directly affecting natural mortality
rate is probably predation; this is implied by a
large body of literature describing changes in
prey community composition and abundance fol-
lowing changes in composition and abundance of
predators (e.g., Carpenter et al. 1985).

Direct evidence that predators account for most
natural mortality in fish stocks is difficult to
gather (Section II). To quantify the fraction of M
due to predation, one must know, not only rela-
tive changes in abundance, but absolute popula-
tion density of all predators and prey together
with consumption rates and prey preferences of
all the predators. Although this is rarely possible,
at least two studies from freshwater systems do
present quantified estimates of predatory mortal-
ity in relation to available prey. Forney (1977)
quantified predation mortality in a relatively
simple, unmanipulated lake system where there
were few species of predator and prey. Combining
stomach-content estimates of prey consumed with
trawl-sample estimates of predator and prey
abundance, he concluded that 30 to 100% of yel-
low perch production was consumed by walleye,
their principal predator. In a manipulated sys-
tem. Stein et al. (1981) assessed predatory mor-
tality of young tiger muskellunge after they were
stocked in a small pond and lake. During the time
of the study, a single predator (largemouth bass)
accounted for 25 to 45% of losses to natural mor-
tality.

In marine systems evidence for the relative im-
portance of predation can be gleaned from com-
paring total natural mortality with estimated
predatory mortality based on abundance of preda-
tors and feeding preferences. For example, multi-
species cohort analyses reported by Pope and

FISHERY BULLETIN: VOL. 86, NO. I

Knights ( 1982) show predatory mortality as 80 to
909c of M for age-0 cod, whiting, and haddock in
the North Sea (the fraction of M due to predatory
mortahty cannot be assessed accurately in the
older ages because predators appropriate to these
sizes were not included in the analysis). In an-
other example, estimating M from energy flow
models, Sissenwine (1984) demonstrated that
predation in the Georges Bank ecosystem can ac-
count for all production by prey fish; nonpreda-
tory mortality was negligible.

Thus a multitude of factors, acting alone or in
concert, can be expected to produce variations in
M between individuals within single groups of
fish, as well as between groups. Differences can
be expected between species, between stocks
within species, and from place-to-place and time-
to-time within given stocks. In the following sec-
tion, I will review more completely existing evi-
dence for, and the extent of, this expected
variability in M .

V. VARIABILITY WITHIN AND
BETWEEN GROUPS

As discussed above (Section III), simulation
studies generally show that effects of choosing a
particular value or set of values for M can range
from insignificant to considerable, depending in
part on the model used, in part on the values
chosen for other parameters, and in part on the
form chosen for the estimate(s) of M. Authors
suggest that in the future, simulations should be
conducted with a range of values for M, to bracket
probable values (e.g., Beverton and Holt 1957;
Tyler et al. 1985).

The problem with this advice is identifying the
appropriate range and distribution of M for any
given group offish. Obviously, wide ranges for M
will lead to great discrepancies between model
predictions based on one end of the range or the
other. It has been shown above, however, that
model output can be relatively insensitive to
small changes in M . This is particularly true if F
is much larger than M (i.e., if the stock is highly
exploited so that losses to fishing far exceed losses
to natural mortality). The problem is determining
whether, for a given stock in situ, changes in M
are in fact large or small. Compensatory changes
in M, in response to changes in F, will further
confound the problem, because variations in M
will then be a function of the value(s) of F, in
addition to the suite of other factors that may be
affecting estimates of M .

M does appear to vary considerably between
groups offish. Estimates of M compiled by Pauly
(1980) (Fig. 1) for 175 stocks and species offish
worldwide differ greatly between groups, ranging
from a minimum of about 0.1 year"^ to several
unusual values as high as 7.0 year ^ Even
within a group as ostensibly homogeneous as the
tunas, the range of estimated mortality constants
spans the majority of the common values (0.2 to
2.0 year"i. Murphy and Sakagawa 1977).

Estimates of variability in M within groups of
fish are much less common, but are actually more
important than the obvious differences between
groups with obviously different characteristics
such as differing lifespans. Most fishery analyses
are directed toward understanding or predicting
dynamics of single stocks (single groups offish).
The most important considerations for natural
mortality parameter values in these single-
species analyses are whether and if so over what
values M varies for the group of fish in question.

But measuring trends or variability in natural
mortality rates withingiven groups (e.g., stocks)
is difficult and, with the exception of trends with
age, rarely attempted. This is primarily because
the only extant methods for estimating M depend
either directly or indirectly on analysis of catch
data (Section II), and catch data are prone to
many well known (but largely unsolved) prob-
lems.

Problems with analysis of catch data fall into
two general categories: 1) problems with sam-
pling procedure, such that fish are caught or
counted out of proportion to their true abundance
and 2) problems with fish appearing or disappear-
ing from the "unit stock" due to causes other than
birth or natural mortality (i.e., migration, fishing
mortality, or tagging mortality), again resulting
in catch data that do not represent the true struc-
ture of the stock. If sampling biases can be over-
come, the problem reduces to partitioning total
disappearance offish into fractions owing to fish-
ing, tag mortality, and migration. The first parti-
tion can be eliminated by studying unfished popu-
lations, the second by quantifying tag mortality,
and the third by studying only closed or tagged
populations.

Unfortunately, very few sampled populations
satisfy completely even one of these criteria. Re-
gardless, we still need at least some crude esti-
mates of M in order to determine whether M truly
varies enough to invalidate the standard assump-
tion in fisheries models that M is effectively con-
stant during exploited ages. The question here

VETTER NATURAL MORTALITY IN FISH STOCKS

UJ
3
O
UJ
QC
U.

28
26-

24-
22-

20-

18-

16-

14-

10-

NATURAL MORTALITY ESTIMATES: FROM PAULY (1980)

(FISH STOCKS)

nni

[,i i"i'"('|"i'i'i I I I I'i'i I I I I I I I'l'i'i'i'n I I I iTi I I I I I I I I I I I I I I I I I I iMi I I I'l'i I
0.5 1.0 1.6 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0

ESTIMATED NATURAL MORTALITY (M)

Figure 1. — Frequency of estimated instantaneous annual rates of natural mortality (M) in 175 different fish stocks, populations, or
species. Estimates include both freshwater and marine species. Data from Pauly 1980.

concerns variability of M within groupings that
would be used commonly to estimate M , such as
stocks of single species, rather than general pat-
terns across species. If M truly varies relatively
little during these ages (so that the log of catch-
at-age decreases linearly with age), if the age
structure has been relatively constant histori-
cally (so that catch curves are actually linear,
rather than curvilinear as seen in stocks with
inconstant age structure, e.g., chapter 2, Ricker
1975), and if catch curves actually reflect rela-
tively accurately this constancy and low variabil-
ity, then most estimates of M derived from analy-
sis of appropriately processed catch curves cannot
help but be relatively close to the true rate. Model
predictions, although in theory sensitive, would
in practice be fairly robust to any particular value
chosen from the true range of values for M .

Despite the potential problems with accuracy
or precision of existing estimates from single
groups of fish, I list in Table 3 most of the esti-
mates available for unexploited populations, and

some of the few existing estimates from exploited
populations. My purpose is to identify the appar-
ent range of variability in M within single stocks.
The estimates are drawn from references cited by
Pauly (1980) and other sources. Only references
that reported multiple estimates for M are in-
cluded, thus excluding most of the references re-
viewed. Because these estimates are derived from
catch data, the stated ranges are "apparent",
rather than demonstrably the "true" values.

Estimated rates of natural mortality are not
particularly constant for either unexploited or ex-
ploited groups, and are only slightly less variable
within stocks than they are within species. Al-
though the range of rates within groups may ap-
pear relatively small compared to the total range
of rates reported for all fish species (e.g., 0.36 to
0.56 for sauger from Lake Nipigon [Table 3] vs.
approximately 0.1 to 3.0 for most species listed
by Pauly 1980), the maximum and minimum
rates reported for single groups differed by at
least 50% in 20 of the 22 comparison listed in

FISHERY BULLETIN: VOL. 86, NO. 1

Table 3. In at least one case maximum and mini-
mum estimates differ by as much as a factor of 7
(i.e., young vs. old whitefish in Shakespeare Is-
land Lake, Ricker 1947, Table 3).

The range of reported estimates of M for species
(rather than single groups or stocks within a spe-

cies as compared above) is even greater. Even the
least variable estimates differed by a factor of
1.75 (75%, male vs. female plaice, Beverton 1964).
In whitefish, the species for which the most esti-
mates exist, maximum estimates are 20 times
greater than minimum estimates (Table 3).

Table 3. — Ranges in estimates of instantaneous rate of natural mortality in unexploited and exploited fisfi populations. 'Wr^ax'''^min is
expressed as ffie ratio between tfie maximum (Mmax) and minimum (/Wmm) values reported for M for that species, Values in parenttieses
are total range of estimates and ratios for those species where multiple reports exist.

Common

Age

Species name

name

Body of water

Sex

(years)

M range

'Wmax^'^mir

, Source

1) Unexploited populations:

Amboplites rupestris

rock bass

Nebish Lake

10-12

1.47-2.1

1.49

Ricker 1947

Nebish Lake

10-14

1.1-1.6

1.45

Ricker 1947

Nebish Lake

both

10-14

1.08-1.56
(1.08-2.1)

1.44
(2.01)

Ricker 1947

Stizostedion canadensis

sauger

Lake Nipigon

8-14

0.36-0.56

1.56

Ricker 1947

Coregonus clupeaformls

whitefish

Lake Opeongo

6-13

0.53

Ricker 1947

Shakespeare Island Lake

11-27

0.08-0.60

7.51

Ricker 1947

Great Slave Lake

17-22 .

0.71-0.99

1.39

Kennedy 1953

Lake Nueltin

13-15

0.84

Kennedy 1963

Lake IVIcDonald

11-14
9-10

1.34

1.66

(0.08-1.66)

(20.75)

Kennedy 1963
Kennedy 1963

Leuachthys sardinella

Ikroavik Lake

6-10

0.2-1.4

7.00

Wohlschlag 1954

Cnstovomer namayacush

Great Slave Lake

1-26

0.31-1.61

5.19

'Kennedy 1954

Great Slave Lake

15-23

0.49-0.92

1.88

'Kennedy 1954

Great Slave Lake

15-23

0.52-0.75
(0.31-1.61)

1.44
(5.19)

2Kennedy 1954

Perca fluviatilus

perch

River Thames

3-8

0.56-0.98

1.75

Williams 1967

River Thames

3-7

0.2-0.64

3.20

3Williams 1967

River Thames

juv

3-5

0.53-1.69
(0.2-1.69)

3.19
(8.45)

3Williams 1967

Leuciscus leuciscus

dace

River Thames

5-11

0.36-1.31

3.64

"Williams 1967

Alburnus alburnus

bleak

River Thames

3-8

0.6-2.4

4.00

■•Williams 1967

Rutilus rutilus

roach

River Thames
River Stour

2-11
3-12

0.22-1.38

0.44
(0.22-1.38)

6.27
(6.27)

"Williams 1967
Mann 1973

Cheilodactylus macropterus

tarakihi

Chatham Islands
Chatham Islands
New Zealand

5-35
5-22

0.03

0.08

0.15

(0.03-0.15)

(5.00)

Vooren 1977
Vooren 1977
Vooren 1977

II) Exploited Populations:

Pleuronectes platessa

plaice

North Sea

5-13

0.08

Beverton 1964

North Sea

5-13

0.14
(0.08-0.14)

(1.75)

Beverton 1964

Brevooiiia petronus

gulf menhaden

Gulf of fvlexico,

1-3

0.7-1.6

2.29

5Ahrenholz 1981

Central (1969, 1971)

Gulf of fvlexico,

1-3

0.88-0.98

1.11

SAhrenholz 1981

Eastern (1969,

1971)

Gulf of IVIexico,

1-3

1.17-1.23

1.05

SAhrenholz 1981

Western (1969,

1971)

Gulf of IVIexico,

1-3

0.95-1.2

1.26

SAhrenholz 1981

all areas (1969,

1971)

(0.7-1.6)

(2.29)

Gadus morhua

cod

North Sea

0.5-1

0.59-1.46

2.47

7Pope and
Knights 1982

Coregonus clupeaformls

whitefish

Lake Huron

3.8

0.34-1.67

4.91

SHenderson et al.
1983

'Increasing with age.

2Year to year variation (1946-52); ages 15-23 combined.

3Not consistent with age.

"Generally increasing with age.

SAssuming 20% tag loss rate,
spor tag loss rates from 10 to 30%.
78 different years (1967-75).
823 different year classes (1947-75).

VETTER: NATURAL MORTALITY IN FISH STOCKS

As discussed previously, these different esti-
mates can lead to at least as great a difference in
results produced by fishery analyses such as yield
models or stock reconstruction analyses (Sec-
tion II).

Reported differences in estimates of M for
whitefish stocks living in Shakespeare Lake com-
pared with other relatively small lakes (e.g., Lake
McDonald, Table 3) are particularly significant.
Because both stocks are of the same species and
living in more or less similar environments
(small lakes), one might easily (and incorrectly)
assume that both have the same rate of natural
mortality; but they did not. The lower rates oc-
curred in the stock occupying a small lake with no
predators. This is a clear example of the effect
that environment, particularly the predator envi-
ronment, can have on the realized rate of natural
mortality in a fish stock. Obviously, choosing a
single appropriate constant for this species would
be difficult. Choosing an appropriate species-
specific constant for some of the other species
with multiple estimates might be difficult as well
(e.g., rock bass, lake trout, perch, roach, tarakihi,
or menhaden. Table 3).

None of these studies from either unexploited
or exploited stocks support the assumption that M
is constant for any given stock or species, nor are
these within-stock ranges particularly narrow. In
addition, treatment of the original catch data
may have in some cases obscured the "true" vari-
ability. Ricker (1947) and Kennedy (1953, 1954,
1963), for example, use a 3-yr smoothing tech-
nique to reduce the effects of unequal recruit-
ment; this also serves to reduce variability that
may actually be due to differences in natural mor-
tality. Also, single estimates from data collected
during only one or two years of sampling (e.g.,
Wohlschlag 1954; Williams 1967; Mann 1973;
Vooren 1977) can be seriously biased by annual
changes in either recruitment or mortality rates.
If the estimates reported above are even approxi-
mately accurate, it is apparent that the range of
possible values for M is wide, and that variability
can be considerable even within single stocks.

A solution to this problem of choosing a reason-
able value for M, at least for long-lived fish, is
suggested by the possibility that variation in M
(not just the mean value) may be related to max-
imum lifespan. Fish that live for many years
must naturally have lower mortality rates than
more short-lived fish. These lower rates may also
be less variable in the longer lived stocks, if as in
many other biological processes, variability is

proportional to the mean. This could account for
the ubiquity and apparent effectiveness of the
constant 0.2 year"\ used almost universally for
the long-lived (20 to 30 years) and well-studied
fish stocks from northern European seas (e.g.,
Beverton 1964). If so, assuming a constant M
might be valid for these longer lived stocks.

Unfortunately, the few studies cited above do
not support this attractive idea. Although in gen-
eral, mortality rates decease as lifespan in-
creases, the variability in estimates does not ap-
pear to follow the same trend. This may be due
partially to the relatively similar lifespans (10 to
20 years) for most of the species for which esti-
mates exist. But the apparent range in rates for
the shortest lived species cited above (Ahrenholz
1981, Brevoortia patronus, ages 1 to 3 years, M
range 0.7 to 1.6 year"^) is certainly not greater
than ranges reported for the longer lived white-
fish (Henderson et al. 1983, Coregonus clu-
peaformis, ages 10+ years, M range 0.34 to 1.67
year"^).

VI. SUMMARY AND
RECOMMENDATIONS

Thus it appears that rates of M, or at least rates
of M derived by existing estimation methods, do
in fact vary widely within many fish stocks. Be-
cause the variations appear to be considerable
and because the results from fishery models can
be sensitive to large variations in M , one must
conclude that assuming constancy without proof
can have serious consequences for fishery man-
agement.

A better approach may be to discard the notion
that a single "best" estimate of M can be found,
and instead try to tailor estimates of M to local
groups, based on some combinations of the meth-
ods discussed in Section III. Obviously, practical
considerations of time and resources will limit the
accuracy and precision with which M can be esti-
mated. Also, the estimates in the studies re-
viewed here are prone to all the artifacts men-
tioned in the previous sections. True rates of
natural mortality, and their variability, are still
very poorly known for even the great stocks of
commercial fish in temperature regions that have
been subject to continuous exploitation for
decades. Careful, repeated tagging experiments
probably hold the most promise for determining
with any reasonable degree of accuracy, rates of
natural mortality in fish stocks. But even these
have inherent problems that are not easily

FISHERY BULLETIN: VOL 86, NO. 1

solved. There remains a great need both for new
methods, and refinements of the old.

ACKNOWLEDGMENTS

This review was initiated in response to a re-
quest for Deus ex machina in mortality estima-
tion for fish stocks, elicited from participants at
the 35th Annual Tuna Conference held at Lake
Arrowhead, CA, 20-22 May 1984, during a discus-
sion organized in an attempt to identify the single
most important but least well-estimated parame-
ter in fishery models. The author apologizes for
the lack of divine inspiration herein, but extends
grateful thanks to Chris Boggs, Andy Dizon, John
Hoenig, Pierre Kleiber, Robert Olsen, and two
anonymous reviewers for their insightful discus-
sion and helpful comments during preparation of
this review.

LITERATURE CITED

Agger, P , I Boetius, and H Lassen

1973. Errors in the virtual population analysis: The
efiects of uncertainties in the natural mortality
coefficients. J. Cons. Int. Explor. Mer 35:98.
Ahrenholz, D W

1981. Recruitment and exploitation of gulf menhaden,
Brevoortia patronus. Fish. Bull., U.S. 79:325-335.
Alverson, D L , AND M J Carney

1975. A graphic review of the growth and decay of
population cohorts. J. Cons. Int. Explor. Mer 36:133-
143.
Anderson, K P , and E Ursin

1977. A multispecies extension to the Beverton and Holt
theory of fishing, with accounts of phosphorus circulation
and primary production. Medd. Dan. Fisk. Havunders,
7:319-435.

BaCKIEL, T , AND E D. LECREN.

1978. Some density relationships for fish population
parameters. In S. Gerking (editor), Ecology of
freshwater fish production, p. 279-302. Blackwell
Scientific, Oxf.

Barlow. J

1984. Mortality estimation: biased results from unbiased
ages. Can. J. Fish. Aquat. Sci. 41:1843-1847.
Baranov. F I

1918. On the question of the biological basis of
fisheries. Reports of the Department of Fisheries and
Scientific and Industrial Investigations, Vol I, Part 1.
(Engl, translat.; Library, Southwest Fisheries Center,
P.O. Box 271, La Jolla, CA.)

BARTOO, N W , AND A L COAN

1979. Changes in yield recruit of yellowfin tuna Thunnus
albacares under the ICCAT minimum size
regulation. ICCAT Coll. Vol. Sci. Pap. VIII:120-138,

BEDDINGTON, J R . AND J G COOKE.

1983. The potential yield offish stocks. FAO Fish. Tech.
Rep. 242, 47 p.
Beverton, R J H

1963. Maturation, growth and mortality of clupeid and

engraulid stocks in relation to fishing. Rapp. P. -v. Cons,
int. Explor. Mer 154-44-6J.
1964. Differential catchability of male and female plaice
in the North Sea and its effect on estimates of
abundance. Cons. Perm. Int. Explor. Mer Rapp. et
P.-v. 155, p. 103-112.
Beverton, R J H , and S J Holt

1957. On the dynamics of exploited fish popu-
lations. Fish. Invest. Minist. Agric, Fish. Food (G. B.),
Ser. II, 19:21-42.

1959. A review of the lifespan and mortality rates of fish
in nature and their relation to growth and other
physiological characteristics. In CIBA Foundation.
Colloquium on aging V. The lifespan of animals, p.
142-180. Churchill, Lond.
Bishop, Y.

1959. Errors in estimates of mortality obtained from
virtual populations. J. Fish. Res. Board Can. 16:73-90.
Blinov, V V

1977. The connection between natural mortality and re-
production of food fishes and modeling of the processes
involved. J. Ichthyol. 16:1033-1036.
BoiKO, E G

1964. Evaluating natural mortality in the Azov Zander
iStizostedion lucioperca ). Fish. Res. Board Can. Transl.
Ser. No. 541, 15 p.

Brownie. C , D. R Anderson, K Burnham, and D. S. Robson.
1985. Statistical inference from band recovery data - a
handbook. 2d ed. U.S. Fish Wildl. Serv. Resour. Publ.
# 156.

BULGAKOVA. T I. AND Y N EFIMOV.

1982. A method for forecasting catch by consideration of
the dependence of natural mortality on age. J. Ichthyol.
22:24-32.
Burnham, K., and D. R Andersen.

1984. Tests of compensatory vs. additive hypotheses of
mortality in mallards. Ecology 65:105-112.
BuRHAM, K , G C White, and D R Anderson

1984. Estimating the effect of hunting on annual survival
rate of adult mallards. J. Wildl. Manage. 48:350-361.
Butler, S A , and L. L MacDonald.

1979. A simulation study of a catch-effort model for esti-
mating mortality rates. Trans. Am. Fish. Soc. 108:353-
357.
Caddy, J F

1984. Method of solving for natural mortality rate from
stocks with different schedules of growth and mortal-
ity. Can. J. Fish. Aquat. Sci. 41:1226-1230.

Carpenter, S R , J F. Kitchell. and J R Hodgson

1985. Cascading trophic interactions and lake productiv-
ity. Bioscience 35:634-639.

Chapman, D G , and G I Murphy

1965. Estimates of mortality and population from survey-
removal records. Biometrics. 21:921-935.

Chatwin, B M

1958. Mortality rates and estimates of theoretical yield in
relation to minimum commercial size of ling cod (.Ophid-
ian elongatus ) from the Strait of Georgia, British Colum-
bia. J. Fish. Res. Board Can. 15:831-849.

Craig. J F

1984. Aging in fish. Can. J. Zool. 63:1-8.
Gushing, D. H

1967. The possible density-dependence of larval mortality
and adult mortality in fishes. In J. H. S. Blaxter (edi-
tor), The early life history of fishes, p. 103-
111. Springer- Verlag, N.Y,

VETTER: NATURAL MORTALITY IN FISH STOCKS

1975. The natural mortality of the plaice. J. Cons. Int.
Explor. Mer 36:150-157.

Deriso. R B

1980. Harvesting strategies and parameter estimation for
an age-structured model. Can. J. Fish. Aquat. Sci.
37:268-282.
1982. Relationship of fishing mortality to natural mortal-
ity and growth at the level of maximum sustainable
yield. Can. J. Fish Aquat. Sci. 39:1054-1058.
DOUBLEDAY. W G

1976. A least squares approach to analyzing catch at age
data. ICNAF Res. Bull. No. 12, p. 69-81.

DOUBLEDAY. W G , AND T D BEACHAM

1982. Southern Gulf of St. Lawrence cod: a review of

multispecies models and management advice. In M. C.

Mercer (editor), Multispecies approaches to fisheries

management advice, p. 133-140. Can. Spec. Publ. Fish.

Aquat. Sci. No. 59,
Efimov, Y N

1976. Assessment of the magnitude of the reserves of
Pacific hake by the virtual population method. [In
Russ.) In Abstracts of All-Union Conference of Young
Scientists, Fisheries Exploitation in the Resources of the
Worlds Oceans.

Forney. J L

1977. Reconstruction of yellow perch iPerca flavescens)
coherts from examination of walleye iStizostedwn vit-
reum vitreum ) stomachs. J. Fish Res. Board Can.
34:925-932.

FOURNIER, D . AND C P ARCHIBALD

1982. A general theory for analyzing catch at age
data. Can. J. Fish. Aquat. Sci. 39:1195-1207.

Francis. R C

1974. TUNPOP. a computer simulation model of the yel-
lowfin tuna population and the surface tuna fishery of the
eastern Pacific Ocean. Inter-Am. Trop. Tuna Comm.,
Bull. 16:234-258.

1976. Relationship of fishing mortality to natural mortal-
ity at the level of maximum sustainable yield under the
logistic stock production model. J. Fish. Res. Board
Can. 31:1539-1542.

GULLAND. J A

1983. Fish stock assessment. A manual of basic meth-
ods. FAOAViley series on food agriculture. Vol 1. John
Wiley and Sons, N.Y., 223 p.

Gunderson, D R

1980. Using R-K selection theory to predict natural mor-
tality. Can. J. Fish. Aquat. Sci. 37:2266-2271.
Heincke, F

1913. Investigations on the Plaice. General report 1. The
plaice fishery and protective regulations. Rapp. Cons.
Explor. Mer 17A, p. 1-153.
Henderson. B A . J Collins, and J. A Reckahn

1983. Dynamics of an exploited population of whitefish
(Coregonus dupeaformis I in Lake Huron. Can. J. Fish.
Aquat. Sci. 40:1556-1567.

Hennemuth.R. C.

1961. Year class abundance, mortality and yield-per-
recruit of yellowfin tuna in the eastern Pacific Ocean,
1954-1959. Inter-Am. Trop. Tuna Comm., Bull. 6:1-51.
Hewitt. R . G. H Theilacker. and N C H Lo

1985. Causes of mortality in young jack mackerel. Mar.
Ecol. Prog. Ser. 26:1-10.
Hochbaum. G S . AND C J Walters.

1984. Components of hunting mortality in ducks: a man-
agement analysis. Can. Wildl. Serv, Occas. Pap. # 52.

Hoenig, J

1983. Empirical use of longevity data to estimate mortal-
ity rates. Fish. Bull., U.S. 82:898-903.

Hunter, J R

1984. Inferences regarding predators in the early life
stages of cod and other fishes. In E. Dahl, D. S.
Danielsen, E. Moksness, and P. Solemdal (editors), The
propagation of cod Gadus morhua . L. Flodevigen Rapp.
1:533-562.

Jensen, A L

1984. Non-linear catch curves resulting from variation in
mortality among subpopulations. J. Cons. Int. Explor.
Mer 41:121-124.
Jones. R

1979. Materials and methods used in marking experi-
ments in fishery research. FAO Fish. Tech. Pap. No.
190, 134 p.
1982. Ecosystems, food chains, and fish yield. In D.
Pauly and G. Murphy (editors). Theory and management
of tropical fisheries, p. 195-240. ICLARM Conf Proc. 9.
Jones. R . and C Johnston

1977. Growth, reproduction, and mortality in gadoid fish
species. In J. H. Steele (editor). Fisheries mathemat-
ics, p. 37-62. Acad. Press, N.Y.
Kennedy, W A

1953. Growth, maturity, fecundity and instability in un-
exploited whitefish iCoregonus dupeaformis) of Great
Slave Lake. J. Fish. Res. Board Can. 10:413-441.

1954. Growth, maturity, and mortality in the relatively
unexploited lake trout, (Cristivomer namaychush),
of Great Slave Lake. J. Fish. Res. Board Can. 11:827-
852.

1963. Growth and mortality of whitefish in three unex-
ploited lakes in northern Canada. J. Fish. Res. Board
Can. 20:265-272.
Laevastu. T , F Favorite. H A Larkin.

1982. Resource assessment and evaluation of the dynam-
ics of the fisheries resources in the northeastern Pacific
in numerical ecosystem models. In M. C. Mercer (edi-
tor), Multispecies approaches to fisheries management
advice, p. 70-81. Can. Spec. Publ. Fish. Aquat. Sci. 59.
Lander. R H

1962. A method of estimating mortality rates from
changes in composition. J. Fish. Res. Board Can.
19:159-168.
Larkin, P. A , and W Gazey.

1982. Applications of ecological simulation models to
management of tropical multispecies fisheries. In D.
Pauly and G. Murphy (editors). Theory and management
of tropical multispecies fisheries, p. 123-140. ICLARM
Conf Proc. 9.

Lester. R J G

1984. A review of methods for estimating mortality due to

parasites in wild fish populations. Helgol. Meersunters.

37:53-64.
MA.JK0WSKI, J.

1981. Application of a multispecies approach for assessing
the population abundance and age-structure of fish
stocks. Can. J. Fish. Aquat. Sci. 38:421-431.

Mann. R H K

1973. Observations on the age, growth, reproduction, and

food of the roach, Rutilus rutilus (L.), in two rivers in

England. J. Fish. Biol. 5:707-736.
Mercer, M C

1982. Multispecies approaches to fisheries management
advice. Can. Spec. Publ. Fish. Aquat. Sci. 59, 169 p.

FISHERY BULLETIN; VOL. 86, NO. 1

MUNRO, J L

1982. Estimation of biological and fishery parameters in
coral reef fisheries. In D. Pauly and G. I. Murphy (edi-
tors), Theory and management of tropical multispecies
fisheries, p. 71-83. ICLARM Conf Proc. 9.

Murphy, G. I.

1965. A solution to the catch curve. J. Fish. Res. Board
Can. 22:1965.

Murphy. T. C, and G T. Sakagawa

1977. A review and evolution of estimates of natural mor-
tality rates of tunas. In ICCAT Coll. Vol. Sci. Pap.
VI(SCRS-1976):117-123.

Myers, R A., and Doyle, R W.

1983. Predicting natural mortality rates and
reproduction-mortality trade offs from life-history
data. Can. J. Fish. Aquat. Sci. 40:612-620.

Paloheimo, J E

1961. Studies on estimation of mortalities I. Comparison
of a method described by Beverton and Holt and new
linear formula. J. Fish. Res. Board Can. 18:645-662.
1980. Estimation of mortality rates in fish popula-
tions. Trans. Am. Fish. Soc. 109:378-386.
1982. Estimation offish mortality. Trans Am. Fish. Soc.
111:535-537.
Paloheimo, J E , and L M Dickie.

1966. Food and growth of fishes III. Relations among food,
body size, and growth efficiency. J. Fish. Res. Board.
Can. 23:1209-1247.

Parks, W W

1977. Cohort and equilibrium yield-per-recruit analyses
for the Atlantic bluefin tuna fisheries system, accounting
for two system configurations and two natural mortality
models. ICCAT Coll. Vol. Sci. Pap. VI:385-401.

PAULIK, J , AND D S ROBSON.

1969. Statistical calculations for change-in-ratio estima-
tors of population parameters. J. Wildl. Manage. 33:1-
27.
Pauly, D

1980. On the interrelationships between natural mortal-
ity, growth parameters, and mean environmental tem-
perature in 175 fish stocks. J. Cons. Int. Explor. Mer
39(2):175-192.
1982. Studying single-species dynamics in a tropical mul-
tispecies context. In D. Pauly and G. Murphy (editors).
Theory and management of tropical fisheries, p. 33-
70. ICLARM Conf Proc. 9.
Pauly, D., and G. Murphy (editors).

1982. Theory and management of tropical fish-
eries. ICLARM Conf Proc. 9, 360 p. International
Center for Living Aquatic Resources Management,
Manilla, Phillipines.
PELLA, J J . AND P K TOMLINSON.

1969. A generalized stock-production model. Inter-Am.
Trop. Tuna. Comm., Bull. 13:416-496.
Peterson. J , and J S Wroblewski

1984. Mortality rate of fishes in the pelagic ecosys-
tem. Can. J. Fish. Aquat. Sci. 41:1117-1120.

Pope. J G

1971. Note on cohort analysis and age-specific fishing
mortality. ICCAT Tuna Meeting Working Paper, 2 p.

1972. An investigation of the accuracy of virtual popula-
tion analysis using cohort analysis. ICNAF Res. Bull. 9,
p. 65-74.

Pope, J G , and D J Garrod.

1973. A contribution to the discussion of the effect of error
on the action of catch quotas and effort quotas. Ann.

Meet. Inst. Comm. Northwest Atl. C. Fish. 1973, Res.
Doc. 119, Serial #3074.
Pope, J. G , and B J Knights.

1982. Simple models of predation in multispecies fisheries
for considering the estimation of fishery mortality and its
effects. In M. C. Mercer (editor), Multispecies ap-
proaches to fisheries management advice, p. 64-
69. Can. Spec. Publ. Fish. Aquat. Sci. 59.
Pope, J. G., and J, G Shepard

1982. A single method for the consistent interpretation of
catch-at-age data. J. Cons. Int. Explor. Mer 40:176-
184.
Reed, R. B., and W D Davies

1980. Estimating rate of exploitation from tag returns
and fishing effort. Proc. Conf Southeast. Assoc. Fish.
Wildl Agencies 34:151-157.
RICKER, W. E

1945. Natural mortality among Indiana bluegill sun-
fish. Ecology 26:111-122.

1947. Mortality rates in some little-exploited populations
of fresh-water fishes. Trans. Am. Fish. Soc. 71:114-128.

1969. Effects of size-selective mortality and sampling bias
on estimates of growth, mortality, production, and
yield. J. Fish. Res. Board Can. 26:479-541.

1971. Derzhavin's biostatistical method of population
analysis. J. Fish. Res. Board Can. 28:1666-1672.

1975. Computation and interpretation of biological statis-
tics offish populations. Bull. Fish Res. Board Can. 191,
382 p.

1977. The historical development, /n J. A. Gulland (edi-
tor). Fish population dynamics, p. 1-26. Wiley-
Interscience, Lond.

RoBSON. D S., AND D G Chapman

1961. Catch curves and mortality rates. Trans. Am.
Fish. Soc. 90:181-189.
ROFF, D A

1984. The evolution of life history parameters in teleosts.

Can. J. Fish. Aquat. Sci. 41:989-1000.
1986. Predicting body size with life history model.
Bioscience 6:316-323.
Sandland, R. L.

1982. Estimation, inference, and data analysis for log-
linear models in tagging studies. J. Cons. Cien. 40:290-
303.
Schaefer, M B

1954. Some aspects of the dynamics of populations impor-
tant to the management of commercial marine fish-
eries. Inter-Am. Trop. Tuna. Comm., Bull. l(2):27-56.
Schupp. D

1978. Walleye abundance, growth, movement and yield in
disparate environments in a Minnesota lake. Am. Fish.
Soc. Spec. Pub. 11:58-65.

Seber, G a F.

1973. The estimation of animal abundance and related
parameters. Griffin, Lond., 506 p.
Sette, O E

1943. Biology of the Atlantic mackerel (Scomber scom-
brus) of North America. U.S. Fish. Wildl. Serv., Fish.
Bull. 50:149-237.
Silliman, R p.

1943. Studies on the Pacific pilchard, or sardine
iSardinops caerula ) 5. A method of computing mortali-
ties and replacements. U.S. Fish Wildl. Serv., Spec. Sci.
Rep. 24, 10 p.
Sims. S. E

1982a. The effect of unevenly distributed catches on stock

VETTER: NATURAL MORTALITY IN FISH STOCKS

size estimates using virtual population analysis (cohort
analysis). J. Cons. Int. Explor. Mer 40:47-52.

1982b. Algorithms for solving the catch equation forward
and backward in time. Can. J. Fish. Aquat. Sci. 39:197-
202.

1984. An analysis of the effect of errors in the natural
mortality rate on stock-size estimates using virtual popu-
lation analysis (cohort analysis). J. Cons. Int. Explor.
Mer 41:149-153.
SISSENWINE. M P

1984. Why do fish populations vary? In R. M. May (edi-
tor). Exploitation of marme communities, p. 59-94.
Springer-Verlag, N.Y.

Smith. P E

1985. Year-class strength and survival of 0-group clu-
peoids. Can, J. Fish. Aquat. Sci. 42:69-82.

Stein. R A , R F Carline. and R S Hayward.

1981. Largemouth bass (Micropterus salmoides) preda-
tion on stocked tiger muskellunge. Trans. Am. Fish.
Soc. 110:604-612.
Taylor, C C

1958. Natural mortality rate of Georges Bank had-
dock. U.S. Fish Wildl. Serv., Fish Bull, 58:1-7,
Theilacker. G H

1986. Starvation-induced mortality of young sea-caught
jack mackerel, Trachurus synimetricus, determined with
histological and morphological methods. Fish, Bull,,
U,S, 84:1-17,

Tyler, A V . and V Gallucci

1980. Dynamics of fished stocks. In R. T. Lackey and

L. A. Nielsen (editors). Fisheries management, p. 111-

149, John Wiley & Sons, NY,

Tyler, A V , L L Sebring, M C Murphy, and L F. Murphy.

1985. A sensitivity analysis of Deriso's delay-difference

equation using simulation. Can. J. Fish. Aquat. Sci.

42:836-841.
Ulltang. O

1977. Sources of errors in and limitations of virtual popu-
lation analysis (cohort analysis). J. Cons. Int. Explor.
Mer 37:249-260.
Ursin, E

1967, A mathematical model of some aspects of fish
growth, respiration and mortality, J, Fish, Res, Board
Can, 24:2355-2452,
1982, Multispecies fish stock and yield assessment in
ICES, In M, C, Mercer (editor), Multispecies approaches
to fisheries management advice, p, 39-47, Can, Spec.
Publ, Fish, Aquat, Sci, 59,
VOOREN.C M

1977, Growth and mortality of tarahiki (Pisces: Cheilo-
dactylidae) in lightly exploited populations. N.Z. J.
Mar. Freshwater Res. ll(l):l-22.
Walters, C J

1969, A generalized computer model for fish population
studies. Trans, Am, Fish, Soc, 98:505-512,
Ware, D M.

1975. Relation between egg size, growth and natural mor-
tality of larval fish. J. Fish. Res. Board Can. 32:2503-
2512,
Williams, W P

1967, Growth and mortality of 4 species of fish in the
Rivers Thames and Reading. J, Anim, Ecol, 266:695-
720,
Wohlschlag, D E

1954, Mortality rates of whitefish in an Arctic
lake. Ecology 35:388-396,
Woodhead, a D

1979, Senescence in fish. In P, J, Miller (editor), Fish
phenology: Anabolic adaptations in teleosts, p, 179-
205, Symp, Zool, Soc, Lond, 44,

NATURAL HISTORY OF THE RAYS OF THE GENUS MOBULA

IN THE GULF OF CALIFORNIA

Giuseppe Notarbartolo-di-Sciara1

ABSTRACT

Mobulid rays, which abound during summer in the southern Gulf of California, southern Baja
California, Mexico, were monitored for a period of four years during a study of their biology. A total
of 262 specimens belonging to four species of Mobula were examined. Mobula thurstoni was the most
abundant l58'/f of the catch), followed by M. japanica (30%), M. munkiana (9'5f ), and M. tarapacana
(3'^). The study area served as a nursery ground for M. thurstoni, a summer feeding and mating
ground for M. thurstoni and M. japanica, and a wintering ground for M. munkiana and young
M. thurstoni ; M. tarapacana was rare. Data on size, weight, sex ratio, life history, seasonality, feeding
habits, behavior, habitat, and symbionts are presented for each species. Size segregation was a
common feature of M. thurstoni , M. japanica , and M. munkiana ; sex segregation was not evident. An
extreme degree of feeding specialization was noted: summer prey were almost exclusively the eu-
phausiid Nyctiphanes simplex: the mysid Mysidium sp. dominated in the winter. A key to the genus
Mobula in the Gulf of California is presented as an aid for species identification.

This paper reports on natural history aspects of
rays of the genus Mobula (Mobulidae), a poorly
known group of elasmobranchs commonly called
manta rays or devil rays, frequent in the Gulf of
California. A good early overview of the family
Mobulidae was given by Gill (1908). Cadenat
(1960) described the natural history of the mobu-
lids of tropical west African waters, based on
specimens which were occasionally captured by
the local fishermen. However, with the exception
of observations carried out with some regularity
by Coles (1910, 1913, 1915, 1916a, 1916b) of Mo6-
ula olfersi (= M. hypostoma ) and Manta birostris
off North Carolina, most of the available litera-
ture is purely anecdotal and deals with occasion-
ally encountered or harpooned specimens. Long-
term field investigations of devil ray ecology and
behavior are wanting. As a result, mobulids are
among the least known of the batoid taxa. This
was recognized by Bigelow and Schroeder (1953)
in their comprehensive review of the knowledge
of this family. No major contribution to the
understanding of any aspect of mobulid biology
has since been published.

Regular fisheries for mobulids were not known
to exist, because mobulid meat is generally con-
sidered of little market value. However, in 1981,
during a reconnaissance trip to the southern Gulf

iScripps Institution of Oceanogi'aphy, University of Califor-
nia, San Diego, La Jolla, CA 92093; present address: Museo
Civico di Storia Naturale, corso Venezia 55, 20121 Milano,
Italy.

Manu.script accepted October 1987.
FISHERY BULLETIN: VOL. 86, NO. 1. 1988.

of California (Mexico), in the vicinity of La Paz,
Baja California Sur, a regular, seasonally impor-
tant fishery was discovered. This activity afforded
the opportunity to study several aspects of the
natural history and the ecology of these batoids.
Preliminary oral interviews revealed that local
fishermen in the Gulf of California knew of, and
routinely captured, four species of devil rays, in
addition to the well-known giant manta ray,
Manta birostris. This information contrasted
with the scientific literature, where only two
mobulid species, Manta birostris and Mobula lu-
casana, were reported for the area (Beebe and
Tee-Van 1941; Fowler 1944; Castro-Aguirre
1965). The confusing state of mobulid taxonomy
demanded a revisionary work of the genus Mob-
ula (Notarbartolo-di-Sciara 1987), and a discus-
sion of the systematics of Manta in the eastern
Pacific (Notarbartolo-di-Sciara in press). Such ef-
fort permitted designation of names for all species
of Mobula found in the Gulf of California: M.
thurstoni (Lloyd 1908), of which M. lucasana
Beebe and Tee Van (1938) is a junior synonym;
M. japanica (Miiller and Henle 1841); M. tara-
pacana (Philippi 1892); and M. munkiana
Notarbartolo-di-Sciara (1987), which had not
been described before. Many of the reports of
M. lucasana { = M. thurstoni ) from Central and
South America (Beebe and Tee- Van 1941; Fowler
1944; Nichols and Murphy 1944; Barton 1948;
Castro-Aguirre 1965; Chirichigno 1974; Pequeno
1983) undoubtedly refer to other species of Mob-
ula . A key to the genus Mobula in the Gulf of

F1SHP:RY BULLETIN: VOL 86. NO 1

D
O

NOTARBARTOLO-DI-SCIARA: NATURAL HISTORY OF MOBULA

California is presented as an aid to future studies
of mobulids from this region.

Working relationships were established with
the local fishing communities, and their activities
were intermittently monitored between 1981 and
1984. Captured rays were examined and mea-
sured before their pectoral fins were filleted;
stomach contents and reproductive organs were
examined later. Information was gathered on
size, weight, sex ratios, life history, seasonality,
feeding habits, habitat, behavior, and symbionts
of four species of rays belonging to the genus Mob-
ula (M. thurstoni, M. japanica, M. munkiana,
and M. tarapacana ). Detailed descriptions and
morphometries of those species are given by
Notarbartolo-di-Sciara (1987). The manta ray,
Mania birostris, was also occasionally captured
(Notarbartolo-di-Sciara in press), but is not
treated in the present study.

METHODS

Although mobulids are locally said to be abun-
dant on both sides of the southern Gulf of Califor-
nia, for logistic convenience collecting trips were
made only to the peninsular coast (Fig. 1). The
fishing cooperative based at Punta Arena de la
Ventana was selected as the prime collecting site,
because mobulids were caught there more consis-
tently than at other localities. Fishing camps on
Isla El Pardito, at Cueva de Leon, Ensenada de
los Muertos, and Bahia de los Frailes were also
sources of study material. Other fishing commu-
nities, such as Juncalito, San Evaristo, El Sar-
gento. La Ventana, and San Jose del Cabo were
occasionally visited, but yielded no data because
mobulids were not specifically sought by the fish-
ermen. Seven field trips were made. Six were
short-term (24 January-8 February 1981, 25
November 1981, 16-21 December 1981, 20-23
December 1982, 19-26 January 1984, and 28 Oc-
tober-1 November 1984); one lasted almost six
months (26 January-15 July 1983).

Mobulids of all available species and sizes are
caught with nets and harpoon; their meat is fil-
leted out of the pectoral fins for human consump-
tion and used as shark bait. Gill nets are either
strung just under the surface or are set on the
bottom perpendicular to shore, usually at depths
between 10 and 200 m. Fishing vessels were 5-7
m fiberglass launches, locally called "pangas",
powered by an outboard engine. Fishing occurred
within a radius of about 15 km from a base camp.
Nets are checked once a day, early in the morn-

ing. Rays weighing up to approximately 100 kg
were hauled on board, larger specimens were
towed ashore. Rays that were dead in the nets,
after several hours (i.e., three unsexed specimens
of Mobula thurstoni), were often partially de-
stroyed by gammarid amphipods (locally called
"plaga"), and were unmarketable.

Specimens were weighed and measured before
being processed by the fishermen. Weights (WT)
were taken with calibrated spring-scales. Rays
lighter than 20 kg were weighed to the nearest
pound with a 50-lb scale; weights were subse-
quently converted to kg. Heavier rays were
weighed to the nearest kg with a 150 kg scale.
Specimens which exceeded 150 kg (all postnatal
M. tarapacana ) were cut in four pieces and
weighed separately. Ten percent was then added
to the total weight to compensate for body fluid
loss. All the specimens could not be weighed, as
occasionally a large number of rays were beached
simultaneously, and because of the intense heat
the fishermen could not delay their processing.

A set of 29 measurements was taken for mor-
phometric analysis and systematic purposes.
Methods and results are presented in
Notarbartolo-di-Sciara (1987). Measurements
relevant to the present paper were disc width
(DW), greatest dimension between outermost tips
of pectoral fins, pelvic fin length, from anterior
margin of vent to tip of pelvic, and clasper length,
from anterior margin of vent to tip of clasper.
Most specimens were discarded after measuring
and sampling. All preserved specimens were
deposited in the Marine Vertebrate Collection of
the Scripps Institution of Oceanography. Raw
data listing all specimens examined and pre-
served can be found in Notarbartolo-di-Sciara
(1985).

The size and shape of the testes were inspected
in male specimens, and the ducti deferentes were
cut slightly above the genital papilla. Presence or
absence of seminal fluid was determined by
running a finger in the caudal direction over the
ducts anterior to the cut. Clasper length in
thousandths of disc width (DW) was plotted
against DW to determine size at maturity, and
the presence or absence of seminal fluid was
noted. Relative size and contents of uteri and
nidamental glands were examined in female spec-
imens, and right and left ovaries were compared.
The diameter of the largest ovum was plotted
against DW to determine size at maturity of fe-
male Mobula. Eggs were extracted from the
germinative epithelium and their greatest di-

KISHKRY [BULLETIN: VOL 86, NO 1

ameter was measured to the nearest 0.1 mm with
a steel dial caliper.

Stomach contents, if not larger than approxi-
mately 200 cc, were sampled whole; otherwise,
the bolus was made homogeneous by stirring, and
about 200 cc were preserved. Stomach content
samples were fixed and preserved in 10% buffered
formalin. The stomach content of each sample
was thoroughly agitated, separated with a plank-
ton strainer (mesh size 0.5 mm), rinsed of forma-
lin in deionized water, and blotted for 30 seconds
on blotting paper. The lump was then molded into
a cylindrical shape, and a portion of one end was
separated to make up 1 g of wet weight, measured
to the nearest 0.1 g. The subsample was then
placed with water in a gridded tray, and exam-
ined under a dissecting microscope. Contents of
the spiral intestine were discarded, because the
small crustacean prey was rapidly digested.

Feeding habits were analyzed quantitatively
by computing the Index of Relative Importance
(%IRI) (Pinkas et al. 1971; Hyslop 1980) for each
prey species. The IRI combines percentage by
number iN), mass (M), and frequency of occur-
rence (F) in the formula:

mi = (%N + %M) X %F

Prey items were identified, when possible, to
lowest taxa or species, then the %N of all prey
species within each subsample was calculated.
When more than one species was present, all
items were individually counted. To obtain the
%M term of the equation, mean mass was calcu-
lated for each species by measuring the length of
each item contained in five randomly selected
squares on a tray, calculating the average length
of each prey species, and obtaining mass values
from Miller's (1966) Plankton Conversion Tables,
where mass is related to length for all main
planktonic taxa. The %N from all subsamples
were summed, and the percent from the new sum
was calculated, to calculate %IRI for each prey
species. The same procedure was applied to %M
and 7cF. The total 7(N was then added to the total
7cM, and that sum was multiplied by the total
%F, to obtain total IRI for each prey species, from
which the %IRI was calculated. When few items
(e.g., copepods) were found among a large amount
of partially digested euphausiid or mysid
shrimps, the possibility of reconstructing the
shrimp number within the subsample by count-
ing the digestion-resistant eye pairs was dis-
carded to avoid bias in favor of the shrimp frac-

tion. The following method was adopted instead:
all odd prey items were counted and measured,
and their total mass was obtained from Miller's
tables; this was subtracted from the total weight
(1 g) of the subsample. The remaining weight was
divided by the mean weight of each individual
item, calculated by averaging the lengths of all
available intact specimens, and obtaining the cor-
responding weight in Miller's tables. A potential
biasing factor existed, when only a few prey re-
mains were found (e.g., when a relatively uncom-
mon item occurred alone in a stomach, therefore
contributing a value of 100 7(N and 7<M to the
total IRI). This factor was avoided by considering,
for quantitative treatment, only those stomachs
which contained more than 1 g (wet weight) of
recognizable food. All stomachs that had <1 g of
contents were in fact virtually empty, and the few
items found in them were treated only qualita-
tively.

Remoras were fixed in 10% formalin, preserved
in 50% isopropanol, and deposited in the Marine
Vertebrate Collection, Scripps Institution of
Oceanography. Parasitic copepods were fixed and
preserved in 50% isopropanol and sent to the
Long Beach State University for taxonomic iden-
tification and study.

RESULTS AND DISCUSSION

A total of 262 mobulid rays, belonging to four
species iMobula thurstoni, M. japanica, M.
miinkiana, and M. tarapocana) were examined
between 1981 and 1984. Of these, M. thurstom
was the most abundant species, constituting 58%
of the total catch, followed by M. japanica (30%),
M. munkiana (9%), and M. tarapacana (3%).

KEY TO THE SPECIES OF MORULA IN
THE GULF OF CALIFORNIA

Mobula can be distinguished from Manta by the
mouth on the lower surface of the head rather
than being subterminal, and by the presence of
toothbands in both jaws. Moreover, Manta grows
to a greater size, and the size of its head, relative
to the body, is much greater than in Mobula.

la. Branchial filter plates fused

M. tarapacana

(Spiracle in an elongated longitudinal
slit, dorsal to plane of pectoral fins. Teeth
tessellated, surface of crown pitted, buc-
cal edge comblike. Thick cover of acumi-

NOTARBARTOLO-DI-SCIARA: NATURAL HISTORY OF MOHri.A

nate denticles. Large adult size, often
exceeding 3 m in width. Dorsal side
greenish brown. Ventral side anteriorly
white, posteriorly gray).
lb. Branchial filter plates separate 2

2a. Spine on base of tail M. japanica

(Tail very long with a line of white tuber-
cles on both sides. Spiracle a short
transversal slit, dorsal to plane of pec-
toral fins. Teeth not in contact with each
other, at least twice as high as the crown
is wide; crown subtriangular in apical
view. Medium-large adult size, approach-
ing 2.5 m in width. Dark blue to black on
dorsal side, apex of dorsal fin white; ven-
tral side white. Skin rough to the touch).

2b. No spine on base of tail 3

(Spiracle small, subcircular, ventral to
plane of pectoral fins. Skin smooth to the
touch).

3a. Base of tail dorsally depressed

M. thurstoni

(Double curvature of anterior margin of
pectoral fin. Surface of tooth crown
rugose. Medium adult size, exceeding 1.8
m of width. Dark blue to black on dorsal
side, apex of dorsal fin white; ventral side
white, with a dark greenish patch near
the posterior margin of each pectoral fin,
and with a pattern of dark and shiny sil-
very pigmentation on distal half of pec-
torals).

3b. Base of tail laterally compressed

M. munkiana

(Anterior margin of pectoral fin straight
to weakly convex. Surface of tooth crown
smooth. Small adult size, barely exceed-
ing 1 m of width. Dorsum mauve gray;
ventral side white, tips of pectoral fins
gray).

Mobula thurstoni (Lloyd 1908)
Local name: cubana de lomo azul

Eighty one males (disc width range 630-1,770
mm), 69 females (210-1,801 mm), and three spec-
imens of undetermined sex (941-1,494 mm), were
caught at four stations (Punta Arena de la Ven-
tana. Cueva de Leon, Ensenada de los Muertos,
and Isla El Pardito) and their adjacent waters,
between 7 February 1981 and 30 October 1984.
Overall and seasonal size-frequency distributions

for M. thurstoni are given in Figure 2. Mean size
varied with season, smaller rays being dominant
in winter catches, medium sizes prevailing in the
summer. The difference between mean disc width
(DW) in winter and overall mean DW is highly
significant (T-value = 5.189, df=169, P
<< 0.001). There were no significant differences
between male and female DWs (T-value
= 0.3767, df = 145, P > 0.5), with the exception of
the November to February period, when females
were larger ( T-value = 2.331, df = 12, P < 0.05).
A total of 105 specimens (210-1,770 mm DW)
were weighed. The WT/DW relationship is best
described by the equation:

WT - 4.817 X 10-8 ^Y)^^)2.^8
r = 0.99

WT is given in kg, DW in mm. The largest speci-
men in the sample was a female; DW was not
measured because the fishermen had already
started filleting the pectoral fins. Calculated DW,
regressed from disc length, cranial width, and
upper toothband length, was 1,801 mm (multiple
correlation coefficient = 0.99). The second largest
specimen was also a female, 1,799 mm DW. The
largest male had a DW of 1,770 mm and weighed
53 kg. The smallest freshly caught specimen was
876 mm DW and weighed 6.4 kg. The smallest
postnatal specimen was a male, the carcass of
which was found drying on the beach in Ensenada
de los Muertos. Its calculated DW, regressed from
toothbands length, was 864 mm (multiple correla-
tion coefficient = 0.99).

Overall ratio of males to females caught was
1.18 {N = 148). Catch sex ratios varied with sea-
son. Females appeared to be dominant in winter
(ratio of males to females 0.27; A'^ = 14). The re-
verse was true in March, in favor of males. A
significant difference from a 1:1 ratio (chi square
testP > 0.05) was not noted. Geographical segre-
gation, either of sex or size, was not apparent for
M. thurstoni during the warmer months when a
wide array of size classes and both sexes were
found in the same fishing area. Males and fe-
males were occasionally harpooned from the same
group basking at the surface. This fact argues
against behavioral sex segregation. Winter data,
however, were suggestive of size segregation at
that time of year. It was common knowledge
among local fishermen that during the winter
months all M. thurstoni caught are small. The
bimodal size-frequency distribution for early

FISHERY BULLETIN: VOL. 86, NO. 1

OVERALL
Xtot= 13456 ±4022 (N=I48)

V,f

13449+5513 {N-79)

30 -] H% - I356.3±59.27(N = 66)

20-

10-

q:
u.

lij

V)
CD

30nx,<„:

20-

10-

'•%%

MARCH -APRIL

1211 1±I52 30 (N = I5)
II28.81I6I 34(N = I0)
I346.0±370.37(N = 4)

£22.

rr-irT^rrr-iEin. i .h^.

20-

10-

JUNE

Xtot= 1401 22+39.45 (N=76)
X^j.:l38l42±55 22(N^43)
Xjj:l42703±55 I0(N = 33)

NOVEMBER -FEBRUARY

Xtot = 95969±42 75(N-I6)
Xj.jt = 86733±48l(N-3)
X^j =987 46±52 26(N = II)

D TOTAL
E3 ??

□73. 1

MAY

Xtot= 1435 4 + 69 87 (N = 37)
Xj.^ = 1436.61 107.98 (N: 20)
Xj^ = l433.9±87.27(N=l7)

r^m. , .mm

lA.

JULY

X,0T = 15I275±341.72(N = 4)
X^^ = 1409 67 ± 385.40 ( N = 3)

:|80I(N=I)

950 1150 1350 1550 1750 950 1150 1350 1550 1750

DISC WIDTH (mm)

Figure 2. — Size-frequency distributions of Mobula Ihurstoiii (means ±2 SE).

spring (Fig. 2) suggested that larger rays began to
move into the area in the spring from their un-
known wintering grounds.

Mature testes are large, elongated structures
attached by the mesorchia to the upper anterior
wall of the pleuroperitoneal cavity, on either side
of the vertebral column. A large epigonal organ is
associated with each testis. Both testes appeared
to be functional. They were usually about the
same size, although occasionally the left testis
was nearly 25''f larger. The paired ducti defer-
entes remain separated from each other through-

out their length and open into the cloaca at the tip
of the urogenital papilla through two distinct
pores, rather than merging in the urogenital
sinus, as in most elasmobranchs (Daniel 1934),
including Manta ehrcnhcrgi (Gohar and Bayoumi
1959). Clasper length was plotted against DW for
43 M. thurstoni (Fig. 3), to determine the onset of
male maturation. Rapid increase in relative size
of the claspers, beginning at a DW of about 1,500
mm, was concomitant with the incipient presence
of abundant seminal fluid in the lower portion of
each ductus deferens. The pelvic fin area, and

NOTARBARTOLO-ni-SCIARA NATl'RAl. HISTORY OV MORULA

especially the tissue at the bases of claspers, of
the larger males appeared swollen and congested
in May and June, and some of the skin had as-
sumed a pink coloration. Similar observations in
carcharhinid and odontaspid sharks have been
linked with mating activities (Springer 1960;
Gilmore et al. 1983).

The ovaries in M. thursfoni are paired, elon-
gated organs located inside the pleuroperitoneal
cavity, analogously to the testes, and are con-
nected to a large epigonal organ. Eggs are pro-
duced within the germinative epithelium. The
largest eggs were found at the anterior end of the
ovary. Only the left ovary develops and is func-
tional, whereas the size and aspect of the right
ovary remains comparable with those of the im-
mature stage. Asymmetry is also present in the
oviducts, the left uterus being usually the largest
in mature females. It consists of a voluminous,
thick-walled expansion of the lower tract of the
oviduct; its lumen is lined with a highly devel-

oped epithelium consisting of elongated, flattened
villi (trophonemata), a well-known mobulid (Gill
1908; Setna and Sarangdhar 1950; Wourms 1977)
and rhinopterid (Schwartz 1966; Smith and Mer-
riner 1986) feature. In several instances both
uteri were found to contain a viscous, whitish or
greenish substance. Oviducts open separately
into the cloaca. A progression of sexual maturity
in female mobulids was evident from the exam-
ined ovary's developmental condition. In the im-
mature female the germinative epithelium is a
narrow, leaf-shaped band, tapering at both ends,
located opposite to the mesovarium (facing the
center of the cavity) along the ovary's longitudi-
nal axis. In mature females the germinative ep-
ithelium takes over most of the ovary's ventral
side, making room for the mature ova. To deter-
mine the size at maturity of female M. thurstoni,
the diameter of the largest ovarian egg was plot-
ted against DW (19 specimens. Fig. 4). An egg
growth plateau was not evident, because data on

Figure 3. — Relationship between clasper size and body size
in Mobula thurstoni.

I 40

g^30

_i -^
o

UJ "
Q. £ 20
OT ^

< s

i 10

• ••mtn«l fluid pr»««nl
O ••minal fluid aba«nl

o ca>

500 1000 1500

DISC WIDTH (mm)

•

••

•• •

•

••

•

•
•

•

••

•

•
1

]

Figure 4. — Relationship between size of largest ovum and
body size in Mobula thurstoni.

1000 1500

DISC WIDTH (mm)

2000

FISHKKY BULI.KTIN VOL 86. NO 1

the larger sizes were insufficient, and therefore
the maximum egg diameter was not l^^nown. It
appears from the scanty available data that fe-
male M. thurstoni began to mature at a DW of
about 1,500 mm (the point at which the slope of
the curve becomes steeper).

Mobulo thurstoni, like all mobulids, is a vivi-
parous matrotroph ( Wourms 1981 ), the near-term
embryo being three orders of magnitude larger
than the mature egg. Uteri and nidamental
glands of 68 females were inspected. No adujt
female exammed from March through June iN
= 55) was pregnant. Embryos were found in July
and October. Two embryos from females caught
in July were near-term. Four embryos found in
October were in an early stage of development.
All females (A^ = 4) inspected in October were
pregnant, and all embryos were in the same de-
velopmental stage, suggesting coordinated breed-
ing activity. The largest female (DW 1,801 mm)
had a single embryo, which appeared to be in the
final stage of fetal development, fully pigmented,
the yolk sac completely absorbed, and the umbili-
cus a mere scar (Fig. 5A). The embryo's DW was
630 mm; its WT, 3.4 kg. It occupied the left uterus
with the rostrum pointed forward. Its pectoral
fins were folded dorsally, the right pectoral on
top. The cephalic fins were almost totally un-
rolled and extended ventrally towards the mid-
line of the body. The uterus, with the embryo
inside, occupied roughly one third of the female's
pleuroperitoneal cavity. Lack of space inside the
cavity and the distended skin on the abdomen
made it apparent that no other embryo had been
recently expelled or aborted. Uniparity appears to
be a common pattern within the genus Mobiila
(Hill 1862; Gill 1908; Coles 1913, 1916b; Barnard
1925; Setna and Sarangdhar 1950; Cadenat 1960;
Wallace 1967; Capape and Zaouali 1976;
Figueiredo 1977). Only Risso (1826) asserted that
Cephaloptera giorna (= M. mohular) may have
one or two young, but his statement was not doc-

umented. Since the smallest free-swimming spec-
imen noted had a DW of 864 mm, the average size
at birth for M. thurstoni is probably between 650
and 850 mm DW, and a WT of approximately 4.5
kg. The second largest female (DW 1,799 mm)
had also only one embryo in her left uterus. The
embryo was unpigmented, with disc 210 mm
wide, and weighed 173 g. Like the term-embryo,
its rostrum was pointing forward; unlike it, how-
ever, its pectoral fins were folded ventrally.

Mating, parturition, and early mobulid life his-
tory take place in the shallower portion of a popu-
lation's range, not an uncommon elasmobranch
feature. McLaughlin and O'Gower (1971) dis-
cussed inshore movements in the mating Port
Jackson shark Heterodontus portusjacksoni , as
did Springer (1960) for the sandbar shark Eu-
lamia milberti (= Carcharhinus plumbeus). One-
year-old gray reef sharks, Carcharhinus am-
bhrhynchos, were observed in French Polynesia
in shallower waters than adults by Nelson and
Johnson (1980). A similar result was reported for
the hammerhead shark, Sphyrna lewini , by
Clarke (1971) in Hawaii and by Klimley ( 1983) in
the Gulf of California. Bullis (1967) hypothesized
an upward movement to shallower depths for
newborn marbled cat sharks, Galeus arae. There
is likely an advantage for juveniles to remain in
relatively protected areas during the earlier
stages of their life, when they are most vulnera-
ble to predation (Springer 1967).

Examination of 139 stomachs indicated that
M. thurstoni was extremely specialized in its
feeding habits. Eighty one (58.3%) stomachs were
empty, or had only traces of food (<1 g wet
weight). The remaining 58 stomachs (41.7%) had
quantifiable contents. All recognizable prey
items were planktonic crustaceans (with the ex-
ception of a few fish eggs, one nematod, and a
small coleopteran, probably ingested accidentally
when it was floating). They were listed, ranked by
decreasing ^IRI, in Table 1. Mobula thurstoni

Table 1 —Prey species found m 57 stomachs of Mobula thurstoni, ranked by decreasing Index
of Relative Importance (IRI). N = percentage of prey species by number; M = percentage of
prey species by mass; F = percent frequency of occurrence of prey species.

Prey species

°/oM

IRI ,

%IRI

Nyctiphanes simplex

4,940.8

86 70

4.982.0

87 40

877

15,268 1

97,90

Mysidium sp.

6352

11 10

631.2

11.10

12.3

273 '

1.75

Copepoda

99.9

1 80

74.5

1.31

15.8

49 ;

Megalopa larvae

12.0

0.21

4.1

0.07

14.0

3.95 1

Hyperiid amphipods

6.8

0.12

5.8

0.10

5.3

1.18 '

[ 0,35

Fisfi eggs

1,4

0.03

1,8

0.03

5.3

0.302 \

Nematoscelis difi.

0.5

001

1.1

002

1.8

0.050 1

Stomatopod larvae

0.7

0.01

0.6

0.01

1.8

0.041 'J

NOTARBARTOLO-DI-SCIARA: NATURAL HISTORY OF MOBULA

B i

Figure 5. — A: term-embryo of Mobula thurstoni . The scale in the photograph is in centimeters. B; embryo of M.

tarapacana . The scale bar equals 5 cm.

KISHEKY BULLETIN: VOL 86, NO. 1

fed mostly on adult and juvenile euphausiids,
Nyctiphanes .s///?p/c.v. The niysid Mysidium sp.
(underscribed, Thomas E. Bowman^) was second
in order of importance. Mysids and euphausiids
were never found together in the same stomach.
The overall importance of the two food items dif-
fered by two orders of magnitude. All other prey
species found in the stomachs were rare and prob-
ably fortuitously ingested. These included one
zoea larva and the following copepod species:
Undinula vulgaris, Eucalanus subcrassus, E.
subtenuis, Ternora discaudata, Scolecithrix
danae, Nannocalanus minor, Euchaeta remana,
Euchaeta sp., and Labidocera diandra. Diet
varied with season (Fig. 6A), with mysids being
dominant from December through March, and eu-
phausiids during the warmer months. Diet varied
with predator size (Table 2): smaller individuals
fed both on euphausiids and mysids; the larger
rays fed only on euphausiids. This result probably
reflects the predominance of smaller rays during
winter, when fewer euphausiids are available,
rather than an ontogenetic change in food prefer-
ences.

Table 2. — Size differences in the diet of Mobula thurstoni.

%IRI

Prey species

DW < 1 ,300
(n = 21)

mm DW

> 1 ,300 mm
(n = 36)

Nyctiphanes simplex
Mysidium sp.
Other

79.04

18.13

2.83

99.91
0
0.09

Two remoras (Echeneididae) were occasionally
found on large M. thurstoni: Remora remora (3
specimens; range: 98-200 mm SL), and R.
albescens (3 specimens; 93-100 mm SL). Crus-
tacean parasites were encountered: Pupulina
minor (Copepoda: Caligidae), Ecthrogaleus den-
ticulatus (Copepoda: Pandaridae) sparsely on the
skin, and Ecthrogaleus disciarai (Benz and Deets
1987) in large patches on the dorsal surface, En-
tepherus laminipes (Copepoda: Cecropidae) from
the branchial filter plates, Eudactylina oliveri
(Copepoda: Eudactylinae) from the gill lamellae,
and Kroeyerina sp. (Copepoda: Kroyeriidae) from
the olfactory lamellae.

Mobula thurstoni was usually observed at the
surface in coastal waters of Bahia de la Ventana,
Cueva de Leon, and Bahia de los Muertos, often
within a few hundred meters of land and occa-

2Thoma.s E. Bowman, Smithsonian Institution, Washington,
D.C., pers. commun. 1984.

sionally as far as 6 km. When sighted offshore, it
was sometimes found over considerable depths
O500 m), although it appeared to be more abun-
dant in shallower, neritic waters. Mobula
thurstoni was always caught in the shallower
part of the nets, usually at a depth of < 100 m. The
greatest part of the catch, however, was surface-
dwelling rays. Beginning in mid-April, numerous
M. thurstoni were consistently seen in the early
morning hours cruising slowly at the surface.
They would frequently pause, conspicuous on
calmer days, with the tips of their pectoral fins
protruding out of the water. This behavior is well
known in mobulids (Norman and Eraser 1937); it
has been observed also in connection with mating
activities in M. olfersi ( = M. hypostoma ) by Coles
(1910). During such occasions, fishermen could
easily approach the rays and harpoon them, be-
fore startling them and causing them to dive. Re-
peated captures within the same aggregation re-
vealed that rays of various sizes and both sexes
could be found together. While at the surface,
M. thurstoni was usually solitary or in small,
nonpolarized groups (2-6), rather than in larger
aggregations or schools. The species was fre-
quently seen jumping out of the water in spectac-
ular, often reiterated somersaults; it was recog-
nized by the distinctive ventral markings.

It is not known to what extent mobulids make
use of the sea bottom. Beebe and Hollister (1935)
observed a group of 12 small devilfish (most likely
Mobula ) lying on the sandy substrate off Frigate
Islet, in the British West Indies. Bigelow and
Schroeder (1953) speculated that Manta spends
much of its time resting quietly on the seafloor.
During an experiment organized in conjunction
with Sea World of San Diego, aimed at establish-
ing whether M. thurstoni could survive in a con-
fined environment, five young specimens were
captured with gill nets and kept in a large pen (6
m in diameter) anchored in 2.5 m of water in
Ensenada de los Muertos. None of the rays sur-
vived 24 hours of captivity; the reasons for their
deaths were not clear, although the particularly
stressful capturing method appeared as a likely
cause. During that experiment the negatively
buoyant rays (sinking tail-first as soon as they
stopped swimming) spent a great deal of time
resting on the bottom, and were able to circulate
water through their gills while resting, by a syn-
chronized maneuvering of the oral valve and of
the gill covers (as judged from the flow made vis-
ible by the numerous particles suspended in the
water). A frequent method of turning around

NOTARBARTOLO-Dl-SCIARA: NATURAL HISTORY OF MOBULA

XII -II

VI VII VIM

Figure 6. — A. seasonal variation of the relative importance ofNyctiphanes simplex and Mysidium sp.
in the diet of Mobula thurstoni. B. seasonal variation of the abundance of adult and juvenile A^.
simplex in the coastal areas of the southwestern Gulf of California (modified from Brinton and
Townsend 19801. C. captures of M. tarapacana . D. captures of M. munkiana . E. mean number of
daily captures of M. japanica; bars represent 2 SE on either side of the mean. F. mean number of
daily captures of M. thurstoni: error bars as in E. (*): although specimens were also captured
between December and February, data are not comparable with spring and summer captures because
catch effort was minor and inconsistent in the colder months.

FISHKKY BULI.KTIN: VOL 86, NO 1

(e.g., when swimming towards the wall of the
pen) was to dive vertically in a tight circle until
swimming in the opposite direction in an inverted
position, and then spinning around the body axis
to brmg the dorsal side up, rather than turning by
banking to the right or to the left.

An indication of the seasonal abundance of
M. thiirstoni in the surface waters of the study
area was obtained by the mean number of rays
caught daily from March to July 1983 (Fig. 6;
Table 3). Mean daily catch should be taken as a
rough indication of the relative abundance of M.
thurstom rather than as a precise index because
the fishing effort was difficult to quantify. Mean
effort, however, was roughly constant from March
through July because the mean monthly number
of working boats (about 20) and the number and
size of the nets set then was constant. Further-
more, the fishermen would harpoon a ray every
time they had the opportunity to do so. Peak of
abundance was in June, a result which appears to
be consistent with the fishermen's past experi-
ence, despite the 1983 abnormally high water
temperatures (Cane 1983). In July the number of
M. thurstom caught had dropped drastically, and
most of the catch consisted of M. japanica . No
information was obtainable for the August-Sep-
tember period. Eighteen specimens were cap-
tured during six fall and winter field trips (24
January-8 February 1981; 25 November 1981;
16-21 December 1981; 20-23 December 1982;
19-26 January 1984; 28 October-1 November
1984), but that figure was not comparable with
other data because part of the fishing cooperative
migrated south to Los Frailes during the cooler
months. It is common knowledge, however,
among the local fishermen, that M. thurstom in
the colder season is present, but in fewer numbers
than during the summer.

The study area constitutes a feeding, mating
and nursery ground for M. thurstoni. The eu-
phausiid Nyctiphanes simplex , the main diet item
and the only food of the adults when in the area,
is the most abundant and widespread euphausiid
in the Gulf of California, and has been observed in
dense swarms (Brinton and Townsend 1980). Al-
though it is found in the study area year-round,
its juvenile and adult stages are most abundant
between February and August, peaking in June
on the west side of the Gulf of California (Brinton
and Townsend 1980). The seasonal abundance of
M. thurstoni in the southern Gulf thus seems to be
closely related to the seasonal abundance of its
main prey. It is impossible to describe the general

Table 3— Mean number of daily captures of Mobula thurstom
a - total monthly number of captures; b monthily number of days
of monitonng; X = mean number of daily captures; SE = standard
error of tfie mean; so = standard deviation.

Month

a
6

2SE

range

March

038

054

1.1

0-4

April

0.69

1.3

0-4

May

5.00

1.95

2.6

1-9

June

5.10

332

6.4

0-23

July

0.56

0.34

1.0

0-3

movement and life history pattern of M. thurstom
in the Gulf of California from the fragmentary
information available. The scanty data, however,
suggest the following: 1 ) adult male and nonpreg-
nant adult female M. thurstoni enter the area in
spring to feed and to mate, 2) pregnant females
segregate from the rest of the population in
spring (as is also suggested by the slight predom-
inance of males in spring and early summer),
3) gestation period is one year and females give
birth to one young every two or more years, 4) the
young are born in the study area or near it in
midsummer and remain there throughout their
early life, and 5) in late summer, when the num-
bers of adult and juvenile Nyctiphanes simplex
decline due to intense heating of the water (Brin-
ton and Townsend 1980), adult M. thurstom leave
the area, whereas the young switch their diet
from euphausiids to mysids. Further investiga-
tions are needed for additional corroboration of
these hypotheses.

Mobula japanica

(MuUer and Henle 1841)

Local name: cubana de lomo bianco

A total of 78 specimens, 34 males (DW range
1,316-2,386 mm) and 44 females (1,470-2,302
mm), were caught at three stations (Punta Arena
de la Ventana, Cueva de Leon, and Ensenada de
los Muertos) and adjacent waters, between 16 De-
cember 1981 and 13 July 1983. Overall and sea-
sonal size-frequency distributions for M. japanica
are shown in Figure 7. With the exception of
April, when females were larger (T -value
= 4.697, df = 3, P < 0.02), there are no significant
size differences between the sexes (T -value
- 0.535, df= 76, P > 0.5). Most of the rays in the
sample were large; only three were < 1,900 mm
DW.

Twenty-seven specimens (size range: 1,316-
2,285 mm DW; 18.6-115 kg) were weighed. The
WT/DW relationship is described by the equation:

NOTARBARTOLO-DI-SCIARA: NATURAL HISTORY OF MOBULA

WT = 4.29 X 10-10 (DW)'^^

r = 0.98

where WT is given in kg, DW in mm. The overall
male to female ratio was 0.89 (A'^ = 781. Females
dominated the June through December period;
males were predominant in April and May. Sex
ratios, however, never significantly differed from
1 (x" test P > 0.05). Sex segregation, behavioral
or geographical, was never observed in M. japan -
ica (both sexes were caught together in nets and
by harpoon); sampling bias (N = 3) may explain
why only males were caught in May. Geographi-

cal size segregation, by contrast, was an evident
feature of sexually mature specimens in the Gulf
of California (Fig. 7).

No pregnant females were found, although in
some specimens the left uterus had a flabby and
dilated appearance, suggesting recent delivery.
Tissues at the base of the claspers of most of the
larger males were swollen and reddened in June
and July when the tips of the claspers were flex-
ible and the rhipidion could be easily spread, and
in doing so a white, viscous fluid would ooze from
the hypophyle; all this suggested mating activity.
The clasper length-DW relationship for M Japan -
ica (Fig. 8) did not exhibit a clear pattern as in M.

30 -I

10 -

OVERALL

Xtot = 2I25 I±3674 (H-7Q)
^^^ = 21016167 12 (N-34)
Xjj -2143.3139 28 (N = 44)

frrn r-[7^ r-TH , i

.^^

DECEMBER

Xtot = I86I4± 385 74 {H-5)
X^^ = I3I6(N = I)
Xj^:l997.8± 352 22(N = 4)

rmrTTi. i ., i .. i ., i .. i ,. i .1 ^. i ,. i

>
u

o
en
m

30 n

20 -

10 -

APRIL

X,o,^ 2156 0 166 58 (N^5)
X^^ = 2l050i28.84(N-3)

= 22325 + 53.00(N = 2)

mm,

MAY
Xtot = X^^= 2057 33 1122 56 (N^ 3)

.■ I ■■ I ■■ I

■■ I .Fxl— I,

20-

10 -

JUNE

Xtot=2I45 5131.32 (N^26)
X^<?^2I40 4 150 78(N=I0)
5(jo =2I48.814I04(N-I6)

D TOTAL
E3 ^^
0 %%

I ■■ I ■■ I

n]za_

JULY

XT0T = 2l466i42.94 (N = 39)
7^^ = 2132.2 + 88.06 (N= 17)
X^j =2157713598 (N = 22)

I .n-i.

I ■■ I .r-iT^

hi
4.

1450 1650 1850 2050 2250 1450 1650 1850 2050 2250

DISC WIDTH (mm)

Figure 7. — Size-frequency distribution.s of Mobula japanica (means ±2 SE).

FISUKKY BULLETIN: VOL H6. NO 1

Figure 8. — Relationship between clasper size and body size
in Mobula japan ica .

^150

O o
Z.<2
liJ^125

UJ "
CO "O

^«

O §100

# s«mlnal fluid preteni
O seminal fluid absent
A no data

.\.

1500 2000

DISC WIDTH (mm)

2500

thurstoni, since most of the specimens of M.
japanica were of larger size classes. It was appar-
ent, however, that at a DW between 2,050 and
2,150 mm there is a high degree of variability in
the relative size of claspers, which tended to sta-
bilize at higher values (DWs >2,150 mm), indi-
cating that male sexual maturity in M. Japanica
begins at a DW of 2,100 (±50) mm. Lack of infor-
mation on smaller specimens prevented a clear
understanding of the onset of female sexual ma-
turity (Fig. 9). Large eggs were found in speci-
mens as small as 2,070 mm DW, possibly indicat-
ing that female M. japanica began to mature at
that size.

Only 19 (24%) of 78 specimens had quantifiable
stomach contents (>1 g wet WT). The remaining
59 stomachs (76%) were empty or had only traces
of food. All M. japanica fed largely on the eu-
phausiid Nyctiphanes simplex (Table 4); no
mysids were found. Other species occurring in the
stomachs, including copepods, megalopa larvae,
stomatopod larvae, hyperiid amphipods, caridean
decapods (Crangon sp., Pasiphaea sp., and one
alpheid decapod), and one cumacean, had an over-
all relative diet importance of only 0.38%>. Mobula
japanica were therefore very similar in feeding
habits to large M. thurstoni. Information on the

Figure 9. — Relationship between size of largest ovum and
body size in Mobula japanica .

•
•

^ 30

•

• •

•

3 20

•-

•

• •

•

Sio

• •

2000 2200

DISC WIDTH (mm)

2400

Table 4. — Prey species found in 19 stomachs of Mobula japanica, ranked by decreas-
ing Index of Relative Importance (symbols as in Table 2).

Prey species

%/W

IRI

%IRI

Nyctiphanes

simplex
Copepoda
Other

1,890.64
4.61
4.74

99.51
0.24
0.25

1,869.85

1.15

29.00

9841
0.06
1.53

100.00
31 58
36.84

19,792
9.47
65.58

99 62
0.05
0.33

NOTARBAKTOI.O-DI SCIARA: NATl'HAI. HISToHY OF MOBVLA

winter diet of M.Japanica was lacking, as four of
the five specimens collected then had empty stom-
aches; the fifth, a large female, contained a small
fragment of a partially digested fish carcass.
Since quantifiable stomach contents were found
only in large rays between April and July, no size
or seasonal differences in the diet of M.japanica
could be detected.

Mohiila Japanica was often found carrying
Remora remora , usually seen clinging to the out-
side of body, but found once inside a spiracle. Six
specimens of R. remora (range 109-217 mm SL)
were collected from M. japanica . Only one speci-
men of/?, albescens (97 mm SL) was found, in the
mouth cavity of a M.japanica. A pilot fish, Nau-
crates ductor (Carangidae) also associated with
M. japanica, swam alongside a harpooned ray
that was being towed inshore and remained for
some time at the water's edge, where the ray was
beached. Mobula japanica was parasitized by the
following crustaceans: Nerocila acuminata
(Isopoda: Cymothoidae), Pupulina brevicauda
and P . minor (Copepoda: Caligidae) on the skin;
Eudactylina oliveri (Copepoda: Eudactylinae) in
the gills; and Kroeyerina sp. (Copepoda: Kroyeri-
idae) among the olfactory lamellae. Unidentified
trypanorhynch cestodes were occasionally found
within the pleuroperitoneal cavity.

Habitat preference of M . japanica did not ap-
pear to differ from that of M. thurstoni . However,
the use of the habitat differed seasonally: in April
and May, when M. thurstoni was abundant at the
surface, M. japanica was never seen, and few
specimens were bottom gillnetted during those
months. Conversely, M.japanica, in June and
July, was seen in the late morning hours at the
surface in groups of several individuals swim-
ming parallel to the shore. Occasionally speci-
mens were seen in water <1 m deep. Mobula
japanica is not known to school, and I never ob-
served schooling.

Coles (1910) reported that M . olfersi (= M. hy-
postoma ) utters a "musical, bell-like bark" when
dying. A similar account was given by Risso
(1810) of Cephalopterus massena {= M . mobular).
This information led subsequent authors (Nor-
man and Fraser 1937; Bigelow and Schroeder
1953) to wonder whether mobulid rays are capa-
ble of producing sounds while in the water. Sound
production is a fairly widespread phenomenon
among bony fishes (Fish and Mowbray 1970;
Tavolga 1971); however, elasmobranchs lack the
traditional structures used by teleosts to generate
sound, i.e., the swim bladder and bony skeletal

parts (Marshall 1962), and recognizable sounds
have not been recorded from these animals
(Backus 1963). Sound production among elasmo-
branchs has been reported only for the Atlantic
cownose ray, Rhinoptera bonasus (Myrberg
1981); in that case clicks and scraping sounds
were presumably produced with the dental plates,
elicited when strongly prodding three rays which
were confined in a tank (Fish and Mowbray
1970). Mobula japanica, when beached alive,
often emitted a distinctive noise which could have
been the equivalent of Coles' "bark". This noise,
however, was apparently caused by the periodic,
spasmodic contractions of the mandibular, pha-
ryngeal, and hypobranchial musculature of the
asphyxiating ray, which forced air from the
mouth cavity out of the gill openings through the
meshlike branchial filter plates. Although under-
water sonic recordings have never been made, it
seems unlikely that under normal circumstances
any audible sound could be produced in this fash-
ion by submerged mobulids.

This area served as a spring and summer feed-
ing and mating ground for adult M. japanica,
rather than as a pupping or nursery ground, as
indicated by the lack of small-sized specimens.
Seasonal abundance of M .japanica in the surface
waters was indicated by the catch data (Table 5)
and is comparable to the seasonal abundance of
M . thurstoni (Fig. 6). No M . japanica were ob-
served in March; in April and May they occurred
occasionally. By mid-June large numbers ap-
peared in the nearshore surface waters near
Punta Arena de la Ventana, and were easily har-
pooned. Most of the July mobulid catch consisted
of M .japanica, when the numbers of M. thurstoni
had declined. Data are lacking for the August-
October period, therefore it was impossible to tell
whether the peak of abundance occurred in July
or later. Fishermen's reports were not clear, al-
though there was agreement on an overall decline
of mobulid abundance in late summer. Mobula
japanica fed exclusively on the euphausiid A^yc-
tiphanes simplex, and its numbers apparently de-
clined concomitant with the late summer decline

Table 5. — Mean number of daily captures of Mob-
ula japanica (symbols as in Table 1).

Month

2SE

range

March

0-0

April

0.4

0.36

0.7

0-2

May

0.4

0.40

0.5

0-1

June

1.7

1.02

2.0

0-6

July

4.3

2.89

4.3

0-14

FISHERY BULLETIN: VOL 86, NO. 1

of abundance oC their prey (Brinton and
Townsend 1980). Some M . japanica , however, in-
cluding individuals both large and small, were
found in this region throughout December. Like
M. tharstoni , winter catch data are not compara-
ble because of differences in fishing effort, when
M . Japanica is apparently caught less frequently.

Mobula miinkiatia

NotJirbartolo-di-Sciara 1987

Local name: tortilla

Twenty-four specimens, 10 males (DW range
686-900 mm) and 14 females (719-1,097 mm),
were caught at four stations (Punta Arena de la
Ventana, Ensenada de los Muertos, Bahia de los
Frailes, and Isla El Pardito) and adjacent waters,
between December 1982 and October 1984. Size-
frequency distributions for M. munkiana (Fig. 10)
revealed that female mean size was greater than
male, although not significantly (T-value
= 1.724, df = 22,P > 0.1). Seasonal differences in
size-frequency distribution could not be examined
because M . munkiana were only collected during
the fall and winter. All 10 freshly captured speci-
mens were weighed (size range: 686-1,097 mm
DW; 4.1-11.8 kg). Their WT/DW relationship is
described by the following equation:

WT= 1.041 X 10-«(DW)2 34
r = 0.95.

WT is given in kg, DW in mm.

The largest specimen in the sample, a female,
was one of the largest "tortillas" ever seen. There
are no data on size at birth, as no embryos were
found. Lack of knowledge of the size of the young
of the year also prevented insight on size segrega-

>-
o

LlJ

q:
u.

LlI

CE
<

5 -

p?rn r-p^

Xtot = 886.5 ±33.4 mm (M = 24)
X^^ = 853.8 ± 39.5mm (N = IO)
X^^ =909.9 ±47. 1 mm (N:|4)

D TOTAL
m is
0 ??

650 750 850 950 1050
DISC WIDTH (mm)

Figure 10. — Size-frequency distribution.s oi' Mubula munki-
ana (means ±2 SE).

tion. Term-embryos in M . rochebrunei , a closely
related, similar sized species from west Africa,
were 340-350 mm wide (Cadenat 1960). Since
Mobula at birth has a DW of about 1/3 of the
adult, size at birth would be about 350 mm DW.
This information argues in favor of size segrega-
tion in M. munkiana. Male to female ratio was
0.71 (N = 24), insignificantly different from 1 (x"
test P > 0.05). Both sexes were caught in the
same net sets, indicating that males and females
school together, and that there was no sex segre-
gation, either geographic or behavioral.

A dried, twisted male carcass, for which mor-
phometries could not be obtained, with a calcu-
lated DW of 895 mm, had long, well-developed
claspers, markedly protruding beyond the pelvic
fins. Based on other mobulid species, this condi-
tion indicates sexual maturity. A second speci-
men, with a disc 686 mm wide, had small and
pliable claspers, and the ratio between clasper
length and pelvic fin length, both measured from
tip to anterior margin of vent, was 0.88. At this
ratio, both M. thurstoni and M . japanica are im-
mature. Two M . munkiana , 871 and 872 mm DW,
differed greatly in the relative size of their
claspers: one possessed slightly longer claspers
than the pelvic fins (ratio = 1.10), and an incipi-
ent hardening of the cartilage was apparent; in
the other specimen the claspers were much
shorter than the pelvics (ratio = 0.84), and still
soft. This information suggested that male sexual
maturity in M. munkiana began at about 870 mm
DW.

The largest female specimen was sexually ma-
ture, as it had a large, flaccid left uterus, and the
enlarged left ovary consisted mainly of about 30
macroscopic ova (size range 2-15.2 mm).

Ten stomachs were examined: four were empty,
three contained unidentifiable whitish matter,
and three were full of planktonic crustaceans. A
list of prey species ranked by decreasing 9rIRI is
given in Table 6. Mysidium sp. appeared to be the
main staple of M. munkiana's diet. Mobula
munkiana is thus similar to the wintering young
of M. thurstoni. One of the stomachs contained
coarse coral fragments and small gastropod
shells, perhaps ingested by the ray while foraging
on mysids near a sandy substrate. The west
African species M . rochebrunei appeared to have
similar feeding habits, as Cadenat (1960) found
mysids and a few larger postlarval stomatopods
in the stomachs of several specimens.

I have no record of remoras associating with
M. munkiana. The only parasitic crustacean

NOTARBARTOLO-Dl-SCIARA: NATURAL HISTORY OF MOBULA

found was Pupulina cf. minor (Copepoda: Caligi-
dae) from the skin.

Among M. miinkiana's distinguishing features
are its neritic preferences combined with its so-
cial habits. This is the only mobulid species in the
Gulf of California that was consistently seen in
schools. It is not known whether this is a seasonal
behavior, or a permanent ethological feature of
the species. Schools appear as a conspicuous dark
patch, sometimes a few tens of meters in diame-
ter, as they slowly cruise along the coastline in
shallow water. The presence of the school is often
also highlighted by the frequent, simultaneously
leaping individuals, which betray its position
from a long distance. Similar behavioral traits
(schooling and leaping) have been reported for
two closely related species from the Atlantic,
M. hypostoma (Bancroft 1829; Coles 1910, 1916a)
and M. rochebrunei (Cadenat 1960). During
leaps, M. munkiana occasionally reached a
height of about two DWs. Two types of leaps were
observed: rising vertically head first and landing
flat with the belly on the sea surface with a loud
clap (breach), and spinning one to three times
around the main transverse body axis (somer-
sault).

A salient feature of M. munkiana's ecology in
this area is its winter occurrence when all other
mobulids are absent or at their lowest numbers.
Mobula munkiana apparently subsists then
chiefly on the mysid shrimp, Mysidium sp., which
is also the main food for young wintering M.
thurstoni . However, M . munkiana frequents the
area occasionally in summer: two specimens were
caught by surface gill net in Bahia de la Ventana
in July 1983. Even during the season in which it
is most abundant, M. munkiana is seen in
"pulses", as its occurrence at any particular loca-
tion is spotty. It may occur in large numbers at
one location for a few days, and then be absent for
1 or 2 weeks. This observation suggests the possi-
bility that M . munkiana lives in large concentra-

tions, perhaps composed by several schools, which
travel along the coast. A similar phenomenon was
observed off the Senegal coast by Cadenat (1960)
in M . rochebrunei, a species which is closely re-
lated to M. munkiana both morphologically and
ecologically.

It is conceivable that mobulids in the northern
half of the eastern tropical Pacific mate and give
birth in summer, based on the few term and near-
term embryos found in summer in M . thurstoni
and M . japanica, and from anatomical evidence
of mating activity in adult males M . thurstoni
and M .japanica . That such a hypothesis can also
be extended to M . munkiana is supported by lack
of reproductive activity in any of the specimens
collected during the fall or winter, and that new-
born and young-of-the-year are missing from the
sample. This evidence corroborates the hypothe-
sis that the local waters are a wintering ground
for M. munkiana, which then migrates into an
unknown area (perhaps the northern Gulf of Cali-
fornia) during the warmer season for mating and
pupping.

The possible causes of this ecological difference
between M. munkiana and the other mobulids
are many, and open to speculation. Mysid abun-
dance may be declining in summer in the south-
ern Gulf of California, and M . munkiana perhaps
migrates to areas where this crustacean or re-
lated species abound during the warmer season.
Alternatively, M. munkiana could be excluded
from this region in spring by competition with the
incoming, larger M. japanica and adult M.
thurstoni. Finally, M. munkiana may be moving
during the summer into an area which is more
suitable for its reproductive needs. Unfortu-
nately, this recently discovered species is very
little known, and it has been reported only from
the Gulf of California and Ecuador, although its
distribution probably extends to other coastal
areas of the tropical east Pacific (Notarbartolo-di-
Sciara 1987).

Table 6^ — Prey species found in three stomachs of Mobula munkiana ranked by decreasing
Index or Relative Importance (symbols as m Table 2). Identifiable copepod species included
Undinula vulgans, Rhincalanus nasutus, and Scolecithnx danae. The stomatopods found
were "erhithrus" larvae. One unidentified food item was a fragment of a larger crustacean,
probably an euphausnd.

Prey species

°/oN

°/oM

IRI

%IRI

Mysidium sp. 293.40 97.80 287.40 95.80 100.00 19.360

Stomatopod larvae 4.65 1.55 9.06 3.02 66.67 304

Copepoda 1.59 0.53 1.44 0.48 66.67 67.3

Other 0.37 0.12 2.10 0.70 66.67 54.7

97.84
1.54
0.34
0.28

FISHERY BULLETIN: VOL. 86, NO. 1

Mob III (I tarapacatm (Philippi 1892)
Local name: vaquetilla

Mobula tarapacann is not a common species in
the study area. Seven specimens were collected,
one of which, a premature male embryo, was ex-
pelled by a large female while she was being
landed. Of the postnatal individuals, two were
male (DW range 2,476-2,494 mm), and four were
female (2,704-3,052 mm). All were caught in
Bahia de la Ventana between 9 June and 30 Octo-
ber 1983. All but two of the specimens were
weighed. The following equation describes the
WT/DW relationship for M. tarapacana (where
WT is given in kg, DW in mm):

WT = 2.378 X 10'*^(DW)2-92
r = 0.998.

Although all sampled postnatal M . tarapacana
were large, smaller individuals are known from
the area, as can be seen in photographs taken at
Punta Arena de la Ventana in summer 1981
(courtesy Felipe Galvan Magaha, CICIMAR, La
Paz, Mexico; also Greg B. Deets"^). This informa-
tion argues against geographical size segregation
of M. tarapacana . Data on the embryo provide no
indication of size at birth, since it was still far
from term. Pale pigmentation was apparent only
around the head and pelvic regions, and the ex-
ternal yolk sac was present (Fig. 5B). The em-
bryo, expelled tail first, was alive at birth. Judg-
ing by its size it had filled the left uterus
completely and must have been the sole develop-
ing embryo.

The small size of the sample does not permit
any clear inference on size at sexual maturity for
M . tarapacana . Some indication, however, can be
obtained by comparison with similar species. Of
the two postnatal males, the specimen with a disc
2,476 mm wide appeared to be immature: no sem-
inal fluid was found in the ducti deferentes, the
testes were small and apparently little developed,
and the ratio between clasper length and pelvic
fin length was 0.94. Conversely, the second post-
natal male, with a disc of approximately the same
width (2,494 mm), possessed claspers longer than
pelvics (ratio = 1.14), and the testes were well de-
veloped. Thus, sexual maturity in male M . tara-
pacana begins around a DW of 2,400-2,500 mm.

The specimen with a DW of 2,704 mm, one of two
nonpregnant females, had a bulky left ovary, con-
taining numerous large eggs; the largest, 32 mm
in diameter, weighed 12 g. Similar features ap-
peared to be associated with sexual maturity in
female M. thurstoni and M. Japanica. The left
ovary of another specimen, DW 2,831 mm, was
smaller, and the diameter of the largest ovum
was 18.6 mm, indicating that a DW of 2,700-
2,800 mm denotes a transitional stage for female
M. tarapacana , in which both mature and non-
mature individuals can occur.

Twelve echeneidids were recovered from
M . tarapacana . Three were Remora remora (size
range: 108-229 mm SL), and nine were R.
albescens (74-159 mm SL). The following crusta-
cean parasites were also found: one cymothoid
isopod (still in an unidentifiable aegathoid stage)
and Pupulina [lores (Copepoda: Caligidae) on the
skin, Entepherus laminipes (Copepoda: Cecropi-
dae) on the branchial filter plates, and Eu-
dactylina sp. (Copepoda: Eudactylinae) in the
gills.

Mobula tarapacana is strictly a summer and
fall visitor to this region (Fig. 6C). This species is
often found farther from the coast than
M . thurstoni and M. japanica, and may have
more pelagic habits. Four of the five stomachs
examined (all from specimens caught in summer)
were almost empty. Only traces of food were
found among the folds of the stomach epithelium.
Prey included four species of copepods (Acartia
sp. , Pontella sp., Tetnora discaudata , and Undin-
ula vulgaris ), hiperiid amphipods, one brachi-
uran (family Calappidae), one euphausiid, two
caridean decapods (one of which belonged to the
family Alpheidae), megalopa and stomatopod lar-
vae, and a fish egg. The fifth stomach, from a late
October capture, contained the remains of 27
fishes (probably carangids 15-30 cm long, and a
smaller anchovy-like species). Small tetraodon-
tids had been found before in the stomach of a
M . tarapacana caught in Bahia de la Ventana
(Felipe Galvan Magaha'*). On this basis it is im-
possible to determine whether M. tarapacana is a
specialized ichthyophagous ray, with the few
crustacean items accidentally ingested while
swimming, or a generalized feeder. The mesh size
of this species' branchial filter plates is indeed
greater than in other Mobula species
(Notarbartolo-di-Sciara 1987). However, filter-

•*Greg B. Deets, Long Beach State University, CA, pens, com-
mun. 1984.

^Felipe Galvan Magana, CICIMAR, La Paz, Mexico, pers.
commun. 1983.

NOTARBARTOLO-DI-SCIARA NATURAL HISTORY OK MOBrLA

feeding on planktonic Crustacea still appears
to be a feasible foraging technique for M. tara-
pacana, judging from the size of its branchial
sieve as it compares with the average-sized crus-
tacean prey.

SUMMARY AND CONCLUSIONS

Four species ofMobula were found in the south-
ern Gulf of California. The most abundant spe-
cies, M. thurstoni , was present year-round, but
only the smaller individuals were seen during the
winter. The bulk of the population, including the
adults, appeared in early spring. Numbers began
declining in July. Mobula japanica , the second
most abundant species, was comprised of only
large individuals; numbers progressively in-
creased from March throughout July. Large M.
japanica were rare in winter, but were occasion-
ally caught then. Mobula tarapacana is the rarest
mobulid in the area, yet its presence as a summer
and fall visitor is well known and predictable; it
is believed by the local fishermen to be more
abundant farther offshore. All three species share
a similar pattern of peak summer seasonal abun-
dance. The reverse is true for Mobula munkiana ,
it being most abundant in winter, and almost to-
tally absent during the rest of the year. It is not
known where any species goes when not seen in
the area. Seasonal migrations within the
epipelagic habitat to different areas of the
Panamic region are likely, but unverifiable be-
cause of the present lack of knowledge of the oc-
currence of identified Mobula species south of the
Gulf of California. Alternatively, devil rays may
spend part of the year in midwater, or near the
sea bottom, therefore disappearing from sight and
reach.

There is a striking similarity between the
array of mobulid species found in the Gulf of Cali-
fornia (and probably along the Pacific coast of
tropical America) and the mobulid fauna from the
tropical waters off west Africa. The family is rep-
resented in both areas by Manta birostris and by
four species oi Mobula : M . thurstoni; M Japanica
(reported from west Africa as M. rancureli by
Cadenat 1959); M . tarapacana (reported as M.
coilloti for African waters by Cadenat and Ran-
curel 1960 and Stehmann 1981); and a small gi-e-
garious form, represented in the Gulf of Califor-
nia by M . munkiana and off west Africa by the
closely related M . rochebrunei (Notarbartolo-di-
Sciara 1987). Tropical coastal areas off west
America and west Africa are known to be among

the most productive tropical waters in the world,
because of comparable large-scale atmospheric
and oceanographic circulation patterns (Sverdrup
et al. 1942). It is conceivable that the ecological
similarity between these two regions is reflected
in similar faunal associations, especially as far as
low levels of the tropic chain (e.g., plankton-
feeding vertebrates) are concerned.

The Gulf of California presents a unique envi-
ronment in the eastern Pacific Ocean, with ex-
treme annual water temperature ranges, wind-
induced mixing and upwellings, and subsequent
great productivity (Roden 1964; Brusca 1980).
Upwelling is caused along the peninsular coast by
the southerly winds prevailing during the
warmer months. This environment apparently
creates optimal conditions for the existence of the
euphausiid shrimp A^yc^/p/?a/?es simplex , which is
found in great abundance in the neritic habitat
between spring and midsummer, before the in-
tense August heat causes a decline in its numbers
(Brinton and Townsend 1980). The following data
are combined in Figure 6 to provide an overview
of the possible relationship between the seasonal-
ity of predator and prey in the study area: a) the
relative importance of Nyctiphanes simplex and
Mysidium sp. in the diet of M. thurstoni; b) the
relative abundance of A'^. simplex; the occurrence
of M. tarapacana (c) and M . munkiana (d) in the
catch; and the relative abundances of M .japanica
(e) and M . thurstoni (f) (no data on the biology of
Mysidium sp. are available).

Young M . thurstoni and all M . munkiana ex-
amined in winter appeared to subsist largely on
Mysidium sp., whereas adult M. thurstoni and
M .japanica caught during the warmer months
fed exclusively on N . simplex . An extreme degree
of feeding specialization was evident in all mobu-
lid species in which quantitative analyses of the
stomach contents was possible; most prey forms,
other than N . simplex and Mysidium sp., were so
rare that they were probably ingested acciden-
tally. Stenophagy was linked to feeding special-
ization in another myliobatiform species, the
mollusk-feeder Rhinoptera bonasus (Schwartz
1966; Smith and Merriner 1985). These results
suggest that devil rays are highly efficient in lo-
cating and selecting their preferred food. They
may be aided during this behavior by their prey's
habit of swarming. Competitive interaction is to
be expected between sympatric species-pairs
which are closely related both taxonomically and
ecologically. Food-resource partitioning is known
to occur in sympatric species-pairs of skates

FISHERY BULLETIN: VOL. 86, NO. 1

(McEachran et al. 1976). This condition, however,
is not necessarily true when the sought-after re-
sources are not in short supply (Zaret and Rand
1971). This may be the case of M. thurstoni and
M . japanica feeding together on A^. simplex when
the abundance of euphausiids is at its peak. Com-
petition should occur, however, in late summer,
when prey numbers decline. It would be interest-
ing to determine whether the slight morphologi-
cal and behavioral differences between poten-
tially competing species pairs (M. thurstoni/
M. munkiana in winter, M. thurstoni IM . japan-
ica in spring and summer) influence or reflect
partitioning of their habitat when food resources
become limiting, as was described for both fresh-
water (Werner and Hall 1977) and marine
teleosts (Hixon 1980; Larson 1980).

This overview of the ecology and natural his-
tory of mobulids in the Gulf of California is based
on field investigations made chiefly in 1983, a
year in which the El Nino perturbation was par-
ticularly severe (Cane 1983). Although in terms
of fishermen's experience the year 1983 was not
unduly different, as far as mobulid relative abun-
dance and seasonality are concerned, the abnor-
mally high water temperatures resulting from El
Nino may have affected the devil rays studied in
subtle ways; therefore this investigation should
be repeated in a normal year.

According to the fishermen, the abundance of
sharks (mostly carcharhinids and sphyrnids) on
which their activity is based is declining. This
decline will probably result in an increase of mob-
ulid fishing effort. It is of concern that 12% of the
specimens of M. thurstoni caught were immature
(DW <1,500 mm).

ACKNOWLEDGMENTS

I owe deep gratitude to the many persons who
assisted me in this investigation: Richard H.
Rosenblatt, Theodore H. Bullock, Paul K. Dayton,
William E. Evans, and Walter H. Munk, mem-
bers of my doctoral committee; Edward Brinton,
Robert Cowen, Abraham Fleminger, Nicholas
Holland, Margaret Knight, Spencer Luke,
William Newman, Mark Grygier, Jeff
Schweitzer, George Shor, George Snyder, and
Fred White of the Scripps Institution of Oceanog-
raphy; Thomas Bowman (United States National
Museum, Washington, D.C.); Daniel Brooks (Uni-
versity of British Columbia, Vancouver); Greg
Deets (Long Beach State University); Dennis
Bedford and Robert Lea (California Department

of Fish and Game); Fay Wolfson (Hubbs Marine
Research Institute); Alexis Fossi (Institut Na-
tional des Techniques de la Mer, Cherbourg,
France); Felipe Galvan Magaiia (Centro Inter-
diciplinario de Ciencias Marinas, La Paz, Mex-
ico); Lalo Cuevas, Marcelo Geraldo, Juan Lucero,
and their colleagues of the Cooperativa Pesquera
de Punta Arena de la Ventana; Steven Kramer
(National Marine Fisheries Service, Honolulu,
Hawaii); Carl A. Jantsch and Steven D. Kamol-
nich (Sea World, Inc., San Diego). The Hubbs
Marine Research Institute (San Diego) loaned a
sailing vessel, the "Fling", for the field study;
Rodney Black helped in the outfitting of the ves-
sel and in sailing it to the Gulf of California. This
investigation was supported in part by a grant
from the Foundation for Ocean Research (San
Diego).

LITERATURE CITED

Backus. R H

1963. Hearing in elasmobranchs. In P. W. Gilbert
(editor), Sharks and survival, p. 243-254. D. C. Heath &
Co., Lexington, MA.
Bancroft, E. N

1829. On the fish known in Jamaica as the
sea-devil. Zool. J. 4:444-457.
Barnard. K H

1925. A monograph of the marine fishes of South Africa.
Part 1. Ann. S. Afr. Mus. 21:1-415.
Barton, O

1948. Color notes on Pacific manta, including a new
form. Copeia 1948:146-147.

BEEBE, W , AND M A HOLLISTER

1935. The fishes of Union Island, Grenadines, British
West Indies, with the description of a new species of
stargazer. Zoologica, N.Y. 19:209-224.
Beebe. W., and J Tee- Van

1938. Eastern Pacific Expeditions of the New York Zo-
ological Society. XV. Seven new marine fishes from
Lower California. Zoologica, N.Y. 23(15):299-301.

1941. Eastern Pacific expeditions of the New York Zoolog-
ical Society. XXVIII. Fishes from the Tropical Eastern
Pacific. (From Cedros Island, Lower California, south to
the Galapagos Islands and northern Peru). Part 3. Rays,
mantas and chimaeras. Zoologica, N.Y. 26(26):245-
280.
Benz. G W . AND G B Debts

1987. Echthrogaleus disciarai sp. nov. (Siphonostoma-
toida: Pandaridae), a parasitic copepod of the devil ray
Mobula lucasana Beebe and Tee- Van, 1938 from the Sea
ofCortez. Can. J. Zool. 65:685-690.

BiGELOW. H B , AND W C SCHROEDER

1953. Sawfishes, guitarfishes, skates and rays. In J.
Tee- Van, C. M. Breder, A. E. Parr, W. C. Schroeder, and
L. P. Schultz (editors), Fishes of the western north At-
lantic, Memoir 1, part 2, p. 1-514. Sears Foundation for
Marine Research, New Haven.

BrINTON, E . AND A W TOWNSEND

1980. Euphausiids in the Gulf of California - the 1957

NOTAKliAIMOI.O 1)1 SflARA NATl'KAl. HISTORY OF .\f()IUl.A

cruises. Calif. Coop Oifaiiic Fi.sh Inve.st. Rt'p. 21:211-
•J.'i.'').

Bki-.sca. R C

1980. Common intertidal invcrti-brate.-i of thf (iulf of
California Univ. Arizona Pres.s, Tust<in. 513 p

BLl.l.lS. H H ,Jk

1967. Depth segregations and distribution of sex-
maturity groups in the marbled catshark. Galena
arav. In P. \V. Cilbert, R. F. Mathewson, and I). P. Rail
(editors). Sharks, skates, and rays. p. 141-148. Johns
Hopkins Press. Baltimore.

Cadknat. J

1959. Notes d'lchtyologie ouest-africaine. XXV. Descrip-
tion dune Mohiila de grande taille. a aiguillon caudal, de
Cote dlvoire: Mobula rancurvli. sp. nov. Bull. Inst. F"r.
Afr. Noire Ser. A 21 :i;326-1331.

19(i(). Notes d'ichtyoJogie ouest-africaine. XXIX. Les
Mobulidae de la cote occidentale d'Afrique. Bull. Inst.
Fr. Afr. Noire Ser. A 22:1053-1084.

CaHENAT. J . A.M) P Ra.MLRKI,

1960. Notes dichtyologie ouest-afncaine. XXVI. Descrip-
tion dune nouvelle espece de Mobulidae de la Cote
divoire: Mobula cotllnti - Bull. Inst. Fr. Afr. Noire Ser.
A 22:283-288.

Cank. M a

1983 Oceanographic events during El Nino. Science
222:1189-1195.
CAPAPE. C . AND J ZaoL'AL!

1976. Notes on the presence of the devilfish Mobula rnob-
ular (Bonnaterre. 1788) (Selachii. Rajiformes) in
Tunisian waters. Doriana 5i 223): 1-8.
CastroAci'IRRE. J L

1965. Feces sierra, rayas, mantas y especies afines de
Mexico. An. Inst. Nac. Invest. Biol.-Pesq. 1:171-256.
Chirkhicno. N.

1974. Clave para identificar los peces marinos del Peru.
Informe n. 44. Callao: Institute del Mar del Peru, 387 p.
Clarke, T A

1971. The ecology of the scalloped hammerhead shark,
Sphyrna Icuini. in Hawaii. Pac. Sci. 25i2):133-144.
Coi.ES R J

1910. Observations on the habits and distribution of cer-
tain fishes taken on the coast of North Carolina. Bull.
Am. Mus. Nat. Hist. 28:337-348.
1913. Notes on the embryos of several species of rays, with
remarks on the northward summer migration of certain
tropical forms observed on the coast of North Caro-
lina. Bull. Am. Mus. Nat. Hist. 32:29-35.
1915. Notes on the sharks and rays of the Cape Lookout,

N.C. Proc. Biol. Soc. Wash. 28:89-94.
1916a. Natural history notes on the devil fish, Manta
btrostna (Walbaumi and Molbula olfersi iMuller). Bull.
Am. Mus. Nat. Hist. 35:649-657.
1916b. Notes on Radcliffe's sharks and rays of Beau-
fort. Copeia 32:45-47.
Daniel J F

1934. The elasmobranch fishes. Univ. California Press,
Berkeley. 332 p.
FlC.l'EIREDO, L

1977 Manual de peixes marinhos do sudeste do Brasil.
Vol. 1. Cacoes, rayas e quimeras. Universidade de Sao
Paulo. Museu de Zoologia. Sao Paulo, 105 p.
Fish. M P . a.nu W H Mowbray

1970. Sounds of western north Atlantic fishes: a reference
of biological underwater sounds. Johns Hopkins Press,
Baltimore, 207 p.

Fowler. H W

1944. Results of the fifth George Vanderbilt South Pacific
Expedition il941i. The fishes. Monogr. Acad. Nat. Sci.
Phila. 6:57-583.
GlI.L.T

1908. The story of the devil-fish. Smithson. Misc. Col-
lect. 52(1816):155-180.

GlLMORE.R G J W DoDRILL. ANDP A LiNLEY

1983. Reproduction and embryonic development of the
sand tiger shark, Odontaspis taurus iRafinesque). Fish.
Bull., U.S. 81:201-225.
GoHAR. H a F , AND A R Bayou.mi

1959. On the anatomy of Manta eherenbergi with notes on
Mobula kuhlii. Publ. Mar. Biol. Stn. AI Ghardaqa
10:191-238.
Hill, R

1862. The devil-fish in Jamaica. Intell. Obs. 2:167-176.
Hixon. M a

1980. Competitive interactions between California reef
fishes of the genus Embiotoca . Ecology 61:918-931.
Hyslop. E J

1980. Stomach contents analysis - a review of methods
and their application. J. Fish. Biol. 17:411-429.
Klimley. a P

1983. Social organization of schools of the scalloped ham-
merhead shark, Sphyrna lewini (Griffith and Smith), in
the Gulf of California. Ph.D Thesis, University of Cali-
fornia, San Diego, Scripps Institution of Oceanography.
Larson. R J

1980. Competition, habitat selection, and the bathymetric
segregation of two rockfishes iSebastes) species. Ecol.
Monogr. 50i2):221-239.

Lloyu. R E

1908. On two species of eagle-rays with notes on the skull
of the genus Ceratoptera . Rec. Indian Mus. (Calcutta)
2(2):175-180.
McEachran. J D., D F BoESCH, and J A MUSICK

1976. Food division within two sympatric species-pairs of
skates (Pisces: Rajidaei. Mar. Biol. (Berl.) 35:301-317.
McLaughlin. R H . and A K OGower.

1971. Life history and underwater studies of a heterodont
shark. Ecol. Monogr. 41(4):271-289.
Marshall. N B

1962. The biology of sound-producing fishes. Symp.
Zool. Soc. Lond. 7:45-60.
Miller, J K

1966. Biomass determination of selected zooplankters
found in the California Cooperative Oceanic Fisheries
Investigations. Scripps Inst. Ocean. Ref Ser. 66-15. 16
P-

MULLER. J , AND J HENLE.

1841. Systematische Beschreibung der Plagiostomen.
Verlag von Veit & Co., Berlin, 204 p.
Myrberg, A A . JR

1981. Sound communication and interception in
fishes. In W. N. Tavolga, A. N. Popper, and R. N. Fay
(editors). Hearing and sound communication in fishes, p.
397-425. Springer- Verlag, New York. Heidelberg, and
Berlin.

Nelson, D R., and R. H Johnson.

1980. Behavior of the reef sharks of Rangiroa, French
Polynesia. Natl. Geogr. Res. Rep. 12:479-499.
NICH0L.S. J T . AND R C Ml rphy

1944. A collection of fishes from the Panama Bight,
Pacific Ocean. Bull. Am. Mus. Nat. Hist. 83(4):217-
260.

FISHKKY BULLKTIN: VOL. 86, NO. 1

NOKMAN, J K , AND F C FKASKR

1937. Giant fishes, whales, and dolphins W. W Norton,
N.Y., .361 p.

NOTARBARTOLO-DISC'IAKA, G

1985. A revisionary study of the genus Mohula
fiafinesque 1810, with the description of a new species,
and natural history notes on east Pacific mobu-
lids. Ph.D Thesis, Univ. California, San Diego.
University Microfilms International No. 8510912, Ann
Arbor, Michigan, 346 p.

1987. A revisionary study of the genus Mohula
Rafinesque, 1810 (Chondnchthyes: Mobulidaei, with the
description of a new species. Zool. J. Linn. Soc. 91:1-91.

In press. Myliobatiform rays fished in the southern Gulf

of California (Baja California Sur, Mexico) (Chon-

drichthyes: Myliobatiformesl. Proc. Fifth Symposium of

Marine Biology, La Paz, B.C.S., Mexico. 24-26 Oct. 1984.

Pequeno. G

1983. La condrictiofauna de las regiones de Chile y
California-Oregon: comparacion preliminar. In P. M.
Arana (editor). Proceedings of the International Confer-
ence on Marine Resources, p. 253-267. Pacific, Vina del
Mar, Chile.
Phii.ippi. R A

1892. Algunos peces de Chile. An. Mus. Nac. Chile, Sec.
1, Zool. 3:1-17.
PiNKAS. L.MS Oliphant. and I L K Iverson

1971. Food habits of albacore, bluefin tuna, and bonito in
Californian waters. Calif Dep. Fish Game Fish Bull.
1.52, p. 1-82.
RiSSO. A

1810. Ichtyologie de Nice, ou histoire naturelle des pois-
sons du departement des Alpes Maritimes. Paris F.
Schoell, 388 p.

1826. Histoire naturelle des principales productions de
1 Europe meridionale et particulierement de celles des
environs de Nice et des Alpes Maritimes. Vol. 3. Paris:
F. G. Levrault, 480 p.
RODEN, G I

1964. Oceanographic aspects of Gulf of California. Am.
Assoc. Pet. Geol., Symp. Mar. Geol. Gulf Calif 3:30-58.
Schwartz. F J

1966. Embryology and feeding behavior of the Atlantic
cownose ray, Rhmoptera honasus. Assoc. Islands Mar.
Lab. Caribb. 7th Meeting, Barbados, 24-26 August 1966,
p. 15.

SETNA, S. B , AND P, N, Sarangdhar

1950. Breeding habits of Bombay elasmobranchs. Rec.
Indian Mus. (Calcutta I, 47:107-124.
Smith. J W . and J. V. Merriner,

1985. Food habits and feeding behavior of the cownose
ray, Rhmoptera honasus, in lower Chesapeake
Bay. Estuaries 8:305-310.

1986. Observations on the reproductive biology of the
cownose ray, Rhmoptera honasus, in Chesapeake
Bay. Fish. Bull., U.S. 84:871-877.

Springer, S

1960. Natural history of the sandbar shark Eulamia mil-

berti. U.S. Fish Wildl. Serv., Fish. Bull. 61:1-38.
1967. Social organization of shark populations. In P. W.
Gilbert, R. F. Mathewson, and D. P. Randall (editors).
Sharks, skates and rays, p. 149-174. Johns Hopkins
Press, Baltimore.
Stehmann, M

1981. Batoid fishes. /« W. Fischer, G. Bianchi, and W. B.
Scott (editors), FAO species identification sheets for fish-
ery purposes. Eastern central Atlantic, fishing areas 34,
37 (in part), Vol. 5. Rome: FAO.
Sverdrup, H. U , M W Johnson, and R H Fleming

1942. The oceans. Prentice-Hall, N.Y., 1087 p.
Tavolga. W. N.

1971. Sound production and detection. In W. S. Hoar
and D. J. Randall (editors). Fish physiology. Vol. 5, p.
135-205. Academic Press, N.Y.
Wallace, J H

1967. The batoid fishes of the east coast of southern
Africa. Part 2. Manta, eagle, duckbill, cownose, butterfiy
and stingrays. S. Afr. Assoc. Mar. Biol. Res., Ocean.
Res. Inst. Invest. Rep. 16:1-55.
Werner, E E , and D J Hall

1977. Competition and habitat shift in two sunfishes
(Centrarchidae). Ecology 58:869-876.
Wourms. J P

1977. Reproduction and development of chondrichthyan

fishes. Am. Zool. 17:379-410.
1981. Viviparity: the maternal-fetal relationship in
fishes. Am. Zool. 21:473-515.
Zaret, T M . AND A S Rand

1971. Competition in tropical stream fishes: support for
the competitive exclusion principle. Ecology 52:336—
342.

NOTES ON DECAPOD AND EUPHAUSIID CRUSTACEANS,
CONTINENTAL MARGIN, WESTERN ATLANTIC, GEORGES BANK

TO \XTSTERN FLORIDA, USA

Austin B Williams'

ABSTRACT

Twenty-six species of decapod crustaceans in 16 families and 1 species of euphausiid are reported from
the outer continental shelf, submarine canyons, and nearby slope of the eastern and southeastern
United States. Station data are given for all collections made with the aid of submersible and surface
vessels. Bathymetric and geographic distributions are summarized for six species (Lithodes maja,
Munida forceps. M. longipes, Chacellus filiformis, Dissodactylus juvenilis . Euchirograpsus ameri-
canus ) whose ranges are extended. Comparative descriptive notes are given for other forms that have
uncertain identities and need further study (Alpheus cf. amblyonyx, sp. near Ligur. Mumdopsis cf.
transtridens ).

Records of decapod crustaceans from the outer
continental shelf, submarine canyons, and
nearby slope off the eastern United States have
accumulated in my files to the point that it seems
appropriate to publish them with notes on the
samples taken in that habitat. Specimens were
collected with the aid of deep and shallow water
submersibles and surface vessels operated cooper-
atively by the Lamont-Doherty Geological Labo-
ratory, Columbia University, Palisades, NY. with
sponsors listed hereinafter, and between Rutgers
University, Center for Coastal and Environmen-
tal Studies, and Department of Horticulture and
Forestry, New Brunswick. NJ; the Northeast
Fisheries Center. National Marine Fisheries Ser-
vice (NMFS). NOAA, Woods Hole. MA; and the
Southeast Fisheries Center Laboratories, NMFS.
NOAA, at Panama City, FL, and Pascagoula, MS
(Able et al. 1982, 1987; Grimes et al. 1980a. 1986;
and Lamont-Doherty records). Vessels (see Table
1 1 and their sponsoring institutions were DSRV
Alvm, Woods Hole Oceanographic Institution,
Woods Hole, MA; RV JSL I and JSL II, Harbor
Branch Oceanographic Institution, Inc., Fort
Pierce, FL; RV Cape Henlopen , University of Del-
aware, Lewes, DE; RV Eastward , Duke Univer-
sity, Beaufort. NC; RV Endeavor, University of
Rhode Island, Kingston, RI; RV Gyre, Texas
A&M University, Galveston, TX. The specimens
have been deposited in the crustacean collection

'Systematics Laboratory, National Marine Fisheries Service,
NOAA. U.S. National Museum of Natural History, Washing-
ton. DC 20560

Manuscript accepted August 1987.
FISHERY BULLETIN: VOL. 86. NO. 1, 1988.

of the United States National Museum of Natural
History (USNM).

Bathymetric and geographic distributions of
many decapod crustacean species from the North
American continental shelf in the western At-
lantic were reviewed by Squires (1965), Williams
and Wigley (1977), Wenner (1982), and Williams
(1984). Wenner and Boesch (1979) and Wenner
and Windsor (1979) included a deeper dwelling
component in their treatments of epibenthic deca-
pods collected from the continental shelf and
slope. Collections by workers from the contribut-
ing institutions listed above include 26 species of
decapod crustaceans in 16 families, some uniden-
tified fragments, and 1 species of euphausiid
listed in Table 1. Named localities are listed from
north to south (see Figures 1 and 2); successive
visits are arranged chronologically, and species
present in each collection are inventoried alpha-
betically.

Species whose reported bathymetric or geo-
graphic ranges are extended by presence in these
collections are discussed below. Some of the forms
have uncertain identities that may be clarified
after more thorough study of samples from the
outer shelf-upper slope environment.

SUPERFAMILY ALPHEOIDEA

Family Alpheidae

Alpheus cf. amblyonyx Chace 1972. Three lots of
specimens key out to A. amblyonyx (see Chace
1972), but they differ in several respects from it.

FISHERY BULLETIN: VOL. 85, NO 1

Table 1 . — Records of decapod crustaceans from submarine canyons, outer
listed from norlti to south, successive visits are arranged chronologically, and
in footnote. ' = Bathymetric or geographic range extension.

Locality

Depth Cruise/
(m) station

Date

■w

39.97'W

Lydonia Canyon
40^21 'N, 67 41
40°22'N, 67' 41
40°21.69N, 67

700-790 Gyre
462 Alvin 1269
906 Alvin 1270

8 June 1982

18 Sept. 1982

19 Sept. 1982

Oceanographer Canyon

40"23.7'N, 68'07.8'W

415-680

Eastv/ard
35982

15 IVIay 1979

40^6 3'N, 68'07.2'W

660-1,424

Eastward
35991

19 IVIay 1979

Slope Area III between Hydrographer and Veatch Canyons

39°50'N, 69°25'W

500-2,000

Endeavor

21 Oct. 1981

Slope Area II betv^^een Toms and Meys Canyons

39°13.05'N, 72°30.93'W

174

JSL 1 1081

1 Aug. 1981

39°12.64'N, 72"30.87'W

200

JSL 1 1081

1 Aug. 1981

39 11.06' N, 72°33.66'W

219

JSL 1 1082

1 Aug. 1981

Hendnckson Canyon

39°03.52'N, 72°28.45W

1,420-1,425

Alvin 1118

19 July 1981

Baltimore Canyon

38 09.6'N, 73 49.2'W

155-160

Eastward
35940

2 May 1979

38°07.9'N, 73°48.8'W

185-190

Eastward
35938

2 IVIay 1979

38°08.5'N, 73°49.9'W

280-570

Eastward
35944

2 IVIay 1979

38°09.6'N, 73°49.2'W

155-160

Eastward
35940

2 IVIay 1979

38°08.7N, 73°49.3'W

148-174

Eastward
35939

2 IVIay 1979

38°11.7'N, 73°52.8'W

1 50-245

Eastward
35942

2 IVlay 1979

38°03.6'N, 73°46.2'W

910-925

Eastward
35946

3 IVIay 1979

38 07.6'N, 73°46.8'W

152-163

Eastward
35947

3 May 1979

38''06'N, 73=50'W

200-800

Eastward
35979

7 May 1979

38°04.93'N, 73°47.79'W

1,024

Alvin 1109

11 July 1979

38°08'N, 73°50.7'W

600-700

B10-1

28 May 1981

34°04'N, 73°46'W

1,000-1,200

Cape Henlopen
CM02

13 June 1981

38°09'N, 73°51'W

171-381

JSL 1 1083

2 Aug. 1981

38°09.99'N, 73'51.86'W

381

JSL 1 1083

2 Aug. 1981

38°11'N, 73°51'W

165

JSL 1 1088

4 Aug. 1981

38°11'N, 73°51'W

165-244

JSL 1 1088

4 Aug. 1981

38°10.1'N, 73°52.2'W

225

JSL 1 1084

4 Aug. 1981

WILLIAMS: DECAPOD AND EUPHAUSIID CRUSTACEANS

continental shelf and upper slope off the eastern United States. Named localities are
species inventoried in each collection are listed alphabetically. Family abbreviations

Species

N. Sex

Notes, familyi

■ Lamont-

Doherty --

Munida valida Smith
Pagurus politus (Smith)
'Munidopsis cf. transthdens

Pequegnat and Pequegnat
Pandalus propinquus

G 0. Sars

Dredge. Ga

Pag

Pan

Panadalus propinquus

Id, 39

Day dredge.

Metacrangon lacqueti agassizii

Dredge, Cr

(Smith)

Nematocarcinus ensifer (Smith) 1 spec.

Camera. N

Euprognatha rastellifera

1 9 ovig.

Stimpson

'Euchirograpsus americanus

Eumunida picta Smith

Munida ins ins

A. Milne Edwards

(parasitized)

'Munidopsis cf. transtndens

Cancer borealis Stimpson

1 juv.

Can

'Euchirograpsus americanus

1 d , 3 9

A. Milne Edwards

Euprognatha rastellifera

Day dredge. M

Cancer borealis

Day dredge, Ca

Euprognatha rastellifera

Day dredge, M

'Munida longipes

1d, 29 ovig.

Day dredge, Ga

A. Milne Edwards

Pagurus politus (Smith)

Day dredge, Pa

'Chacellus filiformis Guinot

Day dredge. Go

Euprognatha rastellifera

Day dredge, M

Munida ins ins

3d, 39

Day dredge, Ga

Plesionika edwardsii (Brandt)

1 9 ovig.

Day dredge. Pan

Cancer borealis

2 juv.

Day dredge, Ca

Cancer irroratus Say

Day dredge, Ca

Euprognatha rastellifera

Day dredge, M

Cancer borealis

Day dredge, Ca

Sergestes arcticus

Day dredge, S

Collodes robustus Smith

1 9 ovig.

Day dredge, M

Eumunida picta

4 juv.. Id, 59

Camera sled, Ch

Lebbeus polaris (Sabine)

Camera sled, H

'Munidopsis cf. transtridens

Airplane wreck, Ga

Sergestes arcticus Kroyer,

Bongo net, S

euphausiid and copepods

Sergestes arcticus

Bongo net, S

Bathynectes longispina

Stimpson

'Lithodes maja (Linnaeus)

Cancer irroratus

Munida ins ins

Pagurus politus

Pag

FISHERY BULLETIN: VOL. 86, NO. 1

Table 1. — Continued.

Depth

Cruise/

Locality

(m)

station

Date

38' 09.51 'N, 73°51.25'W

338

JSL 1 1089

5 Aug. 1981

38'11'N, 73 SrW

402

JSL 1 1090

5 Aug. 1981

38 02N, 73 45'W

200-1,400

Gyre

26 May 1982

Norfolk Canyon

37°03'N, 74°37'W

215
Rutgers

JSL 1 1093

7 Aug. 1981

Lydonia Canyon

40'27.25'N, 67°42.30'W

160-178

JSL 1 1070

23 July 1981

Veatch Canyon

40 03'20"N, 69 45'23"W

124-126

JSL 1 1074

25 July 1981

40TO00"N, 69°45'25"W

123-128

JSL 1 1075

26 July 1981

39 59'51"N, 69°35'00"W

213-244

JSL 1 1076

26 July 1981

40"00.91'N,

70°50.79'W

243-304

JSL II 901

29 July

1984

40 0 1.29 'N,

7050.63'W

4003.24N,

71 05.01 'W

193-209

JSL II 902

29 July

1984

40"02.96N,

70 20.84'W

181-195

JSL II 909

2 Aug.

1984

40'12.73'N,

70'19.77'W

103-104

JSL II 910

2 Aug.

1984

40 12.51 'N,

70'20.22'W

103-104

JSL II 910

2 Aug.

1984

40"02.79'N,

70 11.95'W

183-337

JSL II 911

3 Aug.

1984

40°03.23'N,

70''12.0rW

213-327

JSL II 912

3 Aug.

1984

Off eastern Florida
28=43'N, 80 02'W 137 JSL I 1565 1 Oct. 1984

28°42.5'N, 80°02.8'W 98-114 JSL I 1566 2 Oct. 1984

Off western Florida
27=51 .2N, 84°53.7'W 250-260 JSL I 1673 24 Sept. 1985

27°56.2'N, 84°43.9'W 167-175 JSL I 1676 26 Sept. 1985

'A, Alpheidae; Cal, Calappidae, Can. Cancndae; Ch, Cliirostylidae; Cr, Crangonidae; Ga,
M, Majidae. N, Nematocarcinidae: Pag, Paguridae; Pan, Pandalidae; Pi, Pinnotheridae; Po,

WILLIAMS: DECAPOD AND EUPHAUSIID CRUSTACEANS

Species

N. Sex

Notes, family'

Rochinia crassa

A Milne Edwards
Pandalus propmquus
Sergestes arcticus

29
36

Pan

Camera, S

Bathynectes longispina

1 juv.

Rutgers

'Munida forceps

A. Milne Edwards

IcJ

Bot t 1 1 2°C, Ga

Munida ins ins
Pagurus politus
Acanthocarpus alexandn

Stimpson
'Chacellus filiformis

'Munida forceps

Munida ins ins

'Alphieus cf. amblyonyx Chace
'Alpfieus cf. amblyonyx
Munida ins ins
'Munida forceps
'Goneplacid'' crab
'Chacellus filiformis
'Munida forceps
'Munida forceps
Meganyctipfianes norvegica

(M. Sars)
'Munida forceps
'Chtacellus filiformis'^
Meganyctiptianes novegica
Meganyctiphanes novegica
Meganyctiptianes novegica

Alpfieus cf. amblyonyx

Munida forceps

Crab (xanthid or goneplacid

frags.)
'Alpfieus cf. amblyonx

Munida forceps

Id, 42 ovig.

Bot. t.

1 1 .4'C, Ga

Bot. t.

11.4 C, Pag

Bot. t.

11.5 C, Ca

36 (1 JUV.)

Bot. t.

11.2 C, Go

49 (1 juv.,

3 ovig.)

Bot. t.

12.2'C, Ga

1 (5 , 1 2 ovig.

Bot. t.

12.2'C, Ga

1 JUV,

Ga
Go

Id, 19

1 juv., 1 I ovig

Frag, of chela

2d, 19, 1 JUV.

Tilefisfi burrow, A

Tilefisl

T burrow, Ga

Tilefish burrow, Ga

12.1

C, A

Tilefishi burrow, Ga

*Near Ligur Sarato
'Munida forceps

' Dissodactylus juvenilis
Bouvier

2d. 1 9 Tilefish burrow. Hi

Id, 25 ovig. Tilefish burrow, Ga

1 juv.
1 V ovig. With Clypeaster

ravenellii A.

Agassiz, 13.8'C, Pi

Galatheidae. Go. Goneplacidae; Gr, Grapsidae: E. Euphausiidae; H, Hippolytidae; L. Lithodidae;
Ponunidae; S, Sergestidae.

FISHERY BULLETIN: VOL. 86, NO. 1

68 45

67 44 66 65 43 64 42

V J--*

.\ :.;;/ ^^ -^-^ . X ..• M \ /

^ i;i

,-'-\_-J ,'' NORTHEAST CHANNEL

r ■■■'

//'"' ,,^•''■''^^

./■

/^■'

;"' '.•■■ vVCORSAIR

J ;'••■ * '^ /

\\)k

' 1 •' <n ■ '

1 ^

■: S ■■ :j

i^v^S. •■•.... . ,-J^LYDONIA

/A .•■^V'Tgilbert

-^/ ; ,'TOCEANOGRAPHER /

66
39

43
73

-V

VL ^ ..'JrWELKER

,^° Y

:• ~3jrHYDR0GRAPHER \

... •■ ,-_/VEATCH

A(f

,'_/^TLANTIS y

.■/ /block ^

V-f^ HUDSON y

iC./ —

. //hendrickson

M ATLANTIC

;

.■ '

40
76

';^ WILMINGTON OCEAN <^
f:i BALTIMORE

36
70

* !)

.it WASHINGTON

^y\ NORFOLK

■\ \

^^X VI

34
72

x'S

/ -^Xj p. CAPE HATTERAS \ y \

77 36

76 35 75 74 34 73

Figure l. — East coast of United States showing continental shelf, slope and submarine canyon
areas from which species listed in Table 1 were collected. Base map adapted from Uchupi ( 1965) and
Veatch and Smith (1939). Contours m m: dotted = 100, dashed = 200, solid = 1,000.

WILLIAMS: DECAPOD AND EUPHAUSIID CRUSTACEANS

83 31

30 81

80 29

28 78

\ ^ A

■\K

78
27

y 26

25
80

27 84

83 26

25 81

Figure 2. — Florida peninsula including continental platform showing slope localities from which species listed
in Table 1 were collected. Base map adapted from Uchupi (1965). Contours in m; dotted = 100, dashed = 200,
solid = 1.000.

Some of the more obvious differences for the
Veatch Canyon material are: Rostrum relatively
shorter in relation to basal antennal article. Dor-
sal spines of telson more distally positioned, ante-
rior pair at about midlength of telson but subject
to some variation. Antennal scale with distal
spine exceeding antennular peduncle. Major
chela of mature male missing, but juvenile with
fingers bent mesad; dactyl moderately arched in
profile and greatly overreaching fixed finger,
somewhat twisted, compressed proximally and
dorsally producing thin dorsal margin, external
surface somewhat concave, occlusive surface
lacking plunger but broadened and strongly calci-
fied distally, fitted to obliquely flattened occlu-
sive surface of fixed finger lacking socket but
provided with small stout tooth on mesial sur-
face; palm with obsolescent dorsal and ventral
notches.

In the Florida material, the differences are:
Chela relatively stout, fingers stout and thick,
dactyl opening and closing in oblique plane, tip
rounded, bearing short plunger fitting into shal-
low socket on occlusive surface of fixed finger,
latter with 2 short spines on mesial surface; palm
with shallow notch on dorsal margin and slight

offset on ventral margin, outer surface smooth
but base of dactyl flanked by distodorsal groove
and longer mesial groove. Second pleopod of male
with appendix masculina exceeding appendix in-
terna. Uropodal exopod with lateral margin end-
ing in single sharp tooth and rather long, uncol-
ored movable spine; endopod lacking distal spines
but bearing subterminal tuft of setae on dorsal
surface.

Alpheus amblyonyx is distributed from Quin-
tana Roo (type locality, Bahia de la Ascen-
sion), Yucatan Peninsula, Mexico, to Puerto Rico,
Saint Thomas and Dominica; sublittoral (Chace
1972).

Family Hippolytidae

Three small shrimps, two males and one female
from JSL 1673 in the Gulf of Mexico off Florida,
represent an undescribed species resembling
members of the genus Ligur Sarato, 1885 from
the western Mediterranean and Indo-Pacific re-
gion (see Holthuis 1947, 1955). The specimens
were associated with burrow systems of the blue-
line tilefish, Caulolatilus microps Goode and
Bean.

FISHERY BULLETIN: VOL. 86, NO. 1

SUPERFAMILY PAGUROIDEA

Family Lithodidae

Lithodes maja (Linnaeus 1758). Southern limit
extended from Sandy Hook, NJ (see Williams
1984) to Baltimore Canyon.

SUPERFAMILY GALATHEOIDEA

Family Galatheidae

Munida forceps A. Milne Edwards 1880.
Geographic range extended from south of Norfolk
Canyon, 36°43.2'N, 74°38.0'W, 252 m (Wenner
1952), to Veatch and Lydonia Canyons off south-
ern New England, 103-337 m.

The distinctive color pattern of this species was
described on 14 October 1981 from specimens pre-
served in formalin 25 July 1981. Carapace (Fig.
3), salmon color with lavender submesial spots on
gastric region and interrupted U-shaped bands of
same color in nested series on mesogastric region,
posterior to cephalic groove, and arching across
posterior and posterolateral parts. Oblique red
lines on lateral wall of carapace below suture,
most prominent band along anterior edge, contin-
ued dorsally anterior to antennal peduncle and
ending on lateral side of supraocular spine. Me-
dian band of same intensity on epistome and
labium. Paler oblique lateral band on basal an-
tennular article. Some flecks of red on merus of
chelipeds and cross banding on fingers of some
individuals.

Munida longipes A. Milne Edwards 1880.
Northern limit extended from off Cape Lookout,
NC (Williams 1984) to Baltimore Canyon.

Munidopsis cf. transtridens Pequegnat and Pe-
quegnat 1971. Munidopsis transtridens is
known only from the holotype female taken in the
southeastern Gulf of Mexico at 1,280 m. The spec-
imens reported here from Baltimore, Hendrick-
son, and Lydonia Canyons off New Jersey and
southern New England, 906-1,425 m, are all
males. They resemble M. transtridens but differ
from it in rostral characters (both longer and
shorter, variably narrower or broader, in degree
of lateral convexity) and in having chelae strik-
ingly larger than the slender ones of the holotype.
Although these differences may be attributable to
sexual dimorphism, provisional identification
seems best until more material is available for
study.

Figure 3. — Munida forceps, male. Dorsal view of carapace, dia-
grammatic representation of lavender bands on salmon ground
color, carapace length 17.7 mm to base of supraocular spine.

SUPERFAMILY XANTHOIDEA

Family Goneplacidae

Chacellus filiformis Guinot 1969. Geographic
range extended from the northern Gulf of Mexico
east of the Mississippi River Delta and off the east
coast of Florida, 328-400 m (Guinot 1969), to Bal-
timore and Lydonia Canyons, 160-244 m.

Goneplacid crabs were driven out of secondary
burrows in walls of larger burrows constructed by
tilefish, Lopholatilus chamaeleonticeps Goode and
Bean, with rotenone. The poison did not kill the
crabs but caused them to emerge from the burrow
systems enough that they could be collected by

WILLIAMS: DECAPOD AND EUPHAUSIID CRUSTACEANS

"slurp gun". These burrow systems in Pleistocene
clay, referred to by Warme et al. (1978) and
Cooper and Uzmann (1980) as "Pueblo Villages",
shelter a number of invertebrate and vertebrate
species (Able et al. 1982; Bowman 1986; Grimes
et al. 1980a, 1980b, 1986). Goneplacids are rare in
collections made from surface vessels probably
because trawls or grabs cannot efficiently sample
the burrow systems in which these crabs have
been observed. Galatheids from shallower bur-
rows are more open to capture by conventional
means (Churchill B. Grimes^).

To the brief color description quoted from
Chace by Guinot (1969), the following can be
added from notes made 14 October 1981 on ma-
ture males, females, and juveniles that were pre-
served in formalin 25 July 1981, and personal
communication from Churchill Grimes (fn. 2).
Carapace dorsally spotted with red on off-white
background. Same type of spots on pterygosto-
mian, subocular, epistomial, and subbranchial
areas, on external maxillipeds, and on merus, car-
pus, and propodus of chelipeds (dorsally, later-
ally, and mesially). Spots tending to coalesce
along front of carapace and on chelae. Red color
more diffuse on dorsal or exposed surfaces of
walking legs, becoming more distinct and intense
with increasing size. Fingers of chelae black.
Dactyls of walking legs white except for darkened
tips, but setae pinkish. There is some variation in
pattern on individual crabs.

There is variation also in the length of the male
first pleopod, both in the USNM series of speci-
mens studied by Guinot (1969) and in the new
material reported here. In some specimens of the
latter, this appendage exceeds or at least reaches
the distal edge of the telson, whereas it is shorter
in specimens previously reported from localities
further south. In the latter, the third abdominal
segment is more angled laterally than in speci-
mens from the north. Thus, there seem to be some
differences between the northern and southern
populations.

SUPERFAMILY GRAPSIDOIDEA

Family Grapsidae

Euchirograpsus americanus A. Milne Edwards
1880. Geographic range extended north from off

^Churchill B. Grimes, Southeast Fisheries Center Panama
City Laboratory, National Marine Fisheries Service, NOAA,
.3500 Delwood Beach Road, Panama City, FL .32407-7499, pers.
commun. February 1982.

Oregon Inlet, NC (Williams 1984) to Oceanogra-
pher Canyon at the edge of Georges Bank and
nearby continental slope at 155-200 m.

SUPERFAMILY PINNOTHEROIDEA

Family Pinnotheridae

Dissodactylus juvenilis Bouvier 1917. The
ovigerous female from the Gulf of Mexico off west-
ern Florida, though similar in general features to
D. juvenilis, is very large for that species. In a
recent review of the genus Dissodactylus , Griffith
(1987) reported D. juvenilis from north of Yu-
catan and the Mississippi Delta. Members of the
genus are found in association with clypeastroid
echinoids (Schmitt et al. 1973), as was this speci-
men in a sample that included Clypeaster ravenel-
lii A. Agassizi.

ACKNOWLEDGMENTS

The following persons brought these records to
my attention through requests for identifications:
Barbara Hecker and Dennis T. Logan, Lamont-
Doherty Geological Laboratory, Columbia Uni-
versity, Palisades, NY; Kenneth W. Able, Center
for Coastal and Environmental Studies, and
Churchill B. Grimes, Department of Horticul-
ture and Forestry, Rutgers University, New
Brunswick, NJ (CBG now with Southeast Fish-
eries Center Panama City Laboratory, National
Marine Fisheries Service, NOAA, Panama City,
FL). Maureen E. Downey identified the echinoid.
Ruth E. Gibbons drafted the maps and Keiko Hi-
ratsuka Moore figured the galatheid. The
manuscript was critically reviewed by B. B. Col-
lette and D. L. Felder.

LITERATURE CITED

Able, K W , C B Grimes, R A Cooper, and J. R Uzmann

1982. Burrow construction and behavior of tilefish,
Lopholatilus chamaeleonticeps, in Hudson Submarine
Canyon. Environ. Biol. Fishes 7(3):199-205.
Able. K. W., D. C. Twitchell. C. B. Grimes, and R. S. Jones.
1987. Tilefishes of the genus Caulolatilus construct
burrows in the sea floor. Bull. Mar. Sci. 40:1-10.
Bowman. T H.

1986. Tridentella recava , a new isopod from tilefish
burrows in the New York Bight (Flabellifera: Tri-
dentellidael. Proc. Biol. Soc. Wash. 99(2):269-273.
Chace. F A, Jr

1972. The shrimps of the Smithsonian-Bredin Caribbean
Expeditions with a summary of the West Indian
shallow-water species (Crustacea: Decapoda:
Natantiaj. Smithson. Contrib. Zool. 98:1-179.

FISHERY BULLETIN: VOL. 86, NO 1

Cooper. R A , and J R Uzmann.

1980. Ecology of juvenile and adult Homarus. In J. S.
Cobb and B. F. Phillips (editors), The biology and man-
agement of lobsters. Vol. II, Ecology and management,
ch. 3, p. 97-142. Acad. Press, NY.
Griffith. H

1987. Taxonomic revision of the genus Dissodactylus
(Crustacea: Brachyura: Pinnotheridae). Bull. Mar. Sci.
40:397-422.
Grimes, C B , K W Able, and R S Jones.

1986. Tilefish, {Lopholatilus chamaeleonticeps), habitat,
behavior and community structure in mid-Atlantic and
southern New England waters. Environ. Biol. Fishes
15(41:273-292.
Grimes, C B . K W Able, and S C Turner

1980a. A preliminary analysis of the tilefish, Lopholatilus
chamaeleonticeps fishery in the Mid-Atlantic
Bight. Mar. Fish. Rev. 42(11):13-18.
Grimes, C G , K W Able, S C. Turner, and S J Katz

1980b. Tilefish: Its continental shelf habitat. Under-
water Nat. 12(41:34-38.
Guinot, D

1969. Recherches preliminaires sur les groupements na-
turels chez les Crustaces Decapodes des Brachyoures.
VII. Les Goneplacidae (suite et fin). Bull. Mus. Natl. Hist.
Nat. Ser. 2, 41(3):688-724.
HOLTHUIS, L B

1947. The Decapoda of the Siboga-Expedition. Part IX.
The Hippolytidae and Rhynchocinetidae collected by the
Siboga and Snellius Expeditions with remarks on other
species. Siboga-Exped. Monogr. 39aS, 100 p.
1955. The recent genera of caridean and stenopodidean
shrimps (Class Crustacea, Order Decapoda, Supersection
Natantia) with keys for their determination. Zool.
Verb. (Leiden) 26:1-157.
PEQUEGNAT, W E , AND L H Pequegnat

1971. New species and new records o{ Munidopsis (Deca-
poda: Galatheidae) from the Gulf of Mexico and
Caribbean Sea. Texas A & M Univ. Oceanogr. Stud., 1
(suppl.):3-24.
Schmitt. W L , J C McCain, and E S Davidson

1973. Decapoda I, Brachyura I, Fam. Pinnotheridae. In

H.-E. Gruner and L. B. Holthuis (editors), Crustaceorum
Catalogus 3, 160 p. Dr. W. Junk B. V.-Den Haag.

Squires, H J

1965. Decapod crustaceans of Newfoundland, Labrador
and the Canadian eastern Arctic. Fish. Res. Board
Can., Manuscr. Rep. Ser. (Biol.) 810, 212 p.

UCHUPI, E.

1965. Map showing relation of land and submarine topog-
raphy. Nova Scotia to Florida. U.S. Geol. Surv. Misc.
Geol. Invest., Map 1-451 (Sheets 1-3).

Veatch, a C, and P a Smith

1939. Atlantic submarine valleys of the United States
and the Congo Submarine Valley. Geol. Soc. Am. Spec.
Pap. 7, 101 p.

Warme. J E , R A Slater, and R A Cooper

1978. Bioerosion in submarine canyons. In D. J. Stanley
and G. Kelling (editors). Sedimentation in submarine
canyons, fans, and trenches, ch. 6, p. 65-70. Dowden,
Hutchinson & Ross, Inc., Stroudsburg, PA.

Wenner, E L

1982. Notes on the distribution and biology of Galathei-
dae and Chirostylidae (Decapoda: Anomura) from the
Middle Atlantic Bight. J. Crustacean Biol. 2(3):360-
377.

Wenner, E L , and D F Boesch

1979. Distribution patterns of epibenthic decapod Crus-
tacea along the shelf-slope coenocline. Middle Atlantic
Bight, USA. Bull. Biol. Soc. Wash. 3:106-133.

Wenner, E L , and N. T Windsor

1979. Parasitism of galatheid crustaceans from the Nor-
folk Canyon and Middle Atlantic Bight by bopyrid
isopods. Crustaceana 37(3):293-303.
Williams, A B.

1984. Shrimps, lobsters, and crabs of the Atlantic coast of
the eastern United States, Maine to Florida. Smithson.
Inst. Press, Wash., DC, 550 p.
Williams, A B , and R L Wigley

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

AGE AND GROWTH OF LARVAL GULF MENHADEN,
BREVOORTIA PATRONUS, IN THE NORTHERN GULF OF MEXICO

Stanley M. Warlen'

ABSTRACT

Experiments on laboratory-spawned and -reared larval gulf menhaden, Brevoortia patronus , showed
that they formed one otolith growth increment per day and that the increments could be used to
estimate their age. Wild larvae from collections in the northern Gulf of Mexico along three transects
I Cape San Bias, Florida; Southwest Pass, Louisiana; and Galveston, Texas) were aged. Gompertz
growth equations were used to describe the relationship between age and standard length for larvae
collected at various locations, and in different seasons and years. MANOVA tests and subsequent
pairwise tests were used to test for differences among these growth curves. For the most extensive
data set (Southwest Pass, Louisiana), there were significant differences in growth between early
season (December) and late season (February) larvae. Early season larvae grew faster than late
season larvae. Growth of larvae also differed among December collections and among February
collections. The growth model for the pooled data for all wild larvae predicted that they grew from 2.4
mm SL at hatching to 20.4 mm SL at 62 days.

Gulf menhaden, Brevoortia patronus , is the most
abundant commercial finfish in the Gulf of Mex-
ico and, with 883,500 metric tons (t) landed in
1985 (U.S. National Marine Fisheries Service
1986); it constitutes the largest fishery in the
United States. Some aspects of the oceanic early
life history of this clupeid are known and are re-
viewed by Turner (1969), Christmas and Waller
(1975), Lewis and Roithmayr (1981), Govoni et al.
(1983), and Shaw et al. (1985a). However, virtu-
ally nothing is known about the age and growth
of the larvae, much less how these parameters
vary spatially and temporally. Daily growth in-
crements on otoliths of larval fishes can be used
as an indicator of their age, and once the use of
this technique, first described by Pannella ( 1971 ),
is validated for the larvae of an individual spe-
cies, their ages can be estimated with confidence
and growth rates can be determined. Intraspecific
growth may be compared for larvae from different
areas and seasons (Lough et al. 1982), and from
this it may be possible to ascertain how biotic and
abiotic environmental variables affect larval
growth and survival. The objectives of this study
are to 1) validate the periodicity of increment for-
mation in otoliths of larval gulf menhaden, 2)
estimate larval growth rates, 3) compare growth
rates of larvae from different locations and times,
4) estimate spawning times, and 5) examine pos-

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

Manuscript accepted September 1987
FISHERY BULLETIN. VOL. 86, NO. 1, 1988.

sible relationships between larval growth and
surface water temperature. This work was part of
a larger project designed to investigate the early
life history of several economically important
fishes and the marine planktonic food webs that
support their growth and survival in the northern
Gulf of Mexico.

METHODS

Spawning and Larval Rearing

Adult gulf menhaden were collected near Gulf
Breeze, FL, and transported to the Beaufort Lab-
oratory, Beaufort, NC (Hettler 1983). After a pe-
riod of acclimation, adults were induced to spawn
in the laboratory. The resultant larvae were used
in experiments to validate the periodicity of in-
crement formation on their otoliths and the age at
first increment formation.

Beginning February 1983, several thousand
newly spawned gulf menhaden eggs were trans-
ferred to a tank containing 90 L of filtered sea-
water. The static water in this tank, kept at
20.5° ± 0.5°C throughout the experiment, was
continuously aerated and the salinity maintained
at 31 ± 17cc. Photoperiod was 12 hours light:12
hours dark. A food concentration of 25 rotifers
(Brachionus plicatilis) mL"^ was maintained. A
green alga, Nanochloris sp., was added periodi-
cally as food for the rotifers and to aid in remov-
ing toxic metabolites. The otoliths of larvae sam-

KISHERY BULLETIN; VOL. 86, NO. 1

pled at 10-, 17-, 24-, and 31-d posthatch were ex-
amined.

In January 1984 additional larvae were reared
to compliment results of the earlier experiment.
Smaller tanks with 60 larvae in 10 L of filtered
water were used. Experimental conditions were
the same as for the first experiment. The otoliths
of larvae sampled at 7-, 14-, and 20-d posthatch
were examined.

Larval Collections

Larval gulf menhaden were collected in the
northern Gulf of Mexico during six cruises of the
RV Oregon II. Sampling stations (Fig. 1) along
transects LA (off Louisiana) and FL (off Florida)
were occupied during 11-19 December 1979, 5-
15 February 1980, and 2-12 December 1980 and
along transects LA, FL, and TX (off Texas) during
9-24 February 1981, 2-13 December 1981, and
4-16 February 1982. Transect LA is near the
Mississippi River outflow off Southwest Pass, LA;
transect FL is southwest of Cape San Bias, FL;
and transect TX is located off Galveston Bay, TX.
Sampling stations were in water depths of 18, 91,
and 183 m except off Texas where only the 18 and
91 m depths were sampled.

A multiple opening-closing net and environ-
mental sensing system (MOCNESS) as described
by Wiebe et al. ( 1976) were the primary sampling
gear used to capture larvae. Additional samples
were taken in oblique tows with a 60 cm bongo

frame also fitted with 505 fjim mesh nets. Samples
were collected day and night and were preserved
in 95Vf ethanol (final concentration ^757^ ) within
5 minutes of collection. The ethanol was changed
in all samples at least once after initial preserva-
tion to prevent dissolution of otoliths in fish from
any samples that may have been inadequately
preserved. Data from larvae collected at all sta-
tions within a transect were combined for that
transect.

Estimating Age and Growth

All gulf menhaden larvae were measured to the
nearest 0.1 mm standard length (SL). The largest
otolith pair (sagittae) was teased from the sur-
rounding tissue, cleaned in distilled water, and
then placed on a glass microslide under a thin
layer of Flo-Texx-^ mounting medium.

Otoliths were viewed with a compound micro-
scope fitted with a television camera. Growth in-
crements were counted from otolith images on a
video monitor at magnifications of at least 400 x.
An increment appeared as a light, wide incremen-
tal band and a dark, narrow, discontinuous band
(Tanaka et al. 1981). Increments were generally
clearly discernable and easily counted (Fig. 2).
Estimated age was the number of increments
counted plus an empirically derived value for the

2Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.

96 00' 94° 00' 92° 00' 90° 00' 88° 00' 86° 00' 84° 00'

30° 00'

28° 00'

26° 00'

Figure l. — Location of sampling sites from which larval gulf menhaden were collected during crui-ses of the RV Oregon II in

December 1979-81 and February 1980-82.

WARLEN; AGE AND GROWTH OF LARVAL GULF MENHADEN

Figure 2. — Photomicrograph of a saggital otohth with 22 increments from a 17.4 mm SL
field collected larval gulf menhaden. Scale bar represents 10 pim. Growth increments
appear as pairs of wide incremental and narrow discontinuous bands.

number of days from spawning to first increment
formation. Results of the laboratory experiments
established the periodicity of otolith increment
formation.

A spawning date was assigned each ageable
larva by using the estimated age of the fish in
days to back-calculate from the date of capture. It
was assumed that there were no differences in
either the age at initial increment deposition or
the otolith increment deposition rate between lo-
cations and seasons and that the rate was not a
function of temperature, food, or photoperiod.

Average growth of larvae was described by the
Laird version (Laird et al. 1965) of the Gompertz
growth equation (Zweifel and Lasker 1976) fitted
to estimated age and size at time of capture for
fish from all cruises and transects. To stabilize
the variance of length over the observed age in-
terval, length data were log-transformed and
model parameters were estimated from the log-
transformed version of the growth equation. The
model was fit to data for each transect within
each cruise and for pooled data from all cruises.

Potential differences in the overall growth
curves among years and between seasons for lar-
vae caught off Louisiana and between years ( 1981
and 1982) for larvae caught off Louisiana and

Texas were examined by treating the parameters
of the Gompertz equation as dependent variables
in two-way multivariate analysis of variance
(MANOVA) designs. A one-way MANOVA de-
sign was used to test for differences
among transects (LA, FL, TX) within one season
(February 1982). Following significant
MANOVA results, prespecified pairwise
Hotelling's T'^ test comparisons (Bernard 1981, as
modified by Hoenig and Hanumara 1983) were
made using the Bonferroni procedure (Harris
1975) to provide conservative tests of statistical
significance. Bonferroni critical values for these
individual tests were equal to the overall error
rate (significance level = 0.05) divided by the
number of possible comparisons in the particular
MANOVA design. The emphasis in the compari-
sons was to look for overall differences in the
growth of larvae using these statistics as a guide
and not to look for differences in individual
parameters of the growth models.

Hotelling's T^ test and MANOVA both require
that the data fit a multivariate normal distribu-
tion and that the variance-covariance matrices of
the populations are not different (Harris 1975).
These assumptions are difficult to test and are
almost certainly not valid for real data sets (par-

FISHERY BULLETIN: VOL 86. NO 1

ticularly field data), but they may be nearly valid
for many sets of data (Harris 1975). No direct test
of normality in a trivariate, joint probability dis-
tribution is available (Bernard 1981), but bias
arising from nonnormal, multivariate, joint dis-
tributions is minimized with large sample sizes
(Bernard 19811. While methods are available to
test the hypothesis of equal variance-covariance
matrices (e.g.. Box's modification of Bartlett's
test), these methods are very sensitive and even
minor differences between group dispersions will
likely be discovered (Pimentel 1979). In any
event, the use of MANOVA in this paper relies on
variance-covariance matrices estimated from
nonlinear regressions, and these are not
amenable to testing. However, both MANOVA
and Hotel ling's tests are extremely robust even
under violation of the assumptions of homo-
scedasticity and multivariate normality (Harris
1975).

RESULTS

Increment Formation

The age of gulf menhaden at formation of the
first otolith growth increment was estimated
from laboratory-reared larvae. The intercept (2.6

days) of the regression of the number of growth
increments on known posthatch age of 36 larval
gulf menhaden (Fig. 3) was used to estimate
posthatch age at formation of the first increment.
This value was added to the time from spawning
to hatching which at 20°C is 2 days (Powell'M.
This sum (4.6 days) is the estimated time from
spawning to formation of the first increment.
Hence, it was necessary to add 5 days to each
increment count to estimate the age of larval gulf
menhaden from spawning.

The periodicity of increment formation was as-
certained from the regression of the number of
growth increments on the known age (Fig. 3). The
slope did not differ significantly (/-test, P < 0.05)
from 1.0, and thus, on the average, one otolith
growth increment was formed per day in
laboratory-reared larvae up to 31 days after
hatching. Results of a second experiment (Table
1 ) confirmed this periodicity. The age of gulf men-
haden larvae estimated from otolith increment
counts ( + 5) closely approximated the known ages
of 51 laboratory-reared larvae. Mean estimated
age of larvae differed by < 1 day from the known

3A. B. Powell, Southeast Fisheries Center Beaufort Labora-
tory, National Marine Fisheries Service, NOAA, Beaufort, NC
28516-9722, pers. commun. February 1986.

H
Z
111

liJ
DC
O

No. increments=-2.617+ 0.921(known age)
r= 0.928
n= 36

10 15 20 25

KNOWN AGE (days)

Figure 3. — Regression of the number of
growth increments on the known
posthatch age of 36 laboratory-reared
gulf menhaden. Standard error of the
slope IS 0.108.

WARLEN; AGE AND GROWTH OF LARVAL GULF MENHADEN

Table 1. — Standard length (mm) and estimated age (num-
ber of otolith growth increments +5) of laboratory-reared
larval gulf menhaden. Values in parentheses are 95% inter-
val estimates.

Known age Number of Mean estimated
(days) fish age

Mean SL

7
14
20

13
16
22

7.8 (±0.38)

13.2 (±1.14)

19.3 (±0.77)

4.7 (±0.28)
6.4 (±0.38)
8.0 (±0.42)

ages and the 957c confidence intervals included
the known age in each of the three groups. Some
of the variation in the number of gi'owth incre-
ments observed in known age larvae (Fig. 3,
Table 1) may have been due to 1) poor growing
conditions during rearing that could have re-
sulted in reduced growth in underfed larvae, 2)
variation in the inception of increment formation
as has been observed in other species (Laroche et
al. 1983; Fives et al. 1986), and 3) faintness of
growth increments in some larvae. In contrast,
increments on otoliths of field collected larvae
(Fig. 2) were usually very regular and distinct
and were more easily observed than those on
otoliths of laboratory-reared larvae. I assumed
that the gi'owth increment deposition rate was
also daily in wild larvae examined in this study.

Age and Growth of Larvae

Average growth of larval gulf menhaden dur-
ing their first two months of life was described by
the Gompertz growth model for pooled length at
age data for 2,003 fish representing collections
from all six RV Oregon II cruises (Table 2, Fig. 4).
Larvae ranged in age from 5 to 62 days (x=24A
days) and in SL from 3.4 to 28.0 mm (x = 12.6
mm). In the log-transformed model, age ac-
counted for 82% of the variation in length. Gulf
menhaden were predicted to have grown from 2.4
mm SL at hatching to 20.4 mm at age 62 days.
The size at hatching estimated from the Gom-
pertz equation was only slightly less than the
hatching size, 2.6-3.0 mm SL, observed in the
laboratory (Hettler 1984). Age-specific growth
rates declined from — 79f/day at age 10 days to
<0.4%/day at age 60 days. Maximum absolute
growth rate occurred when gulf menhaden larvae
were 7.9 mm SL and 13 days old.

The asymptotic length of larvae (21.5 mm SL),
determined from the variables of the growth
equation, is approximately the size when larvae
begin to transform into juveniles. This transfor-
mation, described by Lewis et al. (1972) for the

closely related Atlantic menhaden, B. tyrannus,
apparently ends when the fish reach 28-30 mm
SL (Suttkus 1956).

In all instances except one transect (TX Decem-
ber 1981, where there was no convergence in the
parameter values in the computer fitting proce-
dure and the model would not fit the data), the
Gompertz growth model could be used to describe
the growth of gulf menhaden larvae from each
cruise and transect (Figs. 5-7, Table 2). The
growth model for the FL December 1980 larvae
approximates an exponential form because of the
exceptionally low value for a. This may be due to
the preponderance of small, young larvae.

GROWTH COMPARISONS

Louisiana - Seasons and Years

There were statistically significant differences
(MANOVA, P < 0.001) in the growth curves for
larvae caught off Louisiana for two seasons (De-
cember, February) and three years (1979-80,
1980-81, 1981-82). To determine if differences ex-
isted between seasons in each year and among
any two years within each season, I selected 9 of
the possible 15 pairwise comparisons for testing.
The Bonferroni critical value in these tests was
0.0033 (0.05/15). The inability to fit a Gompertz
growth model to the TX December 1981 data pre-
cluded a comparison with the larvae collected off
Texas in February 1982.

Pairwise comparisons for within years data for
Louisiana larvae showed significant differences
(P < 0.003) in growth curves between early sea-
son (December) and late season (February) for
each year. Faster growth of early season larvae is
evident if the respective curves (Figs. 5; 6a, c; 7a,
c) are compared. For any age, the predicted size is
greater for early season than for late season lar-
vae. Only for the third year did the length at age
40+ days of February-caught larvae exceed that
for December-caught larvae.

In similar comparisons for larvae caught off
Louisiana in December of all three years, there
were significant differences (P< 0.003) in the
growth curves (Figs. 5a, 6a, 7a) for any two years.
As judged by the predicted size at any age, larvae
appeared to grow faster in 1979 than in either

1980 or 1981. While the curves for the 1980 and

1981 larvae overlapped, larvae from 1980 were
larger at 30+ days than were the 1981 larvae.
Significant differences were also found among the
curves (Figs. 5b, 6c, 7c) for larvae caught in

F1SHF:KV BUI.l.KTlN VOL Hli, NU 1

Table 2.— Estimates of Gompertz growth model parameters and mean age (days) and mean SL (mm) for
larval gulf menhaden collected m the northern Gulf of Mexico during the winters 1979-80, 1980-81, and
1981-82. fl2 is the coefficient of determination for the respective models.

Number
fish

Growth model parameters^

fvlean
estimated

tVlean

Date

Transect'

aged

'-(O)

A(0)

age (d)

SL (mm)

Winter 1979-80

Dec. 1979

0863

2 768

(1.270)

0 1701
(0,0711)

00809

(00190)

28 1

174

Feb. 1980

324

0954

2.131
(0 138)

0 1592
(00105)

00710
(0 0031)

302

12.7

Winter 1980-81

Dec. 1980

191

0931

2888
(0.188)

0 1125
(0,0097)

00496
(0.0044)

220

118

Dec. 1980

0701

3418
(0.994)

00561
(0 0370)

0.0001
(0,0407)

14 9

Feb. 1981

338

0849

2702
(0.231)

0 1159
(0 0120)

00577
(0,0049)

246

11.7

Feb. 1981

223

0526

5839
(0.579)

00305
(00088)

0,0125
(0,0110)

21,5

10.3

Winter 1981-82

Dec. 1981

370

0921

0384
(0.076)

04780
(0,0433)

0 1240
(0,0054)

21 7

13.3

Dec. 1981

114

—

25,4

11.9

Feb. 1982

191

0,736

0807
(0,337)

02729
(0,6093)

0,0851
(0,0090)

31 2

15.5

Feb. 1982

0.624

2278
(0,824)

0 1067
(0,0454)

0,0394
(0,0205)

17,6

Feb 1982

0946

1 798
(0,392)

0,1276
(0,0340)

00500
(0,0142)

225

100

All years

All data pooled

2,003

0822

2,355
(0.098)

0,1345
(0,0059)

0,0608
(0,0020)

24.4

12.6

iLA=IVIississippi River Delta (Southwest Pass Louisiana); FL = Cape San Bias, FL; TX = Galveston, TX.
2L|0| = length at hatching, A(0) = specific growth rate at hatching, « = exponential decline in A,o)- Values in
parentheses are estimated standard errors from the nonlinear regressions.
3Gompertz growth model did not fit the data.

E
E

X
»-

Z
lU

Figure 4.— Growth of gulf menhaden
larvae collectccl in the winters of 1979-
80, 1980-81, and 1981-82 m the north-
ern Gulf of Mexico. The Gompertz growth
model was used to describe the pooled
data. Two through nine coincident data
points are labelled with their numeral.
Coincident points of 10 and above are la-
belled A, B, etc.

1 1

3
1 1 1
1 1

1 1

2 2231 1 1111 1 11 1121

1 112 11 13313221331
21 4 222635231 8 3687ia<T 2 1
224235B253223445«7B32 2
8562 22 82 5 I 2J-9r536831 221
122 B8C49221AS3251 623352 11
1 1 780DB425*35253342 1214 1
4 5238 1 CBie58 31314 23
327586F9NBC65 6272331 12
34EB9A56Q7H561311 11 1

1 274E6E>C6K5H28 111 1
2194EaC6546383331 1
55E5BaC665A2A2 1

" ::: _ 31 1 i

L59422

-0.0608t

n= 2,003

ESTIMATED AGE (cJays)

WARLEN; AGE AND GROWTH OF LARVAL GULF MENHADEN

E
E

I
(-

cr
<

1 1 >

' 12^
''^1 1 1

1
1

A 1 1

' \

Louisiana

Dec 1979

1 1

50 60

10-

11 11 12

1 2 2232221 1 123 11

114"

" ' 1

322 1 1
833122
1 32
212 1

223
121

Louisiana
Feb 1980

ESTIMATED AGE (days)

Figure 5. — Growth of larval gulf menhaden collected in the winter
1979-80 in the northern Gulf of Mexico. The Gompertz growth model
was used to describe the data. Coincident data points are labelled as in
Figure 4.

February in three years. Larvae caught in 1980
grew faster than larvae caught in 1981 and
larvae caught in 1982 up to 25 days, thereafter
1980 and 1982 had very similar size at age esti-
mates.

Louisiana vs. Florida vs.
Texas - February 1982

There were no significant differences
(MANOVA, P = 0.212) among the growth curves
for larvae caught in February 1982 off Louisiana,
Texas, and Florida (Figs. 7c, d, e), and hence no
pairwise comparisons were necessary.

Louisiana vs. Texas - February
1981-82

Statistically significant differences (MAN-

OVA, P < 0.002) in larval growth were observed
in the LA and TX transects from February 1981-
82. Pairwise comparisons indicated significant
differences (P < 0.008) in the growth of larvae
collected off Texas in 1981 (Fig. 6d) and 1982 (Fig.
7e) and in growth between LA 1981 (Fig. 6c) and
TX 1981 collections. The earlier pairwise com-
parisons had already shown a significant differ-
ence in growth of larvae from LA February 1981
and February 1982 collections (Fig. 7c), but none
for growth of larvae caught in the LA February
1982 and TX February 1982 collections. Two
other potential tests, between transects of differ-
ent areas and different years, were not considered
to be meaningful.

Larvae caught off Louisiana in February 1981
were larger at age 18+ days than were larvae
caught off Texas in February 1981, and might be
considered to be faster growing fish. There was a

FISHERY BULLETIN: VOL 86, NO 1

E
E

cr
<

Q
Z
<

I-

Louisiana
Dec 1980

Louisiana
Feb 1981

Florida
Dec 1980
-L, I

1 1
1 1

2 2
31J3512 1
1 4 244
2 126263M?22
3122ttaft?7311
511 1
__J3323 23 1
2 12131
11112

Texas
Feb 1981

ESTIMATED AGE (days)

Figure 6.— Growth of larval gulf menhaden collected in the winter 1980-81 in the northern Gulf of Mexico. The Gompertz growth
model was used to describe the data. Coincident data points are labelled as in Figure 4.

statistical difference in growth of larvae caught
off Texas in February 1981 and 1982, and it ap-
pears that the 1982 larvae grew at a faster rate.
Conclusions from these statistical differences in-
volving TX February 1981 larvae collections
should be viewed with caution because of the rel-
atively poor fit (r^ = 0.526) of the model. Inade-
quacies, such as the lack of larvae <13 or >31
days old, in that data set probably resulted in the
relatively poor parameter estimates (Table 2).
Additional sampling would be necessary to fur-
ther test the hypotheses of differences in growth
between geographic areas in the northern Gulf of
Mexico and between years for Texas larvae.

Estimated Spawning Times

Gulf menhaden larvae observed in this study
were estimated to have been spawned from mid-
October to mid-February (Fig. 8). The limited ex-
tent of seasonal sampling precluded estimation of
the probable total range of the spawning season.
Most larvae captured in December and February
had been spawned in November and January re-
spectively (Fig. 8). The considerable overlap in
spawning times of larvae caught the same month
in different years is a reflection of the similarity
of sampling dates. The relatively narrow distribu-
tion of spawning dates for larvae caught off Flor-

WARLEN; AGE AND GROWTH OF LARVAL GULF MENHADEN

15-

10-

Louisiana
Dec 1981

41 1
173 721
1 1252925
1 21435131
1 13262

112
122
611
3

Texas
Dec 1981

1 1

1
1 1 1

1 1 1512 21;

21246212 "
2 1611
12 51 i^21 1 t
r4121 1 1
1212 ^\ 14 31
nil 222

21 1
2 1 1

Louisiana
Feb 1982

Texas
Feb 1982

J-

10 20 30 40

ESTIMATED AGE (days)

Florida
Feb 1982

10 20 30 40 50

ESTIMATED AGE (days)

Figure 7. — Growth of larval gulf menhaden collected in
the winter 1981-82 in the northern Gulf of Mexico. The
Gompertz model was used to describe the data from all
transects except Texas December 1981 where it could
not be made to fit the data. Coincident data points are
labelled as in Figure 4.

ida in both December 1980 (Fig. 6b) and in Febru-
ary 1982 (Fig. 7d), off Louisiana in February 1982
(Fig. 7c), and off Texas in February 1981 (Fig. 6d)
represent larvae from fewer cohorts.

DISCUSSION

Laboratory observations indicate that larval
gulf menhaden on the average form one growth
increment per day on their otoliths and that
counts of these increments can be used to esti-
mate age. Otoliths of larval gulf menhaden are

thin and round, and the increments are generally
easily counted and consequently are ideally
suited for ageing. The most closely spaced incre-
ments, those occurring near the focus, were at
least 1.5 |xm wide and were above the 0.2 jxm
resolution of the light microscope (Campana and
Neilson 1985). First increment formation occurs
about 5 days after spawning and probably coin-
cides with first exogenous feeding. This is sup-
ported by Hettler ( 1984) who found that gulf men-
haden eggs hatched at about day 1.7 at 19°-20°C.
Four days after hatching larvae had functional

FISHERY BULLETIN: VOL. 86, NO. 1

(42)

1979

D ILA

(324)

I — C

1980

LU
CC

Q.
<
O

(191)

(80

)l— {JH

rn — HLA

1980

(338)h
(223)1 —

1981

c±i ^L/

-Cp ITX

(88)1 IJHFL

1982

DEC I Jan I FEB '

SPAWNING DATE

Figure 8. — Schematic plots of the spawning times of larval gulf menhaden collected in the northern
Gulf of Mexico during 6 cruises of the RV Oregon II from December 1979 to February 1982. In each
distribution the vertical line is the median value and 50% of the data points fall within the block. Lines
beyond the boxes represent the range of data points. The value in parentheses to the left of each
distribution is the number of fish.

mouths and were 4.5 mm SL. However, develop-
mental rates are probably temperature depen-
dent (Powell and Phonlor 1986), and hence larvae
at lower temperatures would be older at first feed-
ing.

The Gompertz growth model appears to ade-
quately describe the growth of larval gulf men-
haden in most cases. Except where data are some-
what limited (Figs. 6b, d; 7b, d) the fit of the
model is relatively good and the r'^ is >0.73 for
each transect (Table 2). Gompertz gi'owth models
have been used (Zweifel and Lasker 1976; Methot
and Kramer 1979; Laroche et al. 1982; Warlen
and Chester 1985) to describe growth of larval
fishes where the length-age plots are nonlinear
and upper asymptotes were apparent.

Average growth rate of larval gulf menhaden to
day 60 was 0.30 mm/day throughout its oceanic

existence. This rate was very similar to that, 0.28
mm/day (estimated from figure 2 of Hettler 1984),
for larvae reared in the laboratory at 20° ± 2°C
for 60 days. However, wild-caught larvae were
from wider extremes in water temperature, with
mean early season (December) temperatures
from 17.4° to 21.2°C and late season (February)
12.9° to 16.4°C. The growth rate of larval Atlantic
herring, Clupea harengus , up to 50 days old was
similar and varied between 0.23 and 0.30 mm/day
(Lough et al. 1982). However, gulf menhaden lar-
vae grew slower than the fast growing but rela-
tively short-lived engraulids — bay anchovy, An-
choa mitchilli (Fives et al. 1986) and northern
anchovy, Engraulis mordax (Methot and Kramer
1979).

Only a small number of larvae from all the
collections were 2^50 days old. Larvae of this age

WARLEN; ACE AND GROWTH OK LARVAL C.rLF MENHADEN

were either not in the sampling area or were inac-
cessible to the fishing gear used. Although the
latter cannot be fully discounted, the former pos-
sibility is most likely, since larvae as they g:-ow
are known to be transported (Shaw et al. 1985b)
toward estuaries. Larvae are about 15-25 mm SL
(estimated from Suttkus 1956) when they enter
estuaries in Louisiana, and the smallest immi-
grating larvae are estimated from the growth
model <Fig. 4) to have been at least 30 days old.
Larvae then are probably 30-70 days old when
they enter Gulf of Mexico estuaries. This range is
very similar to that for Atlantic menhaden, B.
tyranniis, entering North Carolina estuaries (S.
M. Warlen, unpubl. data). The so-called "larval
drift" period for gulf menhaden is probably closer
to 4-10 weeks than the 3-5 weeks surmised by
Reintjes (1970).

Growth of larval gulf menhaden in the north-
ern Gulf of Mexico varied both spatially and tem-
porally. For three consecutive years there were
significant differences in gi'owth for early season
(December) and late season (February) larvae
caught off the Mississippi River Delta. The in-
crease in length for early season larvae was
greater than for larvae hatched in late season.
Environmental conditions in this area differed
between early season and late season. Mean
water temperature measured during the Decem-
ber 1979, 1980, and 1981 cruises were 17.4°, 19.4°,
and 21.2°C, respectively, while in February 1980,
1981, and 1982 the temperatures were 13.8°,
15.7°, and 14.7°C, respectively. Although not
shown experimentally for gulf menhaden larvae,
there is evidence that larvae of some marine
fishes grow faster at higher temperatures (Lau-
rence 1978; Laurence et al. 1981). Jones (1985)
also associated higher water temperatures with
higher growth of larvae and found that increase
in length of Atlantic herring larvae hatched early
in the season was greater than for larvae hatched
late in the season.

Growth rate of larvae caught in the same sea-
son but in different years was inversely related to
mean water temperature. Larvae caught in De-
cember off Louisiana showed a trend of higher
growth (1981 < 1980 < 1979) at lower respective
mean water temperatures (21.2°, 19.4°, 17.4°C);
similarly, the growth rates for larvae caught in
February (1981 < 1982 < 1980) was higher at
lower respective mean water temperatures ( 15.7°,
14.7°, 13.8°C). Other environmental factors in ad-
dition to temperature may also affect the growth
rate of larval menhaden. Food availability that

can be an important growth-limiting factor for
larval fishes, may be determining the relative
growth rates at the lower temperatures in Febru-
ary. On the basis of limited data on the zooplank-
ton (pelecypod larvae, copepod nauplii, cope-
podites, and adult copepods) that could serve as
food for gulf menhaden larvae (Govoni et al.
1983), food availability (number/100 m'^) is
highest in February 1980, lower in 1982, and low-
est in 1981. Analogous food abundance data is not
available for the December cruises, but levels
would probably need to be higher on a per fish
basis to support the higher metabolism concomi-
tant with the higher December mean water tem-
peratures (17.4°-21.2°C). The main and interac-
tion effects of growth-limiting (food abundance)
and growth-regulating (temperature) factors on
larval gulf menhaden growth still must be deter-
mined experimentally, preferably using labora-
tory-spawned and -reared larvae.

The apparent growth advantage enjoyed by
menhaden larvae spawned early in the season
(November) is only typical for a small part of the
population. The largest segment of the popula-
tion, those spawned in the peak months of Janu-
ary and February (Christmas and Waller 1975)
and that immigrated to estuaries in February-
April, typically had slower growth. Although
slower growing, larvae spawned later in the sea-
son may be more successful in reaching estuaries.
Guillory et al. (1983) found a negative relation-
ship between temperature and gulf menhaden re-
cruitment into Louisiana estuaries. The larger
estuarine recruitment later in the season may be
related to winter-early spring (January-April)
predominant west-northwest longshore flow of
coastal water within and just outside the coastal
boundary front producing longshore advective
transport and of lesser importance by episodic,
short-term cross-shelf advection associated with
cold fronts (Shaw et al. 1985b). They hypothesized
that longshore currents facilitated the movement
of larvae toward shore. Only for December did
they note a reverse flow (eastward) that would
not allow larvae to be transported toward estuar-
ies west of the Mississippi delta.

The between transect comparisons of growth
rate of larvae caught off Texas and Louisiana in
February showed a difference in 1981 but not in
1982. Again higher growth rates were associated
with higher mean water temperatures. Larvae
from the LA February 1981 sample (x water tem-
perature = 15.7°C) grew faster than larvae from
the TX February 1981 sample U water tempera-

FISHERY HULl.KTIN: VOL. 86, NO 1

ture = 12.0°C). Where no significant differences
in growth were found, i.e., between LA and TX
February 1982 samples, the respective mean
water temperatures were 14.7° and 14.4°C. Nei-
ther of those growth curves were significantly dif-
ferent from the curve for larvae from the FL
February 1982 sample where the mean water
temperature was 16.4°C. However, the paucity of
larvae >23 days old caught off Florida in Febru-
ary 1982 (Fig. 7d) suggests that comparisons of
that data set with the data sets for larvae caught
off Louisiana and Texas in February 1982 would
be of little value.

The estimated spawning period for gulf men-
haden extended from mid-October to mid-
February (Fig. 8). These results agree with Fore
(1970) and Christmas and Waller (1975) who,
using the occurrence of eggs and larvae, esti-
mated that gulf menhaden spawned from mid-
October through March. Gonad weight-body
weight ratios of adults (Lewis and Roithmayr
1981) and morphological and physiological fea-
tures of ovarian tissue (Combs 1969) also indicate
that spawning extends from October to early
March. Based on the movement of late larvae into
Lake Pontchartrain, Suttkus (1956) presumed
that gulf menhaden spawning began in October
and ceased in February. He suggested that the
beginning and end of the spawning period fluctu-
ates from year to year, and that there is no spawn-
ing activity during the spring and summer
months as Higham and Nicholson (1964) have
reported for the closely related Atlantic men-
haden.

Most larvae caught in December were spawned
in November (Fig. 8) regardless of the year. Lar-
vae caught in February were spawned mostly in
January but estimated spawning dates extended
from mid-December to mid-February. For any
given cruise, larvae from off Texas and Louisiana
were spawned at about the same time. There was
also considerable overlap in the spawning dates
in any cruise off Florida and the other areas. The
distribution of the central 507^ of spawning dates
from the Louisiana sample in February 1980 ex-
tended over a 29-d period and was wider than for
any other data set. This unusually wide distribu-
tion may have been due to the presence of two
distinct cohorts, one spawned in late December
and one in late January, collected on the Febru-
ary 1980 cruise. Combs (1969) found that this
species had intermittent total spawning. Lewis
and Roithmayr (1981) inferred that gulf men-
haden were intermittent, or fractional spawners.

Christmas and Waller (1975) noted a modal tem-
poral distribution of eggs in the region from the
Mississippi delta to east of Cape San Bias. Bal-
dauf sampled young menhaden in the lower
Neches River, TX, from November through April
and found two incoming populations from which
he suggested that there may have been two
spawning peaks. Only in the larval collections of
December 1981 did spawning date distribution
appear to be bimodal; 7 and 20 November for Lou-
isiana and 8 and 19 November for Texas. Future
sampling throughout the spawning season will be
necessary to determine the seasonal periodicity
and peaks of gulf menhaden spawning. Relative
numbers of larvae in cohorts within the spawning
season could then be compared with measure-
ments of environmental conditions as a test of the
match-mismatch hypothesis (Cushing 1975) and
to further test, as Methot (1983) has done,
whether larvae spawned during favorable envi-
ronmental periods constitute the greatest per-
centage of the year class.

ACKNOWLEDGMENTS

I thank the following persons of the Beaufort
Laboratory: M. Boyd who extracted, mounted,
and aged otoliths, W. Hettler who spawned gulf
menhaden and furnished eggs for the laboratory
experiments, A. Chester who advised on statisti-
cal procedures and problems, D. Ahrenholz and
A. Powell who reviewed early drafts of the
manuscript, and the crew and scientists on the
RV Oregon II cruises. P. Ortner of the Atlantic
Oceanographic and Meteorological Laboratories,
NOAA, Miami, FL, provided raw data from which
zooplankton counts were summarized. This re-
search was supported by a contract from the
Ocean Assessments Division, National Ocean
Service, NOAA.

LITERATURE CITED

Bernard. D R

1981. Multivariate analysis as a means of comparing
growth in fish. Can. J. Fish. Aquat. Sci. 38:233-236.
Campana, S. E., and J D. Neilson,

1985. Microstructure of fish otoliths. Can. J. Fish.
Aquat. Sci. 42:1014-1032.
Christmas. J Y , and R S Waller.

1975. Location and time of menhaden spawning in the

4Baldauf R. J. 1954. Survey and study of surface and sub-
surface conditions in and around Beaumont, Texas. Biological
survey of the Neches River in the region of Beaumont,
Texas. Texas A&M Res. Found., Mimeo. Rep., 184 p.

WARLEN; AGE AND GROWTH OF LARVAL GULF MENHADEN

Gulf of Mexico. Gulf Coast Res. Lab.. Ocean Springs,
MS, 20 p.
Combs. R M

1969. Embryogenesis, histology, and organology of the
ovary of Breioortia patronus . Gulf Res. Rep. 2:333-
434.

CUSHINO. D H

1975. Marine ecology and fisheries. Cambridge Univ.
Press. Cambridge, Eng., 278 p.
Fives. J M , S M Warlen. and D E Hoss

1986. Aging and growth of larval bay anchovy, Anchoa
mitchilli, from the Newport River estuary, North Caro-
lina. Estuaries 9(4B):362-367.
Fore. P L

1970. Eggs and larvae of the Gulf menhaden. In Report
of the Bureau of Commercial Fisheries Biological Labo-
ratory, Beaufort. N.C., p. 11-13. U.S. Fish Wildl. Serv.
Circ. 341.

GovoNi. J J , D E Hoss, AND A J Chester

1983. Comparative feeding of three species of larval fishes
in the northern Gulf of Mexico: Brevoortia patronus,
Leiostomus xanthurus, and Micropogonias undula-
tus. Mar. Ecol. Prog. Ser. 13:189-199.
GUILLORY. V . J GEAGHAN. AND J. ROUSSEL

1983. Influence of environmental factors on gulf men-
haden recruitment. La. Dep. Wildl. Fish. Tech. Bull. 37,
32 p.
Harris. R J

1975. A primer of multivariate statistics. Acad. Press,
N.Y., 332 p.
Hettler, W F

1983. Transporting adult and larval gulf menhaden and
techniques for spawning in the laboratory. Prog. Fish-
Cult. 45:45-48.

1984. Description of eggs, larvae, and early juveniles of
gulf menhaden, Brevoortia patronus. and comparisons
with Atlantic menhaden, B. tyrannus, and yellowfin
menhaden, B. smithi . Fish. Bull.. U.S. 82:85-95.

HiGHAM, J R . AND W R NICHOLSON

1964. Sexual maturation and spawning of Atlantic men-
haden. Fish. Bull., U.S. 63:255-271.

Hoenig. N a . AND R G Hanumara

1983. Statistical considerations in fitting seasonal growth

models for fishes. Cons. Int. Explor. Mer, CM. 1983/

D:25, 25 p.
JONES. C

1985. Within-season differences in growth of larval At-
lantic herring, Clupea harengus harengus . Fish. Bull.,
U.S. 83:289-298.

Laird. A K . S A Tyler, and A D Barton

1965. Dynamics of normal growth. Growth 29:233-
248.

Laroche. J L . S L Richardson, and A A Rosenberg

1982. Age and growth of a pleuronectid, Parophrys ve-
tulus, during the pelagic larval period in Oregon coastal
waters. Fish. Bull., U.S. 80:93-104.
Laurence. G C

1978. Comparative growth, respiration and delayed feed-
ing abilities of larval cod (Gadus morhua ) and haddock
{Melanogrammus aeglefinus) as influenced by tempera-
ture during laboratory studies. Mar. Biol. (Berl.) 50:1-
7.
Laurence, G C . A S Smigielski. T A Halavik. and B R
Burns

1981. Implications of direct competition between larval

cod {Gadus morhua I and haddock i Melanogrammus ae-
glefinus ) in laboratory growth and survival studies at
different food densities. Rapp. P. -v. Reun. Cons. int. Ex-
plor. Mer 178:304-311.

Lewis. R M . and C M Roith.mayr

1981. Spawning and sexual maturity of gulf menhaden,
Brevoortia patronus . Fish. Bull., U.S. 78:947-951.

Lewis. R M . E P H Wilkins, and H R Gordy

1972. A description of young Atlantic menhaden, Bre-
voortia tyrannus, in the White Oak River estuary, North
Carolina. Fish. Bull., U.S. 70:115-118.

Lough. R G . M Pennington. G R. Bolz. and A A Rosenberg

1982. Age and growth of larval Atlantic herring, Clupea
harengus L., in the Gulf of Maine-Georges Bank region
based on otolith growth increments. Fish. Bull., U.S.
80:187-199.

Methot. R D. JR

1983. Seasonal variation in survival of larval northern
anchovy, Engraulis mordax. estimated from the age dis-
tribution of juveniles. Fish. Bull., U.S. 81:741-750.

Methot, R.. D., Jr , and D Kramer

1979. Growth of northern anchovy, Engraulis mordax,
larvae in the sea. Fish. Bull., U.S. 77: 413-423.
Pannella. G

1971. Fish otoliths: Daily growth layers and periodical
patterns. Science (Wash., D.C.) 173: 1124-1127.
Pimentel. R a

1979. Morphometries, the multivariate analysis of biolog-
ical data. Kendell/Hunt Publ. Co., Dubuque. lA, 276

P-
Powell. A B , and G Phonlor

1986. Early life history of Atlantic menhaden, Brevoortia
tyrannus , and gulf menhaden, B . patronus . Fish. Bull.,
U.S. 84:991-995.
Reintjes. J W

1970. The gulf menhaden and our changing estuar-
ies. Proc. Gulf Caribb. Fish. Inst. 22:87-90.
Shaw. R F . J H Cowan. Jr , and T L Tillman

1985a. Distribution and density of Brevoortia patronus

(gulf menhaden) eggs and larvae in the continental shelf

waters of western Louisiana. Bull. Mar. Sci. 36:96-

103.

Shaw. R F . W J Wiseman, Jr . R E Turner. L J Rouse. Jr .

R E Condrey. and F J Kelly. Jr

1985b. Transport of larval gulf menhaden Brevoortia pa-
tronus in continental shelf waters of western Louisiana:
A hypothesis. Trans. Am. Fish. Soc. 114:452-460.
Suttkus, R D

1956. Early life history of the gulf menhaden. Brevoortia
patronus. in Louisiana. Trans. N. Am. Wildl. Conf
21:390-407.
Tanaka. R. Y Mugiya. and J Yamada

1981. Effects of photoperiod and feeding on daily growth
patterns in otoliths of juvenile Tilapia nilotica. Fish.
Bull., U.S. 79:459-466.
Turner. W R

1969. Life history of menhadens in the eastern Gulf of
Mexico. Trans. Am. Fish. Soc. 98:216-224.
US National Marine Fisheries Service

1986. Fisheries of the United States, 1985. U.S. Natl.
Mar. Fish. Serv. Curr. Fish. Stat. 8380, 121 p.
Warlen, S. M.. and A. J Chester.

1985. Age, growth, and distribution of larval spot, Leios-
tomus xanthurus, off North Carolina. Fish. Bull., U.S.
83:587-599.

FISHERY BULLETIN: VOL. 86, NO. 1

WiEBK. P H . K H Burt. S H Boyd, and A W Morton Zweifei,. J R . and R Lasker

1976. A multiple opening/closing net and environmental 1976. Prehatch and posthatch growth of fishes — a general

sensing system for sampling zooplankton. J. Mar. Res. model. Fish. Bull., U.S. 74:609-621.

34:313-326.

SOURCES OF VARIATION IN CATCH PER UNIT EFFORT

OF YELLOW! AIL FLOUNDER, LIMANDA FERRUGINEA (STORER),

HARVESTED OFF THE COAST OF NEW ENGLAND

LORETTA O BrIEN AND RALPH K. MaYO^

ABSTRACT

Factors affecting variability in commercial catch per unit effort (CPUE) of yellowtail flounder were
examined in order to establish a basis for standardizing fishing effort. Analysis of variance ( ANOVA)
procedures were employed to test for differences in CPUE among vessel tonnage class, fishing area,
and depth zone and the interactions between tonnage class and area, and tonnage class and depth.
Vessel tonnage class and fishing area accounted for highly significant IP < 0.01) sources of variation
in CPUE whereas depth was not significant (P > 0.05) in most cases. Interactions between tonnage
class and stock area were also highly significant in all cases. A series of annual fishing power
coefficients was computed for each tonnage class relative to a standard for each stock based on
parameter estimates obtained by fitting the CPUE observations to a linear model with tonnage class
as the independent variable. Deviations of annual fishing power coefficients from the 20-year mean
were found to exhibit significant first order autocorrelations. Consequently, annual coefficients were
computed over the entire 1964-83 period by incorporating tonnage class, annual and seasonal effects
as independent variables in a three-way linear model. Although the standardized CPUE estimates
obtained from this procedure are similar to those obtained by previous methods, the revised proce-
dures described in this paper insure adequate representation of all vessel classes engaged in the
yellovfcrtail fishery in the CPUE calculations.

Fishing effort and resulting catch per unit effort
(CPUE) indices are routinely used in assessing
the impact of commercial fishing operations on
stock abundance. The traditional concept that ag-
gregate CPUE indices may be used to measure
annual changes in relative stock abundance is
based on the principal assumption that the catch-
ability coefficient (q) either remains constant
over all fleet components, or that nominal effort is
adjusted to account for differences in relative effi-
ciencies (Pope and Parrish 1964; Kimura 1981).
Variation in q may be due to persistent differ-
ences in fishing power of various types of gear or
to technological innovations which may be gradu-
ally introduced over time (Gulland 1964; Sis-
senwine 1978). Biological interactions such as
changes in availability of a species due to sea-
sonal distribution patterns or to annual changes
in abundance may also affect the overall catcha-
bility of demersal species (Garrod 1964; Pope and
Garrod 1975). Variability in catchability coeffi-
cients may be taken into account by relating nom-
inal fishing effort of each fleet component to some
chosen standard category.

'Northeast Fisheries Center Woods Hole Laboratory, Na-
tional Marine Fisheries Service. NOAA, Woods Hole, MA
02543.

Manuscript accepted October 1987.
FISHERY BULLETIN; VOL. 86. NO. 1, 1988.

Numerous authors have described the basic
procedures for calculating relative fishing power
of various fleet components. Beverton and Holt
(1957) provided evidence to suggest that the dis-
tribution of logarithms of fishing power factor/
vessel tonnage ratios could be described by a nor-
mal curve while Gulland (1956) employed an
analysis of variance (ANOVA) model of log
CPUE. The properties of the ANOVA model were
further examined by Robson (1966) who extended
the techniques developed by Gulland (1956) and
formally specified the analysis of Beverton and
Holt (1957) as a two-way multiplicative ANOVA
model. Stern and Hennemuth (1975) employed
the method of Robson (1966) in their analysis of
fishing effort in the U.S. Georges Bank haddock
fishery using depth fished and vessel tonnage as
classification variables. In a previous study,
Rounsefell (1957) computed standardized log
CPUE indices to determine relative abundance of
several co-occurring species on Georges Bank ac-
cording to depth. More recently, Gavaris (1980)
and Kimura (1981) have developed modifications
of the ANOVA model to estimate annual stan-
dardized CPUE indices from time series of catch
and effort data by incorporating a year effect in
the model.

Standardized annual CPUE indices based on

I-ISIIKKY lUU.l.KTIN: VOL, 86, NO. 1

criteria established by Lux (1964) have been rou-
tinely used to monitor relative abundance of
three stocks of yellowtail flounder, Limanda fer-
ruginea (Storer), in the commercial fishery off the
New England coast (Fig. 1 ). Lux calculated CPUE
indices for otter trawlers ranging from <26 gross
registered tons (GRT) to 100 GRT based on trips
in which yellowtail flounder accounted for bQ^/< or
more of the total landed weight between 1942 and
1961. A fishing power coefficient was then com-
puted for each of several GRT categories as the
ratio of CPUE to a standard GRT category CPUE
for the entire timespan. A separate set of fishing
power coefficients was computed for each of the
three stocks. Lux's (1964) work improved upon an
earlier analysis of yellowtail flounder CPUE by
Royce et al. ( 1959) which was based only on rela-
tively small vessels ranging in size from 5 to 50
GRT that dominated the fishery during the
1940's.

Since 1964, numerous technological innova-
tions have drastically changed the character of
the New England fishing fleet as traditional side
trawlers have gradually given way to larger,
more efficient stern trawlers equipped with so-
phisticated electronic navigation and hydroa-
coustic devices. This gradual alteration in the
fleet characteristics over time suggests that pre-
viously documented relationships among vessel

categories may no longer be applicable to the cur-
rent fishery, and that use of nominal effort in
CPUE calculations will tend to overestimate rela-
tive abundance in the more recent years
(Westrheim and Foucher 1985). Long-term de-
clines in yellowtail flounder abundance on each of
the principal fishing grounds (Clark et al. 1984)
also indicate that current catchability coefficients
may differ from previous values. Accordingly, up-
dated fishing power coefficients are required to
adequately assess changes in effective fishing ef-
fort and CPUE which have occurred during the
past two decades. Further, to obtain annual effort
and CPUE estimates over such a broad period of
years, techniques for computing relative fishing
power should incorporate a time element in the
analysis.

In this paper we examine variation in CPUE
with respect to fishing area, depth, vessel tonnage
class, season, and year for three stocks of yellow-
tail flounder on Georges Bank, Southern New
England, and Cape Cod grounds between 1964
and 1983. Before evaluating differences in rela-
tive fishing power among vessel classes, we inves-
tigate potential interactions between tonnage
class and area and tonnage class and depth
within each year, and partition the data to mini-
mize tonnage class-area interactions. For each
stock, fishing power coefficients are examined for

68 67 66

Figure 1.— Yellowtail flounder stocks off the coast of New England (After Lux 1963).

O'BRIEN AND MAYO: CPUE OF YELLOWTAIL FLOUNDER

annual and seasonal interactions. A three-way
linear model, incorporating annual and seasonal
components, is employed to compute relative
fishing power coefficients over the entire 20-yr
period and estimate annual standardized CPUE
indices.

DISTRIBUTION OF CATCH AND
EFFORT

Commercial exploitation of yellowtail flounder
began in the late 1930's following the decline of
the winter flounder fishery (Royce et al. 1959).
Nominal catches^ for the three grounds combined
rapidly increased to 31,500 metric tons (t) in 1942
but subsequently declined to 5,500 t in 1954.
Landings by U.S. vessels gradually increased to a
record high of 36,900 t in 1963, but declined again
to 10,500 t in 1978 (Fig. 2). Distant water fleet
(DWF) catches were also substantial during this

2Nominal catch defined as live weight equivalent of landings,
excluding discards.

period, peaking at 20,700 t in 1969. Overall
catches from the three fishing grounds have re-
cently increased to 30,800 t in 1983, although
1984 landings declined to 15,500 t (Clark et al.
1984). The decline in catch during the 1940's was
not due to overfishing (Royce et al. 1959) but may
have been related to a warming trend in the re-
gion which affected recruitment (Sissenwine
1974). The more recent decline between 1969 and
1978, however, has been attributed to increased
fishing effort by both domestic and distant water
fleets (Brown et al. 1980).

In the early 1940's the size of vessels fishing for
yellowtail flounder varied from 5 to 75 GRT. The
predominant vessels on Southern New England
and Cape Cod grounds ranged from 26 to 50 GRT;
on Georges Bank, the dominant vessels were in
the 51-75 GRT range. By the mid-1960's larger
vessels had begun to enter the fishery, increasing
the maximum size to 215 GRT. During this period
the size range of the dominant vessels on South-
ern New England grounds and on Georges Bank
had increased to 51-72 and 73-104 GRT, respec-

55 n

O
O

C
<0

50-

'cJj
S 40

30-

Total

DWF

nVja-yy

1940

1945

1950

1955

1960 1965

Year

Figure 2. — Yellowtail flounder landings (metric tons) by United States and distant water fleet (DWF) vessels
from the combined Georges Bank, Southern New England, and Cape Cod grounds, 1940-84.

FISHKKY MUI.LKTIN: VOL. 86, NO 1

tively. Vessels fishing the relatively nearshore
Cape Cod grounds remained in the 34-50 GRT
range. Larger vessels continued to enter the fish-
ery during the 1970's and, by 1983, several ves-
sels were in the 311-400 GRT range. Since 1964,
vessel CRT's have been categorized by tonnage
class (TO as given in Table la.

A review of Lux's (1964) data from 1942 to 1961
and the distribution of more recent yellowtail
flounder landings from 1964 to 1983 (Figs. 3-5)
reveal that vessels of similar size have continu-
ally fished the same general areas over the past
40 years. The TC 21-24 vessels fish primarily on
Southern New England and Cape Cod grounds
(Fig. 3 1, although TC 24 vessels occasionally
enter the southwest part of Georges Bank. The
TC 25-33 vessels fish on both Georges Bank and
Southern New England grounds (Fig. 4), while
TC 41-43 vessels concentrate on Georges Bank
(Fig. 5). Although TC 41 vessels operate at times
on the eastern part of the Southern New England
grounds, the TC 42 and 43 vessels fish exclusively
on Georges Bank.

The distribution plots (Figs. 3-5) reveal a grad-
ual phaseout of smaller (TC 21-24) vessels on the
inshore Southern New England and Cape Cod
grounds and a concurrent increase in the activity
of large (TC 41-43) vessels on Georges Bank,
Southern New England, and, to a lesser extent,
on the Cape Cod grounds. In evaluating trends in
CPUE we must ask whether these changes in the
yellowtail fishery over the past 20 years (i.e., the
shift in the predominant vessel size on two of the

Table 1. — Definition of vessel tonnage classes (a) and depth
ranges and corresponding zones (b) included in analysis of vari-
ance of yellowtail flounder CPUE.

a Gross registered

tonnage

Vessel tonnage

(range)

class

5-10

11-15

16-22

23-33

34-50

51-72

73-104

105-150

151-215

216-310

311-400

b Depth range

Depth

(m)

zone

1-55

56-110

111-183

three grounds and the addition of larger vessels to
the fieet on all three grounds) affect CPUE as
calculated by the traditional method (Lux 1964).
If the same size range of vessels (5-100 GRT) had
fished for yellowtail flounder throughout the
years, a shift in the dominant vessel class would
not affect CPUE estimates since effort would be
standardized against the same class and is, there-
fore, relative. However, the maximum vessel size
has increased and the predominant TC now repre-
sents vessels larger than 100 GRT. Since landings
and effort data contributed by these larger vessels
were not incorporated into previous CPUE calcu-
lations, CPUE estimates will not necessarily rep-
resent overall fleet performance in recent years.
The following procedures, therefore, were devel-
oped to calculate new fishing power coefficients
that encompass the entire size range of vessels
currently in the fishery.

METHODS OF ANALYSIS

Catch and effort data recorded by trip were ob-
tained from Northeast Fisheries Center (NEFC)
detailed commercial landings files. Fishing effort
or days fished (df) is defined on a 24-h basis as
number of hours of actual fishing time divided by
24. Only trips landing 50% or more of yellowtail
flounder were analyzed; trips included within the
qualification level generally accounted for 70-
90% of the total yellowtail landings over the en-
tire period, except on Cape Cod grounds where
qualified trips accounted for 40-60% of the total.
Catch per day fished (CPUE) was computed for
each trip and transformed to In CPUE since pre-
liminary analysis indicated a positive correlation
of the arithmetic mean CPUE with the variance.
Use of the log transformation, however, stabilized
the variance and created a lognormal distribution
as noted by Gulland (1956) and Steel and Torrie
(1980).

Trips landing between 1964 and 1983 are clas-
sified in the data base by vessel tonnage class,
statistical area, and depth zone fished. Vessels
ranging in size from 5 to 310 GRT (Table la) and
statistical areas corresponding to the three major
stocks were selected for analysis as follows:
Georges Bank (areas 522-525), Southern New
England (area 526 and 537-539), and Cape Cod
(areas 514 and 521) (Fig. 6). Because of their spo-
radic representation, TC 21-23 vessels have been
excluded from the Georges Bank analyses and
have been combined as one category on the South-
ern New England grounds. Depth zones 1, 2, and

O'BRIEN AND MAYO CPUE OF YELLOWTAIL FLOUNDER

O — — — ■

■ "X^

to o

IT) O

o o o o
f\j m =r u

— LJ

o —
z a:
a: «-

O —

. . . .

o \r>

o o o
— nj m

• •»!*

ST^'^

1— (

bo
aj

c
c
o

en
>

CO
00
05

(8
05

O
05

CO
05

bo
C

■3

FISHERY BULLETIN: VOL 86, NO. 1

U-)

LlJ

—i

(_) rn

■w

Lu cn

O —

cr -■

z cc

z cr

O LiJ

t— >-

1—1

-a

r-l

oj o in o in o

— f\j i/l r- IT)

_ (_,

o —

Z (T

cr "-

—1 UJ

O l/i o o o

oo O Ul O

— Aj in

O UJ

03 ts^

z a

O UJ

bo
S

-a

3
o

c
_o
"-^

'u
in

D
O

O'BRIEN AND MAYO: CPUE OF YELLOWTAIL FLOUNDER

c
c

<u
>

CO
00
Oi

>H
a>
-a

P
2

FISHERY BULLETIN: VOL. 86. NO. 1

Figure 6. — Yellowtail flounder fishing grounds defined by U.S. statistical area are as follows: Southern New
England. 526-539; Georges Bank, 522-525; Cape Cod, 514 and 521.

3 (Table lb) were also selected from six possible
zones based on the bathymetric distribution of
yellowtail flounder.

Trip data were aggi'egated at different levels of
spatial resolution to examine variability in
CPUE over the entire region and within each of
the three established stocks. Two-way ANOVAs
with interaction were performed on annual data
sets using the BMDP statistical software progi'am
P4V (Dixon 1981). Given the large number of ob-
servations in each analysis, the more rigorous
99% significance level was chosen to test the null
hypothesis (no significant differences) since
relatively small differences in mean CPUE can
produce statistically significant results. The
ANOVA was performed initially to test for differ-
ences in CPUE among all tonnage classes and
statistical areas and to determine the overall ex-
tent of tonnage class-area interactions. Second-
ary analyses were performed to examine the ef-
fect of tonnage class-area and tonnage class-depth
interactions among and within each of the stocks.
All subsequent tests for significance of tonnage
class, area, and depth main effects were per-
formed with the interaction effects absorbed in
the error sum of squares. Estimates of annual
geometric mean CPUE were obtained by combi-
nations of tonnage class, stock area, and depth
from the row and column means provided by the
P4V software (Tables 2, 3).

A standard vessel class was selected for each of
the three stocks for use in calculation of fishing
power coefficients based on the prevalence of the
vessel class in the fishery and its relative contri-
bution to the landings over the 20 years. The TC
32 category was chosen as the standard for both
Georges Bank and Southern New England stocks,
and the TC 25 class was chosen as the standard
for the Cape Cod stock. Within each stock annual
fishing power coefficients were derived for each
tonnage class relative to the standard by fitting
In CPUE to a one-way linear model using the
GLM procedure of the Statistical Analysis Sys-
tem (SAS Institute 1982) as follows:

U = CC + ^ [^jXj] + ^.

Annual deviations of the coefficients from the
20-yr mean were tested for first order autocorre-
lation using the Durbin-Watson test statistic
(Neter and Wasserman 1974). A time component
was subsequently incorporated in the linear
model to account for annual trends in the data;
seasonal effects were also included by classifying
the data according to calendar quarter. The ini-
tial year (1964) and the fourth quarter were se-
lected as reference categories. The general model
is specified as:

O'BRIEN AND MAYO: CPUE OF YELLOWTAIL FLOl'NDER

Table 2. — Geometric mean CPUE^ (landings per day fished, metric tons) for Georges Bank, Southern New England, and Cape Cod

yellowtail flounder trips by vessel tonnage class, 1964-83.

Vessel tonnage

class

Year

Vessel tonnage class

Year

41 42

Georges Bank

1964

—

3.65

4.17

4.88

5.04

4.58

4.52

2.14

1974

—

1.08

1 62

231

2.14

2.07

2.64

1.30 —

1965

—

2.41

3.06

3.62

3.55

3.39

4.97

—

1975

—

0.57

1.57

1.13

1.29

1.33

1.55

1.72 —

1966

—

1.18

1.25

2.27

276

2.78

2.76

2.71

—

1976

—

0.72

0.90

0.53

0.86

1.13

1 32

1.69

2.06 —

1967

—

2.20

2.74

2.96

278

265

—

1977

—

0.93

1.73

1.44

1.77

1.61

1.82

1.98

1.81 —

1968

—

3.88

2.99

3.47

3.55

384

3.53

2.66

1978

—

2.72

1.15

1.58

1.52

2.04

2.26

282 —

1969

—

3.28

2.82

3.06

3.30

3.02

2.81

3.88

1979

0.79

1.74

2.28

1.59

2.17

2.32

2.93

2.88

3.64 —

1970

—

3.74

2.40

2.88

308

2.81

2.70

—

1980

—

1.02

246

1.91

2.54

2.68

2.81

3.07

3.00 —

1971

—

1.66

2.07

2.70

252

2.24

2.35

—

1981

—

1.88

2.50

1.43

2.58

2.76

2.61

3.28

2.23 —

1972

—

1.96

224

239

2.29

1.96

—

1982

249

1.96

2.42

2.29

3 10

3.20

3.72

3.98

3.97 —

1973

—

1.38

2.64

2.61

2.81

3.05

2.76

3.20

1983

8.81

1.74

2.51

3.35

3.43

3.27

3.65

3.57

3.79 —

1974

—

2.52

—

2.79

2.13

2.23

2.19

1.30

3.42

pane Cnd

1975

—

1.93

1.91

1.64

1.93

2.04

1.72

3.19

*»

^i^l^f u »

1976

1.35

1.69

1.77

1.84

2.16

2.06

1964

—

1.65

1.38

1.94

2.64

2.85

2.76

—

— —

1977

1.17

1.66

1.77

226

1.81

1965

—

1.45

1.20

1.59

2.16

3.56

2.12

2.07

— —

1978

1.67

1.50

1.72

2.03

282

1966

1.13

1.07

1.23

1.48

2.55

3.64

1.76

—

— —

1979

2.69

1.31

1.80

2.41

2.37

2.82

3.64

1967

—

0.78

0.90

1.56

2.46

3.21

1.95

2.62

— —

1980

0.30

1.09

0.90

2.88

2.62

3.27

3.00

1968

—

1.16

0.99

1.89

2.65

4.11

2.42

—

1.13 —

1981

1.12

1.24

1.98

2.48

2.61

223

1969

—

1.66

1.33

202

2.77

3.64

2.47

—

— —

1982

1.88

0.48

2.02

1.71

2.10

2.53

397

2.07

1970

—

1.03

1.05

2.34

2.33

3.64

26.96

234.18

2.91 —

1983

—

0.79

2.46

1.97

2.21

3.79

—

1971
1972

1.71

1.17
1.63

1.89
1.16

2.43
2.02

2.04
1.85

2.24
2.02

1.90
1.46

2.45
1.79

3.36 —
1.93 —

Southern New England

1973

1.02

1.32

1.06

1.87

1.99

1.90

1.91

1 63

284 —

1964

234

1.94

6.15

4.88

4.66

4.44

5.04

4.59

4.09

—

1974

0.83

1.15

0.90

1.59

2.08

1.74

1.82

1.87

1.37 —

1965

1.77

1.55

3.87

3.62

3.12

3.56

3.81

4.05

1.50

—

1975

0.77

1.13

1.23

1.42

1.92

1.45

1.64

1.09

1.19 —

1966

1.21

306

3.19

2.93

2.52

261

2.56

2.35

3.03

—

1976

0.34

1.31

1.38

1.69

1.93

1.42

1.54

1.88

2.15 —

1967

1.80

1.14

294

3.38

323

2.67

2.66

2.31

1.80

—

1977

0.28

0.86

1.25

1.50

2.03

1.45

1.42

2.19

2.57 —

1968

2.24

4.15

4.69

4.42

4.02

3.41

3.62

3.63

3.53

—

1978

—

0.84

1.59

1.90

2.11

1.96

1.53

1.92

5.99 —

1969

265

372

5.17

3.81

4 10

3.39

3.43

4.01

2.81

—

1979

0.98

0.91

1.42

1.83

2.10

2.24

1.74

2.23

3.12 —

1970

7.74

3.28

4.76

3.77

3.96

3.08

3.64

2.87

2.70

—

1980

0.26

0.58

1.30

1.73

2.18

2.24

2.07

2.19

3.34 —

1971

11.72

1.60

4.06

3.14

3.32

2.85

2.87

3.16

2.35

—

1981

0.70

0.76

1.17

1.69

2.24

2.40

1.22

1.85

2.99 —

1972

—

1.96

3.19

2.62

3.30

3.01

2.96

3.04

1.96

—

1982

0.54

0.71

1.22

1.79

2.02

2.41

1.57

1.94

2.37 —

1973

—

1.01

1.05

2.26

2.66

2.29

2.22

2.12

2.76

—

1983

0.69

1.51

1.23

1.96

1.91

1.29

1.53

1.46

0.83 —

^Calculated as exp

1 nX In

' landings I
effort I

20nly one trip by tonnage classes 32 and 33 in 1970 on Cape Cod grounds.

a + 2! lP#(/] + e

where U = \n CPUE,

a = intercept estimate,

P,Y = model parameter estimates in loga-
rithmic units for category 7 for ton-
nage class, season, and year,

X,j = dummy variable for tonnage class,
season, and year ( = 1 when category
j occurs; = 0 otherwise), and

e — error term.

All tests for significance of main effects were
based on the above model without interaction.
Separate ANOVAs were also performed to exam-

ine first order interactions. Parameter estimates
obtained from the model without interaction were
retransformed following methods described by
Bradu and Mundlak (1970) to derive unbiased
fishing power, seasonal, and annual coefficients.
Annual coefficients corresponding to the 1965-83
period were multiplied by the reference year
CPUE to compute annual standardized CPUE es-
timates.

RESULTS

Smaller vessels (TC 21-24) generally exhibited
the lowest CPUE indices in all three areas, al-
though TC 21-23 vessels were not represented on
Georges Bank (Table 2). Catch rates of medium
vessels (TC 25; 31-33) were similar to each other,
and were generally greater than those for TC 21-
24 vessels. Mean CPUE indices for the largest
vessels (TC 41 and 42) were more variable, but
generally were equal to or greater than those

FISHERY BULLETIN: VOL. 86, NO. 1

Table 3. — Geometric nriean CPUE^ (landings per day fished, metnc tons) for Georges Bank,
Southern New England, and Cape Cod yellowtail flounder tnps by depth zone, 1964-83.

Georges Bank

herr

New England
pth zone

Cape Cod

Depth zone

Year

Combined 1

Combined

3 Combined

1964

3.92

4.91

2.74

4.80

5.05

4.96

4.71

503

2.13

2.36

—

2.25

1965

3.76

3.48

—

3.55

3.61

3.88

2.28

366

1.61

1.80

2.36

1.71

1966

3.02

2.62

1.29

2.71

2.66

0.71

2.71

1.99

2.13

1.70

2.07

1967

3.19

2.69

2.81

2.79

2.97

2.77

1.14

2.91

1.52

2.16

—

1.80

1968

3.38

3.51

—

3.49

3.70

3.97

3.24

3.76

2.15

2.31

—

2.24

1969

3.36

3.11

—

3.16

3.62

3.93

1.58

3.66

2.16

2.66

—

2.27

1970

2.81

2.99

1.00

2.93

3.44

3.68

1.62

3.49

1.10

2.24

234.18

1.40

1971

2.20

2.30

2.75

2.28

3.26

2.96

—

3.14

1.03

1.94

—

1.46

1972

2.28

2.37

1.34

2.35

3.28

3.25

2.74

3.27

1.56

1.88

—

1.82

1973

2.30

3.06

2.85

2.91

2.41

2.55

4.40

2.47

1.75

1.83

2.27

1.81

1974

2.10

2.32

2.90

2.30

2.31

2.02

—

2.19

1.86

1.78

—

1.82

1975

1.85

1.99

1.29

1.98

1.59

1.63

—

1.60

1.53

1.38

1.32

1.46

1976

1.75

2.06

—

2.03

1.34

1.56

—

1.43

1.54

1.88

2.93

1.70

1977

1.92

2.15

1.67

2.14

1.92

1.97

—

1.94

1,70

1.89

1.27

1.77

1978

1.78

2.09

1.98

2.07

2.45

2.17

—

2.32

1.90

1.71

3.49

1.86

1979

2.47

2.79

2.04

2.74

2.88

2.39

—

2.76

2.03

2.47

2.03

1980

2.58

3.50

1.53

3.37

2.71

3.51

1.13

3.04

2.06

2.31

2.38

2.16

1981

2.26

2.88

1.54

2.76

2.46

2.78

—

2.58

1.74

1.96

5.67

1.79

1982

2.34

2.66

2.53

2.62

3.20

3.54

4.53

3.31

1.81

1.44

0.91

1.67

1983

1.98

2.28

3.04

2.26

3.08

3.13

5.91

3.10

1.70

1.41

—

1.62

1 Calculated as exp

n-:i In (

andings
effort

)]

20nly one tnp in depth zone 3 in 1970 on Cape Cod grounds.

corresponding to medium and small vessels, par-
ticularly in the later years (Table 2). The initial
ANOVAs performed over all statistical areas re-
vealed highly significant iP < 0.01) differences in
CPUE for tonnage class and area main effects in
each of the 20 years (Table 4). The interaction of
tonnage class and area was also highly signifi-
cant in all years, suggesting that relative fishing
power of the individual vessel classes varies ac-
cording to area. ANOVA results for the compari-
son of CPUE among stocks were highly signifi-
cant for area main effects in 19 out of 20 years,
and the tonnage class-stock area interaction term
was highly significant in all years (Table 4).
Grouping the data according to stock tended to
reduce the amount of significant tonnage class-
area interaction within each stock, although dif-
ferences among tonnage classes remained highly
significant.

On Georges Bank the differences in CPUE were
highly significant for statistical area and tonnage
class main effects in 80 and lOO'/r of the years,
respectively, while the tonnage class-area inter-
action was highly significant in only 40% of the
years. Differences due to area on Southern New
England grounds were highly significant in all
years except 1978, and differences due to tonnage
class were highly significant in all years. The in-
teraction term was highly significant in 70% of

the years. Differences due to area on the Cape Cod
grounds were highly significant in all years ex-
cept 1975, and differences due to tonnage class
were highly significant for all years. The interac-
tion was highly significant in only 35% of the
years (Table 4).

Differences in CPUE by depth zone were gener-
ally not significant. Depth main effects yielded

Table 4. — Frequency with which highly significant (P < 0.01) re-
sults were obtained from analysis of vanance (ANOVA) tests of
yellowtail flounder annual CPUE data. (Total number of years
tested = 20.) N/A = Not applicable (tests not performed).

Main effects

Area

Tonnage class

Depth

All areas

2020

20/20

N/A

Among stocks

19 20

N/A

N'A

Within stocks

Georges Bank

16 20

20/20

10 20

So. New England

1 9, 20

20/20

7/20

Cape Cod

19/20

20/20

3/20

Interactions

Tonnage class Tonnage class

■ area

■ depth

All areas

20/20

N/A

Among stocks

20/20

N/A

Within stocks

Georges Bank

8/20

4/20

So. New England

14/20

2/20

Cape Cod

7/20

1/20

100

O'BRIEN AND MAYO: CPUE OF YELLOWTAIL FLOUNDER

highly significant differences in only 50, 35, and
IdVc of the years for Georges Bank, Southern New
England, and Cape Cod grounds, respectively
(Table 4), while the tonnage class-depth interac-
tion was highly significant in no more than 20%
of the years on each of the three grounds.

Interaction between tonnage class and statis-
tical area throughout the region was highly sig-
nificant in all cases. Further analyses yielded
highly significant differences in CPUE among the
three stocks and highly significant tonnage class-
stock interactions which suggests that relative
fishing power among vessel classes is not consis-
tent from stock to stock, implying a need for com-
puting a separate set of fishing power coefficients
for each stock. Within each stock, differences in
CPUE among statistical areas and tonnage
classes were also highly significant in most cases,
although the tonnage class-area interaction was
not.

Standardized CPUE

Annual fishing power coefficients, obtained by
retransforming linear model parameter esti-
mates for each tonnage class, are presented in
Table 5 by stock. Cursory examination of the coef-
ficients reveals distinct trends throughout the 20-
yr series. On Georges Bank and Southern New
England grounds, fishing power coefficients for
the smaller vessels (TC 23-25) relative to the
standard declined over time, whereas coefficients
for the larger vessels increased over time. On
Cape Cod grounds the coefficients increased for
the smaller vessels (TC 23 and 24) although
trends were less pronounced. These trends are
illustrated graphically by plotting annual devia-
tions from the 20-yr average in Figures 7-9. A
Durbin-Watson test for first order autocorrela-
tion of the annual deviations (Neter and Wasser-
man 1974) was significant for most tonnage
classes within each of the stocks, suggesting the
presence of a substantial tonnage class-time in-
teraction.

The three-way linear model, modified to in-
clude interaction terms, also revealed highly sig-
nificant tonnage class-year as well as tonnage
class-season and year-season interactions within
each of the three stocks (Table 6). When interac-
tions are significant, they can be examined in
detail or absorbed in the error term when testing
for main effects. Since tonnage class effects have
already been examined on an annual basis, the
interaction terms were excluded from the three-

way model used to obtain parameter estimates for
tonnage class, season, and year. The model is
specified as follows:

f/ = a + X [Pi, Xy + ^2j X2j + Paj Xsj] +

where ^y, P2/, P3; = model parameter estimates

in logarithmic units for
category j for tonnage
class, season, and year,

Xij , X2J, X2,j

dummy variables for ton-
nage class, season, and
year ( = 1 when category 7
occurs; = 0 otherwise).

ANOVA results obtained without interaction
are presented in Table 6 for each of the three
stocks. For Georges Bank and Southern New
England stocks, year accounts for the greatest
reduction in error sums of squares; on Cape Cod
grounds tonnage class accounts for the greatest
overall reduction.

Coefficients for tonnage class, year, and season,
derived from model parameter estimates for the
combined 1964-83 period are presented in
Table 7. Tonnage class coefficients for Georges
Bank and Southern New England are relatively
homogeneous, as compared with those obtained
for Cape Cod grounds, owing to the narrower
range of vessel tonnage classes which have con-
sistently exploited these fisheries. Seasonal coef-
ficients exhibit the same pattern on Georges
Bank and Southern New England with the
highest catch rates occurring during the third
quarter; on Cape Cod grounds the highest catch
rates occur during the second quarter. Trends in
annual coefficients are similar on all three
grounds. Standardized CPUE indices based on
the annual coefficients are illustrated in Figure
10, and traditional indices based on the methods
of Lux (1964), as given by Clark et al. (1984), are
also presented for comparative purposes.

Although each series indicates similar trends,
CPUE indices obtained from the linear model for
Georges Bank and Southern New England have
remained slightly higher than the traditional in-
dices since the early 1970's. Prior to this, the tra-
ditional CPUE indices were greater than the re-
vised indices. On Cape Cod grounds, differences
between the two series are considerably greater,
particularly in the early years.

101

Georges Bank

1-l

TC24

0.5-

•

■

-0.5-

•

c ■^^
o

65 67 89 71 73 75

77 79 81 83

4-»

- ^1

TC25

> 0.5-

•

£J 0-

, '

— -0.5-

Zj ■' 1 I i I I I 1
J- 65 87 89 71 73 75

1 1 I 1

77 79 81 83

TC31

0.5-

-0.5-

* '

1 1 1 1 1 I
65 67 89 71 73 75

1 1 1 1
77 79 81 83

Table 5.

KIS51EKY BULLETIN: VOL 86, NO 1

-Annual fishing power coeflicients calculated by vessel tonnage

Georges Bank, Southern New England,

Vessel tonnage class^

Year 21 22

23 24

31 32

Georges Bank

1964

073

083

0.97

0.91

0,90

043

1965

068

0.86

1.02

095

1,40

—

1966

0.45

0.81

0.99

099

097

—

1967

—

0.74

0.93

094

0,89

—

1968

1.09

084

098

1 08

088

0,75

1969

0.99

0.85

0.93

0.92

0,95

1.17

1970

1.21

0.78

0.94

0.91

094

0.92

1971

0.66

0.82

1.07

089

1,12

—

1972

—

082

0.94

096

0,93

—

1973

not

0.49

0.94

0.93 1.00

1 09

1,09

1.14

1974

calculated

—

1 25

096

098

097

1.53

1975

1.00

0.99

0.85

1.06

1 04

1 66

1976

0.73

0.92

0.96

1.17

1,01

—

1977

—

0.66

0.94

1.28

1,67

—

1978

—

0.97

0.87

1,18

1,54

—

1979

0.55

0.76

1.02

1.19

1.35

—

1980

0.42

0.35

1.10

1.25

1.43

—

1981

0.45

0.50

0.80

1.05

1.16

—

1982

0.23

0.96

0.82

1.21

1.41

099

1983

—

0.40

1.25

1 12

1.26

—

Southern New England

1964

- 20.95

— 0.97

0.92

0.88

0.91

0.81

1965

— 0.86

— 0.95

0.82

0.93

1 06

0.39

1966

— 1.12

— 1.14

0.98

1.02

092

1.18

1967

— 0.94

— 1.27

1.22

1.00

0.87

068

1968

— 1.21

— 1.23

1.11

0.94

1,01

098

1969

— 1.32

— 1.11

1.19

0.99

1.17

0.82

1970

— 1.23

— 1.04

1.09

085

0.79

0.74

1971

— 1.06

— 1.10

1 16

0.99

1.10

082

1972

— 0.99

— 088

1.12

1.01

1,03

0.66

1973

— 0.46

— 1.01

1.20

1 .03 1 .00

0,95

1.24

'Standard vessel class on Georges Bank and Southern New England
rounds = 32. Standard vessel class on Cape Cod grounds = 25.

TC33

0.5-

0.5-
-1-

[ I 1 1 1 1 1 1 I 1

0.5
0

66 67 69 71 73 75 77 79 81 83

TO 41

~~i I I I I I — I — I — I — I

65 67 69 71 73 75 77 79 81 83

1
0.5

TC42

~i I I I — I — I — I — I — I — I

66 67 69 71 73 75 77 79 81 83

Figure 7. — Deviations in annual fishing power from the
1964-83 20-yr mean for major vessel tonnage classes fish-
ing on Georges Bank.

Year

102

O'BRIEN AND MAYO: CPUE OF YELLOWTAIL FLOUNDER

class relative to a standard class vessel fishing for yellowtail flounder on
and Cape Cod grounds, 1964-83.

Vessel

tonnage class^

Year

41 42

Southern New England

1974

—

0.67

—

1.11

1 03

1 00

1.27

0.62

1975

—

072

—

0.73

0.83

085

1.00

1.11

1976

—

0.66

—

0.40

0.66

086

1.28

1.56

1977

—

087

—

0.79

0.97

088

1.09

0.99

1978

—

1 33

—

0.56

0.78

0.75

1.11

1.38

1979

—

069

—

0.54

0.74

079

098

1.24

1980

—

082

—

0.68

0.90

0.95

1.09

1.07

1981

—

0.92

—

0.55

099

1.06

1.26

0.86

1982

—

0.64

—

0.62

0.84

0.86

1.07

1983

—

0.68

—

092

0.94

0.89

0.98

1.04

Cape Cod

1964

—

062

0.52

0.73

1.08

1.04

—

1965

—

067

0.55

0.73

1.65

0.98

0.96

—

1966

044

042

048

0.58

1.42

0.69

—

1967

—

032

0.36

0.64

1.30

0.79

1.07

—

1968

—

044

0.37

0.71

1.55

0.91

0.99

0.43

1969

—

060

048

0.73

1.31

0.89

—

1970

—

0.44

0.45

1.01

1.56

(3)

1.25

1971

—

0.57

0.93

1.19

1.10

0.93

1.20

1.65

1972

0.93

088

063

1 09

1.09

0.79

0.97

1.05

1973

0.51

066

053

0.94

1.00

0.95

0.96

0.82

1.43

1974

0.40

0.55

0.43

0.76

083

0.87

0.90

0.66

1975

0.40

059

0.64

0.74

0.75

085

0.57

0.62

1976

0 18

068

072

088

0.73

0.80

0.97

1.12

1977

0.14

042

061

0.74

0.71

0.70

1.08

1.26

1978

—

040

075

0.90

0.93

0.73

0.91

2.84

1979

0.47

043

0.68

0.87

1.06

0.83

1.06

1.48

1980

0 12

027

0.60

0.80

1.03

095

1 01

1.53

1981

0.31

0.34

052

0.75

1.07

0.54

0.83

1.34

1982

0.27

035

060

088

1.19

0.77

0.96

1.17

1983

0.36

0.79

0.64

1.02

0.68

0.80

0.77

0.43

2Vessel classes 21, 22, and 23 combined on Southern New England
grounds
3|nsuf1icient data for these categories.

Southern New England

c
o

05-

TC 21-22-23

0
0.5-

-1-

1 1 ; 1 \ 1 1 1 ^ 1

0.5-

0
-0.5

65 67 69 71 73 75 77 79 81 83

TC31

— I 1 1 1 1 1 \ 1 1 I

65 67 69 71 73 75 77 79 81 83

TC33

<0
D
C
C
<

Figure 8. — Deviations in annual fishing power from the
1964-83 20-yr mean for major vessel tonnage classes fish-
ing on Southern New England grounds.

0.5 H

-0.5-

-I — I — I — I — I — I — I — I r
65 67 69 71 73 75 77 79 81 83

TC25

— I 1 1 1 1 1 1 I I I

65 67 69 71 73 75 77 79 81 83

—I — 1 1 1 1 1 1 1 1 I

66 07 69 71 73 76 77 79 81 83

0.5-

TC41

0.5-

-1-

— 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1

65 67 69 71 73 75 77 79 81 83

Year

103

FISHERY BULLETIN VOL 86. NO 1

Cape Cod

0 5
0

-0 5-

TC23

— I — I — I — I — 1 — I — I — I — I — I

65 67 69 71 73 75 77 79 81 83

0.5

-0.5

-1

TC32

—1 — I — I — I — I — I — I — I — I — I

65 67 69 71 73 75 77 79 81 83

Figure 9.— Deviations in annual fishing power from the
1964-83 SO-yr mean for major vessel tonnage classes fish-
ing on Cape Cod grounds.

Q
<o

C
C

1
0.5-

0
-0 5

1-1
0.5
0

-0.5-
-1

TO 24 '

0.5-

—\ 1 1 1 1 1 1 1 1 1

65 67 69 71 73 75 77 79 81 83

0
-0.5

-1

TO 31

—\ — I — I — I — I — I — I — I — I — I

65 67 69 71 73 75 77 79 81 83

0.5

-0.5

-1

Year

TC33

— I — I — I — I — I — I — I — I — I — I

65 67 69 71 73 75 77 79 81 83

•TC41

— I \ 1 1 1 1 1 1 1 I

65 67 69 71 73 75 77 79 81 83

Georges Bank

I ' I ' I ' I ...

64 66 68 70 72 74 76 78 80 82 84

Year

>•

c
,o

Southern New England

n ' I •—] I |— ' I I I — ■— I — I — I 1 I I — I
64 66 68 70 72 74 76 78 80 82 84

Year

4-1

■D
0)

JO.

(f)

a3
Q.

c
,o

Cape Cod

I ' I ' I ' I ' I ' I
64 66 68 70 72 74 76 78 80 82 84

Year

Traditional CPUE index

Linear Model CPUE Index

Figure 10. — Trends in annual yellowtail flounder CPUE (metric tons per day fished) calculated with
traditional CPUE based on Lux (1964) and annual coefficients (linear model CPUE).

104

O'BRIEN AND MAYO; CPUE OF YELLOWTAIL FLOUNDER

Table 6. — ANOVA results obtained from a three-way linear model
incorporating year, quarter, and tonnage class for Georges Bank,
Southern New England, and Cape Cod stocks of yellowtail floun-
der.

Table 7. — Tonnage class, quarter, and year coefficients derived
from a three-way linear model for the 1964-83 period for yellowtail
flounder stocks on Georges Bank, Southern New England, and
Cape Cod grounds.

Sum of

Mean

Georges

Southern

Source

squares
Georges

square
Bank

value

Tonnage

Bank

New England

Cape Cod

With interaction

class

Model

211

1.432 92

6.79

29.53

<0.01

0.92

0.32

Year

908.42

47.81

207.88

<0.01

0.92

0.50

Qtr

107.36

35.79

155.59

<0.01

0.92

0.57

84.24

9.36

40.70

<0.01

057

0.93

0.80

Yr'Tc

103

116.81

1.13

4.93

<0.01

0.80

0.96

1.00

Yr*Qtr

202.29

3.55

15.43

<0.01

0 96

0.92

1.03

Qtr'Tc

13.80

0.69

3.00

<0.01

1.00

0.84

Error

18,437

4.171.68

023

1.08

1.09

1.00

Total

18,648

5.604.60

030

1.13

1.05

1.24

Without inte

iraction
31

1.100.02

35.48

147.83

<0.01

1.19

—

Model

Quarter

Year

908.42

47.81

199.21

<'0.01

Qtr

107.36

35.79

149.11

<0.01

0.96

1.00

1.07

84.24

936

39.00

<0.01

0.89

1.06

1.15

Error

18,617

4.504.58

0.24

1.09
1.00

1.18

1.00

0.90
1.00

Total

18,648

5 604 60

n .^n

Southern New England

Year

With interaction

1964

1.00

Model

217

2,85995

13.18

33.79

<0.01

1965

0.73

0.72

0.84

Year

2,085.95

109.79

281.50

<0.01

1966

0.55

0.54

0.84

Qtr

111.50

37.17

95.30

<0.01

1967

0.54

0.60

0.81

59.80

9.97

25.56

<0.01

1968

0.68

0.79

0.93

Yr*Tc

114

257.12

2.26

5.78

<0.01

1969

0.61

0.77

1.06

Yr-Qtr

304.89

535

13.72

<0.01

1970

0.58

0.75

1.00

Qtr-Tc

40.69

2.26

5.80

<0.01

1971

0.49

0.63

0.95

Error

26,879

10.612.21

0.39

1972

0.45

0.65

0.88

Total

27,096

13,472.16

0.50

1973
1974

0.55
0.42

0.49
0.45

0.84
0.77

Without interaction

1975

0.36

0.30

0.72

Model

2.257.25

80.62

196.62

<0.01

1976

0.37

0.25

0.76

Year

2,085.95

109.79

267.77

<0.01

1977

0.37

0.36

0.74

Qtr

111.50

37.17

90.66

<0.01

1978

0.34

0.39

0.83

59.80

9.97

24.31

<0.01

1979

0.48

0.52

0.85

Error

27,068

11.214.91

0.41

1980

0.55

0.56

0.78

Total

27,096

13.472,16

0.50

1981

0.46

0.54

0.80

Cape Cod

1982
1983

0.44
0.40

0.71

0.70

0.77
0.78

With interaction

Model

247

2.438.73

987

2468

<:0.01

Year

166.49

8.76

21.91

<0.01

Qtr
TC

174.44
1,331.92

58.15
147.99

145.37
369.98

<0.01

DISCUSSION

Yr'Tc

135

526.18

3.90

9.74

<0.01

Yr'Qtr

167.62

2.94

7.35

<0.01

The analytical approach adopted

in this paper

Qtr'Tc

72.09

3.00

7.51

<0.01

is based on _ _

the hypothesis that CPUE of yellow-

Error

19,097

7,731.24

0.40

tail flounder differs among the various tonnage

Total

19.344

10.169.97

0.53

classes of vessels and ge _ _

sographic regions associ-

Without interaction
Model 31

1,672.84

53.96

122 64

<0.01

ated with the fishery. In all of the analyses, the

Year

166.49

876

19.92

<0.01

null hypothesis (i.e., no

1 significant differences)

Qtr
TC

3
9

174.44
1,331.92

58.15
147.99

132.15
336.34

<0.01
<0.01

was rejected only when the probabil

iity of obtain-

Error

19.313

8.497.13

0.44

ing a greater F statistic

was <0.01.

Even at this

Total

19.344

10.169.97

0.53

probability

level, statistically signi

ficant results

were often obtained when difi'erences among vari-
able levels appeared to be minimal due primarily
to the large number of observations included in
most analyses.

The initial series of ANOVAs, based on pooled
CPUE data from all statistical areas encompass-

105

FISHEKY BULLETIN; VOL. 86, NO. 1

ing the three stocks, provided sufficient evidence
to reject the null hypothesis. In each of the 20
years analyzed, the main effects of tonnage class
and statistical area represented highly signifi-
cant (P < 0.01) sources of variation. The tonnage
class-area interaction term was also highly sig-
nificant in all cases, implying that vessels of var-
ious tonnage classes exhibit different CPUE
trends relative to each other in different areas.

The initial results established the basis for fur-
ther investigations. In subsequent analyses, the
data were grouped to test the null hypothesis that
no significant differences in CPUE existed among
the three traditionally accepted stock definitions
(Georges Bank, Southern New England, and
Cape Cod). The highly significant differences ob-
tained in 19 out of 20 years indicate that catch
rates differ among the three stocks. The resulting
highly significant tonnage class-stock area inter-
action term obtained from the ANOVAs in all
years suggests that standardization of CPUE
among tonnage classes should be performed sepa-
rately for each stock.

Analysis of variance within each stock provided
the final basis for performing the standardized
CPUE calculations. In these tests, the rejection of
the null hypothesis for tonnage class main effects
in all years for each stock suggests that separate
fishing power coefficients must be calculated for
each tonnage class even though the coefficients
are similar in many cases. The ANOVA results
also indicated that differences in CPUE among
statistical areas within each stock were highly
significant in SO^f or more of the years implying
that, within each stock region, yellowtail abun-
dance is not homogeneous. This is not surprising
since yellowtail flounder are prevalent only on
certain grounds within each geographic region.
Further analyses of the data by depth indicated
no overall significant differences in CPUE be-
tween the two primary depth zones (1-55 m and
56-110 m) where yellowtail flounder are consis-
tently caught.

The infrequent number of significant interac-
tions on Georges Bank and Cape Cod grounds
relative to Southern New England (Table 4) sug-
gests a greater independence of the tonnage class
and area main effects with respect to CPUE, i.e.
both large and small vessels exhibited relatively
similar changes in mean CPUE among statistical
areas within each stock. In choosing data sets for
computing fishing power coefficients we sought to
minimize the amount of interaction among the
vessel tonnage classes and geographic areas in-

volved. This criterion was met to a greater extent
for the Cape Cod and Georges Bank stocks than
for the Southern New England stocks. It appears
that yellowtail flounder inhabiting this region
are subject to a more complex set of interactions
perhaps due to temperature and bottom type. We
decided, however, to accept the results for each of
the three stocks and proceed with the calculations
of fishing power coefficients.

Annual fishing power coefficients computed for
each vessel tonnage class fishing on Georges
Bank, Southern New England, and Cape Cod
grounds provided a basis for examining the con-
sistency in relative fishing power of individual
tonnage classes over time. Annual deviations for
Georges Bank and Southern New England
grounds indicated a gradual change in relative
fishing power of most tonnage classes between
1964 and 1983, and tests for autocorrelation of
residuals indicated significant time effects. On
these grounds, larger vessels exhibited higher
catch rates relative to the standard in the later
years as compared with the earlier years. Since
many of the larger vessels have been replaced in
recent years by newer vessels which are, pre-
sumably, equipped with more sophisticated elec-
tronics, any attempt to relate CPUE to stock
abundance must account for such technological
advances.

Similarly, changes in seasonal availability are
often great enough to mask interannual variation
in stock abundance. Thus, the presence of signifi-
cant tonnage class-season interactions may be ex-
plained by the ability of certain vessel classes to
effectively target seasonal concentrations. Since
peak spawning of yellowtail flounder occurs dur-
ing late spring (Lux 1964), the presence of high
seasonal coefficients during the second and third
quarters is not surprising.

By specifying the model to include tonnage
class, annual, and seasonal components, we have
attempted to account for technological and sea-
sonal availability factors which interact with
temporal changes in abundance. Although other
factors could be incorporated in the model to ac-
count for a larger portion of the variation in
CPUE, analyses of historical commercial fishing
operations of this type are often limited to those
attributes which can be directly linked to land-
ings records (Kimura 1981; Westrheim and
Foucher 1985). An alternate approach adopted by
Stern and Hennemuth (1975) involved the use of
a study fleet of selected vessels whose characteris-
tics and fishing practices were closely monitored.

106

O'BRIEN AND MAYO CPUE OF YELLOWTAIL FLOUNDER

In our study, factors were selected for inclusion
in the model based on prior knowledge of fleet
characteristics and seasonal and spatial distribu-
tion patterns of the species. Despite this, the
three attributes incorporated in the final model
accounted for 15-25% of the total variation in
CPUE, depending on the stock. Undoubtedly,
other factors such as experience of the captain,
net design and rigging, and variation in local fish
abundance contribute substantially to overall
variation in catch rates.

Differences between the annual CPUE esti-
mates based on Lux's original fishing power coef-
ficients and the recalculated indices occur in
many cases because of shifts in the vessel compo-
sition of the fleet over the past 20 years. The in-
clusion of larger vessels in the more recent years,
particularly on Georges Bank and Southern New
England grounds, may account for the consis-
tently higher CPUE estimates obtained for these
areas since the mid-1970's. On Cape Cod grounds,
CPUE estimates differ substantially prior to this
time. Lux (1964) has stated that a relatively low
proportion of the landings from this area were
used in his CPUE computations and, conse-
quently, the indices were not considered to be as
valid a measure of relative abundance as those
obtained for Georges Bank and Southern New
England. Our analyses for Cape Cod grounds,
based on data for the period since 1964, are sub-
ject to the same concerns since a large proportion
of the yellowtail flounder landings continues to be
taken incidentally.

Although the revised standardized CPUE esti-
mates presented in this paper are based on a dif-
ferent standardization technique, trends are gen-
erally similar to those obtained previously. The
revised procedure, however, accounts for seasonal
and technological influences and insures com-
plete representation of all vessel classes engaged
in the yellowtail fishery.

ACKNOWLEDGMENTS

We wish to sincerely thank Stephen H. Clark
for his advice throughout this study, and for crit-
ically reviewing the manuscript. Michael J. Fo-
garty reviewed the final draft and advised on
statistical procedures. We are also grateful for the
suggestions provided by two anonymous referees.

LITERATURE CITED

Beverton, R J H . AND S J Holt

1957. On the dynamics of exploited fish popula-
tions. Fish. Invest., Lond., Ser. 2, 19:1-533.

Bradu, D. andY Mundlak

1970. Estimation in lognormal linear models. J. Am.
Stat. Assoc. 65(329):198-211.
Brown. B E . M P Slssenwine. and M M McBride.

1980. Implication of yellowtail flounder stock assessment
information for management strategies. U.S. Dep.
Commer., NMFS, NEFC, Woods Hole Lab. Ref Doc. No.
80-21, 12 p.

Clark, S. H., M. M. McBride, and B Wells.

1984. Yellowtail flounder assessment update - 1984.

U.S. Dep. Commer., NMFS, NEFC, Woods Hole Lab. Ref.

Doc. No. 84-39, 30 p,
Dixon, W J (editor),

1981. BMDP statistical software, Univ. Calif. Press,
Berkeley, p. 388-412.

Garrod. D J

1964. Effective fishing effort and the catchability coeffi-
cient q. Rapp. Cons. int. Explor. Mer 155, No. 14, p. 66-
70.

Gavaris. S

1980. Use of a multiplicative model to estimate catch rate
and effort from commercial data. Can. J. Fish. Aquat.
Sci. 37:2272-2275.

Gulland. J A

1956. On the fishing effort in English demersal fish-
eries. Fish. Invest., Lond., Ser. 2, 20(51:1-41.

1964. Catch per unit effort as a measure of abundance.
Rapp. Cons. int. Explor. Mer 155, No. 1, p. 8-14.
KiMURA. D K

1981. Standardized measures of relative abundance based
on modelling log (c. p. u.e.), and their application to Pacific
ocean perch (Sebastes alutus). J. Cons. int. Explor. Mer
39:211-218.

Lux. F E

1963. Identification of New England yellowtail flounder
groups. Fish. Bull, U.S. 63:1-10.

1964. Landings, fishing effort, and apparent abundance
in the yellowtail flounder fishery. Int. Comm. North-
west Atl. Fish. Res. Bull. No. 1, p. 5-21.

NETER. J . AND W WaSSERMAN

1974. Applied Linear Statistical Models. Richard W.
Irwin, Inc., Homewood, IL, 843 p.

Pope. J A . and B B Parrish

1964. The importance of fishing power studies in abun-
dance estimation. Rapp. Cons. int. Explor. Mer 155, No.
17, p. 81-89.

Pope, J G . and D J Garrod

1975. Sources of error in catch and effort regulations with
particular reference to variations in the catchability coef-
ficient. Int. Comm. Northwest Atl. Fish. Res. Bull. 11,
p. 17-30.

Robson, D S.

1966. Estimation of the relative fishing power of individ-
ual ships. Int. Comm. Northwest Atl. Fish. Res. Bull. 3,
p. 5-14.

Rounsefell. G a

1957. A method of estimating abundance of groundfish on
Georges Bank. Fish. Bull., U.S. 113:264-278.

RoYCE, W. F., R J Buller, E. D. Premetz

1959. Decline of the yellowtail flounder (Limanda fer-

ruginea) off New England. Fish. Bull., U.S. 59:169-

267.
SAS Institute

1982. SAS User's Guide: Statistics. 1982 ed. SAS Insti-
tute Inc., Gary, NC.

107

FISHERY BULLETIN: VOL. 86, NO. 1

SissENWiNE. M P Stern. H , ,Jk , and R C Hknnemuth

1974. Variability in recruitment and equilibrium catch of 1975. A two-way model for estimating standardized fish-

the Southern New England yellowtail flounder fish- ing crfort applied to the U.S. haddock fleet. Rapp. Cons.

ery. J. Cons. int. Explor. Mer 36:15-26. int. Explor. Mer 168:44-49.

1978. Is MSY an adequate foundation for optimum Westrheim, S J , AND R. P. FOUCHER.

yield? Fisheries 3(6):22-42. 1985. Relative fishing power for Canadian trawlers land-

Steel. R G D, AND J H TORRIE. ing Pacific cod iCiadus macrocephalus) and important

1980. Principles and procedures of statistics. McGraw- shelf cohabitants from major offshore areas of western

Hill, N.Y., 633 p. Canada, 1960-1981. Can. J. Fish. Aquat. Sci. 42:1614-

1626.

108

REDUCING THE BYCATCH IN A COMMERCIAL TROTLINE FISHERY

Lawrence W McEachron,' Jeff F Doerzbacher,- Gary C Matlock,^
Albert W. Green,^ and Gary E, Saul-

ABSTRACT

Reducing the bycatch of red drum, Sciaenops ocellatus, and spotted seatrout, Cynoscion nebulosus, in
the Texas commercial trotline fishery is desirable. Hook placement within the water column was
examined as a means of accomplishing this objective. The commercial trotline fishery was simulated
in the Laguna Madre during February 1985 through January 1986. Requiring placement of trotline
hooks on bottom will reduce bycatch of red drum, spotted seatrout, and other nonmarketable fishes
and improve operational efficiency of commercial fishermen without significantly reducing catch of
black drum, Pogonias cromis, a target commercial species. Other than crab and shrimp being more
effective baits than oleander leaves, no other generalization could be made concerning baits and
seasons.

Longlines catch species unwanted or legally non-
retainable by fishermen and have been regulated
to reduce the bycatch of nontargeted species
(South Atlantic Fishery Management Council
1985). Trotlines (Figs. 1, 2) are a specialized long-
line used in shallow (<4 m) Texas estuaries to
catch fish (Simmons and Breuer 1962; Breuer
1973, 1974, 1975; Matlock 1980). Red drum,
Sciaenops ocellatus, and spotted seatrout,
Cynoscion nebulosus, were the primary targets
until 1981 when their sale was prohibited be-
cause of overfishing (Matlock et al. 1979; Anony-
mous 1979, 1981, 1983). The effort has since been
redirected toward black drum, Pogonias cromis.
Regulations requiring the use of circle hooks and
placement of the mainline under water were en-
acted to reduce the bycatch of red drum and spot-
ted seatrout. However, a bycatch still occurs. This
study was conducted to determine if the bycatch
could be further reduced by additional regulation
of where in the water column hooks are fished and
bait types.

MATERIALS AND METHODS

The catch on trotlines with hooks placed on the
bottom or in the top of the water column was
compared by simulating commercial fishing tech-
niques in the Laguna Madre, TX (Fig. 3). Bottom
trotlines were set with the mainline on the bot-
tom. Top trotlines had the mainline floated with

iTexas Parks and Wildlife Department, P.O. Box 1717, Rock-
port, TX 78382.

2Texas Parks and Wildlife Department, 4200 Smith School
Road, Austin, TX 78744.

the hooks suspended in water >0.6 m deep to
insure hooks fished in the water column. Texas
Parks and Wildlife Department (TPWD) trotlines
were set in the same area as commercial trotlines.
Commercial fishermen were contacted by tele-
phone within 24 hours prior to TPWD sets to de-
termine areas of commercial activity. All TPWD
trotlines were at least 15 m apart.

Trotlines with 100 hooks each were built ac-
cording to commercial fishermen specifications
(McEachron et al. 1985). The mainline (182.9 m
long) consisted of #36 nylon twine, knotted twice
every 1.8 m for swivel (1/0 black brass barrel)
placement (Figs. 1, 2). Hooks (#8 Mustad
39960ST) were attached by a 610-686 mm long
staging (56.7 kg test monofiliament) to the swivel
at 1.8 m intervals. Stakes (51 cm x 76 mm) and/or
anchors were placed on each end to stretch the
mainline. Floats (3.8 L) were attached to the
mainline every 15 hooks for navigation identifi-
cation.

Eighteen trotlines were set overnight each
month in both the upper and lower Laguna Madre
during 1 February 1985 through 31 January
1986. Six (3 top; 3 bottom) were set during each of
two monthly sampling periods (first and last 15
days of the month). Another six sets were made in
either the first or last half of each month; the
period was randomly selected each month. Each
trotline was baited completely with one of three
bait types — cut portions of blue crab, Callinectes
sapidus; dead shrimp, Penaeus sp.; or oleander,
Nerium sp. leaves — so that all bait types were
used on both top and bottom trotlines during
every period. These baits represented the most

Manuscript accepted October 1987.
FISHERY BULLETIN; VOL. 86. NO 1, 1987.

109

<5-*

5.1 cm X 7.6 cm
WOODEN STAKES

Figure 1.— Top trotline.

1 cm X 7.6 cm
WOODEN STAKES

Figure 2. — Bottom trotline.

commonly used baits by commercial trotline fish-
ermen (McEachron et al. 1980, 1986).

Fishes caught were identified (Hoese and
Moore 1977; Robins et al. 1980), counted (Table
1), and total length (TL) was measured to the
nearest 1 mm. Data were pooled into fall (Septem-
ber-November), winter (December-February),
spring (March-May), and summer (June-Au-
gust) to examine seasonal variation.

A catch rate (No. /line • h) for black drum; red
drum; spotted seatrout; hardhead catfish Arius
felis ; and total fishes was computed for each trot-
line set by dividing the number caught by the
number of hours fished. Catch rates were trans-
formed to log (catch rate + 1) and analyzed using
a four-factor fixed-effects model analysis of vari-
ance (AOV). The four factors were 1) hook place-

ment, at two levels — top and bottom; 2) bait, at
three levels — crab, shimp, and leaves; 3) bay, at
two levels — upper Laguna Madre and lower La-
guna Madre; 4) season, at four levels — fall, win-
ter, spring, and summer.

Diff'erences in main effect means were evalu-
ated with Ducan's multiple range test. However,
when significant first-order interactions were
found, comparisons were made within levels of
the interacting factors using the mean square
error (MSE) from the AOV.

Total lengths of each species were analyzed in
a nested AOV to investigate differences among
the four factors. However, because fish were not
caught in all factor level combinations, factors
and/or factor levels for each species were elimi-
nated from analyses. Spotted seatrout lengths

110

McEACHRON ET. AL: BYCATCH IN TROTLINE FISHERY

Figure 3.— Texas coast.

Table 1 —Number of fishes caught on trotlines in the upper and
lower Laguna Madre during February 1985-January 1986.

Upper

Lower

Laguna

Species

Madre

Total

Arlus felis

977

1,652

2,629

Sciaenops ocellatus

352

658

1,010

Pogonias cromis

265

332

Cynoscion nebulosus

103

132

Micropogonias undulatus

Opsanus beta

Archosargus probatocephalus

Dasyatis americana

Dasyatis sabina

Elops saurus

Orthopristis chrysoptera

Bagre marinus

Lagodon rhomboides

Paralichthys lethostigma

Rhinoptera bonasus

Chilomycterus schoepfi

Ophichthus gomesi

Cynoscion arenanus

Negapnon brevirostris

Trachinotus carolinus

All species

1.512

2,812

4,324

were pooled for both bay systems because an in-
sufficient number of spotted seatrout were caught
for individual bay analyses. Factor levels elimi-
nated from length analyses were leaves and win-
ter from hardhead catfish, leaves and crab from
spotted seatrout, and fall, spring, and summer
from black drum. Spring and winter red drum
lengths were pooled. Lower Laguna Madre data
only were used for red drum, black drum, and
hardhead catfish length analyses. Each measured
fish length was an observational unit of a trotline
set. Sets were a random factor nested within fixed
main effect combinations. The nested set effect
mean square was used for testing other effects
when the set effect was significant. However, the
AOV yields approximate F values because un-
equal numbers of fish were caught among sets.

SAS procedures (SAS Institute, Inc. 1980, 1982)
were used for all analyses. The significance level
for each AOV test was a = 0.01 because the AOV
used to examine catch rates of each species had 15
potential F tests. This alpha value assured that

111

KKSHEKY BULLETIN: VOL. 86, NO. 1

the family level of significance would not exceed
0.15. All other tests were made with a signifi-
cance level of u = 0.05. Mean catch rates and con-
fidence intervals computed from transformed
data were back-transformed for tabular and
graphic presentation (Elliott 1979).

RESULTS

Fishing trotlines on the bottom reduces bycatch
without affecting catches of black drum, the
target species. A significant difference could not
be detected between black drum catch rates on
top and bottom trotlines regardless of bait, sea-
son, or bay (Tables 2, 3). Catch rates for hardhead
catfish, red drum, spotted seatrout and total
fishes were significantly lower on bottom trot-
lines than on top trotlines (Tables 2, 3). Differ-
ences in catch rates between top and bottom trot-
lines for hardhead catfish, red drum, and total
fishes did not vary significantly among seasons
and baits but did vary between bays based on the
first-order interactions (Tables 4, 5). A significant
second-order interaction of position x bay x bait
for spotted seatrout revealed that differences be-
tween top and bottom trotlines were affected by
both bait and bay but not by season (Fig. 4).

No significant difference was found in red drum
catch rates among baits nor in spotted seatrout
catch rates among seasons (Tables 2, 3). All other
main effects were significant for catch rates of all
species and total fishes. Of the possible first-order
interactions involving bait, season, and bay, only
season x bait for hardhead catfish, red drum, and
total fishes, bait x bay for black drum and total
fishes, and season x bay for black drum, hard-
head catfish, and total fishes were significant (Ta-
bles 4, 5). The second-order interaction of
bait X bay x season for total fishes was signifi-
cant (Fig. 5).

No significant differences were found in mean
lengths of black drum, hardhead catfish, red
drum, and spotted seatrout between top and bot-
tom trotlines (Tables 6-8). Significant differences
in mean length of hardhead catfish were detected
for main effects of bait and season (Table 7).

DISCUSSION

Management objectives could be better met by
requiring placement of trotline hooks on bottom
than by allowing hooks to be fished from the sur-
face. Red drum and spotted seatrout mortality
would be reduced without significantly affecting

4.^

If)

•o

O
O

^o
o^

()

CO
<0

I/I

■n

(/)

Q.
Ifl

2
•o

(1)

<fi

5
p

-n

Ji:

V-

(1)

■D

,->"

—i

■D

in
(»

^—

o
o

£:-

(TJ

(/I

■D

—

CO
CO

~ —

(11

(fl

■D

to
(J

*-'

</)

■^

(11

E
E

C/5

Q.
C/)

(0
LL

i_ (0

Q) C

S CD (0

S (0-5

^ (tl

Q. 3 T3
0.01(0
^(0 3

(/)

(0
6

E
o

o
(D
Q.
CD

(JD

t~-

0°

(£)
CM
CO

•^

■^

T- 0

CD 0

-* 1^

0 0

0 CD

—

, —

T—

t^ °

r^®

CO 0

r^ 0

0 ^

'^co

d c\j

■t

0 in

£"

CJ)

■*

CO ""

cr>

t-"

r- <D

C\J

i- 0

■^ CO

r~- ^

0 r-

d r-

cnI

£0

—

cm"

(J)

0)9

C\J

■^

CNJ^

■.- 0

r--

in 0

N-

0 ••-

•^

0 r~~-

£■

•^

^ 0

■^

in®

r- (D

tt 0

q .;.

■<T

-"-cfi

d in

0 TJ-

r-.

1^ 0

(J)

00^

■.- 0

0 0

CvJ

^CD

0 T-

1 —

0 tt

CvJ

M-

t-°

■^

■-- 0

CO 0

p ^.

^co

0 0

0 r-.

—

(j>

^°

coP

■<a- 0

r^ 0

P r^

"^co

0 Cvl

0 O)

T—

■*

05"

cm"

—

T—

CM°

inP

t^ 0

CM *-

°(0

C^liS

0 •*

0 <y>

r-~

■*

r~-

0)0

a)P

CO 0

•.- 0

°in

.,^

Pco

0 eg

0 CD

<5"

—

■^

0)°

OlP

'^ 0

CO 0

°co

'«■

^O)

0 CO

, —

0 CO

■<t

l-~

3
0

■5
01

"O
(0
0)

E
3

TO
0

■5

(11

•<a-

-t

^^^

(NJ

112

McEACHRON ET AL: BYCATCH IN TROTLINE FISHERY

£>
m

C
(0
-p

00
CD

k_
J3
0)

"D
(0

3
O)
TO

(/5
CO

E
o

o
n

a.
o

c
o

QJ
TO

sz
o

TO
O

"CJ

c
TO

O
<

^_
3
O

CD
.<

O q;

O 0)

C/5

^i;r;i:'5'~ir'~'~'~oowcD'-r--
ooooininoooocDoooo-^

OOCDOOOCDOcioCDOOOcb

»— OCDTj-cooOCXJCOOCDCMCDlfl'l-CO

r--coa)inior-.cococDLnoT}--t-coo>

C\J

lO CO o o

r^'^Lrii-ocbcNJcNJo
CO -^

00 OCVJinCMCDCD'— COO'J-CDO'3''-T-r—

CO cocO'-cDTrojcocDininocnocococo

o ocDLnr-'.-coc\iooa)ooocoa5cot-^r-^

CO h^'-OJCOOOini-Tj-^OOOCMOCO

oj "-cvJocDdooocbcbcbczicbocbcri

■^'-'— '— lOt-i-t-i— ,-^C\JCDCOCO

oooocjTj-ocoooair-^-'-Ttoo
cbc3cbc)C)cbocboc3ooodcD

V V V V V V V

cococD''-a5<DcotDcooO'-ir)cor^
CMcoT-LncocDC\j'<j-Oin'.-tor^eD'3;
T-iri-.-rO'-^dtbcb'^'
rr CD T- •■- •^

CO o o T- o o

c\ia)ocDocoa5LniOLOcoa50T-r}-T-
LO'-cD'— oocDC\icoLnr~-c\jcDC\icDco^
cDcDmr^cococ\ji-cocDO'^r-r-.co(£)
u7r--^i-ooc\jococ\ioooooc\j

CDi-dcbdcbdocbdcDdcbdcbiri

■.- ,- O T- CO

o o ■>- o o

■'-maioo-^'j-oo-^Ln
cDcbcbcDcJcbcbcbcodcDcbcridd

COOO'-S-CDt^OOCDCMCOOfOitCD
OOCDi-CpOOr^CO-^COOOCOCOOOCOCD

Lric\icMihc\ic>cb->-T-cbuSc3CDdc)

CM -^ •>- CM

CD tncOOOOCDOOOCMO-r-OOO-^CDCO-r-

co ■^■^CMr^cM-^'^'-in-^LnO'-ocM'-

00 ■^'-oooo-'-oooooooor--

CM ooooooooooooooo---

d cDdcbcDdcDdcDcbcbcbcbcDdcbd

'-oo-^T-cDCM-.-co-'-r^ococDLnco
p-r-oo'^oooooococD-'j-or-

CDOcDcricJdcocricDCDebcbcDCDcb

CM''--^cocMa)t~^rgcDcoooocMO
i-r--.T3-cpt^T}-r~.-<ircMCMCMC\jcoi-cD
co-^iricDCDcocbcoLrid-^ddcMd

>- cDi^CM''--'-'<a-o-^Tj-mcMcococor^oo

LO coi-rO'-cj3CDoocO'-Ti-Lnt^LnocM(0

CD t^CMOJCOOCO-^'^OO-^Oi-COCMCO

in CMOCMOOOOOOdOOOOOO-^

CO cbcDCDcbdcbcDcbdocDdcbdcbcNJ

[i^z:i;°^^f^'~'^'^'^cr5oococD
coooocqo5LnO'-oooa)<DOCD

cbddddcbdcocbdcbcDdcDd

V V V V V

CMO-'-T-COunCOh-COTtOOCDCO'-CM
COCD'<J;00'----COCMCOh-->-'.-0'>^-.-

dcd"^dcbcb<3r^'r^cricDC3C3c\id

^ CM T-

cnCMtDOLnCOCOinOTOOTCOCO^COCM

'-inoocoooocot^ooo'— ocD'-'-

OCMt~^C0OOOC\|T-LnOOOCMOO
OOOOOOOOOOOOOOOh^

cbdcDocbcDdcbdcricDcbddcbd

'-CMCO'-CMCO-'-CMCOCOCMCOCOCOCO'V

c
o
(/)

(1)
</>

o §

~ 00 "' -5

CO tr TO >. in

O TO Q> TO O

Q- CD en m Q.

c
o
uo

</5

•*^

Q)
00

TO
XI

c
o

TO
0)
en

c
o

U)
TO
0)

E .

If)

>.
en

>.
TO

E
B

<^ B B 5 ^ ^ ,_

TO TO TO O
CD m CD Q.

O TO
CL CD

O ^

g A

JO TO

C3)

(/>

.^^

•^^

TO
J3

T3
TO

•o

Q.
05

■D

■o

■D

LLI

—i
CO

TO
0)

< CD

o o

< <

CM CO

■•- CO

O -r-

CD CD

■* O

•<t in

o o

< <

in CO

oo CO

■■- CM

d d

< CD

CD r^

■■- CD

C\J ■.-

d d

< <

LO in

•>- CD

00 CD

(U QJ

■O JJ

TO TO

C C

3 3

O) 03

TO TO

_l _l

k_ k_

03 03

1=1- g

Q. o

D -J

C
TO
0)

03
CL
Q.

c3 =>

03 -.^

? i

g^
c o

TO •
O

^ CL
C ^-'
03 >,

.£ TO

>..y

TO ^

X -^

^ 03

S 2?

05 TO

C
TO

03
C
T3
TO 03

03
O
03

<u <n 03

05 TO -o

;^ ™

c
o

03
TO

TO
O

t 03 O

— TO — i

C 03 -I

TO 03 iu

03 ^ il

5 — "O

I >- TO

I -Q 2

^"S

TO
- I §

m o o3

2 P "1

c
o

03
TO
03

< CD m <

CO Tj- r^ CD

03 tT ■.- tT

OJ -"t CD 00

d d d d

< < < <

CD ■>- CO in

CM O •* O

CO O -^ CO

c:i d d> d

< CD < <

o in in CD

00 O CO ■•-

o c\j o o
cs d d d

< < < <

CO oo CO o
O CM CO O

o o o o
d d d d

< CD < <

h- 00 r-- CD

■^ 00 t^ o

o ■.- o 1-

d d d d

< CD < <

oo ■^ in TT

CD CD CO •*

O TJ- o o

d d d d

< < < <

O TJ- CD O

o CM r~. o

■•- C\J o ••-

d d d d

O < O O

CM oo CO in

•,-■,- r-- ■*

CM T- T- o

d d d d

CD < CD CD

O O ■* -r-

"- o ■* o

CO CM ■^ CM

d d d d

< < < <

■* CD CD r^

•■- CM o in

CD CO O 1^

d d y-^ d

w g'E
= -E i E

TO •> CL 3
LL S CO Cn

113

0.12

0.10-

O.OB

3 0.06

< 0.04
a.

I
u

< 0.02
o

0.00-

— I —

ULM

FISHERY BULLETIN; VOL. 86, NO. 1

SHRIMP

TOP

CRAB

LEAVES

TOP

BOTTOM

— I

LLM

— I

ULM

— I

LLM

— I

ULM

— I

LLM

BAY SYSTEM

Figure 4.— Significant second-order interaction among positions, baits, and bay system for spotted
seatrout catch rates. LLM = lower Laguna Madre, ULM = upper Laguna Madre.

Figure 5. — Significant second-order interac-
tion among bait types, bay system, and sea-
sons for total fishes catch rates. LLM = lower
Laguna Madre, ULM = upper Laguna
Madre.

- 1.5

• 1.2

I 0.9

<
K

I
O

I-
<

uj 0.6 •
0.3
0.0 ,.
l.B
1.5 ■
1.2
0.9
0.6
0.3
0.0-
FALL

CRAB

SHRIMP

LLM

ULM

LEAVES

LLM

ULM

I I

WINTER SPRINS

SEASON

SUMMER

114

McEACHRON ET. AL: BYCATCH IN TROTLINE FISHERY

Table 6. — Mean length (mm t 1SE) of black drum, hardhead catfish, red drum, and spotted
seatrout caught on top and bottom trotlines by bait, bay and season during February 1985-
January 1986 Number in parentheses is number of fish measured.

Position

Bait

Species

Top

Bottom

Crab

Shrimp

Leaves

Black drum

539 ± 10

550 ± 13

566 ± 1 1

493 ± 12

588 i 21

(177)

(146)

(172)

(108)

(43)

Hardhead catfish

338 ± 1

343 ± 2

346 ± 1

328 ± 2

337 ± 3

(1.363)

(675)

(1,201)

(651)

(186)

Red drum

521 ± 4

504 ±9

540 ± 6

481 ±6

542 ± 7

(657)

(173)

(260)

(336)

(234)

Spotted seatrout

439 ± 9

420 i 24

471 ± 23

426 ± 10

445 ± 19

(111)

(18)

(17)

(86)

(26)

Bay

Season

Upper

Lower

Laguna

Species

fvladre

(vladre

Fall

Winter

Spnng

Summer

Black drum

484 ± 19

560 ±8

464 ± 20

538 ±8

630 ± 21

416 ± 19

(67)

(256)

(35)

(201)

(71)

(16)

Hardhead catfish

327 ± 1

348 ± 1

329 ±2

368 ±2

333 ±2

345 ±2

(824)

(1,214)

(538)

(277)

(724)

(499)

Red drum

506 ± 7

523 ± 5

493 ± 1 1

506 ± 5

546 ±9

551 ± 10

(308)

(522)

(133)

(427)

(114)

(146)

Spotted seatrout

411 ± 15

443 ± 10

410 ±21

404 ± 1 2

482 ± 25

478 ± 13

(29)

(100)

(25)

(51)

(16)

(37)

Table 7. — Summary of results of the AOV s of mean length for black drum (winter season only), hardhead catfish (excludes leaves and
winter season), and red drum (winter and spring seasons combined) on top and bottom trotlines in lower Laguna Madre during February
1985-January 1986. NA = not analyzed.

Black drum

Hardhead catfish

Red drum

Source of

Sum of

variation

squares

PR > F

squares

PR>F

squares

PR>F

Total

176

1,941,401

1,050

2,310,898

521

6,484,793

Position

5,978

0.15

0.70

0.02

0.90

6,122

0.26

0.61

Bait

15,082

0.19

083

33,650

6.60

0.01

89,151

1.90

0.15

Season

89,202

5.83

<0.01

133,190

2.85

0.06

Position X bait

49,063

0.69

0.55

8,242

1.62

0.21

10,913

0.23

0.79

Season ^ position

1,517

0.10

0.96

789

0.02

0.98

Season x bait

6,132

0.40

0.75

68,535

0.73

0.57

Season '; position x bait

10,269

0.67

0.57

63,388

0.68

0.61

Set (position x bait)

10,906,616

8.26

-O.OI

Set (season '- position

■ bait)

453,750

3.22

<0.01

2,129,491

2.61

<0.01

Error

144

704,611

946

1,499,474

413

3,706,235

Table 8. — Summary of results of the AOV of mean length for spot-
ted seatrout (shnmp bait only) on top and bottom trotlines in upper
and lower Laguna Madre combined dunng February 1985-
January 1986.

Source of
variation

Sum of
squares

PR>F

Total

750,362

Position

601

0.04

0.84

Season

58,796

1.32

0.29

Position X season

58,462

1.31

0.29

Set (position ^ season)

416,855

3.42

<0.01

Error

217,744

115

FISHERY BULLETIN: VOL 86, NO 1

black drum catches. Operational efficiency of
commercial fishermen should improve with less
handling of nontarget species. Mortality of non-
target fishes would decrease because they would
not be caught and subsequently handled. For red
drum and spotted seatrout that are caught, sur-
vival would be high for those released back into
the water. Survival of released red drum caught
on trotlines in winter and summer and of spotted
seatrout in winter was 100^^ (Martin et al. 1987).
About 509^ of the spotted seatrout died in summer
cage studies; but few commercial trotlines are
fished during this period (TPWD unpubl. data).
Thus, the goal of reducing the catch of nontarget
species and reducing mortality due to trotlines
can be achieved with minimal impact on the com-
mercial fishermen.

Interactions between bay system and the other
three factors for some species probably reflect dif-
ferences in relative abundance. Fewer fish were
available to be caught in upper Laguna Madre
than in lower Laguna Madre (Crowe et al. 1986).
The effects of bait and season on trotline catches
cannot be determined in bay systems where the
fish abundance approaches zero.

No spotted seatrout were caught on crab bait on
bottom in either bay; but they were caught on all
baits on top in the lower Laguna Madre leading to
the significant second-order interaction of posi-
tion X bay X bait. This condition was not unex-
pected because spotted seatrout are predomi-
nately sight feeders (Vetter 1977), and might not
take baits on bottom as readily as baits suspended
in the water column.

Crab and shrimp were more effective baits than
oleander leaves for all four species. No other gen-
eralizations could be made concerning baits and
seasons. Selection of crab or shrimp as the bait of
choice for reducing bycatch while maximizing
black drum catch is unclear because catch rates
for black drum and red drum were greater on crab
than shrimp, especially in winter.

ACKNOWLEDGMENTS

We would like to thank all Laguna Madre field
personnel who diligently collected the samples.
Tom Heffernan, Ed Hegen, Lynn Benefield,
Maury Ferguson, and Tony Maciorowski re-
viewed the manuscript.

LITERATURE CITED

Anonymous

1979. Saltwater finfish research and management in

Texas. A report to the Governor and the 66th
Legislature. Tex. Parks Wildl. Dep., Coastal Fish.
Branch, PWD Rep. No. .3000-59, 21 p.

1981. Saltwater finfish research and management in
Texas. A report to the Governor and the 67th
LegLslature. Tex. Parks Wildl. Dep., Coastal Fi.sh.
Branch, PWD Rep. No. ;J000-10K, ,31 p.

1983. Saltwater finfish research and management in
Texas. A report to the Governor and the 68th
Legislature. Tex. Parks Wildl. Dep., Coastal Fish.
Branch, PWD Rep. No. 3000-154, 48 p.
Breuer, J P

1973. A survey of the juvenile and adult food and game
fish of the Laguna Madre. Tex. Parks Wildl. Dep.,
Coastal Fish. Branch, Proj. Rep. 173-202.

1974. Juvenile and adult food and game fish of the
Laguna Madre. Tex. Parks Wildl. Dep., Coastal Fish.
Branch, Proj. Rep. 109-130.

1975. Biological studies in the lower Laguna Madre of
Texas, 1975. Tex. Parks Wildl. Dep., Coastal Fish.
Branch, Proj. Rep. 158-196.

Crowe, A L , L W McEachron, and P C Hammerschmidt.

1986. Trends in relative abundance and size of selected
finfish in Texas bays: November 1975-December
1985. Tex. Parks Wildl. Dep., Coastal Fish. Branch,
Manage. Data Ser. No. 114, 259 p.

Elliott. J. M

1979. Some methods for the statistical analysis of
samples of benthic invertebrates. Freshwater Biol.
Assoc, Sci. Publ. No. 25, 160 p.

HoESE. H D , AND R. H Moore

1977. Fishes of the Gulf of Mexico, Texas, Louisiana, and

adjacent waters. Texas A&M Univ. Press, College

Station, 327 p.
Martin. J H , K W Rice, and L W McEachron

1987. Survival of three fishes caught on trotlines. Tex.
Parks Wildl. Dep., Coastal Fish. Branch, Manage. Data
Ser. No. Ill, 21 p.

Matlock, G. C

1980. History and management of the red drum fishery.
In Proceedings Colloquium on red drum and sea-
trout, p. 37-54. Gulf States Mar. Fish. Comm. No.
5.

Matlock, G C , P L Johansen, and J P Breuer

1979. Management of red drum in a Texas estuary - a
case study. Proc. Annu. Conf Southeastern Assoc.
Fish Wildl. Agencies 33:442-450.

McEachron, L W , A W Green, G. C Matlock, and G E. Saul.

1985. A comparison of trotline catches on two hook types
in the Laguna Madre. Tex. Parks Wildl. Dep., Coastal
Fish. Branch, Manage. Data Ser. No. 86, 44 p.

1986. Evaluation of the commercial trotline fishery in
the Laguna Madre during fall 1984. Tex. Parks Wildl.
Dep., Coastal Fish. Branch, Manage. Data Ser. No. 93,
25 p.

McEachron. L W . G C Matlock. A R Martinez, and J P.
Breuer

1980. Evaluation of natural, leaf, vegetable, worm and
cork baits used on trotlines in upper and lower Laguna
Madre, Texas (September 1977-October 1978). Tex.
Parks Wildl. Dep., Coastal Fish. Branch, Manage. Data
Ser. No. 8, 68 p.

Robins, C R , R M Bailey, C E Bond. J R Brooker, E A
Lachner, R N Lea, and W. B Scott.

1980. A list of common and scientific names of fishes
from the United States and Canada. (4th ed. ) Am.

116

McEACHRON ET AL: BYC'AIVH IN TROTLINE FISHERY

Fish. Soc. Spec. Pub. No. 12, 174 p. SOUTH ATLANTIC FISHERY MANAGEMENT COUNCIL.

SAS Institute. iNC 1985. Source document for the swordfish fishery

1980. SAS supplemental library user's fjuide. SAS management plan. South Atl. Fish. Manage. Counc,

Institute Inc. Cary, NC, 202 p. Charleston, SC.

1982. SAS users guide: Statistics. SAS Institute Inc. Vetter, R D

Cary, NC, 584 p. 1977. Respiratory metabolism of, and niche separation

Simmons, E G . and J P Breuer between two co-occurring congeneric species, Cynoscion

1962. A study of redfish, Sciaenops ocellata Linnaeus nebulosus and Cynoscion arenarius in a south Texas

and black drum, Pogonias cromis Linnaeus. Publ. estuary. MA Thesis, Univ. Tex, 113 p.
Inst. Mar. Sci., Univ. Tex. 8:184-211.

117

LARVAL DEVELOPMENT OF BLUE GRENADIER, MACRURONUS
NOVAEZELANDIAE (HECTOR), IN TASMANIAN WATERS

B D BruceI

ABSTRACT

The development of Macruronus novaezelandiae is described and illustrated from both reared speci-
mens and larvae from Tasmanian waters. Eggs of M. novaezelandiae are pelagic, spherical (1.08-1.18
mm diameter), and have a single oil droplet (0.36-0.42 mm diameter). Eggs hatch after 55-60 hours
at 14°-19°C. Larvae are 2.2-2.3 mm at hatching. Characteristic pigmentation, a myomere count of
78-80, and the sequence of fin development separate M. novaezelandiae from other known gadiform
larvae. Development is direct, with no marked change in body morphology. Fin development proceeds
in the sequence: second dorsal, anal, first dorsal, pelvic, caudal, pectoral. However, adult fin comple-
ments are reached in the sequence: first dorsal, pelvic, anal, second dorsal, caudal, pectoral.

Caudal development is late in Macruronus. Flexion begins at 20 mm and is not complete until 28
mm. The caudal fin is based on two ural centra, four hypurals, two epurals, and a parhypural. X and
Y bones are present although they are not readily distinguishable from dorsal and anal pterygio-
phores.

The genus Macruronus comprises four nominal
species, which occur in southern temperate conti-
nental shelf and slope regions. Two species,
Macruronus novaezelandiae and M. magellanicus
support commercial fisheries. The blue grenadier,
M. novaezelandiae , forms the basis of fisheries in
New Zealand and Australia where total annual
catches range up to 97,750 and 1,100 t respec-
tively (Patchell 1982; Wilson 1981, 1982).
Macruronus magellanicus is fished commercially
off South America. The remaining species, M.
maderensis and M. capensis , are known only from
a limited number of specimens (Svetovidov 1948;
Cohen 1986). Despite their economic importance
and widespread distribution, very little is known
of the early life history of any member of the
genus. Patchell (1982) identified winter spawn-
ing grounds on the west coast of the South Island
for New Zealand populations of M. novaeze-
landiae and similarly Wilson (1981, 1982) has
suggested a winter spawning, on the west coast of
Tasmania, for Australian M. novaezelandiae.
This paper presents the first published informa-
tion on the larvae of Macruronus .

In 1984, the Division of Fisheries Research of
the Commonwealth Scientific and Industrial Re-
search Organization established a multidisci-

ICSIRO Division of Fisheries Research, GPO Box 1538, Ho-
bart, Tasmania 7001, Australia; present address: South Aus-
tralia Department of Fisheries, GPO Box 1625, Adelaide, South
Australia 5001, Australia.

Manu.scnpt accepted September 1987.
FISHERY BULLETIN: VOL. 86. NO. 1, 1988.

plinary program to investigate the biology and
ecology of blue grenadier in Tasmanian waters.
An integral part of this program was a study of
larval ecology. As such, it was first necessary to
establish criteria for the identification of blue
grenadier larvae. This paper describes the larval
development of M. novaezealandiae from Tasma-
nian waters.

MATERIALS AND METHODS

Specimens were obtained from samples col-
lected aboard the CSIRO Fisheries Research Ves-
sel Soela between April 1984 and September
1985. Details of sampling strategies, locations,
and procedures will be described in a subsequent
manuscript. Larvae were obtained by sampling
with a rectangular midwater trawl (RMT 1+8;
Baker et al. 1973), aim diameter ring net (500
[im mesh), and free-fall, vertical drop nets of 64
|jLm and 200 fxm mesh (Heron 1982). Juvenile
specimens were obtained with an Engels 352
pelagic trawl fitted with a 10 mm liner.

Newly hatched larvae were reared from eggs
stripped and fertilized at sea. Eggs and milt
stripped from ripe adults trawled from 500 m
were mixed in 1 L plastic jars filled with sea-
water. Despite the jars being located in a sea-
water bath, incubation temperatures varied con-
siderably (14°-19°C). On return to the
laboratories at Hobart, the eggs were transferred
to 2 L glass jars and placed in a constant temper-

119

FISHERY BULLETIN: VOL. 86, NO. 1

ature incubation chamber set at 14.0° ± 0.2°C. In-
cubation jars were not aerated, and no attempt
was made to feed the larvae.

All specimens used for description were fixed in
a IQ% formalin-seawater solution buffered with
sodium p-glycerophosphate and later transferred
to a 59^ solution.

This description is based on a series of 74 lar-
vae, 2.2-34.2 mm in length, although comments
on pigment and meristic variability stem from
routine examination of several hundred speci-
mens. A representative series of larvae is de-
posited with the South Australian Museum, Ade-
laide, South Australia.

Developmental terminology follows Ahlstrom
et al. (1976). Body measurements follow Matarese
et al. (1981). Length measurements are reported
as notochord length, NL (i.e., from the snout tip to
the end of the notochord) in preflexion and flexion
larvae, and standard length, SL (i.e., from the
snout tip to the posterior margin of the superior
hypural elements) in postflexion larvae and juve-
niles. Larvae were measured under a dissecting
microscope fitted with an ocular micrometer and
a camera lucida. Juveniles were measured with
vernier calipers.

Meristic counts and examination of ossification
sequences were made on specimens cleared and
stained using Alizarin Red S-KOH-glycerine
(Hollister 1934). Caudal osteology follows Inada
(1981), Marshall and Cohen (1973), and Monod
(1968).

Vertebral counts include the first vertebrae,
the neural spine of which is fused to the supraoc-
cipital crest (Marshall 1966), and both ural cen-
tra. Vertebral centra were counted as ossified
only when a complete band of stain connected
both neural and haemal spines.

RESULTS

Identification of M. novaezelandiae larvae was
based on their typical gadiform morphology
(large head, compact gut, tapering body form),
myomere count, and the development of confluent
dorsal-caudal-anal fins (see section on Distin-
guishing Features). Identification of field-
collected specimens was confirmed by comparison
with reared larvae.

Distinguishing Features

Prior to median fin development, myomere
counts are useful in separating M. novaezelandiae

larvae (78-80) from similarly pigmented morid
(41-72), macrourid (10-16 + 70 > 100), gadid
(39-64) and other known merlucciid larvae (48-
58) which they superficially resemble (Marshall
and Iwamoto 1973; Fahay and Markle 1984;
present study).

Both M. novaezelandiae and most morid larvae
show moderately pedunculate pectoral fins, a fea-
ture common in gadiform larvae with delayed
caudal development (Fahay and Markle 1984).
Macrourid larvae, in contrast, have very promi-
nently stalked pectorals and can further be sepa-
rated from M. novaezealandiae and most morids
by precocious development of the pelvic fin.

Size at caudal flexion and the sequence of fin
development are also useful in separating M.
novaezelandiae from all Merluccius species. In
Merluccius , notochord flexion generally begins at
about 9 mm and the caudal fin is the first to form
(Dunn and Matarese 1984; Fahay and Markle
1984). Macruronus novaezelandiae larvae do not
begin caudal flexion until approximately 20 mm,
and the caudal fin is the second last to form.

Macruronus novaezelandiae larvae have 1-3
prominent melanophores along the ventral mid-
line of the tail (although variable in appearance,
see section on Trunk and Tail Pigmentation)
and a double series of dorsal melanophores. When
expanded, melanophores in these two regions
coalesce to give the appearance of a broad
postanal band. Postanal banding patterns are
widespread in gadoid larvae (Fahay and Markle
1984); however, unlike many gadoid larvae, M.
novaezelandiae lacks pigment at the notochord
tip.

At larger sizes M. novaezelandiae larvae de-
velop long-based dorsal and anal fins confluent
with the caudal fin. Other gadoid larvae with this
configuration have markedly different pigmenta-
tion (see Fahay and Markle 1984 for details).
Ophidiiform larvae have confluent dorsal, caudal,
and anal fins but can be separated from M. no-
vaezelandiae by their lack of a separate first dor-
sal fin and general lack of body pigment (see Gor-
don et al. 1984).

Development

Embryonic development has not been treated
in detail here as it is the subject of a manuscript
in preparation by G. Patchell (Fisheries Research
Centre, Wellington, New Zealand).

The pelagic eggs of blue grenadier are spheri-
cal, with an unsegmented yolk and a smooth cho-

120

BRUCE: LARVAL DEVELOPMENT OF BLUE GRENADIER

rion. Late-stage eggs are 1.08-1.18 mm in diame-
ter with a single oil droplet of 0.36-0.42 mm di-
ameter (Fig. lA). Reared larvae hatch at 2.2-2.3
mm after 55-60 hours ( 14'-19"C). Newly hatched
larvae have a posteriorly positioned oil droplet
and adopt a head down position in rearing con-
tainers. Yolk absorption was incomplete in speci-
mens reared to 3.7 mm (6 days posthatch), al-
though the smallest field-collected larvae (3.6
mm) had already completed yolk absorption. The
anus opened laterally to the right in all reared
larvae and 95*^ of field-collected larvae. The anus
becomes symmetrical by 5.1 mm. A lateral anal
opening in M. novaezelandiae is consistent with
the developmental pattern reported for other gad-
iform species (Marak 1967; Matarese et al. 1981;
Fahay and Markle 1984; Dunn and Vinter 1984).
Field-collected larvae are moderately elongate
with the greatest body depth (16-229? body
length) occurring at or near the pectoral fin base
(Table 1). Head length as a proportion of body
length (BL) remains relatively constant at about
22% BL throughout the larval phase, decreasing
to about \1'7( BL in juveniles. Eye diameter de-
creases from 99f BL in preflexion larvae to A'7c BL
in juveniles. Depth at anus remains relatively
constant at about 139^ BL in larvae and juveniles.
Distances from the snout tip to the first dorsal fin
and from the snout tip to the anal fin decrease
slightly during development from about 279?^ BL
to 2l7c BL and 519r BL to 469? BL respectively.

in the number of melanophores and their degree
of expansion. Although Badcock and Merrett
(1976) suggested the appearance of melanophores
can change on a diurnal rhythm, in the blue
grenadier larvae examined, there was no conspic-
uous relationship between time caught and me-
lanophore expansion.

Head Pigmentation

Newly hatched larvae (2.2-2.3 mm) have
melanophores concentrated in front and behind
the eye (Fig. IB). Melanophores increase in num-
ber and extend over the sides of the head and
snout by 3.3-3.5 mm (reared larvae, Fig. IC).
Melanophores migrate dorsally to the top of the
head by 3.6 mm (Fig. ID). Eyes become pig-
mented at this size in reared larvae. By 4.5 mm,
the dorsal pigment on the head consists of a group
of 3-11 melanophores scattered over the hind-
brain and posteriorly to above the cleithrum. Pig-
ment gradually extends over the midbrain, with 1
or 2 melanophores usually present between the
eyes by 5.3 mm. Melanophores develop externally
over these initial mid- and hindbrain spots and
extend posteriorly as a double row to the dorsal
fin anlage by 7.2 mm. Dorsal pigment gradually
intensifies: melanophores increase in number
and form a cap over mid- and hindbrains by 16.0
mm. Melanophores extend down between the eyes
to the tip of the maxilla by 12.0 mm. Internal

Table 1 — Body proportions of larvae and juveniles of Macruronus novaezelandiae (expressed as percentage NL or SL): mean, standard

deviation, range.

Preflexion

Flexion

Postflexion

Juvenile

Body proportions
sample size

range

length (mm)

8.9

4.5

(3.6-19.0)

23.5

2.3

(20.6-26.1)

300

2.9

(27.6-34.2)

1890

1.4

(188.0-190.0)

head length

227

1.7

(18.3-24.7)

23.7

1.0

(223-24.7)

22.3

1.4

(20.5-23.6)

17.6

0.4

(17.3-17.9)

eye diameter

0.7

(8,1-10.3)

7.9

0.4

(7.3-8.3)

7.3

0.7

(6.5-8.0)

4.2

0.4

(4.4-5.2)

snout length

6.1

0.9

(4.6-7.7)

6.2

0.6

(5.7-7.0)

5.9

0.5

(5.4-6.5)

4.9

0.1

(4.8-5.0)

depth at pectoral

223

1.7

(21.0-24.4)

17.9

0.6

(17.2-18.5)

16.6

2.0

(13.7-18.5)

13.1

0.1

(13.0-13.2)

depth at anus

120

2.4

(8.2-15.2)

13.1

0.3

(12.6-13.4)

13.0

0.8

(12.3-13.8)

12.7

0.6

(12.3-13.1)

snout to first

dorsal fin

27.5

1.3

(25.2-29.3)

26.6

0.8

(26.0-27.7)

25.3

0.8

(24.2-26.0)

20.7

0.2

(20.6-20.9)

snout to anal fin

51.4

0.9

(50.0-52.6)

50.4

0.5

(49.6-50.6)

46.6

1.5

(45.0-48.4)

46.5

1.2

(45.7-47.4)

Pigmentation

Although pigmentation in M. novaezelandiae is
variable, certain features persist that, when com-
bined with meristic and morphometric informa-
tion, enable identification. Variation in the ap-
pearance of pigmentation is a result of differences

pigment expands over the forebrain in larvae
from 9.0 to 15.0 mm.

Ventral pigment on the head first develops in
4.2 mm larvae as 3-5 melanophores between the
dentaries. The number of melanophores increases
to 10-12 by 12.0 mm.

121

FISHERY BULLETIN: VOL, 86, NO. 1

Figure 1. — Development of Macruronus novaezelandiae: A)
Late stage egg 1.08 mm diameter, oil droplet 0.37 mm di-
ameter, B) 2.2 mm.; C) 3.5 mm.; D) 3.6 mm.; E) 5.3 mm.; F)
7.2 mm.; G) dor.sal view of above; H) 12.0 mm.; I) 24.2 mm.
postanal myomeres omitted. A-C reared specimens, D-I
field-collected.

BRUCE: LARVAL DEVELOPMENT OF BLUE GRENADIER

123

FISHERY BULLETIN; VOL. 86, NO. 1

The onset of dentary pigment is variable; no
pigment may be present on some larvae as large
as 7.0 mm. Most larvae develop 1 or 2 melano-
phores over the posterior section of the dentary by
5.3 mm and add melanophores anteriorly along
its length, with 5 or 6 usually present by 7.1 mm.

Two melanophores are often present around the
otic capsule by 7.0 mm, but they are obscured by
overlying tissue in 10.0 mm larvae. Scattered
melanophores develop over the pterotic region by
25.0 mm, but the operculum and preoperculum
remain largely unpigmented, even in the largest
specimen examined (34.2 mm).

Trunk and Tail Pigmentation

Newly hatched larvae have melanophores on
the body above the yolk sac and ventrally on the
tail. Some pigment is also present on the yolk sac
near the developing gut and scattered over the oil
droplet. Pigment forms a cap over the gas bladder
by 4.2 mm. Melanophores are gradually added to
the lateral surfaces of the gut throughout the lar-
val period until the entire gut (including the ven-
tral surface) becomes pigmented by 30.0 mm.

Dorsal pigment first appears on larvae 3.8-4.5
mm as scattered melanophores at approximately
607c NL. Melanophores rapidly increase in num-
ber and form a double row, extending from 51% to
67% NL in larvae of 5.0 mm. Lateral melano-
phores may also develop above the body midline
in this region. Concurrently, a similar double row
of melanophores appears and extends posteriorly
from the head (Fig. IG). The head and tail rows
join by 10.5 mm. Melanophores appear posteri-
orly in the caudal region by 29.0 mm forming a
twin series one either side of the developing dor-
sal fin. Pigment also appears internally on the
dorsal surface of the vertebrae in larvae of 9.5 mm
and extends anteriorly to approximately 50% SL
and posteriorly to the last vertebrae by 34.0 mm.

Single melanophores appear on the dorsal fin
ray bases by 14.0 mm and are present on all bases
by 29.0 mm.

Pigment along the ventral midline of the tail
appears in newly hatched larvae as a diffuse re-
gion that extends posteriorly from the yolk sac to
75-82% NL. This contracts to 1-3 melanophores
(most commonly 2) located 52-65% NL in larvae
of 3.8-4.0 mm. Additional melanophores (up to 6)
may appear later, but the initial 1-3 melano-
phores persist throughout the larval period. In
larvae larger than 7.0 mm, the initial 1-3
melanophores appear internally above the anal

fin ray bases and are gradually obscured by both
overlying musculature and external melano-
phores. These ventral melanophores on the tail
are a useful diagnostic character, although their
appearance varies, depending on their degree of
expansion. This variability in melanophore ap-
pearance is particularly evident in small larvae
where expanded ventral melanophores may ex-
tend over the lateral surfaces of the body to al-
most the dorsal area (Figs. ID, 2).

Lateral pigment gradually intensifies through-
out the larval period, excepting the area immedi-
ately above the gut, which remains largely devoid
of pigment even in the largest specimen (34.2
mm).

Morphological Variability

Macruronus novaezelandiae larvae showed
some size variation in development. In general,
specimens captured in ring net and RMT samples
appeared to develop features at slightly smaller
sizes than those taken from drop net samples.
This is likely a result of difi"erential shrinkage of
specimens caught by the different capture sys-
tems. Hay (1981) reported that considerably more
shrinkage occurred in Pacific herring when lar-
vae were killed prior to fixation and that shrink-
age increased with tow length. Ring net and RMT
tows varied in duration from 15 to 110 minutes,
with most larvae dead by the time the net was
retrieved and the catch fixed. Drop net sampling,
in contrast, lasted for, at most, 3 minutes dura-
tion, and many larvae were still alive on fixation.
Some variability in development can also be ex-
pected in field-collected larvae as a reflection of
past history (e.g., feeding success), although it is
unlikely such variations would account for the
observed differences between larvae caught by
different techniques.

Meristics and Osteology (Table 2)
Head and Axial Skeleton

In laboratory-reared larvae, jaw development
was first visible after 3.5 days (posthatch) with a
functional mouth present in larvae of 5.5 days
(3.7 mm). Pigmentation of the eyes also occurred
at this time suggesting that larvae were ready for
first feeding. The smallest larva stained was a
field-collected specimen 3.7 mm NL. The maxilla,
premaxilla, dentary, and cleithrum were all ossi-
fied in this specimen.

124

BRUCE: LARVAL DEVELOPMENT OF BLUE GRENADIER

Figure 2. — Variability of ventral pigment on the tail in 4.9 mm larvae of Macruronus novaezelandiae.

Ossification of branchiostegals begins in larvae
of 4.6 mm, with the full complement of 7 ossified
by 11.5 mm. Gill rakers are first discernible in
larvae of 9.4-9.9 mm with the full complement of
7 + 22-23 present by 28.9 mm.

Ossification of neural and haemal spines gener-
ally precedes that of the vertebral centra. Ossifi-
cation of centra, neural spines, and haemal spines
occurs sequentially from anterior to posterior pro-
ceeding slowly in larvae less than 9.0 mm in
length and then more rapidly until the full com-
plement is ossified by 23.2 mm. Elements associ-

ated with the caudal complex are the last to os-
sify.

Fins

Completion of fin development in M. novaeze-
landiae occurs in the sequence: first dorsal and
pelvic (almost simultaneously), anal, second dor-
sal, caudal, pectoral.

Pelvic fins first appear in larvae of 5.7— 5.8 mm
as slight swellings either side of the gut. They do
not form distinct buds until 6.9 mm. Ossification

125

FISHERY BULLETIN: VOL 86, NO 1

Table 2.— Meristic counts from cleared and stained larval and juvenile Macruronus novaezelandiae Specimens between dashed lines are

undergoing notochord flexion, a = specimen damaged; b juveniles not stained.

Length
(mm)

3.9

4.2

4.6

4.8

5.2

6.0

9.4

9.9

11.5

16.3

17.4

19.8

dorsal

4 + 28

0 + 19

9 + 74

12 + 84

12 + 86

13 + 87

Fin rays

anal pectoral pelvic

Branchi-

ostegal

rays

Gill rakers

upper lower total

Total
centra

Neural
spines

Haemal
spines

18
4
60
73
76
86

4
8
8

1
1
3
3
5
6
6
7
7
7
7

8
6

12
15
15
15

8
6
12
16
18
20

5
42
38
55
70
71
73

37
34
42
55
55
57

Caudal
elements

3
3

23.2

12+ 100

26.1

13 + 99

28.9

13 + 99

188

13 + 94

190

13 + 96

may start as early as 9.4 mm with the full comple-
ment (8 rays) present by 16.3 mm. Ossification
proceeds from the outer to the innermost rays.

The second dorsal fin anlage is visible in larvae
of 5.7 mm. Bases are first visible by 6.9 mm, with
ray ossification commencing by 7.3 mm. Al-
though the anal fin anlage does not form until 6.9
mm, complete ossification is reached before that
of the second dorsal. Distinct anal fin bases are
first visible in 7.2 mm larvae and ossification has
consistently begun by 9.9 mm. The full comple-
ment of anal rays is present by 21.0 mm and for
the second dorsal, by 23.2 mm.

The first dorsal starts development slightly
later than the second dorsal, although it is the
first fin to complete ossification. The full comple-
ment of 12 or 13 elements is present by 16.3 mm.

Pectoral buds were first observed in larvae 4.5
days posthatch (3.2 mm). However, the pectoral
fin is the last to complete development. Ossifica-
tion of pectoral rays starts by 16.3 mm; a 34.2 mm
specimen had only 15 ossified rays, still short of
the 20 rays of juveniles. Sequence of ossification
is from upper to lower.

The caudal fin anlage first appears on the ven-
tral surface of the notochord just anterior to the
tip in larvae of 10.4 mm. Flexion begins at 20 mm
and is usually complete by 28 mm. Ossification of
all caudal elements was incomplete in a 34.2 mm
specimen. Insufficient material of the appropriate
size was available to define the completion of cau-
dal ossification.

The caudal complex in M. novaezelandiae is
based on two ural centra, two epurals, a superior
hypural (HP3 + 4), inferior hypural (HPl + 2),
and a parhypural (Fig. 3). Eight to nine rays ar-
ticulate with these elements — one or two rays on
the second epural, three rays on the superior hy-
pural, two on the inferior hypural, and one ray
each on the first epural and the parhypural. Sin-
gle rays also articulate with the elongate neural
and haemal spines of the first preural centrum. X
and Y bones are present although they are not
readily distinguishable from dorsal and anal
pterygiophores. Total caudal fin ray counts are
low (12 or 13).

Additional caudal structures occurred in one of
the six specimens examined. This specimen had a
twin haemal spine on the first preural centrum
and greatly elongated haemal spines on preural
centra 3-8 (1.3-1.4 times the length of corre-
sponding neural spines, Fig. 3).

DISCUSSION

The general morphology and pigmentation of
M. novaezelandiae larvae show broad similarities
to Merluccius and to gadine gadids. Characteris-
tic differences between M. novaezelandiae and
Merluccius species occur in fin structure and the
sequence of fin development. In Merluccius, the
caudal fin is the first to form, followed by the
pelvic. In Macruronus, caudal development is
late with the caudal fin being the second last to

126

BRUCE: LARVAL DEVELOPMENT OF BLUE GRENADIER

Additional
Haemal Spine

Figure 3. — Caudal osteology of a juvenile Macruronus novaezelandiae (181 mm SL).
X = X bone, Y = Y bone, EP = epural, SH = superior hypural (hypurals 3 + 4), IH
= inferior hypural (hypurals 1 + 21, PH = parhypural, U = ural centra, PU = preural
centra.

form. The pectoral fin in Macruronus larvae is
more markedly stalked than in Merluccius.
Fahay and Markle ( 1984) suggested that this pec-
toral modification in larvae with delayed caudal
development may be a compensatory response as-
sociated with swimming.

Although the larvae of the remaining mer-
lucciid genera (Lyconus and Lyconodes) are cur-
rently unknown, fin structure and position should
be useful in separating these from Macruronus.
Based on adult features, pelvic insertion should
distinguish Macruronus (pelvics inserted behind
pectorals) from Lyconus (opposite) and Lyconodes
(abdominal). Additionally, Lyconus has only a
single dorsal fin and no caudal fin.

The caudal fin of M. novaezelandiae is similar
to Muraenolepis in its confluence with dorsal and
anal fins (Fahay and Markle 1984). This similar-
ity extends to the undifferentiated X and Y bones
and the total caudal fin ray count (12 or 13) re-
ported by these authors. However, unlike Mu-
raenolepis, M. novaezelandiae has radials fused to
the spines of the first preural centrum, which is
the more typical gadoid condition.

Variability in the structure and appearance of
bones associated with the caudal fin has been re-
ported for other Macruronus species. Marshall
( 1966) observed double neural arches and "super-
numary elements" in M. magellanicus. Indeed,
variability in gadiform caudal structure ap-
pears not to be unusual with examples in several
taxa (Markle 1982). Unfortunately, insufficient
specimens in the appropriate 35-150 mm size
range were available to assess developmental

characteristics of these variations in blue
grenadier.

ACKNOWLEDGMENTS

I thank R. Thresher, J. Gunn, J. Leis, and A.
Miskiewicz for their reviews of the manuscript. I
also thank D. Furlani for sorting the samples and
for her considerable patience in the laboratory.
This work was supported by a grant from the
Fisheries Industry Research Trust Account.

NOTE: Since the acceptance of this paper, the
embryological work by A. Patchell (see section on
Development) has been published in New
Zealand Journal of Marine and Freshwater Re-
search Vol. 21, No. 2. That paper includes a simi-
lar larval developmental sequence to that re-
ported here.

LITERATURE CITED

Ahlstrom. E H , J L Butler, and B Y Sumida

1976. Pelagic stromateoid fishes (Pisces, Perciformes) of
the eastern Pacific: kinds, distributions and early life
histories and observations on five of these from the north-
west Atlantic. Bull. Mar. Sci. 26:285-402.
Badcock, J R . AND N R Merrett

1976. Midwaterfishes in the eastern North Atlantic. I.
Vertical distribution and associated biology in 30° N, 23°
W, with developmental notes on certain myctophids.
Frog. Oceanogr. 7:3-58.
Baker. A C, M R Clarke, and M. J Harrls

1973. The N.I.O. combination net (RMT 1 + 81 and further
developments of rectangular midwatertrawls. J. Mar.
Biol. Assn. U.K. 53:167-184.

127

FISHERY BULLETIN: VOL. 86, NO, 1

Cohen. D M

1986. Merlucciidae. /a; M. M. Smith and P. C. Heemstra
(editors). Smiths' sea fishes, p. 324-326. Springer-
Verlag, N.Y.
Dunn. J R . and A C Matarese

1984. Gadidae: Development and relationships. In H. G.
Moser, W. J. Richards, D. M. Cohen, M. P. Fahay, A. W.
Kendall, Jr., and S. L. Richardson (editors). Ontogeny
and systematics of fishes, p. 283-299. Am. Soc. Ichth.
Herpet. Spec. Publ. 1.
Dunn. J R . and B M Vinter.

1984. Development of larvae of the safiron cod, Eleginus
gracilis, with comments on the identification of gadid
larvae in Pacific and Arctic waters contiguous to Canada
and Alaska. Can. J. Fish. Aquatic Sci. 41:304-318.
Fahay. M P . and D F Markle

1984. Gadiformies: Development and relationships. In
H. G. Mo.ser. W. J. Richards, D. M. Cohen, M. P. Fahay,
A. W. Kendall, Jr., and S. L. Richardson (editors), On-
togeny and systematics of fishes, p. 265-283. Am. Soc.
Ichth. Herpet. Spec. Publ. 1.
Gordon. D J . D F Markle, and J E. Olney.

1984. Ophidiiformes: Development and relationships.
In H. G. Moser, W. J. Richards, D. M. Cohen, M. P. Fahay,
A. W. Kendall, Jr., and S. L. Richardson (editors). On-
togeny and systematics of fishes, p. 308-319. Am. Soc.
Ichth. Herpet. Spec. Publ. 1.
Hay, D E

1981, Effects of capture and fixation on gut contents and
body size of Pacific herring larvae. Rapp. P. -v. Reun.
Cons. int. Explor. Mer 178:395-400.

Heron. A C

1982. A free fall plankton net with no mouth obstruc-
tions. Limnol. Oceanogr. 27:380-383.

Hollister, G.

1934. Clearing and dyeing fish for bone study. Zoologica
12:89-101.
Inada. T

1981. Studies on merlucciid fishes. Bull. Far Seas Fish.

Res. Lab. 18:1-172.
Marak.R R

1967, Eggs and early larval stages of the offshore hake,
Merluccius albidus. Trans Am. Fish. Soc. 96:227-228.

Markle, D F

1982. Identification of larval and juvenile Canadian At-
lantic gadoids with comments on the systematics of gadid
subfamilies. Can. J. Zool. 60:3420-3438.
Marshall. N B

1966. The relationships of the anacanthine fishes,
Macruronus, Lyconus and Steindachneria. Copeia
1966:275-280.
Marshall, N B , and D M Cohen.

1973. Order Anacanthini (Gadiformes). Fishes of the
western North Atlantic. Mem. Sears Found. Mar. Res.
l(6):479-495.
Marshall, N, B,, and T Iwamoto,

1973, Family Macrouridae, In D. M. Cohen (editor).
Fishes of the western North Atlantic, Part 6, p. 496-
665. Mem. Sears Found. Mar. Res. Yale Univ. 1.
Matarese, A C . S L Richardson, and J R Dunn

1981. Larval development of Pacific tomcod, Microgadus
proximus, in the northeast Pacific Ocean with compara-
tive notes on larvae of walleye pollock, Theragra chalo-
gramnia, and Pacific cod, Gadus macrocephalus (Gadi-
dae). Fish. Bull., U.S. 78:923-940.

Monod, T

1968. Le complexe urophore des poissons teleosteens.
Mem. Inst. Fond. Afr. Noire. 81:1-705.

Patchell, G J

1982. The New Zealand hoki fisheries 1972-82. Fish.
Res. Div. Occas. Publ. 38, 23 p.

Svetovidov, a N

1948. Fauna of the USSR, Fishes, Gadiformies. Zool.
Inst. Acad. USSR 9(4):l-222.
Wilson, M A

1981. Blue grenadier spawning grounds. Fintas 4:9-10.

1982. Spawning blue grenadier caught off Cape Sorell.
Fintas 4:13.

128

THE DISTRIBUTION, ABUNDANCE, AND TRANSPORT OF

LARVAL SCIAENIDS COLLECTED DURING WINTER AND EARLY SPRING

FROM THE CONTINENTAL SHELF WATERS OFF WEST LOUISIANA^

James H Cowan, Jr.^ and Richard F. Shaw^

ABSTRACT

The larvae of six species of Sciaenidae were collected in continental shelf waters off west Louisiana
on five midmonthly ichthyoplankton cruises from December 1981 to April 1982. Ranked in order of
abundance these species were sand seatrout, Cynoscion arenarius; Atlantic croaker, Micropogonias
undulatus; spot, Leiostomus xanthurus; black drum, Pogonias cromis; southern kingfish, Menticir-
rhus amencanus; and banded drum. Larimus fasciatus. Total larva density was highest in April, and
the high densities were associated with the coastal boundary layer, a horizontal density front caused
by an intrusion of fresher water onto the inner shelf that probably issued from the Atchafalaya River
east of the study area. Spawning by sand seatrout began in January, two months earlier than
previously reported, and first occurred offshore of midshelf but moved shoreward as the season
progressed. Analysis of length-frequency data suggest that spot probably began to spawn in Novem-
ber, one month earlier than once thought. Both sand seatrout and Atlantic croaker larvae were
captured at higher rates at night than during the daytime. Sand seatrout larvae appear to be
somewhat surface oriented while spot may undergo vertical migration. Interpretation of the sciaenid
data support a previously developed transport hypothesis involving gulf menhaden larvae and west-
northwest alongshore advection within and just outside of a horizontally stratified coastal boundary
layer.

Members of the perciform family Sciaenidae are
an important sport and commercial fishery re-
source along the United States coast of the Gulf of
Mexico and are perhaps the most prominent
group of northern Gulf inshore fishes. Sciaenids
exceed all other families in numbers of species
(18) and in numbers of individuals or biomass;
they are among the top four families with Mugili-
dae, Engraulidae, and Clupeidae (Gunter 1938,
1945; Moore et al. 1970; Franks et al. 1972; Hoese
and Moore 1977). Of the six species of sciaenids
captured during this study, only the banded
drum, Larimus fasciatus, is not commonly sought
by both sport and commercial fishermen.

Many of Louisiana's sciaenids spawn in coastal
or offshore waters. They have pelagic eggs and
young which are then transported into estuaries
(Johnson 1978 for review). The seasonal impor-
tance of Louisiana's estuaries as nursery grounds

'Louisiana State University Contribution No. LSU-CFI-86-
08.

^Center for Wetland Resources, Louisiana State University,
Baton Rouge, LA 70803-7503; present address: Center for Envi-
ronmental and Estuarine Studies, University of Maryland,
Chesapeake Biological Laboratory, Box 38, Solomons, MD
20688-0038.

^Center for Wetland Resources, Louisiana State University,
Baton Rouge, LA 70803-7503.

for postlarval and juvenile sciaenids is well docu-
mented (Cowan 1985 for review), and several
summary works are available which contain tax-
onomic and biological information on adult
sciaenids (Pearson 1929; Suttkus 1955; Guest and
Gunter 1958; Hoese and Moore 1977; Johnson
1978; Powles and Stender 1978; Barger and John-
son 1980; Barger and Williams 1980; Mercer
1984a, b). In contrast, there is little information
about sciaenid ichthyoplankton assemblages in
Gulf continental shelf waters, their offshore and
coastal distribution, or the oceanic current sys-
tems which influence their estuarine recruit-
ment.

This study provides such early life history in-
formation by determining larva distribution,
abundance, and length frequency; by document-
ing spawning location (depth and distance from
shore) of winter and early spring-spawned
sciaenids off west Louisiana; and by analyzing
larval sciaenid distribution with respect to known
water circulation patterns and a larval gulf men-
haden, Brevoortia patronus, transport hypothesis
in the shelf waters of the northwestern Gulf of
Mexico (Shaw et al. 1985b). Recruitment implica-
tions of the observed distribution, larva age struc-
ture, and transport of sciaenids in Louisiana
waters are also discussed.

Manuscript accepted September 1987.
FISHERY BULLETIN: VOL 86. NO. 1, 1988.

129

FISHERY BULLETIN: VOL 86. NO 1

METHODS AND MATERIALS

Detailed sampling methodology has been pre-
sented elsewhere (Shaw et al. 1985 a, b). Briefly,
larval sciaenids were collected off west Louisiana
on a sampling grid consisting of 37 stations on 5
transects (Fig. 1) during 5 midmonthly cruises
from December 1981 to April 1982. Ichthyoplank-
ton samples were analyzed from the 335 |xm mesh
net side of an opening and closing, 60 cm, paired
"bongo type" plankton sampler fitted with Gen-
eral Oceanics' flowmeters (model no. 2030). Most
plankton collections (125 of 187 total) consisted of
10-min stepped oblique tows from near bottom to
surface. Nets were set closed and opened just
prior to the stepped ascent. Each tow had five
steps with a retrieval rate between steps of 20
m/minute; towing speed was about 1 m/second
(2 knots). The object of the 10-min tow was to
filter approximately 100 m*^ of water. This process
increased the water volume filtered per unit
depth at the shallow stations relative to deeper
stations. This discrepancy is acceptable since the
alternative would be to compare 17-s shallow-
station oblique tows with 9-min deep-station tows
at a uniform retrieval rate (Houde 1977). At se-
lected stations (A-3, 6, 9; B-1; C-6; D-1; E-3, 6, 9;
Fig. 1), only 10-min simultaneous surface and
near-bottom horizontal tows (31 surface and 31
near-bottom) were made to determine if sciaenid
larvae were vertically stratified. Larva total
length (TL) was measured to the nearest 0.1 mm.
Larva densities are reported as standardized
catch rates at a station (density = larvae/100 m"^).

A four-way analysis of variance (ANOVA) was
performed on logio transformed [(no. larvae/100
m^) + 1] data to determine the spatial (vertical
and horizontal), temporal, and diel patterns of
species density and distribution. The four main
effects tested were month (January-April); sta-
tion depth group (d.g.) (d.g. 1 < 10 m, 10 m < d.g.
2 < 14 m, 14 m < d.g. 3 < 24 m and d.g. 4 > 24 m);
day-night (2000 hours < night < 0500 hours);
and horizontal tow type (surface vs. near-bottom).
Data from the December cruise were not included
as only the A transect was completed due to ad-
verse weather conditions.

Two methods of current estimates were utilized
(following Shaw et al. 1985b): 1) instantaneous
current profiles taken at each station and
2) continuous surface and near-bottom current

■♦Reference to trade names does not imply endorsement by the
National Marine Fisheries Services, NOAA.

meter measurements at two sites (H and S; Fig.
1). The instantaneous number of larvae trans-
ported on each transect was calculated by using
the equation D x U x M = number of larvae per
meter per second where D = larva density
darvae/m') from either oblique tows or from the
mean of the horizontal tows (i.e., average of sur-
face and near-bottom catch rates), U = depth-
averaged water velocity (m/s) determined from
instantaneous current meter profiles at each sta-
tion, and M = water depth (m) at each station.

Distribution diagrams and length-frequency
histograms were generated for each cruise for the
three most abundant sciaenid species. Inspection
of these data along with current measurements
allowed a comparison with the previously men-
tioned transport hypothesis.

RESULTS AND DISCUSSION
Total Sciaenids

A total of 5,225 larval sciaenids accounted for
9.1% of the fish larvae collected. In December
through February, samples were dominated by
Atlantic croaker, Micropogonias undulatus, and
spot, Leiostomus xanthuriis. In March and April
samples contained mostly sand seatrout,
Cynoscion arenarius. In all, six species of sciaenid
larvae were collected: sand seatrout (N = 4,100);
Atlantic croaker {N = 567); spot (A^ = 264); black
drum, Pogonias cromis {N = 68); southern king-
fish, Menticirrhus americanus (N = 53); and
banded drum (A'^ = 13). Additional Menticirrhus,
not identifiable to species, accounted for 160 more
specimens (Table 1). A more detailed examina-
tion of the data on the three most abundant
sciaenid species follows.

Sand seatrout,
Cynoscion arenarius

A total of 4,100 sand seatrout larvae was col-
lected making it the most abundant sciaenid
taken during the study. Larval sand seatrout den-
sities were highest in April (Table 1) with a mean
of 46.1 larvae/100 m"^; mean density in February
and March was 0.3 and 2.9/100 m'^, respectively,
and 1 larva was collected in January. Larvae
were distributed mostly over the midshelf in
February but highest concentrations were later
found inshore and towards the east (Fig. 1). Over
the course of study, larvae were found in temper-
atures and salinities ranging from 14° to 21°C and

130

COWAN AND SHAW: LARVAL SCIAENIDS COLLECTED OFF WEST LOUISIANA

Figure l. — Density distribution of sand seatrout, Cynoscion arenarius, larvae by month,
February-April 1982. Densities are as follows: o = 0: • >0-10; • >10-50; • >50-99; •
>99-250; 0 250 100 m3 of water filtered from all plankton tow types. Also shown is the
station sampling grid with moored current meter sites (H and S) and selected iso-
baths. SP = Sabine Pass, CR = Calcasieu River, and MR = Mermentau River.

131

KISHKRY BULLETIN: VOL, 86, NO. 1

Table 1. — Total number, months of occurrence and monthly density of
sciaenid larvae collected in west Louisiana shelf waters from December 1981
to April 1982.

Months of

occurrence and

Total
number

density (No. /1 00

m3)

Taxa

Dec.

Jan.

Feb.

IVIar.

Apr.

Cynoscion arenarius

4,100

—

1 larva

0.3

2.9

46.1

Micropogonias undulatus

567

2.2

Leiostomus xanthurus

264

7.8

1.3

0.9

0.1

—

Menticirrhus sp.

160

—

Pogonias cromis

—

1 larva

0.5

0.2

Menticirrhus americanus

—

0.6

Larimus fasciatus

—

0.2

from 15 to 36 ppt (Table 2) and at station depths
ranging from 5 to 70 m, but most were collected
inside the 18 m isobath. Larval sand seatrout
density increased in April with many stations ex-
hibiting densities in excess of 250 larvae/100 m'^;
the larvae appeared to be associated with a
freshet of water on the shelf, probably issuing
from the Atchafalaya River east of the study area
(Shaw et al. 1985a). The presence of riverine
runoff on the shelf, which was most evident in

March and April 1982, caused the development of
an oceanic salinity front referred to as the coastal
boundary layer, 10-35 km from shore (Wiseman
et al. 1987).

Observed spawning seasonality and location
for sand seatrout are in part consistent with pre-
viously published information (Shlossman and
Chittenden 1981 for review). The presence of a 4
mm TL larva in January indicates some spawn-
ing had taken place at least 2 months earlier than

Table 2. — l^onthly data summaries at time of capture for three sciaenid larvae
(Cynoscion arenarius, Micropogonias undulatus. and Leiostomus xanthurus) col-
lected in west Louisiana shelf waters from December 1981 to April 1982.

Total
Species'month number

Length range
(mm TL)

Temperature
range ( C)

Salinity
range

(ppt)

Depth

range

(m)

Cynoscion arenarius
January 1

February 20

2.5-4.5
(mode = 2-3)

14-20

34-36

15-70

March 203

1.5-10.5
(mode = 2-3)

14-18

25-36

5-40

April 3,876

1.5-20.5
(mode = 2-3)

20-21

15-36

5-70

Micropogonias undulatus
December 28

2.5-10.5
(mode = 3-4)

12-20

30-36

10-65

January 158

2.5-10.5
(mode = 4-5)

10-18

30-36

5-70

February 221

2.5-17.5
(mode = 14-15)

11-17

27-36

5-40

March 144

2.5-19.5
(mode =14-15)

14-20

25-36

5-115

April 16

11.5-18.5
(mode= 17-18)

20.5

Leiostomus xanthurus
December 110

2.5-7.5
(mode = 3-4)

14-18

30-36

16-65

January 89

2.5-13.5
(mode = 4-5)

10-18

30-36

5-40

February 62

3.5-15.5
(mode= 12-13)

10-17

28-36

5-40

March 3

3.5-16.5

14-17.5

26-36

11-40

132

COWAN AND SHAW: LARVAL SCIAENIDS COLLECTED OFF WEST LOUISIANA

previously reported. Monthly length-frequency
data for sand seatrout show that larvae as large
as 11 mm TL were first present in March samples
(Fig. 2A). Based on the estimated growth rate
determined for sand seatrout (Cowan 1985; Shaw
et al., in press), an 11 mm larva could be as old as
65 days; this further supports January spawning.
Sand seatrout are reported to spawn from March
to August, during two discrete periods — one in
March-May, the other August-September, with
little spawning between the two peak periods
(Hoese 1965; Daniels 1977; Shlossman and Chit-
tenden 1981).

An examination of distribution and length-
frequency data (Figs. 1, 2A) suggests that most
spawning initially took place in midshelf to off-
shore waters at depths ranging from 15 to 80 m or
to about 175 km from shore. As the season pro-
gressed into March and April, spawning location,
as determined by the presence of larvae <3.0 mm
TL, was more inshore (5-18 m) with few small
larvae occurring at depths >25 m.

Other than the indication that spawning may
move from offshore to inshore waters as the sea-
son progresses, this spatial information agrees
with the limited life history data available on
sand seatrout. Most spawning has been shown to
occur in the shallow waters of the Gulf of Mexico,
primarily between 7 and 15 m in depth (Gunter
1945; Moffet et al. 1979; Shlossman and Chitten-
den 1981). Running ripe C. arenarius have been
captured in deepter waters (70-90 m) in February
and March, but no spawning was indicated
(Franks et al. 1972; Perry 1979).

In a four-way ANOVA employed to determine
patterns of larval sand seatrout density and dis-
tribution, month was a highly significant main
effect (P < 0.01; Table 3) reflecting spawning sea-
sonality and the magnitude of the density in-
crease in April. The test for interaction between
month and day-night was employed to determine
if daytime gear avoidance was evident as size and
mobility of larva increased. Most sand seatrout
collected, however, were small and no clear
monthly modal increase in larva size was evident
(Fig. 2A). The significant interaction (P <0.01)
was probably due to an increased catch in oblique
tows at night as the season progressed (0.0 in
January, 64.9/100 m"^ in April). The significant
interaction between month and depth group and
the highly significant depth group main effect
(P < 0.01) represents the shift in larva concentra-
tion from midshelf early in the study, to a more
coastal distribution in March and April (Fig. 1).

Mean larva density was greatest in depth group 1
(23.7/100 m'^) followed by depth groups 2, 3, and 4
(12.7, 9.2, and 0.3/100 m^, respectively). The third
main effect, day vs. night tows, was highly sig-
nificant (P < 0.01); many more sand seatrout
larvae were collected at night (averge catch rates
in all night (74) tows combined = 21.6/100 m"^
vs. day (113) tows = 7.4/100 m^). Highest night-
time catches occurred in oblique (49) tows (26.9/
100 m'^) while the day-oblique-catch rate aver-
aged 7.6/100 m'^ in 76 tows. Overall, average
catch rate was highest in oblique (125) tows
(14.6 larvae/100 m-^), followed by surface (31)
tows (9.2/100 m^), and then bottom (1.9/100 m^;
31 tows). Intrepretation of the data suggests that

Table 3. — Summary data from four-way analysis of variance done
on logio transformed [(larvae 100 m3) + 1] data from ichthiyoplank-
ton samples collected from January to April 1982. Tfie results are
for A. Cynoscion arenarius. B, Micropogonias undulatus, and C.
Leiostomus xanthurus. The four main effects tested were months
(Jan. -Apr.), station depth group (d.g. 1 < 10 m, 10 m < d.g. 2 < 14
m, 14 m < d.g. 3 < 24 m and d.g. > 24 m), day - night (2000
hours s night < 0500 hours) and horizontal tow type (surface vs.
near-bottom).

Source

r2 = 0.75

A. Dependent variable:

Log 10 [(Cynoscion arenarius 1^00 m3) + 1]

Model 21 O.OOOr*

r^^onth 3 0.0001"

Depth group 3 0.0001"

Day-night 1 0.0026"

Horizontal tow type 1 0.2574 (NS)

IVIonth vs. Day-night 3 0.0001"

l\/lonth vs. Depth group 9 0.0001"

Day-night vs. Tow type 1 0.4180 (NS)

Error 177

Corrected Total 1 98

B. Dependent variable:

Logio [(Micropogonias undulatusHOO m3) + 1]

Model 21 0.0001" r2 = o.63

Month 3 0.0045"

Depth group 3 0.3551 (NS)

Day-night 1 0.0001"

Horizontal tow type 1 0.4448 (NS)

Month vs. Day-night 3 0.2168 (NS)

Month vs. Depth group 9 0.0001"

Day-night vs. Tow type 1 0.1288 (NS)

Error 177

Corrected Total 1 98

C. Dependent variable:

Logio [(Leiostomus xanthu^us/^00 m3) + 1]

Model 21 0.0001" /-s = 0.51

Month 3 0.0001"

Depth group 3 0.0033"

Day-night 1 0.1875 (NS)

Horizontal tow type 1 0.3216 (NS)

Month vs. Day-night 3 0.2138 (NS)

Month vs. Depth group 9 0.0001"

Day-night vs. Tow type 1 0.0324*

Error 177

Corrected Total 198

* = Statistically significant (P < 0.05).
" = Highly significant (P < 0.01).
(NS) = Not significant.

133

FISHERY BULLETIN: VOL. 86, NO. 1

00
CM
II

O
ill

!• I ■ I ■ I ' I ' I

8Q Q O p O o
ift ^ r^ 04 ^

00
U1

z'
<

-J

[

r°

i
E

CM
CM

S 2

I ■ I ■ I ■

S § CO

8 2

8
II

z
m

[
[

— I—

-T —
O

i:
[

<
z

Cvj

CM
II

z
if

<
S

r ■ I ■ 1

So Q O p O O
u^ ^ r-1 "N --

w'::.

^Kis;.

( :....

r:::„

r— ' — 1 — •— I — •"

z
of

<
Z

' E

II
z

2 I

UJ o

-I w

-I 9

I-
O

nVlOl dO iN30a3d

•o

3
(O

c
o
en
o

QJ "O

fe iJ

S wi

Z
O

.O CD
U 0)

5 "O

-J

O^S

-1

< -c

Cvi S

CO "

^ o
^

^ E

Q- D

< c

j|_

£ ,

■o

134

COWAN AND SHAW: LARVAL SCIAENIDS COLLECTED OPT WEST LOUISIANA

sand seatrout larvae v^'ere somewhat surface
oriented.

Atlantic croaker,
Micropogon ias undiilatus

The second most abundant sciaenid taken was
Atlantic croaker (A'^ = 567). Larval Atlantic
croaker density was greatest in February at 3.2/
100 m'^, but density was relatively constant from
December through March (Table 1). Mean densi-
ties for December, January, March, and April
were 2.0, 2.2, 2.2, and 0.2/100 m^, respectively.
Atlantic croaker was the only sciaenid collected
in all months of the study. Their overall distribu-
tion (all sizes combined) was generally uniform
over most of the shelf (Fig. 3) except in March and
April when they were more often found inshore.
Recently spawned larvae (<3.0 mm TL) were also
collected over much of the shelf at station depths
ranging from 15 to 115 m or from about 20 to 200
km from shore. However, most small larvae were
collected near midshelf about 65-125 km from
shore. In December and January the majority of
the larvae were small. By April, no recently
spawned individuals were collected (Fig. 2B). For
the study overall, larvae were found in salinities
and temperatures ranging from 22 to 36 ppt and
from 10 to 20.5°C (Table 2).

Spawning by Atlantic croaker in Gulf of Mexico
waters is reported to occur from September to
March, with a distinct peak in October (Hoese
1965; Sabins and Truesdale 1974; White and
Chittenden 1977; Benson 1982) and to occur pri-
marily offshore over a wide area (Pearson 1929;
Hildebrand and Cable 1930; Wallace 1940; Haven
1957; Bearden 1964; Hoese 1965; Nelson 1967).
Atlantic croaker larvae, however, have been
taken on the outer continental shelf off Texas
from September to May (Finucane et al. 1979).

As with sand seatrout, a four-way ANOVA was
used to determine patterns in larva density and
distribution (Table 3). Larval Atlantic croaker
density by month was a highly significant main
effect (P <0.01). Densities at the end of their
spawning period were low, increased only slightly
in February, and then dropped off by April (Table
1). The interaction between month and day-night
was not significant. The highly significant inter-
action between month and depth group was not
surprising. Larvae were in more offshore waters
early in the study while later becoming more
abundant inshore (Fig. 3). However, as a main
effect, depth gi'oup was not significant. Larval

Atlantic croaker mean densities for depth groups
1 through 4 were 3.9, 0.8, 0.4, and 0.7/100 m^,
respectively. Day-night, as a main effect, was
highly significant (P < 0.01). Larval Atlantic
croaker density was over 5 times higher at night
(all tow types combined) than during the day (3.7
vs. 0.7/100 m'^). However, the interaction between
day-night and horizontal tow type was not signif-
icant. The fourth main effect tested, horizontal
tow type, was not significant. Average catch rates
at the surface and near-bottom were similar (1.0
and 1.8/100 m'\ respectively).

Spot, Leiostomiis xcmthurus

The third most abundant sciaenid collected was
spot (A^ = 264). Density of spot larvae was highest
in December at 7.8 larvae/100 m'^ (Table 1). How-
ever, the high December value must be viewed
with a consideration of the abbreviated cruise
track for that month and the resultant reduction
in spatial coverage. Mean densities for January to
March were 1.3, 0.9, and 0.1/100 m"^, respectively.
No spot larvae were collected in April. In general,
larva density was low and their distribution was
uniform over the shelf out to the 40 m isobath,
about 130 km offshore (Fig. 4). Spot were col-
lected in temperatures and salinities ranging
from 10° to 18°C and from 26 to 36 ppt (Table 2),
and at stations with depths ranging from 5 to 65
m.

Larvae >7 mm TL in our mid-December (Fig.
2C) collections and small larvae (<3.0 mm TL) in
all but the last cruise indicate that spawning
probably began by at least November and contin-
ued through March. Spawning occurred from
near midshelf (about 65 km) out to 175 km from
the coast. Data presented here partly concur with
previously published information on spot spawn-
ing periodicity. In the northern Gulf, spawning
reportedly occurs from late December to March,
peaking in January, and takes place well offshore
in moderately deep water (Pearson 1929; Kilby
1955; Townsend 1956; Dawson 1958; Springer
and Woodburn 1960; Pacheco 1962; Nelson 1967;
Joseph 1972; Music 1974; Sabins and Truesdale
1974).

A four-way ANOVA indicated that month, as a
main effect for spot larvae, was highly significant
(P < 0.01 ), which probably reflects the decreasing
catch rates seen from January to March (Table 3).
The interaction between month and depth group
was also highly significant (P < 0.01) as was
depth group as a main effect. Larval spot

135

FISHERY BULLETIN VOL 86. N(J 1

DEC

-yT ,4^-

^?^IV^

° • :

--^.^^^^

0 o

o o o

o
o

MAR

APR

Figure 3— Density distribution of Atlantic croaker, Micropogonias undulatus, larvae by month, December 1981— April 1982. Densities are
as follows: o = 0; • >0-10; • > 10-50; • >50;-99; • >99-250; # >250/100 m3 of water filtered from all plankton tow types.

136

COWAN AND SHAW: LARVAL SCIAENIDS COLLECTED OFF WEST LOUISIANA

i.Tr'if"

^u^

*^^\

-^ 0 o ■

° O O .

O 0 o

o o
o

FEB

1 1

o
o

■ o o 0
o o

o o

O 0

o o o

o
o

MAR

Figure 4 — Density distribution of spot, Leiostomus xanthurus. larvae by month, December
1981-Marcti 1982. Densities are as follows: o = 0; ■ >0-10; • > 10-50; • >50-99; • >99-
250; 9 >250/100 m3 of water filtered from all plankton tow types.

137

FISHERY BULLETIN: VOL. 86, NO. 1

densities were higher offshore in the early part of
the study and then greater inshore during Febru-
ary and March. Depth group 4 had the highest
mean density (1.3/100 m-'^) followed by depth
groups 1 and 3 (0.4/100 m'^ each) and depth
group 2 (0.1/100 m^). Day-night comparisons
proved nonsignificant as a main effect for spot
larvae. The average catch rates for all day and
night tows were identical (0.5/100 m^). However,
the interaction between day-night and horizontal
tow type was statistically significant (P < 0.05).
In this case, vertical migration and stratification
may be indicated. Average catch rate of spot lar-
vae during the day at the surface was 0.1/100 m^,
while near the bottom it averaged 1.6. Con-
versely, nighttime average catch rate at the sur-
face was 1.0/100 m-^ while the near-bottom rate
averaged 0.04. These are very low densities but
the vertical differences are an order of magnitude
and their reversing pattern suggests that spot
larvae were stratified and undergoing diel verti-
cal migration. Daytime bottom and nighttime
surface average catch rates were higher than for
oblique (O) tows (day, 0 = 0.45/100 m^; night,
O = 0.49/100 m^). Average catch rates for surface
and near-bottom tows, regardless of time of day,
were 0.5 and 0.9/100 m\ respectively.

As previously mentioned, other larval sciaenid
species (i.e., black drum, banded drum, southern
kingfish) were collected during these cruises.
However, relatively few individuals were cap-
tured (Table 1), making information on their dis-
tribution inconclusive.

TRANSPORT ANALYSIS

Alongshore advection within and just outside
the coastal boundary layer in the northwestern
Gulf of Mexico has been hypothesized as the
major mechanism transporting gulf menhaden
larvae to the estuaries in western Louisiana,
rather than across-shelf transport from directly
offshore. In contrast, such direct across-shelf
transport has been demonstrated for sciaenids
and other species along the U.S. mid-Atlantic
coast (Nelson et al. 1976; Norcross and Austin
1981; Miller et al. 1984). The data collected for
sciaenid larvae (all species combined) were exam-
ined in light of this Gulf hypothesis. Larval
sciaenid densities were less than those for gulf
menhaden but similarities in distribution were
evident. Both larval sciaenid and gulf menhaden
densities were highest at midshelf early in the
study. By March and April the highest densities

were found towards the east and inshore and were
associated with a horizontal density front (coastal
boundary layer) caused by an intrusion of fresher
water onto the shelf.

The along-transect length-frequency patterns
exhibited by larval sciaenids and gulf menhaden
were also similar. No apparent increase in size
was seen until gulf menhaden larvae were on the
inner shelf or sciaenids were on the mid- to inner
shelf. The expected pattern of a gradual increase
in larva size from offshore to inshore, which
would result if there were significant across-shelf
(south to north) transport, was not evident in ei-
ther data set. Off the North Carolina coast, War-
len (1981) and Miller et at. (1984) showed that
ages and lengths of both spot and Atlantic
croaker larvae increased systematically toward
shore in an area where winter water currents fa-
vored across-shelf (west to east) transport.

During the winter of 1981-82, moored current
meter data from sites H and S (Fig. 1) indicated
that flow was directed primarily alongshore in
the west-northwest direction. Several researchers
have reviewed the circulation in the northwest-
ern Gulf (Nowlin 1971; Kelly et al. 1982; Crout
1983). It was not until Cochrane and Kelly (1986)
developed their comprehensive circulation model
for the Louisiana-Texas continental shelf, how-
ever, that the ocean current patterns, which led to
the hypothesized larva transport model, were
fully documented. Flow in nearshore coastal
waters is westward all year except in summer
when it usually reverses, while farther offshore
flow is eastward all year (Cochrane and Kelly
1986).

To quantify transport, larval sciaenid densities
were combined with the vertically averaged, in-
stantaneous current measurements. The resul-
tant curves present the number of larvae trans-
ported per unit time at each station (Fig. 5). Early
in the winter, highest sciaenid larva transport
(mostly Atlantic croaker and spot) was located
midshelf. Later (March and April), transport val-
ues were higher inshore and reflected the in-
crease in larval sciaenid density (primarily sand
seatrout). Overall, larva transport was primarily
westward and ranged from about 0.05 to 4.0 lar-
vae/meter per second.

Although the oceanographic data collected
were insufficient to precisely quantify onshore
transport rates, an estimate was obtained by
using the mean current vectors from the near sur-
face meter at site H (Fig. 1) from 24 January to 12
May 1982 (14.33 cm/second alongshore westward

138

COWAN AND SHAW LARVAL SCIAENIDS COLLECTED OFF WEST LOUISIANA

■«

o I—-—

MAR

(/)

r5Csi»-0'-fMr)'*

1-' I

»- o »- oj r)
09999

t-8,uj avAbvn

a.
<

3 8 8 3 8 3
009990

-I — ' — »■ o

3
CO
CO
<D

k_
k.

3
O

5 uj

T3 -^

(1)

(/>

CO O)

<D 01

8 g

CO c

(0 0)

~ a:
^^

O o

E c
o I

I"
o

0)
Q

•o
LU a>

- U5

O S
< 2

o
a.

2 <o

2 a)
<-> 2
<" to

5 o

JO (D

t: <2
o ^
Q- 9

<o £

^ CO

1 1

to a>

y CD

139

FISHERY BULLKTIN: VOL. 86, NO. 1

and 1.75 cm/second shoreward). Based on that av-
erage shoreward advection rate we calculated
that larvae could be passively transported 98 km
in the onshore direction in 65 days. Examination
of length frequency and age at capture data
(Cowan, in press) suggest that larval Atlantic
croaker arrive in nearshore coastal waters, on the
average, 60-90 days after hatching. Most small,
newly hatched Atlantic croaker larvae were col-
lected approximately 100 km offshore. Although
the onshore component of advective transport is
small in comparison with the average alongshore
component, the estimate of shoreward transport
rate is reasonable when age of larvae is consid-
ered.

CONCLUSIONS-
RECRUITMENT IMPLICATIONS

Across-shelf transport appears to be an order of
magnitude smaller than alongshore advective
transport in the northwestern Gulf shelf waters
during winter and spring. Sciaenid larvae col-
lected offshore in the study area, at midshelf and
beyond, would probably be lost to the estuaries in
western Louisiana. Those offshore larvae would
be transported towards north Texas estuaries, or
back to the east if they were far enough offshore,
since there is evidence for an easterly counter
current (Kelly et al. 1982; Cochrane and Kelly
1986).

Sand seatrout are common in west Louisiana
estuaries (Herke et al. 1984) and were the most
abundant sciaenid larvae collected in this study.
They spawn, in general, more inshore (Fig. 1)
than Atlantic croaker or spot. Conceivably, many
of the sand seatrout collected in the study area
inside the coastal boundary layer on the inner
shelf would have recruited to Louisiana estuaries.

Still, large numbers of postlarval sciaenids,
other than sand seatrout, enter the estuaries in
west Louisiana each year. Atlantic croaker and
spot were the 3rd and 21st most abundant fish, of
117 species collected, in the Calcasieu River
Basin, the largest estuary in west Louisiana
(Herke et al. 1984). However, the distribution and
transport analyses indicate that most spot and
Atlantic croaker larvae directly offshore at least
would not have recruited to the Calcasieu Basin.
Interpretation of these data suggests that the
source of the sciaenid postlarvae shown to season-
ally recruit to the Calcasieu estuaries must be
east of the study area. Interpretation of data sum-
marizing several years of northern Gulf shrimp-

trawl collections suggests that, during the spawn-
ing season, a sufficient concentration of adults
exists to the east of our study area (Darnell et al.
1983).

In the fall and winter, high concentrations of
Atlantic croaker, and to a lesser extent spot, have
been found between the 20 and 40 m depth con-
tours (65 and 125 km offshore) in an area east of
the sampling grid. The area and timing of high
concentration coincides with the reported spawn-
ing location and period for both Atlantic croaker
and spot. If indeed this concentration represents a
spawning distribution, it would help explain why
so few Atlantic croaker and spot larvae, relative
to the number of juveniles seen in estuaries, were
collected in this and previous Gulf of Mexico
ichthyoplankton studies. Unless collections were
made in or near the spawning area, single-station
densities would be low as eggs and larvae were
dispersed. Furthermore, this study demonstrates
the need for understanding both biological (verti-
cal distribution, age and growth, behavior, etc. of
larvae) and physical (ocean currents, estuarine-
shelf exchange, etc.) processes which may influ-
ence estuarine recruitment.

ACKNOWLEDGMENTS

We would like to thank Wm. Wiseman,
L. Rouse, Jr., and S. Dinnel for their discussion
and assistance in interpretation of the physical
oceanographic data. We gratefully acknowledge
E. Turner, J. Geaghan, M. Fitzimmons,
B. Thompson, and W. Herke for critically review-
ing this manuscript.

Funding was provided by a Louisiana Depart-
ment of Wildlife and Fisheries, U.S. Department
of Energy and LSU Center for Wetland Resources
cooperative agreement No. DE-FC96-81P010313.
Additional support was given by the Department
of Marine Sciences, Louisiana Sea Grant College
Program and the Coastal Fisheries Institute.

LITERATURE CITATIONS

Barger, L. E., and a. G. Johnson

1980. An evolution of marks on hardparts for age
determination of Atlantic croaker, spot, sand seatrout,
and silver seatrout. U.S. Dep. Commer., NOAA Tech.
Memo. NMFS-SEFC-22, 5 p.

Barger, L E , and M. L. Williams.

1980. A summarization and discussion of age and growth
of spot, Leiostomus xanthurus Lacepede, sand seatrout,
Cynoscion arenarius Ginsburg, and silver seatrout,
Cynoscion nothus (Holbrook), based on a literature
review. U.S. Dep. Commer., NOAA Tech. Memo.

140

COWAN AND SHAW: LARVAL SCIAENIDS COLLECTED OFF WEST LOUISIANA

NMFS-SEFC-14, 15 p.
Bearden, C. M.

1964. Distribution and abundance of Atlantic croaker,

Micropogon undiilatus, in South Carolina. Bears Bluff

Lab. Contrib. 38, 27 p.
Benson, N G (editor).

1982. Life history requirements of selected finfish and
shellfish in Mississippi Sound and adjacent areas. Biol.
Serv. Program, U.S. Dep. Inter., Fish Wildl. Serv.
FWS OBS-78/12, 314 p.

Cochrane, J D., and F. J Kelly, Jr

1986. Low-frequency circulation on the Texas-Louisiana
continental shelf. J. Geophys. Res. 91(C9): 10,645-
10.659.

CoWAN, J H. jR

1985. The distribution, transport and age structure of
drums (family Sciaenidae) spawned in the winter and
early spring in the continental shelf waters off west
Louisiana. Ph.D. Thesis, Louisiana State Univ., Baton
Rouge, 182 p.
In press. Age and growth of Atlantic croaker, Micro-
pogonias undulatus. larvae collected in the coastal
waters of the northern Gulf of Mexico as determined by
increments in saccular otoliths. Bull. Mar. Sci. 42(3).

Crout, R L

1983, Wind-driven, near-bottom currents over the west
Louisiana inner shelf. Ph.D. Thesis, Louisiana State
Univ., Baton Rouge, 117 p,

Daniels. K

1977. Description, comparison, and distribution of larvae
oi Cynoscion nebulosus and Cynoscion arenarius from the
northern Gulf of Mexico. M.S. Thesis, Louisiana State
Univ., Baton Rouge, 48 p.
Darnell. R N , R E Defenbaugh. and D Moore

1983. Northwestern Gulf shelf bio-atlas; a study of the
distribution of demersal fishes and panaeid shrimp on
soft bottoms on the continental shelf from the Rio Grande
to the Mississippi River Delta. Open File Rep. No.
82-04, Miner. Manage. Serv.. Gulf Mex. OCS Reg. Off.,
Metairie, LA, 438 p.
Dawson, C E

1958, A study of the biology and life history of the spot,
Leiostomus xanthurus Lacepede, with special references
to South Carolina. Bears Bluff Lab. Contrib. 28, 48 p.
Finucane. J H , L A Colllns. L E Barger. and J D
McEachran

1979. Environmental studies of the south Texas outer

continental shelf. Ichthyoplankton'mackeral eggs and

larvae. NOAA Final Rep. to Bur. Land Manage. Wash.,

D.C., 504 p.

Franks. J S.J Y Christmas, W L Siler, R Combs, R Waller,

and C Burns

1972. A study of nektonic and benthic faunas as related to
some physical, chemical and geological factors. Gulf
Res. Rep. 4:1-147.
Guest. W. C, and G. Gunter.

1958. The seatrout or weak fishes (genus Cynoscion) of
the Gulf of Mexico. Gulf States Mar. Fish. Comm. Tech.
Summ. No. 1, 40 p.
Gunter. G

1938. Seasonal variations in abundance of certain
estuarine and marine fishes in Louisiana, with
particular reference to life histories. Ecol. Monogr.
8:313-346.
1945. Studies on marine fishes of Texas. Publ. Inst. Mar,
Sci., Univ. Tex. 1:1-190.

Haven, D S

1957. Distribution, growth, and availability of juvenile
croaker, Micropogon undulatus, in Virginia. Ecology
38:88-97.
Herke. W H , B D Rogers, and J. A Grimes.

1984. The study of the seasonal presence, relative
abundance, movements, and use of habitat types by
estuarine-dependent fishes and economically important
decapod crustaceans on the Sabine National Wildlife
Refuge. La. Coop. Fish. Res. Unit Final Rep., two vols.,
603 p.
HiLDEBRAND. S F . AND L E CABLE

1930. Development and life history of fourteen teleostean
fishes at Beaufort, N. C. Bull. U.S. Bur. Fish. 46:383-
488.
HOESE. H D

1965. Spawning of marine fishes in the Port Aransas,
Texas area as determined by the distribution of young
and larvae. Ph.D. Thesis, Univ. Texas, Austin, 144 p,
HOESE, H. D.. AND R. H MOORE

1977. Fishes of the Gulf of Mexico, Texas, Louisiana and
adjacent waters. Texas A&M Univ. Press, College
Station, 327 p.
HOUDE.E D

1977. Abundance and potential yield of the round
herring, Etrumeus teres, and aspects of its early life
history in the eastern Gulf of Mexico. Fish. Bull., U.S.
75:61-89.

Johnson, G. D.

1978. Development of fishes of the Mid-Atlantic Bight.
An atlas of egg, larval and juvenile stages. Vol. IV,
Carangidae through Ephippidae. Biol. Serv. Program,
U.S. Dep. Inter., Fish Wildl. Serv. FWS/OBS-78/12,
314 p.

Joseph, E B

1972. The status of the sciaenid stocks of the middle
Atlantic coast. Chesapeake Sci. 13:87-100.
Kelly. F J . Jr . J E Schmitz, R E. Randall, and J. D.
Cochrane

1982. Physical oceanography. In R. W. Hann, Jr. and
R. E. Randall (editors). Evaluation of brine disposal from
the Bryan Mound site of the Strategic Petroleum Reserve
Program, p. 1-144. Final report of the 18 month
post-disposal studies. Texas A&M Univ. Res. Found.,
College Station.
KiLBY, J

1955. The fishes of two Gulf coastal marsh areas of
Florida. Tulane Stud. Zool. 2:175-247.
Mercer, L P

1984a. A biological and fisheries profile of spotted
seatrout, Cynoscion nebulosus. N.C. Dep. Nat. Resour.
Community Dev., Div. Mar. Fish., Spec. Sci. Rep. 40,
87 p.
1984b. A biological and fisheries profile of red drum,
Sciaenops ocellatus. N.C. Dep. Nat. Resour.
Community Dev., Div. Mar. Fish., Spec. Sci. Rep. 41,
89 p.
Miller, J M , J P Reed, and L J. Pietrafesa.

1984. Patterns, mechanisms and approaches to the study
of migrations of estuarine-dependent fish larvae and
juveniles. In J. D. McCleave, G. P. Arnold, J. J. Dodson,
and W. H. Neill (editors). Mechanisms of migrations in
fishes, p. 209-225. Plenum Publ. Corp., N.Y.
Moffet, a W., L W. McEachron, and J. G. Key.

1979. Observations on the biology of sand seatrout
(Cynoscion arenarius) in Galveston and Trinity Bays,

141

FISHERY BULLETIN VOL 86, NO. 1

Texas. Contrib. Mar. Sci. 22:163-172.
Moore. D . H A Brusher. and L Trent.

1970. Relative abundance, seasonal distribution, and
species composition of demersal fishes ofT Louisiana and
Texas, 1962-1964. Contrib. Mar. Sci. 15:45-70.

Music. J L.Jr

1974. Observations of the spot (Leiostomus xanthurus ) in

Georgia's estuarine and close inshore ocean waters. Ga.

Dep. Nat. Resour., Coastal Fish. OfT. Contrib. Ser. 28,

29 p.
Nelson. W R

1967. Studies on the croaker, Micropogon undulatus

Linneaus, and the spot, Leiostomus xanthurus Lacepede

in Mobile Bay, Alabama. M.S. Thesis, Univ. Alabama,

Tuscaloosa, 85 p.
Nelson, W R , N C Ingham, and W E Schaaf

1976. Larval transport and year-class strength of Atlantic

menhaden, Brevoortia tyrannus. Fish. Bull., U.S.

75:23-41.
NoRCRoss. B L AND H M Austin

1981. Climate scale environmental factors affecting year

class fluctuations of Chesapeake Bay croaker

Micropogonias undulatus. Va. Inst. Mar. Sci. Spec. Sci.

Rep. No. 110, 78 p.
NOWLIN, W D, Jr.

1971. Water masses and general circulation of the Gulf of
Mexico. Oceanol. Int. 6(2):28-33.

PACHECO, a. L

1962. Age and growth of spot in lower Chesapeake Bay,
with notes on distribution and abundance of juveniles
in the York River system. Chesapeake Sci. 3:18-
28.
Pearson, J C

1929. Natural history and conservation of redfish and
other commercial sciaenids on the Texas coast. Bull.
U.S. Bur. Fish. 44:129-214.
Perry, A

1979. Fish of Timbalier Bay and offshore Louisiana
environments collected by trawling. Rice Univ. Stud.
65:537-545.

POWLES, H , AND B W STENDER

1978. Taxonomic data on the early life history stages of
sciaenids of the South Atlantic Bight of the United
States. S.C. Mar. Resour. Comm. Tech. Rep. No. 31,
64 p.
Sabins, D S . AND F. M Truesdale.

1974. Diel and seasonal occurrence of immature fishes in
a Louisiana tidal pass. Proc. Annu. Conf. Southeast.
Assoc. Game Fish Comm. 28:161-171.

Shaw, R F , J H Cowan, Jr , and T L Tillman

1985a. Distribution and abundance of Brevoortia

patronus (gulf menhaden) eggs and larvae in the

continental shelf waters of western Louisiana. Bull.

Mar. Sci. 36:96-103.
Shaw, R F , B D Rogers, J H Cowan, Jr , and W H Herke
In press. Ocean-estuary coupling of ichthyoplankton and

nekton in the northern Gulf of Mexico. Am. Fish. See.

Symp. 3:77-89.
Shaw, R F , W J Wiseman, Jr , R E Turner, L J Rouse, and

R E CONDREY

1985b. Transport of larval gulf menhaden Brevoortia
patronus in continental shelf waters of western
Louisiana: a hypothesis. Trans. Am. Fish. Soc.
114:452-460.
Shlossman, P a , AND M E Chittenden, Jr.

1981. Reproduction, movements and population dynamics
of the sand seatrout, Cynoscion arenarius. Fish. Bull.,
U.S. 79:649-669,
Springer, V G , and K D Woodburn

1960. An ecological study of the fishes of the Tampa Bay
area. Fla. Board Conserv. Mar. Lab. Prof. Pap. Ser.
No. 1, 104 p.
Suttkus, R. D.

1955. Seasonal movements and growth of the Atlantic
croaker (Micropogon undulatus ) along the east
Louisiana coast. Proc. Gulf Caribb. Fish. Inst.
7:151-158.

TOWNSEND, B C , Jr

1956. A study of the spot, Leiostomus xanthurus
Lacepede, in Alligator Harbor, Florida. M.S. Thesis,
Florida State Univ., Tallahassee, 43 p.

Wallace, D H

1940. Sexual development of the croaker, Micropogon
undulatus, and distribution of the early stages in
Chesapeake Bay. Trans. Am. Fish. Soc. 70:475-482.
Warlen, S. M

1981. Age and growth of larvae and spawning time of
Atlantic croaker in North Carolina. Proc. Annu. Conf.
Southeast. Assoc. Game Fish. Comm. 34:204-214.
White, M. L., and M E Chittenden

1977. Age determination, reproduction and population
dynamics of the Atlantic croaker, Micropogon
undulatus. Fish. Bull., U.S. 75:109-123.
Wiseman, W J , Jr , R E Turner, F J Kelly, Jr, L J Rouse, Jr.,
AND R F Shaw.

1987. Analysis of biological and chemical associations
near a turbid coastal front during winter 1982. Contrib.
Mar. Sci. 29:141-151.

142

BEHAVIOR OF SOUTHERN RIGHT WHALES, EUBALAENA AUSTRALIS ,
FEEDING ON THE ANTARCTIC KRILL, EUPHAUSIA SUPERBA

William M Hamner,^ Gregory S. Stone,^ and Bryan S. Obst^

ABSTRACT

Southern right whales, Eubalaena australis, were observed in 3 successive years on the western side
of the Antarctic Peninsula. These whales do not appear to be from the well-documented Valdes,
Argentina population. The whales we observed were feeding on Antarctic krill, Euphausia superba.
When krill were at the surface right whales surface-skimmed at high speed, with upper jaw lifted
above the water surface. In heavy weather one right whale "tail-sailed" at slow speed, with head
submerged and apparently feeding. When krill were organized in subsurface schools, right whales
engaged in subsurface feeding, diving repeatedly in place, apparently working a given school. One
whale swam directionally to the only known large school of krill in the area and fed intensively,
rested on the surface, then began a second feeding bout. Whales hyperventilated, false fluked, and
fluked prior to feeding dives. These are the first detailed observations of feeding behavior of right
whales in Antarctic waters and suggest that coastal Antarctica may have been (and may become
again) a regular part of the summer feeding range of the species.

Right whales are among the rarest of the great
whales, having been hunted almost to extinction
a century ago. The southern right whale, Eubal-
aena australis, has been studied only recently
and only during the austral winter when the
whales aggregate inshore to bear calves and to
mate (Clarke 1965; Payne 1976, 1986; Best 1981;
Aguayo and Torres 1986). Because right whales
were commercially extinct by the mid-1850's,
very little has been learned about their ecology
from the 20th century whaling industry. Informa-
tion on feeding, migration, stock structure, and
reproductive biology was collected for most other
Antarctic mysticete whales during the heyday of
whaling in this century (e.g.. Mackintosh 1965;
International Whaling Commission reports 1964-
present). The small number of surviving right
whales (ca. 29c of historic levels in the Southern
Hemisphere, Breiwick and Braham 1984) has
made it difficult for researchers to study this spe-
cies. Our current understanding of its feeding and
calving ecology in the Southern Hemisphere
comes from observations made primarily off
Peninsula Valdes, Argentina (Payne 1986).

In Antarctic waters south of lat. 60°S, more
than 30 sightings of right whales have been re-
ported previously (Berzin and Vladimirov 1981;
Goodall and Galeazzi 1986; Ohsumi and Kasa-

iDepartment of Biology, University of California, Los Ange-
les, CA 90024.
2College of the Atlantic, Bar Harbor, ME 04609.

Manuscript accepted September 1987.
FISHERY BULLETIN: VOL. 86. NO. 1. 1988.

matsu 1986). Most were in the vicinity of the
South Orkney Islands, 9 were near the Antarctic
Peninsula, 6 were in the Pacific sector of the
Antarctic, and 2 were south of Africa. We have
observed southern right whales during 3 consecu-
tive austral summers near the western shore of
the Antarctic Peninsula, and the sightings re-
ported herein and in Stone and Hamner (in press)
are the most southerly as well as the most de-
tailed observations.

We sighted one individual during the 1983-
84 austral summer (also recorded by Captain
P. Lenie in the log of the RV Hero ; Goodall
and Galeazzi 1986), two individuals in 1984-85,
and eight in 1985-86, four of which we indi-
vidually identified. In 1986 a fortunate combina-
tion of fair weather and available ship time
permitted us to make the first extended uninter-
rupted observations on the behavior of right
whales feeding on the Antarctic krill, Euphausia
superba.

METHODS

Right whales are distinguished by the absence
of a dorsal fin and regions of cornified skin (cal-
losities) on the head, jaws, and chin. Individual
whales were identified by standard methods,
using video tapes and telephotographs of head
callosities and scarring patterns on the head and
back (Payne et al. 1983; Kraus et al. 1986). When
possible, we dropped large disks of plywood of

143

FISHERY BULLP:TIN: VOL 86, NO 1

known diameter next to the whale and included it
in the photograph to measure whale size.

Behavioral patterns were observed from the
ship's bridge and recorded on a portable computer
as they occurred, using a program which timed
entries of encoded behaviors or coments to the
nearest second. Krill schools were recorded on a
Simrad echosounder and the identity of the or-
ganisms causing the echograms was verified by
net samples taken with an Isaacs-Kidd midwater
trawl and by divers' visual confirmation of krill
schools near the surface.

RESULTS

7 January 1984: At 0500 hour, north of Cape
Murray near Two Hummock Island, one right
whale was feeding at the surface with its upper
jaw lifted above the water, swimming at high
speed (estimated at 8-9 knots by the ship's cap-
tain) in feeding runs of 15-20 seconds, which we
recorded on video. Three humpback whales
nearby were diving in one specific location. Th,e
right whale repeatedly changed direction be-
tween surface runs so that its horizontal direc-
tional feeding excursions did not take it far from
the vicinity of the vertically diving humpbacks.
During these powerful filter-feeding runs enor-
mous amounts of water were displaced, cascading
beside and behind the right whale and producing
a large wake.

15 January 1985: A cow and a calf were swim-
ming slowly at the surface some 500 m from the
eastern shore of Anvers Island. On approach by
the ship the whales swam slowly into shallow
water where we could not follow. It was near dusk
and we could not get photographs for future iden-
tification.

7 January 1986: We encountered one southern
right whale and six humpbacks at 1830 hour at
lat. 63°46'S, long. 61°13'W, between Trinity and
Hoseason Islands. We photographed the head and
body of the right whale for subsequent identifica-
tion. We followed the whale for approximately 2
hours, recording diving times, surface intervals,
and breathing rates. The whale frequently
changed directions underwater and consequently
we often failed to see the whale immediately
when it resurfaced, so breathing rate data for this
behavioral sequence are incomplete. The whale
appeared to have captured krill on at least one
dive because when the whale surfaced it repeat-
edly and briefly opened and closed its mouth, with
baleen visible, a behavior presumably associated

with separation of krill and water prior to swal-
lowing the prey (Watkins and Shevill 1976).
About 50 cape petrels, Daption capensis, alighted
on the water and fed at the surface around the
whale. When the whale's jaw movements ceased,
the birds soon stopped feeding, but they remained
on the water and did not follow the whale when it
swam away at the surface.

2 March 1986: We observed one right whale at
the northern end of the Neumayer Channel,
where a large iceberg was grounded on a 93-
fathom rise 2 miles east of Iceberg Point. The
wind was blowing from the north at 20-24 knots
and a strong surface current was flowing south,
producing a bow wave on the grounded iceberg.
The right whale repeatedly swam NE of the berg,
raised its tail high out of the water at 90° to the
wind, submerged its head, and "sailed" downwind
past the iceberg, a behavior previously noted for
right whales in Argentina (Payne 1976). Soon
after we first saw the whale, it stopped tail-
sailing and began diving, still along the N-S tran-
sect near the iceberg where it had been sailing.
The presence of the ship did not cause the whale
to alter its back-and-forth swimming rhythm or
direction. We waited until the whale began one of
its N-S transects past the iceberg and followed
about 100 m behind it with the ship. A large
school of krill was present on the east side of the
iceberg. We recorded no other schools in the vic-
nity. The whale was accompanied by three female
fur seals. The seals constantly darted about the
head of the whale when it surfaced after long
dives and appeared to annoy the whale, because
several times the whale repeatedly slashed its
head sideways when the seals swam too close.

3 March 1986: At 1030 hour, we spotted a single
southern right whale near the mouth of Andvord
Bay on the Antarctic Peninsula. The whale was
swimming SSW at about 3 knots, making short
dives that lasted about 19 seconds {N = 12,
SD = 9.0 seconds), with brief surface intervals
that averaged 6.1 seconds (A^ = 12, SD = 2.6 sec-
onds) (Fig. 1). The whale then stopped diving but
continued to swim SSW toward the NE tip of
Lemaire Island, swimming mostly at the surface
for approximately 90 minutes. During this period
the whale appeared unconcerned with the ship,
which remained 50-100 m behind it, but when
the whale neared an iceberg that was hard
aground near Lemaire Island, it turned suddenly
at a right angle to its prior course and swam be-
tween the iceberg and the rocks. The ship was
nonetheless able to follow the whale through the

144

HAMNER ET AL : FEEDING BEHAVIOR OF SOUTHERN RIGHT WHALES

120

150

180

1M1

125

155

185

100

105

115

130

135

140

145

160

165

170

175

190 195 200

Time in minutes

205

120

150

180

210

Figure 1. — Dive record of the right whale constantly observed for 3.5 hours on 3 March 1986. 44 minutes
(arrowl: The ship approached to within 10 m of the whale for i.d. photographs and size measurement. 54
minutes: The whale hyperventilated and dove in an area without krill, then swam northward. 92 minutes:
The whale began diving on scattered small krill schools while still traveling N and NNW. 140 minutes: The
whale stopped and began diving on one large concentration of krill. 182 minutes: The whale rested on the
surface, moving its jaws. 194 minutes: The whale began a second feeding bout ca. 400 m further south.

narrow channel. Thereafter the whale ignored
the ship and altered course to NNW, still swim-
ming at the surface. The ship pulled ahead of the
whale to measure its length and this may have
caused the short dive noted at ^ = 43 minutes.
However, the whale calmly surfaced again within
10 m of the ship and watched us while we pho-
tographed it next to the wooden disk. The whale
was 11m long. The whale then continued to swim
at the surface to the NNW. During the period that
the whale swam at the surface without diving the
average time between breaths was 50.8 seconds
(n = 55; SD = 10.8). At ^ = 54 minutes the whale
began to hyperventilate, took three short dives,
lifted the flukes partially clear of the water (false
fluking I on the third dive, and then fully raised
the flukes on the fourth dive, which lasted 210
seconds. The whale then remained at the surface
for about 30 minutes while swimming northward.
This pattern of hyperventilation prior to a long
dive occurred before every long dive sequence
which was preceded by a surface interval of at

least 4 minutes (Fig. 2). We used this criterion to
restrict our dive selection for this plot because
there is some indication that there is also a brief
recovery period involving hyperventilation after
long dives. Of the last 19 breaths that were taken
during the 90 seconds before the 5 long dives plot-
ted in Figure 2, 18 were less than 30 seconds
apart, with a mean interval of 15 seconds
(A^ = 19, SD = 7.3). During the time preceding
this 90-s hyperventilation period, the whale aver-
aged 1 breath every 48 seconds iN - 17, SD =
12.4), not significantly different from the time of
50.8 seconds recorded between breaths during
long surface intervals without dives. During hy-
perventilation, therefore, breathing rate in-
creased by about 3 x .

At ^ = 92 minutes we began to record scattered
small krill schools on the echosounder and the
whale began to dive erratically, with some rea-
sonably long dives, but most quite short. Of 23
dives, 13 (56%) were shorter than 10 seconds; the
time averaged for all dives was 75.3 seconds

145

FISHERY BULLETIN: VOL 86, NO 1

70 r

3 2

Minutes prior to dive

Figure 2. — Breath sequences prior to 5 dives lasting from 93 to 279 seconds. During the 90 seconds
immediately preceding dives, the mean interbreath interval was 15 seconds. Prior to this 90-s period
of hyperventilation the mean interbreath interval was 48 seconds.

(A^ = 20, SD = 74.6). During this period of erratic
diving the whale continued to swim N, then
NNW. At about t = 140 minutes the whale
stopped traveling NNW and began a series of 10
dives, the second being the longest, followed by
successively shorter dives. These dives were fol-
lowed by rather uniform surface intervals and all
10 lasted significantly longer (x = 183.5 seconds,
N = 10, SD = 89.2) than the previous mean of 75
seconds. At the end of the dive sequence, at 182
minutes, the whale stopped swimming entirely
and floated motionless at the surface, occasion-
ally moving its jaws. This inactivity lasted about
12 minutes and then the whale began a second
series of dives. We called these dive sequences,
which consisted of regular length surface inter-
vals interspersed with a series of longer dives of
progressively decreasing duration, "feeding
bouts" because at ^ = 140 seconds the whale had
reached the only major aggregation of krill in the
vicinity, as verified by sonic records (Fig. 3). We
saw dense schools of krill and isolated krill at the
surface from the bow of the ship, and we captured
krill in three successive hauls with the 1 m
Isaacs-Kidd midwater trawl.

During the first feeding bout the whale slowed

from its steady 3-knot swimming speed and
moved slowly at about 1 knot, but no longer in
any specific direction, finally swimming some
400 m to the south before beginning the second
feeding bout. The whale had fed on krill earlier
also, because it defecated during the feeding se-
quence and the feces, as determined by later mi-
croscopic examination, were composed entirely of
well-digested euphausiids.

The display of a false fluke prior to the high
fluke initiating a long dive did not necessarily
indicate presence or absence of krill, but when
krill were present, it was a highly significant pre-
dictor of the length of the dive. In the presence of
prey, when the whale showed its flukes once (i.e.,
did not false-fluke preceding the dive), dive dura-
tion averaged 91.2 ± 13.0 seconds {N ^ 5),
whereas when both a false fluke and a high fluke
preceded the dive, the dive averaged 234.7 ± 64.4
seconds (N = 1).

In March and April 1986 we saw a total of eight
southern right whales. Of these we distinguished
four as individuals on the basis of video tapes and
photographs of callosities and body markings.
One of these whales was the same individual that
we had observed in January some 70 miles fur-

146

HAMNER ET AL : FEEDING BEHAVIOR OF SOUTHERN RIGHT WHALES
0( ■

I/)
(V
<D

Cl
O)

S5:<s,<?

.J4%?*

<»

=:Sf-

kilometer

Figure 3. — Echogram of krill school on 3 March 1986 on which right whale stopped swimming and began

feeding bout.

ther north, as confirmed by comparing video
recordings of both sightings. Photographs of these
four individuals were subsequently compared
with the catalog of southern right whales that
overwinter near Peninsula Valdes, Argentina
(Payne and Rowntree 1984), but none of the
Antarctic individuals were in that compendium of
some 623 individuals.

DISCUSSION

Our observations of southern right whales
along the Antarctic Peninsula suggest that
coastal Antarctica may have been (and may well
become again) a regular part of the summer feed-
ing range of the species. These sightings indicate
a more extensive distribution in the Southern
Ocean than heretofore reported, an extension cor-
roborated by the comparison of recent and his-
toric summer distributional patterns by Ohsumi
and Kasamatsu (1986). Southern right whales
were hunted almost to extinction in the shallow
coastal embayments where they overwinter more
than 100 years before whaling began in the
Southern Ocean (Harmer 1928; Townsend 1935).
It is not surprising, therefore, that there are only
a few scattered records of southern rights in
Antarctic waters, fluke observations so to speak.
During the past 50 years southern rights have

been protected, and several domes of southern
right whales have increased (Best 1981; Ohsumi
and Kasamatsu 1986; Whitehead et al. 1986), and
a commensurate extension of the summer feeding
range into areas formerly occupied by these
whales may be occurring.

It is interesting that none of the 4 individual
whales for which we have photographic identifi-
cations appear in the catalog of 623 individual
southern rights for the Valdes, Argentina breed-
ing population (Payne and Rowntree 1984). We
believe that the right whales we encountered in
Antarctic waters probably winter elsewhere, per-
haps along the Chilean coast, which has been less
well surveyed than has the east coast of South
America (Cardenas et al. 1986).

Payne (1976) described tail-sailing among the
Valdes population of southern right whales and
suggested that this may be a form of play behav-
ior. In Antarctic waters this behavior may be a
method used to forage on krill. The right whale
that we observed tail-sailing had three fur seals
in attendance, constantly darting about its head,
apparently feeding on the same prey as the
whale. Furthermore, tail-sailing occurred next to
a grounded iceberg at a specific location that was
repeatedly transected by the whale, the only spot
in the vicinity where krill were detected on the
ship's sonar.

147

FISHERY BULLETIN: VOL. 86, NO I

Northern right whales feed by both surface-
skimming and subsurface feeding, apparently
holding the jaws agape below the surface much
like they do when at the surface (Watkins and
Schevill 1976, 1979). Surface-skimming by north-
ern right whales that feed on copepods and juve-
nile euphausiids occurs at speeds of about 3 knots
with relatively little turbulence (Watkins and
Schevill 1976, 1979). The one southern right
whale which we observed surface-skimming, and
which was also mentioned in Goodall and
Galeazzi (1986), swam at about 8 knots, with an
immense amount of turbulence. This whale un-
doubtedly was feeding on Euphausia superba.
The extremely high speed of this whale when
surface-skimming may be related to type of prey.
Euphausia superba is a large euphausiid that
swims up to 30 cm/second, and it is quite adept at
avoiding scuba divers and other dark objects
(Hamner et al. 1983; Hamner 1984). It may be
that right whales adjust the speed of their surface
feeding runs to overcome the swimming speed of
the particular prey that they pursue. Fin whales
also exhibit differences in breathing and diving
rhythms when feeding on euphausiids and on
schooled fish (Watkins et al. 1984).

The southern right for which we have a 3.5-h
record in 1986 fed exclusively below the surface.
It swam for about an hour at the surface almost
directly NNW toward the only patch of krill that
we had located previously via the echo sounder
during repeated transects across the mouth of
Andvord Bay. It is not understood how baleen
whales locate prey. They might echolocate, they
might listen for krill noises, or they might re-
member where they had prior success and return
to the general vicinity and then begin to hunt
randomly. Although our data do not allow us to
choose among these possibilities, it appears that
this particular southern right whale did navigate
directly to patches of prey, as suggested by
Watkins and Schevill (1979) for northern right
whales. However, some subsurface searching
probably occurs as well. The whale that we fol-
lowed for the longest period of time made a rea-
sonably long dive of 210 seconds in the middle of
its surface swim toward the krill patch. It appar-
ently did not find anything to eat (nor did the
sonar indicate the presence of krill), and it made
no more significant dives for about 30 minutes.

When the whale reached an area with small
scattered krill schools, its dives were generally
short, and the temporal spacing and duration of
dives and surface intervals were erratic. When

the whale reached a large school of krill it stopped
swimming between dives, dives lasted longer,
surface intervals between dives became more reg-
ular, and successive dives were progressively
shorter. After 10 of these dives the whale floated
without any swimming movements at the surface
for about 12 minutes while occasionally rattling
its baleen and slightly lifting its upper jaw.

The whale clearly anticipated its dives. It hy-
perventilated for about 90 seconds prior to all of
the long dives. Hyperventilation is a common pre-
cursor to diving in air-breathing vertebrates
(Kooyman et al. 1981), but this respiratory pat-
tern apparently has not been reported previously
for a baleen whale.

The data that we present in support of this feed-
ing pattern are admittedly modest, consisting of
only one continuous 3.5-h data sequence for one
individual. Nonetheless, this data set is truly
unique and it includes a series of statistically sig-
nificant changes of behavior at the surface and at
depth which are correlated with prey distribu-
tion. Since the changes of behavior are real, they
merit interpretation. One hypothesis to account
for the increase and then the decrease in dive
length after encountering prey is that the whale
dove to its full capacity in order to maximally
exploit an opportunity to feed. With repeated
dives its dive capacity may have diminished, pro-
ducing first long, then shorter and shorter dives
(Fig. 1). In contrast, this breathing pattern has
not been observed in grey whales, which exhibit a
sustained capacity for repetitive, nearly
monotonic diving while feeding (Obst et al. in
prep.). It is possible that right whales, being more
specialized for surface-skimming, become ex-
hausted more readily during sustained diving.

An alternative hypothesis to account for this
pattern of behavior is that the whale captured
krill with such ease on the first long dive of 348
seconds that a "digestive bottleneck" (Kenward
and Sibly 1977) developed; subsequent dives
would be progressively shorter because the stom-
ach (or some storage chamber) filled more rapidly
than the food could be processed. The long surface
interval of immobility followed by a second feed-
ing bout might then represent the time taken to
clear such a chamber (e.g., the first of the three
stomach pouches) in preparation for the next bout
of feeding. If this hypothesis is correct, it implies
that once a right whale has found a particularly
favorable school of krill, it fills up very rapidly.
This interpretation contrasts markedly with a
popular impression of great whales as continuous

148

HAMNER ET AL.: FEEDING BEHAVIOR OF SOUTHERN RIGHT WHALES

feeders that harvest prey while swimming lazily
through diffuse planktonic resources. Herwig et
al. (1984) presented preliminary evidence that
microbial fermentation of ingesta may occur in
the forestomach of baleen whales, a process which
may be expected to limit the throughput rate of
food in accordance with continuous-flow, stirred
tank reactor kinetics (Penry and Jumars 1987). It
is tempting to speculate that the feeding behavior
of right whales, among the largest endotherms on
earth, may be constrained by a digestive bottle-
neck analogous to that described for humming-
birds (Karasov et al. 1986), the smallest of en-
dotherms. These hypotheses warrant subsequent
testing through more extensive time-budget
analyses of right whales and other balaenopter-
ans on their feeding grounds, such as described
here and in Watkins et al. (1984).

ACKNOWLEDGMENTS

We thank P. Hamner, M. Lang, and the crews
of the RV Hero and RV Polar Duke for field assis-
tance. R. Jedas wrote the computer program for
recording behavior in real time. R. Payne and V.
Rowntree were of great assistance in comparing
photographs of individual Antarctic whales and
of right whales occurring off Peninsula Valdes,
Argentina. P. Hamner provided editorial assis-
tance. This research was supported by NSF
grants DPP83-02852 and DPP86-14821.

LITERATURE CITED

Aguayo. L a , AND N D Torres

1986. Records of the southern right whale, Eubalaena
australis (DesmouHns, 1822), from Chile between 1976
and 1982. Rep. int. Whaling Comm., Spec. Issue 10,
p. 159-160.
BERZIN, a a . AND V L Vladimirov

1981. Changes in the abundance of whalebone whales in
the Pacific and Antarctic since the cessation of their ex-
ploitation. Rep. int. Whaling Comm. 31:495-499.
Best, P

1981. The status of right whales (Eubalaena glacialis) off
South Africa, 1969-1979. S. Afr. Dep. Agric. Fish., In-
vest. Rep. 123, 44 p.
Breiwick, J M . AND H W Braham (editors).

1984. The status of endangered whales. Mar. Fish. Rev.
46(4):l-64.
Cardenas, J C , S S Stutzin. C Cabello, and J Oporto

1986. The first steps to cetacean conservation and man-
agement in Chile. Final Report to the World Wildlife
Fund, Project No. WH-445. World Wildl. Fund U.S.,
Wash., DC.
Clarke, R

1965. Southern right whales on the coast of Chile. Nor.

Hvalf.-Tid. 54:121-128.
Goodall, R N P . AND A R Galeazzi

1986. Recent sightings and strandings of southern right

whales off subantarctic South America and the Antarctic

Peninsula. Rep. int. Whaling Comm., Spec. Issue 10, p.

173-176.
Hamner, W M.

1984. Aspects of schooling in Euphausia superba. J.

Crust. Biol. 4 (Spec. No. l):67-74.
Hamner, W M , P P Hamner, S W Strand, and R W Gilmer.

1983, Behavior of Antarctic krill, Euphausia superba:
chemoreception, feeding, schooling, and molting.
Science 220:433-435.

Harmer, S. F.

1928. A history of Whaling. Proc. Linn. Soc. Lond.
140:51-95.
Herwig, R P , J T Staley, M K. Nerini, and H W Braham

1984. Baleen whales: preliminary evidence for forestom-
ach microbial fermentation. Appl. environ. Microbiol.
47:421-423.

International Whaling Commission

1964-present. Reports. Office of the Commission, Cam-
bridge.
Karasov, W H , D Phan, J M Diamond, and F L Carpenter
1986. Food passage and intestinal nutrient absorption in
hummingbirds. Auk 103:453-464.
Kenward. R E , and R M Sibly.

1977. A woodpigeon (Columba palumbus) feeding prefer-
ence explained by a digestive bottle-neck. J. Appl. Ecol.
14:815-826.
Kooyman, G L . M A Castellini, and R W Davis.

1981. Physiology of diving in marine mammals. Ann.
Rev. Physiol. 43:343-356.
Kraus, S D , K E MooRE, C. A Price, M J Crone, W A.
Watkins, H E Winn, and J. H. Prescott.

1986. The use of photographs to identify individual North
Atlantic right whales (Eubalaena glacialis ). Rep. int.
Whaling Comm., Spec. Issue 10, p. 145-151.
Mackintosh, N A.

1965. The stocks of whales. Fishing News (Books) Ltd.,
Lond.
Ohsumi, S . AND F Kasamatsu

1986. Recent off-shore distribution of the southern right
whale in summer. Rep. int. Whaling Comm., Spec.
Issue 10, p. 177-185.
Payne, R S

1976. At home with right whales. Natl. Geogr. Mag.
149:322-339.

1986. Long term behavioral studies of the southern right
whale (Eubalaena australis ). Rep. int. Whaling Comm.,
Spec. Issue 10, p. 161-167.

Payne, R. S., O Brazier, E M. Dorsey, J. Perkins, V. Rowntree,
AND A. Titus.

1983. External features in southern right whales (Euba-
laena australis) and their use in identifying individu-
als. In Payne, R. S. (editor), Communication and be-
havior of whales. Westview Press, Boulder, CO,
p. 371-445.

Payne, R. S., and V. J. Rowntree.

1984. Un catalogo fotografico de ballenas individuales
avistadas en las aguas alrededor de la Peninsula Valdes,
Argentina. 1st ed. World Wildlife Fund, Center for
Long Term Research, Lincoln, MA.

Penry, D L , and P A Jumars.

1987. Modeling animal guts as chemical reactors. Am.
Nat. 129:69-96.

149

FISHERY BULLETIN: VOL. 86, NO. 1

Stone. G S . and W M HamnKR baleen whales: Eubalaena glacialis. Balaenoptera bore-
In press. Humpback and right whales in the Gerlache alis, Megaptera novaeangliae , and Balaenoptera
Strait on the Antarctic Peninsula. Polar Rec. physalus. J Mammal. 60:155-163.
TowNSEND. C H Watkins, W A . K E Moore. J Sigur-jonsson, D Wartzok, and
1935. The distribution of certain whale species as shown G N DI S( IARA
by logbook records of American whaleships. Zoologica 1984. Fin whale (Ba/aenop<erap/!vsa/us 1 tracked by radio
(N.Y.) 19:1-50. in the Irminger Sea. Rit Fiskideildar 8:1-14.
Watkins. W A. AND W E Schevill Whitehead, H. R Payne, and M Payne

1976. Right whale feeding and baleen rattle. J. Mam- 1986. Population estimate for the right whales off Penin-

mal. 57:58-66. sula Valdes, Argentina, 1971-1976. Rep. int. Whaling

1979. Aerial o'bservation of feeding behavior in four Comm., Spec. Issue 10, p. 169-171.

150

NOTES

AN ECONOMETRIC ANALYSIS OF

IVfET INVESTMENT IN
GULF SHRIMP FISHING VESSELS'

The major capital inputs in the Gulf shrimp fish-
ery are vessels and gear. Early vessels, built
mostly of wood, employed drag seines, cast nets,
and fixed traps to catch shrimp. Today, much
larger and more powerful vessels, equipped with
sophisticated fishing gear and accessories, trawl
the Gulf of Mexico. The last 20 years has seen a
remarkable substitution of steel and fiberglass
vessels for wooden vessels. The factors underlying
this aggregate investment trend in the Gulf of
Mexico (hereafter referred as the Gulf) shrimp
fishery, however, have yet to be examined.

Economic analysis of the fishing industry has
increased in recent years because of the growing
importance of world fish stocks. However, these
studies have been focused predominantly at the
micro or firm level (Thompson et al. 1970^;
Wilson et al. 1970; Juhl 1974; Watson 1977;
Griffin et al. 1978; Prochaska and Cato 1981).
Moreover, accounting for a comprehensive mean
of the cost of capital has been ignored in past
research efforts. Yet, the cost of capital is likely a
major factor in fishing vessel investment deci-
sions in the Gulf shrimp fishery. The long run
profitability of the sector and its exposure to fi-
nancial risk depends, to a large extent, upon its
capital structure and fluctuations in the cost of
debt and equity capital. Futhermore, this deter-
minent of aggregate investment behavior repre-
sents a major channel through which macroeco-
nomic policy actions are transmitted to the Gulf
shrimp fishery.

The purpose of this study is to estimate an
econometric model of annual real net investment
in fishing vessels in the Gulf and to determine the
sensitivity of investment decisions in the indus-
try to fluctuations in the cost of equity and debt
capital. This study begins by examining the indi-
vidual factors that affect the expansion of the

iTechnical Article No. TA-20803 of the Texas Agricultural
Experiment Station, Texas A&M University System. This re-
search was supported in part by the Texas A&M University Sea
Grant College Program under Grant No. NA81AA-D-00092.

2Thompson, Russel G., R. W. Callen, and L. C.
Wolken. 1970. Optimal investment and financial decisions
for a model shrimp fishery firm. Unpubl. Rep. Texas A&M
Univ., TAMU-SG-70-205.

stock of steel, wood, and fiberglass vessels in the
Gulf fieet. The effects of alternative macroeco-
nomic policies on investment expenditure trends
in the Gulf shrimp fishery are then studied. The
final section of this paper presents some conclud-
ing remarks.

Determinants of Net Investment

The aggregate investment model used in this
study is based upon the neoclassical theory of ag-
gregate investment behavior. The determinants
of the desired capital stock of fishing vessels as
well as the relationship between this desired cap-
ital stock and net investment behavior of Gulf
shrimp fishermen are specified in this section.

Desired Stock of Fishing Vessels

In making investment decisions, competitive
firms add to their existing stock of capital as long
as the present value of the periodic net cash flows
generated by an additional unit of capital exceeds
its net purchase price. This condition for any par-
ticular type of fishing vessel (e.g., wood, steel,
fiberglass) can be stated algebraically as follows^:

2 pidX/dKj) - (dTi/dKj) - r{dDt/8Kj}
t=i

idPt/dKj) (1 + p)-' > Qjia - ic)

X U + ^ dRjt/dKj]a + p)-'] (1)

where variable p represents the real price fisher-
men expect to receive per unit of output in the
Gulf of Mexico shrimp industry, X represents the
quantity of shrimp expected to be harvested, Kj is
the real stock of thej'^ category of fishing vessels,
Tt and P/ represent the tax payment and principal
payment due in period t expressed in constant

3The variables without any subscripts are expected at the
time the investment is made to remain constant over time.

FISHERY BULLETIN: VOL. 86, NO. 1, 1988

151

dollars, D, represents debt outstanding in period
t, r is the real rate of interest on debt capital, p is
the real after-tax opportunity rate of return on
equity capital desired by fishermen, Qj is the real
price paid for the^'^ category of fishing vessels at
the retail level, a is the proportion of investment
financed with equity capital, i^. is the investment
tax credit rate, and Rjt represents the real level of
replacement investment in the j^^ category of
fishing vessels required in period t to offset losses
in productive capacity due to wearout.

The entire term on the left-hand side of the
inequality sign in Equation (1) represents the
present value of the additional net cash flows gen-
erated by a permanent addition to the^'^ category
of fishing vessels. It is assumed that both the in-
terest and the principal payments, ((tP/SK^) and
r(dD/BK, ), vary over time as further expenditures
are required to maintain the productive capacity
of this addition to the capital stock at its original
level. The right-hand side of Equation (1) repre-
sents the initial downpayment minus the invest-
ment tax credit plus the present value of all fu-
ture cash outlays required to maintain the stock
of the j^^ category of fishing vessels at its new
level.

To maximize the present value of their equity.
Gulf shrimp fishermen would continue to add to
the stock of the j^^ category of fishing vessels
until Equation (1) holds as an equality. Equiva-
lently, maximization of the present value of
owner equity requires thaf*

pidX/dKj) = j^^

ic + Z - A

(1 - iy)

(2)

where F, represents the present value of the
stream of capacity depreciation of they '^^ category
of vessels and iy is the average income tax rate.
The term Z represents the present value of the
stream of after-tax interest payments and princi-
pal payments on debt while A represents the
present value of the stream of tax depreciation
allowances that can be claimed for each dollar of
investment as the stock of vessels is maintained
at its new level. ^ The right-hand side of this ex-

4Equation (1) as well as the derivation of the implicit rental
price of vessels (c) assume that fishermen expect real prices (p)
and the marginal physical product of vessels dX/dK to remain at
current levels. These and other assumptions which allow us to
treat many components of Equation (1) as consoles are consis-
tent with those employed in Penson et al. (1981) and Coen
(1975).

pression thus represents the implicit rental price
of the J^^ category of fishing vessels.

The concept of the implicit rental price of capi-
tal has been widely employed in previous studies
of investment behavior as a determinant of the
capital stock which firms desire to hold (e.g., Coen
1968, 1975; Penson et al. 1981). Equation (2) sug-
gests that the implicit rental price of they*^^ cate-
gory of fishing vessels will increase if their pur-
chase price, the cost of debt, and equity capital, or
income tax rates increase. These effects, however,
will be offset to some extent by an increase in
the investment tax credit rate, the deductibil-
ity of tax depreciation allowances, and interest
expenses.

Let us assume that output in this industry is a
function, in part, of the stock of fishing vessels
and that this production relationship is of the
Cobb-Douglas form. Letting p^ represent the par-
tial production elasticity associated with the
stock of the 7'^ category of fishing vessels (Kj),
the marginal physical product for these vessels
can be expressed as follows:

idX/dKj) = i^jiX/Kj).

(3)

Substituting Equation (3) into Equation (2), the
desired stock of they'*^^ category of fishing vessels
at the end of year t can be expressed as follows:

K*jt^^j{pX/Cjh

(4)

where Cj represents the expected implicit rental
price of they '^'^ category of fishing vessels given by
the right-hand of Equation (2). Equation (4) im-
plies that the desired stock ofj^^ category of fish-
ing vessels is directly related to the expected real
gross income from Gulf shrimp fishing and is in-

^The nominal value of Z (the present value of the stream of
after-tax loan payments) in Equation (2) is equal to

il -iyW 2j d,(l + ^)''{1 + p)-'

i = 1
n

+ ^ ^ (e - d,)(l -I- (J))-'(l -h p)-'
( = 1
while the real value of A (the present value of the stream of
depreciation allowances is equal to

i^(hl(p + A> + pi> + ^))

where d^ is equal to the nominal interest payment on a loan of
one constant dollar, (J) is the inflation rate, e is the nominal
amortized loan payment on a loan of one current dollar, and 8 is
the tax depreciation rate given by 2ln where n is the service life
of the vessel.

152

versely related to the expected implicit rental
price of these vessels. Similar equations could be
developed for other inputs used in the shrimp
fishing effort.

Desired Net Investment in Fishing Vessels

New fishing vessels are acquired by Gulf
shrimp fishermen both to expand their productive
capacity and to replace losses in the productive
capacity of existing vessels. This partitioning of
observed gross investment into net investment
and replacement investment for the j^^ category
of fishing vessels can be expressed definitionally
as follows:

Njt = Kj, - K,,

h - ^jt

(5)

where 7,7 represents the level of real gross invest-
ment in the 7'^ category of fishing vessels in year
t while Rf is the real replacement investment
needed to offset annual capacity depreciation of
these vessels. The variables K; and K/- 1 represent
the productive capital stock of the 7'^ category of
fishing vessels the end and the beginning of the
year, respectively. Given Equations (4) and (5),
the following relationship between the desired
stock of the j^^ category of fishing vessels and
current real net investment in these durable
inputs can be defined:

QjiK*j,

Kj,-i)

(6)

where 0 < 0^ < 1 and where 0^ represents the par-
tial adjustment coefficient that describes the
speed of adjustment of actual stocks to desired
levels for thej'^ category of fishing vessels. Sub-
stituting Equation (4) into Equation (6) and as-
suming an adaptive expectations hypothesis for
(pXICj)t, the following compound geometric ex-
pression is obtained:

A^^, = 0^ \^jUpXICj)t + (1 - ^j)Njt-i

+ %{l - \j)Kjt-2

Jt-l + \^Jt

(7)

where \j is the adaptive expectations coefficient
and [Xji represents the error term. Since Kjt-2 is
equal to Kjt^i - Nf-i, Equation (7) reduces to the
following estimating equation:

A^,, = bjo + bji (pXICj)t + bj2 Kjt-i

where 60 is the intercept, 61 = 8 px, 62 = -8X,
63 = (1 - \) (1 - 0) and \xt is once again the ran-
dom disturbance term. The estimates of the 61
and 63 coefficients are expected to be positive
while the value of 62 is expected to be negative.^
Equation (8) thus represents the general form of
the equations to be econometrically investigated
in this study.

Data

The time series data used in this study consist
of annual observations for each variable in Equa-
tion (8) over the 1965-77 period. This time period
represents the only period for which investment
expenditure information is available.

The productive capital stock of the^'^'^ category
of fishing vessels is comprised of a series of differ-
ent vintages of vessels or

Kj, = Ijt + (1 - hj^)Ijt.^ + (1 - A,i - hj2)\^_^ + . . .

+ (\-hj^-hj2- .. .- hjr,)Ijt-n (9)

where /i^, is the fraction of thej*^^ category of fish-
ing vessel's original productive capacity lost in
the i^^ year of its service life. The value of A, is
represented by (1 - <}>)'" ^, where <t> = 2/n and n is
the assumed service life.^ In a related matter,
the present value of the stream of capacity de-
preciation of a vessel {Fj) was computed as fol-
lows:

+ bj3Njt-i + iXjt

(8)

6The net investment model expressed in Equation (8) can be
seen as a part of a simultaneous equation system that includes
other investment equations as well as supply equations for all
inputs and the production function for the fishing industry. The
specification of the complete simultaneous system of equations
and measurement of time series data needed to simultaneously
estimate the 6, coefficients in Equation (8) are beyond the scope
of this study. Since the disturbance terms for this set of invest-
ment equations are likely correlated, the seemingly unrelated
regression equations estimator was employed. The disturbance
terms given by this estimator were also examined for autocorre-
lation. The estimated rho coefTicient in this small sample was
shown to be insignificant in all cases. Finally, the predicted
rather than actual value of A'^, _ j was used in estimating Equa-
tion (8) to address the pyossibility of correlation between the
lagged dependent variable and the disturbance term.

''While a geometric decline in productive capacity has been
assumed for fishing vessels, recent studies indicate that the
productive capacity of equipment and machinery deteriorates at
a lower rate in the early period than in latter years. Coen (1975)
suggested that equipment and machinery deteriorate as they
age, though not necessarily at a geometric rate. For farm trac-
tors, a concave decay pattern represents the best proxy for the
capacity depreciation pattern suggested by engineering consid-
erations (Penson et al. 1981). The true pattern which underlines
actual capital spending decisions in the fishing industry could
not be examined due to inadequate data.

153

Fj = ^hj,a + p)

(10)

(=1

Data from the National Marine Fisheries Ser-
vice were used as annual observations on the
nominal value of gross investment in Gulf fish-
ing vessels (U.S. Department of Commerce 1965-
77^). These values were deflated to real terms
using the industrial price index. The quality of
the time series for real net investment in fishing
vessels, N/, depends on how well the annual
values of/, reflect quality changes in vessels over
time.

The annual levels of the implicit rental price of
vessels (c) were computed using the definition
outlined on the right-hand side of Equation (2).
Coen (1975) assumed that the real after-tax rate
of return desired on equity capital, p, is constant
over the economic life of the investment. Follow-
ing the lead of Coen, a value for p of 5% was
employed in this study.

The real rate of interest on nonreal estate loans
at commercial banks, r, along with annual rate of
inflation equals the nominal rate of interest on
debt capital. Annual values for all these variables
were obtained from US. Department of Com-
merce publications. The annual values for the
fraction of investment expenditures that are debt
financed (v];) used in computing Z were found by
dividing the annual change in total debt in the
fishing industry by the annual level of gross
investment in durable inputs provided by the Na-
tional Marine Fisheries Service. The time series
for a, the fraction of capital expenditures fi-
nanced with internal equity capital, was equal to
one minus the percentage debt financed
(l-il/).

Investment tax credit rate, i^, was equal to 7%
during the 1965-68 period, 0% during the 1969-70
period, and I09c during the 1971-77 period. The
maximum corporate income tax was assumed to
represent i^ for the Gulf shrimp fishery. The
double-declining balance method was assumed in
determining the present value of the stream of
annual tax depreciation allowances in A .

The time series data on prices paid for vessels,
a component of the rental price, were measured
using cost data collected from shrimp vessel
builders. Griffin et al. (1978) have shown that

vessel length, material of construction, and year
of purchase were the most significant factors de-
termining the price of a vessel. The equation esti-
mated in that study was used to extend available
vessel price information over the entire time pe-
riod covered by this study.

Econometric Results

Statistical as well as economic criteria can be
employed to evaluate the estimated equations for
the various categories of fishing vessels. The eco-
nomic criteria include the reasonableness of the
elasticities for the economic variables and as well
the partial production elasticities implied for the
production function.

Empirical estimates of the annual real net in-
vestment model for steel, wooden, and fiberglass
vessels indicate statistically significant coeffi-
cients for all but one of the explanatory variables
at the 10% level or less (Table 1). The lone excep-
tion was the coefficient associated with the lagged
capital stock variable in wooden vessel model,
which was not significantly different from zero at
less than the 20% level. All the coefficients associ-
ated with the explanatory variables have the
signs hypothesized earlier in this paper. Finally,
the coefficients on the lagged dependent variable
satisfy the constraint of being both greater than
zero and less than one.

Table 1 . — Estimated coefficients for the annual net investment
model for fishing vessels. Gulf Shrimp Fleet, 1965-77.

Vessel

Constant

(pX/c),

K,-:

N,-^

type

(bo)

(bi)

{b2)

(b3)

Steel

-95.3895

4.3770

-9.1302

0.6318

M6.21)

(6.23)

(2.34)

Wooden

384.3848

0.1132

-3.1944

0.9089

(1.79)

(0.93)

(1.21)

Fiberglass

0.3819

0.2990

-0.9529

0.2765

(5.37)

(1.74)

(2.88)

8U.S. Department of Commerce. 1965-77. Vessel charac-
teristics data. National Marine Fisheries Service, NOAA,
Wash., D.C.

1 Numbers in parentheses indicate absolute values of f -statistic.

Economic criteria employed in evaluating the
reasonableness of the empirical results and in
comparing the investment behavior among vessel
types include the partial production elasticity of
fishing vessels (p). This elasticity is given by
-61/62 which is computed using the estimated
beta coefficients in Equation (7). It appears that
steel vessels, with a partial production elasticity
of 0.479, are highly productive and play an impor-
tant role in the supply of fish. The fishing sector,
unlike other sectors of the economy, depends pri-

154

marily on one major capital input — vessels.
Therefore, the high partial production elasticities
recorded for steel vessels and fiberglass vessels
(0.314) are no surprise. Even though wooden ves-
sels appear to incur more repair and maintenance
costs, attract a lower quality crew, and, for that
matter, are less efficient than the other vessel
types, the low partial production elasticity of
0.033 is surprising. This low partial production
elasticity may have been caused by the fact that
the instrument for A'^^ - i in the wooden vessels
equation used to address the issue of the relation-
ship between the lagged dependent variable and
the error term did a poor job of explaining Nf - i.
An examination of the elasticities associated
with the (pX/c )/ term, computed at the mean, re-
veals that real net investment in steel vessels is
the most sensitive to changes in prices, interest
rates, taxes, and the other factors captured in this
variable. An elasticity of 7.28 associated with this
economic variable was computed for steel vessels.
This means that a 1% change in the (pX/c)t vari-
able causes real net investment in steel vessels to
change by 1.289c. This high investment response
to changes in these economic relationships could
be attributed to the fact that steel vessels, by far
the most productive (as evidenced by the high
partial production elasticity reported earlier), are
the most durable and the most capital intensive.
The elasticity associated with the (pX/c)t term in
the fiberglass and wooden vessel equations were
5.35 and 3.11, respectively. This would suggest
that macroeconomic policy actions would have a
substantially greater effect on real net invest-
ment in steel vessels than, say, wooden vessels.

Impact of Changes in
Cost of Capital

The impact of high real interest rates on the
growth of selected sectors in the economy has
been of great concern in the 1980s. The sensitiv-
ity of annual real net investment in fishing ves-
sels to changes in the real rate of interest is exam-
ined in this section by simulating the estimated
equations under annual real rates of 5 and 109?^.
In the short run (3 years), an increase in the real
rate of interest on debt capital from 5 to 10%
would cause real net investment in fishing ves-
sels to decrease by 3.04%. Annual real net invest-
ment in these fishing vessels would decrease by
15.88% in the long run. As the real rate of interest
on debt capital increases, it becomes more diffi-
cult to justify the purchase of additional vessels

owing to their rising marginal factor cost. Given
the fact that 67% of the cost of new fishing vessels
is normally financed with debt capital, it is not
surprising that rising real interest rates on debt
captial have a significant negative effect on the
long run expansion of the Gulf fleet.

The real cost of equity capital, which reflects
the opportunity cost of the fisherman's own funds,
has a less dramatic effect on annual real net in-
vestment in fishing vessels. This can be at-
tributed to the fact that only 33% of the cost of
new fishing vessels are financed with equity cap-
ital. The short run impact of an increase in the
real cost of equity capital from 5 to 10% translates
into only a 1.76% decrease in annual real net
investment in fishing vessels in the short run.
This same change in the cost of equity capital
would result in a 12.32% decrease in annual real
net investment in the long run.

Summary and Conclusions

This study evaluated aggregate investment be-
havior by fishermen for steel, wooden, and fiber-
glass fishing vessels in the Gulf of Mexico shrimp
fishery and examined the implications of changes
in the cost of acquiring debt and equity capital on
the industry's investment response. This study
showed statistical justification for the theoretical
model of aggregate investment behavior for all
three vessel types.

It is quite evident that the cost of capital plays
an important role in influencing the investment
decisions in the Gulf shrimp fishing industry.
Macroeconomic policies that lead to high real in-
terest rates depress real net investment in this
fishery. Capital expenditures for steel vessels are
the most sensitive to changes in real interest
rates while wooden vessels are the least sensitive.
While low real interest rates are desirable for
stimulating investment activities in the general
economy, they would add to the overcapitaliza-
tion problem which currently exists in the Gulf
shrimp fishing industry. Finally, this study un-
derscores the need to reinitiate efforts to collect
data on gross investment expenditures for differ-
ent categories of fishing vessels in the Gulf fleet.

Literature Cited

COEN, R M

1968. Effects of tax policy on investment in manufactur-
ing. Am. Econ. Rev. 58:200-211.

1975. Investment behavior, the measurement of deprecia-
ble and tax policy. Am. Econ. Rev. 65:59-74.

155

Griffin, W L.J P Nichols, R G Anderson. J E Buckner.and
C M Adams

1978. Cost and returns data: Texas shrimp trawlers, Gulf
of Mexico, 1974-75. Dep, Agr, Econ. Tech. Rep,, TAMU-
SG-76-601, Texas A&M Univ.
JUHL. R

1974. Economics of the Gulf of Mexico industrial and food-
fish trawlers. Mar. Fish. Rev. 36(ll):39-42.
Kmenta. J

1971. Elements of econometrics. MacMillan Company,
N.Y.
Penson, J B . R G Romain, and D W Huches

1981. Net investment in farm tractors: An econometric
analysis. Am. J. Agr. Econ. 63:629-635.
Prochaska. F J. andJ C Cato

1981. Economic conditions in the Gulf of Mexico shrimp
industry: 1960-81. Food Resour. Econ. Staff
Rep., Univ. Florida, Gainesville, FL.
Watson, J W . Jr . and C McVea, Jr

1977. Development of a selective shrimp trawl for the
southeastern United States shrimp fisheries. Mar.
Fish. Rev, 39(101:18-24,
Wilson, R R , R G Thompson, and R W Callen,

1970, Optimal investment and financial strategies in
shrimp fishing, Dep. Agr. Econ. Tech. Rep, TAMU-SG-
70-218. Texas A&M Univ.

John B Penson, Jr,
Ernest O, Tetty
Wade L Griffin

Department of Agricultural Economics
Texas A&M University
College Station, TX 77843-2124

APPENDAGE II^JURY IN DUNGENESS

CRABS, CANCER M AGISTER, IN

SOUTHEASTERN ALASKA

The Dungeness crab, Cancer magister, is com-
mercially important along the western coast of
the United States. Like many decapod crus-
taceans, it can autotomize and regenerate ap-
j>endages to heal wounds and limit injury.

Studies of appendage injury may be useful in
assessing the physical condition of crustacean
populations and the impact of fishing on commer-
cially important species. Incidences of appendage
loss in the field have been reported for species of
crabs other than C. magister (McVean 1976;
McVean and Findlay 1979; Needham 1953). Ap-
pendage loss was studied in adult Dungeness
crabs in Washington (Cleaver 1949) and Oregon
(Waldron 1958) and for juvenile crabs in the Co-
lumbia River estuary (Durkin et al. 1984).

In this study we examined adult Dungeness

crabs in southeastern Alaska to determine the
incidence of missing, regenerating, and damaged
appendages. Temporal incidence of appendage in-
jury was compared to the molting and mating
periods of the crabs and to the commercial fishing
season for Dungeness crabs.

Materials and Methods

Adult Dungeness crabs were collected from Icy
Strait and the Excursion Inlet fjord near Glacier
Bay, AK (lat. 135°30'N, long. 58°25'W), from May
through November 1984-85. Data were obtained
by monthly surveys of commercially caught
crabs. Crab pots (Waldron 1958) were set at
depths of 7 to 20 m and remained in the water for
3 to 11 days. All crabs were held in live tanks
(<24 hours) before they were measured on board
ship. In southeastern Alaska, pots are equipped
with escape rings to permit release of crabs with
carapace widths <165 mm, but sublegal-sized
crabs are often found in the catch.

Carapace width (excluding the 10th anterolat-
eral spines), wet weight, and sex were recorded
for each crab. Carapace condition was graded as
soft-shell (recent molt), new-shell, worn-shell, or
skip-molt (Somerton and Macintosh 1983). The
number and identities of missing, damaged, or
regenerating chelipeds and walking legs were
recorded. An appendage with a cracked cuticle or
missing dactyl was considered damaged. Ap-
pendages smaller in length and diameter than
intact appendages were designated regenerating.
Combined missing, damaged, and regener-
ating appendages are referred to as injured
appendages.

Interrelationships between variables were de-
termined with Pearson correlations (SAS 1985).
Means were compared with Student's ^ -tests, and
chi-square analyses were used to determine if
multiple autotomies occurred by chance (Steel
and Torrie 1960). Data are presented as means
± 1 standard error of the mean.

Results

Males comprised 65% and females 35% of the
878 Dungeness crabs examined. Average cara-
pace widths were 169 ± 0.6 and 159 ± 0.7 mm for
males and females, respectively. Wet weights
were 1,102 ± 9 g for males and 884 ± 14 g for
females. The greatest number of female crabs was
caught in July, and the greatest number of males
in August.

156

FISHERY BULLETIN; VOL 86, NO, 1, 1988.

Seventy-five percent of the crabs were intact,
with no appendage injuries. Twenty-five percent
of all crabs had injured limbs; 18% had missing,
5% had regenerating, and 2% had damaged ap-
pendages. No relationship existed between cara-
pace width and appendage injury.

Most of the Dungeness crabs sampled were in
the worn-shell condition (67%). Twenty-eight per-
cent of all crabs were new-shelled. Only 1% were
soft-shelled and 4% were skip-molts. Correlations
between carapace condition and appendage in-
jury were not significant.

No significant differences existed in appendage
injury between male and female Dungeness
crabs. Injuries were bilaterally symmetrical ex-
cept for the 3d walking leg which was missing
more frequently on the left side (P < 0.05). Con-
sidering only those crabs with missing legs, a
total of 246 legs were missing with a mean of
1.5 ±0.1 missing legs/crab. Ninety-seven crabs
had legs missing on the right side and 98 had legs
missing on the left side. The maximum number of
missing legs per crab was 5. Sixty percent of the
crabs had 1 leg missing, 20% had 2 missing legs,
and 12% were missing 3 or more legs. Of the crabs
with missing legs, 63% were males and 37% were
females.

Forty Dungeness crabs had regenerating legs,
with a mean of 1.2 ± 0.1 regenerating legs per
crab. Sixty percent of those crabs had 1 regenerat-
ing leg, 10% had 2 regenerating legs, and 3% had
3 or more regenerating legs. The maximum num-
ber of regenerating legs per crab was 4. Of the
crabs with regenerating legs, 73% were males
and 27% were females.

Seventeen crabs had damaged appendages with
a mean of 1.1 ± 0.1 damaged appendages/crab. Of
the crabs with damaged appendages, 82% were
males and 18% were females.

The observed number of Dungeness crabs with
2 or more missing appendages was significantly
higher (P <0.01) than expected for both sexes,
indicating that appendage loss was not due only
to chance.

Appendage injury was significantly correlated
with date, with more injuries occurring later in
the year. The number of Dungeness crabs with
missing appendages was significantly correlated
with date for both males and females (P < 0.01).
The lowest percentage of crabs with injured ap-
pendages occurred in July (4.8%, both sexes com-
bined) and increased to a maximum of 34.3% in
November. The percentage of male crabs with re-
generating appendages did not vary significantly

over time and was about 6% for all months. How-
ever, the percentage of female crabs with regener-
ating legs increased from 0% in May to 10.5% in
November (P < 0.01). Male crabs with damaged
appendages increased from 0% to 8.5% from May
to October (P < 0.05) and then decreased to 1.7%
in November.

Chelipeds and 1st and 4th walking legs were
injured most frequently. The hierarchy for fre-
quency of injury for female crabs (chelipeds > 4th
walking legs > 1st, 2d, and 3d walking legs) dif-
fered slightly from the hierarchy for males (che-
lipeds > 1st walking legs > 4th walking legs > 2d
and 3d walking legs). Months in which high per-
centages of crabs had injured chelipeds also had
high percentages with injured 1st (males) and 4th
(females) walking legs (Fig. 1).

The temporal incidence of appendage injury in
Dungeness crabs was compared to life history
events and to the commercial crab fishing season
in southeastern Alaska (Fig. 2). The season
opened 15 June and closed 15 August, reopened
1 October and closed 28 February 1986. Ap-
pendage injuries were low in July and increased
157% from July to August, a period of simulta-
neous molting, mating, and fishing. An addi-
tional increase in appendage injury of 43% oc-
curred in October, even though the fishery was
closed from 16 August to 30 September.

Discussion

Pot samples are biased towards larger sized
Dungeness crabs because of the size of the mesh
on the pot and the presence of two escape rings
with diameters of 11 cm. However, 62% of the
crabs collected for this study were either male
crabs with carapace widths <165 mm or were
females. Very few soft-shell crabs were caught,
even though molting was occurring during part of
the sampling period. Our dependence on commer-
cial crabbers for data collection restricted us to
sampling mostly during the open fishing season
when most of the crabs were not in the soft-
shelled condition.

Twenty-five percent of the Dungeness crabs
sampled in southeastern Alaska had appendage
injuries. In other studies of Dungeness crabs in
Washington, Oregon, and the Columbia River es-
tuary, 18%, 32%, and 62%, respectively, of the
crabs were injured (Cleaver 1949; Waldron 1958;
Durkin et al. 1984). The crabs examined in our
study were held for up to 24 hours in crowded
tanks on board ship before being measured and

157

Appendage Injury in Male Crabs

a.
a.

A ppendage Injury in Fe m ale Crabs

01
C31

ID
T3

C
01
Q

Figure 1. — Monthly percentages of male (A) and female (B) Dungeness crabs with injured (missing + regenerating + damaged)

chelipeds and walking legs.

•o

60-

50-

4 0-

30-

20-

1 0-

FISHING

d" MOLTING

9 MOLTING

FISHIlie

1 15 31
May

—I 1 — T"^ — I n — I 1 1

15 30 15 31 15 31 15 30 15 31
Jun Jul Aug Sep Oct

s
s

s
S
s
s
s

s
s
s
\
\
\
\

15 30
Nov

Males

l\\l Females

Figure 2. — Temporal relationships of percentages of male and female Dun-
geness crabs with injured (missing + regenerating + damaged) append-
ages to the crabs' molting and mating periods and the commercial crab
fishing season in southeastern Alaska.

were sometimes observed grasping other crabs,
but very few autotomized limbs were found in the
tanks.

The estimate of appendage injury may be low if
Dungeness crabs with injured appendages were
less likely to enter pots than intact crabs. In an-
other study, the observed number of Carcinus

maenas missing 2 or more legs was higher than
expected if multiple autotomies occurred by
chance, which was interpreted to mean that in-
jured C. maenas enter pots as readily as intact
crabs (McVean 1976). Because there were more
Dungeness crabs with 2 or more missing legs in
our collections than would be expected if multiple

158

autotomies occurred by chance alone, our data
could be similarly interpreted to suggest that
Dungeness crabs with injured appendages
showed little decrease in pot-entering ability and
that our estimate of injury was accurate. For this
interpretation of the chi-square results to be
valid, one must assume that all injuries occurred
before the crabs entered pots and that injury did
not occur within the pots.

Appendage injury in Dungeness crabs was bi-
laterally symmetrical except for the 3d walking
leg. Interestingly, Easton (1972) demonstrated
that 3d walking legs of Hemigrapsus oregonensis
were the most easily autotomized. In other stud-
ies, both bilateral symmetry and asymmetry have
been reported for different species of crabs
(Durkin et al. 1984; Needham 1953). Asymmetri-
cal appendage loss has been associated with crabs
that move predominantly in one direction, while
symmetrical leg loss occurs in crabs that move
randomly (Needham 1953).

The chelipeds, followed by the 1st and 4th
walking legs, were most vulnerable to injury.
Limb loss has been correlated in other studies
with degree of exposure of the limb; the outermost
limbs, the longest limbs and limbs with postures
that afford little protection are most frequently
lost (Needham 1953). Anterior limbs are lost
more frequently than posterior limbs (Needham
1953). The chelipeds are the most anterior and
one of the most exposed appendages on Dunge-
ness crabs and are frequently used in aggressive
threat postures. After loss of chelipeds, the 1st
walking legs remain as the most anterior, ex-
posed limbs and therefore, the most vulnerable.
The 4th walking legs are the most posterior and
also very exposed limbs on an intact crab.

A significantly greater number of Dungeness
crabs with 2 or more missing legs was observed
than expected if multiple autotomies occurred by
chance, indicating an increased susceptibility to
subsequent appendage loss after initial injury
(Needham 1953; Easton 1972).

The correlations between appendage injury and
date were significant but may not be biologically
important. Although these correlations were sig-
nificant, the r^ values (square of the coefficient of
variation) were low. Increased appendage injury
later in the year may be related to other factors.
Soak times, the length of time pots were left in
the water, were longer later in the year. Dunge-
ness crabs may cannibalize other crabs while con-
fined in pots (Waldron 1958). There may also be
delays between time of injury and subsequent au-

totomy and regeneration. Regeneration of legs in
Dungeness crabs is usually completed after 2 or 3
molts (Cleaver 1949).

Over the sampling period, only b^c of all crabs
had regenerating appendages while 187^ had
missing appendages. The discrepancy may be due
to increased mortality of the crabs following in-
jury (McVean and Findlay 1979), or by the effi-
cient, yearly removal of legal-sized, injured crabs
by the commercial fishery.

When temporal incidence of appendage injury
was compared with the opening and closing of the
commerical Dungeness crab fishing season, con-
siderable appendage injury occurred when the
fishery was closed. Closure of the fishery tradi-
tionally occurs during the crabs' mating period,
when a high percentage of soft-shelled female
crabs are present in the population. There was,
however, some overlap in late July and early
August in fishing, molting, and mating. Exclud-
ing damage by humans, potential causes of ap-
pendage injury are aggression between males
competing for females, the cheliped-to-cheliped
mating embrace of Dungeness crabs that can last
up to a week, cannibalism, and the increased vul-
nerability of females which molt prior to mating
(Butler 1960; Durkin et al. 1984). Damage to
Dungeness crabs can also result from other fish-
ing gear such as trawls (Reilly 1983), but no other
commercial fisheries occurred in the study area
while the Dungeness crab fishery was closed.

The results of our studies indicate that Dunge-
ness crabs in southeastern Alaska are in com-
parable condition to adult populations of Dunge-
ness crabs examined in Washington and Oregon,
in terms of appendage injury. Further studies are
needed to investigate the effect of appendage in-
jury on survival of Dungeness crabs and the con-
tribution of handling injury and mortality of
crabs in the commercial fishery.

Acknowledgements

T. Meyers and D. Erickson provided technical
assistance for which we are very appreciative. We
would like to thank T. Olsen, C. Kondzela, and
D. Sterritt for assistance in data collection. This
research was funded by Alaska Sea Grant project
R/06-20.

Literature Cited

Butler, T. H.

1960. Maturity and breeding of the Pacific edible crab,

159

Cancer magister Dana. J. Fish. Res. Board Can. 17:641-

646.
Cleaver, F. C.

1949. Preliminary results of the coastal crab {Cancer

magister) investigation. Wash. Dep. Fish., Biol. Rep.

No. 49A, 82 p.
DuRKiN, J. T., K. D. Buchanan, and T. H. Blahm.

1984. Dungeness crab leg loss in the Columbia River estu-
ary. Mar. Fish. Rev. 46(l):22-24.

Easton, D. M.

1972. Autotomy of walking legs in the Pacific Shore crab
Hemigrapsus oregonensis. Mar. Behav. Physiol. 1:209-
217.
McVean, a.

1976. The incidence of autotomy in Carcinus maenas
(L.) J. Exp. Mar. Biol. Ecol. 24:177-187.
McVean, A., and I. Findlay.

1979. The incidence of autotomy in an estuarine popula-
tion of the crab Carcinus maenas. J. Mar. Biol. Assoc.
U.K. 59:341-354.
Needham, a. E.

1953. The incidence and adaptive value of autotomy and
of regeneration in Crustacea. Proc. Zool. Soc. Lond.
123:111-122.
Reilly, p. N.

1983. Effects of commercial trawling on Dungeness crab
survival. P. W. Wild and R. N. Tasto (editors), In Life
history, environment, and mariculture studies of the
Dungeness crab. Cancer magister, with emphasis on the
central California fishery resource, p. 165-174. Calif
Dep. Fish Game Fish Bull. 172.
SAS Institute, Inc.

1985. SAS user's guide: statistics. 5 ed. Cary, NC, 956

P-
Somerton, D. a., and R. A. MACINTOSH.

1983. Weight-size relationships for three populations in
Alaska of the blue king crab, Paralithodes platypus
(Brandt, 1850) (Decapoda, Lithodidae). Crustaceana
45:169-175.
Steel, R. G. D., and J. H. Torrie.

1960. Principles and procedures of statistics with special
reference to the biological sciences. McGraw-Hill Book
Co., N.Y. 481 p.
Waldron, K. D.

1958. The fishery and biology of the Dungeness crab
(Cancer magister Dana) in Oregon waters. Fish Comm.
Oreg., Contrib. 24, 43 p.

Susan M. Shirley
Thomas C. Shirley

Juneau Center for Fisheries and Ocean Sciences
University of Alaska- Fairbanks
11120 Glacier Highway
Juneau. AK 99801

REEXAMINATION OF THE USE OF OTOLITH
NUCLEAR DIMENSIONS TO IDENTIFY

JUVENILE ANADROMOUS AND

NONANADROMOUS RAINBOW TROUT,

SALMO GAJRDNERI^

Otoliths are a potential source of taxonomic char-
acteristics for identifying stocks offish (Ihssen et
al. 1981). Differences in dimensions of the otolith
nucleus have provided a basis for separating win-
ter from summer races of steelhead, anadromous
rainbow trout, Salmo gairdneri. In addition,
otoliths provided data from which to distinguish
steelhead from resident nonanadromous forms as
well (McKern et al. 1974; Rybock et al. 1975).
Neilson et al. (1985) studied the development of
sagittal otoliths in resident rainbow trout and
steelhead from south-central British Columbia,
and identified sources of variability in the size of
otolith nuclei. However, they were unable to find
morphometric differences between the two forms
of trout. They concluded that the usefulness of
dimensions of the otolith nucleus for separating
steelhead from resident rainbow trout was much
more limited than that suggested by Rybock et al.
(1975) for rainbow trout in the Deschutes River,
Oregon.

The difference in mean length of the otolith
nuclei between the rainbow trout studied by Ry-
bock et al. (1975) and those studied by Neilson et
al. (1985) suggested either population differences
or differences in defining the nuclear boundary.
These disparate results, which led to opposite con-
clusions, limit the usefulness of measurements of
otolith nuclei for the racial identification of juve-
nile rainbow trout until the source of these differ-
ences is better understood. Consequently, to de-
termine whether juveniles of the two forms could
be distinguished by differences in dimensions of
otolith nuclei, we measured the nuclei in sagittae
from steelhead and resident rainbow trout col-
lected from the same Deschutes River, OR, loca-
tions used by Rybock et al. (1975). We used the
definitions proposed by Rybock et al. and by Neil-
son et al. (1985), and compared our measure-
ments for the two forms with each other and with
published values.

Methods

Resident rainbow trout and steelhead were col-

iQregon State University Agricultural Experiment Station
Technical Paper No. 8279.

160

FISHERY BULLETIN: VOL. 86, NO. 1, 1988.

lected from three locations in the Deschutes
River, OR. Resident rainbow trout, which were
collected from the main stem near the mouth of
Nena Creek in March 1985, were mature and
smaller (280-450 mm FL) than the steelhead,
and, based on analyses of scales and otoliths
(McKern et al. 1974), had never entered salt-
water. Juvenile progeny of steelhead were col-
lected from Round Butte Hatchery on the
Deschutes River in 1984. Wild juvenile rainbow
trout (<200 mm FL) of unknown parental origin
were collected in 1984 and 1985 from Bakeoven
Creek, an important spawning tributary for steel-
head in the Deschutes River.

Sagittae removed from rainbow trout were
stored in 90"^^ ethanol for up to two months. Be-
fore they were viewed, one otolith from each pair
was mounted (concave face up) with epoxy on a
glass slide. The back of the slide was blackened
with indelible ink. The otolith was ground by
hand with 600 grit wet sandpaper and periodi-
cally inspected under a light microscope at 100 x
until the microstructure of the nucleus, as de-
scribed by Neilson et al. (1985), was visible. The
otolith was rinsed with 59f HCl for several sec-
onds to remove scratches and improve resolution.

To reduce bias, we coded each slide with a ran-
dom number and ordered the slides sequentially
for viewing. Otoliths were examined with a Zeiss^
dissecting microscope at 125 x. A camera lucida
attachment enabled us to use a computer digitizer
to measure three dimensions of the otolith. In
measuring length and width of the central nu-
cleus, we used the first growth increment encir-
cling all the central primordia, which was the
nuclear boundary defined by Neilson et al. (1985).
In addition, we measured the maximum length
along the longest axis through an area defined by
the first metamorphic check, a narrow hyaline
ring surrounding an opaque ring with a hyaline
center, to replicate the measurements of Rybock
et al. (1975).

We used analysis of variance (ANOVA) to test
for significant differences in each dimension of
the otolith nuclei among groups in our study.
Where adequate data were available, we tested
for significant differences between groups in our
study and similar groups described by Rybock et
al. (1975) and Neilson et al. (1985) for mean di-
mensions of otolith nuclei. Neilson et al. (1985)
showed that the mean length of otolith nuclei for

rainbow trout incubated at 6.5°C was signifi-
cantly less than those for trout incubated at 9.5°
or 15.0°C. Because of this discrepancy, we evalu-
ated the potentially confounding effects of incuba-
tion temperature on the comparisons of otolith
dimensions between our samples and those of Ry-
bock et al. (1975), by testing the hypothesis that
water temperatures during 1967-69 were higher
than those during 1982-83. We used a paired t-
test of average daily water temperatures recorded
by the U.S. Geological Survey on the 1st and 15th
day of each month from 1 January to 1 August
during 1967-69 and 1982-83 (U.S. Department
of the Interior Geological Survey 1967, 1968,
1969, 1982, 1983 1. These dates represent the incu-
bation periods for most of the resident rainbow
and steelhead trout sampled in our study and by
Rybock et al. (1975). Incubation temperature for
steelhead at Round Butte Hatchery is from hatch-
ery records. We estimated spawning and incuba-
tion periods for resident rainbow and steelhead
trout on the basis of reports of the Oregon Depart-
ment of Fish and Wildlife (Fessler 1972) and per-
sonal observations.

Results

For each dimension, we failed to reject the hy-
pothesis (a = 0.05) that rainbow trout collected
from different populations for our study had
otolith nuclei of the same size (Table 1). There-
fore, we concluded that these dimensions could
not be used to discriminate between the resident
and steelhead forms of rainbow trout sampled in
our study.

Water temperatures during 1967-69 were
slightly greater than those during 1982-83
(t = 2.03, df = 14, P = 0.03). Mean difference be-
tween the two periods was 0.8°C. Spawning dates
for resident rainbow trout and steelhead differ;
steelhead spawn from January to April and resi-

Table 1. — Means, standard errors (in parentheses), and sample
size for three otolith dimensions in resident rainbow trout and steel-
head from three Deschutes River populations.

2Reference to trade name does not imply endorsement by the
National Marine Fisheries Service, NOAA.

Dimensions of nuclei

Nucleus

Check

Populations

No of

length

width

length

compared

fish

(mm)

Resident

0.173

0.070

0.323

rainbow trout

(0.006)

(0.003)

(0.012)

Hatchery

0.190

0.070

0.349

steelhead

(0.006)

(0.002)

(0.009)

Suspected

0.178

0.069

0.312

wild steelhead

(0.006)

(0.002)

(0.007)

161

dent rainbow trout spawn from May to mid-July
(Fessler 1972). Mean water temperature during
the period of steelhead egg incubation was 8.4°C
for 1967-69 and 7.6°C for 1982-83. Mean water
temperature during the period when resident
rainbow trout eggs were incubating in the main
stem of the river was 12.6°C in 1967-69 and
11.9°C in 1982-83. Incubation temperature for
steelhead at Round Butte Hatchery was 10°C and
did not vary.

The dimensions of otolith nuclei from resident
rainbow trout and steelhead in our study were
indistinguishable from those in fish from British
Columbia. No significant difference (a = 0.05) in
mean length of otolith nuclei existed between the
British Columbia steelhead incubated at 9.5° or
15°C and suspected wild steelhead from Bakeoven
Creek or Round Butte Hatchery steelhead incu-
bated at 10°C. Among resident rainbow trout, the
mean length of otolith nuclei for fish from the
Deschutes River was also not significantly differ-
ent from that for fish from British Columbia incu-
bated at 9.5° or 15°C. Because Rybock et al. ( 1975)
did not provide variances, we were unable to test
the hypothesis that means from our study coin-
cided with theirs. However, mean length and
width of otolith nuclei in our study were 29 and
559c less, respectively, for resident rainbow trout
and 49-70% less, respectively, for steelhead than
those studied by Rybock et al. (1975).

Discussion

The similarity of our results to those of Neilson
et al. (1985), who used similar methods, might be
expected for different populations under similar
genetic and environmental control. The disparate
results of our study and that of Rybock et al.
(1975) for the same populations after little ge-
netic change (based on comparisons of unpub-
lished, biochemical genetic data for these popula-
tions from 1972 to 1974 and 1984 to 1986) and
little environmental change partly reflected the
use of different definitions for the nucleus. We
defined the nuclear boundary as the first growth
ring surrounding all the fused primordia,
whereas Rybock et al. (1975) defined the nucleus
as the hyaline area in the center of the otolith
that is bounded by a metamorphic check formed
at hatching; they resolved the check by rendering
the otolith with HCl. We also measured the
length of the check surrounding the nucleus, as-
defined by Rybock et al. (1975), which we found
either to correspond with the area enclosed by the

first check or to increase in density of growth
increments surrounding both the central and ros-
tral primordia. The close similarity between our
estimate for Round Butte Hatchery steelhead
(0.349 mm) and the mean calculated by Rybock et
al. (1975) for steelhead (0.354 mm) suggested sim-
ilar checks. It is unclear, however, why values for
resident rainbow trout for this dimension and the
results of tests to discriminate races differed be-
tween the two studies. Rybock (1973) noted that
the nuclear check could not be distinguished in
29% of the otoliths and that the use of HCl may
have caused the frequent confusion between the
metamorphic check and other groups of daily
growth rings. The grinding and polishing of
otoliths greatly reduce this source of error. Neil-
son et al. (1985) also discouraged the use of meta-
morphic checks as boundaries because the causal
links between checks and developmental events,
such as hatching, have not yet been established.

Neilson et al. (1985) demonstrated that nuclear
length increased significantly with increase in in-
cubation temperature from 6.5° to 9.5°C but not
from 9.5° to 15°C. Although average water tem-
peratures in the Deschutes River were 0.8°C
lower during 1982-83 than in 1967-69, it is un-
likely that such differences completely explain
the greater estimates of mean length and width of
otolith nuclei in the earlier study by Rybock et al.
(1975). Rybock et al. (1975) calculated mean nu-
clear lengths and widths of 0.354 and 0.230 mm
for steelhead and 0.243 and 0.154 mm for resident
rainbow trout in the Deschutes River. Our esti-
mates were 29-70% less than their estimates for
a 0.8°C difference; whereas under controlled con-
ditions in British Columbia, mean nuclear length
for resident rainbow trout at 6.5°C was 18% less
for resident rainbow trout and 21% less for steel-
head than the nuclear length for fish incubated at
9.5°C, a difference of 3°C (Neilson et al. 1985).

Comparisons of otolith nuclear dimensions be-
tween resident rainbow trout and steelhead incu-
bated at similar temperatures would establish
whether significant differences exist for these
measurements between the two races from the
Deschutes River. The use of a common definition
of nuclear boundaries would allow better com-
parisons between studies. However, given the dis-
parate results of our study, which were similar to
the results of Neilson et al. (1985), and the origi-
nal study for steelhead and resident rainbow
trout in the Deschutes River, as well as our fail-
ure to discriminate between races using both nu-
clear definitions proposed by Neilson et al. (1985)

162

and Rybock et al. (1975), we believe that popula-
tion differences do not explain the differences in
results between the studies of Rybock et al. (1975)
and Neilson et al. (1985). Furthermore, our study
provided strong evidence to support the conclu-
sion of Neilson et al. (1985) that the usefulness of
measurements of otolith nuclei to identify sym-
patric juvenile progeny of resident rainbow trout
and steelhead reared in the wild may be limited.

Acknowledgments

We thank Jeff Light for his advice on grinding
and polishing otoliths to resolve their nuclear di-
mensions and Eric Volk for his review of this
manuscript. This research was funded by
Bonneville Power Administration, U.S. Depart-
ment of Energy, Agreement No. DE-A179-
83BP13499.

Literature Cited

Fessler, J L.

1972. An ecological and fish cultural study of summer
steelhead in the Deschutes River, Oregon. Fed. Aid
Fish. Prog. Rep. Proj. No. F-88-R-1. Oreg. State Game
Comm., Portland, OR, 47 p.

IHSSEN. P E . H E BooKE. J M Casselman, J M McGlade, N. R.
Payne, and F M Utter

1981. Stock identification: materials and methods. Can.
J. Fish. Aquat. Sci. 38:1838-1855.

McKern. J. L., H. F HoRTON, and K V. Koski

1974. Development of steelhead trout (Salmo gairdneri)
otoliths and their use for age analysis and for separating
summer from winter races and wild from hatchery stocks.
J. Fish. Res. Board Can. 31:1420-1426.

Neilson. J D . G L Geen, and B Chan

1985. Variability in dimensions of salmonid otolith nu-
clei: implications for stock identification and microstruc-
ture interpretation. Fish. Bull., U.S. 83:81-89.

Rybock. J. T

1973. Use of otoliths to differentiate juvenile steelhead
trout from juvenile rainbow trout in the lower Deschutes
River, Oregon. M.S. Thesis, Oregon State University,
Corvallis, OR, 44 p.

Rybock. J T , H F Horton, and J L Fessler.

1975. Use of otoliths to separate juvenile steelhead trout
from juvenile rainbow trout. Fish. Bull., U.S. 73:654-
659.

United States Department of the Interior Geological
Survey

1967. Water resources data for Oregon, Pt. 2. U.S. Dep.
Inter., Geol. Surv., Water quality records, 163 p.

1968. Water resources data for Oregon, Pt. 2. U.S. Dep.
Inter., Geol. Surv., Water quality records, 145 p.

1969. Water resources data for Oregon, Pt. 2. U.S. Dep.
Water quality records, 137 p.

1982. Water resources data for Oregon water year 1982.
U.S. Dep. Inter., Geol. Surv., Eastern Oregon, Vol. 1,
206 p.

1983. Water resources data for Oregon water year 1983.

U.S. Dep. Inter., Geol. Surv., Eastern Oregon, Vol. 1,
202 p.

Kenneth P. Currens

Carl B Schreck

Hiram W. Li

Oregon Cooperative Fishery Research Unit
Oregon State University
Corvallis. OR 973313

3Cooperators are Oregon State University, Oregon Depart-
ment of Fish and Wildlife, and U.S. Fish and Wildlife Service.

AGE-SPECIFIC VULNERABILITY OF

PACIFIC SARDINE, SARDINOPS SAGAX, LARVAE

TO PREDATION BY

NORTHERN ANCHOVY, ENGRAULIS MORDAX

To a large degree interannual variability in re-
cruitment determines the size of pelagic fish pop-
ulations. Recruitment to the Pacific sardine,
Sardinops sagax, population off California varies
from year to year over several orders of magni-
tude and is unrelated to spawning stock size
(Murphy 1966; MacCall 1979). Variable mortal-
ity rates in the first year of life must determine
year-class strength, although the sources of this
variability are unknown. Mortality rates in the
earliest stages are size specific with highest rates
in the egg and yolk-sac stage (Ahlstrom 1954;
Butler 1987) and may contribute to variability in
year-class strength (Smith 1985).

The sources of mortality of sardine larvae have
yet to be investigated. In other pelagic larvae,
mortality is due to either starvation or predation,
and starvation is significant only during the brief
period after the onset of feeding (O'Connell 1980;
Hewitt et al. 1985; Theilacker 1986; Owen et al.
1987). In sardines, significant mortality occurs
during the egg and yolk-sac stages (Ahlstrom
1954) and this mortality can only be due to preda-
tion. Variable mortality in older larval and juve-
nile sardines may also contribute to variability in
recruitment, and this mortality, as in other
fishes, may also be due to predation (Hunter
1984).

The objective of this paper was to determine the
size-specific vulnerability of Pacfiic sardine lar-
vae to predation by adult northern anchovies,
Engraulis mordax. The vulnerability of cape
anchovy and northern anchovy larvae to

FISHERY BULLETIN: VOL. 86, NO. 1, 1988.

163

cannibalism has been investigated by Brownell
(1985) and Folkvord and Hunter ( 1986) and found
to be an important source of mortality. In this
paper the vulnerability of sardine larvae will be
compared with that of anchovy larvae and differ-
ences in the biology of sardines and anchovies
will be discussed.

Our approach was to observe the avoidance be-
havior of Pacific sardine larvae in response to
predatory attacks by northern anchovy adults.
Adult northern anchovy were chosen as a preda-
tor because the northern anchovy was the most
abundant pelagic fish in the California Current
region during the waning years of the sardine
fishery and because its planktivorous diet in-
cludes fish eggs and larvae (Loukashkin 1970;
Hunter and Kimbrell 1980).

Materials and Methods
Experimental Fishes

The Pacific sardine larvae used in the experi-
ments were reared from eggs spawned in the lab-
oratory. Adult Pacific sardines were collected off
San Diego and held in 175 m"^ aquarium for six
months. Males and females with developing go-
nads were isolated in spawning tanks and in-
jected with 250 mg human chorionic go-
nadotropin and on the following day injected with
200 units pregnant mare serum and 20 mg
salmon pituitary extract. On the third day fertil-
ized eggs were collected from the spawning tank.
Larval rearing procedures follow those described
by Hunter (1976). Temperature in the rearing
tanks was maintained at 21°C.

Apparatus and Procedures

Experimental apparatus and procedures were
the same as those described by Folkvord and
Hunter (1986) but will be briefly outlined here.
Experimental predators were two groups of 5
adult northern anchovy (range of standard
lengths 8.4-9.2 cm). Predators were maintained
in two rectangular fiberglass tanks (0.75 x 2.15
X 0.83 m = 1.35 m"^) and fed adult brine shrimp
except on days of experimental observation. Sea-
water was supplied continuously to the tanks ex-
cept during experiments. The, temperature in the
observation tank ranged from 16.2° to 22.8°C
(mean = 20.1°C). Two 100 W incandescent lamps
produced 2,000-3,000 mc at the surface of each
tank. A black plastic tent enclosing a window on

one side of the tank provided a darkened observa-
tion chamber.

Each trial consisted of the encounter of three
prey with the predators. Prey were introduced
into the observation tank with a clear glass
beaker. Initial feeding behavior of the predators
is quite variable but becomes less variable as the
predators become accustomed to prey in the tank.
For this reason, prior to each experiment adult
Artemia were introduced as prey for five consecu-
tive trials to standardize predator behavior. After
the preliminary trials with Artemia, three trials
with sardine larvae were alternated with one
trial with brine shrimp until 18 trials with sar-
dine larvae were completed. Each experiment
was concluded with a trial of brine shrimp to test
for satiation.

The number of observations for each larval size
class was the total number of predator-prey inter-
actions observed among larvae in that size class.
The mean standard length was determined from
20 larvae sampled randomly from the rearing
tank on the day of each behavior experiment. The
numbers of observations for each size class (mean
SL) were 8.0 mm, 41; 11.3 mm, 51; 12.1 mm, 114;
12.7 mm, 46; 14.1 mm, 81; 17.6 mm, 104; and 19.6
mm, 69. Experiments were not extended to larger
sizes due to insufficient numbers of larvae.

Classification of Behavior

Prey behavior was scored only when the preda-
tor attacked a prey. Four measures of predator-
prey interactions were calculated: predator at-
tack distance, the distance from which the
predator responded to the prey and initiated its
attack; frequency of avoidance response; fre-
quency of escapes; and predation rate (percentage
of larvae captured during the 5-min trials). An
avoidance response was a change in speed or tra-
jectory of a larvae before the predator had com-
pleted its attack by closing its mouth. An escape
was defined as a larval response in which the
predator failed to capture the larvae in a single
attack. Typically adult anchovy make a single
attack on a prey item and do not pursue a prey
that escapes (Folkvord and Hunter 1986) but
rather continue searching the tank. Thus attacks
on one prey item were recorded twice if the first
attack was unsuccessful. Although predator at-
tack distance was recorded, this measurement is
highly subjective and comparison with the meas-
urements of other observers is suspect. We did not
analyze predator attack distance for this reason.

164

To compare response and escape behaviors of
Pacific sardine with those reported by Folkvord
and Hunter (1986) for northern anchovy at simi-
lar stages of development, mean lengths were
converted to ages using field growth rates back-
calculated from otolith increment widths for each
species. Confidence limits of the percentage of lar-
vae responding or escaping attack were estimated
assuming the binomial distribution.

Results

Probability of Response to Attack

The youngest larval stages of both sardine and
anchovy were the most vulnerable to predation.
Only n7c of 8 mm sardine larvae (smallest size
tested) responded to attack by adult anchovy.
With increasing size more sardine larvae re-
sponded to attack. At 20 mm, the largest size
tested, 61% of the larvae responded to attack. The
response rate of Pacific sardine larvae was consis-
tently lower than that of northern anchovy larvae
of similar lengths (Fig. 1). Although this differ-
ence in responsiveness could be due to differences
in the observer, it may also be explained by the
difference in age of anchovy and sardine at the
same length. Sardine larvae are about 6.2 mm
when they begin feeding (age = 5 days from fertil-
ization at 17°C), whereas first-feeding anchovy
larvae are only about 4.3 mm (age = 5 days from
fertilization at 17°C) (Zweifel and Lasker 1976).
Sardine larvae also grow faster than anchovy lar-

100

o 80 1-

O 60

U
OC

20 -

ANCHOVY

V.--

SARDINE

8 12 16

LENGTH (mm)

Figure 1. — Increase by size of the percentage of Pacific sardine
larvae, Sardinops sagax, and northern anchovy larvae, En-
graulis mordax, responding to attack by adult northern an-
chovy and 95% confidence intervals. Data on anchovy larvae
from Folkvord and Hunter (1986).

vae at the same temperature (Butler and Rojas de
Mendiola 1985). Thus, sardine larvae are younger
at a given size than anchovy larvae.

Since the latency of response to attack must be
related to the development of the central nervous
system (Webb 1981; Webb and Corolla 1981), it
may be more appropriate to compare sardine lar-
vae with anchovy larvae of the same age. For that
reason lengths of the larvae of both species were
converted to age using growth rates measured in
the field (Methot and Kramer 1979; Butler 1987).
Comparison of the percentage of larvae respond-
ing to attack at a given age (Fig. 2) reveals no
significant difference in the rate of development
of response to attack. Thus, the escape response
develops at the same rate in Pacific sardine and
northern anchovy, and the difference in propor-
tion of larvae responding at a given size (Fig. 1)
is due to the difference of size at hatching and
the difference in growth rates of the two
species.

100

a.
<
O
If)

OC
UJ
Q.

80 -

60 -

_ h

ANCHOVY . -'-

- -|

r .••■■

..?'

r .■■•••

r J

, [

•■••n

1 '

SARDINE

20
AGE (days)

Figure 2. — Increase by age of the percentage of Pacific sardine
larvae, Sardinops sagax, and northern anchovy larvae, En-
graulis mordax, responding to attack by adult northern an-
chovy and 95% confidence intervals. Size categories of reared
larvae have been converted to ages using growth rates esti-
mated from the field. Data on anchovy larvae from Folkvord
and Hunter (1986).

Probability of Escaping Attack

The ability to successfully avoid attack in-
creased with size of Pacific sardine as well as
northern anchovy. Few small larvae of either spe-
cie escaped attack by adult northern anchovy.
Only 3% of 8 mm sardine larvae escaped attack
and the percentage of larvae escaping increased
to only 11% for 17 mm larvae and 13% for 20 mm

165

larvae (Fig. 3). The proportion of small anchovy
larvae escaping attack was also low {67c ) but in-
creased with size to 737r of 22 mm larvae
(Folkvord and Hunter 1986). The numbers of lar-
vae escaping attack were significantly different
between anchovy and sardines at sizes larger
than about 13 mm. Conversion of lengths to age
using field growth rates does not eliminate the
differences between sardine and anchovy (Fig. 4).
Sardine larvae older than 20 days were more vul-
nerable to predation than anchovy larvae of the
same age (Fig. 4).

100 I—

z
o
a

O
a.

UJ
Q.

ANCHOVY .f

V..-- ■

-SARDINE

8 12 16

LENGTH (mm)

Figures. — Increase by size of the percentage of Pacific sardine
larvae, Sardinops sagax, and northern anchovy larvae, En-
graulis mordax, escaping attack by adult northern anchovy and
95'^ confidence intervals. Data on anchovy larvae from
Folkvord and Hunter (1986).

1 \J\J

i 80

.,.•'1

z
o

Q. 60
V)

\u
a.

^ 40
o

ANC

i- <

r>-'

UJ 20

~ <

■.C-^'

J-

SARDINE

, ^

-l_. 1

1 F t r

) 1

20 30

GE (

days)

Figure 4. — Increase by age of the percentage of Pacific sardine
larvae, Sardinops sagax, and northern anchovy larvae, En-
graulis mordax, escaping attack by adult northern anchovy and
95% confidence intervals. Size categories of reared larvae have
been converted to ages using growth rates estimated from the
field. Data on anchovy larvae from Folkvord and Hunter (1986).

Discussion

The proportion of Pacific sardine larvae re-
sponding to attack and escaping attack increased
with size and with age. Our results differ from
those reported by Folkvord and Hunter (1986) for
anchovy larvae in the rate at which sardine lar-
vae respond and escape attacks at given sizes and
ages. It should be noted that, although the
methodology was the same, the observers were
different. This difference could affect rate of re-
sponse to attack. It also should be noted that the
size of adult anchovy used by Folkvord and
Hunter (1986) ranged from 83 to 89 mm SL,
whereas the size range was 84-95 mm SL in our
study and that the size of predator influences the
number of larvae escaping (Folkvord and Hunter
1986). The slightly larger size of predators used in
this study is not sufficient to explain difference in
escapement, nor is the difference in observer
likely to affect the rate of escapement since the
observer's task is to examine whether the larvae
are escaping or are being eaten.

The greater vulnerability to predation of sar-
dine larvae than anchovy larvae has interesting
implications. In general, larger larvae are less
vulnerable to predation than small larvae. Bailey
(1984) and Bailey and Batty (1983) compared the
vulnerability of cod, flounder, plaice, and herring
larvae to predation by invertebrate predators.
They found that herring larvae were the least
vulnerable larvae because herring were more re-
active and had the greatest escape speeds. Sar-
dine larvae are larger at hatching and at a given
age are larger than anchovy larvae. In our exper-
iment sardine larvae react to predatory attacks at
similar rates as anchovy larvae, but escape attack
at a much lower rate.

This difference in vulnerability to attack may
be due to differences in swimming behavior. An-
chovy larvae swim using beat and glide locomo-
tion (Hunter 1972). The escape behavior is usu-
ally a burst of swimming from a motionless
position (Folkvord and Hunter 1986). We ob-
served that sardine larvae, however, swim contin-
uously and they respond to attack by changing
direction and increasing speed. This difference in
swimming mode may affect escape behavior in
two ways. The escape behavior of sardine larvae
may be less flexible than that of anchovy larvae
because the direction the sardine larvae takes is
largely determined by its trajectory. Since sar-
dine larvae cruise, their scope for activity (escape
behavior) may be limited. Anchovy larvae accel-

166

erate from a standing start and have the possibil-
ity of moving in a number of directions. We spec-
ulate that the beat and glide behavior of anchovy
larvae may not only be hydrodynamically more
efficient (Weihs 1974) but also may reduce the
vulnerability to predation.

Acknowledgments

Roderick Leong maintained sardines in the lab-
oratory and spawned the adults after manipula-
tion with hormones. Sandor Kaupp reared sar-
dine larvae in the laboratory. Arild Folkvord and
Clelia Booman elaborated experimental proce-
dures. John R. Hunter commented on the
manuscript.

Literature Cited

Ahlstrom, E H

1954. Distribution and abundance of egg and larval popu-
lations of the Pacific sardine. Fish. Bull., U.S. 56:83-
140.
Bailey, K. M.

1984. Comparison of laboratory rates of predation on five
species of marine fish larvae by three planktonic inverte-
brates; effects of larval size on vulnerability. Mar. Biol.
(Berl.) 79:303-309.

Bailey. K M . and R S. Batty

1983. A laboratory study of predation on Aurelia aurita on
larval herring (Clupea harengus ): experimental observa-
tions compared with model predictions. Mar. Biol.
(Berl.) 72:295-301.

Brownell, C. L.

1985. Laboratory analysis of cannibalism by larvae of the
Cape anchovy Engraulis capensis. Trans Am. Fish. Soc.
114:512-518.

Butler, J. L.

1987. Comparisons of the early life history parameters of
Pacific sardine and northern anchovy and implications
for species interactions. PhD Thesis, University of
California at San Diego, San Diego, 242 p.

Butler, J L., and B. Rojas de Mendiola

1985. Growth of larval sardines off Peru. CalCOFI Rep.
XXVI:113-118.

Folkvord. A., and J R Hunter

1986. Size-specific vulnerability of northern anchovy, En-
graulis mordax, larvae to predation by fishes. Fish.
Bull., U.S. 84:859-869.

Hewitt, R P . G H Theilacker. and N C H Lo

1985. Causes of mortality in young jack mackerel. Mar.
Ecol. Prog. Ser. 26:1-10.
Hunter. J R

1972. Swimming and feeding behavior of larval an-
chovy, Engraulis mordax. Fish. Bull., U.S. 70:821-
838.
1976. Culture and growth of northern anchovy, Engraulis
mordax, larvae. Fish. Bull., U.S. 74:81-88.

1984. Inferences regarding predation on the early life
stages of cod and other fishes. Flodevigen rapportser.,
1., ISSN 0333-2594 The propagation of cod Gadus
morhua L, p. 533-562.

Hunter. J R . and C A. Kimbrell

1980. Egg cannibalism in the northern anchovy, En-
graulis mordax. Fish. Bull., U.S. 78:811-816.
Loukashkin, a S

1970. On the diet and feeding behavior of the northern
anchovy, Engraulis mordax (Girard). Proc. Calif Acad.
Sci. 37 (Ser. 4):419-458.
MacCalL, A. E.

1979. Population estimates for the waning years of the
Pacific sardine fishery. Calif Coop. Oceanic Fish. In-
vest. 10:72-82.
Methot, R D , JR , AND D Kramer.

1979. Growth of northern anchovy, Engraulis mordax,
larvae in the sea. Fish. Bull., U.S. 77:413-423.

Murphy, G I.

1966. Population biology of the Pacific sardine (Sardinops
caerulea). Proc. Calif Acad. Sci. 34:1-34.
OConnell, C, p.

1980. Percentage of starving northern anchovy, En-
graulis mordax, larvae in the sea as estimated by histo-
logical methods. Fish. Bull., U.S. 78:475-484.

Owen, R W , J L Butler, N C H Lo. A Alvarlno. G H Thei-
lacker, J R. Hunter, J. Hakanson, and Y Watanabe.
In press. Spawning and survival patterns of larval an-
chovy in contrasting environments — a site intensive ex-
periment. Cal. Coop. Fish. Invest. Rep.
Smith, P E.

1985. Year-class strength and survival of 0-group clu-
peoids. Can. J. Fish. Aquat. Sci. 42 (Suppl. 11:69-82.

Theilacker. G H

1986. Starvation-induced mortality of young sea-caught
jack mackerel, Trachurus symmetricus , determined with
histological and morphological methods. Fish. Bull.,
U.S. 84:1-15.

Webb, P. W

1981. Responses of northern anchovy, Engraulis mordax,
larvae to predation by a biting planktivor, Amphiprion
percula. Fish. Bull., U.S. 79:727-735.

Webb. P W . and R T Corolla

1981. Burst swimming performance of northern anchovy,
Engraulis mordax, larvae. Fish. Bull., U.S. 79:143-
150.
Webb. P W , and D Weihs.

1986. Functional locomotor morphology of early life his-
tory stages of fishes. Trans. Am. Fish. Soc. 115:127.
Weihs, D.

1974. Energetic advantages of burst swimming of
fish. J. Theoret. Biol. 48:215-229.
ZWEIFEL, J E , AND R. LaSKER

1976. Prehatch and posthatch growth of fishes — a general
model. Fish. Bull., U.S. 74:609-621.

John L. Butler
Darlene Pickett

Southwest Fisheries Center

National Marine Fisheries Service, NOAA

P.O. Box 271

LaJolla.CA 92117

167

INDUCTION OF SPAWNING IN
THE WEAKFISH, CYNOSC/ON REGALIS

Reproductive activity in the weakfish, Cynoscion
regalis, is associated with extensive, north-south
migrations that result in spawning in the estuar-
ies of the Middle Atlantic Bight during the late
spring and early summer. Spawning is appar-
ently related to increasing water temperature
and day length, but there have been no experi-
mental investigations of specific factors that con-
trol this process in the weakfish. In contrast, gen-
eral aspects of the reproductive biology of the
species are well known (reviewed by Mercer
1983). Both males and females become sexually
mature at 1 year of age, and remain sexually
active throughout their lifespan (10^ years).
Spawning involves external fertilization of eggs
by pairs or small aggregations of fish.

There has been limited study of larval develop-
ment in the laboratory, and descriptions of
growth and development of larval weakfish come
entirely from field investigations (Lippson and
Moran 1974). Weakfish larvae resulting from
gametes stripped from sexually mature adults
captured in the field have been reared for a few
days on natural zooplankton diets (Public Service
Electric and Gas Company 1984), but no informa-
tion is available on mortality or growth rates on
prescribed diets and rations. In contrast, Houde
and Taniguchi (1981) and Taniguchi (1981, 1982)
have conducted extensive investigations of the ef-
fects of diet, ration, and temperature on growth
and survival of larval spotted seatrout, Cynoscion
nebulosus, under laboratory conditions. Similar
studies have been conducted with other sciaenids
including red drum, Sciaenops ocellatus (Holt and
Arnold 1983; Holt et al. 1981), spot, Leiostomus
xanthurus (Powell and Gordy 1980), and
bairdiella, Bairdiella icistia (May 1974).

The present paper describes a technique for in-
duction of spawning in a laboratory population of
weakfish and provides preliminary information
on early development and growth of weakfish lar-
vae.

Methods

Sixteen adult weakfish, C. regalis, (approxi-
mately 30-45 cm and 0.5-1.5 kg) were captured
in September 1984 in Delaware Bay by hook-and-
line. Five of these fish were dissected and found to
have regressed gonads. The remaining 11 fish
were placed in two large tanks (2,000 L) con-

nected to a 20,000 L recirculating system that
delivered temperature-controlled seawater to
each tank at 10 L/minute. Water in the recircu-
lating system was replaced approximately
monthly. Ordinary white room light was provided
by two 1.25 m fluorescent lamps positioned 1 m
above the surface of the water in each tank. Ini-
tial conditions in the tank were similar to ambi-
ent conditions in Delaware Bay in September:
18°-19°C, 30%c, and 12 hours light:12 hours dark.
Temperature and salinity in the tanks were mea-
sured daily and pH approximately weekly. The
pH was maintained between 7.0 and 7.6 by
additions of new seawater to the system; this was
accomplished by replacing 40% of the water in the
system with new seawater approximately
monthly.

Fish began to feed 5-7 days after capture. Diet
consisted of an ad libitum ration of sliced squid
with weekly additions of penaeid shrimp or fresh
calf liver. After approximately one month in the
laboratory, the fish were subjected to a prescribed
regimen of temperature and photophase (Fig. 1).
Temperatures were lowered and light phase
shortened over a period of three weeks until con-
ditions reached 8 hours light, 13°-14°C; this ap-
proximated winter conditions on the continental
shelf off Cape Hatteras where adult fish are
known to overwinter (Merriner 1976). Fish were
held under these conditions for 11 weeks after
which temperature was gradually raised and
light phase increased until spring conditions of 14
hours light and 22°-23°C were reached. Fish were
held at these conditions throughout a period of
extended spawning activity. After spawning ac-
tivity ceased, six newly captured fish were placed
in the system to replace fish that had died during
the previous year, and the process of gradually
changing temperature and photoperiod to winter
conditions was repeated (Fig. 1).

Fertilized eggs were buoyant and exited the
tanks at the surface via stand pipes that emptied
into a sump tank. The drain was located at the
bottom of the sump tank allowing eggs to accu-
mulate at the water surface. The presence or ab-
sence of eggs in the sump was determined daily
with a fine-meshed dip net. After collection, eggs
from each spawning were allowed to hatch in 20 L
plastic aquaria filled with 5.0 ixm filtered sea-
water at 30%c and 23°C. Eggs exposed to gentle
aeration under these conditions hatched in 24-36
hours.

Larvae were cultured in 2 L beakers filled with
filtered seawater (25°C, 30%f, and 5 mg/liter

168

FISHERY BULLETIN: VOL 86, NO. 1, 1988.

MAY JUNE JULY AUG SEPT

LU
IT

Ld
Q_

25
24
23

20 H
19
18
17
16
15-1
14
13
12
II
10

▼ T

C/)

-z.

OCT NOV

DEC

JAN

FEB MAR APR

MAY

Figure 1. — Temperature-photophase regime used to condition northern weakfish, Cynoscion regalis, to spawn
in the laboratory. Values are 7-d means. Standard deviations are plotted for temperature. Arrows on time axis
indicate dates of spawning. A = 1984-85. B = 1985-1986.

169

chloramphenicol) at a density of 25 larvae/liter.
Larvae were fed rotifers, Brachionis plicatilis, be-
ginning two days after hatching; on the seventh
day after hatching, brine shrimp Artemia sp.
were added to the diet.

Early larval development from spawning to
first feeding was determined for larvae from three
separate spawnings. Newly hatched larvae (6-15
hours old) were pipetted into an aquarium and
10-15 removed immediately and preserved in
70% ethanol. Additional larvae were sacrificed
daily for nine days. These samples were examined
to relate early development and age.

Results and Discussion

Three adult fish died of undermined causes dur-
ing the period of winter conditions in 1984-85.
Four additional fish died after jumping from the
tanks. The initial spawning by the remaining
population of four fish occurred five weeks after
spring conditions were achieved. This was fol-
lowed the next day by another spawning. Spawn-
ing continued for nine weeks at 10-14 d intervals.
Spawning episodes usually consisted of produc-
tion of fertilized eggs over 2 successive days.
While actual spawning was never observed, it al-
ways occurred between sunset and 08:00 the fol-
lowing day.

After cessation of spawning, the four fish were
removed from the tanks and their sex deter-
mined. Two of the fish drummed when handled
and were clearly males. At least one (and proba-
bly both) of the two remaining fish were females.
The fish were returned to the system after exam-
ination.

The temporal sequence of the spawning events
(two successive days at 10-14 d intervals) sug-
gests that each female may have spawned as
many as four times during the 9-wk period. This
is in contrast to reports for natural populations of
C. regalis (e.g., Merriner 1976), but has been
reported for laboratory populations of other
sciaenid fishes. For example, Arnold (1984) ob-
served 82 spawning events in a laboratory popu-
lation of 12 C. nebulosus over a 27-mo period and
52 spawning events in a similar population of six
S. ocellatus over a 3-mo period. It is not clear why
the C. regalis in our investigation ceased spawn-
ing after several months of long days and high
temperatures while the C. nebulosus in Arnold's
system continued to spawn over a much longer
period. Perhaps this is related to the smaller an-
nual variation in photophase and temperature

typical of C. nebulosus habitats. However, go-
nadal resorption has been reported for at least
one other sciaenid species iB. icistia) held in the
laboratory for extended periods of long day-length
(May 1974).

Spawning from the 1985-86 conditioning pe-
riod first occurred four weeks after spring
conditions were reached with a second spawning
nine days later. Further spawning in the 1985-
86 population did not occur because of an
unidentified infection resulting in the death of all
10 fish in the system over a period of a few weeks.
Autopsies revealed that all fish had highly
developed ovaries or testes at the time of
death.

Newly hatched larvae (6-15 hours old) had a
yolk sac, no mouth, and little development of the
eyes. By 24-36 hours after hatching, the yolk sac
had been virtually absorbed, the mouth was just
beginning to form, and the eyes were not yet pig-
mented. By 48-60 hours, larvae had a completely
formed mouth and digestive system, and the eyes
were pigmented. Larvae were capable of feeding
at this stage.

Chloramphenicol improved larval survival,
which was as high as 24% over 11 days. This
survival is comparable to that reported for other
sciaenid larvae (Holt et al. 1981; Houde and
Taniguchi 1981; Holt and Arnold 1983). However,
growth was less than the maximum seen in the
laboratory for C. nebulosus. After 11 days C. re-
galis larvae in the present experiments had
grown from a mean, posthatching size of 2.7 mm
(36 fxg dry weight) to 4.5 mm (235 \i.g dry weight).
In contrast Houde and Taniguchi (1981) found
that one group of C. nebulosus larvae reached a
size of 13.6 mm (7,082 fxg dry weight) in 12 days
when reared on a concentrated ration of natural
zooplankton at very low stocking density and
high temperature (32°C). However, when fed a
rotifer diet at comparable temperatures and
stocking densities, growth of C. nebulosus larvae
was somewhat less than that of C. regalis larvae
in the present experiments.

Our results show that the spawning cycle of
weakfish can be manipulated to produce repeated
spawnings without the aid of hormone injections.
While the fish appear to resorb their gonadal tis-
sue after several months of exposure to long day
length and high temperature, the differential ma-
nipulation of several groups of fish could allow
year-round production of fertilized eggs. Further-
more, survival and growth of larvae produced in
this manner appear comparable to survival and

170

growth of other sciaenids reared in the labora-
tory.

Acknowledgments

This study was supported by funds provided by
the University of Delaware Sea Grant College
Program and by the U.S. Department of Fish and
Wildlife through the Delaware Department of
Natural Resources and Environmental Control.
Paul Cosby, Big John Ellsworth, Anne Marie Ek-
lund, John Ewart, Paul Grecay, Anne Hastings,
Kathleen Little, Leslie Picoult, and Peter Rowe
were indispensible in the capture, care, and feed-
ing of the adult and larval weakfish used in the
investigation.

Literature Cited

Arnold, C R

1984. Maturation and spawning of marine finfish. In
Carl J. Sindermann (editor). Proceedings of the seventh
U.S. -Japan meeting on aquaculture, marine finfish cul-
ture held at Tokyo, Japan, October 3-4, 1978, p. 25-
27. U.S. Dep., NOAA Tech. Rep. NMFS 10.
Holt, G J , and C R. Arnold

1983. Effects of ammonia and nitrite on growth and sur-
vival of red drum eggs and larvae. Trans. Am. Fish. Soc.
112:314-318.
Holt, G J , R Godbout, and C R. Arnold

1981. Effects of temperature and salinity on egg hatching
and larval survival of red drum, Sciaenops ocel-
lata. Fish. Bull., U.S. 79:569-573.
Houde, E D.. and a K Taniguchi

1981. Marine fish larvae growth and survival. Effects of
density-dependent factors: spotted seatrout iCynoscion
nebulosus ) and lined sole iAchirus lineatus ). Report to
U.S. Environmental Protection Agency, Environmental
Research Laboratory, Narragansett, R.I. EPA 600/3-81-
052. Natl. Tech. Inf Serv. PB82-101395, 66 p.
LiPPSON, A J , and R L. Moran

1974. Manual for identification of early developmental
stages of fishes of the Potomac River estu-
ary. Environmental Technology Center, Martin Mari-
etta Corporation, Baltimore, MD, 282 p.
May,R C

1974. Effects of temperature and salinity on fertilization,
embryonic development, and hatching in Bairdiella icis-
tia (Pisces:Sciaenidae), and the effect of parental salinity
acclimation on embryonic and larval salinity toler-
ance. Fish. Bull., U.S. 73:1-22.
Mercer. L P.

1983. A biological and fisheries profile of weakfish,
Cynoscion regalis. N.C. Dep. Nat. Resour. Comm. Dev.
Div. Mar. Fish., Spec. Sci. Rep. No. 39, 107 p.
Merriner, J V.

1976. Aspects of reproductive biology of the weakfish,
Cynoscion regalis, in North Carolina. Fish. Bull., U.S.
76:18-26.
Powell, A B , and H R Gordy

1980. Egg and larval development of the sp)ot Leiosto-
mus xanthurus (Sciaenidae). Fish. Bull., U.S. 78:701-
714.

Public Service Electric and Gas Company.

1984. Salem Generating Station 316 (b) Demonstration.
Appendix XI. Weakfish (Cynoscion regalis): A synthesis
of information of natural history with reference to occur-
rence in the Delaware River and Estuary and involve-
ment with the Salem Generating Station. Public Ser-
vice Electric and Gas Company, Newark, NJ.

Taniguchi, A K.

1981. Survival and growth of larval spotted seatrout
(Cynoscion nebulosus ) in relation to temperature, prey
abundance and stocking densities. Rapp. P. -v. Reun.
Cons. int. Explor. Mer 178:507-508.

1982. Growth efficiency estimates for laboratory-reared
larval spotted seatrout fed microzooplankton or roti-
fers. In C. F. Bryan, J. V. Conner, F. M. Truesdale (ed-
itors), Proceedings of the Fifth Annual Larval Fish Con-
ference, p. 6-11. Louisiana State University,
Cooperative Fishery Research Unit, Baton Rouge, LA.

Charles E Epifanio

David Goshorn

Timothy E. Targett

College of Marine Studies
University of Delaware
Lewes, DE 19958

171

ERRATA

Fishery Bulletin: Vol. 85, NO. 4

Botton, Mark L., and John W. Ropes, "Populations of horseshoe crabs, Limulus
polyphemus , on the northwestern Atlantic continental shelf," p. 809.

Page 809, Table 3, footnote was omitted:

'Depth at the end of this tow was 439 m

Henwood, Tyrrell A., and Warren E. Stuntz, "Analysis of sea turtle captures and
mortalities during commercial shrimp trawling," p. 814.

Page 814, paragraph 1, line 3, the equation should read:

E = (nets * length/30.5 m) * (min/60)

INFORMATION FOR CONTRIBUTORS TO THE FISHERY BULLETIN

Manuscripts submitted to the Fishery Bulletin will reach print faster if they conform to the following instructions.
These are not absolute requirements, of course, but desiderata.

CONTENT OF MANUSCRIPT

The title page should give only the title of the paper,
the author's name, his/her affiliation, and mailing ad-
dress, including ZIP code.

The abstract should not exceed one double-spaced
page.

In the text, Fishery Bulletin style, for the most part,
follows that of the U.S. Government Printing Office
Style Manual. Fish names follow the style of the Amer-
ican Fisheries Society Special Publication No. 12, A List
of Common and Scientific Names of Fishes from the
United States and Canada , Fourth Edition, 1980.

Text footnotes should be typed separately from the
text.

the literature cited. The accuracy of the literature

cited is the responsibility of the author. Abbreviations
of names of periodicals and serials should conform to
Serial Sources for the BIOSIS Data Base. (Chemical Ab-
stracts also uses this system, which was developed by
the American Standards Association.)

Common abbreviations and symbols, such as mm,
m, g, L, mL, mg, °C (for Celsius), %, %o, and so forth,
should be used. Abbreviate units of measure only when
used with numerals. Periods are only rarely used with
abbreviations.

We prefer that measurements be given in metric
units; other equivalent units may be given in parenthe-
ses.

FORM OF THE MANUSCRIPT

Figures and tables, with their legends and headings,
should be self-explanatory, not requiring reference to
the text. Their placement should be indicated in the
right-hand margin of the manuscript.

Preferably figures should be reduced by photography
to 53/4 inches (for single-column figures, allowing for
50% reduction in printing), or to 12 inches (for double-
column figures). The maximum height, for either width,
is 14 inches. Photographs should be printed on high
quality or glossy paper.

Do not send original drawings to the Scientific Editor;
if they, rather than the photographic reductions, are
needed by the printer, the Scientific Publications Office
will request them.

Each table should start on a separate page. Consis-
tency in headings and format is desirable. Vertical rules
should be avoided, as they make the tables more expen-
sive to print. Footnotes in tables should be numbered
sequentially in arable numerals. To avoid confusion
with powers, they should be placed to the left of numer-
als.

Acknowledgments, if included, are placed at the
end of the text.

Literature is cited in the text as: Lynn and Reid
(1968) or (Lynn and Reid 1968). All papers referred to in
the text should be listed alphabetically by the senior
author's surname under the heading "Literature Cited."
Only the author's surname and initials are required in

The original of the manuscript should be typed,
double-spaced, on white bond paper. Please triple space
above headings. We would rather receive good dupli-
cated copies of manuscripts than carbon copies. The se-
quence of the material should be;

TITLE PAGE

ABSTRACT

TEXT

LITERATURE CITED

TEXT FOOTNOTES

APPENDIX

TABLES (Each table should be numbered with an

arable numeral and heading provided).
LIST OF FIGURES (Entire figure legends)
FIGURES (Each figure should be numbered with

an arable numeral; legends are desired)

ADDITIONAL INFORMATION

Send the ribbon copy and two duplicated or carbon
copies of the manuscript to:

Dr. Andrew E. Dizon, Scientific Editor

Fishery Bulletin

Southwest Fisheries Center La Jolla Laboratory

National Marine Fisheries Service, NOAA

P.O. Box 271

La Jolla, CA 92038

Fifty separates will be supplied to an author free of
charge and 50 supplied to his organization. No covers
will be supplied.

Contents — Continued

CURRNES, KENNETH P., CARL B. SCHRECK, and HIRAM W. LI. Reexamina-
tion of the use of otolith nuclear dimensions to identify juvenile anadromous and -^
nonanadromous rainbow trout, Salmo gairdneri 160

BUTLER, JOHN L., and DARLENE PICKETT. Age-specific vulnerability of
Pacific sardine, Sardinops sagax, larvae to predation by northern anchovy, En-
graulis mordax 163

EPIFANIO, CHARLES E., DAVID GOSHORN, and TIMOTHY E. TARGETT.
Induction of spawning in the weakfish, Cynoscion regalis 168

GPO 791-008

kt^'

Co.

Fishery Bulletin

•^'■^rEs o<

P^^^^^^^^

Vol. 86, No. 2

ocr^8

April 1988

Migratibns- of jjoho' sa^lmon,

PEARCY, WILLIAM G., and JOSEPH P. FISHI

Oncorhynchus kisutch, during their first summer in the ocifta'n-?'*.^.^*^.;^^. T?: . . . 173

DUTIL, J.-D. and J.-M. COUTU. Early marine Hfe of Atlantic salmon/Sa?7m>
salar, postsmolts in the northern Gulf of St. Lawrence 197

MURPHY, MICHAEL L., JOHN F. THEDINGA, and K V. KOSKI. Size and diet
of juvenile Pacific salmon during seaward migration through a small estuary in
southeastern Alaska 213

BOLZ, GEORGE R., and R. GREGORY LOUGH. Growth through the first six
months of Atlantic cod, Gadus morhua , and haddock, Melanogrammus aeglefinus ,
based on daily otolith increments 223

NYMAN, ROBERT M., and DAVID O. CONOVER. The relation between spawn-
ing season and the recruitment of young-of-the-year bluefish, Pomatomus salta-
trix , to New York 237

JAHN, A. E., D. M. GADOMSKI, and M. L. SOWBY. On the role of food-seeking
in the suprabenthic habit of larval white croaker, Genyonemus lineatus (Pisces:
Sciaenidae) 251

WILLIAMS, AUSTIN B. New marine decapod crustaceans from waters influenced
by hydrothermal discharge, brine, and hydrocarbon seepage 263

MARTIN, JOEL W., FRANK M. TRUESDALE, and DARRYL L. FELDER. The
megalopa stage of the Gulf stone crab, Menippe adina Williams and Felder, 1986,
with comparison of megalopae in the genus Menippe 289

SHENKER, JONATHAN M. Oceanographic associations of neustonic larval and
juvenile fishes and Dungeness crab megalopae off Oregon 299

DAGG, M. J., P. B. ORTNER, and J. AL-YAMANI. Winter-time distribution and
abundance of copepod nauplii in the northern Gulf of Mexico 319

HERRNKIND, WILLIAM F., MARK J. BUTLER IV, and RICHARD A. TANKERS-
LEY. The effects of siltation on recruitment of spiny lobsters, Panulirus
argus 331

KIRKLEY, JAMES E., and DALE E. SQUIRES. A limited information approach
for determining capital stock and investment in a fishery 339

POLACHECK, TOM. Analyses of the relationship between the distribution of
searching effort, tuna catches, and dolphin sightings within individual purse seine
cruises 351

(Continued on back cover)

Seattle, Washington

U.S. DEPARTMENT OF COMMERCE

C. William Verity, Jr., Secretary

NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION
William E. Evans, Under Secretary for Oceans and Atmosphere

NATIONAL MARINE FISHERIES SERVICE

James W. Brennan, Assistant Administrator for Fisheries

Fishery Bulletin

SCIENTIFIC EDITOR, Fishery Bulletin

Dr. Andrew E. Dizon

Southwest Fisheries Center La Jolla Laboratory

National Marine Fisheries Service, NCAA

P.O. Box 271

La Jolla, CA 92038

Editorial Committee

Dr. Jay Barlow

National Marine Fisheries Service

Dr. William H. Bayliff

Inter-American Tropical Tuna Commission

Dr. George W. Boehlert

National Marine Fisheries Service

Dr. Robert C. Francis
University of Washington

Dr. James R. Kitchell
University of Wisconsin

Dr. William J. Richards
National Marine Fisheries Service

Dr. Bruce B. CoUette

National Marine Fisheries Service

Dr. Tim D. Smith

National Marine Fisheries Service

Mary S. Fukuyama, Managing Editor

The Fishery Bulletin (ISSN 0090-0656) is published quarterly by the Scientific Publications Office, National Marine Fisheries Service,
NOAA, 7600 Sand Point Way NE, BIN C15700, Seattle, WA 98115. Second class postage is paid in Seattle, Wash., and additional
offices. POSTMASTER send address changes for subscriptions to Fishery Bulletin, Superintendent of Documents, U.S. Government
Printing Office, Washington, DC 20402.

Although the contents have not been copyrighted and may be reprinted entirely, reference to source is appreciated.

Fishery Bulletin

CONTENTS / LIBRARY

OCT 2 8 )98g '

Vol. 86, No. 2 I April1988

PEARCY, WILLIAM G., and JOSEPH P. FISHERi..,Migrfcns of coho salr^^n, :
Oncorhynchus kisutch , during their first summer in thr nrrTHl i i , J.73

DUTIL, J.-D. and J.-M. COUTU. Early marine life of Atlantic salmon, Salmo
salar, postsmolts in the northern Gulf of St. Lawrence 197

MURPHY, MICHAEL L., JOHN F. THEDINGA, and K V. KOSKI. Size and diet
of juvenile Pacific salmon during seaward migration through a small estuary in
southeastern Alaska 213

BOLZ, GEORGE R., and R. GREGORY LOUGH. Growth through the first six
months of Atlantic cod, Gadus morhua , and haddock, Melanogrammus aeglefinus ,
based on daily otolith increments 223

NYMAN, ROBERT M., and DAVID O. CONOVER. The relation between spawn-
ing season and the recruitment of young-of-the-year bluefish, Pomatomus salta-
trix , to New York 237

JAHN, A. E., D. M. GADOMSKI, and M. L. SOWBY. On the role of food-seeking
in the suprabenthic habit of larval white croaker, Genyonemus lineatus (Pisces:
Sciaenidae) 251

WILLIAMS, AUSTIN B. New marine decapod crustaceans from waters influenced

by hydrothermal discharge, brine, and hydrocarbon seepage 263

MARTIN, JOEL W., FRANK M. TRUESDALE, and DARRYL L. FELDER. The
megalopa stage of the Gulf stone crab, Menippe adina Williams and Felder, 1986,
with comparison of megalopae in the genus Menippe 289

SHENKER, JONATHAN M. Oceanographic associations of neustonic larval and
juvenile fishes and Dungeness crab megalopae off Oregon 299

DAGG, M. J., P. B. ORTNER, and J. AL-YAMANI. Winter-time distribution and

abundance of copepod nauplii in the northern Gulf of Mexico 319

HERRNKIND, WILLIAM F., MARK J. BUTLER IV, and RICHARD A. TANKERS-
LEY. The effects of siltation op recruitment of spiny lobsters, Panulirus
argus 331

KIRKLEY, JAMES E., and DALE E. SQUIRES. A limited information approach

for determining capital stock and investment in a fishery 339

POLACHECK, TOM. Analyses of the relationship between the distribution of
searching effort, tuna catches, and dolphin sightings within individual purse seine
cruises 351

(Continued on next page )

Seattle, Washington

1988

For sale by the Superintendent of Documents, U.S. Government Printing OfTice, Washing-
ton DC 20402 — Subscription price per year: $16.00 domestic and $20.00 foreign. Cost per
single issue: $9.00 domestic and $11.25 foreign.

Contents — Continued

WATSON, CHERYL, ROBERT E. BOURKE, and RICHARD W. BRILL. A compre-
hensive theory on the etiology of burnt tuna 367

BROWN-PETERSON, NANCY, PETER THOMAS, and CONNIE R. ARNOLD.
Reproductive biology of the spotted seatrout, Cynoscion nebulosus, in South
Texas 373

Notes

CHEN, CHE-TSUNG, TZYH-CHANG LEU, and SHOOU-JENG JOUNG. Notes
on reproduction in the scalloped hammerhead, Sphyrna lewini, in northeastern
Taiwan waters 389

COLLINS, MARK R., and CHARLES A. WENNER. Occurrence of young-of-the-
year king Scomberomorus cavalla, and Spanish, S. maculatus, materials in
commercial-type shrimp trawls along the Atlantic coast of the southeast United
States 394

DEW, C. BRAXTON. Stomach contents of commercially caught Hudson River
striped bass, Morone saxatilis , 1973-1975 397

VERNET, MARIA, JOHN R. HUNTER, and RUSSELL D. VETTER. Accumula-
tion of age pigments (lipofuscin) in two cold-water fishes 401

MULLIN, M. M., and E. R. BROOKS. Extractable lipofuscin in larval marine
fish 407

Notices: NOAA Technical Reports published during the last 6 months of 1987 . . . 416

MIGRATIONS OF COHO SALMON, ONCORHYNCHUS KISUTCH ,
DURING THEIR FIRST SUMMER IN THE OCEAN

William G. Pearcy and Joseph P. Fisher'

ABSTRACT

Marked juvenile coho salmon caught in fine-meshed purse seines during the summers of 1981-84 off
Oregon and Washington generally demonstrated northward migrations from their rivers of ocean
entrance. Northward movements in summer were preceded by southerly movements during spring,
probably caused by southerly advection. Catch rates and sizes offish caught in different months and
regions of the coast also indicated northerly movements of both yearling and subyearling coho
salmon. Despite this movement, the average catch of juvenile coho salmon per purse seine set along
the coasts of Washington and Oregon in late summer, including marked fish from the Columbia
River, was still a substantial proportion of that in May and June soon after ocean entrance, suggest-
ing that many coho did not migrate great distances. Additionally, recoveries of marked juvenile coho
salmon by sports and commercial fishermen from Alaska to California and by scientists in Alaska
were generally in the region of release. These data indicate that migrations of juvenile coho are of
limited extent during their first summer in the ocean and are not strong support for an earlier
conclusion that juvenile coho salmon from the Columbia River, Oregon, and California may form a
large proportion of the stocks of this species that migrate northward along the coastal belt in Cana-
dian and Alaskan waters each summer.

Although there has been Httle research on juve-
nile salmon during their first summer at sea, this
phase of the life history may be critical to survival
and recruitment to fisheries (Hartt 1980). High-
est ocean mortality is thought to occur early in
marine life (Foerster 1968; Parker 1968; Ricker
1976). Production (catch and escapement) of adult
coho salmon, Oncorhynchus kisutch, in the Ore-
gon Production Index (OPI) Area (from Leadbet-
ter Point, WA, to Monterey Bay, CA) is usually
accurately predicted in one year by the number of
precocious males (jacks) returning to index
streams in the previous year (Gunsolus 1978; Or-
egon Department of Fish and Wildlife 1982;
Pacific Fishery Management Council 1986).
Hence survival from jacks to adults is fairly con-
stant from year to year. Because survival rates
from smolt to adult are variable (Nickelson 1986),
however, variable year-class survival must occur
before the time that jacks return, after only a few
months in the ocean. This relationship, and the
positive correlation between coastal upwelling
and survival of OPI coho salmon (Gunsolus 1978;
Scarnecchia 1981; Nickelson 1986), strongly sug-
gest that the first few months in the ocean consti-
tute the "critical period" in determining subse-
quent adult survival.

iCollege of Oceanography, Oregon State University, Corval-
lis, OR 97331.

Between 1976 and 1985 the production of coho
salmon in the OPI area drastically declined, de-
spite large increases in the number of public and
private smolt releases (Oregon Department of
Fish and Wildlife 1982; Nickelson 1986). Reduced
upwelling and ocean productivity, perhaps cou-
pled with density-dependent mortality, is one of
the hypothesized causes for this decrease in sur-
vival (Scarnecchia 1981; Peterman and Rout-
ledge 1981; McCarl and Rettig 1983; McGie 1984;
Nickelson 1986). To understand the mechanisms
affecting survival of juvenile salmonids at sea, we
must first know where salmon reside at the time
of their high and variable mortality. Are the
smolts highly migratory, immediately leaving
local coastal waters and migrating into waters of
the Gulf of Alaska (Hartt and Dell 1986), or are
they nonmigratory, spending their early ocean
life in local coastal waters?

This paper summarizes research on the move-
ments and migrations of coho salmon during their
first summer in the ocean in the northeastern
Pacific Ocean based on purse seine catches made
mainly in coastal waters off Oregon and Wash-
ington. A few records of migrations of tagged ju-
venile (age .0)^ coho salmon were given by God-

Manuscript accepted November 1987.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.

2The numeral preceding and following the decimal indicate
the number of winters spent in fresh water and in the ocean,
respectively.

173

FISHERY BULLETIN: VOL. 86, NO. 2

frey (1965) and French et al. (1975), but by far the
most comprehensive data were provided by Hartt
(1980) and Hartt and Dell (1986). All of these
studies, however, were based on recovery of ma-
ture or maturing coho in the year following tag-
ging and on tagging in northern waters from the
Strait of Juan de Fuca to the Alaska Peninsula.
The only other studies of juvenile salmon in the
ocean off Washington and Oregon have been re-
stricted to within 24 km of the Columbia River
(Dawley et al. 1981) or to coastal waters during
1980 (Miller et al. 1983). Our 1979-85 research,
covering large areas along the coast, provides ex-
tensive and unique data on the movements of ju-
venile coho salmon during their first summer in
the ocean.

METHODS

Purse seines, our primary sampling method,
were used to sample juvenile salmonids during
1979-85 (Table 1). Cruises were in coastal waters
off Oregon in 1979-80, off Oregon and southern
Washington in 1981, and off Oregon and the en-
tire Washington coast during the summers of
1982-85 (Fig. 1). During July 1984, sets were also
made from northern California (lat. 40°32'N) to
northern Vancouver Island (50°26'N). Except for
the exploratory cruises off Oregon in 1979 and
1980, purse seine sets were usually made along
east-west transect lines (Fig. 1). Sets started at
the 37 m (20-fathom) contour, and continued at
9.3 km (5-mi) intervals farther offshore, usually
until no salmonids were captured. Repeat sets

were sometimes made when fish with missing
adipose fins were common, indicating the pres-
ence of coded wire tagged (CWT) fish. In 1985,
special sets were made in the vicinity of the Co-
lumbia River plume. Detailed sampling data are
provided in Pearcy (1984) and our cruise reports
(Wakefield et al. 1981; Fisher et al. 1982, 1983;
Fisher and Pearcy 1984, 1985).

The mesh size of the seines were the same dur-
ing all years, 32 mm (stretch), with 32 mm or
smaller mesh in the bunts of the seines. The seine
was 495 m long except in 1981 (457 m). Depths
that seines fished, sometimes measured with a
depth gauge on the lead line, varied among years
from about 20 m to 65 m (Table 1).

Generally, purse seine sets were "round hauls",
where the seiner and the skiff made a circle with
the net. The seine was fully pursed after about
one-half its length was aboard (half-purse sets).
All sets were "blind". We attempted to use sonar
on some cruises to locate concentrations of sal-
monids but were unsuccessful. Radar was some-
times used to determine the distance between the
seiner and the skiff when a semicircle was made
with the net. Each round haul encompassed about
17,000 m2 (1981) or 19,000 m2 (1979-85). To de-
termine the direction of movement of fish, eight
"half-round" hauls, or "semicircular" sets, were
made in 1979, where the entire net formed an
open semicircle. Paired sets were made in close
succession, with sets open in a northern and a
southern direction, at four locations. The seine
was open for the same duration (15-45 minutes,
depending on location) in each paired set while

Table 1. — Summary of number of purse seine sets and latitudinal range of sampling, 1979-85.

Dates of

No. of

Purse

seine

Lengtfi

Depthi

Year

cruises

setsi

Latitudinal range of sampling

(m)

1979

18-29 June

Cape Disappointment to Cape Arago

46°20'

-43°18'

495

1980

20-28 June

Cape Disappointment to Alsea River

46°20'

-44°30'

495

1981

16-25 May

Willapa Bay to Alsea River

46°35'

-44°25'

495

9-18 June

Willapa Bay to Cut Creek

46°35'

-43°1 1 '

495

9-19 July

Willapa Bay to Alsea River

46°35'

-44°25'

457

8-19 Aug.

Willapa Bay to Cut Creek

46°36'

-44°1 1 '

457

1982

19 May-2 June

Waatch Point to Siuslaw River

48°2r

-44°00'

495

265

7-22 June

Quinault River to Yachiats

47°21 '

-44°20'

495

265

4-14 Sept.

Quinault River to Yachats

47°20'

-44°19'

495

265

1983

16-27 May

Waatcfi Point to Yachats

48°2r

-44°20'

495

249

9-27 June

Waatcfi Point to Four Mile Creek

48°20'

-43°00'

495

249

15-24 Sept.

Waatch) Point to Coos Bay

48°20'

-43°28'

495

249

1984

4-20 June

Waatcfi Point to Coos Bay

48''20'

-43°27'

495

249

9 July-3 Aug.

Winter Harbor, B.C. to False Cape. CA

50°26'

-40°32'

495

249

1-15 Sept.

Waatcfi Point to Siuslaw River

48°20'

-44°00'

495

249

1985

29 May-25 June

112

Sea Lion Rock to Coos Bay

48°00'

-43°27'

495

225

^Quantitative sets.
^Measured with depth guage.

174

PEARCY and FISHER: MIGRATIONS OF COHO SALMON

Figure l. — Locations of purse seine transects
off the Oregon and Washington coasts.

WAATCH POINT
SEA LION ROCK
DESTRUCTION ISLAND

QUINAULT RIVER

GRAYS HARBOR

WILLAPA BAY

CAPE DISAPPOINTMENT

SEASIDE

NEHALEM BEACH

CAPE LOOKOUT

WECOMA BEACH

YAOUINA HEAD

YACHATS
SIUSLAW RIVER

COOS BAY
FOUR MILE CREEK

125

124"

the vessel and skiff towed the seine only fast
enough to maintain a constant net opening.

The purse seine catches were either dip-netted
from the bunt of the seine while it was alongside
the vessel, brailed aboard, or hauled aboard in the
bunt, depending on the composition and size of
the catch.

In 1979 and 1980, juvenile salmon preserved in
formalin were identified ashore. In 1981-85, ju-
venile salmon were identified to Species at sea,
fork length (FL) was measured to the nearest mil-

limeter and then they were individually wrapped
in labelled plastic bags and frozen. All salmonids
with marks or missing adipose fins were frozen.
When large numbers of juvenile salmonids were
caught in a set, most unmarked fish were re-
leased after they were measured.

In order to increase the numbers of marked fish
released into our study area we marked about 1.5
million coho smolts in 1981 and 835,000 in 1982
using fluorescent pigment propelled by com-
pressed air (see Phinney et al. 1967) prior to their

175

FISHERY BULLETIN: VOL. 86. NO. 2

transport from Oregon Aqua-Foods, Inc. (OAF)
hatchery to their ocean release facilities at
Yaquina Bay or Coos Bay, OR.

In the laboratory ashore, species identifications
were confirmed and individuals remeasured and
reexamined for both fluorescent marks (under ul-
traviolet light 1981-82) and missing adipose fins
or other marks (1979-85). Coded-wire tags from
the heads of salmonids with missing adipose fins
were decoded by personnel from the Oregon De-
partment of Fish and Wildlife, Clackamas Labo-
ratory.

Juvenile or age .0 (first year in the ocean) coho
salmon were distinguished from adult or age .1
(second year in the ocean) coho salmon by exami-
nation of size-frequency histograms and scales.
The division between age .0 and .1 coho pro-
gressed from approximately 300 to 420 mm FL
from May to September, in most years. Most coho
salmon migrated to the ocean a little over one
year after hatching (age 1.0), but OAF released
large numbers of subyearling (age 0.0) smolts
into Yaquina Bay and Coos Bay. These two age
groups of smolts were distinguished by the radial
distance to the 21st circulus on scales removed
from the preferred area (Clutter and Whitesel
1956) of the fish. The accuracy of this method for
distinguishing known age 0.0 and 1.0 fish was
approximately 85-90%. In the years 1981-85,
scales from 52% of the 4,222 juvenile coho sam-
pled were analyzed. The estimated numbers of
age 0.0 and 1.0 fish represented in different
geographic areas and cruises were then extrapo-
lated from their proportions in each 10 mm length
interval.

Distances traveled and movement rates were
estimated from actual distances between sites of
release and entry into the ocean, and from
straight-line distances between ocean entry and
recapture locations for CWT or fluorescent
marked juvenile coho salmon that were recovered
in the ocean within 10 days of release. These dis-
tances and swimming speeds are minimal esti-
mates.

In addition to purse seining, fine-meshed
monofilament gill nets were used off the Oregon
coast (ca. lat. 45°00'N, long. 124°21'W) during 24
and 25 July 1985, from the training vessel Oshoro
Maru , to determine depth and direction of swim-
ming of juvenile salmonids. Surface and subsur-
face nets were used. The surface gill nets were
2,050 m long, and fished from depths of 0-6 m
with 11 mesh sizes ranging from 29 to 121 mm
(stretch). The subsurface nets were 500 m long

and consisted of four mesh sizes ranging from 29
to 42 mm; they were suspended below large (300-
400 mm) mesh to fish at depths of 5-12 m. Four
sets were made in an east-west direction with
soak-times of about 4-9 hours. As the gill nets
were hauled, the direction that each juvenile
salmonid was heading when caught, and its depth
in the net (upper, middle, or lower section) were
noted. Each juvenile salmonid was given a con-
secutive number and frozen for later identifica-
tion. Comparisons of catch rates in the surface
and subsurface nets were based on equal lengths
of the four mesh sizes of the subsurface net, stan-
dardized to 10 hours of fishing time.

Information on the location of landings of
marked juvenile coho by commercial and sports
fishermen was provided by the Pacific Marine
Fisheries Commission (PMFC), (PMFC 1980,
1981, 1984a, b, c, 1985a, b), from lists of non-
standard recoveries (Johnson PMFC unpubl.
data), and from state agencies. The actual num-
bers of tagged fish, and the total numbers of
tagged fish estimated from the proportions of the
catch sampled are reported.

To determine if juvenile coho salmon were sex-
ually precocious "jacks", we examined testes from
542 juvenile males caught in July 1981 and 1984,
in August 1981, and in September 1982, 1983,
and 1984. All developed and some undeveloped
testes (ribbonlike, with no thickening), as deter-
mined by visual inspection, were weighed (123)
and gonadal-somatic indices (GSI = testes wt./
body wt. X 100) were determined.

RESULTS AND DISCUSSION

Swimming Direction

Of the 106 juvenile coho salmon captured dur-
ing June in paired, half-round purse seine sets,
all but two were in the sets open to the south
(Table 2). This suggests that juvenile coho salmon

Table 2. — Catches of coho salmon In semicircular
purse seine sets open to the south (S) and north (N) off
Oregon, June 1979.

Location

Km
offshore

Age
S

.0
N

Age .1
S N

Clatsop Spit
Clatsop Spit
Clatsop Spit
Newport

12.6

18.5

9.4

57
37

7
3

0
0
0
2

2 6

15 37

6 8

0 9

98%

28% 72%

176

PEARCY and FISHER: MIGRATIONS OF COHO SALMON

were swimming to the north during this sampHng
period. Maturing fish over 300 mm FL (age .1
echo) showed the opposite trend.

Miller et al. (1983) made several hundred
paired purse seine sets open to the south and
north during three cruises off the northern
Oregon-southern Washington coasts in 1980.
During their May-June cruise, they caught 76%
of the juvenile coho salmon, 80% of the chinook
salmon, and almost all the steelhead trout in sets
open to the south, indicating northward move-
ment. We note a positive relation between the
proportion of juvenile salmon caught in south-
facing sets in their three cruises and strength of
upwelling during these cruises (mean daily
Bakun indices of 52, 39, and 19 m^ s"! 100 m'^
coastline at 45°N, 125°W in May, July, and Au-
gust, respectively (Mason and Bakun 1986)), sug-
gesting that surface currents to the south result-
ing from Ekman transport may be cues for
orientation of salmon smolts.

Hartt (1980) and Hartt and Dell (1986) found
that 83% of the combined species of juvenile
salmonids caught in 19 paired purse sets along
the coast from Cape Flattery, WA to Yakutat, AK
were caught in sets held open to the southeast and
only 17% in sets open to the northwest and north.
They concluded that juvenile salmonids tended to
migrate in a northwest direction along the coast
during July-September.

Of the 100 juvenile coho salmon (135-315 mm
FL) caught in the gill nets set in an east-west
direction off the Oregon coast in July 1985, 90
coho were caught as they approached the south-
ern face of the gill net (heading north) and 10 in
the northern face (heading south). Jaenicke et al.
(1984) reported that 63% of the juvenile coho
caught in a surface gill net fished off southeastern
Alaska in July moved north at night, but only 6%
moved to the north during the day.

Available data indicate that most juvenile coho
salmon caught off Oregon and southern Washing-
ton, as well as juveniles farther to the north, are
predominantely swimming in a northerly direc-
tion during summer months.

Depth Distribution

One-half of the juvenile coho salmon caught in
gill nets set off the Oregon Coast in 1985 were in
the upper 2 m of the surface gill net (Table 3).
Catches in the surface net exceeded those in the
subsurface net, except for the last set that fished
during daylight hours, indicating that juvenile

coho salmon were most common in the upper 4 m
of the water column.

Other information on the vertical distribution
of maturing coho and other species of salmon
caught in gill nets or with longlines in oceanic
waters also indicates that they usually swim near
the surface, between 0 and 20 m (Manzer 1964;
Godfrey 1965; Godfrey et al. 1975). Machidori
(1966), for example, fished gill nets from the sur-
face to 50 m and caught 79% of the coho salmon in
the upper 10 m of the gill net. Although catches in
gill nets at different depths may be biased by ver-
tical differences in avoidance reactions to the net
or swimming speeds (Hartt 1975), acoustical
methods have also shown that salmon are usually
distributed near the surface (Susuki and Sonoda
1972; Lord et al. 1976). We conclude that most
juvenile coho salmon in coastal waters and ma-
turing coho in oceanic waters reside at depths
above 20 m, the minimum depth that our purse
seine fished. We recognize, however, that matur-
ing coho and other species of salmon may feed in
deeper water. Some salmon (including coho
salmon) caught in surface gill nets in the oceanic
waters of the Gulf of Alaska contained prey in
their stomachs characteristic of mesopelagic
depths (200-1,000 m), suggesting that some indi-
viduals may feed well below the thermocline
(Pearcy et al. in press).

Table 3. — Catches of juvenile coho salmon in four gill net sets
in 50 m lengths of 29, 33, 37, and 42 mm mesh at different
depths and times 24-25 July 1985, each set adjusted to 10-h
fishing duration.

Depth

in meters

Surface net

Subsurface net

Times of set

0-2

2-4

4-6

5-7

7-9

9-12

0913-1702
2001-0104
0248-0701
0830-1737

2.5
25.0
42.8

1.1

3.8
11.5
2.4
0

0
0

2.4
0

1.3
0

9.5
6.6

1.3
0

16.6
7.7

1.9
0
5.5

Total catch

Percent of
total catch

51
50.3

16
12.5

1
1.7

11
12.3

15
18.0

6
5.2

North-South Trends in Catch per Set
and Sizes of Juvenile Coho

Variations in the average catches and sizes of
juvenile coho salmon in purse seine sets in differ-
ent regions of the Oregon-Washington coast dur-
ing the summer provide indirect evidence for
north-south coastal movements. Histograms
showing average catches per set for 10 mm size

177

FISHERY BULLETIN: VOL. 86, NO. 2

groups of juvenile coho, classified as age 0.0 or 1.0
from scale analysis, are shown in Figure 2, for
1981-84 in three regions: (A) Cape Flattery, WA
to Grays Harbor, WA (called Washington), (B)
Willapa Bay, WA to Nehalem Bay, OR (Columbia
River region), and (C) Cape Lookout, OR to Coos
Bay, OR (Oregon) (Fig. 1). In May of 1981, 1982,
and 1983, average catch per set of yearling (age
1.0) coho generally decreased from the southern
to the northern regions. Catches v^ere highest off
Oregon (Area C) or the Columbia River (Area B)
and lowest off Washington (Area A) in May of
1982 and 1983. This trend was reversed later in
the summer. In June of 1981, 1983, and 1984,
lowest catches were found in the Oregon region.
By August or September 1981-84, highest
catches consistently occurred off the Columbia
River or Washington and few yearling fish were
caught off Oregon. These shifts in abundance sug-
gest a northerly movement of age 1.0 smolts dur-

ing the summer. Highest catch rates occurred in
May and June of 1981 and 1982 when an average
of over 10 juvenile coho salmon were caught in
most sets.

Subyearling or age 0.0 coho salmon released
from private facilities at Yaquina and Coos Bays
provide more direct evidence on movements. Sub-
yearling coho salmon clearly demonstrated
northward dispersal. They were most common in
our catches of July 1981 and September 1982,
1983, and 1984 (Fig. 2). They were apparently
more numerous than age 1.0 coho salmon in the
Oregon region during June-August 1981 and
September 1982, and in the Oregon and Columbia
River regions in September 1983. The catches and
proportions of age 0.0 coho salmon increased off
Oregon during the summer because they were
released from hatcheries later in the summer
than yearling coho salmon. They were found in
the most northern region sampled late in the

1981

MAY

/SET

-TV.

X 6
O

^ 5
" 4

wW.ljb.nv ,..y

20 30

JUNE

JULY

....Mljlito^

i^i,i^»i

AUGUST

11*11(^1

FORK LENGTH (cm)

Figure 2. — Catch per purse seine set of age 1.0 (open) and age 0.0 (shaded) juvenile coho by 10 mm length groups during different
months, 1981-84, for three regions of the Oregon-Washington coast: (A) Cape Flattery to Grays Harbor, WA, (B) Willapa Bay, WA
to Nehalem Beach, OR, (C) Cape Lookout to California.

178

PEARCY and FISHER: MIGRATIONS OF COHO SALMON

summer of all years. Their abundance, as that of
age 1.0 coho salmon, also increased during the
summer off Washington where they intermingled
with age 1.0 coho salmon. Because the Oregon
region included the release locations of all age 0.0
coho salmon, our figures provide no information
on southward movements of these fish.

The mean lengths of both age 0.0 and 1.0 coho
salmon increased from the southern to the north-
ern areas during most months. Larger age 0.0 and
1.0 coho salmon were caught off Washington than

Oregon during the late summer, 1981-84 (Fig. 2),
providing corroborative evidence for northward
migration of coho salmon. Larger, and pre-
sumably older, fish were found farther to the
north than smaller fish.

Despite northward movements, many yearling
coho salmon did not migrate out of the sampling
area, but remained in coastal waters off Oregon
and Washington during the entire summer. Mean
coastwide catch per set of yearling coho salmon in
August 1981 and September 1982, 1983, and 1984

1982

I-

LJ
if)

6
5
4

21-

6
5
4
3
21-

51- J
4

MAY

TTfllTTTTTTtlll l^[

JUNE

t^t^ri^iM^rT^f rr

SEPTEMBER

¥.«

¥^

iWiT'i I

T^T*T^ IllllTTTfftfl

^ .UJJlmnifii

^JMpMk

T^P»

<L¥

»'ri^i f I

IM^

.p..Q.,,,n,f:T;i, ,f77.
20 30 10 20 30 10 20 30 40

FORK LENGTH (cm)

Figure 2. — Continued.

179

FISHERY BULLETIN: VOL. 86, NO. 2

1983

6
5

21-

MAY

■ I nfXb. .yi iGv

JUNE

■ ■f;w{ttlkbp.

SEPTEMBER

fffllfflVIIVI

^T^f^^WP^^^'^^^''''' T y T

D ■

5 .

3 ■

1- r

,,dJ,U.i.b|

i*t f »'f^^*yT^ I

■

•

■

JOCII.

-•

•

10 20

30 10 20 30 10

FORK LENGTH (cm)

Figure 2.— Continued.

1^1 I I I t [ I I I I [I

20 30 40

was 61%, 42%, 81%, and 77% respectively, of that
in June of the same years. In September 1983, a
strong El Nino year (Pearcy and Schoener 1987),
almost all yearling juvenile coho salmon were
caught at the extreme northern transect off Cape
Flattery, but in August or September of other
years they were more evenly distributed off the
Columbia River and the Washington coast.

In July 1984, we sampled both north (15 sets off
the west coast of Vancouver Island) and south (5

sets off northern California) of Washington and
Oregon. Catches per set of yearling and subyear-
ling juvenile coho salmon were higher off the
Columbia River (5.1), the area of greatest smolt
production, and off Washington (3.5) than off
Vancouver Island (1.8), Oregon (1.7), and Califor-
nia (1.2). This shows that as late as July juvenile
coho salmon occurred in coastal waters of all re-
gions and were not concentrated off Vancouver
Island or California.

180

PEARCY and FISHER: MIGRATIONS OF COHO SALMON

1984

H
UJ
C/)

5
4

3|-

2
I

6
5

JUNE

rftflMru,

JULY

1^1 1 iiTMpiiiriii'JiiiriVpi^i I 'i -rf^'^^

SEPTEMBER

Ml^l I p ■ I I

pir^Tr7i.»,r>|

.oM^

rp.iP, ■ ,

1 — f^
30 10

-r-

10 20 30 10 20 30 10 20 30

FORK LENGTH (cm)

Figure 2.— Continued.

The Question of "Jacks"

Are the juvenile coho salmon off Oregon and
Washington in late summer relatively nonmigra-
tory because they are sexually precocious? GSI
were almost always <0.1% for those fish visually
classified as "undeveloped". GSI's from fish with
"developed" testes ranged from 0.29?^ to 1.0% in
July 1981 and 1984 in fish >250 mm FL; from
0.3% to 5.6% in August 1981, mostly in fish >280
mm FL; and from 2.4% to 6.6% (except for one
value at 0.6) in September 1982, 1983, and 1984
in fish >300 mm FL (Fig. 3). In August 1981, and

clearly in September 1982, 1983, and 1984, two
distinct groups of fish were evident: "jacks", with
developing testes (GSI >0.3% August or GSI
>2.0% September), and "nonjacks", which
showed no development (GSI <0.1%).

The total numbers of jacks and nonjacks in each
50 mm length group were estimated for the catch
during August 1981 and September 1982, 1983,
and 1984 from the ratio of jacks and nonjacks in
the sample (Table 4). Only 8.4%, 4.8%, 5.2%, and
2.8% of all juvenile fish (male and female) were
"jacks" in August 1981 and September 1982,
1983, and 1984, respectively. However jacks com-

181

FISHERY BULLETIN: VOL, 86, NO. 2

LU
Q

-z.
o

D 1981 n= 12
A 1984 n=9

A A A

,D ^ * °

A
1 1 1 1 1

1 1 1

6
5
4
3
2
I

0
9
8
7
6
5
4
3
2
I
0

D 1981 n=55

D
D

°D

° a

.iiniLtniiM imiiiiiriijiiinjii .Viiiiiim ii. i

i 1

■ 1 t 1 1

1 1

0 1982 n=9

• 1983 n=27
A 1984 n=ll

•

A
^ A O

^ •a

• 0°

, M . , r . m^m

• m^}A

^AaftMB . an . . .

, 1 , • ,

FORK LENGTH (cm)

Figure 3. — Gonadal-somatic index (testis wt/total body wt) x 100 of juvenile coho salmon vs.
length of fish for July, August, and September 1981-84. Only data for those testes actually
weighed are shown.

prised a higher percentage of fish larger than 300
mm FL in August 1981 and larger than 350 mm
FL in September 1982, 1983, and 1984. These
results indicate that most juvenile coho salmon
caught off Oregon and Washington were not sexu-
ally precocious. Thus, the relatively large catches

of juvenile coho salmon in late summer are ex-
plained by lack of strong migrational tendencies
of juvenile coho salmon in this region and not by
a high proportion of precocious "jacks" that re-
sided in this region as a prelude to re-entry of
streams for spawning.

182

PEARCY and FISHER: MIGRATIONS OF COHO SALMON

Table 4. — The percent of coho salmon jacks and males, by length groups, in the total catch,

Augjst 1981 and September 1982-84.

Fork

Number

Est.

Est. %

length

of fish

Total

total

jacks

Date

(mm)

examined

males

jacks'

catch

jacks

of total

Aug. 1981

<200

60.6

111

201-250

60.0

115

1.7

251-300

47.3

104

7.7

301-350

81.3

27.3

351-420

80.0

78.9

Total

215

59.5

371

8.4

Sept. 1982

<200

55.4

125

201-250

76.2

251-300

63.6

109

301-350

69.7

9.3

351-420

53.8

39.3

Total

145

62.8

413

4.8

Sept. 1983

<200

62.5

201-250

47.8

251-300

56.5

301-350

61.5

10.4

351-420

66.7

Total

103

58.3

194

5.2

Sept. 1984

<200

33.3

201-250

47.4

251-300

50.0

128

2.3

301-350

55.6

6.5

351-420

33.3

28.6

Total

167

49.1

250

2.8

'Jack Is defined as a male whose testes wt. /total body wt. x 100 > 0.3% in August and >2.0% in Septem-
ber.

Movements of Marked Fish

Direct evidence of movements of juvenile coho
salmon was obtained from capture of marked fish
containing coded wire tags or marked with fluo-
rescent pigment. The generalized pattern of
movements that emerges for 1981-85 is an initial
movement of most juvenile coho salmon to the
south soon after ocean entry in May and June and
then a reversal of movement with most fish mi-
grating to the north by August and September
(Figs. 4-8). These trends are discussed for fish
originating from the Columbia River, Oregon
coastal, Washington coastal, and private hatch-
eries.

Columbia River

Juvenile coho salmon originating from hatch-
eries on the Columbia River were usually recov-
ered south of the Columbia River in May. This
trend was especially obvious in May 1982 when
all 22 marked fish which were recovered moved
south, some as far as 175 km (Fig. 5). In May
1981, all but one of 14 marked Columbia River
fish were caught to the south, three as far as 180

and 204 km (Fig. 4). In May 1983, all four fish
were taken south of the mouth of the Columbia
River (Fig. 6).

During June and July of all years, marked Co-
lumbia River coho salmon were recovered in
nearly equal proportions both north and south of
the river mouth, except in June 1982 when 15 of
17 fish were found to the south (Figs. 4—8). By
September, all marked Columbia River coho
salmon were captured north of the river, includ-
ing fish captured off the Quinault River in Sep-
tember 1982 and off Cape Flattery in September
1984. Fish were also caught close to the mouth of
the Columbia River in July, August, and Septem-
ber, indicating that some marked juvenile coho
salmon did not undertake extensive migrations at
sea.

In two sets on the Wecoma Beach Transect on 1
June 1982 we caught 17 marked juvenile coho
salmon released between 30 April and 6 May
from six hatcheries on the Columbia River. Based
on downstream migration rates for these groups
to Jones Beach (Dawley et al. 1985) and assuming
similar rates from Jones Beach to the ocean, these
fish had probably been in the ocean for <10 days
before recapture. This indicates that some juve-

183

FISHERY BULLETIN: VOL. 86, NO. 2

AUG JULY JUNE

1981

Figure 4. — North-south movements of marked juvenile echo salmon captured
in purse seines, May-August 1981. The width of the lines are approximately
proportional to the number offish. Numbers at end of arrows indicate number
offish captured. Arrows without numbers and thin lines represent single fish.
Inshore-offshore movements are not shown. Dashed lines indicate latitudinal
extent of sampling.

nile coho salmon released from hatcheries at
about the same time tended to stay together dur-
ing their downstream migration in the Columbia
River and during early residency in the ocean.

Oregon Public Coastal Hatcheries

We captured marked fish originating ft-om pub-
lic Oregon coastal hatcheries both north and
south of the latitude of ocean entrance in May. A
total of five fish were found to the south, while 11

fish were found to the north in May (Figs. 4-8).
With the exception of one coho salmon from the
Umpqua River in June 1983 and two from the
Rogue River in July 1984 (Figs. 6, 7), the other 25
fish taken after May were captured north of
where they entered the ocean. Northerly move-
ments into Washington waters occurred by June
1983 and 1985 (Figs. 6, 8).

The southward movements of two juvenile coho
salmon released from the Rogue River (south of
Cape Blanco) and captured off northern Califor-

184

PEARCY and FISHER: MIGRATIONS OF COHO SALMON

I I I I Ku I I I I 1 IV q

-CAPE FLATTERY

SEPT JUNEV MAY
1982

-rCAPE FLATTERY

Figure 5. — North-south movements of marked juvenile coho
salmon captured in purse seines, May, June, and September
1982. The width of the lines are approximately proportional to
the number of fish. Numbers at end of arrows indicate number
offish captured. Arrows without numbers and thin lines repre-
sent single fish. Inshore-offshore movements are not shown.
Dashed lines indicate latitudinal extent of sampling.

JUNE
1983

Figure 6. — North-south movements of marked juvenile coho
salmon captured in purse seines, May, June, and September
1983. The width of the lines are approximately proportional to
the number of fish. Numbers at end of eirrows indicate number
offish captured. Arrows without numbers and thin lines repre-
sent single fish. Inshore-offshore movements are not shown.
Dashed lines indicate latitudinal extent of sampling.

nia during July 1984 are notable (Fig. 7). They
were captured in our only cruise into California
waters and represent the only recoveries of
marked juvenile coho salmon originating from
hatcheries south of Cape Blanco in all six years of
sampling. Although ocean sampling was limited
south of Coos Bay, if juvenile coho salmon from
southern Oregon and northern California hatch-
eries had migrated north of Coos Bay, we

would expect them to be represented in our
catches. The fact that they were not caught in this
northern region, but two were caught after swim-
ming to the south, suggests that juvenile coho
salmon originating in streams south of Cape
Blanco may migrate south, possibly occupying
the region of intense coastal upwelling off north-
ern California during their first summer in the
ocean. The catch of over 70% of the adult coho

185

FISHERY BULLETIN: VOL. 86, NO. 2

SEPT

UMPQUA
RIVER

JUNE 10-25

.coos BAY

MAY 29 - JUNE 5
1985

CAPE BLANCO
I I I I I

— 43=

JULY
1984

Figure 8. — North-south movements of marked juvenile coho
salmon captured in purse seines, 29 May-5 June and 10-25
June. The width of the lines are approximately proportional to
the number of fish. Numbers at end of arrows indicate number
offish captured. Arrows without numbers and thin lines repre-
sent single fish. Inshore-offshore movements are not shown.
Dashed lines indicate latitudinal extent of sampling.

Figure 7. — North-south movements of marked juvenile
coho salmon captured in purse seines, June-September
1984. The width of the lines are approximately propor-
tional to the number of fish. Numbers at end of arrows
indicate number of fish captured. Arrows without num-
bers and thin lines represent single fish. Inshore-
offshore movements are not shown. Dashed lines indi-
cate latitudinal extent of sampling.

186

PEARCY and FISHER: MIGRATIONS OF COHO SALMON

salmon from the Rogue River in the troll fishery
off California (R. Garrison'^) is further evidence
for a southern distribution of this stock. Marked
juvenile coho salmon from California hatcheries
were reported from the sports fishery off southern
Oregon, however, as will be shown later.

Washington Coastal Hatcheries

Juvenile coho salmon from Washington public
hatcheries demonstrated southerly movements,
sometimes into Oregon waters, during May 1981
and 1982 (Figs. 4, 5). During June 1982, 1984,
and 1985, Washington coastal fish were found
both north and south of ocean entry. Except for
one fish that moved north in September 1984, no
Washington coastal fish were taken in August or
September of other years, suggesting that most
Washington fish may have migrated out of our
sampling area by late summer.

Oregon Private Hatcheries

All marked juvenile coho salmon originating
from Yaquina and Coos Bays that we captured at
sea were from private hatcheries. Those from
Yaquina Bay were mainly age 0.0 smolts from
OAF, those from Coos Bay were either age 1.0
smolts from Anadromous, Inc. or age 0.0 smolts
from OAF. Forty-one recoveries of marked OAF
fish released from Yaquina Bay were caught to
the north while only 4 were to the south of
Yaquina Bay. In general, more juvenile coho
salmon from Yaquina Bay were captured in late
than early summer, and distances traveled to the
north were largest (up to 413 km) for fish caught
in later summer (Figs. 4-8). All recoveries of
marked Anadromous, Inc. and OAF fish released
into Coos Bay were to the north in all years.
Large northerly movements were demonstrated
by some of these fish (Figs. 4, 6, 8). Since our
sampling in the ocean usually did not extend
south of Coos Bay, recoveries of these fish are
biased to the north; however, strong northward
movements of these stocks were indicated.

Rates of Movement

Recoveries of marked juvenile coho salmon in
the ocean provided information on the minimum
rates of movement from hatchery release to cap-

3R. Garrison, Oregon Department of Fish and Wildlife, Cor-
vallis, OR 97330, pers. commun. December 1983.

ture in the ocean. Some fish moved rapidly
through estuaries into the ocean. We captured
some tag-groups in the ocean only a few days after
the median date of capture at Jones Beach (75 km
from the ocean) as reported by Dawley et al.
(1985): 5 fish after 2 days in 1981, 6 fish after
3-11 days in 1981, 8 fish after 1-14 days in 1982,
and 5 fish after 3-8 days in 1983. Dawley et al.
(1986) found average rates of movement of 14-23
km d"^ for marked groups of coho smolts from
areas of release on the Columbia River to river
km 75, and rates of movement that were 40%
faster from river km 75 to the lower Columbia
River estuary and to the ocean plume. These re-
coveries and those reported by Miller et al. (1983)
for yearling chinook salmon and steelhead trout
indicate rapid movements of juvenile salmonids
of over 20 km d^^ through the Columbia River
estuary.

Some juvenile coho salmon released from
Yaquina Bay and Coos Bay also demonstrated
rapid movements into and in the ocean, e.g., 17
Anadromous, Inc. fish were captured 11 km north
of Coos Bay only two days after release in June
1983 (Table 5). Myers (1980) described an expo-
nential decrease in the catches of juvenile coho
salmon released from the OAF facility into
Yaquina Bay; about one-half the fish from
marked groups remaining in the bay after 1.7-9.0
days. Juvenile coho salmon apparently emigrate
rapidly from estuaries into the ocean.

Some of the marked fish recovered within 10
days of release demonstrated rapid movements
down-rivers or in the ocean. Twenty-four fish
traversed 10 km d~^ or more largely in the ocean,
in both north and south directions (Table 5). Four
fish released in bays or in the ocean moved over
18.8 km d~^ Two of these swam to the north,
presumably against coastal currents. These
speeds are equivalent to 1.7 body lengths (BL) per
second or more and suggest that some fish must
be traveling nearly straight courses during
24-h days, since 1-3 BL s"^ are thought to be
optimal cruising speeds for small (<20 cm)
pelagic fishes (Weihs 1973; Ware 1978). These
maximum rates of movement for purse seine
caught juvenile coho salmon are similar to
those estimated by Hartt (1980) and Hartt and
Dell (1986) for tagged sockeye salmon during
their first summer in the ocean: 14-27 km d"^
for 11 Fraser River fish and 6-14 km d"^ for
10 Skeena River fish that were between about
8 and 23 cm in length during the migration
period.

187

FISHERY BULLETIN: VOL. 86. NO. 2

Table 5. — Release Information and mean travel speeds in kilometers per day and body length (BL)
per second for CWT and fluorescent-pigment marked age .0 coho salmon recovered in the ocean
within 10 days of release. CR ^ Columbia River: OAF = Oregon Aqua Foods, Inc., (OAF Yaquina
is 3.7 km from ocean; OAF Coos Is 14 km from the ocean), Anad. - Anadromous Inc. (7.4 km from
the ocean).

Median

Direction

Release

Date

days to

Mean

Year

No.

location

released

recovery

km/d

BUS

movement

1981

Big Cr. (CR)

8 June

153

22.2

1.7

Tanner Cr. (CR)

6 July

138

41.0

3.4

OAF Yaquina

11 May

140

4.0

0.3

OAF Yaquina

12 June

124

14.5

1.4

Anad. Coos

8 June

179

2.3

0.2

OAF Yaquina

10-15 June

124

3.7

0.3

OAF Yaquina

10-15 June

123

10.2

1.0

OAF Yaquina

10-15 June

126

18.8

1.7

OAF Coos

5-9 June

122

1.9

0.2

OAF Yaquina

10-15 June

124

24.0

2.2

OAF Coos

5-9 June

136

20.1

1.7

1983

Anad. Coos

26 June

156

11.2

0.8

1984

OAF-Offshore

7 June

143

11.3

0.9

1985

Tongue Pt. (CR)

24 IVIay

151

9.5

0.7

Offshore-22 km

30 l^ay

0.9

134

22.0

1.9

Naselle River

20 l^ay

143

7.8

0.6

Cowlitz River

31 IVIay-6 June

2.5

144

88.0

7.1

Effects of Ocean Currents

The tendency for juvenile coho salmon to move
to the south early in the summer and to the north
later in the summer (Figs. 4—8) may be related to
advection of water and the size, orientation, and
swimming speeds of fish. Generally, surface cur-
rents are to the south off Oregon and Washington
in the early summer owing to prevailing north-
westerly winds (Hickey 1979; Huyer 1983).
Southward flow averaging 17-34 km d"^ (Huyer
et al. 1975, 1979) has been measured near the
surface. May and June are periods of peak outflow
of the Columbia River, so fish entering the ocean
at this time, especially in the Columbia River
plume, may be displaced to the south by advection
of surface waters. Southward flow is at a maxi-
mum in the coastal jet which is strongest (—22 km
d~^) during the spring about 15-20 km from
shore (Kundu and Allen 1976; Huyer et al. 1979).
Since currents can be equivalent to 1.7 BL s"^ for
a 15 cm smolt, advection alone could explain the
southward movement of most marked Columbia
River fish during May and June but not the rapid
northward movement of fish during this period
(see Figures 4-7).

Coastal Oregon fish were often found to the
north in May and June, but these fish were usu-
ally substantially larger and generally released
much earlier in the spring than Columbia River

fish, and were presumably better able to swim
against the current. Later in the summer when
Columbia River hatchery fish had grown larger,
movement was also predominately northward. In
August and September southward velocities of
surface coastal currents are diminished and the
mean may be near zero (Huyer et al. 1975).
Northward movements during the summer off
Oregon and Washington generally cannot be ex-
plained by passive drift and in most years must
entail active, oriented swimming.

The northern El Nino of 1982-83, which had
severe effects on the growth and survival of adult
and jack coho salmon (Pearcy et al. 1985; Johnson
1984; Pearcy and Schooner 1987; Fisher and
Pearcy in press), also appeared to affect the distri-
bution of juvenile coho salmon. During Septem-
ber of 1983 nearly all the seine-captured juvenile
coho were taken along our northernmost transect,
off Cape Flattery, WA (Fig. 2). In other years
juvenile coho salmon during late summer were
common and more equally distributed from the
Columbia River northward. In the summer of
1983 juvenile coho salmon may have moved far-
ther north, or more likely those to the south may
have experienced higher mortality, as a result of
northerly currents (Huyer and Smith 1985),
warm temperatures and low productivity (Pearcy
et al. 1985; Chung 1985) that prevailed off Ore-
gon.

188

PEARCY and FISHER: MIGRATIONS OF COHO SALMON

Recoveries of CWT Juvenile Fish
in Ocean Fisheries

Data on ocean location of landings of juvenile
CWT coho salmon reported in sports and commer-
cial fisheries in the ocean along the west coast of
North America, 1977-83, provide valuable infor-
mation on ocean migrations of marked fish, al-
though these data are biased by differences in
legal minimum sizes, time and duration of open
season, and effort in the different regions. The
summary of all years shows that, except for Cali-
fornia, most of the recoveries of juvenile coho
salmon during their first summer in the ocean
were in the general region of their ocean entry
location (Table 6). Both the actual number offish
reported and the estimated total numbers (in
parentheses) support our earlier conclusion that
many juvenile coho salmon off Oregon and Wash-
ington are not highly migratory. All (20) of the
actual recoveries of marked juvenile coho salmon
that were released in southeastern Alaskan
waters were from southeastern Alaska. Ninety-
seven percent of the recoveries of marked fish
released in British Columbia waters were landed
in British Columbia; only two were landed in
Alaska. Most (86%) marked juvenile fish from
Puget Sound hatcheries were caught in the

Sound, and more were recovered in British Co-
lumbia fisheries (13%) than in coastal Washing-
ton fisheries (<1%), probably due to the smaller
size limits for coho in British Columbia as well as
migratory patterns. Half of the actual numbers of
recoveries of juvenile coho salmon liberated into
Washington coastal waters were landed in
Washington coastal ports, followed by British Co-
lumbia (29%) and the Columbia River region
(17%). Only one was landed in Alaska and two in
Oregon ports (Garibaldi and south). Juvenile coho
salmon originating from Columbia River hatch-
eries had a broader distribution of recoveries in
other regions. Only 40% of Columbia River fish
were caught in this region, 41% were taken in
northern regions, including two (1%) in Alaska.
The remaining 19% were captured off Oregon.
The majority (73%) of Oregon coastal fish were
recovered off Oregon, followed by the Columbia
River region, Washington coast, and British Co-
lumbia. None was reported from Alaska, but 10
(2%) were from California ports. All marked Cali-
fornia fish were recovered from the Columbia
River region and farther south. Most (87%) were
landed in Oregon. The few recoveries of Califor-
nia fish off California is undoubtedly influenced
by the larger size limits in this than other fish-
eries.

Table 6. — Recoveries of coded wire tagged juvenile

coho in the ocean fisheries 1 977-83. Estimated

total numbers are

in parentheses

Landing area

»«..

#«"^

Release area

CO'

/I"

S.E. Alaska

20
(39)

0 0 0

British

1,086

0 0 0

Columbia

(2)

(1,735)

(90)

(8)

Puget Sound

201

1,352

1 2 0

(729)

(5,262)

(40)

(2) (9)

Washington

125

42 2 0

Coast

(5)

(316)

(42)

(451)

(151) (77)

Columbia

67 31 0

River

(133)

(14)

(162)

(310) (164)

Oregon Coast

62 308 10

(107)

(13)

(83)

(213) (1,137) (45)

California

1 138 19
(4) (552) (200)

'Sports catches are not expanded. The estimated total number
CWTs recovered in the sports fisheries. Preliminary data.

expanded commercial catch + actual number

189

FISHERY BULLETIN: VOL. 86, NO. 2

In general, the legal size limits increased from
north to south, which could result in more recov-
eries of juvenile coho salmon in northern than
southern regions. Thus, these data do not provide
evidence that a large proportion of juvenile coho
salmon from British Columbia and waters to the
south made northward migrations into Alaskan
waters before or during the commercial and
sports salmon seasons. Movements of Washing-
ton, Columbia River, and Oregon fish into British
Columbian waters were common however.

Hunter (1985) expanded the catches of CWT
juvenile coho salmon caught along the west coast
of North America during 1978-80 to the total
landed plus estimated "drop-off' mortality (fish
that were hooked and died without being landed)
of both tagged and untagged hatchery groups.
Calculations of the percentage returns from dif-
ferent release and recapture areas are similar to
ours (Table 6). The highest percentage of returns
were from the areas of release for all areas except
for California, and a higher proportion of the
catches of Washington coastal, Puget Sound, Co-
lumbia River, and Oregon coastal stocks were re-
ported north than south of the area of release.

Are Juvenile Coho Highly Migratory?

Based on our observations on movements of
marked fish, north-south and seasonal trends in
abundance and size, and directional purse seine
sets during the summer, we conclude that many
juvenile coho salmon from Oregon and Washing-
ton coastal streams and the Columbia River are
transported by currents to the south in May and
June but then migrate north later in the summer.
The mean catches per set of yearling coho salmon
in August and September are a large fraction of
those in June, indicating that in the years studied
many juvenile coho salmon in coastal waters of
Oregon and Washington were not highly migra-
tory. Moreover, more marked juvenile hatchery
coho salmon were caught in ocean fisheries in the
region of release than in distant waters. Recover-
ies of juvenile coho salmon released from hatch-
eries south of Cape Flattery were rare in northern
waters off Alaska and relatively few were recov-
ered in British Columbia (Table 6). In addition,
the positive correlation between upwelling off Or-
egon and survival of hatchery coho salmon from
the Columbia River, Oregon, and California
(Nickelson 1986) also argues for a close coupling
of OPI coho salmon with a local, not a distant,
environmental event during the time that year-

class survival is determined. All of these trends
suggest that most juvenile coho salmon from this
area are not highly migratory and that many usu-
ally remain in coastal waters near their sites of
ocean entry during their first summer in the
ocean, and perhaps during their entire ocean life.
In years of unfavorable ocean conditions, how-
ever, movements may be more extensive or mor-
tality may be higher, as suggested by the very low
catches of juvenile coho salmon in purse seine sets
south of Cape Flattery during September 1983,
the year of the recent strong El Nino.

Although Pacific salmon are renown for their
long foraging migrations in the subarctic Pacific,
coho salmon demonstrate both nonmigratory and
highly migratory behavior. Milne (1950) found
immature coho salmon almost year-round in
Georgia Strait and concluded that two types of
coho salmon exist in British Columbia waters:
"ocean" and "inshore" types, the "ocean" type
spending most of its ocean life in coastal and off-
shore waters and the "inshore" type in inside
waters such as Georgia Strait. Healey (1978)
caught "inshore" juvenile coho salmon in purse
seines in Georgia Strait during summer, fall, and
winter months. Similarily, large numbers of coho
salmon originating from streams of Puget Sound
remain in the Sound throughout their marine life
(Haw et al. 1967). Young coho salmon have also
been found in the winter and spring, many
months after seawater entry in Yaquina Bay
(Myers 1980) and other Oregon estuaries
(J. Nicholas^). Hartt and Dell (1986), in their im-
pressive study of juvenile salmonids of the north-
eastern Pacific during 1956-70, recognized these
two migratory patterns of coho salmon. They
found juvenile coho salmon in waters off Vancou-
ver Island and in the Strait of Juan de Fuca
throughout the summer and fall, and concluded
that some coho salmon spend their entire marine
life in "inside" waters and make only limited
ocean migrations.

What Proportion of Juvenile Coho

from Oregon and

Washington Migrate North?

The tagging experiments reported by Hartt and
Dell (1986) and Godfrey (1965) provide convinc-
ing evidence for long-distance migrations of coho

■IJ. Nicholas, Oregon Department of Fish and Wildlife, Cor-
vallis, OR, 97331, pers. commun. May 1986.

190

PEARCY and FISHER: MIGRATIONS OF COHO SALMON

salmon during their first summer in the ocean.
Based on recoveries of maturing coho salmon that
were tagged a year earlier at sea during April-
October 1956-70, Hartt and Dell (1986) con-
cluded that juvenile coho salmon from the Colum-
bia River, Oregon, and California may form a
large proportion of the coho stocks that migrate
north along the coast each summer. Of the 70
recoveries of tagged fish that were released be-
tween Kodiak Island and 56°N, 37% were recov-
ered the following year in the area of the Colum-
bia River and Oregon-California; of the 59
recoveries offish released between 56°N and Cape
Flattery, 47% were recovered in these southern
regions. In all, 25% of the recoveries were from
Oregon-California, 16% from the Columbia
River, 14% from Washington, 33% from British
Columbia, and 12% from Alaska.

Loeffel and Forster (1970) concluded that pat-
terns of radioactive ^^Zn in juvenile coho salmon
collected in the northeastern Pacific supported
the concept of a northerly migration from Oregon
and Washington into the Gulf of Alaska during
the summer. They found that juvenile coho
salmon captured off the west coast of Vancouver
Island in June and July 1967 contained ^^Zn, pre-
sumably originating from neutron activation of
Columbia River water used to cool the nuclear
reactors at Hanford, WA. ^^Zn levels decreased
in fish caught farther to the north (54°42'N-
58°24'N) in July-September of 1967. The authors
thought the low concentrations in northern sam-
ples represented background levels and that fish
with relatively high levels of ^^Zn had associated
with the Columbia River plume and subsequently
migrated north from the Oregon- Washington re-
gion. They found low ^^Zn levels in 1968, how-
ever, and no pronounced latitudinal gradients.
Furthermore, they reported none of the many fin-
marked juvenile coho salmon released from Ore-
gon and Washington hatcheries in 1967 and 1968
north of Juan de Fuca Strait. Hence their evi-
dence for northward movements of Columbia
River or Oregon-Washington coho salmon was
equivocal.

During June and July 1984, research was con-
ducted with the NMFS Auke Bay Laboratory in
waters from northern California to southwest
Alaska from the FV Pacific Warwind and Bering
Sea , both making round hauls with the same size
of purse seine, to sample juvenile coho in waters
north of Oregon and Washington: 37 sets were
made in coastal waters of British Columbia, and
39 were made in coastal waters and 29 in inland

waters (bays, inlets, and ^ords) of southeastern
Alaska. Of the 371 juvenile coho salmon captured
in these regions, 77% were caught in inland
waters of southeastern Alaska. The seven CWT
juvenile coho salmon captured were all from
Alaska inland waters and all originated from
Alaska hatcheries (Auke Bay Laboratory 1984a).
A later cruise in southeast Alaska by the Auke
Bay Laboratory in August 1984 caught eight
CWT coho salmon, also all from inland waters
and from Alaskan hatcheries (Auke Bay Labora-
tory 1984b).

Of the 14 CWT juvenile coho salmon collected
in other purse seines, gill nets, and special troll
gear in waters of southeastern Alaska during
1982, 1983, and 1985, 12 originated from Alaska
hatcheries and 2 originated from Washington
hatcheries (Auke Bay Laboratory 1983; Jaenicke
et al. 1984; Orsi et al. 1987). Table 6 shows that
only 5 of 25 CWT juvenile coho salmon caught in
Alaskan waters during 1977-83 were from hatch-
eries south of Alaska, indicating that most juve-
nile coho salmon caught in southeastern Alaska
during the summer originated from Alaska and
not from southern regions.

Hartt (1980) and Hartt and Dell (1986) recog-
nized that their data did not indicate the propor-
tion of southern stocks that made northerly mi-
grations, but they concluded that a large
proportion is probable, since juvenile coho salmon
were consistently caught in most seine sets
throughout the area sampled. They estimated
that the average density of juvenile salmonids in
coastal waters between 56°N and 60°N off south-
eastern Alaska during August and September
1964-68 was 1,500 km"^. The average density of
juvenile coho salmon in this area during these
two months was only 82 km~^ (Hartt and Dell
1986, app. A). During August and September
1981-84, the average density of juvenile coho
salmon in our round hauls between Cape Flat-
tery, WA and Cape Arago, OR to 37 km offshore
was 350 km"^, several times the estimates of
Hartt and Dell for the same months of the year.
This suggests that juvenile coho salmon may be
found in higher densities off Oregon and Wash-
ington than southeastern Alaska during late
summer, assuming that distributions and abun-
dances in the late 1960s and early 1980s were
similar. This trend for higher abundances of juve-
nile coho salmon off Oregon and Washington than
in coastal waters farther north was also found
during July 1984 (Table 7), although average
catches off Washington and Oregon were not as

191

FISHERY BULLETIN: VOL. 86, NO. 2

Table 7. — Average catches of juvenile coho salmon in purse
seines sets in coastal waters along the west coast of North America
during July 198412

Area

No. per set

No. per km2

Sitka-Juneau

Ketchikan-Sitka

Cape Scott-Dixon Entrance

Vancouver Is.

Washington

Oregon No. California

1.29
0.58
1.90
1.80
3.76
2.59

100

95
198
136

'Cruise Report. Drum Seiner FV Bering Sea, Coastwide NWAFC/OSU
Cooperative Study, Ecology of Juvenile Salmon in Coastal and Inside Waters
of Soutfieast Alaska, 28 June-26 July 1984, NWAFC Auke Bay Laboratory,
National Marine Fisfierles Service, NCAA, P.O. Box 115, Auke Bay, AK
99821 .

2Fisher and Pearcy (1984).

large as in some earlier years owing to low sur-
vival (Fig. 2; Pearcy 1984; Fisher and Pearcy in
press).

Comparisons of the estimates of total juvenile
yearling coho salmon abundances off Oregon and
Washington with the production of coho smolts in
the Oregon Production Area (Columbia River to
California) also suggests that many juvenile coho
resided off Oregon and Washington during the
summer. By expanding our catches per m^ to the
region sampled, we estimated that the numbers of
juvenile yearling coho salmon in areas surveyed
by our purse seine sampling during August or
September 1981-84 were 6.3%, 6.5% 5.1%, and
5.2%, respectively, of the numbers of hatchery
and wild smolts released in the Columbia River
and in Oregon (T. Lichatowich^). The areas in-
cluded in these estimates were roughly 83%, 62%-,
51%, and 68% of the total area from Cape Flattery
to Cape Arago out to 37 km offshore. Recognizing
that the entire area was not sampled, that year-
class strength of coho salmon in this region is
probably established soon after ocean entrance
(Fisher and Pearcy in press), and that early
marine mortality may be inversely related to size
(Parker 1968; Ricker 1976) so that much of the
ocean mortality has occurred by late summer,
these percentages probably represent a substan-
tial portion of the surviving OPI coho smolts. In
fact, they are several times higher than the
smolt-to-adult survival of 1.3-2.8% for OPI public
hatchery coho salmon (excluding Rogue River
and California hatcheries) during 1981-84 (R.
Kaiser^).

5T. Lichatowich, Oregon Department of Fish and Wildlife,
P.O. Box 59, Portland, OR 97207, pers. commun. September
1987.

6R. Kaiser, Oregon Department of Fish and Wildlife, Hatfield
Marine Science Center, Newport, OR 97265, pers. commun.
September 1987.

We conclude, therefore, that a major fraction of
the juvenile coho salmon from Oregon and Wash-
ington hatcheries did not undertake distant mi-
grations into the Gulf of Alaska in recent years.
This is not necessarily in conflict with Hartt and
Dell's (1986) data, since they established the pres-
ence of Oregon and Washington coho salmon in
northern waters but not the proportion of total
production that undertakes this migration. On
the other hand, neither the stocks of coho nor
oceanographic conditions have remained con-
stant over the period from 1956 to 1985 when
these two studies were conducted. Wild coho
smolts exceeded hatchery smolts in the Oregon
Production Area before 1961 (Nickelson 1986) but
comprised <12% of the smolts in 1980-85 (R.
Kaiser fn. 6). Perhaps wild smolts from the OPI
area had different migratory patterns than hatch-
ery smolts do today and migrated rapidly into
northern waters soon after ocean entrance. This
may explain Nickelson's (1986) finding that sur-
vival of hatcherv, but not wild coho smolts, was
significantly correlated with coastal upwelling off
Oregon.

Ocean conditions have also changed over this
period. The late 1960's were accompanied by
strong upwelling along the coast compared to
weak upwelling in the early 1980's (Nickelson
1986; Mason and Bakun 1986). McLain (1984),
Norton et al. (1985), and Royer (1985) illustrated
that sea surface temperatures and sea levels in-
creased in the northeastern Pacific between 1976
and 1984. These factors and associated changes in
ocean circulation could explain differences in mi-
gratory behavior of coho salmon between 1960's
and 1980's. If currents provide orientational cues
to migration, cues facilitating northerly move-
ments may be reduced during years of weak up-
welling and weak Ekman transport from the
north. Ocean conditions may have modified mi-
gratory patterns, as they possibly have for the
migration of Fraser River sockeye salmon around
Vancouver Island (Groot et al. 1984; Hamilton
1985).

ACKNOWLEDGMENTS

We thank D. Larden and his crew of the FV
Pacific Warwind and Bering Sea for their cooper-
ation and competence during purse seining opera-
tions at sea; A. Chung, R. Brodeur, J. Shenker,
W. Wakefield, D. Gushee, C. Banner, J. Long, K.
Krefft, and C. Wilson for their help on cruises;
J. Norton and the Oregon Department of Fish and

192

PEARCY and FISHER: MIGRATIONS OF COHO SALMON

Wildlife Clackamas Laboratory for decoding
coded-wire tags; K. Johnson for providing data on
marked fish; W. McNeil and R. Severson for as-
sisting in fluorescent spray-marking Oregon
Aqua-Food's Inc. smolts; Northwest and Alaska
Fisheries Center for the loan of a purse seine;
Northwest and Alaska Fisheries Center Auke
Bay Laboratory for the loan of electronarcosis
equipment; and the Faculty of Fisheries, Hok-
kaido University, and the TV Oshoro-Maru for
conducting gill net research. A. Hartt,
H. Jaenicke, R. Brodeur, R. Gowan, D. Hankin,
and L. Botsford all provided helpful comments on
the manuscript.

This research was supported by Oregon State
University Sea Grant College Program (Grant
No. NA 81AA-D-00086, Project R/OPF-17), the
Northwest and Alaska Fisheries Center (Con-
tract 81-ABC-00192, 83-ABC-00102, 84-ABC-
0009, and NA-85-ABH0002J), the Oregon De-
partment of Fish and Wildlife, and Oregon
Aqua-Foods, Inc.

LITERATURE CITED

Auke Bay Laboratory.

1983. Cruise report. NOAA fishery research vessel John
N. Cobb, JC-83-03. Juvenile salmon purse seining proj-
ect in coastal and inside waters of southeastern
Alaska. NWAFC Auke Bay Lab., Natl. Mar. Fish.
Serv., NOAA, 30 p.
1984a. Cruise report drum seines. F/V Bering Sea.
Coastwide NWAFC/OSU Cooperative study. Ecology of
juvenile salmon in coastal and inside waters of south-
eastern Alaska, 28 June-26 July 1984. NWAFC Auke
Bay Lab., Natl. Mar. Fish. Serv., NOAA, 9 p.
1984b. Cruise report, NOAA RV John N. Cobb Cruise
84-03. Juvenile salmon purse seining project in coastal
and inside waters of southeastern Alaska, July 31-Au-
gust31, 1984. NWAFC Auke Bay Lab., Natl. Mar. Fish.
Serv., NOAA.

Chung, A

1985. Relationships between oceanographic factors and
the distribution of juvenile coho salmon (Oncorhynchus
kisutch) off Oregon and Washington, 1982-83. M.S.
Thesis, Oregon State Univ., Corvallis, 116 p.

Clutter. R I., and L E. Whitesel.

1956. Collection and interpretation of sockeye salmon
scales. Int. Pac. Salmon Comm. Bull. 9:1-159.
Dawley. E. M , R. D Ledgerwood, T H Blahm, C. W. Sims, J. T.
DuRKiN, R A Kirn, A E Rankis, G E Monan, and F J.

OSSIANDER

1986. Migrational characteristics, biological observa-
tions, and relative survival of juvenile salmonids enter-
ing the Columbia River estuary, 1966-1983.
Final Report, Bonneville Power Administration Project
No. 81-102. Northwest and Alaska Fish. Cent., Natl.
Mar. Fish. Serv., NOAA, Seattle, 256 p.

Dawley. E M . R. D. Ledgerwood, and A. Jensen

1985. Beach and purse seine sampling of juvenile

salmonids in the Columbia River estuary and ocean
plume, 1977-1983. Vol. 2: Data on marked fish recover-
ies. U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/
NWC-75, 397 p.
Dawley, E M., C. W Sims. R. D. Ledgerwood, D. R. Miller, and
J G Williams

1981. A study to define the migrational characteristics of
chinook and coho salmon in the Columbia River estuary
and associated marine waters. Final Rep. Northwest
and Alaska Fish. Cent., Natl. Mar. Fish. Serv., NOAA,
68 p.

Fisher, J P , and W G Pearcy

1984. Studies of juvenile salmonids off northern Califor-
nia, Oregon, Washington and Vancouver Is-
land. 1984. Oreg. State Univ. Sea Grant Coll. Pro-
gram (ORESU-T-OGl), 24 p.

1985. Studies of juvenile salmonids off the Oregon and
Washington coast, 1985. Oreg. State Univ. Sea Grant
Coll. Program (ORESU-T-004), 31 p.

In press. Growth of juvenile coho salmon [Oncorhynchus
kisutch ) off Oregon and Washington, USA in years of
differing coastal upwelling. Can. J. Fish. Aquat. Sci.
Fisher, J P , W G Pearcy, and A. W Chung

1982. Studies of juvenile salmonids off the Oregon and
Washington coast, 1982. Oreg. State Univ. Sea Grant
Coll. Program (ORESU-T-83-003), 41 p.

1983. Studies of juvenile salmonids off the Oregon and
Washington coast, 1983. Oreg. State Univ. Sea Grant
Coll. Program (ORESU-T-001), 29 p.

FOESTER, R. E.

1968. The sockeye salmon Oncorhynchus nerka. Fish
Res. Board Can. Bull. 162, 422 p.
French, R. R., R G Bakkala. and D. F. Sutherland.

1975. Ocean distribution of stocks of Pacific salmon,
Oncorhynchus spp., and steelhead trout, Salmo gaird-
nerii, as shown by tagging experiments. U.S. Dep.
Commer., NOAA Tech. Rep. NMFS SSRF 689, 89 p.
Godfrey, H

1965. Salmon of the North Pacific Ocean. Part IX Coho,
chinook and masu salmon in offshore waters. 1. Coho
salmon in offshore waters. Int. North Pac. Fish. Comm.
Bull. No. 16, 39 p.
Godfrey, H., K A Henry, and S. Machidori.

1975. Distribution and abundance of coho salmon in off-
shore waters of the North Pacific Ocean. Int. North
Pacific Fish. Comm. Bull. No. 31, 80 p.
Groot, C , L Margolis, and R. Bailey.

1 984 . Does the seaward migration of Eraser River sockeye
salmon (Oncorhynchus nerka) smolts determine the
route of return of adults? In J. D. McCleave, G. P.
Arnold, J. J. Dodson, and W. H. Neill (editors). Mecha-
nisms of migration in fishes, p. 283-292. Plenum Press,
N.Y.

GUNSOLUS, R. T.

1978. The status of Oregon coho and recommendations for
managing the production, harvest, and escapement of
wild and hatchery reared stocks. Oreg. Dep. Fish.
Wildl. Fish Div. Proc. Rep., 59 p.

Hamilton, K.

1985. A study of the variability of the return migration
route of Eraser River sockeye salmon (Oncorhynchus
nerka). Can. J. Zool. 63:1930-1943.

Hartt, A C

1975. Problems in sampling Pacific salmon at sea. Int.
North Pac. Fish. Comm. Bull. 32:165-231.

193

FISHERY BULLETIN: VOL. 86, NO. 2

1980. Juvenile salmonids in the oceanic ecosystem — the
critical first summer. In W. J. McNeil and D. C.
Himsworth (editors), Salmonid ecosystems of the North
Pacific, p. 25-57. Oreg. State Univ. Press.
Hartt, a. C , AND M. B Dell.

1986. Early oceanic migrations and growth of juvenile
Pacific salmon and steelhead trout. Int. North Pac.
Fish. Comm. Bull. 46:1-105.
Haw. F . H. O. Wendler, and G Deschamps.

1967. Development of Washington State salmon sports
fishery through 1964. Wash. Dep. Fish. Res. Bull. 7,
191 p.
Healey, M. C.

1978. The distribution, abundance, and feeding habits of
juvenile Pacific salmon in Georgia Strait, British Colum-
bia. Dep. Fish. Environ. Can., Fish. Mar. Serv. Tech.
Rep. No. 788, 49 p.

HiCKEY, B M

1979. The California Current System - hypotheses and
facts. Prog. Oceanogr. 8:191-279.

Hunter. M. A.

1985. The 1976-1978 brood coho model. Wash. Dep.
Fish. Prog. Rep. No. 222, 146 p.
HUYER. A

1983. Coastal upwelling in the California Current Sys-
tem. Prog. Oceanogr. 12:259-284.

HUYER, A , R O PiLLSBURY, AND R L. SMITH.

1975. Seasonal variation of the alongshore velocity field
over the continental shelf off Oregon. Limnol.
Oceanogr. 20:90-95.

HuYER, A., AND R. L. Smith.

1985. The signature of El Niiio off Oregon, 1982-1983.
J. Geophy. Res. 90:7133-7142.

HUYER, A., E. J. C. SOBEY, AND R. L. SMITH.

1979. The spring transition in currents over the Oregon
continental shelf J. Geophy. Res. 84:6995-7011.
Jaenicke. H. W., R D Brodeur, and T Fujii.

1984. Exploratory gillnetting from the Oshoro-Maru for
juvenile salmonids off southeastern Alaska, 24-25 July
1982. Bull. Fac. Fish. Hokkaido Univ. 35:154-160.

Johnson, S L.

1984. The effects of the 1983 El Nirio on Oregon's coho and

Chinook salmon. Oreg. Dep. Fish Wild. Info. Rep. 84-8,

40 p.
KuNDU, P. K., AND J S Allen

1976. Some three-dimensional characteristics of low-
frequency fluctuations near the Oregon coast. J. Phys.
Oceanogr. 6:181-199.

LOEFFEL, R. E., AND W. O FORSTER

1970. Determination of movement and identity of stocks
of coho salmon in the ocean using the radionuclide zinc-
65. Res. Rep. Fish Comm. Oreg. 2:1-12.
Lord. G . W C Acker. A. C Hartt. and B J Rothschild.

1976. An acoustic method for high-seas assessment of mi-
grating salmon. Fish. Bull., U.S. 74:104-111.
Machidori, S

1966. Vertical distribution of salmon (Genus Oncor-
hynchus) in the North-western Pacific. Hokkaido Reg.
Fish. Res. Lab. 31:11-17.
Manzer, J I

1964. Preliminary observations on the vertical distribu-
tion of Pacific salmon (Genus Oncorhynchus ) in the Gulf
of Alaska. J. Fish. Res. Board Can. 25:1085-1089.
Mason, J. E , and A Bakun

1986. Upwelling index update, U.S. West Coast, 33N-
48N latitude. U.S. Dep. Commer., NOAA Tech. Memo.

NMFS No. 67, 81 p.

McLain, D R

1984. Coastal ocean warming in the Northeast Pacific,
1976-83. In W. G. Pearcy (editor), The infiuence of
ocean conditions on the production of salmonids in the
North Pacific, p. 61-86. Oreg. State Univ. Sea Grant
Coll. Program (ORESU-W-001).

McCarl, B. a., and R. B. Rettig

1983. Influence of hatchery smolt releases on adult
salmon production and its variability. Can. J. Fish.
Aquat. Sci. 40:1880-1886.

McGlE.A M

1984. Evidence for density dependence among coho
salmon stocks in the Oregon Production Index area. In
W. G. Pearcy (editor). The influence of ocean conditions
on the production of salmonids in the North Pacific, p.
37-49. Oreg. State Univ. Sea Grant Coll. Program
(ORESU-W-83-001).

Miller, D R., J G. Williams, and C. W. Sims.

1983. Distribution, abundance and growth of juvenile

salmonids off Oregon and Washington, summer 1980.

Fish Res. 2:1-17.
Milne, D J

1950. The difference in the growth of coho salmon on the

east and west coasts of Vancouver Island in 1950. Fish.

Res. Board Can. Prog. Rep. Pac. Coast Biol. Sta. 85:80-

82.
Myers, K, W. W.

1980. An investigation of the utilization of four study

areas in Yaquina Bay, Oregon, by hatchery and wild

juvenile salmonids. M.S. Thesis, Oregon State Univ.,

Corvallis, 234 p.
NiCKELSON, T. E.

1986. Influences of upwelling, ocean temperature, and
smolt abundance on marine survival of coho salmon
(Oncorhynchus kisutch ) in the Oregon Production Area.
Can. J. Fish. Aquat. Sci. 43:527-535.

Norton, J , D McLain. R Brain, and D Husby

1985. The 1982-83 El Nirio event off Baja and Alta Cali-
fornia and its effect on ocean climate context. In W. S.
Wooster and D. L. Fluharty (editors). El Niiio North, p.
44-72. Univ. Wash. Sea Grant Program.

Oregon Department of Fish and Wildlife.

1982. Comprehensive plan for production and manage-
ment of Oregon's anadromous salmon and trout. Part II.
Coho salmon plan. Oreg. Dep. Fish Wildl., var. p.
Orsi, J A., A G Celewycz, D. G Mortensen. and K A
Herndon

1987. Sampling juvenile chinook salmon (Oncorhynchus
tshawytscha) and coho salmon (O. kisutch) by small
trolling gear in the northern and central regions of south-
eastern Alaska, 1985. U.S. Dep. Commer., NOAA Tech.
Memo. NMFS F/NWC-115, 47 p.

Pacific Fisheries Management Council

1986. Preseason Report 1. Stock abundance analysis for
1986 ocean salmon fisheries. Pac. Fish. Manage.
Counc, var. p.

Pacific Marine Fisheries Commission.

1980. 1977 Pacific salmonid coded wire tag recoveries.
Regional Mark Processing Center for Pacific Coast
States, Portland, OR.

1981. 1978 Pacific salmonid coded wire tag recoveries.
Regional Mark Processing Center for Pacific Coast
States, Portland, OR.

1984a. 1979 Pacific salmonid coded wire tag recoveries.
Regional Mark Processing Center for Pacific Coast

194

PEARCY and FISHER: MIGRATIONS OF COHO SALMON

States, Portland, OR.
1984b. 1980 Pacific salmonid coded wire tag recoveries.

Regional Mark Processing Center for Pacific Coast

States, Portland, OR.
1984c. 1981 Pacific salmonid coded wire tag recoveries.

Regional Mark Processing Center for Pacific Coast

States, Portland, OR.
1985a. 1982 Pacific salmonid coded wire tag recoveries.

Regional Mark Processing Center for Pacific Coast

States, Portland, OR.
1985b. 1983 Pacific salmonid coded wire tag recoveries.

Regional Mark Processing Center for Pacific Coast

States, Portland, OR.
Parker. R R

1968. Marine mortality schedules of pink salmon of the

Bella Coola River, Central British Columbia. J. Fish.

Res. Board Can. 25:757-794.
Pearcy. W G

1984. Where do all the coho go? The biology of juvenile
echo salmon off the coasts of Oregon and Washington.
In W. G. Pearcy (editor). Influence of ocean conditions on
the production of salmonids in the North Pacific, p. 50-
60. Oreg. State Univ. Sea Grant Coll. Program
(ORESU-W-83-001).

Pearcy, W G , R D Brodeur, J M Shenker, W. W Smoker, and
Y Endo

In press. Food habits of Pacific salmon and steelhead
trout, midwater trawl catches, and oceanographic condi-
tions in the Gulf of Alaska, 1980-1985. Bull. Ocean
Res. Inst. Univ. Tokyo.
Pearcy, W . J Fisher. R Brodeur, and S Johnson

1985. Effects of the 1983 El Niiio on coastal nekton off
Oregon and Washington. In W. S. Wooster and D. L.
Fluharty (editors). El Nino North, p. 188-204. Univ.
Wash. Sea Grant Program.

Pearcy. W G , and A Schoener

1987. Changes in the marine biota coincident with the

1982-1983 El Nino in the northeastern subarctic Pacific
Ocean. J. Geophys. Res. 92, No. 013:14417-14428.
Peterman, R M . and R D Roittledge

1981. Experimental management of Oregon coho salmon
(Oncorhynchus kisutch ): Designing for yield of informa-
tion. Can. J. Aquat. Sci. 40:1212-1223.
Phinney, D E , D M Miller, and M L Dahlberg.

1967. Mass-marking young salmonids with fluorescent
pigment. Trans. Am. Fish. Soc. 96:157-162.
Richer. W. E

1976. Review of the growth rate and mortality of Pacific
salmon in salt water, and non-catch mortality caused by
fishing. J. Fish. Res. Board Can. 33:1483-1524.
Royer, T. C.

1985. Coastal temperature and salinity anomalies in the
northern Gulf of Alaska, 1970-84. In W. S. Wooster and
D. L. Fluharty (editors), El Nino North, p. 107-115.
Univ. Wash. Sea Grant Program.
SCARNECCHIA, D. L.

1981. Effects of streamflow and upwelling on yield of wild
coho salmon (Oncorhynchus kisutch ) in Oregon. Can. J.
Fish. Aquat. Sci. 38:471-475.
SUSUKI, T., AND H SONODA

1972. On a trial fish finder for salmon and its experimen-
tal results. Bull. Jpn. Soc. Sci. Fish. 38:463-469.

Wakefield, W. W., J. P. Fisher, and W G Pearcy.

1981. Studies of juvenile salmonids off the Oregon and
Washington coast, 1981. Oreg. State Univ. Sea Grant
College Program Ref 81-13, 51 p.
Ware, D M

1978. Bioenergetics of pelagic fish: Theoretical change in
swimming speed and ration with size. J. Fish. Res.
Board Can. 35:220-228.
Weihs, D

1973. Optimal fish cruising speed. Nature 245:48-50.

195

EARLY MARINE LIFE OF ATLANTIC SALMON, SALMO SALAR,
POSTSMOLTS IN THE NORTHERN GULF OF ST. LAWRENCE

J -D DUTIL AND J.-M. COUTU'

ABSTRACT

Postsmolts of Atlantic salmon, Salmo salar, having spent some 2-4 months in the marine environ-
ment were collected in the nearshore zone of the northern Gulf of St. Lawrence. From a back-
calculated smolt length of 135 mm they had increased in length to 265 mm (212 g) on 1 September,
and 306 mm (320 g) on 30 September. The rate of increase in length averaged 1.65 mm/day over more
than 2 months. Individuals reached 35 cm and 500 g in late autumn. Postsmolts moved in small shoals
near the surface and were possibly more active at dawn and dusk. In midsummer, stomach contents
changed from insects and gammarids to sand lance, 40-100 mm in length; vertically migrating
crustaceans also occurred in the stomachs in autumn. River origin of these postsmolts is not known.
The possibility of their belonging to a particular subgroup of some north shore stocks is examined in
relation to scale patterns and size of gonads. This occurrence of postsmolts near shore in late summer
and presumably their late movement out of the Gulf of St. Lawrence indicate the directional nature
of smolt migration to distant feeding areas should be reconsidered. Low sea temperature is hypothe-
sized to trigger the movement out of the Gulf of St. Lawrence. Continual presence of postsmolts in a
shallow layer at the surface could prove to be extremely valuable in forcasting movements and
production.

Many papers have been published concerning the
biology of Atlantic salmon, Salmo salar, but very
little has been said concerning the postsmolt
stage. This stage has been defined as ". . .the
juvenile salmon from the time that it leaves the
river as a smolt until the onset of wide annulus
formation on the scales at the end of the first
winter in the sea" (Allan and Ritter 1977). This
paper presents new data on the Atlantic salmon
postsmolts (hereafter referred as postsmolts) in
the northern Gulf of St. Lawrence, reviews our
current knowledge on the biology of postsmolts,
and points to biological and environmental fac-
tors potentially limiting the success of their early
life in marine environments.

Published data on postsmolts are mainly lim-
ited to stocks in the Baltic Sea. Routes of migra-
tion have been described based on the locations
and time of early recapture from smolt releases in
Sweden and Finland (Carlin 1959; Larsson and
Ateshkar 1979; Ikonen and Auvinen 1984, 1985;
Jutila and Alapassi 1985). Data on predators are
limited (Soikkeli 1973; Valle 1985), and most of
the material concerns predation on smolts in
streams and estuaries (Larsson 1985). Many
analyses of stomach contents have been pub-

'Ministere des Peches et des Oceans, Gouvernement du
Canada, 850, route de la Mer, C.P. 1000, Mont-Joli, Quebec,
Canada G5H 3Z4.

lished, particularly on smaller postsmolts (re-
viewed by Christensen and Larsson 1979; Jutila
and Toivonen 1985). Data on rate of growth
(Ikonen and Auvinen 1985) and rate of mortality
(Carlin 1959) are lacking. However, Baltic
salmon spend their entire sea life in the brackish
waters of the Baltic and nearly 80% of smolt pro-
duction originates from hatcheries (Anonymous
1984). Thus the information derived from salmon
in the Baltic should be extended to other stocks
only with caution.

Publications on postsmolts in the northern At-
lantic and Gulf of St. Lawrence mentioned small
salmon as bycatches of commercial fisheries and
described the distribution of recaptures from
smolt release programs. The earliest report on
postsmolts in the Gulf of St. Lawrence claimed
that small salmon, referred to as "ouananiche" by
local fishermen, were regularly taken near shore
in herring nets in autumn (Comeau 1909).
Kendall (1935) also reported such incidental
catches for the New England coast. Elson (1953)
recorded a bycatch of more than 1,000 marked
postsmolts from one locality in the Bay of Fundy
in the period 1951-53 and reported their mean
length. There are also limited records of
postsmolts taken off France (Vibert 1953) and in
the Gulf of St. Lawrence (Caron 1983) from
smolts tagged in streams. Recently, information
on movements has been derived from tag returns

Manuscript accepted December 1987.
FISHERY BULLETIN; VOL. 86, NO. 2, 1988.

197

FISHERY BULLETIN: VOL. 86, NO. 2

of smolts released in spring in New England and
caught in summer in Canada (Meister 1984). Ru-
mors of bycatch in herring nets along the coast of
the northern Gulf of St. Lawrence in autumn pro-
vided an occasion to acquire some knowledge con-
cerning the elusive postsmolt. Production of
salmon in the sea may well be limited by the
success of smolts in the marine environment.

Materials and Methods

Postsmolts were collected between Bale Trinite
and Riviere-au-Tonnerre in the northern Gulf of
St. Lawrence, in 1982, 1983, 1984, and 1985 (Fig.
1). Fishermen contacted in summer 1982 col-
lected postsmolts in late summer and autumn as
bycatch in herring gill nets. Four of them were
asked in 1983 to monitor the catch of postsmolts
in experimental gill nets in late-September in 4
locations (Bale Trinite, Riviere Pentecote, Port
Cartier, and Sept-Iles). We also monitored sta-
tions in Bale Trinite in 1983 (23 September-11
October), in Baie Trinite and Port Cartier in 1984
(21 August-20 October), and in Sept-Iles in 1985
(20 August-4 October). Finally, smolts and early
postsmolts were collected in seines in June and

July, during an eel marking program in the estu-
ary of Grande Trinite River at Baie Trinite (Fig.
1).
Fishermen used standard herring gill nets in

1982, but custom-made gill nets were used in

1983, 1984, and 1985. Custom-made gill nets had
5 sections of increasing mesh sizes (50.8, 57.2,
63.5, 69.9, and 76.2 mm stretched) covering the
range in mesh sizes of herring gill nets in the
northern Gulf of St. Lawrence. Stretched mesh
sizes were determined by measuring 10 meshes
per section. Sections were 6 m deep and 10 m long.
In 1983, 1984, and 1985, postsmolts were
recorded by section individually. In 1984 and
1985, their position in the nets was recorded more
precisely: floating lines had numbered buoys and
a string divided the nets into 2 halves horizon-
tally. Time of the catch was also recorded. The gill
nets were usually visited at 2-h intervals be-
tween 0600 and 1800, as sea conditions allowed.
They were left fishing overnight. Gill nets were
all set at the surface and near shore (<2 km).
Mean air temperatures for 1982 to 1985 were
drawn from Environment Canada meteorological
summaries for Sept-Iles airport. Temperature
recorders were also tied to nets in 1983, 1984, and

R.MOISIE

SEPT-ILES
PORT CARTIER.

R.PENTECdTE*/

NORTH SHORE

GULF
OF
ST. LAWRENCE

Figure 1. — Locations of the areas investigated in the northern Gulf of St.
Lawrence (shaded area).

198

DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON

1985 to get hourly records of temperature near
the surface, but portions of these records are miss-
ing.

Fish were preserved at -15°C for less than
4 months in 1982 and less than 2 months in 1983,
1984, and 1985. Fork length was measured to
nearest mm and weight to nearest g. Condition
index values were calculated as

C/ = sin 1

100 ■ W

0.5

CI = condition index; W = weight (g); L = length
(mm); s = slope of the length- weight relationship
for age groups combined. Scales, usually taken
below the dorsal fin close to the lateral line, were
cleaned and mounted on glass slides. Stomachs
were preserved in formalin (10%) or alcohol (70%)
to be analyzed later for their contents. The sex
was determined visually and checked histologi-
cally on a subsample of postsmolts collected in
1982. Gonads were weighed to nearest mg and
only in 1982. Gonadosomatic index (GSI) was cal-
culated as

GSI =

TW-GW

100

where GW = gonad weight; TW - total weight.

Scales showing no focal regeneration were ex-
amined for age determination by 2-4 readers.
Readings were discussed and a consensus was
reached in most cases. Reported ages are smolt
ages plus 2-4 months. Fork length at smoltifica-
tion (mm) was back-calculated as

LS =

(LC - 34) ■ RS
RC

+ 34

where LS = back-calculated smolt length; LC
= body length of post smolt; RS = scale radius at
smolt check; RC = scale radius of postsmolt.
Length at scale formation is assumed to be 34
mm. Scales were also examined for any mark in
the postsmolt zone that could be of potential in-
terest. Hence the number of individuals showing
a transition zone between the riverine and
marine increments was determined in 1982, and
the number of individuals showing a summer
check was determined in 1982, 1983, 1984, and
1985. Scale radii were measured along the
postero-anterior axis on 4 scales per individual in
1982, and on 1 scale per individual in 1983, 1984,
and 1985. Projected scales were measured at con-

stant magnification on a digitizing pad connected
to a personal computer. Fork length at formation
of a summer check (mm) was back-calculated for
1982 and 1983 as

LF =

(LC - 34) • RM
RC

+ 34

where LF = back-calculated length; LC = body
length of postsmolt; RM = scale radius at summer
check; RC = scale radius of postsmolt.

Stomach contents were examined in detail in
1982. Results are expressed as percentage of oc-
currence (number of stomachs containing an item
divided by the number of stomachs examined)
and percentage by number (the count number of
an item divided by the count number of all items).
Food items that could not be identified are re-
ferred to as remains. In as far as their condition
allowed, prey fish were usually measured to the
nearest cm. Stomach contents examined in 1983,
1984, and 1985 were consistent with the 1982
conclusions, so the results are not presented.

Results
The Age and Sex Composition

Three-year-old female salmon dominated the
catch both in 1982 and 1983 (Table 1). Postsmolts
ranged from 2 to 5 years in 1982 and from 2 to 4
years in 1983. Five-year-old salmon made up less
than 4% of the catch in 1982. Three-year-old and,
to a lesser extent, 4-yr-old salmon dominated the
catch both in 1982 and 1983 (82% of the catch);
mean smolt age was 3.2 ± 0.7 years. Age composi-
tion in 1983 differed from 1982 because of an in-
crease in the percentage of 2-yr-old salmon (12%
to 21%) and a decrease in the percentage of 5-yr-
old salmon (x^ = 15.96, P < 0.01). The sex ratio
(male:female) for 1982 and 1983 combined was

Table 1 . — The age and sex composition of the catch of postsmolt
Atlantic salmon along the north shore of the Gulf of St. Lawrence,
1982 and 1983.

Sex

Smolt

age

Year

Total

1982

Female

114

226

Male

142

1983

Female

Male

Total

Female

172

324

Male

199

FISHERY BULLETIN; VOL. 86, NO. 2

0.62. This ratio did not change between years (x^
= 7.92, P < 0.01) and was significantly less than
1.0 (binomial test: z = -5.37, P < 0.001). There
was no trend between sex ratio and smolt age in
1982 (x^ = 1.21, P = 0.75), but the percentage of
males tended to increase with smolt age in 1983
(X^ - 12.67, P - 0.002).

Length at Smoltification

The mean length at smoltification was esti-
mated by back-calculation at 135 ± 15 mm for
postsmolts collected in 1982 and 1983. Scale ra-
dius was linearly related to body length both in

1982 (4 scales measured per fish, P < 0.0001) and
in 1983 (1 scale measured per fish, P < 0.0001);
probability values are those of F-tests from anal-
ysis of variance. Back-calculated smolt lengths
were normally distributed (P = 0.80): 75% of the
postsmolts ranged between 120 and 150 mm at
smoltification. There were also 7% individuals in
the 160-200 mm range. Differences in mean
smolt length were not significant between the
ages and the sexes both in 1982 (F = 1.84,
P = 0.08) and in 1983 (F = 1.05, P = 0.39), and,
pooling the ages and the sexes, between 1982 and

1983 (^ = 1.32, P >0.15).

Rate of Increase in Size

The rate of increase in size of postsmolts was
very rapid in summer both in 1982 and 1983.
Postsmolts collected in 1982 between mid-August
and mid-October in = 383; mean date is 1 Sep-
tember) averaged 265 ± 25 mm (range 195-328)
and 212 ± 58 g (range 92-389). In 1983,
postsmolts collected between mid-September and
mid-October (n = 155, mean date is 30 Septem-
ber) averaged 306 ± 17 mm (range 258-362) and
320 ± 57 g (range 192-565). There was no differ-
ence in mean size between males and females and
between age-classes in 1982 (P > 0.40) and 1983
(P > 0.20); probability values are those of F-tests
from 2-way analysis of variance. From a mean
smolt length of 135 mm and assuming smolts mi-
grated to estuaries 15 June, postsmolts grew at a
rate of 1.65 mm/day during a 2.5-mo (15 June —
1 September) and a 3.5-mo (15 June-30 Septem-
ber) period in 1982 and 1983, respectively.

This estimate is conservative because the rate
of increase in length tended to decrease late in the
sampling period. Postsmolts steadily increased in
length and weight in the period mid-August to
mid-September 1982 (Figs. 2, 3). From mid-

September, the rate tended to slow down. The
inclusion of data for 1983, collected later in the
autumn, corroborates this observation indicating
that conditions changed in late-September in
1982 and 1983.

Length-weight relationships were examined
for 1982 and 1983 separately. The analysis of co-
variance showed that males and females exhib-
ited the same length-weight relationship, both in

1982 (P - 0.53) and 1983 (P - 0.42). Similarly,
length-weight relationship did not change be-
tween age groups in 1982 (P - 0.06) and 1983
(P = 0.49). The covariance for 1982 was nearly
significant because the slope for 5-yr-old
postsmolts, based on 15 individuals, was larger
than for other age groups. However, mean condi-
tion index values by age revealed no significant
difference between age groups in 1982 (P = 0.11)
or 1983 (P = 0.28).

Since there was also no significant difference in
the length- weight relationship between 1982 and

1983 (P = 0.14), the data were pooled. Thus the
length-weight relationship for postsmolts col-
lected in this study can be described as a single
regression:

log W = (2.8280 • log L) - 4.5336

P < 0.001, n = 539

where W = weight (g); L = length (mm).

Maturation of Gonads

Postsmolts were all immature both in 1982 and
1983, but differences were observed between
males and females in 1982. Testes averaged
48 ± 6 mg (95% C.L., n = 124) for a mean male
gonadosomatic index value of 0.025% ± 0.003%
(95% C.L., n = 124). There was no significant dif-
ference in the mean value of either parameter
between age groups (P > 0.52). Testis weight in-
creased in time and as body length and body
weight increased, but again there was no differ-
ence between age groups (Table 2). However, the
gonadosomatic index did not change in time
(P = 0.10) or as postsmolts' size increased
(P = 0.16 for body length and P = 0.10 for body
weight), suggesting that changes in size of the
testes were not allometric in male postsmolts in
the autumn period. Regressions were tested and
compared following Snedecor and Cochran (1967)
and Sokal and Rohlf (1969).

There was more variability in the data for fe-

200

DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON

360-.

350

340-

330

320H

310

300-

290-

E
E

I-
O

280
270
260
250
240
230
220
210
200
190-
180-

• *♦ • .

* A *

01 06 11 16 21 26 31 05 10 15 20 25 30 05 10 15 20 25 30

AUGUST SEPTEMBER OCTOBER

DATE

Figure 2. — Seasonal changes in length of Atlantic salmon postsmolts collected in 1982 and 1983.

Table 2. — Functional regressions of gonad weight on date, body length, and
body weight of Atlantic salmon male postsmolts {n = 124). Lengths and
weights are log-transformed.

Date

Length (mm)

Weight (g)

Regression
Covariance
(age groups)

P< 0.00011
P = 0.31 n.s.

P< 0.00012
P = 0.56 n.s.

P< 0.00013
P = 0.47 n.s.

1/ = 0.0168 X- 5.41 66
2y = 7.8496 X - 20.3350
3V = 2.6798 X- 7.5298.

201

FISHERY BULLETIN: VOL. 86, NO. 2

460
440

420
400
380
360
340
320
300
280
260
240
220
200
180

160
140
120
100

(473) (565) (464)
(488)

01 06 11 16 21 26 31 05 10 15 20 25 30 05 10 15 20 25 30

AUGUST

SEPTEMBER
DATE

OCTOBER

Figure 3. — Seasonal changes in weight of Atlantic salmon postsmolts coUecteci in 1982 and 1983.

males. Though the size of ovaries did not differ
significantly between age groups (P = 0.07), go-
nadosomatic index values increased in older fe-
male postsmolts (P = 0.0006). Mean values
ranged from 133 mg (0.06%) for 2-yr-old post-
smolts to 205 mg (0.10%) for 5-yr-old postsmolts
(Table 3). The regressions of ovary weight on body
length, body weight, and time were all significant
except for 5-yr-old females (Table 4), and differed
between age groups particularly in elevation
(P < 0.05, n - 213). The regressions of gonadoso-
matic index values on the same variables were

Table 3. — Gonad weight (mg) and gonadosomatic index (GSI) (%)
of Atlantic salmon female postsmolts: mean values and confidence
limits (n = 213).

Gonad
Mean

weight

95% C.L.
interval

GSI

Age
group

Mean

95% C.L.
interval

2
3
4
5

133
152
175
205

124-144
142-163
161-189
194-217

0.064
0.077
0.092
0.095

0.060-0.067
0.073-0.082
0.087-0.098
0.091-0.099

202

DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON

not significant indicating that, in females as in
males, changes in size of the gonads were not
allometric (body length: P = 0.06; body weight:
P = 0.73; time: P - 0.68).

Scale Marks

Scales used for age determination were also ex-
amined for any mark that could be of use in stock
identification. Many individuals exhibited a tran-
sition zone on their scales. Circuli in this zone
were more wide-spaced than circuli laid down in
earlier summers, but they were more narrow-
spaced than circuli formed in summer in a marine
environment. This mark was present on 73% of
the scales in both males and females (72.6 and
73.87f) but tended to decrease as smolt age in-
creased: 84% (age 2 years), 75% (3 years), 68% (4
years), and 62% (5 years). However, the trend was
not significant (x^ = 5.74, P > 0.10).

Scales also exhibited summer checks in 1982,
1983, and 1984. The number of circuli between
the smolt mark and the summer check was
10.9 ± 2.6 (mean ± SD) (range 7-15) in 1982 and
9.7 ± 2.2 (range 5-15) in 1983. The ratio between
the radius to summer check and the radius to
smolt mark averaged 2.00 ± 0.30 (range 1.49-
2.45) in 1982, and 1.67 ± 0.19 (range 1.26-2.32)
in 1983. The overall percentage of occurrence was
low in 1982 (3.6%) and involved only specimens
collected in the area Bale Trinite-Pointe aux
Anglais where 11 out of 13 postsmolts examined
showed a summer check. The incidence of sum-
mer checks on scales increased markedly in 1983
(55%) and 1984 (75%), and summer checks were
no longer restricted in distribution. Examination
of data concerning postsmolts possessing a sum-

mer check showed no relationship between the
check and measured biological variables except
perhaps in 1982. The gonads of males having a
summer check on their scales in = 5) were heav-
ier than in males having no summer check (120
mg and 48 mg respectively). Their gonadosomatic
index was also higher (GSI = 0.042% and 0.025%
respectively). This was not so in females. Infor-
mation on gonads is not available for 1983, 1984,
and 1985.

Back-calculated lengths at summer check for-
mation were larger in 1982: 22 ± 2.5 cm in 1982
and 20 ± 2 cm in 1983. Postsmolts measured 265
mm on 1 September. With a mean rate of increase
in length of 1.65 mm/day, the check must have
been formed some 26 (1982) to 40 (1983) days
earlier, i.e., in late-July. This is confirmed by ex-
trapolation of the length-date plot (Fig. 2). No
salmon were noted possessing 2 summer checks
on their scales.

Food Items

Drastic changes took place in prey selection be-
tween postsmolts collected in summer and in au-
tumn. The stomach contents of 40 salmon ranging
up to 188 mm in length (70 g) collected in the
second half of July 1984 in the estuary of the
Grande Trinite River, near Bale Trinite, revealed
a low diversity in prey items, indicative of a tran-
sition period between the riverine and marine en-
vironment. Small fish remains were observed in
only 5 out of 39 stomachs containing food re-
mains. Invertebrates were observed in 38 stom-
achs, classes Insecta and Crustacea occurring in
87 and 92% of the stomachs respectively. Food
items consisted mainly of 2 families: Chironomi-

Table 4. - Functional regressions of gonad weight on date, body length, and
body weight of Atlantic salmon female postsmolts (n =213). Lengths and
weights are log-transformed.

Significance

Functional regression

Date

Length
(mm)

Weight (g)

2: P = 0.0001

3 : P = 0.0001

4 : P = 0.0058

5 : P = 0.26 n.s.

covariance : P

2
3
4
5

2
3
4

P < 0.0001
P < 0.0001
P < 0.0001
P = 0.34 n.s.
covariance

P < 0.0001
P < 0.0001
P < 0.0001
P = 0.28 n.s.
covariance

Y = 0.0075 X

Y = 0.0044 X

Y = 0.0052 X

2.7127
1.9052
2.0229

0.04

y = 3.8015 X- 10.1089
y = 2.9105 X - 7.8659
/ = 3.1893 X -8.4701

P = 0.02

Y= 1.31 10 X- 3.9428
Y = 1 .0385 X - 3.2227
Y= 1.1763 X- 3.4659

P = 0.04

203

FISHERY BULLETIN: VOL. 86, NO. 2

dae (95% of insects by number) and Gammaridae
(92% of crustaceans by number).

Later in summer and autumn, postsmolts con-
sumed mainly small fish. Stomach contents were
analyzed for 373 out of 385 postsmolts collected in
1982. There were 109 stomachs with no food re-
mains (29%). They were most prevalent in the
first half of August: 46%, 1-15 August; 25%, 16-
31 August; 26%, 1-15 September; 25%, 16 Sep-
tember-31 October. Fishes dominated the list of
prey items as they occurred in 238 out of 264
(90%) stomachs containing food remains, includ-
ing 200 postsmolts (84%) that fed exclusively on
small fish. Fishes could be identified in 157 stom-
achs. Diversity was low, capelin, Mallotus villo-
sus, occurring in 16 stomachs (10%) and sand
lance, Ammodytes americanus, in 145 stomachs
(92%). Ammodytes americanus dominated in
terms of percentage by number (94%). Postsmolts
consumed smaller A. americanus, in the 40-100
mm range (Fig. 4). Stomachs examined in 1983,
1984, and 1985 confirmed these observations.

Invertebrates were found in 69 out of 264 stom-
achs (26%) containing food remains. Eighteen
postsmolts had only invertebrates in their stom-
achs (26%). In contrast with smolts in the estua-
rine environment, postsmolts did not rely on in-
sects; insects occurred in only 8 stomachs (3%),
whereas crustaceans occurred in 61 stomachs
(23%) (respectively, 12 and 88% by number). Fur-
thermore, the class Amphipoda no longer domi-
nated the crustaceans (Table 5).

>-
O

z
u

oc
u.

20 40 60 80 100 120 140 160
LENGTH - CLASS (mm)

Figure 4. — Length distribution of sand lance in the stomachs of
Atlantic salmon postsmolts collected in 1982.

Horizontal and Vertical Position
in the Nets

Postsmolts were not randomly distributed in
the nets in 1984. They occurred most frequently
(78% of the individuals) in the top half of the nets
(binomial test; z = 4.62, P < 0.001; Siegel 1956).
Furthermore, 25 out of 74 salmon occurred alone
in the nets, but many also occurred in clusters.
Positions of postsmolts are indicated by number
of nearest buoy on the head-line for those sets
having taken 2 salmon and more (stations visited
at 2-h intervals usually but also nets set
overnight) (Table 6). Distributions are likely not
random in sets 3, 6, 7, 9, 10, and 11 and most
particularly in sets 13 and 15. The catch was low
in midday: 3 salmon between 0900 and 1200 and
5 between 1200 and 1500. This increased to 12
between 1500 and 1800. The majority were
caught later than 1800 (36) and in the morning
between 0600 and 0900 (20).

Finally, positions in the nets were analyzed in
terms of selectivity. The gear used in 1982 could
not be controlled. Fishermen reported using

Table 5. — Crustacean organisms in the stomachs of Atlantic
salmon postsmolts collected in the penod August-October 1982,
based on 39 stomachs containing identifiable crustaceans.

Crustacean
order

Percentage

of
occurrence

Percent-
age by
number

Main organisms

Euphausia-
cea

Decapoda

Amphipoda

24
8

Meganyctiphanes norvegica
Thysanoessa inermis
Thysanoessa raschi

Chionoecetes opilio
(larvae)

Table 6. — Positions of Atlantic salmon postsmolts in nets by num-
ber of nearest buoy on head-line for catches of 2 salmon and more.

Set

No. of

no.

smolts

Number of nearest buoy to each smolt

42 44

23 23

18 22 42

12 13 13 13 14 14

204

DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON

62 mm stretched mesh nets (range 57-70 mm);
mesh size is regulated. The nets used in 1983,
1984, and 1985, had 5 sections of increasing
stretched mesh sizes: 50.8, 57.2, 63.5, 69.9, and
76.2 mm. Catches were recorded for individual
mesh sizes in 1983 combining the 4 stations.
Modal length increased only slightly as mesh size
increased. Furthermore, the distribution for the
69.9 mm mesh was skewed to the right, indicat-
ing no larger sized postsmolts were present.
There was also no catch in the 50.8 mm mesh, but
this may have been because of the small mesh
section being made of a coarser material in 1983.
Postsmolts were frequent in = 49) in the 57.2 mm
mesh, but the distribution for this mesh does not
suggest the existence of smaller postsmolts as the
size range is similar to that of the 63.5 mm mesh
and covers the size range for the 5 meshes com-
bined. Catches declined from a maximum in the
63.5 mm mesh (56) to 40 and 9 in the 69.9 and
76.2 mm mesh, respectively. Therefore it is un-
likely that there was any bias except perhaps
against smaller postsmolts.

Fall Movement out of Coastal
Reaches

Postsmolts gradually left the nearshore area in
late-September. Fishermen in 1982 started col-
lecting postsmolts in mid-August. Their bycatch
declined from mid-September and was low in Oc-
tober. The catch declined partly because most
commercial fishermen were asked to return no
more than 20 salmon each, and they reached this
limit early (Table 7). Fishing was initiated later
in 1983 and took place over a shorter period (23
September-11 October), but a 68^ decline in the
catch was observed between the period 29 Sep-
tember-5 October and the period 5-11 October.
Finally in 1984, postsmolts came near the coast
near the end of August and moved out in mid-
September (Table 8) so that no salmon were
caught in the period 20 September-20 October.

This movement out of the nearshore zone was
associated with decreasing near-surface tempera-
tures in autumn. Temperatures measured near
the surface closely followed the mean air temper-
atures recorded in Sept-Iles (Fig. 5). Since most
postsmolts travelled near the surface, mean air
temperatures were assumed to reflect prevailing
conditions for postsmolts. In 1982, postsmolts
were abundant until mean air temperatures de-
clined below 5°C, i.e., in early October. The situa-
tion was similar in 3 stations out of 4 in 1983,

particularly in the Bay of Matamek River near
Sept-Iles and in Riviere Pentecote, 2 stations
closer to our monitoring station. In 1984, this de-
cline in mean air temperature occurred earlier
(26 September), but surface temperatures were
also lower than air temperatures, and the catch
declined (Table 8) as soon as surface tempera-
tures fell below 5°C (mid-September.) Hence, low
temperatures were associated with a movement
of postsmolts out of the nearshore zone. However,
the reverse is not true: results for 1983 and 1984
suggest that postsmolts do not necessarily move
towards nearshore zones when prevailing tem-
perature conditions are favorable.

Variations in seasonal abundance are masked
by a general decline in salmon abundance near
shore from 1982 to 1985. Fishing effort could not
be assessed in 1982, but the incidence of
postsmolts in coastal herring nets in 1982 was
such that it can reasonably be concluded they
were more abundant than in 1983, 1984, or 1985.
Relative abundance can be assessed for 1983,
1984, and 1985. Bale Trinite and Port Cartier
stations were operated in 1983 and 1984 using
similar nets at the same locations each year and
showed that postsmolts were more abundant in
1983 than in 1984 (Table 9). Finally, a station
was monitored near Sept-Iles in 1985 in an area
where the best catches were made in 1982, using

Table 7. — Time distribution of the commercial
bycatch of Atlantic salmon postsmolts in 1982.

Period

Number

10-31 August
01-15 September
16-30 September
01-15 October
16-31 October

210

118

Table 8. — Catch of Atlantic salmon postsmolts by
period and locality in 1984.

Locality

Baie

Port

Period

Trinite

Cartier

21-24 August

24-27 August

27-30 August

30 August-02 September

02-05 September

05-08 September

08-11 September

11-14 September

14-17 September

17-20 September

20 September-20 October

205

-5

15
10

P 5

uj 0
cc

£ 20
Q.

-5

FISHERY BULLETIN: VOL. 86, NO. 2

1982

Sept-lles

air temperature

^ 05 10 15 20 25 30 04 09 14 19 24 29 04 09 14 19 24 29

1983

Port Cartier
sea temperature

05 lb i5 20 25 30 04 09 14 19 24 29 04 09 14 19 24 29

1984

Bale Trinite
sea temperature

05 io i5 20 25 30 04 09 14 19 24 29 04 09 14 19 24 29

1985

Sept-lles

sea temperature

I r

I ~

05 10 15 20 25 30 04 09 14 19 24 29 04 09 14 19 24 29

AUGUST

SEPTEMBER
DATE

OCTOBER

Figure 5. — Mean daily air temperature in Sept-lles (— -), and mean daily sea temperature at 0.5 m ( — ) in Port Cartier
(1983), at 6 m (---) in Baie Trinite (1984), and at 0.5 m (---) and 6 m ( ) in Sept-lles (1985).

206

DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON

Table 9. — Relative abundance in terms of catch per unit of eHort
(CPUE) at 2 stations run in 1983 and 1984 using similar nets at tfie
same sites.

Baie Trinite

Port Cartier

Period

Catch

CPUE

Catch

CPUE

1983

23 Sept.- 11 Oct.

38.5

1.4

20.0

2.0

1984

21 Aug. -31 Aug.

20.5

0.1

32.5

0.2

30 Aug. -17 Sept.

54.5

0.3

73.0

0.8

17 Sept.-23 Sept.

22.0

—

17.0

—

23 Sept.- 11 Oct.

55.5

—

45.5

—

11 Oct.-18 Oct.

9.5

—

21.0

—

'CPUE: catch of 1 net in 24 hours.

similar nets to 1983 and 1984. The catch was
nearly nonexistent: 5 salmon for 120 unit effort
( 1 unit effort, is 1 net x 24 hours). Thus, based on
limited observations, numbers of salmon near
shore in summer and autumn seem to be highly
variable from year to year.

DISCUSSION

Postsmolts of Atlantic salmon stay much longer
near our coasts than is usually believed. Though
early months in the marine environment have
been shown to have a marked influence on
salmon runs 1 and 2 years later (e.g., Christensen
and Larsson 1979; Scarnecchia 1983, 1984),
postsmolt biology has been a neglected area of
investigation. North American smolts are as-
sumed to migrate rapidly out of the estuaries of
their home rivers, to feeding areas located far out
in the North Atlantic east of the Grand Bank
(Templeman 1968; Reddin 1985) and north to
Labrador and Greenland (Saunders 1966; Tem-
pleman 1967). They return 1 or 2 years later to
spawn in home rivers. This study, and early
records, indicate that some postsmolts remain in
coastal areas as late as autumn before moving
offshore. This was clearly suggested by Comeau
(1909) who stated that 0.5-1.5 lb postsmolts were
regularly taken in autumn along the north shore
of the Gulf of St. Lawrence. Fishermen inter-
viewed in 1982 on the north shore of the Gulf of
St. Lawrence, from Pointe-des-Monts to Blanc
Sablon, declared incidental catches of postsmolts,
mainly in the months of August and September.
The majority declared taking postsmolts each
year. Smolts were also shown to linger in estuar-
ies of the north shore of the Gulf of St. Lawrence
(this study; Power and Shooner 1966; Randall and
Power 1979). Huntsman (1939) mentioned their

occurrence in autumn in herring nets near the
mouth of Gaspe Bay. Fall catches also occurred in
New England (Kendall 1935). Recently, smolts
released in New England were caught as post-
smolts in coastal areas of Canada (Meister 1984).
There are stocks in the Bay of Fundy (as the
stocks in the Baltic) that do not go on extensive
migrations in the North Atlantic (Huntsman
1939; Jessop 1976): postsmolts of these stocks are
regularly taken in herring nets in Passa-
maquoddy Bay and off Grand Manan Island (El-
son 1953, 1964; Allen et al. 1972). Hence the pres-
ence of postsmolts near shore in autumn (or in
summer and autumn) is a characteristic of the
marine life of Atlantic salmon in North America.
Timing of migration has been described for
hatchery-reared smolts released in Sweden
(Larsson 1974) and Finland (Jutila and Alapassi
1985). Tags were returned mostly from a distance
of less than 10 km between days 0 and 10, 20-50
km between days 10 and 20, and 50-100 km 2
months past their release, in the brackish waters
of the Gulf of Bothnia (Jutila and Alapassi 1985).
Behavior of postsmolts is similar to that of 1-
and 2-sea-year salmon. Postsmolts in this study
occurred mainly near the surface as indicated by
their distribution in the nets. LaBar et al. (1978)
concluded that smolts migrated near the surface
in the Penobscot estuary. Templeman (1967,
1968) also found salmon to occur near the surface
in the Northwest Atlantic: 62% occurred in the
top 0.6 m and 90% in the top 1.5 m in July and
August 1965. Similarly in 1966, most salmon
were taken in the top 1.5 m, the number caught
decreasing sharply below 0.6 m. Similar observa-
tions were made on Baltic salmon (Carlin and
Lundin 1967; Christensen 1968). There is less in-
formation available on schooling. Postsmolts did
not regularly have a clustered distribution in
nets, but considering that a net does not retain all
the salmon striking it, there were still many in-
stances of salmon moving in schools. Thurow
(1968) came to the same conclusion for older
salmon in the Baltic. Templeman (1967) pre-
sented limited evidence for salmon in the North-
west Atlantic, but reached negative conclusions
later (Templeman 1968). Finally there are lim-
ited data in the literature concerning the rhythm
of activity of salmon in the marine environment.
Christensen and Lear (1980) showed that in West
Greenland best catches occurred early in the
morning (0600-0800), decreased sharply between
0800 and 1000, and were low between 1000 and
1400. The nets were not set between 1400 and

207

FISHERY BULLETIN: VOL. 86, NO. 2

0600. Catches in this study were also low in mid-
day. Either this is a reflection of 2 peaks of activ-
ity, dawn and dusk, as is common in salmonids in
freshwater, or this is possibly due to salmon
avoiding the nets in high light levels. Thus
postsmolts and 1- and 2-sea-year salmon appear
to have a similar behavior at sea. They move in
small schools close to the surface and are possibly
more active at dusk and dawn.

Food items in the stomachs of postsmolts
changed markedly in summer and indicated a low
diversity of prey. This is in contrast to findings
reported for salmon in the Northwest Atlantic.
Grande Trinite River postsmolts had fed mostly
on chironomids and gammarids in late-July.
Baltic salmon postsmolts fed mainly on aerial in-
sects though small fish and crustaceans also oc-
curred in the stomachs of postsmolts in the south-
ern Baltic (reviewed by Christensen and Larsson
1979). Jutila and Toivonen (1985) also found
aerial insects to be the dominant food items in the
stomachs of small postsmolts ( <20 cm) in the Gulf
of Bothnia (Baltic). They observed that post-
smolts were not selective and must have fed near
the surface (20 cm surface layer). Postsmolts col-
lected later in the present study relied mainly on
small sand lance. Insects and gammarids had
been replaced by vertically migrating crus-
taceans such as Meganyctiphanes noruegica
(Kulka et al. 1982). Thurow (1968) estimated 25
cm to be the length threshold for piscivorous feed-
ing by Baltic salmon. In the present study, this
size was likely reached in the first half of August
1982. This coincides with a major change in stom-
ach contents and a high percentage of stomachs
containing no food. The data on postsmolts in
July are too limited to suggest that sand lance
abundance could limit the early success of
postsmolts at sea, but potential relationships in
late summer should be tested as was done for
capelin by Reddin and Carscadden (1981). Data
on 1- and 2-sea-year salmon indicate they will
readily feed on a diversity of prey items, main
items including Arctic squid, Gonatus fabricii;
paralepids, Paralepis coregonoides ; and lantern
fishes (Lampenyctus sp., Notoscopelus sp.) (Tem-
pleman 1967, 1968; Lear 1980). Sand lance and
capelin are dominant items in West Greenland
and on the coast of Newfoundland (Lear 1972,
1980), and on the Grand Bank (Reddin 1985).
Reddin ( 1985) observed major changes in stomach
contents between salmon on the Grand Bank
(sand lance and capelin) and east of the Grand
Bank (Bathylagidae, Paralepis sp., and crus-

taceans), emphasizing that salmon are not selec-
tive predators.

The rate of increase in mean length averaged
1.65 mm/day in the Gulf of St. Lawrence over the
summer period in 1982 and 1983. This value is
based on the hypothesis that smolts migrated to
estuaries in mid-June. Smolt migration took
place in the first half of June in Restigouche River
in the southern Gulf of St. Lawrence (Peppar
1982) and in the second half of June in Grande
Trinite River in the northern Gulf of St.
Lawrence (Caron 1984). Downstream migration
of smolts peaked at various dates in June in West-
ern Arm Brook in western Newfoundland (Chad-
wick 1981). The calculated rate of increase is also
based on a mean back-calculated smolt length of
135 mm. Length at smoltification averaged 125-
130 mm in Grande Trinite River (mean and SD:
127.5 ± 12.3, n = 88, in 1981; 125.8 ± 10.9,
n = 92, in 1982; see also Caron 1984). Matamek
River and Moisie River smolts measured 125-150
mm (Schiefer 1972). They measured 150 mm in
Little Codroy River (Murray 1968) and 174 mm in
Western Arm Brook (Chadwick 1981). There are
no data in the literature concerning the rate of
increase in size of smolts and postsmolts in the
marine environment. Postsmolts in the Bay of
Fundy reached a mean length of 296 mm in mid-
August 1952 (Allen et al. 1972), some 3 cm more
than postsmolts in this study: 265 mm and
306 mm on 1 September and 30 September. How-
ever the high value of the power exponent of the
length-weight relationship as compared with
salmon in Newfoundland and Labrador (Lear
1973) indicates postsmolts were not in poor condi-
tion. There was possibly a decline in the rate of
increase in length in mid-summer as suggested
by the large proportion of scales showing a sum-
mer check (false-annulus) in 1983. This occurred
10 circuli from the smolt check in mid-summer in
postsmolts 20-22 cm in length, i.e., prior to this
study period, and may have been produced as a
response to a shortage of prey or to deteriorating
environmental conditions. Elson (1953) also no-
ticed the frequent occurrence of a slowing of
growth 6-10 circuli out from the last parr an-
nulus. The percentage of occurrence of the check
varied between locations (1982) and between
years (1982 < 1983 < 1984). Therefore it is not
likely to be a response to a change in postsmolt
biology such as a scheduled shift in prey selection.
However, the summer check can be thought of as
a potential tool for stock discrimination. Some
26-32 circuli are formed before the first sea an-

208

DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON

nulus is formed (Lear and Misra 1978) at a length
of 46-50 cm in the first half of April (Munro
1970).

Presence of postsmolts near shore in late sum-
mer in the northern Gulf of St. Lawrence, as re-
ported in this study, and presumably their late
movement out of the Gulf of St. Lawrence indi-
cate that the directional nature of the migration
should be reconsidered. There are some smolts
that do not head towards the high seas as soon as
they reach the estuaries. They seem to roam
nearby unless prevailing conditions are not favor-
able. Temperature can be hypothesized as trig-
gering the late movement of postsmolts out of the
northern Gulf of St. Lawrence. Saunders (1986)
reviewed the thermal biology of Atlantic salmon
and suggested the thermal range for salmon in
the sea is lower than for juvenile salmon in fresh-
water. Salmon occur mainly at temperatures
ranging from 4° to 8°C in the Northwest Atlantic
(Templeman 1968; May 1973; Reddin 1985). Post-
smolt movements out of the nearshore area took
place in a short period as temperature was de-
creasing, between mid-September and mid-
October. Postsmolts were more abundant in 1982
and 1983 as mean air temperature ranged
between 4° and 10°C in early autumn. In 1984,
sea temperature decreased rapidly from more
than 15°C in late-August down to 2°C in mid-
September; postsmolts vanished from the near-
shore area as temperature declined below 4°C.
However in 1985, they did not come near the coast
though sea temperature ranged between 8° and
12°C. Saunders et al. (1975) reported the lethal
temperature of salmon in seawater to be -0.7°C.
This precludes the possibility of salmon over-
wintering in the Gulf of St. Lawrence unless they
return to freshwater, as do some salmon in the
Koksoak River (Cote et al. 1984; Robitaille et al.
1984a, b), or move down to midwater, a behavior
described for salmon in the Baltic in response to
high temperatures (>12°C) at the surface (Aim
1958). Comeau (1909) reported postsmolts found
in the stomach of seals off Pointe-des-Monts in
January and February. Low sea temperature has
been hypothesized as limiting the passage of Kok-
soak River smolts (Ungava Bay) to West Green-
land in some years, thereby resulting in an estu-
arine population (Power 1969, 1981). This
situation might also be hypothesized to occur in
the Gulf of St. Lawrence. For instance in 1983,
mean air temperature, not to mention minimal
temperature, decreased from 4° to 0°C and less in
a short period near the end of October. Masses of

seawater carried by gyres east and west of Anti-
costi Island, and presenting momentarily favor-
able conditions, can get surrounded by masses of
seawater at lower temperature. Should salmon
rely on temperature as a cue for their movement
out of the Gulf of St. Lawrence, then late mi-
grants could not escape as conditions deteriorate.

The origin of postsmolts collected in this study
is not known. They may be a particular subgroup
of some north shore stocks. Postsmolts in this
study smoltified earlier and at a smaller size than
stocks in northern Newfoundland (Chadwick
1981). However, their origin cannot be deter-
mined based on smolt length or smolt age distri-
butions. For instance, there is a general tendency
for increasing smolt age with latitude, but there
is much variability in the data at latitudes below
52°. Data for salmon stocks in rivers near 50°
latitude range from 3 to 4 years (Power 1981).
Furthermore, postsmolts in this study had an age
distribution similar to that of salmon in the Port-
aux-Basques (Newfoundland) drift net fishery.
Port-aux-Basques salmon migrated to rivers all
around the Gulf of St. Lawrence (Belding and
Prefontaine 1938).

Postsmolts in this study may also be from a
particular subgroup of individuals, such as late-
migrant smolts. Power and Shooner (1966) and
Randall and Power (1979) observed remnants of
the smolt migration feeding in river estuaries on
the north shore of the Gulf of St. Lawrence in July
and August. Furthermore coho salmon released
in mid- to late-summer did not leave the general
area of release (Mahnken and Joyner 1973).

Since grilse are known not to migrate as far as
2- and 3-sea-year salmon in the Northwest At-
lantic (Ruggles and Ritter 1980), postsmolts in
this study can be thought of as potential grilse.
However, the only indication in that direction
that we have is the observation that some males
having a summer check on their scales in 1982
had a higher gonadosomatic index than males
having no summer check on their scales (0.042
and 0.025% respectively). However, this is based
on a small number of postsmolts as few salmon
exhibited a summer check in 1982 and, unfortu-
nately, no gonads were preserved in 1983-85.
There are some stocks maturing mainly as grilse
among the north shore stocks, but grilse are
nearly exclusively males in these stocks (Schiefer
1972; Caron 1984). Postsmolts in this study were
62% females. Female grilse are common in New-
foundland (Chadwick 1981; Power 1981). There
are no published data on Anticosti stocks. Poten-

209

FISHERY BULLETIN: VOL. 86, NO. 2

tial grilse have been reported from the Grand
Bank in offshore fisheries in the Northwest At-
lantic (Reddin 1985), and it has been suggested
that stocks maturing as grilse in the Bay of
Fundy may not leave the general area throughout
their entire marine life (Jessop 1976).

Knowledge on the marine biology of Atlantic
salmon postsmolts has been a neglected area of
research. Their continual presence in sea surface
waters could prove to be extremely valuable in
forecasting salmon movements and production
(Chadwick 1982; Scarnecchia 1984). Potential
studies include mortality rates in the period of
transition (July) and the relationship between
low temperatures (3°-4°C) and postsmolt migra-
tion out of the Gulf of St. Lawrence. Biological
data in general are needed to be included in mod-
els forecasting salmon runs in the North Atlantic.

ACKNOWLEDGMENTS

The authors wish to thank Fisheries and
Oceans personnel, including many summer stu-
dents, having assisted in field work on the North
Shore, particularly S. Cloutier, M. Fortin, M.
Laverdiere, Y. Lavergne, B. Legare, B. Mercille,
M. Michaud, M. Poirier, and others. J. Boulva
believed in the need to study postsmolt biology
and actively supported this study. G. Shooner,
G. Morin, and team assisted in scale reading and
in examining preserved postsmolts in 1982 and
1983. J. R. Robitaille assisted in scale data analy-
sis. Fishermen between Pointe-des-Monts and
Blanc Sablon kindly discussed the incidence of
postsmolts in their nets and participated in sam-
pling in 1982 and 1983. P. Bertrand of the Minis-
tere du Loisir de la chasse et de la Peche (Quebec)
shared his knowledge of the bycatch in herring
nets. Many thanks also to the population of Bale
Trinite. E. M. P. Chadwick, G. Power, and
R. L. Saunders kindly reviewed an earlier version
of the manuscript.

LITERATURE CITED

Allan. I R H . and J A Ritter.

1977. Salmonid terminology. J. Cons. int. Explor. Mer
37:293-299.
Allen. K. R.. R. L. Saunders, and P. F. Elson

1972. Marine growth of Atlantic salmon (Salmo salar) in
the Northwest Atlantic. J. Fish. Res. Board Can.
29:1373-1380.
Alm, G

1958. Seasonal fluctuations in the catches of salmon in
the Baltic. J. Cons. int. Explor. Mer 23:399-433.

Anonymous

1984. Report of the Baltic salmon and trout working
group. ICES CM. 1984/ Assess:ll.
Belding, D L., and G Prefontaine

1938. Studies on the Atlantic salmon. II-Report on the
salmon of the 1937 Port-aux-Basques (Newfoundland)
drift-net fishery. Contrib. Inst. Zool. Univ. Montreal 3,
58 p.

Carlin, B

1959. Results of salmon smolt tagging in the Baltic area.
Rapp. P.-v. R^un. Cons. int. Explor. Mer 147:89-96.
Carlin, B., and H.-E Lundin

1967. The size and position of the salmon caught in drift
nets in the Baltic. Swed. Salmon Res. Inst. Rep. 14, 6

P-
Caron. F.

1983. Migration vers I'Atlantique des post-saumoneaux
(Salmo salar) du Golfe du Saint-Laurent. Nat. can.
(Rev. Ecol. Syst.) 110:223-227.

1984. Rapport d'op6ration de la riviere Trinite, 1984.
Gk)uvemement du Qu6bec, Ministere du Loisir, de la
Chasse et de la Peche, Direction gen^rale de la faune, 26

Chadwick, E M. P

1981. Biological characteristics of Atlantic salmon smolts
in Western Arm Brook, Newfoundland. Can. Tech. Rep.
Fish. Aquat. Sci. 1024, iv + 45 p.

1982. Dynamics of an Atlantic salmon stock (Salmo
salar) in a small Newfoundland river. Ph.D Thesis,
Memorial Univ. Newfoundland, St. John's, 267 p.

Christensen, O.

1968. Report on an experimental fishery with salmon
drift-nets. ICES CM. 1968/B:11.

Christensen, O , and P.-O Larsson

1979. Review of Baltic salmon research. ICES Coop.
Res. Rep. 89, 124 p.

Christensen O., and W. H. Lear.

1980. Distribution and abundance of Atlantic salmon at
West Greenland. Rapp. P.-v. R6un. Cons. int. Explor.
Mer 176:22-35.

Comeau, N. a.

1909. Life and sport on the north shore of the lower St.
Lawrence and Gulf Telegraph Printing Company, Que-
bec.
COTfi, Y., I. Babos, and J A. Robitaille.

1984. Caract6ristiques scalim6triques des saumons du
Koksoak (Ungava, Quebec). Nat. can. (Rev. Ecol. Syst.)
111:401-409.
Elson, P F

1953. Growth and migration of PoUett River salmon. St.

Andrews Biol. Stn. Annu. Rep., p. 146-148.
1964. Post-smolt Atlantic salmon in the Bay of Fundy.
St. Andrews Biol. Stn. Annu. Rep., E20-E23.
Huntsman, A G.

1939. Conference on salmon problems. Publ. Am. Assoc.
Adv. Sci. 8:86-106.

Ikonen, E., and H Auvinen

1984. Migration of salmon in the Baltic Sea, based on
Finnish tagging experiments. ICES CM. 1984/M:4.

1985. Migration of salmon post-smolts (Salmo salar) in
the Baltic Sea. ICES CM. 1985/M:19.

Jessop. B M

1976. Distribution and timing of tag recoveries from native
and normative Atlantic salmon (Salmo salar) released
into Big Salmon River, New Brunswick. J. Fish. Res.
Board Can. 33:829-833.

210

DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON

JUTILA. E.. AND T ALAPASSI.

1985. Recaptures of salmon post-smolts (Salmo salar)
during the first summer after release in Finnish tagging
experiments. ICES CM. 1985/M:28.
JUTILA. E , AND J TOIVONEN

1985. Food composition of salmon post-smolts (Salmo
salar) in the northern part of the Gulf of Bothnia. ICES
CM. 1985/M:21.
Kendall. W C

1935. The fishes of New England, the salmon family.
Part 2. The salmons. Mem. Bos. Soc. Nat. Hist. 9:1-
166.
KULKA. D W , S Corey, and T D Iles

1982. Community structure and biomass of euphausiids
in the Bay of Fundy. Can. J. Fish. Aquat. Sci. 39:326-
334.
LaBar, G W , J D McCleave. and S M Fried.

1978. Seaward migration of hatchery-reared Atlantic
salmon iSalmo salar) smolts in the Penobscot River estu-
ary, Maine: open-water movements. J. Cons. int. Ex-
plor. Mer 38:257-269.

Larsson. P -O

1974. Migration of the Swedish west coast salmon stocks.

Swed. Salmon Res. Inst. Rep. 3, 11 p.
1985. Predation on migrating smolt as a regulating factor
in Baltic salmon, Salmo salar, populations. J. Fish Biol.
26:391-397.
Larsson, P -O., and S Ateshkar.

1979. Laxsmoltens vandring fraan Lulealven.
Fiskeritidskrift Finl. 1:8-9.

Lear, W H

1972. Food and feeding of Atlantic salmon in coastal areas
and over oceanic depths. ICNAF Res. Bull. 9:27-39.

1973. Size and age composition of the 1971
Newfoundland-Labrador commercial salmon catch.
Fish. Res. Board Can. Tech. Rep. 392, 43 p.

1980. Food of Atlantic salmon in the West Greenland-
Labrador Sea area. Rapp. P. -v. Reun. Cons. int. Explor.
Mer 176:55-59.

Lear, W H , and R K Misra

1978. Clinal variation in scale characters of Atlantic
salmon (Salmo salar) based on discriminant function
analysis. J. Fish. Res. Board Can. 35:43-47.
Mahnken, C, and T Joyner

1973. Salmon for New England fisheries. Ill: Developing
a coastal fishery for Pacific salmon. Mar. Fish. Rev.
35(10):9-13.
May, AW

1973. Distribution and migrations of salmon in the
Northwest Atlantic. Int. Atl. Salmon Found. Spec.
Publ. Ser. 4:373-382.
Meister. a L

1984. The marine migrations of tagged Atlantic salmon
(Salmo salar) of USA origin. ICES CM. 1984/M:27.
Munro. W R

1970. Notes on the salmon long-lining cruise by the R.V.
"Jens Chr. Svabo" off Faroe, April 1969. ICNAF Res.
Doc. 70/40, 8 p.
Murray, A R

1968. Smolt survival and adult utilisation of Little Cod-
roy River, Newfoundland, Atlantic salmon. J. Fish.
Res. Board Can. 25:2165-2218.
Peppar, J L

1982. Atlantic salmon smolt investigations, Restigouche
River system. New Brunswick. Can. MS Rep. Fish.
Aquat. Sci. 1641, vii + 15 p.

Power, G.

1969. Salmon of Ungava Bay. Arct. Inst. N. Am. Tech.

Pap. 22, 72 p.
1981. Stock characteristics and catches of Atlantic
salmon (Salmo salar) in Quebec, and Newfoundland, and
Labrador in relation to environmental variables. Can.
J. Fish. Aquat. Sci. 38:1601-1611.
Power, G., and G. Shooner.

1966. Juvenile salmon in the estuary and lower Nabisipi
River and some results of tagging. J. Fish. Res. Board
Can. 23:947-961.
Randall, R G , and G Power.

1979. Atlantic salmon (Salmo salar) of the Pigou and
Bouleau Rivers, Quebec. Environ. Biol. Fish. 4:179-
184.

Reddin, D. G.

1985. Atlantic salmon (Salmo salar) on and east of
the Grand Bank. J. Northwest Atl. Fish. Sci. 6:157-
164.

Reddin, D. G., and J. E Carscadden.

1981. Salmon-capelin interactions. Can. Atl. Fish. Sci.
Adv. Comm. Res. Doc. 81/2, 38 p.
RoBiTAiLLE, J A . Y. Cote, G. Hayeur, and G. Shooner.

1984a. Particularit^s de la reproduction du saumon atlan-
tique (Salmo salar) dans une partie du reseau Koksoak,
en Ungava. Rapp. tech. can. sci. halieut. aquat. 1313,
vii + 33 p.
1984b. Croisssance estuarienne du saumon atlantique
(Salmo salar) dans le fleuve Koksoak, en Ungava.
Rapp. tech. can. sci. halieut. aquat. 1314, vii + 23 p.
RUGGLES. C P., AND J. A RlTTER

1980. Review of North American smolt tagging to assess
the Atlantic salmon fishery off West Greenland. Rapp.
P.-v. Reun. Cons. int. Explor. Mer 176:82-92.

Saunders, R L.

1966. Some biological aspects of the Greenland salmon
fishery. Atl. Salmon J., Summer: 17-23.

1986. The thermal biology of Atlantic salmon: influence
of temperature on salmon culture with particular
reference to constraints imposed by low tempera-
ture. Inst. Freshwater Res. Drottningholm Rep. 63,
38 p.

Saunders. R L . B C Muise, and E. B. Henderson.

1975. Mortality of salmonids cultured at low temperature
in seawater. Aquaculture 5:243-252.
Scarnecchia. D L

1983. Age at sexual maturity in Icelandic stocks of At-
lantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci.
40:1456-1468.

1984. Climatic and oceanic variations affecting yield of
Icelandic stocks of Atlantic salmon (Salmo salar). Can.
J. Fish. Aquat. Sci. 41:917-935.

SCHIEFER, K

1972. Ecology of Atlantic salmon, with special reference
to occurrence and abundance of grilse, in north shore
Gulf of St. Lawrence rivers. Ph.D Thesis, Univ. Water-
loo, Ontario, 129 p.

SlEGEL, S

1956. Nonparametric statistics for the behavioral sci-
ences. McGraw-Hill Book Co., N.Y., 312 p.
Snedecor. G W . AND W G. Cochran

1967. Statistical methods. Iowa State Univ. Press,
Ames, 10, 593 p.

SOIKKELI, M.

1973. Tagged salmon smolts in the diet of the Caspian
tern. Laxforskningsinst. Medd. 3:1-5.

211

FISHERY BULLETIN: VOL. 86, NO. 2

SoKAL, R R , AND F J RoHLF. 5:62-85.

1969. Biometry. The principles and practice of statistics Thurow, F.

in biological research. Freeman and Co., San Francisco, 1968. On food, behavior and population mechanisms of

CA., 776 p. Baltic .salmon. Swed. Salmon Res. Inst. Rep. 4, 16 p.

Templeman, W. Valle, E

1967. Atlantic salmon from the Labrador Sea and off West 1985. Predation of birds on salmon and sea trout smolts
Greenland, taken during A. T. Cameron cruise, July- and post-smolts. ICES CM. 1985/M:22.

August 1965. ICNAF Res. Bull. 4:5-40. ViBERT, R

1968. Distribution and characteristics of Atlantic salmon 1953. Voyages maritimes des saumons et retour h la riv-
over oceanic depths and on the bank and shelf slope areas iere natale. Bull. Fr. Piscic. 170:5-23.

off Newfoundland, March-May 1966. ICNAF Res. Bull.

212

SIZE AND DIET OF JUVENILE PACIFIC SALMON
DURING SEAWARD MIGRATION THROUGH A SMALL ESTUARY

IN SOUTHEASTERN ALASKA

Michael L. Murphy, John F. Thedinga, and K V. Koski'

ABSTRACT

To assess competition and predation among juvenile Pacific salmon iOncorhynchus spp.) migrating
through the estuary of Porcupine Creek, a small stream in southeastern Alaska, their size and diet
were determined in 1979 and 1981. Mean fork length (FL) during May and June increased from 32
to 73 mm (1.5 mm/day) for pink salmon, O. gorbuscha; from 39 to 51 mm (0.4 mm/day) for chum
salmon, O. keta; and during June and July, from 99 to 165 mm (1.6 mm/day) for coho salmon, O.
kisutch. Prey, in order of importance, included larval fish (mostly Gadidae), larval molluscs (Mesogas-
trofKxla), and calanoid copepods for pink salmon; larval molluscs, larvaceans, and hyperiid amphipods
for chum salmon; and fish (Clupea harengus pallasi, Ammodytes hexapterus, and Gadidae), insects,
and larval decapods (Brachyrhyncha) for coho salmon. No pink or chum salmon were found in the coho
salmon stomachs. Prey size for pink and chum salmon was similar (median, 0.4 mm long for both
species), and much smaller than that of coho salmon (median, 2.3 mm). Diet overlap was greater
between pink and chum salmon than between either species and coho salmon. Pink salmon, however,
ate almost exclusively (95%) pelagic prey, whereas chum salmon ate both pelagic (74%) and epiben-
thic (26%) prey. Rapid ecirly growth and differences in diet probably help minimize predation and
competition among salmon during seaward migration.

The early marine life stage of juvenile Pacific
salmon iOncorhynchus spp.), during transition
from freshwater to seawater, is important in de-
termining brood-year survival and subsequent
adult returns (Manzer and Shepard 1962; Parker
1968); their survival rate is lowest during this
time (Parker 1968; Bax 1983). Salmon often
school in large concentrations in estuaries as they
migrate seaward, and are more likely to deplete
food supplies and compete for food than after they
disperse to the sea (Bailey et al. 1975; Feller and
Kaczynski 1975). Survival depends on size
(Parker 1971; Healey 1982), and competition for
food can depress early growth (Peterman 1984)
and prolong vulnerability to predators (Taylor
1977; Walters et al. 1978). Size and diet of juve-
nile salmon in an estuary, therefore, determine
the potential for predation and competition and
can greatly affect survival.

As salmon aquaculture expands and more juve-
nile salmon are released into estuaries, competi-
tion and predation among salmon may increase
(Johnson 1974). To optimize hatchery production
and avoid adversely affecting wild stocks, an

1 Northwest and Alaska Fisheries Center Auke Bay Labora-
tory, National Marine Fisheries Service, NOAA, P.O. Box
210155, Auke Bay, AK 99821.

Manuscript accepted November 1987.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.

understanding is needed of how different stocks of
salmon interact in estuaries. This paper compares
size and diet of juvenile pink, O. gorbuscha;
chum, O. keta; and coho, O. kisutch, salmon
to assess potential predation and competition
between the species during their seaward mi-
gration through the estuary of a small, pristine
stream.

STUDY AREA

This study was conducted in the estuary of Por-
cupine Creek, the only salmon stream flowing
into Steamer Bay in southeastern Alaska (Fig. 1).
The estuary (about 5.5 km long) consists of a 1.5
km stream reach that is periodically inundated
by tides, and a 4 km series of three estuarine
basins. At low tide, the inner and middle basins
are small (2 and 7 ha, respectively) and shallow
(14 and 16 m, respectively) compared with the
outer basin (120 ha and 42 m deep). The littoral
zone ranges from low-gradient mudflats to steep
cobble. Bottoms of the basins are level and com-
posed of shell, gravel, and mud.

During low tide, the inner and middle basins
are partially isolated from the outer basin and the
main part of Steamer Bay by tidal rapids 1-3 m
deep. Salinity is lower in the inner and middle

213

FISHERY BULLETIN: VOL. 86, NO. 2

Figure 1. — Aerial photo of study site in the inner part of Steamer Bay, southeastern Alaska,
showing the Porcupine Creek estuary at low tide and location of smolt traps used by Thedinga

(1985).

basins (24-29%c) than in the outer basin (28-
30%c), but temperature does not differ between
basins in spring and summer (11°-13°C from May
to September 1981). Heavy tidal flushing, partic-
ularly during spring tides, results in a diverse
community within the estuary; e.g., eel grass,

Zostera; Dungeness crabs, Cancer magister; bull
kelp, Nereocystis; and rock scallop, Hinnetes. A
detailed description of the study area is in Merrell
and Koski (1978) and Koski (1984).

Porcupine Creek, upstream of tidal influence, is
4.5 km long and has an average discharge of

214

MURPHY ET AL.: SIZE AND DIET OF JUVENILE PACIFIC SALMON

about 0.5 m"^/second. Its watershed is forested by
mature western hemlock, Tsuga heterophylla ,
and Sitka spruce, Picea sitchensis. Annually,
5,000-75,000 adult pink salmon and 200-4,000
chum salmon spawn in the creek from late July to
October, and 250-600 adult coho salmon spawn
from late September to November (Koski 1984).
Pink and chum salmon fry typically migrate from
Porcupine Creek from late March to mid-May
(Koski^). Coho salmon smolts migrate from late
April to early June, but over 90% usually migrate
in late May (Thedinga 1985).

METHODS

Six stations, one each on the east and west sides
of the three basins (Fig. 1) were sampled by a
beach seine 37 m long, with 1.6 cm stretch mesh
on the wings, and a central bag of 6 mm stretch
mesh. The seine tapered from 2 m deep at the
central bag to 1 m deep at each end. In 1979, only
one station in each basin was seined about every
4 days from 16 May to 12 June. In 1981, all six
stations were seined biweekly from 26 May to
7 July and monthly thereafter through 11
November. Seines were set parallel to and about
40 m from shore by a skiff, and retrieved from
shore. Setting and retrieval were accomplished
within 10 minutes.

All fish caught were identified and counted.
Fork lengths (FL) were measured to the nearest
millimeter from a random sample of <25 salmon
per species, station, and sampling period. Stom-
ach contents were collected only in 1981 from <10
salmon per species and station in May, June, and
July. Contents were collected from anesthetized
fish by flushing the stomach with water from a
syringe (Meehan and Miller 1978; Koski and
Kirchhofer 1984) and preserved in 5% formalde-
hyde. Prey were later identified, counted, mea-
sured, and weighed.

For diet analysis, the index of relative impor-
tance (IRI) was calculated, where

IRI = {% number + % weight) (% frequency of
occurrence)

(Pinkas et al. 1971). Diet overlap between salmon
species was calculated (McCabe et al. 1983):

2 2 X'Y,

C =

1 = 1

2xf . J^Yf

1=1

1 = 1

2K V. Koski, Northwest and Alaska Fisheries Center Auke
Bay Laboratory, National Marine Fisheries Service, NOAA,
P.O. Box 210155, Auke Bay, AK 99821, pers. commun. October
1985.

where C = overlap coefficient and X, and Y, are
proportions of the total diet of salmon species X
and Y, respectively, contributed by prey taxon i of
s prey taxa. Diet overlap was calculated sepa-
rately for proportions based on prey number and
weight. Prey were also classified as epibenthic or
pelagic to assess overlap in foraging mode (Feller
and Kaczynski 1975). Epibenthic prey were poly-
chaetes, gammarid amphipods, harpacticoid cope-
pods, barnacle cyprids, and cumaceans. Pelagic
prey were calanoid copepods, euphausiids, barna-
cle nauplii, cladocerans, larvaceans, larval deca-
pods (Brachyrhyncha), hyperiid amphipods, and
fish (eggs, larvae, and juveniles).

RESULTS

Size

In May 1979, pink salmon were the size of
newly emergent fry, about 32 mm FL (Fig. 2).
Average length increased 1.5 mm/day, to 73 mm
on 12 June 1979. In 1981, pink salmon averaged
73 mm FL in late May and early June. Changes
in average FL in 1981 could not be calculated
because most migration occurred before sampling
began.

Average FL of chum salmon increased slower
than that of pink salmon. Mean FL of chum
salmon increased 0.4 mm/day in both years, from
39 mm to 51 mm in 1979, and from 60 mm to 78
mm in 1981 (Fig. 2). Chum salmon averaged
about 10 mm FL longer in 1981 than in the same
period in 1979. Chum salmon were not found in
the estuary after early July, except for two fry
caught in the outer basin in November.

Average FL of coho salmon was nearly con-
stant, between 85 and 110 mm, throughout May
and early June in both 1979 and 1981 (Fig. 2).
Average FL of coho salmon in the estuary during
this period was influenced by an influx of Porcu-
pine Creek migrants, which averaged between 75
and 96 mm FL (Thedinga 1985). After the migra-
tion from Porcupine Creek in 1981 (=9 June),
average FL increased 1.6 mm/day to 165 mm by
20 July. Average FL then decreased to 85 mm in

215

FISHERY BULLETIN: VOL. 86, NO. 2

60
40

•
/

/
/
- /

(a) pink

'l 1 1

1 1 1

E 80

O 60

u
-» 50

Z
O 40

< 30
0)

._^k, ^^ (^ CHUM

220
180
140
100
60

3-^-tH^4

4—.^

"n 1 1 1 1 r

15 MAY 1 JUN 15 JUN 1 JUL 15 JUL 1 AUG

Figure 2.— Length of salmon in 1979 and 1981. Data shown are
means and ranges for pooled samples from all stations on each
sampling date. Data for pink salmon in 1981 are omitted be-
cause of small sample sizes. (▲ = 1979, • = 1981.)

August and 106 mm in September, after most
smelts had left and a few new smelts entered the
estuary.

Diet

A wide variety ef prey was eaten by the three
salmon species, but usually only one or two prey
taxa dominated the diet (Table 1). Pink salmon
ate mostly larval molluscs (Mesogastropoda) and
larval fish (mostly Gadidae) in May, and calanoid
copepods in June. Chum salmon ate mostly larval
molluscs in May; larval molluscs, larvaceans, and
cladocerans in June; and hyperiid amphipods and
larval decapods in July. Coho salmon ate mostly
fish and insects in May and June, and fish and
larval decapods in July. The identifiable fish prey
of coho salmon consisted of 53% Pacific herring,
Clupea harengus pallasi; 45% cod (Gadidae); and

2% Pacific sand lance, Ammodytes hexapterus. No
identifiable pink or chum salmon were in the coho
stomachs. Catch of coho, but not that of the other
salmon, was significantly correlated (r = 0.46,
P < 0.001) with aggregate catch of herring, sand
lance, and cod, indicating that coho salmon con-
gregated near prey schools.

Diet overlap was higher between pink and
chum salmon than between either species and
coho salmon (Table 2). Diet overlap between pink
and chum salmon was especially high in May
when both species ate large numbers of larval
molluscs. If based on prey weight, diet overlap
between pink and coho salmon was negligible. If
based on prey number, however, overlap was
>50% in June when both pink and coho salmon
ate large numbers of calanoid copepods. Diet
overlap between chum and coho salmon was con-
sistently low, especially when based on prey
weight.

Of the 12 most important prey taxa in May and
June, when all 3 salmon species were present in
the estuary, only 4 differed significantly
iP < 0.05) in mean number per stomach between
pink and chum salmon, whereas 9-10 differed
significantly between the two species and coho
salmon (Table 3). Compared with pink salmon,
chum salmon ate more harpacticoid copepods,
cladocerans, and insects. Coho salmon ate fewer
small plankton and more fish than did the other
salmon species. Coho salmon averaged fewer than
20 total prey items, compared to more than 100 in
pink salmon and 200 in chum salmon.

Coho salmon ate larger prey than did the other
salmon (Fig. 3). Median prey length for coho
salmon was 2.3 mm, compared with 0.4 mm for
pink and chum salmon. Coho salmon generally
selected larger individuals of each prey taxon —
particularly larger calanoid copepods, gammarid
amphipods, euphausiids, and larval decapods —
than did pink and chum salmon (Table 4). Offish
prey, coho salmon ate mostly juveniles, whereas
pink and chum salmon ate mostly eggs and lar-
vae.

As they grew larger, all three salmon species
selected larger prey. Numbers of hyperiid am-
phipods, euphausiids, and fish larvae — all rela-
tively large prey — were positively correlated
with FL of pink or chum salmon, whereas num-
bers of cladocerans and larvaceans — both rela-
tively small prey — were negatively correlated
with chum salmon FL (Table 5). Numbers of
calanoid copepods and fish were positively corre-
lated with coho salmon FL, whereas the number

216

MURPHY ET AL.: SIZE AND DIET OF JUVENILE PACIFIC SALMON

Table 1 . — Stomach contents of juvenile salmon in Porcupine Creek estuary, 26 May-7 July 1981 . %A/ is percent by
number, %W is percent by wet weight, %F0 is percent frequency of occurrence of fish with prey item /, and %IRI is
percent of total sum of IRI for all prey taxa. IRI = (%/V + %W)%fO. Taxa are omitted if %IRI is s3 for all salmon
species.

Pink salmon

Chum salmon

Coho salmon

Prey taxon

%F0

%IRI

%/V

%F0

%IRI

%/V

%F0

%IRI

26-29 May 1981

Mollusc larvae

100

Calanoida

Harpacticoida

Cladocera

Cumacea

Euphausiacea

Decapod larvae

Shhmp

Crabs

Fish eggs

and larvae

Fish

Insects!

100

Total

100

(Number of stomachs)

(7)

(31)

(27)

9-10 June 1981

Mollusc larvae

100

Calanoida

100

Harpacticoida

Cladocera

Decapod larvae

Shhmp

Crabs

100

Larvacea,

Oikopleura

100

Fish eggs

and larvae

100

Fish juveniles

Insects^

Total

(Number of stomachs)

(4)

(26)

(14)

7 July 1981

Calanoida

—

Hypehidea

—

Decapod larvae

Shrimp

—

Crabs

—

Fish eggs

and larvae

—

Fish juveniles

—

Insects^

—

100

Total

—

(Number of stomachs)

(0)

(7)

(17)

1 Mostly adult Diptera.

of insects was negatively correlated. As a conse-
quence of the selection of larger prey as the
salmon grew, total prey weight increased with
salmon FL, whereas total prey number did not
(Table 5). Although pink and chum salmon ate
prey of similar size, they foraged differently (Fig.
4). Pink salmon consumed about 95% pelagic
prey; chum salmon, only 74%. Individual taxa
changed, but the importance of pelagic prey did
not change significantly between sampling peri-
ods, estuary basins, or salmon FL classes.

Diet of coho salmon, on the other hand, varied
widely depending on salmon FL, date, and loca-
tion (Fig. 4). Pelagic prey increased from 1% of
total prey for coho salmon <80 mm FL to 80% for
those >100 mm FL. Coho salmon ate fewer
pelagic prey in May, when most coho were in the
inner basin and feeding mainly on insects, than
in July when most were in the outer basin and
feeding mainly on fish (Table 1). Analysis of vari-
ance, however, showed that differences between
basins and sampling periods were not significant

217

FISHERY BULLETIN: VOL. 86, NO. 2

Table 2. — Diet overlap (McCabe et al. 1983) based on number (n)
and weight (W) of prey by sampling period for juvenile salmon in
Porcupine Creek estuary, 26 May-7 July 1981. The number of
stomach samples is in Table 1 .

Sampling
period

Pink vs
n

chum
W

Pink vs
n

coho
W

Chum vs. coho
n W

26-29 (vlay
9-10 June
7 July

0.87
0.47

0.40
0.38

0.07
0.54

0.09
0.00

0.13 0.08
0.29 0.00
0.10 0.01

Table 3. — Comparison of mean number of the 12 most important
prey and total of all prey per salmon stomach from Porcupine Creek
estuary, 26 N^ay- 1 0 June 1 98 1 . Means followed by the same letter
are not significantly different (Kruskal-Wallis analysis of variance,
P>0.05) compared within a row and across columns. The number
of stomach samples is in Table 1 .

Prey item

Pink salmon Chum salmon Coho salmon

Mollusc larvae

28 a

Barnacle larvae

Calanoida

40 a

Harpacticoida

Cladocera

10a

Cumacea

Euphausiacea

Decapod larvae

11 a

Larvacea

Fish eggs and

larvae

Fish

0 a

Insects

86 a

12 b

19 b

37 b

1 a

12a

20 a

11 a

Total

115a

216 a

0 b
4 b
6 b

19 b

Table 4. — Mean length (mm) of prey items in salmon stom-
achs from Porcupine Creek estuary, 26 May-7 July 1981.
Mean prey length within prey taxa was significantly greater
for coho salmon than for pink or chum salmon (sign test,
n = 12 means, P=0.03 and P =0.003, for coho salmon vs.
pink and chum salmon, respectively). The number of stom-
ach samples is in Table 1 .

Pink

Chum

Coho

Prey item

salmon

Mollusc larvae

0.5

0.4

Barnacle larvae

0.4

0.7

0.4

Calanoida

1.2

4.8

Harpacticoida

1.3

1.1

1.3

Cladocera

0.6

0.7

Cumacea

2.4

2.5

2.8

Hyperiidea

3.3

2.4

3.0

Gammaridea

1.5

1.8

5.2

Euphausiacea

3.5

3.0

18.0

Decapod larvae

2.8

2.0

3.5

Larvacea

0.8

0.9

Fish, all life

stages

2.0

1.2

23.9

Insects

2.8

2.2

3.5

T — I — I — I — I I I "1 — I — *T — I — T — I — I — r
0.5 2.5 4.5 6.5 8.5 10.5 12.5 >1 4
Prey Length (Interval Midpoint, mm)

Figure 3. — Relative frequencies of length of prey eaten by pink,
chum, and coho salmon in the Porcupine Creek estuary in 1981.
Total prey measured were 687 in 11 pink, 5,634 in 63 chum, and
1,179 in 53 coho salmon.

Table 5. — Spearman rank correlations between number of prey
items and fork length of juvenile salmon. Because of the large
number of correlations tested, significance levels were adjusted by
multiplying the probability P by the number of tests for each salmon
species. The number of stomach samples is in Table 1 .

Prey item

Pink salmon Chum salmon Coho salmon

Mollusc larvae

-0.32

Barnacle larvae

-0.36

Calanoida

-0.51

Harpacticoida

-0.30

Cladocera

-0.06

Cumacea

-0.20

Hyperiidea

-0.35

Euphausiacea

0.84

Decapod larvae

0.43

Larvacea

-0.36

Fish eggs

-0.33

Fish larvae

and juveniles

0.86

Insects

0.61

Total prey

number

-0.40

Total prey

weight

0.85

0.05
-0.02

0.10
-0.03
-0.34*

0.04
-0.37-

0.36*
-0.19
-0.50*
-0.18

-0.04
-0.11

-0.04

0.35*

0.18

0.22

0.33*

0.19

0.04

0.28

0.19

0.15

0.25

0.36*

0.71"

0.09

0.66**

'None present.

'None present in any stomachs.
*Adjusted probability P < 0.05.
**Adjusted probability P < 0.01.

after adjusting for differences in coho salmon FL
(Table 6). Thus, changes in diet were mainly re-
lated to increasing size of coho salmon as they
migrated through the estuary.

DISCUSSION

Both size and diet can affect predation and com-
petition among juvenile salmon in an estuary. A

218

MURPHY ET AL.: SIZE AND DIET OF JUVENILE PACIFIC SALMON

100

60-

40-

-A)

. Pink
' Chum
Coho

^100^

Q.
O

■5. 80
«

<D
Q.

^ 60

« 40-
Q.

26May

9June

7July

100-f
80
60
40
20

Inner Basin

Middle Basin

Outer Basin

O-l

/
/
/

50-59 60-79 80-99

Salmon Length (mm)

> 100

Figure 4. — Number of pelagic prey as percent of total prey
eaten by individual salmon compared between sampling periods
(A), estuary basins IB), and salmon fork length classes (C) in the
Porcupine Creek estuary in 1981. Symbols are means; bars are
±2 SE of the means. Symbols in B and C are the same as in A .
Pelagic prey are defined in the text.

salmon's size mainly influences its vulnerability
to predators, whereas its diet determines poten-
tial competition for food. Size and diet, however,
are not independent. Salmon change their diet as
they grow, which helps relieve competition be-
tween salmon of different size, and a poor diet
slows their growth, which prolongs vulnerability
to predation.

Table 6. — Analysis of variance of percentage pelagic
prey of coho salmon, with sampling period and estuary
basin as factors and salmon fork length as covariate.
Factors, covariate, and interactions were adjusted
simultaneously before assessing significant (Kim and
Kohout 1975).

Source
of variation

Mean
square

Length

Sampling period
Basin
Residual

Total

8,458

1,580

654

900

1,519

9.4
1.8
0.7

0.004
0.184
0.489

Because of similar diets, pink and chum salmon
are potential competitors. Although diets of pink
and chum salmon in the Porcupine Creek estuary
were similar in prey size and some prey taxa,
however, pink salmon fed almost solely on pelagic
prey , whereas chum salmon foraged both pelagi-
cally and epibenthically; such differences may
help reduce competition. Competition probably
also was reduced because, as the salmon grew
larger, they switched to larger prey. Coho salmon
probably did not compete for food with the other
two species because the coho fed on larger, differ-
ent prey.

Rapid early growth of salmon is important in
reducing vulnerability to predators (Parker 1971;
Taylor 1977). For example, hatchery pink salmon
fry raised for 60 days (to 40 mm FL) before release
into an estuary in southeastern Alaska survived
68% better at sea than did fry released immedi-
ately after emergence (Martin et al. 1981).
Marine survival also is higher for year classes of
larger (9-11 cm FL) than for smaller (6-8 cm FL)
sockeye salmon, O. nerka , smolts (Foerster 1954).
Coho salmon smolts from Porcupine Creek in
1978 averaged 99 mm FL and their survival was
twice that of the 1979 smolts, which averaged
only 91 mm FL (Thedinga 1985). Smolt size and
migration timing, however, interact complexly to
influence marine survival of coho salmon (Bilton
1978).

Growth of juvenile salmon in estuaries usually
inferred from changes in mean size

(LeBrasseur and Parker 1964; Healey 1978), but
these estimates are subject to bias. In this study,
changes in mean size of fish in the catches on
successive dates could underestimate real growth
for two reasons: 1) small individuals may have
migrated continuously into the estuary from

219

FISHERY BULLETIN: VOL. 86, NO. 2

freshwater and 2) larger individuals may have
migrated continuously from the estuary to the
sea. Conversely, growth could be overestimated if
predators of salmon selected small individuals
(Parker 1971). In addition, although the inner
and middle basins are semiisolated from adjacent
marine waters during low tide, juvenile salmon
from adjacent waters could easily enter the estu-
ary, especially the outer basin, during flood tide
and mix with salmon from Porcupine Creek.

Estimates of salmon growth in estuaries and
nearshore marine waters are variable, but gener-
ally range between 1 and 2 mm/day. LeBrasseur
and Parker (1964) estimated pink salmon growth
to be 0.9 mm/day during the first 30 days at sea,
and Healey (1978) estimated pink salmon growth
during summer to be 1.0 mm/day; our estimate
was 1.5 mm/day. Our estimate for chum salmon
at 0.4 mm/day was considerably less than that of
Healey (1978) at 1.5 mm/day, also based on
change in mean length; however, our estimate for
coho salmon of 1.6 mm/day was similar to that of
Healey (1978) at 1.2 mm/day. Summer growth
back-calculated from scales of salmon from the
Sea of Okhotsk was about 1.6 mm/day for pink
and chum salmon (Birman 1969).

Because of their initial small size, pink and
chum salmon particularly are vulnerable to
predators including juvenile coho salmon (Parker
1971). Several authors have suggested that a
major share of pink salmon mortality in the first
weeks at sea results from juvenile coho salmon
predation (Parker 1971; Kaczynski et al. 1973;
Hargreaves and LeBrasseur 1985), but such pre-
dation has not been found in field collections.
Parker (1971) demonstrated predation by juve-
nile coho salmon on pink salmon fry in the labora-
tory, and juvenile coho salmon are known preda-
tors of salmon fi:y in fi*eshwater (Hunter 1959;
Koski and Kirchhofer 1984). However, we have
not found any published data that show predation
by juvenile coho salmon on other salmon in estu-
aries or marine waters. Predation by juvenile
coho salmon on pink salmon fiy migrating ft-om
freshwater does occur in the tidal-influenced
reach of Porcupine Creek (Koski and Kirchhofer
1984), but such predation apparently does not ex-
tend into the estuarine basins. Many fishes have
been identified as predators of pink and chum
salmon in estuaries, including Pacific herring
(Thorsteinson 1960), sea-run cutthroat trout,
Salmo clarki; cod, Gadus macrocephalus ; and
sculpin, Leptocottus armatus, (Bax et al. 1977).
We speculate that predation by coho salmon on

salmon fiy may occur only under circumstances
in which the coho salmon are combined with
small fry as they migrate from freshwater.

The period of vulnerability of pink and chum
salmon fry to predation by juvenile coho salmon is
probably relatively short. Within the first 3
weeks after entering the estuary, pink salmon fiy
can grow larger than the prey fish of juvenile coho
salmon. In the laboratory, juvenile coho salmon
ate the smallest pink salmon available and did
not eat any larger than about 50 mm FL (Parker
1971), which coincides with the largest fish eaten
by coho salmon in our study. At a growth of 1
mm/day, pink salmon entering the estuary at 32
mm FL will outgrow predation by coho salmon
smolts in 18 days. In Porcupine Creek, most pink
and chum salmon migrated fi-om the stream sev-
eral weeks before coho salmon, which enables
them to grow large enough to avoid predation by
coho salmon in the estuary. Thus, early migration
and rapid growth of pink and chum salmon fiy
probably are important in reducing predation by
coho salmon.

In the Porcupine Creek estuary, competition
and predation probably were slight. Competition
for food was minimal, as evidenced by the rapid
salmon growth, because of differences in prey and
foraging mode and because regular tidal flushing
probably replenished food supplies, as in Traitors
Cove, AK (Bailey et al. 1975). Natural stocking
levels in the estuary also probably were below
thresholds where competition for food would de-
press survival. Predation by coho salmon on pink
and chum salmon was avoided because the pink
and chum salmon migrated earlier than coho
salmon and rapidly grew too large for the
coho to handle. Thus, in this natural system, com-
petition and predation probably were unimpor-
tant because of moderate stocking levels, rapid
growth, and differences in diet and timing of mi-
grations. In systems with hatchery inputs, how-
ever, stocking levels would probably be higher
and salmon size and timing of migrations differ-
ent than in natural systems, which could increase
competition and predation.

Stocking levels and timing of hatchery releases
of juvenile salmon in estuaries are important in
minimizing competition and predation (Myers
1980). Hatchery releases should avoid combining
large concentrations of pink and chum salmon fi-y
so as not to deplete food supplies. Conversely, re-
leases during low predator abundance and good
growing conditions — high food availability and
warm temperature — could increase grov^rth and

220

MURPHY ET AL.: SIZE AND DIET OF JUVENILE PACIFIC SALMON

survival. Early releases of coho salmon could in-
crease predation on pink and chum fry (Johnson
1974), especially fry <50 mm FL, if the releases
coincide with fry migrations through the estuary.

ACKNOWLEDGMENTS

Our thanks to R. Brodeur, G. Grette,

C. Hawkes, D. Kirchhofer, and R. Walter for help
in study design and field sampling. Thanks also
to J. Hard, J. Landingham, T. Merrell, Jr.,

D. Mortensen, and J. Pella for reviewing the
manuscript. The Fisheries Research Institute of
the University of Washington identified and
measured prey from stomach samples.

LITERATURE CITED

Baily. J. E., B. L. Wing, and C R. Mattson

1975. Zooplankton abundance and feeding habits of fry of
pink salmon, Oncorhynchus gorbuscha , and chum
salmon, Oncorhynchus keta, in Traitors Cove, Alaska,
with speculations on the carrying capacity of the
area. Fish. Bull., U.S. 73:846-861.
Bax, N J.

1983. Early marine mortality of marked juvenile chum

salmon (Oncorhynchus keta) released into Hood Canal,

Puget Sound, Washington, in 1980. Can. J. Fish.

Aquat. Sci. 40:426-435.

Bax, N J , E O. Salo, B. P Snyder, C. A Simenstad, and W. J.

Kinney

1977. Salmon outmigration studies in Hood Canal: a sum-
mary—1977. In W. J. McNeil and D. C. Himsworth (ed-
itors), Salmonid ecosystems of the North Pacific, p. 171-
201. Oregon State Univ. Press, Oregon State Univ. Sea
Grant Program, Corvallis.

BiLTON, H. T

1978. Returns of adult coho salmon in relation to mean
size and time at release of juveniles. Fish. Mar. Serv.
Tech. Rep. 832, 73 p. Dep. Fish. Environ., Pac. Biol.
Stn., Nanaimo, B.C.

BiRMAN, I B

1969. Distribution and growth of young Pacific salmon of
the genus Oncoryhynchus in the sea. Probl. Ichthyol.
9:651-666.
Feller, R. J., and V. W. Kaczynski.

1975. Size selective predation by juvenile chum salmon
(Oncorhynchus keta) on epibenthic prey in Puget
Sound. J. Fish. Res. Board Can. 32:1419-1429.
FOERSTER, R. E.

1954. On the relation of adult sockeye salmon
{Oncorhynchus nerka) returns to known smolt seawEird
migrations. J. Fish. Res. Board Can. 11:339-350.
Hargreaves, N B., and R J LeBasseur

1985. Species selective predation on juvenile pink
(Oncorhynchus gorbuscha ) and chum salmon (O. keta ) by
coho salmon (O. kisutch). Can. J. Fish. Aquat. Sci.
42:659-668.
Healey, M. C

1978. The distribution, abundance, and feeding habits of
juvenile Pacific salmon in Georgia Strait, British Colum-
bia. Fish. Mar. Serv. Tech. Rep. 788, 49 p. Dep. Fish.

Environ., Pac. Biol. Stn., Nanaimo, B.C.

1982. Timing and relative intensity of size-selective
mortality of juvenile chum salmon (Oncorhynchus keta)
during early sea life. Can. J. Fish. Aquat. Sci. 39:952-
957.

Hunter, J G.

1959. Survival and production of pink and chum salmon

in a coastal stream. J. Fish. Res. Board Can. 16:835-

886.
Johnson, R. C.

1974. Effects of hatchery coho on native Puget Sound
stocks of chum salmon fry. In D. R. Harding (editor).
Proceedings of the 1974 Northeast Pacific Pink and
Chum Salmon Workshop, p. 102-109. Dep. Environ.,
Fish. Vancouver, B.C.

PCaczynski, V W , R. J Feller, J. Clayton, and R. J. Gerke.
1973. Trophic analysis of juvenile pink and chum salmon
(Oncorhynchus gorbuscha and O. keta) in Puget
Sound. J. Fish. Res. Board Can. 30:1003-1008.
Kim, J O , AND F J. Kohout.

1975. Analysis of variance and covariance: subprogram
ANOVA. 2d ed. In N. H. Nie, C. H. Hull, J. G.
Jenkins, K. Steinbrenner, and D. H. Brent (editors).
Statistical package for the social sciences, p. 398-
433. McGraw-Hill, N.Y.

KOSKI, K V

1984. A stream ecosystem in an old-growth forest in
southeast Alaska: Part I. Description and characteristics
of Porcupine Creek, Etolin Island. In W. R. Meehan,
T. R. Merrell, Jr., and T. A. Hanley (editors), Fish and
wildlife relationships in old-growth forests. Proceedings,
p. 47-55. Am. Inst. Fish. Res. Biol., Juneau, AK.
Available from J. W. Reintjes, Rt. 4, Box 85, Morehead
City, NC 28557.

KosKi, K V , AND D. A. Kirchhofer.

1984. A stream ecosystem in an old-growth forest in
southeast Alaska. Part FV: Food of juvenile coho salmon,
Oncorhynchus kisutch in relation to abundance of drift
and benthos. In W. R. Meehan, T. R. Merrell, Jr., and
T. A. Hanley (editors). Fish and wildlife relationships
in old-growth forests, Proceedings, p. 81-87. Am.
Inst. Fish. Res. Biol., Juneau, AK. Available from
J. W. Reintjes, Rt. 4, Box 85, Morehead City, NC
28557.

LeBrasseur, R J , AND R R Parker

1964. Growth rate of central British Columbia pink
salmon (Oncorhynchus gorbuscha ). J. Fish. Res. Board
Can. 21:1101-1128.

Manzer, J I., AND M. P. Shepard

1962. Marine survival, distribution and migration of pink
salmon (Oncorhynchus gorbuscha) off the British Colum-
bia coast. H. R. MacMillan Lectures in Fisheries, p.
113-122. Symposium on Pink Salmon 1960, Univ.
British Columbia, Vancouver.

Martin, R. M , W. R. Heard, and A. C. Wertheimer.

1981. Short-term rearing of pink salmon (Oncorhynchus
gorbuscha ) fry: effect on survival and biomass of return-
ing adults. Can. J. Fish. Aquat. Sci. 38:554-558.

McCabe, G. T , JR , W D MUIR, R L. Emmett, and J. T. DURKIN.

1983. Interrelationships between juvenile salmonids and
nonsalmonid fish in the Columbia River estuary. Fish.
Bull., U.S. 81:815-826.

Meehan, W R., and R A Miller.

1978. Stomach flushing: effectiveness and influence on
survival £ind condition of juvenile salmonids. J. Fish.
Res. Board Can. 35:1359-1363.

221

FISHERY BULLETIN: VOL. 86, NO. 2

Merrell, T. R.. Jr., and K V. Koski.

1978. Habitat values of coastal wetlands for Pacific coast
salmonids. In P. E. Greeson, J. R. Clark, and J. E. Clark
(editors). Wetland functions and values: the state of our
understanding, p. 256-266. Proceedings of the Na-
tional Symposium on Wetlands, American Water Re-
source Association, Minneapolis, MN.
Myers, K W. W.

1980. An investigation of the utilization of four study
areas in Yaquina Bay, Oregon, by hatchery and wild
juvenile salmonids. M.S. Thesis, Oregon State Univer-
sity, Corvallis, 234 p.
Parker, R R

1968. Marine mortality schedules of pink salmon of the
Bella Coola River, central British Columbia. J. Fish.
Res. Board Can. 25:757-794.

1971. Size selective predation among juvenile salmonid
fishes in a British Columbia inlet. J. Fish. Res. Board
Can. 28:1503-1510.
Peterman. R M.

1984. Density-dependent growth in early ocean life of
sockeye salmon iOncorhynchus nerka). Can. J. Fish.
Aquat. Sci. 41:1825-1829.

PiNKAS, L., M S OLIPHANT, and I. L. K. IVERSON.

1971. Food habits of albacore, bluefin tuna, and bonito in
California waters. Calif Dep. Fish Game, Bull. 12,
105 p.
Taylor, S. G.

1977. The effect of timing of downstream migration on
meirine survival of pink salmon iOncorhynchus gor-
buscha). M.S. Thesis, Univ. Alaska, Southeastern
Senior College, Juneau, 40 p.

Thedinga, J F.

1985. Smolt scale characteristics and yield of coho
salmon, Oncorhynchus kisutch, smolts and adults from
Porcupine Creek, southeastern Alaska. M.S. Thesis,
Univ. Alaska, Juneau, 93 p.

Thorsteinson, F. V

1960. Herring predation on pink salmon fry in a south-
eastern Alaska estuary. Trans. Am. Fish. Sec. 91:321-
323.

Walters, C. J., R. Hilborn, R. M. Peterman, and M J Staley.

1978. Model for examining early ocean limitation of
Pacific salmon production. J. Fish. Res. Board Can.
35:1303-1315.

222

GROWTH THROUGH THE FIRST SEX MONTHS OF ATLANTIC COD,

GADUS MORHUA , AND HADDOCK, MELANOGRAMMUS AEGLEFINUS,

BASED ON DAILY OTOLITH INCREMENTS^

George R. Bolz and R. Gregory Lough^

ABSTRACT

Daily growth increments of otoliths from larval and juvenile Atlantic cod and haddock were enumer-
ated, and growth curves were derived describing the first six months of life. Growth for both species
was best described by Gompertz-type curves. Inverse regressive methods were employed to construct
general models with confidence limits for predicting age (days) for given standard lengths (mm) from
hatching through the juvenile period. Microstructural analysis of the otoliths did not discern a
settling check at the time when the fish would be expected to leave the pelagic lifestyle for the
demersal one, which indicates that the transition is neither physiologically stressful nor abrupt.

Fluctuations in the year-class strength of fish
stocks are thought to be determined by the rate of
mortality during the first year of life (Moser 1981;
Lough et al. 1985; Neilson and Geen 1986; and
others). Calculation of reliable mortality rates,
assessment of the influences of size-selectivity,
and establishment of precise hatching dates and
times during a given year when loss to recruit-
ment is greatest are dependent upon accurate age
and abundance estimates. Recently, investigators
have suggested that mortality during the postlar-
val and juvenile periods may be as critical as that
occurring in the egg and larval life stages (Cohen
and Grosslein 1982; Sissenwine 1984). Investiga-
tion of this hypothesis by the Northeast Fisheries
Center (NEFC) has been ongoing since 1984.
Enumeration of daily growth increments de-
posited on fish otoliths provides the best method
for the age determination of larvae and juveniles
needed for generating growth curves and estimat-
ing mortality (Essig and Cole 1986). An excellent
review of past and current methodologies em-
ployed in the study and application of otolith mi-
crostructure may be found in Campana and Neil-
son (1985).

Atlantic cod and haddock are both spring
spawners on Georges Bank (Sherman et al. 1984)
and have pelagic eggs and larvae that undergo
similar development. Transformation to the juve-

nile life stage occurs around 20-30 mm SL, or 2-3
months from hatching (Fahay 1983). The transi-
tion from the pelagic to demersal habitat of the
adults takes place sometime after transforma-
tion, usually by 6-8 cm in midsummer, and re-
cent field observations by the NEFC indicates the
transition is a gradual process with considerable
variability.

In an earlier study by Bolz and Lough (1983),
growth curves were developed for larval Atlantic
cod and haddock based on otolith analysis that
defined growth from hatching (4-5 mm SL)
through the first two months of life (ca. 20 mm
SL). Juvenile Georges Bank Atlantic cod and had-
dock are not fully vulnerable to bottom-trawl
gear (Clark et al. 1982), and growth curves based
on groundfish surveys conducted by the NEFC in
the autumn and spring are inaccurate for fish
younger than about six months of age. The pri-
mary goal of the work reported here was to derive
age-at-length curves for field-caught Atlantic cod
and haddock describing their growth from hatch
until they are fully available to capture by
bottom-trawl survey gear. A secondary objective
was to determine if a check ring, a wide incremen-
tal band indicative of physiological or environ-
mental changes, was deposited during the juve-
nile's transition from the pelagic to the demersal
mode of life.

IMARMAP Contribution FED/NEFC 87-15, Northeast Fish-
eries Center Woods Hole Laboratory, National Marine Fish-
eries Service, NOAA, Woods Hole, MA 02543.

^Northeast Fisheries Center Woods Hole Laboratory, Na-
tional Marine Fisheries Service, NOAA, Woods Hole, MA
02543.

Manuscript accepted December 1987.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.

METHODS

Atlantic cod and haddock larvae and juveniles
were collected on six cruises conducted by the
NEFC's RV Albatross IV and RV Delaware II on

223

FISHERY BULLETIN: VOL. 86, NO. 2

Georges Bank during the springs of 1981, 1983,
and 1984 and the summers of 1984 and 1985.
Sample dates and station locations where larvae
and juveniles were collected for otolith analysis
are given in Table 1. The samples were collected
with either 1) a continuous double-oblique haul
using a 61 cm bongo net sampler (0.505 and 0.333
mm mesh) deployed to a maximum depth of 100 m
(Posgay and Marak 1980), 2) a 1 m MOCNESS^
fitted with nine 0.333 mm mesh nets which sam-
pled discrete vertical strata from the bottom of
the water column to the surface, 3) a 10 m MOC-
NESS (3 mm mesh) with five nets fished in the
same manner as the 1 m MOCNESS (Wiebe et al.
1976, 1985), or 4) a Yankee 36 otter trawl towed
for 30 minutes (Grosslein 1974). Stations with

high densities of Atlantic cod and haddock larvae
and juveniles in good condition were selected dur-
ing the cruises for otolith analysis. The fish were
removed immediately following the haul and pre-
served in 95% ethanol.

In the laboratory, larvae and juveniles repre-
sentative of the entire size-range collected were
selected for analysis. The standard length, as well
as several other morphometric measurements of
each larva or juvenile, was measured to the
nearest 0.1 mm prior to removal of their otoliths.
The 2 sagittae, 2 lapilli, and, when possible, 2
asterisci were dissected from the fish and, except-
ing juvenile sagittae, mounted whole on micro-
scope slides with Permounf*. The growth incre-
ments (Fig. lA) on most of these otoliths were

^Multiple Opening/Closing Net and Environmental Sensing
System.

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

Table 1 . — Station information for Atlantic cod and haddock specimens collected for otolith analysis by 61
cm bongo net (0.505 mm mesh) oblique hauls (6B5), 1 m MOCNESS (0.333 mm mesh) discrete vertical
hauls (1M3), 10 m MOCNESS (3.0 mm mesh) discrete vertical hauls (10M), and Yankee 36 otter trawl
(Y36) duhng the 1981, 1983, and 1984 survey seasons.

Time GMT

Bottom

Number of

Lat.

Long.
W

(Night or

depth

M&ll

Station

Date

day)

Gear

(m)

Cod

Haddock

1981

Albatross IV

81-03

4ri0'

67=06'

24 April

1235(D)

6B5

—

4ri3'

67=02'

24 April

1330(D)

6B5

—

4ri8'

66=58'

24 April

1450(D)

685

—

41°22'

66=55'

24 April

1630(D)

6B5

—

41=26'

66=51 '

24 April

1840(D)

6B5

—

160

41°22'

67=00'

26 April

0645(N)

1M3

—

1981

Albatross IV

81-05

190

40°57'

67=19'

22 May

0300(N)

1M3

—

197

40°55'

67=13'

25 May

1200(D)

1M3

—

205

40°55'

67=09'

26 May

1130(D)

1M3

—

211

41-11'

67=35'

27 May

1200(D)

1M3

—

215

4ri2'

67=36'

27 May

2330(D)

1M3

—

1983

Albatross IV

83-03

415

40°54'

67=32'

13 May

1816(D)

1M3

418

40°56'

67=35'

14 May

0456(D)

1M3

421

40°5r

67=34'

14 May

1026(D)

1M3

432

40°47'

67=26'

15 May

1636(D)

1M3

—

434

40°46'

67=24'

15 May

2229(D)

1M3

—

438

41°05'

67=47'

16 May

1147(N)

1M3

—

440

4r09'

67=54'

16 May

1646(N)

1M3

442

4r08'

67=48'

16 May

2222(N)

1M3

—

444

4r09'

67=55'

17 May

0504(N)

1M3

—

1984

Albatross IV

84-05

519

4ri9'

67=19'

18 June

0319(N)

10M

—

1984

Delaware II

84-07

40°53'

66=22'

15 Aug.

1744(D)

Y36

4r50

66=26'

16 Aug.

1045(D)

Y36

—

41=49'

66=23'

16 Aug.

1430(D)

Y36

—

41°47°

66=18'

16 Aug.

1604(D)

Y36

—

4r47'

66=24'

16 Aug.

1715(D)

Y36

—

41=47'

66=30'

16 Aug.

1821(D)

Y36

—

41°45'

66=30'

16 Aug.

2024(D)

Y36

—

41=42'

66=25'

16 Aug.

2124(D)

Y36

—

41=47'

66=1 1 '

17 Aug.

0249(N)

Y36

—

1984

Albatross IV

84-09

41=49'

66=16'

12 Sept.

0845(N)

Y36

—

41=52'

66=21'

12 Sept.

0951 (N)

Y36

224

BOLZ AND LOUGH: GROWTH OF ATLANTIC COD AND HADDOCK

Figure 1. — A. Scanning electron micrograph for a portion of the sagitta from a 5-yr-old Atlantic cod, 79 cm SL, showing daily growth
increments. Bar of photograph represents 100 (xm. B. Sagitta from 47-d-old Atlantic cod larva, 13.3 mm SL (630x). Bar of
photograph represents 20 ji.m. nc = nuclear check, yc = yolk-sac check.

225

FISHERY BULLETIN: VOL. 86, NO. 2

discernible without any further preparation.
Sagittae from fish >25 mm SL were mounted in
epoxy resin and were ground, above and below,
with carborundum paper (600 grit). The resulting
thin section was secured to a microscope slide
with epoxy resin and etched with 6% EDTA (pH
7.0). Both the grinding and etching procedures
were monitored periodically by viewing the
sagitta under a dissecting microscope.

The sagittae were then viewed under a Zeiss
compound microscope with transmitted light. The
number of growth increments were counted from
the image projected by a drawing tube onto a
Zeiss MOP Digital Image Analyzer System.
Under transmitted light each growth increment
was composed of a light and dark ring (Fig. IB),
which corresponded to the heavily calcified incre-
mental zone and the organic-rich discontinuous
zone of Watabe et al. (1982). Depending on the
size of the otolith, magnifications used ranged
from 400 X to 1,000 x . Three counts were made on
one of the 2 sagittae from each larva or juvenile,
and those otoliths with a repeatable increment
count of >90% were used in the growi:h analysis.
The other sagitta was counted once for compari-
son, as were the 2 lapilli. The number of incre-
ments on the 2 asterisci also were enumerated. It
was found in the previous study (Bolz and Lough
1983) that the asterisci were not detectable at
hatching, in contrast to the sagittae and lapilli,
but appeared later in the larval period. This was
reflected in the asterisci having on average 27
fewer growth increments than the sagittae. In
those instances where the sagittae and lapilli
were particularly difficult to read, the number of
asteriscal increments plus 27 was consulted as an
additional check. Maximum and minimum di-
ameters and planar surface area of the entire
otolith were measured routinely on all sagittae,
lapilli, and asterisci.

The differential shrinkage of Atlantic cod and
haddock larvae and juveniles with respect to
standard length was corrected using Theilacker's
algorithm ( 1980), which is specified and discussed
in Bolz and Lough (1983). All lengths referred to
in the results and discussion portions of this
paper are reported as corrected lengths.

RESULTS

Haddock Larval and Juvenile Growth

From analysis of the 189 larval and juvenile
haddock, ranging from 3.5 to 123.4 mm SL, we

found that growth was best described by a
Gompertz-type curve. Previous uses of the Gom-
pertz growth curve and methodology for fitting
the curve are described in Pennington (1979),
Lough et al. (1982), and Messieh et al. (1987). The
variance was stabilized by using the natural log
form of the growth equation, and parameters
were derived by nonlinear estimation techniques
resulting in the relationship:

ln(L) = 1.1987 + 4.8438(1 -e

0.0088R\

(1)

where L = standard length in mm, and

R = number of days (increments) from
hatch.

A plot of the Gompertz curve fitted to the natural
log of standard length vs. age in days is shown in
Figure 2.

The predicted hatch-length fi-om the curve of
3.32 mm falls within the range of previous studies
(Colton and Marak 1969; Fahay 1983). An aver-
age growth rate of 0.24 mm/day (Table 2) for the
first 30 days is also reasonable (Laurence 1978;
Laurence et al. 1981) and agrees with the earlier
study of Bolz and Lough (1983). As a generalized
model the Gk)mpertz equation described haddock
growth through the first six months (175 days), at
which point it intersected (Fig. 3) the von Berta-
lanffy growth curve generated fi'om an analysis of
adult haddock by Clark et al. (1982):

Table 2. — Mean standard length at age, 95% confidence limits,
and growth rate (mm/day and %/day) of larval and juvenile had-
dock from hatch through 200 days estimated from the Gompertz
growth model fit.

Mean

95% confidence
limits

Growth

Age

(d)

length
(mm)

rate

Lower

Upper

(mm/day)

(%/day)

3.32

3.22

3.41

0.14

4.22

4.99

4.88

5.11

0.20

4.01

7.27

7.13

7.41

0.26

3.58

10.25

10.07

10.42

0.34

3.32

14.03

13.80

14.26

0.42

2.99

18.70

18.39

19.02

0.51

2.78

24.34

23.89

24.80

0.61

2.55

30.98

30.34

31.63

0.71

2.32

38.63

37.74

39.53

0.82

2.12

47.27

46.08

48.51

0.91

1.95

100

56.88

55.30

58.50

1.01

1.77

110

67.37

65.35

69.45

1.09

1.63

120

78.67

76.14

81.28

1.17

1.49

130

90.66

87.56

93.86

1.23

1.35

140

103.23

99.52

107.08

1.28

1.23

150

116.26

111.88

120.81

1.32

1.14

160

129.63

124.54

134.92

1.35

1.04

170

143.21

137.38

149.28

1.36

0.95

180

156.88

150.29

163.76

1.37

0.87

190

170.54

163.17

178.25

1.36

0.80

200

184.09

175.92

192.63

1.35

0.73

226

BOLZ AND LOUGH: GROWTH OF ATLANTIC COD AND HADDOCK

JZ
-P

X)
L.
O

o
-p
cn

200.0

100.0 -■

50.0 -■

25. 0 --

12.0

-Q 0088R

ln(L) = 1. 1987 + 4.8438(1 - g " )

-I — I — ^-

-I — \ — U

4 — I — U

J I \ L

J I L

100

125

150

175

AgG in Days

Figure 2. — Gompertz growth curve and equation fitted to plot of In standard length and number of otolith increments (estimated age

in days) for 189 larval and juvenile haddock collected on Georges Bank.

L = 738.0(1 - e-0.3763[(fl+D)/365-0.1649])^ (2)

where D = Julian date of hatch.

Based on the 1981 season, an average hatch-date
of 15 April {D = 105) was employed in the present
model. An average length of 19.9 cm would have
been attained on 1 January, by fisheries science
convention the date at which an individual is con-
sidered to be 1-year-old.

The predicted hatch-length of 4.02 mm was
within known limits (Colton and Marak 1969).
The average growth rate of 0.21 mm/day (Table 3)
through the first month was slightly lower than
that of haddock, which is consistent with previous
findings (Bolz and Lough 1983). At approxi-
mately 192 days the larval and juvenile growth
curve intersected the von Bertalanffy curve calcu-
lated for adult Atlantic cod by Penttila and Gif-
ford (1976):

Atlantic Cod Larval and
Juvenile Growth

Although there were few larger individuals
amongst the 157 larval and juvenile Atlantic cod
examined, the apparent pattern was similar to
that seen in haddock. A Gompertz growth curve
also provided a good fit when the natural log of
standard length (range: 4.6-104 mm) was plotted
(Fig. 4) against age in days (range: 7-151):

ln(L) = 1.3915 + 6.2707(1 - e-o.oo53/?)_

(3)

L = 1481 0(1 - e-01200((i?+D)/365-0.6160)^ (4)

For purposes of the model a mean hatch-date of 15
March (D = 74) was assumed. An average At-
lantic cod would be expected to have achieved a
length of 26.1 cm by 1 January (Fig. 5).

Predictability

Since it is desirable, especially during field sur-
veys when direct analysis of otoliths is impossi-
ble, to be able to predict age from a given length,

227

FISHERY BULLETIN: VOL. 86, NO. 2

SL
-t->

D
-P

200.0

100.0

•

50.0

-. 3783 <(R- 1055/365-
ln<L) = 6.6039 * Ind - e

Adult von Bertalanffy Growth Curve
Clark at al. (1982)

1849)

25.0

12.0

•5P^

jf^' '

6.0

•

-0. 0088R
ln(L) = 1. 1987 * 4.8438C1 - e )

3.0
n

•

Larval and Juvenile Gompertz Growth Curve
1 1 1 I

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dqc

Jan

Figure 3. — Haddock growth model with an assumed mean hatch-date of 15 April projected through mid-January. An average
length of 19.9 cm would have been attained on 01 January, by fisheries science convention the date at which an individual is
considered to be 1-year-old.

inverse regression (Draper and Smith 1966) was
performed on the Atlantic cod and haddock
grovii;h curves to estabhsh confidence intervals
for predicting age from a given standard length.
In its reduced form the equation obtained for had-
dock was

X^ _ ln(l - (Xq ± Q.023^(((Xo - 0.2990)2/7.7959) + (1 + l/n))^^^))
Xi -0.0088

(5)

where X^^ and Xi = upper and lower confidence

limits,
Xo=l-e-ooo88« and

n = sample size.

Figure 6 shows the fitted growth curve bracketed
by 95% confidence intervals.

Performing the same calculations on the At-
lantic cod growth curve yielded the relation-
ship:

X„, _ ln(l - (Xq ± 0.022t{{iXo - 0.1918)^/0.9294) + (1 + l/n))^'^))

0.0053

(6)

where Zo = l-e-o 0053ft

228

BOLZ AND LOUGH: GROWTH OF ATLANTIC COD AND HADDOCK

4->

D
C/)

200.0 -

100.0 -

50. 0 -

25.0 •-

12.0

6.0 -

3.0 -

ln(L) = 1.3915 * 6.2707(1 - g-°-°°53Rj

J I \ L

J I L

J 1 L

J I L

100

125

150

175

AgG in Days

Figure 4. — Gompertz growth ciirve and equation fitted to plot of In standard length and number of otolith increments (estimated age

in days) for 157 larval and juvenile Atlantic cod collected on Georges Bank.

Table 3. — Mean standard length at age, 95% confidence limits,
and growth rate (mm/day and %/day) of larval and juvenile Atlantic
cod from hatch through 200 days estimated from the Gompertz
growth model fit.

Mean

95% confidence
limits

Growth

Age

(d)

length
(mm)

rate

Lower

Upper

(mm/day)

(%/day)

4.02

3.79

4.27

0.13

3.37

5.56

5.31

5.82

0.18

3.24

7.55

7.30

7.82

0.23

3.05

10.11

9.85

10.37

0.29

2.87

13.32

13.03

13.62

0.36

2.70

17.31

16.88

17.75

0.44

2.60

22.19

21.48

22.92

0.54

2.43

28.08

26.95

29.26

0.64

2.31

35.11

33.40

36.91

0.76

2.19

43.39

40.91

46.03

0.90

2.10

100

53.05

49.58

56.77

1.04

1.98

110

64.18

59.48

69.26

1.19

1.87

120

76.90

70.69

83.65

1.35

1.77

130

91.27

83.26

100.05

1.52

1.68

140

107.38

97.24

118.57

1.70

1.58

150

125.27

112.65

139.30

1 88

1.51

160

144.99

129.52

162.30

2.06

1.43

170

166.55

147.85

187.61

2.25

1.35

180

189.95

167.61

215.26

2.43

1.28

190

215.17

188.79

245.24

2.61

1.22

200

242.18

211.34

277.52

2.79

1.16

Figure 7 shows the Atlantic cod growth curve
bracketed by 95% confidence intervals. Tables 4
and 5 provide predicted ages of Atlantic cod and
haddock for given standard lengths with 70% and
95% confidence limits.

Otolith Growth

In the earlier study of larval Atlantic cod and
haddock (Bolz and Lough 1983), it was found that
the sagittal rings (one incremental and one dis-
continuous zone) were segregated into distinct re-
gions separated by thicker, darker discontinuous
zones referred to as checks or check rings. Two
"heavy rings" were noted in the larvae: 1) a nu-
clear check surrounding a central, amorphous
core and 1 or 2 irregular rings, and 2) a yolk-sac
check 2-8 increments farther outward (Fig. IB).
The present study corroborated the existence of
these two checks. Although each otolith was care-
fully examined for the presence of a settling
check, no regularly occurring heavy ring could be
discerned beyond the yolk-sac check in either the
haddock or the Atlantic cod juveniles. It should,

229

FISHERY BULLETIN: VOL. 86, NO. 2

"□

200.0

100.0

50.0

25.0

12.0

6.0

3.0

-0.0053R
In(L) = 1.3915 * 6.2707(1 - e )

Larval and Juvonile Gompertz Growth Cunva

ln(L) = 7.6622 * Ind

-. 1200((R*74)/3e5*.eiBro.

Adult von Bortalanffy Growth Curve
Panttila and Glfford (1976)

Mar Apr

Mav

Jun

Jul

Aug

Sep

Oct

Nov

Dec Jan

Figure 5. — Atlantic cod growth model with assumed mean hatch-date of 15 March projected through mid-January. An average
length of 26.1 cm would have been attained by 01 January, by fisheries science convention the date at which an individual is
considered to be 1-year-old.

D
Q

CD
<

200

;

175

^-'^^.^

150

^ - '^

125

•

lUU

—

/ y
/ /

—

25
n

111)

1 1 . 1 < 1

1 1

1 1 1 1 1

I 1 1

1 1 I 1 1 1 1 1

100

125

150

175

230

Standard Length (mm)

BOLZ AND LOUGH: GROWTH OF ATLANTIC COD AND HADDOCK

200

175

D
Q

150

125

^ .^--"''^

/ ^^^-^""^

/ ^^'''''^ ^ -^

/ '^''■^'^ --

100

■^ ^...^ . ^ '"

■

y /-^ • • -^

/ ^X ^ '^

/ .X /

3r>

- itl

J — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — I — 1_. 1 1 1 1 1

100

— I — I — 1 I I I I I I I

125 150 175

Standard Length (mm)

Figure 7. — Inverse regression of Atlantic cod growth curve with 95% confidence intervals for predicting age in days for a given

standard length (mm).

however, be noted that the size range available
for study limited the search for a settling check to
individuals >90 mm SL and does not preclude the
possibility that thinner, less discernible checks
may be found when greater numbers of juveniles
50-90 mm SL are analyzed.

In both haddock and Atlantic cod diametral
growth (fjim) of the sagittae, lapilli, and asterisci
was linearly related to standard length (mm)
throughout the larval and juvenile periods. The
high correlation (r > 0.98) of this relationship
and its good agreement with measurements made
by Bergstad (1984) would allow the sagittal di-
ameter to be used as a check on the predictability
model outlined in Equations (5) and (6) for esti-
mating age from standard length. Estimated ages
for haddock larvae and juveniles based on maxi-

mum otolith diameters may be obtained with the
following equation:

Figure 6. — Inverse regression of haddock growth curve with
95% confidence intervals for predicting age in days for a given
standard length (mm).

Y = 28.390 + 2.413Xi + 21.561X2 + 73.841X3

(7)

where Y = estimated age in days,

Xi = sagittal diameter in mm,
X2 = lapillus diameter in mm, and
X3 = astericus diameter in mm.

Table 6 provides a comparison of estimated ages
derived from otolith diameters with observed
ages derived from the number of daily incre-
ments. Although multiple regression analysis
using the three otolith diameters yielded a high
correlation coefficient (r = 0.9890) and nonsignif-
icant ^-values, the 95% confidence limits are
quite broad (±2 weeks) and should be used with
caution. Use of the sagittal diameter alone
(Y = 35.945 + 18.484Xi) provided a good fit
(r = 0.9861) for juveniles >90 mm SL but was a
poor age predictor for younger fish. If only the
sagitta is available for analysis, the relationship:

231

Table 4. — Predicted age in days with 70% and 95% confidence
limits of larval and juvenile haddock for a given standard length.

FISHERY BULLETIN: VOL. 86, NO 2

Table 5. — Predicted age in days with 70% and 95% confidence
limits of larval and juvenile Atlantic cod for a given standard length.

Observed

standard

length

(mm)

Predicted
age
(d)

70% confidence

limits

95% confidence
limits

Observed

standard

length

(mm)

Predicted
age
(d)

70% confidence
limits

95% confidence
limits

Lower

Upper

Lower

Upper

Lower

Upper

Lower

Upper

10.0

7.1

13.0

4.4

15.9

6.7

2.1

11.3

-2.1

15.7

29.3

25.8

32.8

22.7

36.2

29.6

24.5

34.8

19.9

39.8

42.3

38.4

46.3

34.9

50.1

44.5

39.0

50.1

34.0

55.5

52.5

48.2

56.9

44.4

61.1

55.8

49.9

61.8

44.7

67.6

61.1

56.5

65.8

52.4

70.4

65.0

58.9

71.3

53.4

77.5

68.6

638

73.7

59.4

78.7

72.9

66.5

79.6

60.8

86.0

75.4

70.3

80.8

65.7

86.1

79.9

73.2

86.8

67.3

93.5

81.7

76.2

87.4

71.4

93.0

86.1

79.2

93.3

73.1

100.3

87.5

81.7

93.5

76.7

99.5

91.8

84.7

99.2

78.3

106.5

92.9

86.9

99.3

81.6

105.6

97.0

89.7

104.7

83.2

112.2

98.1

91.8

104.8

86.3

111.4

101.9

94.3

109.8

87.6

117.6

103.1

96.5

110.0

90.7

117.0

106.4

98.7

114.6

91.8

122.6

107.8

101.0

115.1

95.0

122.4

110.7

102.7

119.1

95.7

127.3

112.4

105.3

120.0

99.0

127.6

114.7

106.6

123.3

99.4

131.8

116.8

109.4

124.8

103.0

132.7

118.6

110.3

127.4

102.9

136.1

121.1

113.5

129.4

106.8

137,7

122.3

113.8

131.3

106.3

140.2

125.3

117.4

133.9

110.4

142.6

125.8

117.1

135.0

109.5

144.1

129.5

121.2

138.4

114.0

147.4

129.2

120.3

138.6

112.5

147.9

133.5

124.9

142.8

117.5

152.2

132.4

123.4

142.0

115.5

151.6

100

137.5

128.6

147.1

121.0

156.9

100

135.6

126.4

145.4

118.3

155.1

105

141.4

132.2

151.4

124.3

161.6

105

138.6

129.2

148.6

121.0

158.6

110

145.2

135.8

155.6

127.6

166.2

110

141.5

132.0

151.7

123.7

161.9

115

149.0

139.3

159.8

130.9

170.8

115

144.4

134.7

154.8

126.3

165.1

120

152.8

142.7

163.9

134.1

175.4

120

147.2

137.3

157.7

128.7

168.3

125

156.6

146.1

168.1

137.2

180.0

125

149.9

139.9

160.6

131.1

171.3

130

160.3

149.5

172.2

140.4

184.6

130

152.5

142.3

163.4

133.5

174.3

135

164.0

152.9

176.3

143.4

189.2

135

155.1

144.8

166.1

135.8

177.3

140

167.6

156.2

180.4

146.5

193.9

140

157.6

147.1

168.8

138.0

180.2

145

171.3

159.5

184.6

149.5

198.5

145

160.0

149.4

171.4

140.2

183.0

150

175.0

162.8

188.7

152.5

203.2

150

162.4

151.6

174.0

142.3

185.7

155

178.6

166.0

192.8

155.5

207.9

155

164.8

153.8

176.5

144.4

188.5

160

182.3

169.3

197.0

158.5

212.7

160

167.1

156.0

179.0

146.4

191.1

165

185.9

172.6

201.2

161.4

217.5

165

169.3

158.1

181.4

148.4

193.7

170

189.6

175.8

205.4

164.3

222.4

170

171.5

160.2

183.8

150.4

196.3

175

193.3

179.0

209.6

167.3

227.4

175

173.7

162.2

186.2

152.3

198.8

Y = 11.875 + 112.654Xi, r = 0.9129

(8)

DISCUSSION

should be used for the larval and postlarval pe-
riod.

The equation for the estimated ages of larval
and juvenile Atlantic cod (Table 7) is as follows:

Y = 48.202 + 8.628Xi - 121.908^2 + 139.733Z3,

r = 0.9292. (9)

When using only the sagittal diameter, the fol-
lowing relationship should be applied to larvae
and postlarvae:

Y = 19.364 + 89.560Xi, r = 0.8659. (10)

Unlike the tedious laboratory methods needed for
the enumeration of otolith increments, gross
measurements on a limited number of juvenile
otoliths could be performed at sea.

Despite the tedious methodology required for
enumerating daily growth increments in larval
and juvenile otoliths, the present work suggests
that it is feasible to construct age-length keys for
Atlantic cod and haddock similar to those com-
monly applied in adult population studies (Clark
et al. 1982). The value of such growth data is
based on several assumptions, however. Since all
conclusions drawn depend upon it, reasonable as-
surance of the day-increment relationship in the
species being investigated is critical (Beamish
and McFarlane 1987; Geffen 1987). Confidence in
the growth models generated here for Atlantic
cod and haddock may be found in the following
inferences: 1) the predicted hatch lengths of 3.32
mm for haddock and 4.02 mm for Atlantic cod fall
within known limits; 2) a high correlation for the
length-at-age data with the rearing experiments
of Laurence et al. (1981); and 3) the smooth-

232

BOLZ AND LOUGH: GROWTH OF ATLANTIC COD AND HADDOCK

Table 6. — Estimated age in days based on otolith diame-
ters with 95% confidence limits for larval and juvenile had-
dock compared with observed age derived from number of
daily increments.

Observed
age
(d)

Estimated
age
(d)

95%

confidence
limits

Lower

Upper

0.0

1.5

-12.3

15.4

10.0

11.3

-2.5

25.1

20.0

21.1

7.4

34.9

30.0

30.9

17.2

44.6

40.0

40.7

27.0

54.4

50.0

50.5

36.8

64.2

60.0

60.3

46.7

73.9

70.0

70.1

56.5

83.7

80.0

79.9

66.2

93.6

90.0

89.7

76.0

103.3

100.0

99.5

85.8

113.1

110.0

109.3

95.6

122.9

120.0

119.0

105.3

132.8

130.0

128.8

115.1

142.6

140.0

138.6

124.8

152.5

150.0

148.4

134.5

162.3

160.0

158.2

144.3

172.2

170.0

168.0

154.0

182.0

180.0

177.8

163.7

191.9

Table 7. — Estimated age in days based on otolith diame-
ters with 95% confidence limits for larval and juvenile At-
lantic cod compared with observed age derived from num-
ber of daily increments.

Observed
age
(d)

Estimated
age
(d)

95%

confidence
limits

Lower

Upper

0.0

7.1

-6.9

21.1

10.0

15.8

2.1

29.5

20.0

24.5

11.0

38.0

30.0

33.2

19.9

46.5

40.0

41.9

28.7

55.1

50.0

50.6

37.5

63.8

60.0

59.3

46.2

72.5

70.0

68.0

54.8

81.2

80.0

76.7

63.4

90.1

90.0

85.4

71.9

98.9

100.0

94.1

80.4

107.9

110.0

102.8

88.8

116.9

120.0

111.6

97.2

125.9

130.0

120.3

105.5

135.0

140.0

129.0

113.8

144.1

150.0

137.7

122.1

153.3

160.0

146.4

130.3

162.5

170.0

155.1

138.5

171.7

180.0

163.8

146.6

180.9

ness with which the larval and juvenile curves
flow into those independently developed for the
adults (Clark et al. 1982; Penttila and Gifford
1976).

The predictive models for Atlantic cod and had-
dock have to be viewed as general in nature, and
the widening of the confidence intervals with in-
creasing length (Tables 4, 5) must be kept in

mind. Natural variability of length-at-age and
difficulty in the preparation and reading of
otoliths increases as the fish becomes older and
makes precise age determinations extremely dif-
ficult. For example, the ability to predict correctly
the age of an individual haddock at the 70% con-
fidence level decreases from ±3 days at 5 mm SL
to ±2 weeks at 175 mm. In spite of this problem,
otolith aging of field-caught larvae and juveniles
provides a degree of precision not possible with
indirect methods based on size-frequency analy-
ses (Ebert 1973). Refinement of the estimated
means and the reduction and stabilization of the
variance should result as a greater number of
otoliths are analyzed in the future.

Microstructural examination of larval Atlantic
cod and haddock otoliths clearly delineated check
rings related to hatching and yolk-sac absorption
(Bolz and Lough 1983). Both of these transitions
are abrupt, and the dark, thick discontinuous
zones readily observable on the otoliths are a re-
flection of metabolic disturbances undergone at
these times. Although additional check rings
were noted in 3 or 4 of the juvenile otoliths, there
was no regularity with respect to age of their oc-
currence. In these individuals the checks were
probably the result of physiological trauma in-
duced by disease or injury since calcium carbon-
ate secretion ceases not only with the metabolic
changes accompanying transitional phases but
during times of stress (Morales-Nin 1987). It was
suspected that a distinct check, similar to the set-
tling check found by Victor (1982) in the bluehead
wrasse, Thalassoma bifasciatum , would be found
demarcating the transition from the pelagic to
the demersal mode of life with its accompanying
changes in diet and activity levels. No check rings
were found in the transition period (50-100 days)
on the otoliths analyzed. This suggests that an
abrupt metabolic disturbance does not occur at
this phase of the fish's life and that settling near
the bottom takes place over an extended period of
time (1-2 months) even for individual fish. This
agrees with a preliminary finding for Scotian
Shelf gadoids by Campana and Neilson (1985).
However, in a recent study by Mahon and Neilson
(1987) on the gut contents of Scotian Shelf had-
dock, they concluded that the transition from
pelagic to demersal life occurred relatively sud-
denly, less that a month for the individual fish.
Apparently, change to the demersal life stage is
not stressful for Atlantic cod and haddock, at
least as a metabolic manifestation recorded in
their otoliths.

233

FISHERY BULLETIN: VOL. 86. NO. 2

When used in conjunction with length-
frequency data collected throughout the year, the
Atlantic cod and haddock growth curves pre-
sented in this report should allow accurate esti-
mates of the following: 1) peak hatching dates, 2)
the number of cohorts produced within a given
season, 3) intraseasonal changes in growth and
mortality rates of cohorts, and 4) which part of the
spawning curve the recruits originated from
(Methot 1983). In the future year-to-year com-
parison of deviations in these estimates could
lead to the construction of viable recruitment
models permitting the early prediction of year-
class strength.

ACKNOWLEDGMENTS

We gratefully acknowledge the technical help
and guidance provided by Susan Houghton with
the scanning electron microscope portions of this
paper and by Michael Pennington with the statis-
tical analyses.

LITERATURE CITED

Beamish, R J . and G. A McFarlane.

1987. Current trends in age determination methodology.
In R. C. Summerfelt and G. E. Hall (editors), Age and
growth offish, p. 15-42. Iowa State Univ. Press, Ames.

Bergstad, 0 A.

1984. A relationship between the number of growth incre-
ments on the otoliths and age of larval and juvenile cod,
Gadus morhua L. In E. Dahl, D. S. Danielssen, E. Mok-
ness, and P. Solemdal (editors). The propagation of cod
Gadus morhua L, p. 251—272. Flodevigen rapportser. 1,
1984.

BoLZ, G R., and R. G. Lough.

1983. Growth of larval Atlantic cod, Gadus morhua, and
haddock, Melanogrammus aeglefinus , on Gteorges Bank,
spring 1981. Fish. Bull., U.S. 81:827-836.

Campana. S. E , and J. D. Neilson.

1985. Microstructure of fish otoliths. Can. J. Fish.
Aquat. Sci. 42:1014-1032.

Clark, S. H . W J. Overholtz, and R. C. Hennemuth.

1982. Review and assessment of the Georges Bank and
Gulf of Maine haddock fishery. J. Northwest Atl. Fish.
Sci. 3:1-27.
Cohen. E. B., and M D Grosslein.

1982. Food consumption by silver hake (Merluccius bilin-
earis) on Georges Bank with implications for recruit-
ment. In G. M. Calliet and C. A. Simenstadt (editors),
Gutshop '81, fish food habits studies, p. 286-294. Proc.
Third Pac. Workshop, Washington Sea Grant, Univ.
Wash. Press, Seattle.
COLTON, J B , AND R R. MaRAK.

1969. Guide for identifying the common planktonic fish
eggs and larvae of continental shelf water. Cape Sable to
Block Island. U.S. Bur. Commer. Fish., Biol. Lab.,
Woods Hole, Mass., Lab. Ref. Doc. 69-9, 43 p.
Draper, N R , and H. Smith.

1966. Applied Regression Analysis. Wiley, N.Y.

Ebert, T a

1973. Estimating growth and mortality rates from size
data. Oecologia 11:281-298.

EssiG. R. J , and C F. Cole.

1986. Methods of estimating larval fish mortality from
daily increments in otoliths. Trans. Am. Fish. Soc.
115:34-40.

Fahay, M P

1983. Guide to the early stages of marine fishes occurring
in the western North Atlantic Ocean, Cape Hatteras to
the southern Scotian Shelf. J. Northwest Atl. Fish. Sci.
4:1-423.

Geffen, a J

1987. Methods of validating daily increment deposition in
otoliths of larval fish. In R. C. Summerfelt and G. E.
Hall (editors). Age and growth offish, p. 223-240. Iowa
State Univ. Press, Ames.

Grosslein, M. D.

1974. Bottom trawl survey methods of the Northeast
Fisheries Center, Woods Hole, Massachusetts, U.S.
ICNAF Res. Doc. 74/86, Ser. No. 3332.

Laurence, G. C.

1978. Comparative growth, respiration and delayed feed-
ing abilities of larval cod (Gadus morhua) and haddock
{Melanogrammus aeglefinus) as influenced by tempera-
ture during laboratory studies. Mar. Biol. (Berl.) 50:1-
7.
Laurence, G. C, A S. Smigielski, T. A. Halavik, and B R
Burns.

1981. Implications of direct competition between larval
cod (Gadus morhua) and haddock (Melanogrammus ae-
glefinus ) in laboratory growth and survival studies at
different food densities. In R. Lasker and K. Sherman
(editors). The early life history of fish, p. 304-311.
Rapp. P.-v. Reun. Cons. int. Explor. Mer 178.

Lough, R. G., G R. Bolz, M. Pennington, and M D Grosslein
1985. Larval abundance and mortality of Atlantic herring
(Clupea harengus L.) spawTied in the Georges Bank and
Nantucket Shoals areas, 1971-78 seasons, in relation to
spawning stock size. J. Northwest Atl. Fish. Sci. 6:21-
35.

Lough, R. G., M. Pennington, G R. Bolz, and A A Rosenberg.

1982. Age and growth of larval Atlantic herring, Clupea
harengus L., in the Gulf of Maine-Georges Bank region
based on otolith growth increments. Fish. Bull., U.S.
80:187-199.

Mahon, R., and J D Neilson.

1987. Diet changes in Scotian Shelf haddock during
pelagic and demersal phases of the first year of life.
Mar. Ecol. Prog. Ser. 37:123-130.

Messieh, S. N., D S Moore, and P. Rubec.

1987. Estimation of age and growth of larval Atlantic her-
ring as inferred from examination of daily growth incre-
ments of otoliths. In R. C. Summerfelt and G. E. Hall
(editors), Age and growth of fish, p. 433-442. Iowa
State Univ. Press, Ames.

Methot, R D, Jr.

1983. Seasonal variation in survival of larval northern
anchovy, Engraulis mordax, estimated from the age dis-
tribution of juveniles. Fish. Bull., U.S. 81:741-750.

MoraleS-Nin, B.

1987. Ultrastructure of the organic and inorganic con-
stituents of the otoliths of the sea bass. In R. C. Sum-
merfelt and G. E. Hall (editors). Age and growth offish,
p. 331-343. Iowa State Univ. Press, Ames.

234

BOLZ AND LOUGH: GROWTH OF ATLANTIC COD AND HADDOCK

MOSER, H G.

1981. Morphological and functional aspects of marine fish
larvae. In R. Lasker (editor), Marine fish larvae, p. 90-
131. Univ. Wash. Press, Seattle and Lond.
NEILSON, J. D., AND G. H. Geen.

1986. First-year growth rate of Sixes River chinook
salmon as inferred from otoliths: effects on mortality and
age at maturity. Trans. Am. Fish. Soc. 115:28-33.
Pennington. M R

1979. Fitting a growth curve to field data. In J. KOrd.,
G. P. Patil, and C. Taillie (editors). Statistical distribu-
tions in ecological work, p. 419-428. Int. Coop. Publ.
House, Fairland, MD.

PENTTILA, J. A , AND V. M. GiFFORD.

1976. Growth and mortality rates for cod from the
Georges Bank and Gulf of Maine areas. Int. Comm.
Northwest Atl. Fish, Res. Bull. No. 12, p. 29-36.

POSGAY. J A , AND R R MARAK

1980. The MARMAP bongo zooplankton sampler. J.
Northwest Atl. Fish. Sci. 1:91-99.

Sherman. K , W Smith, W Morse. M Berman, J Green, and
L. Ejsymont.

1984. Spawning strategies of fishes in relation to circula-
tion, phjloplankton production, and pulses in zooplank-
ton off the northeastern United States. Mar. Ecol. Prog.
Ser. 18:1-19.

SiSSENWINE, M p.

1984. Why do fish populations vary? In R. May (editor),
Exploitation of marine communities, p. 59-94.
Springer-Verlag, N.Y.

Theilacker, G H.

1980. Changes in body measurements of larval northern
anchovy, Engraulis mordax, and other fishes due to han-
dling and preservation. Fish. Bull., U.S. 78:685-692.
Victor, B. C.

1982. Daily otolith increments and recruitment in two
coral-reef wrasses, Thalassoma bifasciatum and Halicho-
eres bivittatus. Mar. Biol. (Berl.) 71:203-208.
Watabe, N . K. Tanaka. J Yamada, and J. M Dean.

1982. Scanning electron microscope observations of the
organic matrix in the otolith of the teleost fish Fundulus
heteroclitus (Linnaeus) and Tilapia nilotica (Linnaeus).
J. Exp. Mar. Biol. Ecol. 58:127-134.
Wiebe, P H , K. H Burt, S H Boyd, and A W Morton.

1976. A multiple opening/closing net and environmental
sensing system for sampling zooplankton. J. Mar. Res.
34:313-326.
Wiebe, P. H., A. W. Morton. A M Bradley, R H Backus.
J. E. Craddock, V Barber, T. J. Cowles, and G R Flierl.

1985. New developments in the MOCNESS, an apparatus
for sampling zooplankton and micronekton. Mar. Biol.
(Berl.) 87:313-323.

235

THE RELATION BETWEEN SPAWNING SEASON AND

THE RECRUITMENT OF YOUNG-OF-THE-YEAR BLUEFISH,

POMATOMUS SALTATRIX, TO NEW YORK^

Robert M. Nyman^ and David O Conover^

ABSTRACT

The association between oceanic spawning season and the recruitment of young-of-the-year (YOY)
bluefish, Pomatomus saltatrix, to the inshore waters of New York was studied by estimating the
spawn dates of recruited fish collected in the shore zone from the number of growth increments in
their otoliths. Field collections on the south shore of Long Island showed that recruitment of 3-6 cm
fork length fish occurred as a distinct pulse during the last week of May in 1985 and the second week
of June in 1986. Length-frequency distributions were generally unimodal and most fish collected later
could be attributed to this one recruitment episode. The frequency of otolith ring deposition in YOY
bluefish was determined by marking the otoliths of field-caged fish with an injection of tetracycline,
and then periodically subsampling these over the ensuing 61-day period. Regression analysis indi-
cated a 1:1 relation between the number of days since marking and the number of rings beyond the
mark. Back-calculation to the time of first ring deposition revealed that field-collected YOY bluefish
from Long Island were spawned primarily in the March-April spawning season reported to occur
south of Cape Hatteras. Relatively few fish were collected from the summer spawning season that
reportedly occurs in the Middle Atlantic Bight. Almost all of these summer-spawned fish were
collected from the Hudson River.

The bluefish, Pomatomus saltatrix, supports a
major recreational fishery along the Atlantic
coast of the United States. In 1985, more bluefish
by weight were caught than any other marine
fish, accounting for over 24% of the total marine
recreational catch (U.S. Department of Com-
merce 1986). Despite the importance of bluefish
to the recreational fishery, very little is known of
its early life history.

Bluefish are found over different portions of the
continental shelf from Florida to Nova Scotia at
various times of the year (Bigelow and Schroeder
1953; Wilk 1977; Gilmore 1985). Based on de-
scriptions of the temporal and spatial abundance
of larvae, Kendall and Walford (1979) suggested
that there are primarily two distinct spawning
periods and regions: a spring spawning in the
South Atlantic Bight at the edge of the Florida
Current (see also Collins and Stender 1987), and
a summer spawning in the Middle Atlantic Bight

iContribution No. 588 of the Marine Sciences Research Cen-
ter, State University of New York, Stony Brook, NY.

^Marine Sciences Research Center, State University of New
York, Stony Brook, NY 11794-5000; present address:
Chesapeake Biological Laboratory, University of Maryland,
Solomons, MD 20688.

^Marine Sciences Research Center, State University of New
York, Stony Brook, NY 11794-5000.

Manuscript accepted November 1987.
FISHERY BULLETIN; VOL. 86, NO. 2, 1988.

midway over the continental shelf (see also Morse
et al. 1987). They further proposed that the
spring-spawned larvae are transported north-
ward in the slope waters and then move inshore,
spending their first summer in the bays and estu-
aries of the Middle Atlantic Bight. Summer-
spawned larvae, according to Kendall and Wal-
ford, spend their first summer at sea or enter the
estuaries of the Middle Atlantic Bight only
briefly before migrating southward with the
onset of winter. A minor spawning season extend-
ing from September to November off the coast of
Georgia and Florida (Collins and Stender 1987)
involves fish resident to the South Atlantic Bight
(Kendall and Walford 1979).

The purpose of this study is to evaluate Kendall
and Walford's hypothesis by back-calculating the
spawn dates of young-of-the-year (YOY) bluefish
that have recruited to inshore waters, from the
number of grov^i;h increments in their otoliths.
First, we describe the timing and pattern of re-
cruitment of YOY bluefish to one segment of the
mid-Atlantic coastline: Long Island, NY. If
spawning is episodic, and if YOY bluefish from
each spavining period enter New York waters,
then length-frequency distributions of field col-
lections should be multimodal. Next, we verify
that otolith increment deposition has a daily peri-

237

FISHERY BULLETIN: VOL. 86, NO. 2

odicity in P. saltatrix. Finally, the spawning sea-
son(s) of YOY bluefish recruiting to New York is
determined by ageing and back-calculating to the
date of first ring deposition.

METHODS

Seine Collections

The temporal abundance and length-frequency
distribution of YOY bluefish was estimated by
seining 2-4 times per month from April to Octo-
ber at several sites on Long Island and in the
Hudson River (Fig. 1). In 1985 and 1986, three
sites in Great South Bay on the south shore of
Long Island were sampled: Smith Point County
Park, Fireplace Neck, and the Carmans River.
Seining was conducted with a 0.6 cm mesh, 30 m
net set from shore, either on foot or from a small
boat. Water temperature was recorded at each
site and date. In 1986, a site on the north shore of
Long Island, Setauket Harbor, was also sampled.
A few samples were taken in the fall by angling
with rod and reel. All specimens were frozen for
later measurement of fork length (FL) and
weight, and extraction of otoliths.

Additional specimens captured in 1986 from
Jamaica Bay and the Hudson River were pro-
vided by the New York Department of Environ-
mental Conservation (NYDEC). Their sampling

was conducted with a 60 m seine (1.2 cm mesh) set
from a boat.

Otolith Preparation and Analysis

The sagittae were mounted concave side down
on a glass microscope slide with cyanoacrylate
(instant glue). Two layers of masking tape were
applied on either side of the otolith. The slide was
then turned upside-down and sanded on a strip of
wet 1200 grit wet-dry sandpaper. The masking
tape ensured that the otolith was sectioned on a
consistent plane and helped prevent grinding
past the nucleus. Once the nucleus was reached,
the otolith was polished on wet felt, using levi-
gated alumina polishing compound. Three repli-
cate counts of each otolith were made under a
Zeiss'* compound microscope with transmitted po-
larized light at 125-312X. If the three counts dif-
fered by more than 10% (which occurred in about
1 out of every 10 otoliths), an additional count
was made and the outlier discarded. The three
final counts were then averaged. The total length
of each otolith was measured (nearest 0.1 mm)
with a dissecting microscope using an ocular
micrometer.

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

7 4 "loo'

73° |00

SOUTH BAY

JAMAICA BAY

ATLANTIC OCEAN

73°]00'

72 °|00'

Figure 1. — Map of the study area with sampling locations cis indicated.

238

NYMAN and CONOVER: YOUNG-OF-THE-YEAR BLUEFISH

Frequency of Ring Deposition

The frequency of growth ring deposition was
determined by marking the otoliths of fish with
tetracycline and then subsampling the marked
fish at various periods of time thereafter (Cam-
pana and Neilson 1982). Sixty YOY bluefish (7-
10 cm FL) were captured by seine in Flax Pond,
Old Field, NY (Fig. 1) and were transported to the
Flax Pond Laboratory of SUNY Stony Brook. The
fish were anesthetized in a solution of MS-222 (30
mg/L) and given an intraperitoneal injection of
tetracycline (100 mg/kg offish). After injection,
all fish were placed in a 1.3 x 1.3 m cylindrical
floating cage constructed out of 5 mm plastic
mesh and anchored in Flax Pond. The fish were
fed chopped Menidia menidia twice a day, and
dead bluefish were removed daily. Samples of 5-
10 healthy fish were periodically taken from the
cage using a dip net and frozen until the otoliths
could be excised. The experiment was terminated
61 days after the injections.

After preparation as described above, the
tetracycline-treated otoliths were viewed on a
Zeiss compound microscope using reflected ultra-
violet (UV) light at 160-400 X. Tetracycline fluo-
resces upon exposure to UV light, thus enabling
the location of the marked ring to be determined.
The UV light was then turned off, and the num-
ber of rings from the mark to the edge of the
otolith was counted under transmitted white
light. Each otolith preparation was coded so that
the reader did not know the true age. Three repli-
cate counts were conducted on each otolith.

RESULTS

Temporal Abundance and Length
Frequency

Great South Bay

The appearance of YOY bluefish in the shore
zone was abrupt in both years of the study. In
1985, no YOY bluefish were caught in weekly
samples until 28 May when a catch per unit effort
(CPUE = no. fish per seine haul) of 14.0 was
recorded (Fig. 2a). Corresponding water tempera-
ture was about 20°C. CPUE declined steadily
thereafter through October with two exceptions:
the large collections on 10 and 28 July were each
due to an unusually large number offish in single
seine hauls in the Carmans River. In 1986, YOY
bluefish were first caught on 10 June when the

water temperature was 24°C. The maximum
CPUE (45.3) was obtained on 16 June and was
followed by a decrease in CPUE in subsequent
collections (Fig. 2b).

Length-frequency distributions in 1985 showed
the progression of a single mode through mid-
August (Fig. 3a). Newly recruited fish in late May
were 3-6 cm FL. Subsequent samples showed an
increase in the mean and range of fish lengths,
probably due to somatic growi;h of the initial re-
cruits. There was no evidence of new 3-6 cm re-
cruits entering the shore zone later in the year
(Fig. 3a). Although seining continued until
November, very few YOY bluefish were caught
after August. An additional sample {n = 8) taken
on 16 September by angling from a pier on Great
South Bay had a mean fork length of 17.8 cm and
a range of 14.6-19.5 cm. Length-frequency data
from 1986 (Fig. 3b) show a very similar pattern to
that in 1985: a single length mode appears in
June and these fish increase in size through the
summer. Few YOY bluefish were caught in Au-
gust, September, or October.

Size at recruitment to the shore zone was simi-
lar in both years of our study: mean length of the
1985 and 1986 year classes at first appearance in
the shore zone was 4.6 and 4.5 cm respectively
(Fig. 4). However, because the 1986 year class
first appeared in the shore zone two weeks later
than did the 1985 year class (Fig. 2), the mean
lengths of 1986 year class were less than those of
1985 on comparable dates in June and early July.
By mid-July, however, this difference in mean
length of the two year classes was no longer ap-
parent. Both year classes reached a size of about
13-14 cm by late August when they rarely ap-
peared in our seine collections.

Setauket Harbor

In 1986, the YOY bluefish did not appear on the
north shore of Long Island at Setauket Harbor
until 3-6 weeks after they first appeared in Great
South Bay. Collections at Setauket Harbor were
small at first with only one individual being
caught on 1 July and three on 8 July. It was not
until 22 July that catches similar in number to
those in Great South Bay were being obtained.
These fish had similar mean lengths (10.2 cm,
n = 87, on 22 July; 11.9 cm, n = 22, on 5 August;
13.9 cm, n = 17, on 20 August) to those on com-
parable dates from Great South Bay (of. Fig. 4).
Length-frequency distributions by date were uni-
modal.

239

^ 20
UJ

'^ 16

^ 12

UJ
CL

FISHERY BULLETIN: VOL. 86, NO. 2

(12)

(2)

-(2)(2)(4)(3)
0 0 0 0

(3)

(13)

(7)

(9)

(5)

(7)

(10)
(6) 0

MAY

1 \ 1

JUNE JULY AUG SEPT OCT

FlGlTRE 2.— Catch per unit effort (CPUE) of YOY bluefish from
Great South Bay, NY, plotted with the mean water temperature

Jamaica Bay and the Hudson River

The length-frequency distributions of YOY
bluefish from Jamaica Bay in June, July, and
August were similar to those from Great South
Bay (Fig. 5). Fish lengths in June were unimodal.
Subsequent collections contained progressively
larger fish that were also unimodal in length dis-
tribution.

Sampling in the Hudson River began on 16
July 1986 and continued through 8 October. The
size ranges of YOY bluefish in the July and Au-
gust samples were similar to those from Great
South Bay, although the length distribution on 30
July appears bimodal (Fig. 6). The length distri-
butions from the 10 and 23 September collections
were especially broad. In particular, the 23 Sep-
tember sample contained a group offish that were
much smaller (10-14 cm FL) than the mean size
at this time in Great South Bay (Fig. 4), together
with a second group of larger fish that correspond
more closely in size with those collected else-
where (18-24 cm).

Frequency of Ring Deposition

In tetracycline-injected YOY bluefish, the
number of rings beyond the tetracycline mark (Y)
and the number of days after injection (X) had a
1:1 correspondence (Fig. 7). The relationship was
described by the equation Y = 0.97 LY - 0.287
in = 27, r = 0.996). The slope did not differ signif-
icantly from 1.0 (^test, P > 0.1).

Growth rate of the caged fish was slightly
greater than that of field fish and survival was
high (80%) with mortalities occurring only in the
first few days of the 61-d experiment. The in-
crease in mean fork length was 1.7 mm/day
among the caged fish, as compared with about 1.3
mm/day for fish from field collections during the
same time period (Fig. 4). Hence, the caged fish
did not appeEir to be adversely affected by confine-
ment.

240

NYMAN and CONOVER: YOUNG-OF-THE-YEAR BLUEFISH

b. '°'-

^ 26
o

a. 22

liJ

< 14
10,

J-

^ 20|-

en 16

iZ 12

(3)
45.3

(9)

"(12) (8) (10(3)
0 OOP

— I r

(3)

(10)

(13)

(7)

(e)(4)

MAY

JUNE

1 r

JULY AUG SEPT OCT

Figure 2. — Continued — on each sampling date, (a) 1985;
(b) 1986. Number of seine hauls is in ().

Back-Calculated Date of
First Ring Deposition

A representative sample of 169 YOY bluefish
(n - 88 from 1985, n = 81 from 1986) captured in
Great South Bay were aged by counting the total
number of otolith rings. The date of first ring
deposition for each aged fish was then calculated
based on the date of capture. In both 1985 and
1986, the dates of first ring deposition for YOY
bluefish were predominantly in March and April
(Fig. 8a, b).

Four fish from each of the two apparent length
modes in the 30 July collection from the Hudson
River (Fig. 6) were aged to determine if these
represented a difference in spawning season. The
fish examined were 7.8—13.8 cm in size, and back-
calculated dates of first ring deposition extended
from 7 to 30 April. Hence, these fish could all be

attributed to the same spring spawning period as
those from the south shore of Long Island.

However, YOY bluefish from the smaller (10-
14 cm) size class caught on 23 September in the
Hudson River (Fig. 6) were also aged and their
back-calculated dates of first ring deposition were
found to be predominantly in June and July, and
to a lesser extent in May (Fig. 8c). These dates
differed greatly from those offish captured earlier
in the year in the Hudson River, and along the
south shore of Long Island.

The relationship of ring number and fork
length for each year was best described by the
following equations: Y = 132.308X - 29.890 in
1985 and Y = 95.532X + 1.186 in 1986, where X
is log fork length and Y is the number of rings
(Fig. 9). The slopes of these regressions differed
significantly (ANCOVA, P < 0.001). Total otolith
length and fork length were highly correlated
(r > 0.99) and increased isometrically. Total
otolith length and ring number also had a high
correlation (r = 0.91).

241

FISHERY BULLETIN: VOL. 86, NO. 2

'Z.
UJ
Z)

15;
10;

28 MAY

5 \

II 1

32 ■

22;

24 JUNE

12 ;

10 JULY

Jl.

15
10;

2 8 JULY

5 '■

. •■iklMI

15 1
10;

2 AUG

5 \

■ Li

3 6 9 12 15 18 21 24

FORK LENGTH (cm)

>-
o

UJ
(T

l_L_

10 JUNE 86

22 JULY 86

20 AUG 86

I illJUili , I

3 6 9 12 15 18 21 24

FORK LENGTH (cm)

Figure 3.— Length-frequency histograms for YOY bluefish (no.) from Great South Bay, NY. (a) 1985; (b) 1986.

242

NYMAN and CONOVER:

YOUNG-OF-THE-YEAR BLUEFISH

1 /

—

-7^-—

—

llJ

—

V i/

/♦^ — ¥

• = 1985

4
2

• = 1986
1 1 1

JUNE

JULY

AUG SEPT

DATE

COLLECTION

Figure 4. — Mean fork length (cm) of YOY bluefish (no.) captured with a beach seine in
Great South Bay in 1985 and 1986. Vertical bars are 95% confidence intervals. The last
sample in 1986 was caught by angling.

01 AUG

3 6 9 12 15 18 21 24 3 6 9 12 15 18 21 24

FORK LENGTH (cm)

Figure 5. — Length-frequency distributions for YOY bluefish (no.) fi^m Jamaica Bay, NY, 1986.

243

FISHERY BULLETIN: VOL. 86, NO. 2

30 JULY

*T*- -T

13 AUG

• f "-y

0 SEPT

.uIL

ii iJ I

23 SEPT

3 6 9 12 15 18 21 24

FORK LENGTH (cm)

The analysis was then expanded to the remain-
ing samples of YOY bluefish from Great South
Bay that had not been aged. The above length-age
equations were used to estimate date of first ring
deposition from the dates of capture for all YOY
captured in each year of sampling from Great
South Bay. This exercise revealed that the vast
majority of YOY bluefish in our collections from
Long Island had dates of first ring deposition in
late March, April, and early May (Fig. 10). The
weighted mean date of first ring deposition was 8
April 1985 and 14 April 1986.

The age-length equations for YOY bluefish
from Great South Bay were not applied to collec-
tions from Jamaica Bay or the Hudson River. Pre-
liminary analyses suggested that the age-length
equation for fish from the Hudson River differs
substantially from those in Great South Bay,
probably owing to a difference in growth rate.
Geographic variation in the pattern of recruit-
ment and in the age-length relationships of YOY
bluefish are being further investigated.

DISCUSSION

Recruitment of YOY Bluefish
to New York

In both 1985 and 1986, the arrival of YOY blue-
fish on the south shore of Long Island was abrupt.
Within about a 1-wk period, CPUE went from 0.0
to 14-18 fish/seine haul. CPUE then remained
high for the next two months until declining in
August and September when the fish probably
became too large to be efficiently sampled by our
techniques. These data suggest that the YOY
bluefish recruit to the shore zone as a sudden
pulse. The timing of this recruitment event is ap-
parently variable, differing by about two weeks
among the two years of our study. The appear-
ance of fish 3-6 weeks earlier on the south shore
(Great South Bay) than on the north shore (Se-
tauket Harbor) of Long Island suggested that
these fish arrive from offshore waters to the
south.

Temperature probably influenced the time of
arrival of YOY bluefish in the shore zone. In both
years of our study, YOY bluefish appeared as
temperatures reached about 20°-24°C. In Octo-
ber, after temperatures dropped to the middle

Figure 6. — Length-frequency distributions for YOY bluefish
(no.) from the Hudson River, NY, 1986.

244

NYMAN and CONOVER: YOUNG-OF-THE-YEAR BLUEFISH

Figure 7. — Relation between number of days
since marking with tetracycline and the number
of growth increments beyond the mark in YOY
bluefish maintained in a field cage.

<
2

>-
UJ
CD

I-
z

40 -

S 20

Y= 0.97IX- 0.287

r = 0.996

10 20 30 40 50 60
DAYS SINCE INITIAL MARKING

O
UJ

on
u.

Figure 8. — Back-calculated date of first ring deposi-
tion for YOY bluefish (no.) as determined by count-
ing daily growth rings, (a) and (b) represent fish
from Great South Bay in 1985 and 1986, respec-
tively, (c) is for fish seined from the Hudson River
on 23 September 1986.

MAR APR MAY JUN JUL MAR APR MAY JUN JUL

DATE OF FIRST RING DEPOSITION

245

FISHERY BULLETIN: VOL. 86, NO. 2

180 -

160 -

■

140 -

■

Ul
CD
Z
1 — 1

o
z

120 -
100 -
80 -
60 -
40 -
20 -

1986

■ Xx , X %^^f0^

1985

**^ ^ X

■ 1985
X 1986

0 -

1 1 1 1 r 1

0.4 0.6 0.8 10 12

LOG^O FORK LENGTH (cm)

Figure 9. — Relation between log fork length (X) and number of otolith rings (7) for YOY
bluefish from Great South Bay, NY. The regression equations are 7 = 132.308X - 29.890
(n = 88) for 1985 and y = 95.532A: + 1.186 (n = 81) in 1986.

teens, we no longer captured YOY bluefish. Oben
(1957) noted that in the Black Sea, YOY bluefish
appeared in the shore zone at temperatures of
18°-24.5°C, and left the shore zone in October and
November when temperatures dropped to 13°-
15°C.

Length-frequency distributions of YOY blue-
fish from the south shore samples showed only a
single mode that attained progressively larger
size through the summer and fall, and corre-
sponded to the initial recruitment offish. If multi-
ple spawning and recruitment events contributed
YOY bluefish to Long Island, multimodal length-
frequency distributions should have been ob-
served. The unimodal distributions suggested
that only one spawning period contributed the
majority of YOY bluefish to Long Island.

Interannual variation in the length-age rela-
tionship of YOY bluefish was observed. Although
recruitment occurred two weeks earlier in 1985
than in 1986, the empirical mean lengths at re-
cruitment were similar (Fig. 4). Postrecruitment
growth, however, was slower in 1985 than 1986 so
that empirical mean lengths became similar by
mid-July. Correspondingly, the slope of the
length-age regressions differed significantly
among years: YOY bluefish at an age of about
50-70 days had greater fork lengths in 1985 than

in 1986 (Fig. 9), but the reverse was true among
older, larger fish. Apparently, the growth rate of
YOY bluefish prior to recruitment was higher in
1985 than 1986, but this pattern among the two
years was reversed during the period of postre-
cruitment growth.

Validation of Daily Otolith Rings

Our experimental results demonstrate that
otolith ring deposition is daily in YOY bluefish. A
regression slope of 0.971 indicates a one-to-one
correspondence between number of days after in-
jection and the number of rings beyond the tetra-
cycline mark. This outcome is not particularly
surprising because numerous studies have shown
that increment production is daily, particularly
in the early life history when somatic growth is
rapid (Brothers et al. 1976; Campana and Neilson
1985; Jones 1985). Cases where ring periodicity is
reportedly less-than-daily have involved subopti-
mal growing conditions (Geffen 1982, 1987; Rice
et al. 1985). In our study, the confinement of YOY
bluefish in a field cage apparently had little effect
on growth rate, or the production of daily growth
increments. The field cage allowed for natural
light, temperature, and salinity variations that
the fish would normally have experienced in na-

246

NYMAN and CONOVER: YOUNG-OF-THE-YEAR BLUEFISH

CD
Z)
O"
QJ

50 -

40 -

30 -

20 -

10 -

150

180

Julian Date of First Ring Deposition

Figure 10. — Estimated date of first ring deposition for all YOY bluefish caught in Great
South Bay in 1985 (n = 561) and in 1986 {n = 868) using the respective age-length equa-
tions in Fig. 9.

ture. Feeding periodicity was probably the pri-
mary artifact of confinement that could have af-
fected the rate of ring production in caged fish.
However, Marshall and Parker (1982) showed
that feeding periodicity did not significantly af-
fect ring production in sockeye salmon,
Oncorhynchus nerka.

We were unable to determine directly the num-
ber of days between spawning and first ring depo-
sition because numerous attempts to capture
running-ripe females for initiating experiments

on eggs and larvae were unsuccessful. However,
most species of fish deposit the first daily growth
increment within a few days of hatching (Broth-
ers et al. 1976; Radtke and Dean 1982; McGurk
1984; Radtke 1984; Davis et al. 1985). Recent ev-
idence suggests this is also true in bluefish. Lar-
vae captured off Long Island in 1987 had about
seven otolith increments at a total body length of
4-5 mm (R. K. Cowen and D. O. Conover, unpubl.
data). Based on the rate of development at 20°C in
the laboratory observed by Deuel et al. (1966),

247

FISHERY BULLETIN: VOL. 86, NO. 2

larvae would reach this size in about 5-7 days
posthatch. If so, first ring deposition would
roughly coincide with hatching. Bluefish hatch in
48 hours at 20°C (Deuel et al. 1966), so the day of
first ring deposition probably follows the date of
spawning by about 2-4 days.

Spawning Seasons
Along the Atlantic Coast

Published studies of larval bluefish distribu-
tions along the Atlantic coast suggest the exis-
tence of three temporally and spatially distinct
spawning seasons: spring and fall spawning sea-
sons in the South Atlantic Bight and a midsum-
mer spawning in the Middle Atlantic Bight. In
the only synoptic study covering most of the U.S.
east coast, Kendall and Walford (1979) described
two periods of high larval abundance: One peak
occurred in March and April on the outer shelf of
the South Atlantic Bight, and the other peak was
in July and August midway over the continental
shelf of the Middle Atlantic Bight. Subsequently,
Powles (1981) and Collins and Stender (1987) also
found the highest abundance of bluefish larvae in
the South Atlantic Bight (Cape Canaveral to
Cape Fear) to be in April and May. Collins and
Stender, however, noted the existence of a lesser
peak in larval abundance during September-
November. This fall spawning season in the
South Atlantic Bight was further confirmed by
Finucane and Collins (in press) based on the
gonad condition of bluefish from Georgia and the
Carolinas. In the Middle Atlantic Bight off Vir-
ginia, Norcross et al. (1974) found that eggs and
larvae of bluefish first appeared in June, peaked
in abundance in July, and persisted into August.
Similar observations on the timing of the summer
spawning season in the Middle Atlantic Bight
were presented by Sherman et al. (1984) and
Morse et al. (1987).

Lassiter (1962) provided additional evidence
of the existence of relatively discrete spawning
seasons in bluefish. He showed that the dis-
tribution of back-calculated lengths at age one
has a distinctly bimodal pattern among adult
fish from North Carolina. Size at age 1 tended
to be either about 14 cm or 28 cm. Lassiter
showed that the bimodal pattern could not be
explained as a difference in growth rate, and
suggested that there must be two distinct
spawning seasons such that one group of fish had
a first growing season about twice as long as the
other.

Spawn Dates of YOY Bluefish
from New York

Back-calculation to the day of first ring deposi-
tion for YOY bluefish recruiting to Great South
Bay in 1985 and 1986 demonstrated that these
fish were spawned primarily in March and April
(Figs. 8, 10). Fish that were spawned in July-
August were rarely captured by us on Long Island
in 1985 or 1986, despite continued sampling into
October.

Recruitment to Jamaica Bay and the Hudson
River in July and August 1986 involved YOY
bluefish of about the same size as those fi-om
Great South Bay. Though the size range of fish
fi-om the Hudson was slightly greater than those
from Long Island, fish aged from each of the two
modes appearing in the July Hudson River sam-
ples (Fig. 6) were all spawned during April within
about three weeks of each other. The apparent
bimodality in July is probably a sampling arti-
fact. Hence, Jamaica Bay and Hudson River fish
collected in July and August can be attributed to
the same spawning season as those fi^om Great
South Bay.

Length-frequency distributions ft"om the Hud-
son River in September, however, contained a
group of unusually small bluefish, and back-
calculation showed that they were spawned pre-
dominately in June and July (Fig. 8c). These fish
probably resulted fi-om the summer spawning
season in the Middle Atlantic Bight. Examina-
tion of gonads from adult fish captured during
1986 suggested that the running-ripe males and
mature females were most abundant during late
June and July off Long Island (L. Chiarella and
D. O. Conover, unpubl. data). Hence, at least,
some summer-spawned YOY bluefish do recruit
to the shore zone of the Middle Atlantic Bight.
They were, however, much less abundant than
spring-spawned YOY bluefish in our 1985 or 1986
samples.

Spawning by bluefish in the spring is known to
occur only in the South Atlantic Bight (Kendall
and Walford 1979; Collins and Stender 1987).
Water temperatures over the shelf north of Cape
Hatteras are probably too low for bluefish to
spawn in March and April: average shelf water
temperatures in the Middle Atlantic Bight range
from 5° to 14°C in March and April (Ingham
1986). Virtually no eggs and larvae (Morse et al.
1987) and comparatively few adult bluefish
(Gilmore 1985) are captured in plankton or trawl
surveys north of Cape Hatteras in March and

248

NYMAN and CONOVER: YOUNG-OF-THE-YEAR BLUEFISH

April. Moreover, the time of arrival of YOY blue-
fish on Long Island actually precedes the summer
spawning season in the Middle Atlantic Bight.
We therefore conclude that YOY bluefish recruit-
ing to the Middle Atlantic Bight in late spring
come from spawnings in the South Atlantic
Bight.

Larval Transport

The physical mechanisms that account for the
transport of bluefish larvae fi"om the South At-
lantic Bight to New York are not clear. Spawning
in the South Atlantic Bight occurs primarily over
the outer half of the continental shelf (Powles
1981; Collins and Stender 1987), and some larvae
may be entrained by the Gulf Stream and carried
northward into the slope waters of the Middle
Atlantic Bight (Kendall and Walford 1979).
Neuston net collections in April have shown that
bluefish larvae are periodically abundant on both
sides of the Gulf Stream-shelf water interface off
Cape Hatteras (Kendall and Walford 1979).
Collins and Stender (1987) found a negative cor-
relation between larval size and latitude in the
South Atlantic, but their sampling may not have
extended far enough north (i.e., they did not sam-
ple above Cape Fear).

If the Gulf Stream is responsible for the north-
ward transport, a mechanism by which larvae
avoid being advected too far offshore would ap-
pear to be necessary. According to our results, the
interval between spawning and recruitment to
Long Island is about 45-60 days, whereas the
surface flow of the Gulf Stream at lat. 36°N is
about 104 km/day (Iselin 1936). Hence, larvae re-
maining in the Gulf Stream for an extended pe-
riod would be transported far off the shelf. Reten-
tion near the shelf could be achieved by entering
the slope waters at an appropriate time.

The abrupt appearance of YOY bluefish in the
shore zone suggests that the onshore migration is
a temporally distinct event, perhaps triggered by
vernal warming of the shelf. Because the circula-
tion of the slope and shelf waters of the Middle
Atlantic Bight is toward the southwest (Sherman
et al. 1984), the cross-shelf migration must to
some extent involve active swimming.

Very few summer-spawned YOY bluefish were
captured in our study. This may not be surpris-
ing, however, because the prevailing currents
over the midshelf off Long Island would carry lar-
vae to the southwest. If so, summer-spawned fish
would be found along the coast fi"om approxi-

mately New Jersey to Cape Hatteras. We caution,
however, against any general conclusion concern-
ing the lack of summer-spawned fish in New
York. There could, for example, be substantial
year-to-year variation in the recruitment level of
spring- and summer-spawned fish along any par-
ticular segment of the U.S. coast. These issues are
now being examined by extending our sampling
to southern latitudes.

ACKNOWLEDGMENTS

We thank Robert Cerrato, Robert K. Cowen,
and Peter Woodhead for reviewing the manu-
script and Stephen Heins, Melanie Meade and
Louis Chiarella for assistance in the field. Byron
Young and Kim McKown of the NYDEC gra-
ciously provided samples from the Hudson River
and Jamaica Bay. An earlier version of this paper
was submitted by R.M.N, to the Graduate School
of the State University of New York at Stony
Brook in partial fulfillment of the requirements
for an M.S. degree in Marine Environmental Sci-
ences. Initial funding was provided by grants
from the Sport Fishery Research Foundation
(D.O.C./R.M.N.) and the Montauk Marine Basin.
Later funding was provided by grants to D.O.C.
fi-om the NYDEC through the Dingell-Johnson
Federal Aid in Sport Fish Restoration Act and
by the New York Sea Grant Institute through
the NOAA Office of Sea Grant, U.S. Department
of Commerce, under Grant No. NA86AA-D-
SG045.

LITERATURE CITED

BiGELOW, H. B., AhfD W. C. SCHROEDER.

1953. Fishes of the Gulf of Maine. U.S. Fish and Wildlife
Service, Fish. Bull. 53:1-577.
Brothers. E. B, C P. Mathews, and R. Lasker.

1976. Daily growth increments in otoliths from larval and
adult fishes. Fish. Bull., U.S. 74:1-8.
Campana, S. E., and J. D. Neilson.

1982. Daily growth increments in otoliths of starry floun-
der (Platichthys stellatus ) and the influence of some envi-
ronmental variables in their production. Can. J. Fish.
Aquat. Sci. 39:937-942.
1985. Microstructure of fish otoliths. Can. J. Fish.
Aquat. Sci. 42:1014-1032.
Collins, M. R., and B. W. Stender.

1987. Larval king mackerel Scomberomorus cavalla,
Spanish mackerel (S. maculatus), and bluefish (Po-
matomus saltatrix ) off the southeast coast of the United
States, 1973-1980. Bull. Mar. Sci. 41:822-834.
Davis, R D., T. W. Storck. and S. J Miller.

1985. Daily growth increments in the otoliths of young-of-
the-year gizzard shad. Trans. Am. Fish. Soc. 114:304-
306.

249

FISHERY BULLETIN: VOL. 86, NO. 2

Deuel. D G , J R. Clark, and A J Mansueti

1966. Description of embryonic and early larval stages of
bluefish, Pomatomus saltatrix. Trans. Am. Fish. Soc.
95:264-271.
FiNUCANE, J. H., AND L A. COLLINS.

In press. Reproductive biology of bluefish, Pomatomus
saltatrix, from the southeastern United States. North-
east Gulf Sci.
Geffen, A J

1982. Otolith ring deposition in relation to growth rate in
herring (Clupea harengus) and turbot (Scophthalmus
maximus) larvae. Mar. Biol. (Berl.) 71:317-326.
1987. Methods of validating daily increment deposition in
otoliths of larval fish. In R. C. Summerfelt and G. E.
Hall (editors). Age and growth offish, p. 223-240. Iowa
State Univ. Press, Ames.
Gilmore, J.

1985. Oceanic distribution, abundance, and migration of
bluefish along the east coast of the United States. M.S.
Thesis, State University of New York, Stony Brook,
60 p.

Ingham, M C

1986. Sea surface temi>eratures in the northwestern At-
lantic in 1985. NAFO SCR Doc. 86/75.

ISELIN, C. O'D.

1936. A study of the circulation of the western North At-
lantic. Pap. Phys. Oceanogr. Meteorol. 4:1-101.
Jones, C. J.

1985. Determining age of larval fish with the otolith in-
crement technique. Fish. Bull., U.S. 84:91-103.
KENDALL, A. W , Jr. and L. a. Walford

1979. Sources and distribution of bluefish, Pomatomus
saltatrix, larvae and juveniles off the east coast of the
United States. Fish. Bull., U.S. 77:213-227.
Lassiter, R. R.

1962. Life history aspects of the bluefish, Pomatomus
saltatrix (Linnaeus), fi-om the coast of North Carolina.
M.S. Thesis, North Carolina State College, Raleigh, NC,
91 p.
Marshall. S L , and S. S. Parker.

1982. Pattern identification in the microstructure of sock-
eye salmon (Oncorhynchus nerka) otoliths. Can. J.
Fish. Aquat. Sci. 39:542-547.
McGURK, M. D.

1984. Ring deposition in the otoliths of larval Pacific her-
ring, Clupea harengus pallasi . Fish. Bull., U.S. 82:113-
120.

Morse, W W , M P Fahay, and W. G. Smith

1987. MARMAP surveys of the continental shelf from
Cape Hatteras, North Carolina, to Cape Sable, Nova Sco-
tia (1977-1984). Atlas No. 2. Annual distribution pat-
terns of fish larvae. U.S. Dep. Commer., NOAA Tech.
Memo. NMFS-F/NEC-47, 215 p.
NoRCROss, J J , S L Richardson, W H Massmann, and E. B.
Joseph

1974. Development of young bluefish (.Pomatomus salta-
trix ) and distribution of eggs and young in Virginian
coastal waters. Trans. Am. Fish. Soc. 103:477-497.
Oben, L C

1957. About the drifting approach of fingerling bluefish
(Pomatomus saltatrix) (Linnaeus) to the shores of the
Black Sea in the Karadaga region (1947-1954). Kara-
dagskaia Biol. Stn. Tr. (4):155-157.
POWLES, H.

1981. Distribution emd movements of neustonic young of
estuarine dependent iMugil spp., Pomatomus saltatrix)
and estuarine independent (Coryphaena spp.) fishes off
the southeastern United States. Rapp. P. -v. R^un.
Cons. int. Explor. Mer 178:207-209.

Radtke, R L

1984. Formation and structural composition of larval
striped mullet otoliths. Trans. Am. Fish. Soc. 113:186-
191.

Radtke, R L., and J M Dean.

1982. Increment formation in the otoliths of embryos, lar-
vae and juveniles of the mummichog, Fundulus heterocli-
tus. Fish. Bull., U.S. 80:201-215.

Rice, J A , L B Crowder. and F P Binkowski.

1985. Evaluating otolith analysis for bloater Coregonus
hoyi: do otoliths ring true? Trans. Am. Fish. Soc.
114:532-539.

Sherman, K , W. Smith, W Morse, M Berman, J Green, and

L EllSYMONT.

1984. Spawning strategies of fishes in relation to circula-
tion, phytoplankton production, and pulses in zooplank-
ton off the northeastern United States. Mar. Ecol. Prog.
Ser. 18:1-19.
US. Department of Commerce.

1986. Recreational fishery statistics for 1985. U.S. Dep.
Commer., NOAA, NMFS, 130 p.

WILK, S J

1977. Biological £md fisheries data on bluefish, Po-
matomus saltatrix (Linnaeus). NMFS, NEFC, Sandy
Hook Lab. Tech. Ser. Rep. No. 11, 56 p.

250

ON THE ROLE OF FOOD-SEEKING IN THE SUPRABENTHIC HABIT
OF LARVAL WHITE CROAKER, GENYONEMUS LINEATUS

(PISCES: SCIAENIDAE)

A. E. Jahn, D. M Gadomski, and M. L Sowby'

ABSTRACT

Fish larvae and their prey were sampled from discrete depths within the bottom meter and at
middepth near the 15 m depth contour off southern California. The smallest white croaker larvae
(<2.7 mm NL) occurred mostly at middepth. Mid-sized larvae (2.7 mm to the beginning of flexion)
were almost all collected at the two depths nearest the bottom. All preflexion-stage larvae ate small
(50-300 (im in length) prey, chiefly rotifers, copepod nauplii, tintinnids, and invertebrate eggs.
Although small and mid-size larvae ate these items in different proportions, this difference could not
be ascribed to vertical distribution. Diet of the largest larvae, flexion and postflexion (roughly 5-15
mm), consisted mainly of copepods and differed by >90% from diets of smaller larvae. Though largest
larvae were only captured 50 cm above the bottom, their prey, with one exception (amphipods), were
more abundant at or above 1 m. It was concluded that the observed suprabenthic concentration of
older white croaker larvae was probably not motivated by food-seeking.

Disparity between concentrations of food re-
quired for survival and growth of laboratory-
reared fish larvae and observations of average
concentrations of food organisms in the ocean has
led to the widely accepted idea that aggregations
offish larvae and their food must frequently over-
lap in nature (see reviews by Theilacker and
Dorsey [1980] and Hunter [1981]). Direct and in-
direct evidence for the importance of overlapping
concentrations of larvae and their prey (Lasker
1975, 1978; Govoni et al. 1985; Buckley and
Lough 1987) comes from sampling at fronts and
discontinuities in the pelagic environment. One
interface that attracts many zooplankters is the
seabed itself (Hamner and Carleton 1979; Wish-
ner 1980; Sainte-Marie and Brunei 1985). On the
southern California continental shelf, the seabed
serves as a surface of aggregation for larvae of
numerous fish species (Brewer et al. 1981; Schlot-
terbeck and Connally 1982; Barnett et al. 1984;
Jahn and Lavenberg 1986) and other zooplankton
(Clutter 1969; Barnett and Jahn 1987) and of
large-zooplankton biomass (Jahn and Lavenberg
1986). While it is tempting to suggest a trophic
advantage to the suprabenthic habit of the fish
larvae, near-bottom concentrations of organisms
actually eaten by larval fishes have yet to be
demonstrated along the open coast.

iNatural History Museum of Los Angeles County, 900 Expo-
sition Boulevard, Los Angeles, CA 90007.

In all cases reported, concentration in the near-
bottom zone was greater in older larvae and,
when observations permitted, greater during the
day than at night (Brewer and Kleppel 1986;
Jahn and Lavenberg 1986). The phenomenon is
therefore thought to be behavioral. Possible ad-
vantages of such behavior, including avoidance of
midwater predators, maintenance of position on
the shelf, and increased encounters with high
concentrations of food, have been discussed else-
where (Barnett et al. 1984; Brewer et al. 1984;
Brewer and Kleppel 1986; Jahn and Lavenberg
1986). In discussing the near-bottom schooling
behavior of a larval clupeoid in Japan, Leis (1986)
stated, "knowledge of the biology of epibenthic
fish larvae is too rudimentary to allow a clear
assessment of the advantages and disadvan-
tages. . . ." Whatever the advantages, a seemingly
more answerable question about the near-bottom
habit is what causes the larvae to behave as they
do? In another study from Japan, Tanaka (1985)
showed that juvenile red sea bream, Pagrus
major, exploited suprabenthic copepod popula-
tions, and he speculated that the distribution of
prey was a template for the descent of the fish
from midwaters and its subsequent migration
into estuaries. The question addressed in the
present study was whether the fine-scale layering
of larval fishes was a direct response to that of
their prey field.

Because of the immediate behavioral aspect of

Manuscript accepted Februar>' 1988.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.

251

FISHERY BULLETIN: VOL 86, NO 2

the question posed, a 1-d study was thought
appropriate. Though environmental conditions
on this day might differ from "average", fish lar-
vae were assumed to be capable of a constant
array of behaviors. In other words, response (if
any) of the larvae to the vertical distribution of
their prey was assumed to be a deterministic
rather than a statistical phenomenon. If their
vertical distribution resembled that of their prey,
then food-seeking would remain a plausible ex-
planation for the near-bottom habit; if not, then
other stimuli must be considered important in
shaping these near-bottom concentrations of fish
larvae.

The sampling was planned for daylight hours,
when most feeding by larvae was expected to
occur (Hunter 1981; Govoni et al. 1983). Late win-
ter was chosen because in this season peak larval
abundances of several species of interest to us
(northern anchovy, Engraulis mordax; white
croaker, Genyonemus lineatus; California hal-
ibut, Paralichthys californicus; and sometimes
queenfish, Seriphus politus, often overlap
(Lavenberg et al. 1986). A survey cruise in late
February found moderate-to-high abundance of
the first three species plus California sardine,
Sardinops sa^ox, (all >0.2 m"'^, Lavenberg un-
publ. data), and so this study was scheduled for 19
March 1985 off Seal Beach, CA (lat. 33°41'N,
long. 118°05'W; for a map, see Jahn and Laven-
berg 1986).

As it happened, we chanced to encounter condi-
tions that were less typical than those found on
the February cruise. Only one fish species, white
croaker, was abundant enough to merit analysis,
and an uncommonly reported prey item, rotifera,
was important for small larvae. The diet of
various-sized larvae with respect to the abund-
ance of prey organisms at an array of heights
above the seabed was nevertheless useful in ques-
tioning whether food-seeking shaped the ob-
served larval distribution.

METHODS

Field

At the hour of 0750 PST, an array of Interocean
model S4^ electromagnetic current meters was
set out over the 15 m isobath, with current meters
1, 4, and 8 m above the seabed. These meters were

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

set to record average current vectors and tem-
perature at 5-min intervals. The vessel (RV West-
wind) was then anchored some 200 m seaward of
the current meter array. A Nielson model NCH
fish pump, rated at 227 m'^ h^ at a 2 m head, was
used to sample fish larvae and zooplankton. The
end of the hose was tethered between a 200 kg flat
steel weight and several subsurface floats, with a
pulley arrangement such that divers could adjust
the distance between hose mouth and seabed. A
similar setup was previously found to give re-
peatable, fine-scale resolution at vertical sep-
arations of 25 cm (Jahn and Lavenberg 1986).
Sampling heights above the bottom were 50 cm,
1 m, and 6.7 m. The 15.2 cm diameter hose was
nearly horizontal at the tether point, so that
nominal sampling strata were z ± 7.6 cm. Ves-
sel surge, transmitted through the stiff hose,
caused occasional downward excursions of some
10 cm.

Accompanying each pump sample was a cast of
water bottles for phytoplankton and microplank-
ton analysis. Rigid arrays of horizontally held 4 L
Niskin bottles (of. Owen 1981) were used to take
water samples simultaneously from 25, 50, and
100 cm above the bottom. The bottle array was
designed to be tripped by messenger, but poor
performance led to diver-implemented use after
the second cast. A midwater sample, 7.5 m below
the surface, was obtained via a single Niskin bot-
tle for each sample set.

The sampling plan thus consisted of duplicate
pump samples from each of three strata, each
pump sample to be accompanied by a set of bottle
samples from four standard heights, three within
1 m of the seabed and one at midwater column.
One-liter samples from the bottles were fixed in
Lugol's solution for later identification of phyto-
plankton and microplankton. Pump samples of
15-min duration (approximately 35 m^) were
mainly directed into an overboard, 330 |xm mesh
plankton net for retention of large zooplankton
and ichthyoplankton. Unexpected problems in
reading an inline flowmeter required that vol-
umes be estimated as 2.4 m"^ min"^, based on pre-
vious experience with the pump under similar
conditions aboard the same vessel. To collect
smaller zooplankton, a 5 cm diameter hose led
from the intake side of the fish pump to a 100 jjtm
mesh plankton net. This small-meshed net was
suspended over a watertight box, which was
marked such that exactly 0.5 m'^ could be subsam-
pled for animals too small to be quantitatively
retained by the large net. This subsample, which

252

JAHN ET AL : FOOD-SEEKING LARVAL WHITE CROAKER

was first seived through 330 p.m mesh, took about
10 minutes to obtain; the portion retained on the
330 |jim mesh was added to the contents of the
large plankton net. All pump samples were pre-
served in 5% formalin.

Laboratory

All fish larvae and eggs were sorted from the
large zooplankton samples and identified. All
specimens of white croaker, the only species
abundant in all six collections, were measured
with an eyepiece micrometer in units of 0.024,
0.062, or 0.159 mm, depending on magnification.
Length was measured from tip of snout to end of
straight (NL) or fiexed (FL) notochord or to the
end of the hypural plate when this margin was
vertical (SL). A further designation of de-
velopmental stage indicated the amount of yolk
present: "free embryos" (Balon 1975) had a
relatively massive yolk sac and may or may not
have had functional eyes and mouths; more
advanced individuals with a much-reduced or
totally resorbed yolk sac, fully pigmented eyes,
and an apparently functional mouth were
designated "feeding-stage" larvae, or simply
"larvae".

All larvae, plus a maximum of 20 free embryos
with apparently functional mouths from each col-
lection, were dissected for gut contents analysis
by methods described in Arthur (1976) and
Gadomski and Boehlert (1984). Length, rather
than width, of prey items was measured, because
it was considered a more conservative property of
often crushed specimens and because our concern
was not so much with what the larvae could eat
(Hunter 1981) as with what they did eat. Lengths
of prey items (of copepods, cephalothorax length)
were recorded in 50 fxm classes up to 200 iJim, by
100 ji-rn classes from 200 \x.m to 1 mm, and by 0.5
mm classes at larger sizes. In a few cases, these
size categories were inconvenient, and more in-
clusive ranges were used.

Water bottle samples of phytoplankton and mi-
crozooplankton were prepared following proce-
dures in Utermohl (1931). From a thoroughly
agitated sample, a 50 mL subsample for net
phytoplankton was taken and placed in a settling
chamber overnight (about 14—18 hours). Cells
were identified and counted in 10 ocular fields,
and mean density (cells per liter) calculated as
the number counted scaled by the proportion of
the area of the 10 fields (20.6 mm^ total) to the
area of the slide (510.7 mm^).

Microzooplankton was filtered from a 500 mL
subsample onto a 35 fxm mesh screen, washed
from the screen into a 50 mL settling tube and
allowed to settle overnight. All organisms >50
fxm were counted and identified to taxon and size
category, using the same system as for larval fish
gut contents. Densities were scaled to number per
liter.

The 100 ^jLm zooplankton samples were concen-
trated to 200 mL, then subsampled twice using a
10 mL Stempel pipette. Organisms were identi-
fied and classified to size categories as described
above for larval fish prey. Counts from two sub-
samples were averaged and expressed as number
per m"^.

Data Analysis

The microzooplankton (from water bottles)
data set consisted of six vertical profiles of four
sampling heights each. Principal components
analysis was used to look for vertical layering and
time-correlated changes in the makeup of these
assemblages. A list of taxa present in three or
more samples from at least one sampling height
was chosen. Abundances were log-transformed
[logio(jc + D], and principal components com-
puted from the covariance matrix. Component
scores for each of the 24 samples were used to
make plots in which two- and three-dimensional
groupings were sought that could be clearly re-
lated to sampling height or to the sequence in
which the samples were taken. The taxa having
high loadings on axes (components) identified
with time and vertical trends were subsequently
scrutinized individually. A similar analysis was
done for phytoplankton, but omitted here in the
interest of brevity.

Gut contents were conveniently analyzed by
lumping taxa into the 10 categories: dinoflagel-
late, tintinnid, rotifer, polychaete larva, lamelli-
branch larva, crustacean nauplius, copepodite
and adult copepod, amphipod, invertebrate egg,
and "other". Unidentifiable matter was ignored
in all comparisons. To test for differences in diet
between subsets of larvae, we used an adaptation
of the "bootstrap" (Efron 1982). The test criterion
was the percentage of prey comprised by a major
item in one of the two groups of guts. The null
hypothesis that two sets were not different was
simulated by combining the two data sets and
then, through repeated sampling, determining
the probability of observing the criterion percent-
age from such a mixture.

253

FISHERY BULLETIN: VOL. 86, NO 2

RESULTS

General Observations

The water column was very weakly stratified,
with temperatures of 12.9 ± 0.1°C at 1 m, 13.0
± 0.2°C at 4 m and 14.1 ± 0.1°C 8 m above the
bottom during the time of biological sampling.
Near the bottom, a turbid suspension limited vis-
ibility to arm's length; the surface of the sediment
was never clearly seen on any of the seven de-
scents during the hours of 0930-1630. The mid-
waters below about 3 m from the surface were
densely populated with larvaceans (visually esti-
mated and later confirmed to be about 10 L"^).
Total diatom cell counts (principally Nitzschia
spp.) were of order 10^ L~^ in all samples, bloom
quantities suggestive of recent upwelling (cf.
Tont 1981).

Currents and Plankton

During the hours of biological sampling, cur-
rents ran steadily alongshore to the southeast,
being deflected counterclockwise near the bottom
and ranging from about 14 cm s"^ at 8 m to 6 m
s"^ at 1 m above the seabed. At these current
speeds, one may expect that the approximately
5-h period from beginning to end of biological
sampling should correspond to a minimum spa-
tial spread of 1-2.5 km. Distances of this order
were previously found to be an important length
scale of variation in larval fish abundance ( Jahn
and Lavenberg 1986). Because the spatial dimen-
sion of interest regarding distribution of larval
fish prey was the vertical, we needed to quantify,
at least partially, the effects of time (vertical
migration?) and distance (advection) on the com-
position and vertical dispostion of the plankton.

Component 4 (11%)
0

2 -

1 -

CM
CM

C
0)

c
o

E
o
o

0 -

1 -

■2 -

FIGURE 1. — Projections of microzooplankton
samples onto the first and fourth principal com-
ponent axes. The initial digit represents profile
number, M = midwater, B = near-bottom, final
digit is proximity to bottom (1 = 25 cm, 2 = 50
cm, 3 = 100 cm), see Figure 3.

254

JAHN ET AL.: FOOD-SEEKING LARVAL WHITE CROAKER

Accordingly, the microplankton data set, repre-
senting six vertical profiles separated in time,
was reduced to principal components for exami-
nation of possible time effects.

Twenty-four taxonomic/size categories of mi-
crozooplankton were used to compute principal
component scores for the 24 samples. The first
four principal components accounted for 60% of
the variance. No clear separation of midwater
from near-bottom samples was seen. The first
component, which accounted for 22% of the vari-
ance, separated the near-bottom samples into two
groups, morning to midday and afternoon (Fig. 1),
leaving the midwater samples at intermediate
projections. The midwater samples were in turn
separated by the fourth component (11% of the
variance) into time groups corresponding to those
of the near-bottom set. No stratification by sam-
pling height was seen within the near-bottom
samples, and none of the other axes provided sep-
aration by time. The highest loading variables on
components 1 and 4 (Table 1) were various sizes of
rotifer and, for component 1, three genera of
tintinnids (Favella, Acanthocystis , and Da-
dayella ). Much of the time-correlated variance
structure depicted in Figure 1 thus appears to be
due to change in the size composition of rotifers,
described in a later section, and a decrease in
these three tintinnids near the bottom in late af-
ternoon (Table 2). An identical analysis of the

Table 1. — Loadings of important variables on tfie first and fourth
principal components of microplankton data.

Component 1

Component 4

Variable

Rotifer
150-200 ^Lm
200-300 ^JLm

Tintinnids
Favella sp.
Acanthocyctis
Dadayella sp.

sp.

0.572
-0.380

0.365
0.331
0.314

Rotifer, 100-150 M-m

Egg, 50-100 \i.m

Copepod nauplii,
150-200 ^JLm

0.388
0.349

-0.252

phytoplankton data found no trends in time or
depth.

Larval Fish Abundance

Of 1,125 total fish larvae taken in the six pump
samples, 666 (59%) were white croaker, a deep-
bodied, robust larva (Watson 1982). More than
half (338) of these had absorbed the yolk sac and
were thus of feeding size. The second most abun-
dant feeding-stage larva was an unidentified gob-
iid type (84 specimens), but this taxon was not
taken above 100 cm of the seabed and so was
excluded from the gut analysis. Feeding-stage
California sardine, northern anchovy, and Cali-
fornia halibut — all relatively abundant O0.2
m""^) in the area three weeks earlier — each repre-
sented <1% of the catch. Although the earlier
survey employed oblique bongo net tows, past
comparison of the Nielsen pump with bongo tows
found no significant differences in diversity or
abundance estimates based on similar-volume
samples (R. Schlotterbeck^). We therefore think
the differences between the February survey and
our March samples were due mainly to a real
change in the ichthyoplankton, from a typical
late winter assemblage (McGowen 1987; Walker
et al. 1987) to a more depauperate one.

Vertical Distribution and Feeding
Incidence of Larval White Croaker

White croaker free embryos ranged in abun-
dance from <0.1 m"'^ at 0.5 m to ~1 m"-^ at 1 m
to >2 m-3 at 6.7 m above the bottom. Of 61 free
embryos dissected, none had gut contents.

Feeding-stage larvae of white croaker were
only slightly more abundant at 6.7 m (1.9-2.2
m"3) than at 1 m and 0.5 m (1.1-1.6 m'^), but

3R. Schlotterbeck, Robert Schlotterbeck, Inc., 18842
Ridgeview Cr., Villa Park, CA 92667, pers. commun. April
1986.

Table 2. — Density (cells per liter) of three tintinnids as a function of time and
sampling height. Each set of three numbers gives the density of Favella spp. (F),
Acanthocystis spp. (A), and Dadayella spp. (D).

Time (PST)

Height

1030
A D

1130

1220

1313

1425

1454

(cm)

FAD

A D

FAD

750

100

6
2
0
2

4 20
4 24
6 18
6 4

0 4 22
30 18 24
26 12 44
14 12 44

0
6
0
0

4 12
18 48

2 24
22 54

2 0 12

0 2 26

0 0 4

10 2 26

46 12 14
8 4 24
0 0 4
0 0 0

42 6 18
0 0 26
0 6 12
0 0 0

255

there was a marked gradient in development with
proximity to the seabed (Fig. 2). All the larvae at
6.7 m had unflexed notochords, and most were
<2.5 mm NL. Feeding incidence (proportion of
larvae with nonempty guts) was 78% at this
height. At 1 m, modal larval length was 2.65 mm,

FISHERY BULLETIN: VOL. 86, NO. 2

with a single postflexion specimen (Fig. 2); feed-
ing incidence was 74%. At 0.5 m there were still
some preflexion larvae, but a second length mode
at 6.8 mm represented postflexion-stage larvae.
Feeding incidence was 90%) at 0.5 m above the
bottom, being somewhat greater among flexion

39 n

>
c5

o 0 J

0 ^

13 -1

G. lineatus

GUTS WITH CONTENTS ■
GUTS EMPTY 0

(excludes yolksac stage)

670 cm

(n = 109)

2.0 2.5 3.0

in'i'i

100 cm

(n=70)

2.0 2.5 3.0

' I ' I n ' I ' I ' I

5 10

0 ^

fl#^

2.0 2.5 3.0

50 cm

(n=83)

Length (mm)

Figure 2. — Length frequencies of feeding-stage larvae of Genyonemus lineatus at three sampling heights.

256

JAHN ET AL.: FOOD-SEEKING LARVAL WHITE CROAKER

and postflexion larvae (95%) than among preflex-
ion larvae (82%).

Gut Contents

The white croaker larvae were divided into
three size classes for analysis of gut contents with
regard to height above the bottom: preflexion lar-
vae <2.7 mm (size 1), preflexion larvae >2.7 mm
(size 2), and flexion and postflexion larvae (size

E
o

84 guts
373 prey

E
o

3). The largest preflexion larva was 4.6 mm NL,
and the smallest flexion stage larva was 5.5 mm
FL. The division of preflexion larvae at 2.7 mm
retained all but one specimen at 6.7 m in size 1
while partitioning the preflexion larvae at 1 m
and 0.5 m about equally into sizes 1 and 2 (Figs.
2, 3). Besides the 2.75 mm specimen at 6.7 m, a
single flexion stage larva at 1 m was excluded by
these criteria from the comparisons.

31 guts
82 prey

21 guts
66 prey

DINOFLAGELLATES

TINTINNIDS

ROTIFERS

POLYCHAETE LARVAE

LAMELIBRANCH LARVAE

NAUPLII

COPEPODS

AMPHIPODS

INVERTEBRATE EGGS

OTHER

Hii

ESS

E
o

o
it)

8 guts

15 guts

52 guts

22 prey

26 prey

194 prey

SIZE 1

SIZE 2

SIZE 3

Figure 3. — Percentage contribution of 10 food categories to the diet of larval white croaker at three heights above the
bottom. Size 1 = preflexion larvae <2.7 mm NL; size 2 = preflexion larvae >2.7 mm NL; size 3 = flexion and
postflexion-stage larvae.

257

FISHERY BULLETIN: VOL. 86, NO. 2

Most identifiable prey items fit into the nine
categories: dinoflagellate, tintinnid, rotifer, poly-
chaete larva, lamellibranch larva, crustacean
nauplius, copepodite and adult copepod, am-
phipod, and invertebrate egg (Fig. 3). The "other"
category applied only to size-2 larvae at 1 m (1
Globigerina sp.) and to size-3 larvae (1 zoea, 3
larvaceans, and two large [1 mm] unidentified
spheres).

Guts of preflexion (sizes 1 and 2) larvae from
the three sampling heights contained an array of
small (<300 jxm) organisms that varied mainly in
proportions from mostly rotifers (88%) in size-1
larvae at 6.7 m to a diverse mix of prey numeri-
cally dominated by nauplii in size-2 larvae at 0.5
m (Fig. 3). Percent similarity (overlap) among the
5 groups of preflexion larvae ranged from 24 to
75%. The gut of the single size-2 larva at 6.7 m,
not included in Figure 3, contained two tintin-
nids.

Size-3 larvae had a diet consisting chiefly of
copepodite and adult copepods that overlapped
only 8-9% with size-2 larvae and 1% or less with
the three groups of size-1 larvae. The copepods
eaten by size-3 larvae were mostly Corycaeus an-
glicus (62% of all copepods), unidentified cope-
podites (cyclopoid and calanoid, 25%), and Para-
calanus parvus (9%). Polychaete larvae were
identified only from the presence of setae in the
guts, so the proportion (nominally 16% of all prey
items) of this taxon in the diet is more an indica-
tion of incidence than of numerical importance.
Amphipods, mostly in the length range 1-1.5
mm, were found in white croaker larvae ranging
from 6.5 mm FL to 10.3 mm SL. The gut of the
flexion-stage larva at 1 m, not included in Figure
3, contained three C. anglicus and traces of poly-
chaete setae.

While there can be no doubt that flexion and
postflexion larvae had a different diet than pre-
flexion larvae, the pattern of decreasing propor-
tion of rotifers with increasing size and proximity
to the bottom among preflexion larvae was of
questionable statistical significance. The first
question asked was whether the very high per-
centage of rotifers in the diet of size-1 larvae at
6.7 m was likely to have arisen by chance from a
random sampling of size-1 larvae. Formally
stated, Hq = "all size-1 larvae had the same per-
centage of rotifers". The 123 nonempty guts were
pooled, and random samples of 84 each were
drawn. In 1,000 iterations, <4% of the samples
had >88% rotifers, so it was concluded that lar-
vae at 6.7 m ate significantly more rotifers than

similar-sized larvae near the bottom. The remain-
ing 75 preflexion larvae (sizes 1 and 2) are divided
into 4 small groups at 0.5 and 1 m, so we next
tested for a size effect by pooling across sampling
height, such that the guts of the 39 near-bottom
size-1 larvae contained 64% rotifers, and the 36
size-2 larvae had 36% rotifers. Bootstrapping as
before, <2% of samples of 36 had <36% rotifers,
so it was concluded that size-1 and size-2 larvae
differed in this regard. Further testing (e.g., of a
height effect within sizes) was not done because of
small sample sizes and multiple testing consider-
ations.

Abundance and
Vertical Distribution of Prey

Rotifers, all identified as the brachionoid Tri-
chocerca sp., figured importantly both in the diet
of preflexion larvae and in the time-related vari-
ance structure of the microplankton. As shown in
Table 3, there was a change in the size spectrum
of these animals that coincided approximately
with the time of changing from near-bottom sam-
pling to midwater sampling with the fish pump. It
was only the largest category of rotifer (200-300
|xm, including the "toe") that was found in the
guts of the larvae. The relative abundance of total
rotifers in the plankton at the times and heights
of pump sampling differed very little (25-33% of
all organisms in the 100-300 ^JLm size class), but
the percentage of rotifers in the 200-300 [xm class
increased from 21% (near-bottom, morning) to
86% of all rotifers (midwater, afternoon). The
dominance of rotifers in the diet of size-1 larvae in
midwaters is thus likely related to the larger size
of rotifer resident in the water column when that
height was sampled.

The most notable dietary difference among the
larval size groups analyzed was the switch from
small (50-300 (xm) to larger (0.5-2.5 mm) prey,
principally the copepod Corycaeus anglicus (0.5-
0.8 mm), upon flexion of the notochord. The abun-
dance of Corycaeus from the 100 ixm mesh pump
samples (Table 4) shows that this prey item was
equally or more abundant in midwater than near
the bottom, where all the flexion and postflexion
larvae were captured. (Within the bottom meter,
the similar-sized but more transparent Para-
calanus parvus outnumbered C anglicus by a fac-
tor of 5-20.) The only prey found in numbers in
these larvae that was restricted to the 0.5 m sam-
ples was gammarid amphipods. Larger crus-
taceans— cumaceans, crab and shrimp zoea,

258

JAHN ET AL.: FOOD-SEEKING LARVAL WHITE CROAKER

Table 3. — Density (rotifers per liter) of Trichocerca sp. as a function of time and sampling heighit. Eachi set of three numbers gives the density
of 100-150 M-m, 150-200 txm, and 200-300 ^.m rotifers. Height of simultaneous pump sample is given.

Time (PST):
Pump height:

Sampling

height

(cm)

Time (PST):
Pump height:

Sampling

1030
0.5 m

1130
0.5 m

1220
1 m

100-
150 pim

150-
200 ^.m

200-
300 pirn

100-
150 \i.m

150-
200 M-m

200-
300 M-m

100-
150 pLm

150-
200 M.m

1313
1 m

1425
6.7 m

1454
6.7 m

200-
300 M-m

750

100

height

100-

150-

200-

100-

150-

200-

100-

150-

200-

(cm)

150 fim

200 (xm

300 M-m

150 ^.m

200 M.m

300 M-m

150 M-m

200 M-m

300 M-m

750

100

102

Table 4. — Abundance (animals m 3) of cope-
podite and adult Corycaeus spp. in 100 [im
mesh samples from the fish pump.

Height
(m)

First sample

Second sample

6.7

0.5

1,080
120
180

460
500

140

mysids, and euphausid furcilia larvae — were all
abundant OlO m""^) in the 0.5 m, 330 ixm mesh
samples but with the exception of the callianassa
zoea mentioned above (from the gut of an 11 mm
larva) were not found in these white croaker lar-
vae.

DISCUSSION

The chief drawback of the pumping system
used was its inability to obtain a simultaneous
vertical profile. The sampling sequence left the
possibility that differences among heights might
be confounded by trends in time, as discussed by
Jahn and Lavenberg (1986). Slight time effects
were found among the vertical profiles of mi-
croplankton, increasing the suspicion that the ap-
parent vertical distributions of fish larvae and
macrozooplanktonic prey might have horizontal
components. To contradict the argument that
food-seeking did not bring postflexion larvae near
the bottom, one would need to invoke either an
afternoon increase of some two orders of magni-

tude in copepod abundance (Table 4) or else the
presence of flexion and postflexion larvae
throughout the water column in morning and
midday followed by their sudden disappearance
in the afternoon.

A two-order-of-magnitude change in copepod
species abundance over a distance of roughly 1
km (2 hours at 14 cm s"M is certainly possible;
though zooplankton structures reported from the
southern California continental shelf are gener-
ally larger than this (Star and Mullin 1981; Bar-
nett and Jahn 1987), there is always the possibil-
ity of sampling the edge of a patch. Since no such
edge was evident in the abundance or overall
composition of microplankton or of phytoplank-
ton, it seems unlikely that a macrozooplankton
change of this order occurred. Moreover, the main
copepod eaten, Corycaeus anglicus, is generally
more abundant in midwater than near the bottom
over the shallow shelf (A. Barnetf^), in accord
with its apparent distribution in this study. As to
a possible midwater abundance of postflexion
white croaker larvae, no such concentration has
ever been reported. In some nine vertical profiles
taken in daylight over a 6-d period. Brewer and
Kleppel (1986) took virtually all specimens >3.5
mm in their near-bottom sampler. White croaker
appears similar to another abundant sciaenid,
queenfish, in this regard (cf Jahn and Lavenberg
1986).

4A. Barnett, Marine Ecological Consultants, 531 Encinitas
Blvd.. Encinitas, CA 92024, pers. commun. July 1987.

259

FISHERY BULLETIN: VOL. 86, NO 2

The only unequivocal instance in which a prey
item of larval white croaker was vertically dis-
tributed similarly to the larvae was the trace of
amphipods found in the guts of competent (flexion
and postflexion) larvae. At the lengths of larvae
sampled (<12 mm) the prey were all planktonic
and nearly all about equally abundant in mid-
waters as near the bottom. The small numbers of
amphipods eaten may indicate an incipient tran-
sition to larger, suprabenthic crustacean prey.
The size gap between the large prey of these com-
petent larvae and the smaller prey of preflexion
larvae is probably an artifact of the bimodal size
distribution of sampled larvae. Though all of the
prey eaten by size-1 (<2.7 mm) larvae were <300
|xm in length, the more varied diet of larger pre-
flexion larvae contained some copepods as big as
500 |xm. There is therefore nothing in these data
to suggest that the switch from microplanktonic
to macroplanktonic prey is anything but a grad-
ual transition as the larvae grow.

Brewer and Kleppel (1986) also reported a
change to copepod prey in white croaker larvae
>6 mm. Our findings are further similar to those
of Brewer and Kleppel in that there was no indi-
cation that food-seeking had anything to do with
the descent of larval white croaker from mid-
waters to the near-bottom zone. The other defin-
able dietary trend in this study (besides ontoge-
netic change) was the high percentage of rotifers
eaten by midwater preflexion larvae. This was
apparently related to subtle but important differ-
ences in the available planktonic prey — signifi-
cantly, to a greater abundance of suitable-size ro-
tifers— at the time the midwater stratum was
sampled.

It seems safest to conclude that white croaker
larvae descend toward the bottom for reasons
quite apart from seeking food (see discussions in
Barnett et al. [1984], Brewer and Kleppel [1986],
Jahn and Lavenberg [1986]) and simply eat what-
ever they find there that suits them. Many poten-
tial macroplanktonic prey also favor the near-
bottom layer (Jahn and Lavenberg 1986; Barnett
and Jahn 1987). Older larvae and their prey may
occupy the near-bottom layer for different rea-
sons, or it may be that a single advantage, or set
of pressures, underlies the behavior of these di-
verse planktonic and semi-planktonic taxa. Some
species need to remain near shore, and living in
the bottom boundary layer helps assure this. The
boundary layer also tends to be more turbid than
overlying waters and so may lessen an animal's
jeopardy to visual (biting) planktivores. (The gen-

erality of the latter explanation only holds if
suprabenthic fish larvae are less important
planktivores than other water-column inhabi-
tants— see Gushing 1983.)

Rotifers have never to our knowledge been re-
ported as an important food of ocean-caught fish
larvae, even though the genus Brachionus is com-
monly cultured for feeding larval fish in the labo-
ratory. Schmitt (1986) reported that small,
laboratory-reared larval northern anchovy read-
ily fed upon (unidentified) wild-caught rotifers.
Rotifers are only occasionally abundant in neritic
waters, and never in oceanic waters (J. Beers^).
Their rarity notwithstanding, rotifers have the
ability very rapidly to dominate marine mi-
croplanktonic assemblages (Hernroth 1983), and
their good food quality (Theilacker 1987) and
high secondary productivity for a period of weeks
might constitute a significant enhancement to
growth and survival of a larval fish cohort.

Our previous experience in handling larval
white croaker specimens agrees with the findings
of Brewer and Kleppel (1986) in that lamelli-
branch larvae, easily seen through the body wall,
are a common food for small white croaker larvae.
In our study, this taxon was a minor constituent
of the plankton and of the larval fish diet. We
cannot say how unusual were the circumstances
we encountered, but we know that in terms of
diatom numbers and larval fish diversity these
conditions were not typical of March on the south-
ern California continental shelf That white
croaker larvae appeared to find these conditions
salubrious may be one reason this species is so
successful in southern California (Love et al.
1984).

ACKNOWLEDGMENTS

Thanks go to S. Caddell, R. Feeney, T. Garrett,
R. Lavenberg, J. McGowen, J. Petersen,
J. Rounds, and Captain L. Nufer for able partici-
pation in the field work. D. Carlson-Oda,
J. Rounds, and S. Shiba helped process larval fish
samples, and H. Schwarz helped prepare the
manuscript. K. Zabloudil generously loaned the
current meters, and R. Erdman assisted in proc-
essing data therefrom. We also thank J. Beers
and R. Brusca for help in accessing literature on
rotifers. D. Cohen, R. Lavenberg, and J. Petersen

5J. Beers, Scripps Institution of Oceanography, La Jolla, CA
92093, pers. commun. November 1986.

260

JAHN ET AL.: FOOD-SEEKING LARVAL WHITE CROAKER

reviewed the manuscript. The Southern Califor-
nia Edison Company funded the study.

LITERATURE CITED

Arthur. D K

1976. Food and feeding of larvae of three fishes occurring
in the California Current, Sardinops sagax, Engraulis
mordax, and Trachurus symmetricus . Fish. Bull., U.S.
74:517-530.
Balon, E K.

1975. Terminology of intervals in fish development. J.
Fish. Res. Board Can. 32:1633-1670.
Barnett. a. M . AND A E Jahn.

1987. Pattern and persistence of a nearshore planktonic
ecosystem off southern California. Cent. Shelf Res. 7:1-
25.
Barnett, A M , A E Jahn, P D. Sertic, and W Watson.

1984. Distribution of ichthyoplankton off San Onofre,
California, and methods for sampling very shallow
coastal waters. Fish. Bull., U.S. 82:97-111.
Brewer, G D , and G S Kleppel

1986. Diel vertical distribution of fish larvae and their
prey in nearshore waters of southern California. Mar.
Ecol. Prog. Ser. 27:217-226.

Brewer, G D , G S Kleppel, and M Dempsey.

1984. Apparent predation on ichthyoplankton by zoo-
plankton and fishes in nearshore waters on southern
California. Mar. Biol. (Berl.) 80:17-28.

Brewer, G. D , R L Lavenberg. and G E. McGowen

1981. Abundance and vertical distribution of fish eggs
and larvae in the Southern California Bight: June and
October 1978. Rapp. P. -v. Reun. Cons. int. Explor. Mer
178:165-167.

Buckley, L J., and R G Lough

1987. Recent growth, biochemical composition, and prey
field of larval haddock tMelanogrammus aeglefinus) and
Atlantic cod iGadus morhua ) on Georges Bank. Can. J.
Fish. Aquat. Sci. 44:14-25.

Clutter, R I

1969. The microdistribution and social behavior of some
pelagic mysid shrimps. J. Exp. Mar. Biol. Ecol. 3:125-
155.
Gushing, D H

1983. Are fish larvae too dilute to affect the density of
their food organisms? J. Plankton Res. 5:847-854.

Efron, B

1982. The jackknife, the bootstrap and other resampling
plans. Soc. Ind. Appl. Math., Philadelphia, PA, 92 p.

GaDOMSKI, D M , AND G W BOEHLERT

1984. Feeding ecology of pelagic larvae of English sole,
Parophrys vetulus , and butter sole, Isopsetta isolepis, off
the Oregon coast. Mar. Ecol. Prog. Ser. 20:1-12.

GOVONI, J. J., D E HOSS, AND A J CHESTER

1983. Comparative feeding of three species of larval fishes
in the northern Gulf of Mexico: Brevoortia patronus,
Leiostomus xanthurus, and Micropogonias undulatus.
Mar. Ecol. Prog. Ser. 13:189-199.

GovoNi, J J , A J Chester, D E Hoss. and P B Ortner.

1985. An observation of episodic feeding and growth of
larval Leiostomus xanthurus in the northern Gulf of
Mexico. J. Plankton Res. 7:137-146.

Hamner, W M , AND J. H Carleton

1979. Copepod swarms: attributes and role in coral reef

ecosystems. Limnol. Oceanogr. 24:1-14.
HERNROTH, L.

1983. Marine pelagic rotifers and tintinnids — important
trophic links in the spring plankton community of the
Gullmar Fjord, Sweden. J. Plankton Res. 5:835-846.

Hunter. J R

1981. Feeding ecology and predation of marine fish lar-
vae. In R. Lasker (editor). Marine fish larvae: morphol-
ogy, ecology, and relation to fisheries, p. 34-77. Wash-
ington Sea Grant Program, Seattle.
Jahn. A E.. and R. J Lavenberg.

1986. Fine-scale distribution of nearshore, suprabenthic
fish larvae. Mar. Ecol. Prog. Ser. 31:223-231.
Lasker. R

1975. Field criteria for survival of anchovy larvae: the
relation between inshore chlorophyll maximum layers
and successful first feeding. Fish. Bull., U.S. 73:453-
462.
1978. The relation between oceanographic conditions and
larval anchovy food in the California current: identifica-
tion of factors contributing to recruitment. Rapp. P. -v.
Reun. Cons. int. Explor. Mer 173:212-330.
Lavenberg, R. J., G. E McGowen, A E Jahn. J. H. Petersen,
AND T C Sciarrotta

1986. Southern California nearshore ichthyoplankton:
a study of abundance patterns. CalCOFI Rep. 27:53-
64.
Leis, J M

1986. Epibenthic schooling by larvae of the clupeid fish
Spratelloides gracilis . Jpn. J. Ichthyol. 33:67-69.

Love, M S , G E. McGowen, W Westphal, R. J Lavenberg, and
L Martin

1984. Aspects of the life history and fishery of the white
croaker, Genyonemus lineatus (Sciaenidael, off Califor-
nia. Fish. Bull., U.S. 82:179-198.

McGowen, G E

1987. Coastal zone ichthyoplankton assemblages off
Southern California. Ph.D. Thesis, Univ. Southern
California, Los Angeles, 304 p.

Owen. R W.

1981. Microscale plankton patchiness in the larval an-
chovy environment. Rapp. P. -v. Reun. Cons. int. Ex-
plor. Mer 178:364-368.

SaINTE-MaRIE, B . AND P BRUNEL.

1985. Suprabenthic gradients of swimming activity by
cold-water gammaridean amphipod Crustacea over a
muddy shelf in the Gulf of Saint Lawrence. Mar. Ecol.
Prog. Ser. 23:57-69.

Schlotterbeck, R E , and D W Connally

1982. Vertical stratification of three nearshore southern
California larval fishes (Engraulis mordax, Genyonemus
lineatus, and Seriphus politus). Fish. Bull., U.S.
80:895-902.

SCHMITT. P D

1986. Prey size selectivity and feeding rate of larvae of the
northern anchovy, Engraulis mordax Girard. CalCOFI
Rep. 27:153-161.

Star. J. L.. and M M Mullin.

1981. Zooplanktonic assemblages in three areas of the
North Pacific Ocean as revealed by continuous horizontal
transects. Deep-Sea Res. 28A:1303-1322.
Tanaka, M.

1985. Factors affecting the inshore migration of pelagic
larval and demersal juvenile red sea bream, Pagrus
major, to a nursery ground. Trans. Am. Fish. Soc.
114:471-477.

261

FISHERY BULLETIN: VOL. 86, NO. 2

Theilacker. G

1987. Feeding ecology and growth energetics of larval

northern anchovy, EngrauUs mordax. Fish. Bull., U.S.

85:213-228.
Theilacker, G , and K Dorsey

1980. Larval fish diversity, a summary of laboratory and
field research. FAO Intergov. Oceanogr. Comm. Work-
shop Rep. No. 28, p. 105-142.

TONT, S. A.

1981. Historical changes of diatom abundance off south-
ern California as reflected in sea surface temperature, air
temperature, and sea level. J. Mar. Res. 39:191-201.

Utermohl. N

1931. Neue Wege in der quantitativen Erfassung des

Planktons. Verb. Int. Ver. Limnol. 5:567-597.
Walker, H J . Jr , W Watson, and A. M Barnett.

1987. Seasonal occurrence of larval fishes in the near-
shore Southern California Bight off San Onofre, Califor-
nia. Estuarine Coastal Shelf Sci. 25:91-109.
Watson, W.

1982. Development of eggs and larvae of the white
croaker, Genyonemus lineatus Ayres (Pisces:Sciaenidae),
off the southern California coast. Fish. Bull., U.S.
80:403-417.
WiSHNER, K F

1980. The biomass of the deep-sea benthopelagic plank-
ton. Deep-Sea Res. 27:203-216.

262

NEW MARINE DECAPOD CRUSTACEANS FROM WATERS INFLUENCED BY
HYDROTHERMAL DISCHARGE, BRINE, AND HYDROCARBON SEEPAGE

Austin B. Williams^

ABSTRACT

Five species of decapod crustaceans new to science are described. These are caridean shrimps of the
family Bresiliidae — Alvinocaris markensis from a Mid-Atlantic Rift Valley hydrothermal field, A.
muricola from a cold brine seep at the foot of the West Florida Escarpment in the Gulf of Mexico, and
A. stactophila from a hydrocarbon seep on the continental slope of the northern Gulf of Mexico, with
a key to the species of Alvinocaris ; a squat lobster of the family Galatheidae — Munidopsis alvisca
from the Guaymas Basin and from the Juan de Fuca and Explorer ridges in the eastern Pacific; and
a brachyuran crab of the family Bythograeidae — Bythograea mesatlantica from a Mid-Atlantic Rift
Valley hydrothermal field. Species of both Alvinocaris and Bythograea are now known from the
eastern Pacific and Mid-Atlantic. Munidopsis species are widely represented in the world oceans.

Deep ocean hydrothermally active fields and
waters influenced by brine and hydrocarbon seeps
continue to yield species new to science. Such en-
vironments were unknown until explored with
the aid of submersible research vessels from
which observations and collections could be ac-
complished. The species of decapod crustaceans
reported here come from hydrothermal fields in
the Mid- Atlantic Rift Valley, the Guaymas Basin
in the Golfo de California, and Juan de Fuca and
Explorer Ridges in the northeastern Pacific, a
cold brine seep at the foot of the West Florida
Escarpment, and a hydrocarbon seep on the con-
tinental slope of the northern Gulf of Mexico.
These are scattered localities that exhibit diverse
environmental conditions but that are bound
together by the common thread of chemotrophic
food chains (Childress et al. 1986; Brooks et al.
1987).

The material from the Mid-Atlantic Rift Val-
ley, West Florida Escarpment, and Guaymas
Basin was observed and collected by scientists
working with the aid of the DSRV Alvin and RV
Atlantis II based at the Woods Hole Oceano-
graphic Institution. That from the northern Gulf
of Mexico came from the Minerals Management
Service Northern Gulf of Mexico Outer Continen-
tal Slope (MMS/NGOMCS) Regional Office Proj-
ect, involving observation and collection of mate-
rial by scientists from LGL Ecological Research

iSystematics Laboratory, National Marine Fisheries Service,
NOAA, National Museum of Natural History, Washington,
D.C. 20560.

Manuscript accepted January 1988.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.

Associates and Texas A&M University, with the
aid of the submersible research vessel Johnson-
Sea-Link and its support vessels. Specimens from
Explorer and Juan de Fuca Ridges were collected
with the aid of the Canadian DSRV Pisces IV and
its support vessels.

All specimens studied are deposited in the
Crustacean Collection of the United States Na-
tional Museum of Natural History, Smithsonian
Institution, Washington, D.C. 20560.

CARIDEA: BRESILIIDAE

Alvinocaris Williams and Chace, 1982

Three species of bresiliid shrimps described
below as new to science are placed in the hereto-
fore monotypic genus Alvinocaris. Certain fea-
tures of these species necessitate minor changes
in the generic diagnosis by Williams and Chace
(1982) as follows: Rostrum with or without ven-
tral teeth. Telson with 2-5 pairs of principal
spines on posterior margin. Strong median ster-
nal spine between posterior pair of pereopods.

Moreover, the branchial formula seems uni-
formly fixed in this genus as well as in the genus
Rimicaris Williams and Rona, 1986. The arrange-
ment, figured in Williams and Chace (1982) and
Williams and Rona (1986) may be described as
follows:

Phyllobranchs extensively developed in 2 se-
ries; asymmetrically Y-branched pleurobranchs
with relatively short ventral and progressively
longer and more expansive dorsal ramus associ-

263

FISHERY BULLETIN; VOL. 86, NO. 2

ated with pereopods 1-5; smaller and more nearly
symmetrical U-branched arthrobranchs of more
nearly uniform size associated with third maxil-
liped and pereopods 1-4.

Key to Known Species oi Alvinocaris

1. Telson with terminal margin slightly con-
cave in midline and bearing 3-5 pairs of

principal spines

A. lusca Williams and Chace

Telson with terminal margin convex and
bearing only 2 pairs of principal spines ... 2

2. Rostrum with ventral margin bearing 0-1

subterminal ventral spines

A. stactophila new species

Rostrum with ventral margin bearing 4 or
more subterminal ventral spines 3

3. Abdominal segment 3 with pleural margin
entire A. markensis new species

Abdominal segment 3 with pleural margin

obscurely serrate

A. muricola new species

Alvinocaris markensis new species

Figures 1, 2, 7

Material— \JSnU 234286, Holotype 9
(crushed), USNM 234287, Paratypes, 2 9 (dam-
aged), Mid-Atlantic Rift Valley about 70 km
south of Kane Fracture Zone (see Leg 106 Ship-
board Scientific Party [1986]; Ocean Drilling Pro-
gram Leg 106 Scientific Party [1986]),
23°22.09'N, 44°57.12'W, 3,437 m, Alvin Dive
1683, MARK vent, Stn. 1, scoop, 30 May 1986,
pilot D. Foster, observers S. Humphris and J. Ed-
mond. From NSF Ocean Drilling Program-Leg
106, NSF Grant OCE-8311201 to J. F. Grassle,
Woods Hole Oceanographic Institution, Woods
Hole, MA.

Measurements in mm. — Holotype 9, postor-
bital carapace length 4.16, rostrum 2.3, maxi-
mum carapace height 3.3, total length about
15.6. Paratype 9, same 2.7, 1.6, 2.4, 10.9.

Description. — Integument extremely thin,
264

membranous, shining, with a few minute puncta-
tions. Rostrum (Fig. la, 6) almost straight,
slightly elevated above horizontal in distal half,
sharply pointed tip usually reaching to between
midlength of second and tip of third peduncular
articles of antennule; dorsal margin raised into
thin serrate crest containing 12-17 teeth,
strongest in central sector of row, with about 1/3
length of crest continued onto carapace; ventral
margin less prominent and armed with 5-8 sub-
terminal teeth; tooth formulas examined, 17/8
(holotype), 17/5, 12/5 (apparently some subtermi-
nal dorsal teeth fused); strong lateral carina
broadened proximally and confluent with orbital
margin. Carapace with acute antennal spine dis-
tinct; pterygostomian spine acuminate and
prominent. Indistinct antennal groove curving
ventrad to intersect associated indistinct groove
at about midlength of carapace and continuing
posteriad.

Abdomen of female (Fig. Ic) apparently
broadly arched dorsally (all specimens examined
are crushed), gradually tapering posteriorly, nar-
rowest part of sixth somite about 1/2 width of first
somite; pleura of 3 anterior somites broadly
rounded, that of fourth somite drawn posterolat-
erally to strong acuminate spine flanked dorsally
by 0 or 1 much more slender and smaller spine;
posterolateral corner of fifth pleuron strongly
acuminate to nearly right angled and flanked
dorsally by 0-2 spines of variable size analogous
to condition on somite 4, spine number possibly
age related; sixth somite with middorsal length
about 1.9 that of fifth, broad based midlateral
spine overlapping base of telson, smaller pos-
terolateral spine acute; fourth somite with small
erect spine on sternite and fifth with analogous
strong, posteriorly directed spine. Telson (Fig.
\d) elongate subrectangular, length about 3.5
anterior width, 5.8 posterior width, and about
1.75 length of sixth somite, not including poste-
rior spines; armed with 6-8 dorsolateral spines
of nearly uniform size, sometimes bilaterally
unequal in number; posterior margin convex,
armed with 2 principal spines at each corner
and 10-12 feathered strong setae on margin be-
tween.

Eyes (Fig. la , 6 ) with cornea imperfectly devel-
oped, unfaceted though diffusely pigmented, glob-
ular to ovate in outline and with prominent spine
on anterodorsal edge.

Antennular peduncle (Fig. la, 6) reaching end
of antennal scale; basal article 1.3 length of sec-
ond and about 3.0 length of third, stylocerite well

WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS

Figure 1. — Alvinocans markensis, holotype $ : a part of cephalothorax and anterior appendages, lateral; b rostrum, eye, anten-
nular peduncle, antennal scale, dorsal of left side; c abdomen, lateral; cf tail fan; cheliped, e mesial, /"lateral; g pereopod 2,
h chela. Scales = 2 mm: I ia, c , d, e , f, g, h)\2 ib).

265

FISHERY BULLETIN: VOL. 86, NO. 2

separated from peduncle, tapering to slender
elongate tip reaching about 1/4 length of second
article, basal article with distodorsal margin
flanked by transverse row of setae, extended into
strong lateral spine reaching level equal to that of
stylocerite and closely appressed to second article;
shorter second article with strong mesiodistal ap-
pressed spine. Dorsolateral flagellum about 1.5
length of carapace, thickened proximal half bear-
ing conspicuous ventral setae; ventromesial flag-
ellum somewhat more slender in lateral view and
shorter.

Antennal scale (Fig. la, 6) about 2.5 as long as
wide, distolateral tooth falling short of distome-
sial apex of broadly rounded distal margin of
blade; basal article with strong ventrolateral
spine; flagellum slightly exceeding length of body
(missing from holotype).

Mandibles (Fig. 2g) similar, with 2-segmented
palp, incisor process broad and armed with 8
marginal teeth, slender molar process simple,
divergent, its narrowly rounded tip minutely
setose.

First maxilla (Fig. 2h) with proximal endite
asymmetrically oval-triangular, distal margin
bearing many long setae; distal endite with nar-
rowed base but broadened distally, armed with
many (about 37) short spines on mesial margin
and with scattered longer spinules marginally
and submarginally beyond either end of spine
row; palp scarcely bifurcated, with long distal
spine on obsolescent proximomesial branch and 1
shorter submarginal spine on distal branch.

Second maxilla (Fig. 2i) with proximal endite
represented by 2 similar lobes; distal endite sub-
triangular, expanded mesiodistally and paral-
leled laterally by narrow somewhat twisted palp,
scaphognathite with anterior lobe rectangulo-
ovate, fringed with uniformly long, silky setae on
anterior and mesial borders, shorter setae along
entire lateral margin; posterior lobe narrowly
ovate-triangular, fringed on blunt tip and adja-
cent mesial margin by strikingly long, tangled,
strong setae preceded proximally by shorter setae
similar to those on lateral margin.

First maxilliped (Fig. 2j, partly flattened view)
with irregularly fusiform endite, short palp much
exceeded in length and size by leaflike exopod,
epipod obscurely bilobed.

Second maxilliped (Fig. 2k, I) somewhat pedi-
form but flattened, mesial margin of articles
bearing long, feathered setae, mesial surface of
terminal article densely setose, tip of exopod ex-
ceeding leaflike epipod.

Third maxilliped (Fig. 2m, n) slender, 5-
segmented, reaching beyond antennular pedun-
cle; terminal segment trigonal in cross section,
tapered distally, bearing 3 terminal spines,
oblique tracts of dense setae along mesial surface;
similar tract of setae on carpus and another less
conspicuous group on merus-ischium, latter with
distolateral spine at articulation with carpus; ex-
opod much reduced, ovate-triangular, without
lash.

First pereopods (Fig. le, f) chelate, subequal;
fingers curved ventrally and slightly laterad;
dactyl much more slender than and slightly
longer than fixed finger; mesial surface of each
finger convex, lateral surface deeply concave; pre-
hensile surfaces uniformly offset, closing without
gape, each armed with row of almost uniform
teeth so closely set as to be almost contiguous,
line of sensory hairs mesial to cutting edges,
acute tip of dactyl slightly spooned by elongate
teeth slanted distad and curving around its exter-
nal edge. Leg not reaching tip of third maxilliped
and exceeding antennal peduncle. Palm of holo-
type female inflated, length slightly greater than
height and shorter than fingers (0.60); low ridge
ending in small hooked spine on proximomesial
surface near articulation with carpus. Carpus
longer than palm; bearing oblique ventral crest
ending in strong distoventral spine and flanked
mesially by patch of setae on triangular raised
area; rectangular distal notch above spine fol-
lowed by oblique distomesial margin leading to
poorly defined spine at condyle articulating with
palm; distolateral margin with rounded ventral
corner leading to sinuous border above it bearing
2 lobes near articulation with palm. Merus some-
what swollen is distal half and bearing small dis-
tomesial spine, distinct from ischium but fused to
it.

Second pereopod (Figs. 1^, h; 11) shorter and
more slender than first, but reaching beyond an-
tennal peduncle by about length of fingers. Fin-
gers slightly shorter than palm, similar in size
and shape; opposed edges without gape, each
spineless proximally, but distal half pectinate
with single row of spines directed obliquely distad
and increasing slightly in size to end in notice-
ably stronger spine crossing opposite member
when closed. Carpus slender, about 0.9 length of
chela; merus and ischium unarmed.

Third to fifth pereopods (Fig. 2a-f) similar in
length and structure, third reaching distal edge of
antennal scale. Length articles of these legs in
holotype 9, mm:

266

WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS

Figure 2. — Alvinocaris markensis , holotype 9: a pereopyod 3, b dactyl; c pereopod 4, d dactyl; e p)ereopK)d 5, f
dactyl; g mandible; h maxilla 1; i maxilla 2; j maxilliped 1 partly flattened; maxilliped 2, k endopod, ventral, I part
of exopod dorsal; maxilliped 3, m ventral, n dorsal, distal articles only. Scales: Uk, I) = 1 mm; 2 (6, d,f,g, A) = 1 mm;
3(a,c,e,i,j,m,n) = 2 mm.

267

FISHERY BULLETIN: VOL. 86, NO. 2

3rd
4th
5th

ischio-
merus

6.14
5.76
5.44

carpus propodus

2.56
2.18
2.30

3.78
4.12
5.44

dactyl

0.64
0.70
0.64

Each short dactyl armed with 5 spines on flexor
surface, grading from small proximally to longest
and strongest distally, often a sensory seta on
extensor surface. Propodi with setae along flexor
surface progressively more crowded distally.
Carpi with distodorsal extension projecting as a
stop along proximal part of propodal extensor sur-
face. Third leg stronger, at least in merus-
ischium, than fourth and fifth; merus of third and
fourth with closely appressed ventral spine at 1/2
and 3/4 length, that of fifth with spines at 1/3 and
2/3 length, distal spine strongest in each case;
ischium of third and fourth leg with 2 spines in
line with those on merus.

Pleopods well developed, pair 1 with endopods
about half length of exopods, tapering to acute
tip; appendices internae simple, that of pair 5
with blunt tip.

Uropod (Fig. Id) with rami subequal in length,
slightly exceeding distal end of telson, lateral
ramus with movable spine mesial to smaller dis-
tolateral tooth, diaeresis sinuous.

Remarks. — Remarks are given in the account
for A. stactophila.

Etymology. — The name is taken from an
acronym for the site of collection in the Mid-
Atlantic Ridge Valley about 70 km south of an
area known as the Kane Fracture Zone, "MARK",
and the Latin genitive suffix "ensis".

Alvinocaris muricola new species

Figures 3, 4, 7

Material .—\JSnU 234288, Holotype c^ (ceph-
alothorax and abdomen broken apart), USNM
234289, Allotype 9, West Florida Escarpment,
26°01'N, 84°54.61'W, 3,277 m, Alvin Dive 1754,
15 October 1986, pilot W. Sellers, observers
R. Carney and B. Hecker. USNM 234290, Para-
type 9 , West Florida Escarpment, same locality,
Alvin Dive 1753, 14 October 1986, pilot P. Tib-
betts, observers R. Carney and G. Knauer. All
from Barbara Hecker, Lamont Geological Ob-
servatory, Columbia University, Palisades, NY.

Measurements in mm. — Holotype 6, postor-
bital carapace length 6.4, rostrum 4.4, maximum
carapace height 4.5. Allotype 9 , same, 6.4, ros-
trum broken, 5.6.

Description. — Integument thin, shining, mi-
nutely punctate. Rostrum (Fig. 3a, b) almost
straight to slightly upturned in distal half,
sharply pointed tip reaching to proximal part of
third peduncular article of antennule; dorsal mar-
gin raised into thin serrate crest containing 17-21
teeth varying from obliquely erect in proximal
part to nearly horizontal, shorter and more dis-
tant distally, about 1/3 length of crest continued
onto carapace; ventral margin much less promi-
nent and armed with row of 6 correspondingly
smaller subterminal teeth, sometimes obscure;
lateral carina broadened proximally and conflu-
ent with orbital margin. Carapace (Fig. 3a ) with
broad based but slender, acuminate antennal
spine; pterygostomian spine correspondingly
acuminate and prominent. Prominent anterior
antennal carina curving posteroventrally to in-
tersect obliquely with carina extending from
pterygostomian spine at about midlength of cara-
pace, associated groove continuing indistinctly
posteriad.

Abdomen (Fig. 3d,e,f) of both male and female
broadly arched dorsally, gradually tapering dis-
tally, narrowest part of sixth somite less than 2/3
(0.60) width of first somite; pleura of 3 anterior
somites broadly rounded, margin of third slightly
serrated, that of fourth somite drawn posterolat-
erally to strong spine flanked dorsally by 0-3
more slender and smaller spines and preceded on
ventral margin by 0-2 small spines; number, po-
sition, and shape of either lateral or ventral
spines may be asymmetrical; posterolateral cor-
ner of fifth pleura acuminate and flanked dorsally
by 1 or 2 spines analogous to those on somite 4;
sixth somite with middorsal length about 1.7 that
of fifth, broad-based midlateral spine overlapping
base of telson, smaller posterolateral spine acute;
fourth and fifth somites each with strong, posteri-
orly directed spine on sternite. Telson (Fig. 3^)
elongate subrectangular, length about 3.0 ante-
rior width, 6.8 posterior width, and about 1.4
length of sixth somite, not including posterior
spines; armed with 7 dorsolateral spines of nearly
uniform size; posterior margin convex, armed
with 2 principal spines at each corner and 10 or 11
feathered strong setae on distal margin between.

Eyes (Fig. 3a, b) with cornea imperfectly devel-
oped, unfaceted though diffusely pigmented.

268

WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS

Figure 3. — Alvinocaris muricola, holotype d: a cephalothorax and anterior appendages, lateral; b rostrum, eye, antennular
peduncle, antennal scale, dorsal of right side; c antennular peduncle, distal articles, mesial; abdomen with variations in spination
of pleural margins fof allotype 1 1, d lateral, e margin of pleuron 3 from opposite side, /"(of holotjT)e d ) segments 4 and 5; g tail
fan; h median sternal spine between fifth pereopods; i median sternal spine on abdominal segment 5; cheliped, j mesial, k
lateral; I, m, n, o pereopods 2, 3, 4, 5. Scales = 1 mm: 1 (a-g,j-o); 2 (h, i).

269

FISHERY BULLETIN: VOL. 86, NO. 2

ovate in outline, though fused to each other
mesially, and each with a short upturned spine on
anterior surface.

Antennular peduncle (Fig. 3a, 6, c) reaching
beyond end of antennal scale; basal article 1.1 as
long as second and about 2.5 as long as third, all
measured on ventral margin; stylocerite well sep-
arated from peduncle, tapering to slender elon-
gate tip reaching tip of distolateral spine on basal
article; latter exceeding distodorsal margin of ar-
ticle, fringed by transverse subdistal row of setae,
and closely appressed to second article, distome-
sial spine much smaller; shorter second article
with stronger mesiodistal spine. Dorsolateral
flagellum about length of carapace, thickened in
basal 2/3, with annulations, except at base, longer
in female than in male and much longer than in
whiplike distal part; ventromesial flagellum
somewhat shorter and with annulations of vari-
able but shorter length.

Antennal scale (Fig. 3a, 6) about twice as long
as wide, distolateral tooth strong, falling slightly
short of broadly rounded distal margin of blade;
basal article with acute distal spine ventrally;
flagellum (broken in material studied) probably
slightly exceeding length of body.

Mandibles (Fig. 4a) similar, with 2-segmented
palp, incisor process broad and armed with 8 mar-
ginal teeth, slender molar process simple, diver-
gent, its narrowly rounded tip minutely setose.

First maxilla (Fig. 46) with proximal endite
asymmetrically oval-triangular, distal margin
bearing many long setae; distal endite with nar-
rowed base but broadened distally, armed with
many (about 30) short spines on mesial margin
and with scattered longer spinules submarginally
and marginally beyond either end of spine row;
palp scarcely bifurcated, with long distal spine on
obsolescent proximomesial branch and 1 shorter
submarginal spine on distal branch.

Second maxilla (Fig. 4c, c?) with proximal en-
dite represented by 2 similar lobes; distal endite
subtriangular, expanded mesiodistally and paral-
leled laterally by narrow, somewhat twisted palp;
scaphognathite with anterior lobe rectangulo-
ovate, fringed with uniformly long, silky setae on
anterior and mesial borders, uniformly shorter
setae along entire lateral margin; posterior lobe
narrow and acuminate, fringed on blunt tip and
adjacent mesial margin by strikingly long,
strong, tangled setae preceded proximally by
shorter setae similar to those on lateral margin.

First maxilliped (Fig. 4e, f) with irregularly
fusiform endite, short palp concealed and much

exceeded in length and size by leaflike exopod,
epipod obscurely bilobed; indistinct mesial lobule
on exopod possibly representing incipient lash.

Second maxilliped (Fig. 4g, h) somewhat pedi-
form but flattened, mesial margin of articles
bearing long, feathered setae, mesial surface of
terminal article densely setose, exopod barely ex-
ceeding leaflike epipod.

Third maxilliped (Fig. 4i, j) slender, 5-
segmented, reaching beyond antennular pedun-
cle; terminal article trigonal in cross section, ta-
pered distally, bearing 3 spines, transverse tracts
of dense setae along mesial surface; similar tract
of setae on carpus and another conspicuous group
mesiodistally on merus-ischium, latter with stout
distolateral spine at articulation with carpus;
exopod much reduced, subtriangular, without
lash.

First pereopods (Figs. 3^, k; If-k) chelate,
subequal and sexually dimorphic, at least in fully
mature individuals; fingers curved ventrally and
slightly laterad; dactyl more slender than and
with level of tip slightly shorter than or equal to
that of fixed finger; mesial surface of each finger
convex, lateral surface concave, with opposed sur-
faces uniformly offset; closing without gape, each
armed on prehensile edge with row of almost uni-
form teeth so closely set as to be almost contigu-
ous, acute tip of dactyl slightly spooned by elon-
gate teeth slanted distad and curving around its
external edge; line of sensory hairs mesial to cut-
ting edges. Leg shorter than to almost equaling
third maxilliped. Palm of holotype male inflated
laterally, but apparently somewhat irregularly
concave mesially, length 1.4 greatest height and
longer than fingers; palm relatively shorter in
allotype female, 0.3 length of fingers. Carpus
shorter than palm, with oblique ventral crest end-
ing in strong distolateral spine, flanked mesially
by patch of setae on polygonal raised area. Merus
somewhat swollen in distal half, distinct from is-
chium but fused to it, neither armed.

Second pereopod (Figs. 3/ ; 7e ) shorter and more
slender than first, reaching about to end of anten-
nal peduncle; fingers slightly longer than palm,
similar in size and shape, opposed edges without
gape, each pectinate with single row of teeth in
distal half directed obliquely distad and increas-
ing slightly in size to end in noticeably stronger
tooth crossing opposite member when closed, but
spineless proximally; carpus slender, about 1.16
longer than chela; merus and ischium unarmed.

Third to fifth pereopods (Fig. 3m, n,o) similar
in length and structure, third reaching to tip of or

270

WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS

Figure 4. — Aluinocaris muricola, holotype 6: a mandible; 6 maxilla 1; maxilla 2, c ventral, d palp dorsal; maxilliped 1,
e ventral, /"dorsal; maxilliped 2, g ventral, /i dorsal; maxilliped 3, j ventral, _/ dorsal; ^ endopod of pleopod 1; /appendix
masculina, pleopod 2. Scales: I (e, f, g,j , o, k, I) = 1 mm; 2 (c, d) = 1 mm; 3 (/) = 0.2 mm; 4 (a, 6) = 0.2 mm.

271

FISHERY BULLETIN; VOL 86, NO 2

slightly beyond antennal scale. Length articles of
these legs in holotype 6 , mm:

ischium merus carpus propodus dactyl

3rd

1.76

3.84

2.33

2.34

0.35

4th

1.60

3.52

1.98

3.04

0.06

5th

1.60

3.20

2.43

4.96

0.42

Each short dactyl armed with about 4-6 corneous
spines on flexor surface, grading from small prox-
imally to longest and strongest distally; carpus of
each with distodorsal extension projecting as a
stop along proximal part of propodal extensor
surface; third leg stronger, at least in merus-
ischium, than fourth and fifth, but propodus suc-
cessively longer from third to fifth; merus of each
with ventral spine at 1/3 and 2/3 length; ischium
of third, fourth, and fifth leg with 2 spines in line
with those on merus.

Pleopods well developed; first pair with en-
dopods about 1/2 length of exopods in both sexes,
narrowed into distomesial projection in male (Fig.
4k) but evenly tapered in female; appendix mas-
culina (Fig. 41) of second pair in male (holotype)
armed with 7 slender spines extending beyond
level of simple slender appendix interna; en-
dopods of third to fifth in male and second to fifth
in female with simple slender appendix interna,
but that of fifth blunt tipped.

Uropod (Fig. 3^) with rami subequal in length,
lateral ramus slightly exceeding distal end of tel-
son, and with movable spine mesial to smaller
distolateral tooth, diaeresis sinuous.

Remarks. — Remarks are given in the account
for A. stactophila.

Etymology. — The name is from the Latin
"muria", brine, and "cola", inhabiting, for associ-
ation of the species with cold brine seeping from
the base of the West Florida Escarpment.

Alvinocaris stactophila new species

Figures 5, 6, 7

Material .—\]SnM 234291, Holotype d,
USNM 234292, Allotype 9, USNM 234293,
Paratypes, 5 c?, 2 9; north central Gulf of Mexico
about 129 km (80 miles) S of Louisiana,
27°46.94'N, 9r30.34'W, 534 m, Johnson-Sea-
Link Dive 1879, 28 September 1986, Bush Hill
hydrocarbon seep. From Linda H. Pequegnat and

Randall Howard, LGL Ecological Research Asso-
ciates, Bryan, TX, supported by partial funding
for Minerals Management Service-Northern Gulf
of Mexico contract 14-12-0001-30212.

Measurements in mm. — Holotype 6, postor-
bital carapace length 7.0, rostrum 2.7, maximum
carapace height 5.3, total length about 25. Allo-
type 9 , same 6.8, 2.0, 5.9, 24. Paratype 6 , same
4.2, 1.9, 3.2, total length not measured; paratype
9, same, 4.9, 2.0, 4.1.

Description. — Integument thin, shining, mi-
nutely punctate. Rostrum (Fig. 5a, b) almost
straight, imperceptibly elevated above horizontal
in distal half, sharply pointed tip usually reach-
ing proximal level of second peduncular article of
antennule, but sometimes to proximal part of
third peduncular article; dorsal margin raised
into thin serrate crest containing 12-17 teeth,
strongest in central sector of row, with about 1/2
length of crest continued onto carapace; ventral
margin less prominent and armed with 0 or 1
subterminal tooth; sample tooth formulas 11/1,
12/0, 14/1, 17/1; lateral carina broadened proxi-
mally and confluent with orbital margin. Cara-
pace (Fig. 5a, b) with buttressed acuminate an-
tennal spine distinct; pterygostomian spine
acuminate and prominent. Prominent anterior
antennal carina curving posteroventrally to in-
tersect obliquely with carina extending from
pterygostomial spine at about midlength of cara-
pace, associated groove continuing indistinctly
posteriad.

Abdomen (Fig. 5d) of both male and female
broadly arched dorsally, gradually tapering dis-
tally, narrowest part of sixth somite less than 1/2
(0.44) width of first somite; pleura of 3 anterior
somites broadly rounded, that of fourth somite
drawn posterolaterally to acuminate spine
flanked dorsally by 0-3 much more slender and
smaller spines; posterolateral corner of fifth
pleuron varying from strongly acuminate to
nearly rectangular and flanked dorsally by 2-5
spines analogous to those on somite 4; sixth
somite with middorsal length about 1.8 that of
fifth, broad-based midlateral spine overlapping
base of telson, smaller posterolateral spine acute;
fourth and fifth somites each with strong, posteri-
orly directed spine on sternite. Telson (Fig. 5e)
elongate subrectangular, length about 2.8 ante-
rior width, 5.2 posterior width, and about 1.7
length of sixth somite, not including posterior
spines; armed with 5-8 dorsolateral spines of

272

WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS

Figure 5 — Alvinocaris stactophila, holotype 6: a cephalothorax and anterior appendages, lateral; 6 rostrum, eye, antennular
peduncle, antennal scale, dorsal of left side; c antennular peduncle, distal articles, mesial; d abdomen, lateral; e tail
fan. Allotype 9: cheliped, /"mesial, g lateral; h, i,j, k pereopods 2, 3, 4, 5. Scale = 2 mm.

273

FISHERY BULLETIN: VOL. 86, NO. 2

nearly uniform size, occasionally unequal in
number on either side; posterior margin convex,
armed with 2 principal spines at each corner and
8-12 feathered strong setae on distal margin be-
tween.

Eyes (Fig. 5a, 6) with cornea imperfectly devel-
oped; unfaceted though diffusely pigmented in
adults, but with internal facetlike pattern in
smaller individuals; ovate in outline though fused
to each other mesially, and each with an up-
turned spine on anterodorsal surface.

Antennular peduncle (Fig. 5a, b, c) reaching
beyond end of antennal scale; basal article 1.3
length of second and about 2.2 length of third, all
measured on dorsal margin; stylocerite well sepa-
rated from peduncle, tapering to slender elongate
tip variably reaching as far as midlength of sec-
ond article; basal article with distodorsal margin
exceeded by rostral tip though extended into
strong lateral spine reaching level equal to that of
stylocerite and closely appressed to second article,
much smaller distomesial spine slightly diver-
gent; shorter second article with stronger
mesiodistal spine. Dorsolateral flagellum about
twice length of carapace, thickened in basal half;
ventromesial flagellum somewhat shorter.

Antennal scale (Fig. 5a, b) about twice as long
as wide, distolateral tooth strong, falling short of
broadly rounded distal margin of blade; basal ar-
ticle with small distal spine ventrally; flagellum
slightly exceeding length of body.

Mandibles (Fig. 6a) similar, with 2-segmented
palp, incisor process broad and armed with 7 mar-
ginal teeth, slender molar process simple, diver-
gent, its narrowly rounded tip minutely setose.

First maxilla (Fig. 66) with proximal endite
asymmetrically oval-triangular, distal margin
bearing about 25 long setae; distal endite with
narrowed base but broadened distally, armed
with many short spines on mesial margin and
with scattered longer spinules submarginally and
marginally beyond either end of spine row; palp
scarcely bifurcated, with long distal spine on ob-
solescent proximomesial branch and shorter adja-
cent submarginal spine and tangled setae on dis-
tal branch.

Second maxilla (Fig. 6c) with proximal endite
represented by 2 similar lobes; distal endite sub-
triangular, expanded mesiodistally and paral-
leled laterally by narrow somewhat twisted palp;
scaphognathite with anterior lobe rectangulo-
ovate, fringed with uniformly long, silky setae on
anterior and mesial borders, uniformly shorter
setae along entire lateral margin, posterior lobe

narrow and acuminate, fringed on blunt tip and
adjacent mesial margin by strikingly long, tan-
gled strong setae preceded proximally by shorter
setae similar to those on lateral margin.

First maxilliped (Fig. 6d, e) with irregularly
fusiform endite, short palp concealed and much
exceeded in length and size by leaflike exopod,
epipod obscurely bilobed; indistinct mesial lobule
on exopod possibly representing incipient lash.

Second maxilliped (Fig. 6f, g ) somewhat pedi-
form but flattened, mesial margin of articles
bearing long, feathered setae, mesial surface of
terminal article densely setose, exopod barely ex-
ceeding leaflike epipod.

Third maxilliped (Fig. 6h, i, j) slender, 5-
segmented, reaching beyond antennular pedun-
cle; terminal article trigonal in cross section,
tapered distally, bearing 3 spines, transverse
tracts of dense setae along mesial surface; similar
tract of setae on carpus and another less conspic-
uous group on merus-ischium, latter with stout
distolateral spine at articulation with carpus;
exopod much reduced, subtriangular, without
lash.

First pereopods (Figs. 5f, g; 7c, d) chelate,
subequal and sexually dimorphic, at least in fully
mature individuals; fingers curved ventrally and
slightly laterad; dactyl more slender than fixed
finger, tips varying slightly in relative length;
mesial surface of each finger convex, lateral sur-
face concave; prehensile surfaces uniformly off-
set, closing without gape, each armed with row of
almost uniform teeth so closely set as to be almost
contiguous, line of sensory hairs mesial to cutting
edges, acute tip of dactyl slightly spooned by elon-
gate teeth slanted distad and curving around ex-
ternal edge. Leg exceeding third maxilliped by
length of fingers in holotype male, but shorter
than third maxilliped in other individuals. Palm
inflated in holotype male, length 1.4 greatest
height and longer than fingers; palm relatively
shorter in allotype female and other individuals
examined, 0.3 length of fingers. Carpus shorter
than palm in holotype but longer than palm in
remainder of specimens examined, bearing
oblique ventral crest ending in strong distolateral
spine and flanked mesially by patch of setae on
polygonal raised area; notch above spine
smoothly concave and opposing low ridge ending
in small rounded spine on heel of palm; shallowly
concave anteromesial margin of carpus leading
dorsally to 2 low rounded lobes. Merus somewhat
swollen in distal half, distinct from ischium but
fused to it, neither armed.

274

WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS

Figure 6. — Alvinocans stactophila, allotype 9: a mandible; b maxilla 1; c maxilla 2; maxilliped 1, d ventral, e dor-
sal; maxilliped 2, /"ventral, g dorsal; maxilliped 3, h ventral, ; dorsal, y exopod. Paratype 6: k endopod of pleopod 1; I appendix
masculina, pleopod 2. Scales: 1 id-g, k) = \ mm; 2 (a, 6) = 0.5 mm; 3 (/) = 0.3 mm; 4 (c) = 1 mm.

275

FISHERY BULLETIN: VOL 86, NO. 2

Figure 7. — Parts of Alvinocaris chelae viewed by SEM. A. stactophila: fingers of small chela, a mesial, b dorsal; fingers of large
chela showing finely toothed opposed edges near tips, c mesial, teeth flush with convex surface, d lateral, teeth marginal on spooned
tips, with points rounded on dactyl, acute on fixed finger. A. muricola: e fingers of small chela, lateral; /large chela and distal
part of carpus, lateral. Scales: 100 p.m, d; 200 ^.m, a-c; 500 |j.m, e; 1 mm, f.

276

WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS

Figure 7. — Continued. — Alvinocaris muricola : fingers of large chela showing finely toothed opposed edges neair tips, g teeth flush
with convex mesial surface, h spooned lateral surface of same, points rounded on dactyl, acute on fixed finger; close-up lateral view
of teeth and associated sensory setae, teeth of fixed finger in foreground and of dactyl in background, i near distal end of fingers, y
near midlength of fingers; k sensory seta showing 2 rows of sensillae on concave surface. A. markensis: I fingers of small chela,
mesial view of distal part. Scales: 3 jjim, k\ 20 (jim, ;; 30 \i^m,j; 50 p.m, h\ 100 M-m, /; 200 ^.m, g.

277

FISHERY BULLETIN: VOL. 86, NO. 2

Second pereopod (Figs. 5/i; 7a, b) shorter and
more slender than first, reaching to between mid-
length and end of antennal peduncle; fingers
slightly longer than palm, similar in size and
shape, opposed edges without gape, each pecti-
nate with single row of teeth in distal half di-
rected obliquely distad and increasing slightly in
size to end in noticeably stronger tooth crossing
opposite member when closed, but spineless prox-
imally; carpus slender, about 1.2 longer than
chela; merus and ischium unarmed.

Third to fifth pereopods (Fig. 5i,j,k) similar in
length and structure, third reaching beyond an-
tennal scale by about 0.3 length of propodus.
Length articles of these legs in allotype 9 , mm:

3rd

4th
5th

ischio-
merus

4.48
4.89
4.16

carpus propodus

2.30
2.18
2.24

3.20
3.39
4.22

dactyl

0.48
0.48
0.48

Each short dactyl armed with about 6 corneous
spines on flexor surface, grading from small prox-
imally to longest and strongest distally; carpus of
each with distodorsal extension projecting as a
stop along proximal part of propodal extensor sur-
face; third leg stronger, at least in merus-
ischium, than fourth and fifth; merus of third and
fourth with ventral spine at 1/3 and 2/3 length,
distal one strongest, fifth without spines; ischium
of third with 2 spines in line with those on merus,
that of fourth and fifth spineless.

Pleopods well developed, pair 1 with endopods
about half length of expods in both sexes, endopod
of male (Fig. 6j) with asymmetrical mesial exten-
sion, that of female tapering to acute tip; pair 2
with appendix masculina of male (Fig. 6^) bear-
ing distal cluster of about 9 strong straight spin-
ules extending beyond level of simple slender ap-
pendix interna.

Uropod (Fig. 5e) with rami subequal in length,
slightly exceeding distal end of telson, lateral
ramus with movable spine mesial to smaller dis-
tolateral tooth, diaeresis sinuous.

Etymology . — The name is from the Greek
"stactos", oozing out or trickling, and "philos", to
love, for association of the species with hydrocar-
bons seeping from the substrate.

Remarks. — Alvinocaris lusca and the three
new species of Alvinocaris described here exhibit

minor differences that are highlighted in the key
to species given above, but their similarities seem
far more significant; i.e., general body appear-
ance and strength of integument, shape of ros-
trum (although that of A. stactophila sometimes
lacks ventral teeth), shape and general armature
of tail fan, blindness, and general structure of
appendages, including mouthparts. Some minor
differences that may be mentioned are features
such as number of incisor teeth on the mandible,
number of spines on the first maxilla, shape of the
second maxilla, lack of meral spines on pereopod
5 in A. stactophila, unequal distribution of spines
on ischia of pereopods 3-5 in the three species,
and shape of the endopod of male pleopod 1 and
appendix masculina (though males of A. marken-
sis are not yet known).

Each of these species lives in a distinctive ben-
thic environment, but all share similarities that
suggest dependence on a chemotrophic bacteria-
based food chain (Childress et al. 1986). Van
Dover et al. (in press) provide evidence from mor-
phology, behavioral and gut content analyses of
the similar Rimicaris exoculata Williams and
Rona that indicates a bacterial diet grazed from
surfaces of hydrothermal chimneys, although di-
rect observations of bacteria within the digestive
tract could not confirm this. The distinctively
spoon-shaped chelae of the first pereopods of both
Alvinocaris and Rimicaris species, with unbroken
comb of exceedingly fine teeth on the prehensile
edges, could be an adaptation for scooping or
sweeping bacteria toward the mouthparts.
Williams and Chace (1982) described the first
chelae of A. lusca as convex on the extensor sur-
face and concave on the flexor surface, but they
also said (p. 142) that the outer surface of the
fingers is convex and the inner surface concave.
The latter is misleading because in full extension
the convex side of the chela is mesial and the
concave side lateral. It is not yet known how these
appendages are used, but certainly the chelae can
be folded compactly against the leg's proximal
articles, and in the related Rimicaris exoculata
and R. chacei (Williams and Rona 1986) these
legs seem very mobile. Sensillae flanking prehen-
sile surfaces of the fingers seem well adapted to
aid feeding on finely particulate matter. More-
over, the species of Rimicaris have exceedingly
setose mouthparts.

In species of both genera, the second pair of
pereopods have much smaller chelae with fingers
bearing long sensory setae and spines on the pre-
hensile edges that are seemingly adapted for

278

WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS

grasping. For mobile animals of this morphologi-
cal makeup, the most likely feeding methods in
the stated environments would seem to be bacte-
rial concentration, along with secondary preda-
tion and scavenging.

ANOMURA: GALATHEIDAE
Munidopsis alvisca new species

Figure 8

Materm/. —USNM 234294, 9 Holotype,
USNM 234301, 9 Paratype, Guaymas Basin,
Golfo de Cahfornia, 27°00'N, lir25'W, 2,008 m,
Aluin Dive 1616, 8 August 1985, pilots J. Hardi-
man and R. Wilkes, observer J. F. Grassle. From
J. F. Grassle, Woods Hole Oceanographic Institu-
tion, Woods Hole, MA.

USNM 234295, 6 Paratype, Explorer Ridge,
Magic Mountain, 49°45.6'N, 130°16.16'W, 1,818
m, Pisces IV Dive P-1494, Coll. No. 1877, Gulati
Gusher-base, 1 July 1984, pilots-observers. Wit-
combe, Johnson, Tunnicliffe. USNM 234296, 9
ovig. Paratype, Explorer Ridge, Upper Magic
Mountain, 49°45.5'N, 130°16.1'W, 1,812 m,
Pisces IV Dive P-1497, Coll. No. 1873, Lunch
Hour Vent, 4 July 1984, pilots-observers. Shep-
herd, Juniper, Johnson. USNM 234297, 9 ovig.
Paratype, same. Coll. No. 1875, Crab Vent.
USNM 234298, 9 ovig. Paratype, same. Coll. No.
1875. USNM 234299, 2 9 ovig. Paratypes, Juan
de Fuca Ridge, Limbo Vent ( = 3 m from Holland's
Hillock Axial Seamount), 45°55'N, 130°03'W,
1,545 m, Pisces IV Dive P-1732, Coll. No. 1934, 2
August 1986, pilots-observers, K. Shepherd,
R. Embley, J. Franklin. From Verena Tunnicliffe,
Biology Department, University of Victoria, B.C.,
Canada.

Measurements in mm. — Holotype 9, carapace
length including rostrum 23.7, margin of orbit to
posterior edge of carapace 18.6, maximum cara-
pace width 15.7; Paratype 9 234301, same, 27.9,
20.8, 17.3; Paratype 6 234295, same, 13.8, 10.2,
8.4.

Description. — Carapace (Fig. 8a, c) exclusive
of rostrum distinctly longer than broad, moder-
ately arched transversely; anterior and posterior
cervical grooves apparent, depression in anterior
part of cardiac region; short moderately devel-
oped rugosities on each anterior branchial region,
but more distinct and transversely developed

rugae on each posterior branchial region, with
tendency to being continuous across anterior and
central part of cardiac region; posterior margin
with median concavity. Rostrum narrowly tri-
angular, concave dorsal surface smoothly curv-
ing to upturned tip exceeding eyestalks by more
than twice their length, distinct carina bearing
almost imperceptible scalelike rugae diminish-
ing to obsolescence on gastric region. Frontal
margin with broad angle lateral to eyestalk
followed by concave raised and sparsely orna-
mented margin ending in antennal spine followed
in turn by almost rectangular but acute antero-
lateral angle. Lateral plate obliquely rugose, pro-
jecting anteriorly below antennal peduncle, its
rather angular tip minutely but bluntly bi-
spinose.

Abdomen (Fig. 86) unarmed; transverse ridge
of segments 2 and 3 smooth, that of segment 4
obsolescent; segments 5 and 6 smooth.

Eyes (Fig. 8a , c ) moderate in size; well exposed,
smoothly ovate cornea cupped within broad based
movable ocular peduncle extended into elongate
mesiodorsal spine, directed obliquely upward at
low angle and ornamented with tiny irregular
obsolescent spinules, and much shorter mesioven-
tral spine.

Basal article of antennular peduncle with dis-
tal margin irregularly crenulate, slender dorso-
lateral spine and broader lateral spine flanked by
cluster of irregular small spinules, an obsolescent
mesiodorsal spine present. Antennal peduncle
with fixed basal article extended into stout, flat
ventral spine and shorter crenulate lateral spine;
succeeding articles short, second bearing stout
lateral angle, third unarmed, fourth with scal-
loped distal margin, its dorsomesial projection
spinelike.

Third maxilliped (Fig. 8e ) with ischium shorter
than merus, bearing mesial crest armed with
finely uniform, evenly spaced corneous teeth.
Basis with 2 low spines in line with crest on is-
chium. Merus with obsolescent spine at pos-
teromesial corner, mesial margin usually with
another at level of propodo-carpal joint, followed
after an interval by an obscure tubercle, and then
by a more prominent spine at base of convex dis-
tal margin; stronger spine at anterolateral cor-
ner; lateral margin broadly arched. Carpus,
propodus, and dactyl folded on merus-ischium
and about as long as those two articles together,
dense setation on dorsal surface of each, and dis-
tally on prehensile surface of propodus and
dactyl. Sternite (Fig. 8d) at base of third maxil-

279

liped with convex crenulate anterior margin on
mesial lobe, lateral lobe angular.

Epipods absent from pereopods.

Chelipeds (Fig. SP subequal, ornamented with
variably ciliate rugosities tending to arrange-
ment in longitudinal tracts; ischium with mesial

FISHERY BULLETIN: VOL. 86, NO. 2

fidge bearing subterminal spine and obsolescent
irregular subsidiary spines; merus rough, bear-
ing 3 mesial spines, 1 distodorsal spine, and a
smaller distoventral spine; carpus with 2 spines
in dorsolateral row paralleled by less prominent
ventrolateral row; palm with spines on prominent

Figure 8. — Munidopsis alvisca , holotype 9 : a carapace, eyes and right antenna, dorsal; b abdomen, somites 2-4 in folded
position; c part of cephalothorax and anterior appendages, lateral; d stemites at base of third maxilliped and chelipeds; e
left third maxilliped, merus and ischium; g left second pereopod. Paratype 9 234301: /"right cheliped. Scales: 1 (a, 6) = 5
mm; 2 (c) = 5 mm; 3 if, g) = 3 mm; 4 (d, e) = 1 mm.

280

WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS

dorsal ridge, stronger on right than on left; fin-
gers longer than palm, spooned at tips, prehensile
edges close fitting, entire, but small basal tooth of
fixed finger opposed by notch in prehensile edge
of dactyl.

Walking legs rather long, first walking leg
(Fig. 8^) reaching almost to tip of chela, second
and third reaching about to base of dactyl on pre-
ceding leg; corresponding articles of respective
legs approximately equal in length except for
meri which decrease posteriorly; each merus with
rounded, rugose dorsal crest ending in distal
spine; each carpus with longitudinal dorsal and
dorsolateral rib ending in more or less well-
developed spine, and often with secondary
spine(s) on distal margin between them; each
propodus slender, bearing small movable spine
distolaterally at base of dactyl; each dactyl slen-
der, acute corneous tip preceded by row of 12 or
more movable spines on prehensile edge. Slender
fifth leg with well-developed cleaning brush on
more or less flattened dactyl opposed by similar
setae on distal end of propodus.

Variation . — There is minor variation in orna-
mentation of the specimens available for study,
but none of it is associated with the disjunct dis-
tribution in the Golfo de California and the north-
eastern Pacific.

Remarks. — The specimens reported here were
taken around hydrothermal vent sites discussed
by Canadian American Seamount Expedition
(1985), ASHES Expedition (1986), and Tunni-
cliffe et al. (1985, 1986). Munidopsis has been
sighted at three other sites along Juan de Fuca
Ridge, but the only specimens collected are those
listed above (V. Tunnicliffe^).

Comparisons of Munidopsis alvisca with previ-
ously described species of the genus are aided by
reference to A. Milne Edwards (1880), Milne Ed-
wards and Bouvier (1897), Chace (1942), Sivert-
sen and Holthuis (1956), and Ambler (1980). Lack
of epipods on the pereopods immediately sepa-
rates M. alvisca from species such as M. crassa
Smith, 1885 and M. subsquamosa Henderson,
1885 which it superficially resembles. Both of the
latter species have relatively prominent rugae
and spines on the cephalothorax and legs whereas
M. alvisca has fairly smooth ornamentation on

2Verena Tunnicliffe, Department of Biology, University of
Victoria, P.O. Box 1700, Victor, B.C., Canada V8W 2R2, pers.
commun. 1987.

these body parts, except for minor development of
spines on the lateral carapace margin anteriorly.
The rostrum of all of these species is narrowly
triangular, curves moderately upward to the tip
and bears a middorsal carina, but the carina in
M. alvisca bears almost imperceptible scalelike
rugae and diminishes to obsolescence on the gas-
tric region whereas in both M. crassa and M. sub-
squamosa the carina is varyingly rugose, rather
strongly so in the former, and maintains this or-
namentation onto the gastric region. Moreover,
M. crassa bears tiny irregular marginal spines on
the rostrum.

Spination of the merus of the third maxilliped
is far weaker in M. alvisca than in the other two
species discussed, and both the anterolateral
spine of the ischium and the crenulate margin of
the crest on the ischium are less developed than
in them. On the other hand, M. alvisca possesses
both mesiodorsal and mesioventral eye spines
whereas M. subsquamosa and M. crassa lack the
mesioventral spine.

More distant comparisons seem inappropriate
because of different body proportions and orna-
mentation, rostral width, length, elevation and
spination, and structure of the eye and third max-
illiped. The keys for identification by both Chace
(1942) and Pequegnat and Pequegnat (1970), for
example, though strictly applicable to species of
the Atlantic basin, would ally M. alvisca with
M. aries (A. Milne Edwards, 1880), a much larger
species with broader cephalothorax and rostrum,
eyes almost hidden from dorsal view, and with
less transverse ornamentation. The revised ver-
sion of this key by Pequegnat and Pequegnat
(1971) would place M. alvisca in the couplet space
occupied by M. sundi Sivertsen and Holthuis,
1956, a species with superficially similar shaped
cephalothorax, but densely clothed with short
setae.

Etymology. — The name is an acronym taken
from names of the deep submersible vessels used
in collecting the species, Alvin and Pisces IV.

BRACHYURA: BYTHOGRAEIDAE

Bythograea tnesatlantica new
species

Figures 9, 10

Materia/.— USNM 234300, Holotype 9, Mid-
Atlantic Rift Valley about 70 km south of Kane

281

FISHERY BULLETIN; VOL. 86, NO. 2

Fracture Zone (see: Kong et al. [19851; Leg 106
Shipboard Scientific Party 1 19861; Ocean Drilling
Program Leg 106 Scientific Party [1986]),
23°22.09'N, 44°57.12'W, 3,437 m, Aluin Dive
1683, MARK vent, Stn. 1, scoop, 30 May 1986,
pilot D. Foster, observers S. Humphris and J. Ed-
mond. From NSF-Leg 106-Ocean Drilling Pro-
gram, NSF Grant OEC-8311201 to J. F. Grassle,
Woods Hole Oceanographic Institution, Woods
Hole, MA.

Measurements in mm. —

Carapace

Length

13.8

Width

23.3

Depth of cephalothorax

8.1

Frontoorbital width

7.7

Propodus lower margin

R 15.5

L 15.2

Dactyl length

P':ilm

8.3

1 aini
Height

7.9

7.8

Thickness

4.9

5.1

Description. — General aspect similar to that of
B. thermydron, cancroid, depressed. Carapace

(Figs. 9, lOd) broad, transversely elliptical, its
rounded lateral angles displaced somewhat ante-
riorly; almost flat in middle dorsally, very
slightly arched from anterior to posterior and
near lateral margins; anterolateral region pro-
duced, margin not toothed; surface finely granu-
late anteriorly and laterally, smooth but
minutely punctate to unaided eye over posterior
2/3 to 3/4; regions indistinct. Frontoorbital width
ca 1/3 carapace width.

Front almost evenly rounded and somewhat de-
flexed, projecting over folded antennules, shallow
median depression continued onto protogastric
region giving faint suggestion of bilobation; mar-
gin beaded with line of fairly uniform granules,
closely packed on anterior and anterolateral parts
but diminishing almost to obsolescence near or-
bits. Arcuate tract of scattered punctations
sweeping across anterolateral, hepatic, orbital,
protogastric, and metagastric regions. Trans-
verse tract of rather prominent granules at rear
edge of protogastric region. Carapace with
smooth part behind these anterior areas micro-
scopically granular and punctate anteriorly,
grading posteriorly into almost featureless sur-

FlGURE 9. — Bythograea mesatlantica , holotype 9: dorsal. Scale = 3 mm.

282

WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS

Figure 10. — Bythograea mesatlantica . holotype 9: right chela, a frontal view, b fingers viewed from tips; c left chela, frontal
view; d left side of cephalothorax in frontal view showing anterolateral pigmented spot, eye, antennules, antennae, and mouthparts
in situ; e mouth field showing third maxilliped turned to side, second maxilliped, first maxilliped with lacinia bearing tiny
"portunid lobe" at its mesial comer, partly hidden mandibles, and palate with patch of fine setae to either side of midline; /"abdomen
showing somites 3-6 and telson; g oviducal openings and parts of associated stemites. Scales: 1 (a-d, f-g) = 2 mm; 2 (e) = 1 mm.

283

FISHERY BULLETIN; VOL. 86, NO. 2

face. Protogastric, mesogastric, metagastric, and
cardiac regions poorly indicated; epibranchial
line indicated by small, light colored blotches
originating posterior to lateral angle at each side
and arching anteromesially over branchial re-
gion, then posteromesially toward mesogastric
region. Posterior margin concave and paralleled
by obsolescent postmarginal groove becoming
more pronounced along posterolateral margin.

Subhepatic and subbranchial areas orna-
mented with small granules, coarsest along upper
part of hepatic region but becoming finer and
more numerous near base of chelipeds.

Orbits sunk into essentially smooth transverse
concavity in anterolateral region confluent later-
ally at either side with a prominent irregularly
oval tan colored spot having very finely granu-
late, shallowly concave surface; somewhat in-
flated and irregularly granular suborbital area
almost fully visible in dorsal view, reaching level
of front, tilted anteroventrally from frontal plane
lateral to and almost at same level as epistome.

Eyestalks projecting anterolaterad, barely
movable, depressed and broadened to fit snugly in
orbit; unpigmented cornea terminal, subcircular,
narrower than eyestalk and anterolaterally ori-
ented.

Epistome (Fig. lOe) projecting well beyond
front, its anterior margin cut into 6 unequal
lobes; rather narrow and advanced submedian
lobes, separated by narrow deep notch, much
broader intermediate lobes and somewhat less
broadened lateral lobes less advanced.

Antennules folding transversely, stouter than
antennae, large bulbous basal articles contigu-
ous, concealed beneath front, interantennular
septum represented by minute remnant at upper
and lower edge of antennular fossa; slender
penultimate and terminal articles of peduncle
nearly equal in length, former slightly hollowed
laterally, latter slightly longer and more slender.
Flagella short; mesial 7-segmented ramus slen-
der; slightly shorter lateral ramus curved, multi-
segmented, thick at base but tapering to point,
dense mesial brush of long sensory setae in chord
of curve.

Antennal insertion mesial to eyestalk; pedun-
cle mesial to eyestalk, extending anteriorly or
anterolaterally in situ; fixed article broad but
short; first free article slender, ca 1.7 length of
second article; latter broadened distally; terminal
article short, its diameter only slightly greater
than that of flagellar base; flagellar length ex-
ceeding midline of front.

Mouth field (Fig. lOe) divergent anteriorly,
sides of its frame broadest posteriorly and some-
what swollen and granular at anterolateral cor-
ners, maximal inside anterior width about 1.4
minimal inside posterior width. Third maxil-
lipeds filling mouth field except for narrow gap of
nearly uniform width between ischia of en-
dognaths and rather irregular gap anteriorly be-
tween meri-carpi of endognaths and epistome; ex-
ognaths overlapping sides of mouth frame.
Endognaths with exposed surface bearing sparse,
sometimes linear, setose punctations; exposed
surface of ischium nearly smooth; elongate polyg-
onal in outline but primarily rectangular, great-
est (distal) width 1.1 narrowed part ca 1/2 length
from base; mesial margin straight through
most of its length but curved at each end, tooth-
less, bearing many stifl" straight setae, submar-
ginal zone somewhat thickened and flanked later-
ally by shallow longitudinal groove; anterior
margin nearly perpendicular to mesial margin
except for anteriorly projecting truncate lobe at
inner corner; lateral margin concave; posterome-
sial margin obliquely convex; basi-ischial suture
line visible posterolaterally. Merus slightly
narrower than ischium; low granules with tips
directed anteromesially along distal margin; ir-
regularly quadrate perimeter flanked by submar-
ginal thickened zone and groove similar to mesial
counterpart on ischium except on straight proxi-
mal side, anterolateral angle broadly rounded,
anteromesial angle at insertion of palp oblique;
mesial margin doubled anteriorly for reception of
folded palp, its ventral (exposed) side broadly
angled proximal to carpopropodal articulation;
posteromesial corner fitted to projecting lobe of
ischium, dorsal (hidden) side produced behind
carpus, its margin setose. Palp large, dactyl
reaching posteriorly about 1/4 length mesial mar-
gin of ischium. Carpus expanded distally, nar-
rowed proximally, bent nearly at right angle near
insertion and obscurely crimped inside angle;
dense tuft of setae on distooral surface. Propodus
wider than carpus, longer than broad, asymmetri-
cally ovate in ventral view; distal (longest) mar-
gin convex, densely beset with rows of strong ser-
rated setae, longest distally; distal tuft of such
setae on dorsal surface. Linear dactyl slightly
bent away from midline in distal part and setose
as propodus, especially on prehensile edge. Ex-
ognath narrow, not extending to full length of
merus; ventral surface slightly curved mesially to
fit closely against lateral side of endognathal is-
chium, with dorsomesial flange (widest distally)

284

WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS

fitting beneath latter; palp conspicuous, flagel-
lum densely beset with setae in hollow of curve.

First maxilliped with lacinia of endopod broad,
its distal edge 3-lobed and conspicuously though
not heavily setose; oblique mesial margin of
strongly advanced anterolateral lobe confluent
with broader gradually rounded and much less
advanced intermediate lobe, latter in turn fol-
lowed by still less advanced tiny mesial lobe, sep-
arated by a notch and directed anteromesially;
tuft of setae preceding notch.

Endostome large, divided by low median sagit-
tal ridge bifurcated somewhat anteriorly and
merging into projecting endostome; each half of
palate shallowly concave, crossed by low longitu-
dinal ridge slightly offset at its midlength and
trending anteromesially from near base of large
mandibular palp; ridge flanked laterally by irreg-
ular patch of velvety pubescence; smooth lateral
2/3 of palate receiving large efferent branchial
channels.

Chelipeds (Figs. 9, 10a, b, c) heavy, subequal;
integument punctuate on upper and extensor sur-
faces, obscure granulation on upper surfaces of
palms and on ridges or raised areas elsewhere;
chelae inflated, lower margin of palm arched
downward, its rather pronounced keel merging
into fixed finger; swollen palm with shallow exca-
vation proximally for reception of carpus in flexed
position, inner surface glabrous but drawn into
moderate and slightly granular elevation slightly
in front of proximal excavation. Fingers tan col-
ored in preservation (70% ethanol) and darkest
proximally, color of fixed finger not extending
onto palm; fingers not gaping, prehensile edges
entire except for obsolescent proximal tooth on
fixed finger of each hand; dactyl longer than rela-
tively straight fixed finger, arching down distally
to close in distal notch of spooned tip of fixed fin-
ger.

Carpus with extensor surface inflated, right
carpus with internal margin rounded, that of left
obscurely angled. Merus broadened mesially into
cristate flange angled distally for reception of car-
pus, strong granules in single line along inner
margin, outer surface rounded, strewn with obso-
lescent punctations and granules, latter most
prominent along distoventral tract.

Walking legs rather long, flattened, length de-
creasing posteriorly in order 3, 2, 1, 4; each with
dense patches of short darkened setae inter-
spersed with sparer longer setae on extensor sur-
face of carpus and propodus (as well as its lateral
side on legs 1 and 2), distoventral corner of car-

pus, and more extensively on dactyl; fifth legs
somewhat more flattened than others, propodi
relatively broader and not densely setose later-
ally. Mean maximum length of propodi about
twice width. Dactyls slightly longer than propodi,
narrowly lanceolate, shallow longitudinal
grooves on anterior and posterior surfaces ob-
scured by dense setae, tip stout, corneous. Merus
of each with upper margin finely granular, ante-
rior lower margin present throughout length but
posterior lower margin obsolescent proximally.

Sternum broadest between legs 1 and 2, nar-
rower posteriorly, glabrous beyond outline of ab-
domen.

Abdomen (Fig. lOf) ovate in outline, fully seg-
mented and densely fringed with plumose setae;
somite 1 slightly arched dorsally to fit contour of
adjacent carapace, somites 2-4 of about equal
length, somites 5 and 6 progressively longer; ab-
domen with greatest width at 4; telson nearly as
broad as somite 6, outline broadly arched distally.
Somites 2-5 bearing large, well-developed bi-
ramous pleopods, outer curved branch lying near
edge of abdomen and heavily beset with short
setae laterally and mesially, inner branch more
sparsely equipped with ovigerous setae and
jointed.

Female openings (Fig. 10^^) large, obscurely
subtriangular in outline.

Color in preservation predominantly off-white
except for fingers, matted setal tracts laden with
brownish finely particulate matter.

Remarks. — Brachyuran crabs that resemble
Bythograea were observed and reported by Rona
et al. (1986).

Bythograea mesatlantica differs in several re-
spects from Pacific members of the genus,
B. thermydron Williams (1980) and B. microps de
Saint Laurent (1984). Among obvious differences
from B. thermydron, the new species has even
less ornamentation on the carapace; it lacks a
distinct suborbital plate separated by a suture,
and the suborbital area is inflated, not flat and
inclined; there is a transverse concavity lateral to
each eyestalk that terminates near the very dis-
tinctive brown spot in the cuticle at either side of
the carapace; the eyestalk itself is shorter and
thicker than in B. thermydron and the shape and
position of the cornea differs. The ischium of the
third maxilliped is relatively shorter than in
B. thermydron and bears only sparse setiferous
punctations on the external surface, it lacks tiny
granules on the truncate lobe at the anterolateral

285

FISHERY BULLETIN: VOL. 86, NO. 2

corner, the submarginal thickened zone and
groove are less distinct; the merus is not tilted
dorsally in normal position, and the palp is rela-
tively shorter and club shaped rather than curved
like a knife edge along the prehensile edge. The
lacinia of the first maxilliped is more angular
anterolaterally and has a smaller "portunid lobe".
The epistome is less lobulate than in B. thermy-
dron and the concave palatal area has much less
setose covering. The nearly toothless chelae have
brown fingers and there is no dense patch of setae
on the inner side of the palms.

Comparisons with B. microps are necessarily
less complete because of the brief description of
the latter. The eyes are certainly not slender and
retracted in B. mesatlantica ; the chelipeds are not
noticeably dimorphic, and they are relatively
smooth rather than strongly granular and pilose
on the external surface as in B. microps.

The distinctive exocular spots on the carapace
seem similar to those noted on the chelipeds of
Hypsophrys noar Williams (1974, 1976) and Mu-
nidopsis lentigo (Williams and Van Dover 1983).
Their function is unknown.

Etymology. — From the Greek "mesos", middle
and "Atlantic", with reference to the Mid-
Atlantic Rift habitat.

ACKNOWLEDGMENTS

Contributors of specimens and crew members of
submersible vessels who helped to make the col-
lections are owed a special debt of gratitude for
securing the rare material described here. Donors
are acknowledged individually in each of the spe-
cies accounts. I am indebted to Keiko Hiratsuka
Moore for rendering the excellent illustrations, to
Ruth Gibbons for helping to produce the SEM
micrographs, and to F. A. Chace, Jr., R. B. Man-
ning, and B. B. Collette for critical review of the
manuscript.

LITERATURE CITED

Ambler, J. W.

1980. Species of Munidopsis (Crustacea, Galatheidae)

occurring off Oregon and in adjacent waters. Fish.

Bull., U.S. 78:13-34.
ASHES Expedition

1986. Pisces submersible exploration of a high-
temperature vent field in the caldera of Axial Volcano,
Juan de Fuca Ridge. Eos 67(44):1027.

Brooks, J M., M C Kennicutt II, C. R Fisher, S. A Macko,
K. Cole, J. J. Childress, R. R. Bidigare, and R. D. Vetter.

1987. Deep-sea hydrocarbon seep communities: Evidence
for energy and nutritional carbon sources. Science

238:1138-1142.
Canadian American Seamount Expedition

1985. Hydrothermal vents on an axis seamount of the
Juan de Fuca Ridge. Nature 313:212-214.

Chace, F A . Jr

1942. Reports on the scientific results of the Atlantis
expeditions to the West Indies, under the joint auspices of
the University of Havana and Harvard University. The
Anomuran Crustacea. I. Galatheidae. Torreia, Havana,
11:1-106.
Childress, J. J , C R Fisher, J M Brooks, M C. Kennicutt II,
R Bidigare, and A E Anderson

1986. Methanotrophic marine molluscan (Bivalvia,
Mytilidae) symbiosis: mussels fueled by gas. Science
233:1306-1308.

DE Saint Laurent, M.

1984. Crustaces Decapodes d'un site hydrothermal actif
de la dorsale du Pacifique oriental (13° Nord), en
provenance de la campagne frangaise Biocyatherm.
C. R. Hebd. Stances Acad. Sci. Ser. 3 Sci. Nat. (Paris)
299(9):355-360, plate 1.

KoNG. L., W B. F. Ryan, L. A. Mayer, R. S. Detrick, P J. Fox, and
K Manchester.

1985. Bare-rock drill sites, O.D.P. Legs 106 and 109;
evidence for hydrothermal activity at 23°N on the
Mid-Atlantic Ridge. EOS 66(46):936.

Leg 106 Shipboard Scientific Party (R. S. Detrick, J.
Honnorez, A. C. Adamson et al.).

1986. Mid-Atlantic bare-rock drilling and hydrothermal
vents. Nature 321:14-15.

Milne Edwards, A

1880. Reports on the results of dredging, under the
supervision of Alexander Agassiz, in the Gulf of Mexico
and in the Caribbean Sea, 1877, '78, '79, by the United
States Coast Survey Steamer "Blake," ....
VIII. — Etudes pr6liminaires sur les Crustac6s. Bull.
Mus. Comp. Zool. Harvard 8:1-68, 2 plates.

Milne Edwards, A., and E. L. Bouvier.

1897. Reports on the results of dredging, under the
supervision of Alexander Agassiz, in the Gulf of Mexico
(1877-78), in the Caribbean Sea (1878-79), and along
the Atlantic coast of the United States (1880), by the U.S.
Coast Survey Steamer "Blake" .... XXXV.
Description des Crustac6s de la famille des Galatheides
recuellis pendant I'exp^dition. Mem. Mus. Comp. Zool.
Harvard 19:1-141.

Ocean Drilling Program Leg 106 Scientific Party.

1986. Drilling the Snake Pit hydrothermal sulfide deposit
on the Mid-Atlantic Ridge, lat 23°22'N. Geology
14(12):1004-1007.

Pequegnat, L H., and W. E. Pequegnat.

1970. Deep-sea anomurans of Superfamily Galatheoidea
with descriptions of two new species. In W. E.
Pequegnat and F. A. Chace (editors). Contributions on
the biology of the Gulf of Mexico. Tex. A&M Univ.
Oceanogr. Stud. 1(5):125-170.

Pequegnat, W. E , and L H. Pequegnat.

1971. New species and new records of Munidopsis
(Decapoda: Galatheidae) from the Gulf of Mexico and
Caribbean Sea. Contributions on the biology of the Gulf
of Mexico. Tex. A&M Univ. Oceanogr. Stud. 1
(Suppl.):l-24.

RoNA, P. A., G. Klinkhammer, T. a. Nelson. J. H. Tefrey, and
H Elderfield.

1986. Black smokers, massive sulphides and vent biota at
the Mid-Atlantic Ridge. Nature 321:33-37.

286

WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS

Saint Laurent, M., de. See De Saint Laurent.
SiVERTSEN, E., AND L. B HOLTHUIS

1956. Crustacea Decapoda (The Penaeidea and

Stenopodidea excepted). Rep. Sci. Results "Michael

Sars" North Atl. Deep-Sea Exped. 1910, 5(12):l-54,

plates 1-4.

tunnicliffe. v , m botros, m e de burgh, a dinet, h p.

Johnson. S. K Juniper, and R E McDuff

1986. Hydrothermal vents of Explorer Ridge, northeast
Pacific. Deep-Sea Res. 33:401-412.
TUNNICLIFFE, V , S K JUNIPER, AND M E DE BURGH

1985. The hydrothermal vent communities on Axial
Seamount, Juan de Fuca Ridge. Bull. Biol. Soc. Wash.
6:453-464.
Van Dover, C. L., B Fry. J F Grassle, S Humphris, and P. A.

RONA.

In press. Feeding biology of the Mid-Atlantic Ridge
hydrothermal vent shrimp: Functional morphology, gut
content analyses, and stable isotopic composition. Mar.
Biol. (Berl.).
Williams, A B

1974. A new species of Hypsophrys (Decapoda:

Homolidae) from the Straits of Florida, with notes on

related crabs. Proc. Biol. Soc. Wash. 87:485-492.
1976. Integumental organs of unknown function on

chelipeds of deep-sea crabs, genus Hypsophrys. J.

Morphol. 150:889-900.
1980. A new crab family from the vicinity of submarine

thermal vents on the Galapagos Rift (Crustacea:

Decapoda: Brachyura). Proc. Biol. Soc. Wash.

93:443-472.
Williams, A B., and F A. Chace, Jr.

1982. A new caridean shrimp of the family Bresiliidae
from thermal vents of the Galapagos Rift. J. Crust. Biol.
2:136-147.

Williams, A B , and P A. Rona.

1986. Two new caridean shrimps (Bresiliidae) from a

hydrothermal field on the Mid-Atlantic Ridge. J. Crust.

Biol. 6:446-462.
Williams, A. B , and C. L. Van Dover.

1983. A new species of Munidopsis from submarine
thermal vents of the East Pacific Rise at 21°N
(Anomura:Galatheidae). Proc. Biol. Soc. Wash.
96:481-488.

287

THE MEGALOPA STAGE OF THE GULF STONE CRAB, MENIPPE ADINA
WILLIAMS AND FELDER, 1986, WITH A COMPARISON OF MEGALOPAE

IN THE GENUS MENIPPE

Joel W. Martin,' Frank M. Truesdale,^ and Darryl L. Felder^

ABSTRACT

The laboratory-reared megalopa stage of the Gulf stone crab, Menippe adina , is described and illus-
trated and compared with megalopae of three other species of Menippe . The megalopa of M . adina
differs from that of Af. nodifrons in having serrate spines on the ventral margin of the dactylus of
pereiopod 5 and from that of M. rumphii in having spines on the dactyli of pereiopods 2-5 and a more
quadrate carapace. The megalopa of the morphologically similar A/ . mercenaria was also reared in the
laboratory, and selected characters are described and compared with the megalopa of A/, adina;
megalopae of the two species differ only slightly. Megalopae of M. adina taken from field collections
made off South Texas, U.S.A., were compared with and were found to be consistent with laboratory-
reared M . adina megalopae.

Stone crabs of the genus Menippe are large xan-
thid crabs common along the eastern coasts of the
United States and Mexico from North Carolina to
Yucatan, the Bahamas, Cuba, and Jamaica
(Rathbun 1930; Felder 1973; Williams 1984;
Williams and Felder 1986). Recently the
"common" stone crab, Menippe mercenaria (Say,
1818), was divided into two species: Menippe mer-
cenaria (Say) (restricted), known from the east
coast of the United States, the Caribbean, and the
west coasts of Florida and Yucatan, and Menippe
adina Williams and Felder, 1986, known from the
northwestern Gulf of Mexico; hybridization of the
two species occurs in northwest Florida (see
Williams and Felder 1986). These two species
(primarily M. mercenaria ) support an important
stone crab fishery in the southern United States
and Mexico (Williams and Felder 1986) and con-
sequently have been the subject of numerous in-
vestigations. Despite this interest, the complete
larval developments of both commercial species of
Menippe remain unknown. For M. mercenaria
(Say), Hyman (1925) described a prezoea and first
zoeal stage, and Porter (1960) described six zoeal
stages reared in the laboratory. Unfortunately,
Porter did not describe the megalopa stage, pre-

iLife Sciences Division, Natural History Museum of Los An-
geles County, 900 Exposition Boulevard, Los Angeles, CA
90007.

2School of Forestry, Wildlife, and Fisheries, and Louisiana
Agricultural Experiment Station, Louisiana State University,
Baton Rouge, LA 70803.

3Department of Biology and Center for Crustacean Research,
University of Southwestern Louisiana, Lafayette, LA 70504.

sumably because he considered it a postlarva and
not a true larval stage. An unpublished but often-
cited report by Kurata"* included descriptions of
the zoeal stages of M. mercenaria and a brief
sketch of the megalopa; Kurata's description of
the megalopa did not include morphology of the
pleopods, pereiopods, or mouthparts.

Because of recent interest in the phylogenetic
significance of the brachyuran megalopa (see Rice
1981a, in press; Martin in press) and postlarval
stages (Martin et al. 1984; Felder et al. 1985), and
because of the potential importance of stone crab
larval biology to aquaculture, it is surprising that
the megalopae of M . mercenaria and M . adina
remain undescribed. The present paper describes
the laboratory-reared megalopa of the Gulf stone
crab, Menippe adina Williams and Felder, and
compares it with field collections of the same spe-
cies from south Texas, laboratory-reared megalo-
pae of M. mercenaria, and all previously de-
scribed megalopae of the genus Menippe: Menippe
mercenaria (Say, 1818) (as described by Kurata
fn. 4); Menippe nodifrons Stimpson, 1859 (as de-
scribed by Scotto 1979); and Menippe rumphii
(Fabricious, 1798) (as described by Kakati 1977).

MATERIALS AND METHODS

A large ovigerous M . adina was collected from

Manuscript accepted February 1988.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.

4Kurata, H. 1970. Studies on the life histories of decapod
Crustacea of Georgia. Part IIL Larvae of decapod Crustacea of
Georgia. Unpubl. rep., 274 p. University of Georgia Marine
Institute, Sapelo Island, GA.

289

FISHERY BULLETIN: VOL. 86, NO. 2

shallow waters of the northern Gulf of Mexico
near Grande Terre, LA, in May 1982 and held in
a small aquarium at room temperature. After the
eggs hatched, the zoeal larvae were given fresh
seawater and newly hatched Artemia nauplii
daily. Exuviae as well as dead and some living
megalopae were preserved in 70% ethanol. Draw-
ings were made with the aid of a Wild^ M-5
stereoscope and a Wild M-11 compound stereo-
scope, both with camera lucida; accuracy was ver-
ified with a Nikon Optiphot. Measurements were
made with an ocular micrometer. Ten laboratory-
reared megalopae were examined, measured, dis-
sected, and compared with megalopae from field
collections made in 1973 off south Texas. Com-
parisons with M. mercenaria are based on
laboratory-reared M . mercenaria megalopae from
two females collected on 13 August 1987 from the
Indian River system, north of Ft. Pierce, FL. Eggs
of these two females hatched on 21 August 1987,
and the megalopa stage was first reached after 17
days in mass culture aquaria (30%o salinity,
25°C, 12h:12h light/dark regime). Descriptions of
setation for all appendages proceed from proximal
to distal. Specimens examined under the scan-
ning electron microscope (SEM) were prepared
according to procedures outlined by Felgenhauer
(1987) but without postfixation in osmium tetrox-
ide and with 100% ethanol, rather than amyl ac-
etate, as the transitional fluid. Sibling megalopae
and field collections have been deposited in the
U.S. National Museum of Natural History,
catalogue No. USNM 229962 (laboratory-
reared M. adina), USNM 229961 (field-collected
M. adina), and USNM 229963 (laboratory-reared
M. mercenaria).

RESULTS

Carapace (Figs., lA, B, C, 3 A).— Length 1.67
mm, width 1.45 mm (A'^ = 10). Subquadrate, with
2 lateral prominences on each side; dorsoven-
trally thick, with minute tubercle centrally lo-
cated. Posterior border fringed with numerous
short setae; lateral margin with few scattered
setae. Rostrum ventrally deflexed, nearly verti-
cal, with deep medial depression, rounded anteri-
orly. Angular interorbital prominences extend
ventrally between orbit and antennule. Chroma-
tophores variable in placement, but almost al-
ways found in areas indicated in Figure IB.

Eyes (Figs. lA, B, C, 3A). — Large, exposed; eye-
stalks sometimes with 2 or 3 short, simple ante-
rior setae, always with posterodorsal chroma-
tophore.

Abdomen (Fig. lA, B). — Subequal in length to
carapace. All pleura with rounded posterolateral
angles. All somites with sparse setae dorsally;
somites 2-5 always with elongated chroma-
tophores.

Telson (Fig. IG). — Broadly rounded with vari-
able setation, occasionally with pair of small pos-
terior spines (as in Figure IB).

Antennule (Fig. IK). — Biramous; peduncle 3-
segmented, with variable setation. Basal seg-
ment of peduncle large, bulbous, always with
large chromatophore; middle segment subcylin-
drical with 0-2 distal setae; distal segment ovoid
with scattered short setae. Lower ramus
1-segmented with 6-8 setae; upper ramus
5-segmented with aesthetascs arranged in tiers,
usually 0, 7, 8, 6, 4 subterminal plus 3 terminal,
with short setae sometimes present on segments
2 and 4 (note: all aesthetascs not illustrated).

Antenna (Fig. IJ). — Flagellum 12-segmented
(sometimes 11), with 3 peduncular articles and 8
or 9 flagellar articles (see Rice, in press, for cor-
rect number of antennal segments in megalopae);
setation variable, usually 2, 3, 2, 0, 0, 2, 4, 0, 4 or
5, 1, 4, 4.

Mandibles (Fig. 2F). — Asymmetrical, with
broadly rounded spade-shaped cutting edges; palp
2-segmented with setation 0, 11-14.

Maxillule (Fig. 2E). — Protopodite with 1 or 2
long plumose setae on posterodorsal margin; en-
dopodite 2-segmented with setation 1, 2 subtermi-
nal plus 2 terminal; basal endite with 29-35
spines and setae; coxal endite with 13-16 spines
and setae.

Maxilla (Fig. 2D).— Scaphognathite with 70-78
fringing setae and 0—6 setae on blade; endopodite
unsegmented with 0 or 1 distolateral seta and 4 or
5 basal plumose setae; basal endite bilobed with
setation variable, usually 8-10, 9-11; coxal en-
dite bilobed with setation usually 7, 9 or 10.

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

Maxilliped 1 (Fig. 20). — Exopodite 2-seg-
mented, with setation 2 or 3, 5-7. Endopodite

290

MARTIN ET AL.: MEGALOPAE OF STONE CRABS iMENIPPE)

A-C

D-K-

Figure l. — Megalopa of the Gulf stone crab, Menippe adina. A, entire animal, lateral view; B, same, dorsal view; C, frontal view
of rostrum and eyes; D, pleopod 1; E, pleopod 4; F, pleopod 5; G, telson and posterior part of sixth abdominal segment; H,
dactylus of pereiopod 3; I, dactylus of pereiopod 5; J, antenna; K, antennule. Both scale bars = 1.0 mm.

291

FISHERY BULLETIN: VOL. 86. NO. 2

Figure 2. — Megalopa oi Menippe adina, mouthparts. A,
third maxilliped; B, second maxilliped; C, first maxil-
liped; D, maxilla; K,- maxillule; F, mandible. Scale
bar = 0.5 mm.

292

MARTIN ET AL.: MEGALOPAE OF STONE CRABS (MENIPPE)

unsegmented with 6-8 setae arranged as shown.
Basal endite setation 28-33; coxal endite setation
15-17. Epipodite with 22 or 23 long, minutely
plumose setae, appearing simple under low mag-
nification.

Maxilliped 2 (Fig. 2B). — Exopodite 2-segmented,
with setation 3, 5-8. Endopodite 4-segmented,
with setation usually 5, 2 or 3, 5 or 6, 9 or 10;
distal segment with 4 or 5 stout serrate setae.
Epipodite with 9 or 10 long minutely plumose
setae.

Maxilliped 3 (Fig. 2A). — Exopodite 2-seg-
mented, with setation 0 or 1, 6-8. Endopodite
5-segmented, with variable setation, usually 18-
20, 15 or 16, 5-9, 6-8, 7-10; ischium with scal-
loped medial border. Epipodite with 18 long
minutely plumose setae on distal two-thirds plus
8-12 plumose setae on proximal one-third. Pro-
topodite setation variable.

Pereiopods (Figs. lA, B, H, I, 3B, C, D).— Che-
lipeds long, stout, subequal; dactylus with 4 irreg-
ular teeth; immovable finger with 3 teeth (Fig.
3B); tips of fingers overlap distally when approxi-
mated. No recurved hook on basi-ischium
(Fig. 3B). Second to fourth pereiopods similar;
dactylus with 5 (rarely 4) serrate spines ventrally
(e.g., Figs. IH, 3C, D); propodus with long ven-
trodistal spine (Fig. IH). Fifth pereiopod dactylus
(Fig. II) with 3 long pectinate setae, 1 markedly
toothed and concave (Fig. 3E), on distal ventral
border and 3 or 4 serrate spines ventrally.

Pleopods (Fig. ID, E, F). — Decreasing in size
posteriorly. Pleopod 1 (Fig. ID) with 19-22
plumose setae; endopodite with 3 or 4 hooked
setae. Pleopod 4 (Fig. IE) with 19-21 plumose
setae; endopodite with 3 or 4 hooked setae
(Fig. 3F). Pleopod 5 (uropod) (Fig. IF) with 12-14
plumose setae; basal segment lacking setae or

with 1 or 2 setae (field collections); endopodite
absent.

Color. — Overall coloration rose-orange, with
dark blue-black chromatophores located as shown
in Figure lA, B.

DISCUSSION

The genus Menippe de Haan, 1833, presently
contains about 8 species, only 3 of which occur in
North America. The megalopa stage is now known
for 3 species in the genus: M. rumphii (Fabricious,
1798), M. nodifrons Stimpson, 1859, and M. adina
Williams and Felder, 1986. In addition, selected
characters of M. mercenaria (Say, 1818) are pre-
sented here for comparison; some characters of
that species are also obtainable from an unpub-
lished report by Kurata (fn. 4) (see Table 1).

Laboratory-reared megalopae of M. adina were
virtually identical to megalopae presumed to be-
long to M. adina that were collected off south
Texas. Even meristic counts of the mouthpart se-
tation agreed exactly, with the only observed dif-
ferences being that field-collected megalopae
were slightly larger and occasionally bore 1 or 2
setae on the basal segment of the uropod. Thus,
we feel that our laboratory conditions have not
adversely affected development or introduced ab-
normal characters, and we have used these field
collections for the SEM figures of M. adina mega-
lopae (Fig. 3).

We expected to find that characters of the
megalopa of M. adina are similar to those de-
scribed by Kurata (fn. 4) for the morphologically
similar (in adulthood) M. mercenaria, a species
known to hybridize with M. adina (see Williams
and Felder 1986). In general this is true. How-
ever, some characters reported by Kurata differ
from our observations on M. adina and from our
laboratory-reared megalopae of M. mercenaria
(Fig. 4). Kurata mentioned (but did not illustrate)

Table 1 . — Comparison of characters in megalopae of the genus Menippe. Dash ( — ) indicates information not available from reference.

Size! (mm)
CL CW

Setation

Spinafion

Setation
pleopod 5

Palp of
mandible

Epipod of

Dactylus of

Menippe

Maxilliped 1

Maxilliped 2

Maxilliped 3

Pereiopods 2-4

Pereiopod 5

Reference

adina

mercenaria

nodifrons

rumphii

1.67 1.45
1.70 1.55
1.7-8 —
1.50 1.31
1.60 1.55

0, 11-14
0, 11-13

0, 10-13
0,0,9

22-23
20-23

12-20, 226
22

9-10
7-10

up to 10
8

18
18-20

18
18

4-5

4
4
0
0

12-13
11-13
11-12

Present study
Present study
Kurata3
Scotto 1979
Kakati 1977

^CL = carapace length; CW = carapace width.

2From a megalops hatched from a stage 6 (rather than the typical stage 5) zoea.

3See text footnote 4.

293

FISHERY BULLETIN: VOL. 86, NO. 2

Figure 3. — Scanning electron micrographs (SEM) of selected characters of Menippe megalopae (presumably M. adina ) collected in
south Texas. A, dorsal view of carapace (x 25); B, ventral view of chelipeds showing dentition of the fingers and lack of recurved
hook on ischium (x 37); C, dactyli of second (upper figure) and third pereiopods (x 230); D, higher magnification of ventral
dactylar spine indicated by arrow in C (x 1,900); E, endopod of third abdominal pleopod showing 4 dentate hooklike setae
(x 2,200); F, serrate setae (only 2 of 3 shown) of dactylus of pereiopod 5 (x 2,300).

294

MARTIN ET AL.: MEGALOPAE OF STONE CRABS iMENIPPE)

Figure 4. — Scanning electron micrographs (SEM) of selected characters of the laboratory-reared megalopa of Menippe merce-
naria. A, dorsal view of carapace (x 37); B, ventral view of chelipeds and third maxillipeds (x 55); C, dactylus of second
pereiopod showing ventral serrate spines (x 270); D, ventral spines on dactylus of fourth pereiopod with fewer spinules than
anterior pereiopod spines (x 2,000); E, dactylus of fifth pereiopod showing four serrate "sensory" setae with one (arrow) more
obviously serrate (x 190); F, higher magnification of concave serrate setae indicated by eutow in E (x 1,500).

295

FISHERY BULLETIN: VOL. 86, NO. 2

"about 9 small spines" on the ischium of the third
maxilliped; we found no spines on M. adina or M.
mercenaria (Figs. 3B, 4B), but it is possible that
Kurata was referring to the acute borders of the
scalloped medial margin (our Figure 2A), in
which case the 2 species are similar. Spination of
the ischium of pereiopods 1-3 differs also; Kurata
described 5 or 6, 2 or 3, and 1 small spine on the
ischia of pereiopods 1, 2, and 3, respectively,
whereas we did not notice this condition in M.
adina or M. mercenaria (see Figures 3B, 4B). Fi-
nally, Kurata (fn. 4: pi. 74, fig. E) illustrated no
spines on the ventral surface of the fifth pereiopod
dactylus of M. mercenaria; these spines are obvi-
ous on both species (Figs. II, 4E). We found few
differences between megalopae of M. adina and
M. mercenaria. General morphology of the cara-
pace and chelipeds, spination of the dactylus of
the pereiopods, and setation of the pleopods
agreed almost exactly (compare Figures 3 and 4).
Ventral dactylar spines on the posterior walking
legs of M. mercenaria were not so serrate as in M.
adina and were sometimes armed with only 2 or
3 large spinules rather than the numerous spin-
ules seen in M. adina (e.g., Fig. 3D) and in the
more anterior legs of M. mercenaria (see Figure
3C, D, 4C). Also, in all but 1 of the 9 megalopae of
M. mercenaria examined there were 4 (rather
than 3) long serrate setae on the dactylus of the
fifth pereiopod (Fig. 4E). As in M. adina, one of
these setae was more serrate and concave than
were the other long setae (Fig. 4E, F). However,
we have not examined mouthpart morphology of
M. mercenaria in the detail in which we described
M. adina, and so it is possible that additional
characters will be found to separate these 2
species at the megalopa stage.

The megalopa of M. adina is very similar to
that of M. nodifrons as described by Scotto (1979).
Although the 2 species differ in setation of some
of the mouthparts, this setation may differ from
side to side in a given individual. The salient
character that serves to separate megalopae of
these 2 species is the presence in M. adina of 4
stout serrate spines on the dactylus of pereiopod
5. Scotto (1979) figured only setae (and no spines)
on the dactylus of the fifth pereiopod in M. nod-
ifrons and the dactylar spines on other pereiopods
apparently are not serrate (Scotto 1979, fig. 9c,
pereiopod 3).

The megalopa of M. rumphii described by
Kakati (1977) differs from that of M. nodifrons,
M. mercenaria, and M. adina in having a more
ovoid carapace with the rostrum only slightly de-

flexed. Kakati did not describe the dactyli of
pereiopods 2-5 for M. rumphii, but his figure of
pereiopod 2 (1977:639, fig. 2, p. 50) does not show
stout ventral spines on the dactylus. Possibly
Kakati overlooked these spines; if not, the ab-
sence of these spines on pereiopod 2 would further
serve to separate the megalopa of M. rumphii
from those of A/, nodifrons, M. mercenaria, and M.
adina. All 4 species have been described as hav-
ing a rose-orange coloration in life.

Although Rathbun (1930) and Monod (1956)
synonymized M. rumphii with M. nodifrons, de-
scriptions of the zoeal stages of M. rumphii and M.
nodifrons by Kakati (1977) and Scotto (1979), re-
spectively, show that larvae of the 2 species differ
considerably. In the first zoeal stage, M. rumphii
exhibits elongated posterolateral processes on ab-
dominal segment 5 that extend posteriorly to
more than half the length of the telsonal furcae,
which lack spines. The first zoea of M. nodifrons
has similar posterolateral processes but these do
not extend posteriorly beyond the fork of the tel-
son; the telsonal furcae bear 1 dorsal and 2 lateral
spines each. These differences are not apparent in
later zoeal stages, but their presence in the first
zoeal stage and the differences noted in the mega-
lopa stage may be reason to question the syn-
onymy of these 2 species.

Xanthid larvae are known to be variable, and it
is often difficult to reconcile larval and adult
groupings based on morphology. Larvae of some
morphologically disparate (as adult) species are
very similar, whereas zoeal stages for species in
some genera differ markedly in their morphology
(see Martin 1984; Martin et al. 1984, 1985; Mar-
tin and Abele 1986). Because of the known mor-
phological variability of xanthid larvae, charac-
ters presented for taxonomic purposes here and
elsewhere (e.g., Martin 1984) must be used with
caution.

It is not our intent to promote descriptions of
single stages in the life cycles of brachyuran
crabs. However, in those cases where a descrip-
tion of a single stage adds appreciably to our
knowledge of phylogeny (e.g.. Rice 1981b) or fills
a gap in the larval biology of a commercially im-
portant species complex (present study), we feel
such a description is justified. A detailed compari-
son of zoeal stages of the two species is planned for
the near future.

ACKNOWLEDGMENTS

We are grateful to D. H. Wilber and A. B. This-

296

MARTIN ET AL.: MEGALOPAE OF STONE CRABS (MENIPPE)

tie for constructive criticism of the manuscript, to

B. E. Felgenhauer for help in preparing the fig-
ures, to N. N. Rabalais and B. Cole for assistance
in rearing larvae of M. mercenaria, and to L. G.
Abele for providing space and facilities for the
research. We thank S. Silvers and K. Riddle of the
Florida State University Electron Microscopy
Center for their expert assistance. We also thank

C. Dugas, Louisiana Department of Wildlife and
Fisheries, for collecting the ovigerous specimen of
M. adina. This work was supported in part by the
National Science Foundation grants No. BSR-
8414347 and BSR-8615018 to J. W. Martin and
L. G. Abele, and by a research grant to D. L.
Felder, N. N. Rabalais, F. M. Truesdale, and

D. W. Foltz from the Louisiana Education Quality
Support Fund under grant No. 86-LUM(2)-084-
09.

LITERATURE CITED

Felder, D L

1973. An annotated key to crabs and lobsters (Decapoda,
Reptantia) from coastal waters of the northwestern Gulf
of Mexico. La. State Univ. Cent. Wetland Resour., Publ.
LSU-SG-73-02. Baton Rouge, LA.
Felder, D. L., J. W. Martin, and J. W. Gov.

1985. Patterns in early postlarval development of deca-
pods. In A. M. Wenner (editor), Larval growth, p. 163-
225. Crustacean Issues, vol. 2. Balkema Press, Rotter-
dam.
Felgenhauer, B E.

1987. Techniques for preparing crustaceans for scanning
electron microscopy. J. Crustacean Biol. 7:71-76.
Hyman. O W

1925. Studies on the larvae of crabs of the family Xanthi-
dae. Proc. U.S. Natl. Mus. 67:1-22.
Kakati, V S.

1977. Larval development of the crab, Menippe rumphii
(Fabricious), as observed in the laboratory (Crustacea,
Brachyura). Proc. Symp. Warm Water Zooplankton,
UNESCO/NIO Spec. Publ., p. 634-641.
Martin. J W

1984. Notes and bibliography on the larvae of xanthid
crabs, with a key to the known xanthid zoeas of the west-
em Atlantic and Gulf of Mexico. Bull. Mar. Sci. 34:220-
239.
In press. Phylogenetic significance of the brachyuran
megalopa: evidence from the Xanthidae. In A. A. Fin-

cham and P. S. Rainbow (editors), Aspects of decapod
crustacean biology. Symp. Zool. Soc. Lond., Vol. 59.

Martin. J W.. and L G Abele

1986. Notes on male pieopod morphology in the
brachyuran crab family Panopeidae Ortmann, 1893,
sensu Guinot (1978) (Decapoda). Crustaceana 50:182-
198.

Martin, J W., D L. Felder, and F M Truesdale.

1984. A comparative study of morphology and ontogeny
in juvenile stages of four western Atlantic xanthoid crabs
(Crustacea: Decapoda: Brachyura). Philos. Trans. R.
Soc. Lond., Ser. B, 303:537-604.

Martin, J. W., F. M. Truesdale. and D. L. Felder.

1985. Larval development of Panopeus bermudensis
Rathbun, 1891 (Brachyura, Xanthidae) with notes on
zoeal characters in xanthid crabs. J. Crustacean Biol.
5:84-105.

MONOD, Th

1956. Hippidea et Brachyura ouest-africains. M6m.
Inst. Fr. Afr. Noire 45:1-674.
Porter, H. J

1960. Zoeal stages of the stone crab, Menippe mercenaria
Say. Chesapeake Sci. 1:168-177.
Rathbun, M. J.

1930. The cancroid crabs of America of the families Eu-
ryalidae, Portunidae, Atelecyclidae, Cancridae, and Xan-
thidae. U.S. Natl. Mus. Bull. 152:1-609.
Rice, A. L.

1981a. The megalopa stage in brachyuran crabs. The

Podotremata Guinot. J. Nat. Hist. 15:1003-1011.
1981b. The zoea of Acanthodromia erinacea A. Milne Ed-
wards; the first description of a dynomenid larva (Deca-
poda, Dromioidea). J. Crustacean Biol. 1:174-176.
In press. The megalopa stage in majid crabs, with a re-
view of spider crab relationships based on larval charac-
ters. In A. A. Fincham and P. S. Rainbow (editors). As-
pects of decapod crustacean biology. Symp. Zool. Soc.
Lond., Vol. 59.
Scotto. L. E.

1979. Larval development of the Cuban stone crab,
Menippe nodifrons (Brachyura, Xanthidae) under labora-
tory conditions with notes on the status of the family
Menippidae. Fish. Bull., U.S. 77:359-386.
Williams, A. B

1984. Shrimps, lobsters, and crabs of the Atlantic coast of
the eEistem United States, Maine to Florida. Smithson.
Inst. Press, 550 p.
Williams, A. B., and D. L. Felder.

1986. Analysis of stone crabs: Menippe mercenaria (Say),
restricted, and a previously unrecognized species de-
scribed (Decapoda: Xanthidae). Proc. Biol. Soc. Wash.
99:517-543.

297

OCEANOGRAPHIC ASSOCIATIONS OF NEUSTONIC LARVAL AND
JUVENILE FISHES AND DUNGENESS CRAB MEGALOPAE OFF OREGON

Jonathan M. Shenker'

ABSTRACT

The larval and juvenile fishes and crabs inhabiting the neustonic zone within 50 km of the coast were
sampled biweekly from April through July 1984, with a Manta net (mouth 1.0 m wide x 0.7 m deep),
and a large neuston trawl (mouth 3.5 m wide x 1.0 m deep). The Manta net was an efficient sampler
for larval fishes and crabs, while the neuston trawl collected larger juvenile fishes that had rarely
been observed in previous studies.

Nocturnal sampling accounted for nearly all the ichthyoneuston and zooplankton taken. Dunge-
ness crab megalopae were the most abundant species. Although present throughout the survey, the
great majority were found in very large aggregations along visible convergence zones or in association
with Velella velella. Discrete groups of abundant larval and juvenile fishes were found prior to
upwelling [Parophrys vetulu3 , Scorpaenichthys marmoratus , Hemilepidotus spinosus, Hexagrammos
sp., and A noplopoma fimbria ) and after its onset (Engraulis mordax and Sebastes spp.). These species
had distinct zonal (east-west) distribution patterns and were generally associated with, or affected by,
hydrographic characteristics such as convergences, upwelling, and the Columbia River plume.

Recent studies of the ichthyoplankton off the
northwest coast of the United States have con-
tributed new information on the temporal and
spatial occurrences of larvae of coastal and
pelagic fish species (Richardson 1973; Richardson
and Pearcy 1977; Laroche and Richardson 1979;
Richardson et al. 1980; Kendall and Clark 1982a,
b; Clark 1984; Bates 1984; Mundy 1983; Brodeur
et al. 1985; Boehlert et al. 1985). These surveys
focused on larvae occurring below the surface
layer of the ocean, although concurrent neustonic
samples were occasionally collected (Kendall and
Clark 1982 a, b; Clark 1984; Bates 1984; Richard-
son^). Comparison of simultaneous surface and
subsurface samples demonstrated that many spe-
cies were found in both depth strata, while an
additional group of species was collected only
from the neustonic zone. Brodeur et al. (1987)
examined the larval fish and invertebrate compo-
nents of the neuston in the northeast Pacific, and
determined that these organisms were frequent
prey items of juvenile salmonids.

^Oregon State University, College of Oceanography, Hatfield
Marine Science Center, Newport, OR 97365; present address:
University of California, Bodega Marine Laboratory, P.O. Box
247, Bodega Bay, CA 94923.

2Richardson, S. L. Oregon's coastal ichthyoneuston - a pre-
liminary report. Unpubl. rep. Presented at American Soci-
ety of Ichthyologists and Herpetologists, Williamsburg, VA,
June 1975.

Manuscript accepted November 1987.
FISHERY BULLETIN; VOL. 86, NO. 2, 1988.

Standard plankton and neuston nets used in
these studies were effective in collecting the rela-
tively slow-moving early larvae. However, net
avoidance by larger larvae and juvenile stages
(Barkley 1972; Murphy and Clutter 1972) sug-
gests that use of traditional collecting gear is in-
appropriate for more mobile fishes. This paper
describes the results of a neustonic survey con-
ducted off the Oregon coast in the spring and sum-
mer of 1984 that focused on the larger ichthy-
oneuston. Conventional sampling gear and a new
net designed specifically for sampling juvenile
fishes were used to characterize the temporal and
spatial distribution patterns of both larval and
juvenile ichthyoneuston and Dungeness crab,
Cancer magister, megalopae.

MATERIALS AND METHODS

Sampling was conducted at approximately
2-wk intervals from early April through July
1984, on an east-west transect along the 44°40'N
parallel, 3 km north of Newport, OR (Fig. 1). Sta-
tions were located at distances of 1, 5, 10, 15, 20,
30, 40, and 50 km from shore. The stations were
occupied twice during each 24-h cruise, once dur-
ing the day and once at night. On one cruise (8-10
June), the 50 km station was occupied for 27
hours to assess diel variation in abundance of
neustonic organisms. In response to several

299

FISHERY BULLETIN: VOL. 86, NO. 2

45°
00'

-1 1 1 r-

440
50'

50 40

15 5
• • • •)

20 10 It

YAQUINA
BAY

125°
00'

124°
30'

124°
00'

Figure 1. — Location of stations along the transect off Newport,
OR. The numbers indicate distance offshore in kilometers.

weeks of strong upwelling, during which very few
larvae were collected close to the coast, the sam-
pling scheme was modified for the two July
cruises by eliminating some of the inshore sta-
tions and extending the transect as far offshore as
90 km. Only night samples were collected during
July.

One of the primary considerations of the survey
was to collect samples while minimizing net
avoidance by the target organisms. Reduction of
vessel-induced disturbance was accomplished by
deploying the collecting gear from the ends of 12
m-long outriggers on the FV Cumberland Trail,
a chartered 23 m commercial scallop-fishing ves-
sel.

Two different nets were used to collect the sam-
ples. Larval fish and zooplankton were collected
with a Manta neuston net (Brown and Cheng
1981), modified to have a mouth 1.0 m wide x 0.7
m deep. The Manta frame was equipped with a
green-colored 0.333 mm mesh net, PVC cod end
bucket, and General Oceanics model 2030 digital
flowmeter'^. A two-point bridle was attached to
the upper corners of the frame. Drag on the net
while towing kept the entire bridle and towing
wire assembly out of the water.

The second net was designed to collect the
larger juvenile fishes that were assumed to avoid
the smaller Manta net. The neuston trawl was
constructed with a mouth 3.50 m wide x 1.05 m
deep (Fig. 2a). The frame consisted of 43 mm (out-

3References to trade names do not imply endorsement by the
National Marine Fisheries Service, NOAA.

side diameter) heavy-duty galvanized pipe, with
towing points welded at the four corners. Flota-
tion for the frame was provided by three inflat-
able spar buoy floats. These 40 cm diameter floats
had hollow tubes running through their centers
and were fitted onto the upper bar of the frame.
The 8.5 m-long net was made of 4.8 mm green-
colored woven mesh, with a 15 cm-wide cloth
collar around the mouth, PVC cod end bucket and
General Oceanics model 2030 flowmeter. The net
had a mouth slightly larger than the frame
(3.70 m X 1.20 m) so the net could be laced around
the outside of the frame and flotation buoys. The
ends of six 12 mm polypropylene rope riblines
running the length of the net were shackeled to
the frame for additional support. A 4-point bridle
of 6.4 mm wire was attached to the corners of the
neuston trawl (Fig. 2b). Drag forces kept the en-
tire bridle and towing wire (except for the two
short segments attached to the bottom of the
frame) out of the water. A length of 5 cm
polypropylene rope was attached between the
upper towing points on the frame, as a bridle for
use in deploying and retrieving the net. A line
attached to the retrieval bridle was passed to a
hydraulic capstan through a block tied in the rig-
ging over the deck. During a tow, this retrieval
line was slackened, but remained attached to the
vessel, and floated in a broad arc behind the net
mouth.

At each station, the neuston trawl and Manta
net were towed simultaneously from the port and
starboard outriggers, at approximately 1-1.5 m/
second. Tows were generally made either against
or with the direction of the prevailing swells. The
Manta net was usually fished for approximately
8-9 minutes/tow and filtered about 300-400 m^.
The neuston trawl generally filtered 2,000-3,000
m'^ during a 10-11 minute tow. Additional tows
with the neuston trawl were frequently made at
a station to assess small-scale patchiness and
to sample visible "structures" in the surface
layer, such as convergence zones marked by foam
lines and rafts of the pleustonic hydroid Velella
uelella.

On several cruises, onset of high winds and
rough sea conditions prevented use of the Manta
net during the daj^ime sampling of the transect
as we returned toward shore. However, we were
able to fish the neuston trawl in winds up to an
estimated 40—45 km/hour, and in white-capped
seas of 2-3 m. Neuston trawl samples were col-
lected at alternate stations during the periods of
adverse weather.

300

SHENKER: OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON

0.4m

1.05m

3.50m

1.0m

6.0m

Figure 2a. — Front view of the neuston trawl; B = inflated 40 cm diameter spar buoys;

C = cloth collar around mouth of net; R = polypropylene rib lines.
Figure 2b. — Side view of the neuston trawl showing construction of the towing bridle.

Floats were removed for clarity.

Several very large catches of Dungeness crab
megalopae in the neuston trawl were subsampled
by volume, with the majority of the megalopae
being returned alive to the ocean. Samples from
both nets were preserved in 10% formalin/sea-
water and sent to the laboratory for analysis. All
larval and juvenile fishes were identified to the
lowest practical taxonomic level. After extrapo-
lating catch sizes from subsample counts, the
data were normalized to produce density esti-
mates of organisms/1,000 m^^. As an indicator of
patchiness of abundant taxa, the coefficient of dis-
persion (CD. = S'^/X ) was calculated for the sam-
ples in which each taxon was present.

Hydrographic and meteorological data were
collected at each station. Surface water tempera-
ture was determined with a handheld bucket
thermometer. Water samples were collected for
laboratory analysis of surface salinity using a
Model 8400 Guildline Autosalinometer. A 200

mL surface water sample was filtered through a
0.3 fxm pore size Gelman A/E glass fiber filter for
determination of chlorophyll a concentration. The
filters were stored in dessicant over dry ice at sea,
and then in a freezer, before acetone extraction of
the chlorophyll a and analysis with a Turner De-
signs Model 10 Fluorometer. During the daytime,
Secchi depths were estimated, using a 30 cm disc,
and surface irradiance was measured with a
handheld General Electric Model 214 Light
Meter. Other data collected included weather and
sea state conditions at each station.

RESULTS

Hydrographic Data

The neustonic realm was characterized by dis-
tinct temporal and spatial patterns that reflect
the highly dynamic springtime oceanographic

301

FISHERY BULLETIN: VOL. 86, NO. 2

processes of the Oregon coastal zone. Daily up-
welling indices derived from barometric pressure
data (Fig. 3; A. Bakun"*) indicated that in April
and May, local winds varied in direction and
strength, with one extended period of stormy
weather with southwest winds (negative index
values). Temperature and salinity profiles during
these months showed relatively little variation
along the transect within a sampling period (Fig.
4a, b, c). Chlorophyll a peaks and low salinity
near shore may have been due to the influence of
water exiting Yaquina Bay. The highest chloro-
phyll a values were observed following periods of
northwest winds, and the highest temperatures
occurred after a period of very light winds and
high insolation.

The hydrographic regime was altered by the
onset of northwest winds in June (Fig. 3), which
induced the upwelling of cold, more saline
nutrient-rich water along the coast. Strong up-
welling [defined by Small and Menzies (1981) as
a daily index value >50] was sporadic in June but
nearly continuous in July. Temperature, salinity,
and chlorophyll a profiles rapidly changed in re-
sponse to the upwelling-favorable winds (Fig. 5a,
b, c).

During the 8-10 June cruise, low temperature
and high salinity water was found within 5 km of
the coast. Low salinity, warm water more than 15
km offshore apparently was the plume of the Co-
lumbia River. By 19-20 June, continued offshore
transport of the surface layer pushed the inshore
edge of the plume to 40 km off the coast. Low
chlorophyll a concentrations, low temperatures,
and high salinity near shore on 19-20 June indi-
cated the occurrence of active upwelling.

Upwelling persisted through the end of July.
Temperature and salinity profiles were similar to
those observed during the second June cruise,
with the patterns of increasing temperature and
decreasing salinity farther offshore continuing to
the limits of sampling of 90 km on 8 and 9 July
and to 70 km on 23 and 24 July. During the up-
welling in July, a dramatic phytoplankton bloom
occurred in the surface waters within 30 km of the
coast. Chlorophyll a concentrations here were
denser and broader in offshore extent than the
surface and subsurface biomass concentrations
previously observed off Oregon (Small and Men-
zies, 1981).

+ 200

3 -200-

I ||'|||l||l ll'l' f |'l|| '1 11 1 T

lll.i„lil

yt4 y^ /N /K yfs yT\ xTs xl\

"•A. Bakun, Pacific Environmental Group, National Marine
Fisheries Service, Monterey, CA 93940, pars, commun. August
1984.

M J J

1984

Figure 3. — Daily upwelling indices for spring-summer 1984.
Positive values >50 indicate occurrence of winds inducing
strong upwelling. Arrows indicate dates of sampling cruises.

Ichthyoneuston

A total of 107 Manta net and 142 neuston trawl
samples collected 48 taxa of larval, juvenile, and
adult fishes. Larvae <10 mm and a few juveniles
were collected by the Manta net, while large
numbers of juvenile fishes up to 60-70 mm were
taken by the neuston trawl. Size-frequency data
for three of the commonest species (Fig. 6 a, b, c)
illustrate the relative ability of the nets to cap-
ture different sizes of fishes.

Nighttime sampling (59% of the Manta net
tows and 62% of the neuston trawl tows) ac-
counted for 93.3% of all fishes in the Manta net
catch and 96.5% of the fishes taken in the neuston
trawl (Table 1). Four Manta net and 8 neuston
trawl tows made during twilight collected 5.4%
and 2.9%, respectively, of the total number of
fishes taken with each gear type (Table 2). Only
1.3% of the Manta net catch and 0.6% of the neu-
ston trawl catch was made during daytime (Table
2).

Only two species were not collected predomi-
nantly at night. Larval Pacific saury, Cololabis
saira, were taken in both day and night Manta
net samples. Larval northern lampfish, Steno-
brachius leucopsarus , were abundant only in the
twilight samples from early June (Tables 1, 2).

Arbitrary criteria on frequency of occurrence
(in >15% of the night samples of either net) or
abundance (peak densities >40/l,000 m"^) indi-
cated that nine taixa of larval, juvenile, and adult
fishes were dominant components of the ichthy-
oneuston. Most taxa were characterized by dis-

302

SHENKER: OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON
14

o
o

<
if)

X
Q_
O

_J
X

O O 4/6

• • 4/16-17

A A 5/4-5

▲ A 5/18-19

10 20 30 40

DISTANCE FROM SHORE (km)

Figure 4a-c. — Nighttime environmental conditions observed on each cruise during the pre-upwelling period,
a = temperature; b = salinity; c = chlorophyll a concentration.

303

FISHERY BULLETIN: VOL. 86, NO. 2

o
o

01
LJ
Q.

o
o

Z
_J
<

>-
X
Q_
O

q:
o

30 40 50 60 70 80
DISTANCE FROM SHORE (km)

100

Figure 5a-c. — Nighttime environmental conditions observed on each cruise during the upwelling period,
a = temperature; b = salinity; c = chlorophyll a concentration.

304

SHENKER OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON

tinct temporal shifts in occurrence: Five taxa
were abundant only prior to the onset of continu-
ous upwelling in mid-June, three taxa were spo-
radically abundant following the beginning of up-
welling, and one species (Ronquilus jordani) was
present tliroughout the survey.

Pre-upwelling Species

The numerically dominant species in the
nearshore region in spring was larval English
sole, Parophyrys vetulus. These larvae were
patchily distributed (CD. = 32.5) within 30 km
offshore, reaching densities of 75/1,000 m'^ in
Manta net collections. Larval abundance declined
rapidly after April (Table 1). Size-frequency dis-
tributions indicated that the Manta net was more
effective than the neuston trawl for collecting fish
<15 mm (Kolmogorov-Smirnov test, P<0.01).
Eye migration in the larger fish (20-24 mm) was
nearly complete, indicative of their impending
shift to a benthic existence. Similar-sized juve-
niles were taken nearshore in April in benthic
tows (B. Mundy^).

Cabezon, Scorpaenichthys marmoratus , larvae
were most abundant in Manta catches in the
spring (Table 1), although newly hatched larvae
were collected near shore through June. The
smallest larvae (4-6 mm.) were taken only at the
1-15 km stations, while larvae up to 10 mm were
found along the entire transect. These larvae
were patchily distributed, with April densities
reaching 207/1,000 m^ (CD. = 92.7). Frequency
of occurrence and larval density declined through
early July. Large juvenile cabezon (26-38 mm)
were collected by the neuston trawl (Fig. 6a).
These juveniles were encountered all along the
transect from April to mid-June.

Two dominant taxa found at all stations prior
to upwelling were greenling (Hexagrammos sp.)
and brown Irish lord, Hemilepidotus spinosus . Ju-
venile greenlings were the most frequently col-
lected species, occurring in AA.9>% of the night-
time neuston trawl samples (Table 1), although
they were rarely taken by the Manta net (Fig. 6b).
They were the most evenly dispersed of the abun-
dant fish collected (CD. = 2.2), and never ex-
ceeded densities of 10/1,000 m^. The two largest
catches of greenling were made in association
with distinct hydrographic features. Nineteen ju-

a) Scorpaenichthys marmoratus

100 -,

50 -

I J Manta net

Neuston trawl

TT'

4 6 8 10 20 26 32 38

b) Hexagrammos sp.

40 —I

30 -

20 -

10 -

c) Anoplopoma fimbria

5B. C. Mundy, Southwest Fisheries Center Honolulu Labora-
tory, National Marine Fisheries Service, NOAA, 2570 Dole
Street, Honolulu, HI 96822, pers. commun. May 1984.

T r

10 20 30 40 50 60 70 80 90 100 IK) 120130

Standard Length (mm)

Figure 6a-c. — Length-frequency data of larval and juvenile
fishes collected by each net. a = Scorpaenichthys marmoratus ;
b = Hexagrammos sp.; c = Anoplopoma fimbria.

veniles were collected from a convergence zone
near shore in mid-May (along with 150,000 Dun-
geness crab megalopae) and 20 juveniles were
taken from Columbia River plume water 50 km
offshore in early June. Mean lengths of greenling
from the neuston trawl samples increased from

305

FISHERY BULLETIN: VOL. 86, NO. 2

£1

(/)

o
o

o
a>

C/5

y- in <M

'-'to

1/5 in CD

CD
CO

••- r- CO r^

1- CO CM C\J

r-- Tt LO CO
d o o d

C\J

CO CD

CO
Cvl

O
CD

C\J

T- -^ r-

CO
CO

t--

CO T •* T-

cD r-- CO Tf
CO in Tt CO
d d d> d

IT) CD O)

>- t- CO

CO
CD

CO
CO

T- •.- CD

in
uo

1 s

'J-

O
CO

Cvj

CO
C\J

- 1

;; 1

CM
'I-

CD
CM

CO
CO

■•-

CO o
CD O

CD
CO

CO
CO

CO CM CD
CM

t r^ r^
r- o tt

■^ I^ CD

CD
CO

T- O

r~ O) CO
o in CD

CM T-

CM
CO

CO
CD

CO
CO

T- CM CM •--

•r- CO O

in t^ CD

o
•<t

C\J 1- •.-
CO CO »-
Cvi T- CO

If)

CO
CO

CD
CO
CM

•,- ,- TJ-

CO
CO

CO
CM
CO

•<t

CM
CO

t- CO •«t
o in CO

T- d d

CD
CO

00
CO
CM

cri

in y- y-

o
in

n o

CD o

■<J CM CO 1
CD ■<* CM 1

1 ?

o
oq

d d d

00 CO

CO 00

CO
CO

00 •■- '-

CO T-

1 <^J

1 CD

" 1

■^

t ■*

in CM

CM CD

O CO
CD CM

■<}■

in CD I -r-
CO t^ I CO

^ 1

in y-
00 CD

CO CM

c
cb

■^ cc
CtJ 13
0) ^
Q. ^
..3" '°

.CO

ID
^ "J

2 "°

[5 c

O) o

- c p

C CO

-ie

c- ■fe 'I'

■O "D

,9> E

O ™ ^,^

55 2

3
C
CO <0

S iS

5 3
CD ^
O
CD

-c

5 c

£l ■^ "^
fc X: -o

O CD

CO o
-5 o

^ CO
CXJ CD

CO
3

'cB
-3J

3
CJ
CD

CO m
? "C

Oi ^
P CO

3 CO qj

C CO
O Q.

<D
<0 (D ^ O)

^ -S t ™

s|tl

-J £ "^ 2
< in

ill
£ 03 ">

(D
CO

>
_ 3

. . <D
Q. Q. <5

CO CO •;-
CO
e fi^ to CO

tf^ CD <D

CO
o a>

-Q

■^ 0)

(0 to

_ __ E
^.fe 9? S S §-■§

-5 c "o

t 3J to to

E S.

CO CO CO </) CO o

Cl Cl —
dj Cb Q.

. 3

o o

ia^l^

- CO- 9 .a

CD >< -C X
to CD Q. ><

306

SHENKER: OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON

c
o
O

<£>

CVJ

CO •* CO c\j r--
•■- C\J

r^ (j> ia n

O r- <£> ID

un o '-

■^ CM

■* CO

d d

uo
I/)

T- uo ■>- OJ •-

■.- O) T- ^ ^

•^ 00 uo CD in
d oj d (b o

(D
(D
CM

UO
00
CO

LO O

1 ^

Tl- CD

t-~^ If)

CO
CM

■■- CO

O CD
CM 00

1— (D
CD

1 1 '^

1 "^

■V

1 "^

CM 1

1 '- ^ CM 1

1 i ^

1 •*

•^ 1

1 in 1- 1- 1

1 1 r^

■*

1 c^

r-^ 1

1 CM in CD 1

T-'

d d (6

- I - CO

in ■<r

CD I T-
CO I T-

d CD

00
CD

C\J

■^ CM
O CO

■■- CO

cvj d
OJ

O)
CO

■^ 00 CM

CO
CD

■^ I h- T- in

CD CM CD
CO CM ••-

r- c\j in

■.- ^ Tf

in
o

1^
o

CO CM CO •— o

■* CO in CD 05

Tf CD TJ- o ■>-

CO in uri CD r--

'- ••-■<«■

CvJ

CD
CD

d d

■•- CM

d ■*

CO
CO

00
CO

Cvj
CM

'" CO

CO ■*

1 °°

CO
CD

1 CO

■*

CM
CM

r--

C3)

O o CO

tT3

5. t^ ,„
CO g «

CO nj cio <li o 3J

o
O

CO
CO

3
i.

'?: OI Jjco o^ g_

3
"O

O
ID C
(0 cq
■g -c

fC
CD

CB :9
E t;

U) O

CTJ
S '^

?15

O -C
^ Q.
m <i,
fc O
CD O

&§
Uj CD

'co
o

■ c
a. CO

^ £ 5
CO

CD -g o O —

lllll

3
CO

(LI

307

FISHERY BULLETIN; VOL. 86, NO. 2

tu c

_l o ■*

CD C 00

,< ^ O)

H CO ^

C7>

3
-3

E
Q

c
(0

Q.
<

en
0)
o

d
CO

C\J
<£>

■^

LT)

CD
CD

T-

<>i

•* y-

■■- •*

C\j

T-

c\i

o
in

CO o

o ••- T-

O £31 O

in in in

CO
CO

CVJ

CO
CO

CM
CO

CJ5
CO

CO
IT)

CO '-

CO CO

h- ■.- ,-

r- CD

CO
'-CD CD

O CO
CO CJ5

CD -^

■>*■

in in

CM Cvj ■■-■•-

CO
CO

CM CM

■^ CM

N- in

o
6

0) CO
CO -^

•D 3
CD Ol

a>uj

i CD

2
<n

— (D
CO CO

3 -g
£ o

- CJ O

CO CO en

■^ O QJ
Q. C: JD

2 Si E
"55 o

S CO

CO O)

CO TJ

CO ^

CO 2

CO CO

o E

o >-

o -o

3 CO

"3" y

C CO

"cD

CO
CU

-c

CO QJ

0) CO

u "9

§ o

^£

-^ E

s '^

C (0

CO jo

^^ . _ ,

O CO CO CO CO
CO

•Q

E
o

E
~ ''J

CD X

CO (U

>< I:

'cO
CD

c:
o
"S

O CD

Ob
o

o
o
o

in
cii

CO £

o <o

S E

CO £

CO
CD

,- CD C35 "-

CD r~ oo

»- CM 1-

CT>
00

CM
CO

^ 3

§o

- c 9f
E o ^ "a

li O i; F

ICO 2.i

CO o w

9 SJ Q.B

a? CD ■§_^;_
- S o ^

2 9- k <i>

3 .° ^ "O

CJ>

■^

•<r

1—

r~-

'—

l~-

'-'

(ii

■0

E
3
c

3
CO

308

SHENKER: OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON

38.8 mm (A^ = 10) in early April to 51.2 mm
{N = 72) during the first June cruise.

Brown Irish lord juveniles were the numeri-
cally most abundant and the second most fre-
quently occurring species (Table 1). Early juve-
niles (10-20 mm) were taken with both nets in
April and May only at the offshore (40-50 km)
stations. Larger juveniles (20-30 mm) were
patchily distributed from April through mid-June
all along the transect (CD. = 14.1). Although
47% of the total number of trawl caught speci-
mens were taken in densities below 15/1,000 m-^,
repeated sampling in one large aggregation ac-
counted for the bulk of the catch. Three sequen-
tial tows at the 50 km station on 5 May had juve-
nile densities of 15-39/1,000 m*^ , but distinctive
hydrographic characteristics at the station were
not detected.

The only species with a characteristic offshore
distribution prior to upwelling was sablefish,
Anoplopoma fimbria. Small juveniles (10-45
mm) were effectively taken by the Manta net,
while the neuston trawl was more efficient at col-
lecting 30-70 mm specimens (Fig. 6c). The 10-50
mm fish were collected primarily at the 50 km
station from April through mid-June, although
several specimens were taken closer inshore. The
50-75 mm fish were captured only on 8-10 June
from the Columbia River plume water at the 50
km station. After the onset of upwelling, two indi-
viduals ( 120-130 mm) were taken 80-90 km from
shore.

Upwelling Species

Three taxa were abundant on only one of three
cruises made after the onset of upwelling in mid-
June (Table 1). Distributions of northern an-
chovy, Engraulis mordax, larvae; rockfish (Se-
bastes spp.) larvae; and adult blue lanternfish,
Tarletonbeania crenularis, overlapped at some
offshore stations on 8 and 9 July. The offshore
portion of the transect was characterized by a
drop in salinity and an increase in temperature
between 60 and 70 km offshore (Fig. 5a, b), indi-
cating a transition from coastal to Columbia
River plume water.

Of the 814 rockfish larvae collected throughout
the survey, 87% were taken at the 50-90 km sta-
tions on this cruise, with a peak density of the
3.5-7.0 mm larvae of 684/1,000 m^. Adult blue
lanternfish were present in densities up to 315/
1,000 m"^ at the 60-90 km stations. Lanternfish
from simultaneous Manta and neuston trawl tows

displayed no difference in length-frequency dis-
tributions, although a significant change in the
size structure of the catches between stations was
observed (K-S test, P < 0.01). Mean size of the
fish decreased approximately 10% between adja-
cent stations, from 50.9 mm at 60 km to 37.8 mm
at 90 km. Anchovy larvae (3-9 mm) were re-
stricted to the lower salinity plume water at the
70-90 km stations, with peak and mean densities
of 368 and 210/1,000 m^, respectively.

Persistant Species

Only northern ronquils, Ronquilus jordani,
were abundant in both pre-upwelling and up-
welling periods. The elongate larvae and juve-
niles were collected with the Manta net starting
in mid-April, with abundances peaking at 40/
1,000 m*^ in July. Mean lengths over this time
interval increased from 8.7 mm (n = 8) to 26.6
mm (n = 43). Ronquils were relatively dispersed
along the transect (CD. = 7.6), and size or sea-
sonal patterns of distribution were not detected.

Dungeness Crab Megalopae

Dungeness crab megalopae were the dominant
component of the catches throughout the survey,
with an estimated total of 350,000 megalopae col-
lected in 249 hauls. The megalopae were found at
most stations during all cruises (Fig. 7), although
they were rare seaward of 50 km in July. Megalo-
pae were present in 71.4% of the Manta net day-
time tows and 93.8% of the nighttime tows, and in
90.9% and 98.9% of the day and night neuston
trawl collections, respectively. All twilight hauls
captured megalopae.

Despite the higher frequency of occurrence of
megalopae in the neuston trawl, the Manta net
was a more accurate estimator of their abun-
dance. For 58 pairs of nighttime catches with den-
sities under 2,000/1,000 m-^, the Manta net caught
approximately 3 times as many megalopae per m^
as the neuston trawl. The 4.8 mm mesh of the
neuston trawl apparently enabled some megalo-
pae to pass through the mesh, and some escape-
ment was observed while the trawl was sitting on
deck after retrieval. More similar estimates be-
tween the nets were obtained when higher densi-
ties of megalopae were sampled, probably because
of clumping of the large (11 mm total length)
spiny megalopae into large masses which clogged
the mesh.

In general, densities of megalopae caught by

309

FISHERY BULLETIN: VOL. 86, NO. 2

D — D Day

1.000

6 April

X A

100
10

O-D /

H 1 1 ffl 1

\
\

1 1 El

1.000

100

10 ■

A A Night

8-10 June

1
\

t >\ r

1 -
0

< 1 1 E — ffl-

-ffl-

-E-

-m

1.000 •■

16-17 April

1.000 ■•

100D

1.000-

8-9 July

^"^A

100-

■

10-

•

■i — 1 1 1 1

— 1

1.000 ■

100•

1 +

t I

H H

A 18-19 May

I.OOOt

100-

0 5 10 15 20 30 40

1 ■•
0

23-24 July
A— i

H 1-

0 5 10 15 20 30 40

Distance From Shore (km)

Figure 7.— Daytime and nighttime densities of megalopae collected by the Manta net. Rough weather prevented the use of the
Manta net during the day of 19 May; neuston trawl data for this portion of the cruise are presented. Circled letters refer to events
discussed in the text.

310

SHENKER; OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON

the Manta net in the daytime were lower than at
night for each station (Fig. 7). The differences
between day and night were particularly appar-
ent on bright sunny days and at the offshore sta-
tions. Only one of 13 samples of megalopae col-
lected when light intensity exceeded 4,000 lux
had a density over 200/1,000 m^, while 12 of 24
day samples collected at lower light levels ex-
ceeded this density. Highest daytime densities
(1,600-2,400/1,000 m^ in neuston trawl samples)
were observed on 19 May when a sudden shift in
weather before daybreak resulted in very dark,
overcast and rainy skies, with 40-45 km/hour
winds and 2-3 m breaking waves.

Water clarity may also affect the daytime
abundance estimates of megalopae in the neuston
along the transect. Daytime densities were al-
ways low seaward of 30 km where Secchi depths
exceeded 9 m, while higher densities were ob-
served in the more turbid near-coastal waters
with Secchi depths of 4-9 m.

The 27-h occupancy of the 50 km station on 8-9
June provided additional evidence of a distinct
diel pattern in utilization of the neustonic habitat
by megalopae (Fig. 8). Abundance of megalopae
peaked during the twilight periods of dawn and
dusk. A significant decrease in abundance was
noted in the middle of the night, while midday
samples did not collect any megalopae.

In addition to the wide distribution of megalo-
pae along the transect, sampling occasionally de-
tected the aggregation of very large numbers of
megalopae (denoted by letters over station blocks

1,000

o
o
o

0)
D

a.
o

100 --

o o Manta Net

• • Neuston //

Trawl //

I
//

1 -: .

0 I ' ' f^m^^^^^m=^ I i I . ii I T-rr '•' i

18:00 24:00 06:00 12:00 18:00

Time of Day

Figure 8. — Density of megalopae collected by h)oth nets during
a 27 h sampling p>eriod 50 km ofTshore on 8-9 June.

in Figure 7). These aggregations were typically
associated with hydrodynamic or biological struc-
tures in the surface layer.

Event A: Simultaneous tows with the Manta
net and neuston trawl caught megalopae in densi-
ties of 373 and 325/1,000 m^. The neuston trawl
was then redeployed and towed along a foam line
marking a surface convergence. An estimated
60,000 megalopae were collected from the conver-
gence zone (24,000/1,000 m^).

Event B: A series of five consecutive tows
with the trawl net was made. Each tow covered a
linear distance of approximately 1 km in a west-
ward direction, with an estimated distance of 300
m between tows. Megalopae were very sparse in
the simultaneous Manta and trawl tows at the
beginning of the series, with densities of 0 and
1.9/1,000 m"^, respectively. The subsequent neu-
ston trawl tows produced density estimates of 79,
3,555, 2,414 and 187/1,000 m^. Apparently, the
megalopae were aggregated over an area extend-
ing at least 5 km in the east-west direction.

Event C: Following a period of southwest
winds, a large raft of Velella velella was found
around the inshore station. Velella velella were
generally found at offshore stations but at densi-
ties far lower than the 36 kg/1,000 m^ level taken
here. Although the overall extent of the raft could
not be determined at night, the density of megalo-
pae was estimated by both nets at 21,500/1,000

Event D: The density of megalopae sampled
by the Manta net was 790/1,000 m^, while a
simultaneous neuston trawl tow along a conver-
gence zone only 30 m away collected 150,000
megalopae (46,400/1,000 m^).

Census estimates of megalopal abundance were
computed from nighttime samples for each date
(Fig. 9). A modification of Smith's regional census
estimate (1972) was used to compute the total
number of megalopae occurring in a 1 m
wide X 1 m deep track along the surface from the
coast to 55 km offshore:

S (^A)

j=i

where C^ = estimate of abundance of megalopae
per meter of coastline during cruise
k
Aj = volume of water in the track sur-
rounding station i, computed from

311

FISHERY BULLETIN: VOL. 86, NO. 2

10'

_o
a

4 ..

-O

O
O Transect Census
• Convergence Zones

10-

4/6 4/16 5/5 5/18 6/10 6/20 7/8 7/23

Sampling Date

Figure 9. — Census estimates of Dungeness crab megalopae at
night from Manta net samples for each cruise in a 1 m wide x 1
m deep track extending from the coast to 55 km offshore (30-95
km offshore on 7/8). Single samples from convergence zones
sampled with the neuston trawl contained more megalopae
than the entire census estimates for the respective cruises.

the midpoints of the distances be-
tween the adjacent stations
Z), = density of megalopae at station i.

Although census estimates varied between
cruises from approximately 4,000-78,000, these
estimates ignored the concentrations of megalo-
pae in convergence zones or other aggregations
that were sampled by the neuston trawl. The con-
vergences alone contained 5-15 times more
megalopae than the entire census estimates for
their respective cruises.

DISCUSSION

The neustonic habitat off Oregon is a dynamic
environment that supports an abundant and di-
verse fauna including Dungeness crab larvae and
numerous fish species. The large mouth size of
the neuston trawl and minimal disturbance from
the vessel probably contributed to larger catches
of surface-dwelling ichthyofauna than have been
made in previous studies in the same region
(Boehlert et al. 1985).

Several species of neustonic organisms were as-
sociated with specific oceanographic features
such as convergence zones and water masses like
the Columbia River plume. These aggregations of
Dungeness crab megalopae and several fish spe-
cies accounted for the major portion of the total
catch of each taxon. Recent studies by Shanks
(1983, 1985) and Kingsford and Choat (1985,
1986) further emphasize the importance of aggre-
gation of neustonic organisms in oceanic conver-

gences, surface slicks, and around floating ob-
jects. This information clearly demonstrates that
randomized or grid sampling plans will often fail
to detect micro- or meso-scale features of the
environment. Surveys may thus severely under-
estimate the abundance of species, and impor-
tant data on ecological characteristics such as
predator-prey interactions may not be obtained.
This is particularly important in understanding
the role of spatial co-occurrence of patches of lar-
val fishes and their prey on larval survival and
growth (Lasker 1975, 1981; Grover and 011a
1986).

Diel patterns in abundance were striking. Most
of the neustonic organisms were collected at
night, and their absence from daytime collections
may be attributed to vertical migration out of the
surface layer or to visual avoidance of the nets.
The fact that very few larval or juvenile sablefish
and greenlings were collected in subsurface sam-
ples in earlier studies (e.g., Richardson and
Pearcy 1977; Richardson et al. 1980; Kendall and
Clark 1982a; Clark 1984), suggests that these
fishes are obligate inhabitants of the neuston,
and their low abundance in daytime hauls indi-
cates substantial visual avoidance of the nets.

Conversely, other species may be facultative
neuston that undergo diel migrations into the
surface layer (Hempel and Weikert 1972), where
they can most easily be assessed. Dungeness crab
megalopae, in particular, disperse to at least 50 m
during the day, but concentrate at the surface at
dawn and dusk (Booth et al. 1985). Still other
species (the pseudoneuston) may have depth
ranges that overlap with the surface layer.

Previous studies on northwest ichthyoplankton
have defined basic coastal, transitional, and off-
shore assemblages of species during different sea-
sons (Richardson and Pearcy 1977; Richardson et
al. 1980; Kendall and Clark 1984a, b; Clark
1984). In general, the transitional region roughly
parallels the shelf break (approximately 30-40
km offshore in the vicinity of Newport, OR).
Richardson et al. (1980) ascribed the consistency
of these zonal assemblages to the spawning habits
of adults and larval transport in the alongshore
coastal circulation. According to Parrish et al.
(1981), the spatial and temporal patterns of
spawning, and durations of pelagic larval stages
of these species, should correspond with the sur-
face drift patterns of the region to minimize lar-
val advection out of suitable habitats. In the
Pacific Northwest, species with larvae adapted to
the nearshore zone should spawn from fall

312

SHENKER: OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON

through early spring when net surface drift is
primarily northward and onshore. Oceanic lar-
vae, however, should be spawned following the
spring-summer transition when upwelling re-
sults in southward and offshore transport.

The onshore/offshore distributions of several
abundant fish species in 1984 were similar to
those described earlier, and their seasonal occur-
rence usually coincided with the constraints dis-
cussed by Parrish et al. ( 1981). However, four spe-
cies had distributions that crossed the previously
described zonal boundaries (Fig. 10). Early
cabezon larvae appeared to disperse offshore as
they grew. In contrast, early juvenile brown Irish
lords were collected offshore in the early spring,
while the larger juveniles spread inshore across
all stations. Larval greenlings have only been col-
lected close to the coast in late fall (Bates 1984),
while the juveniles collected in this study were
found at all stations prior to upwelling. Juvenile
ronquils were also distributed along the transect,
during both non-upwelling and upwelling condi-
tions.

The dispersion of organisms across the transi-
tion zone may have been accomplished by several

mechanisms. Physical transport of the organisms
by mesoscale hydrographic events (e.g., eddies,
offshore jets, and meanders in the alongshore cur-
rents) (Ikeda and Emery 1984; Mooers and
Robinson 1984; Abbott and Zion 1985; Davis
1985) undoubtedly play important roles in the
onshore/offshore dispersal of planktonic organ-
isms. Transport with these events may be acceler-
ated or hindered by diel vertical migration into
water layers with different zonal flow patterns.
Additionally, several species (especially green-
ling and sablefish) are rapid swimmers whose mo-
bility can contribute to their dispersal or aggrega-
tion.

The rapid change in the composition of the
ichthyoplankton fauna following the onset of up-
welling has not been previously observed. Fishes
abundant prior to upwelling may respond to the
change in the environment in different ways. The
disappearance of greenling and brown Irish lord
juveniles suggests upwelling triggered settle-
ment of these fishes to the demersal habitat uti-
lized by older juveniles and adults. A possible
stimulus for this transition is the breakdown of
the thermocline, which has been identified as a

1 '
90 8*0 70 60

UPWELLING

1 1 1 1
PRE-UPWELLING

V."

45°

nn'

Sablefish 1

10-20 nnm

20-30 mm Brown Irish Lord

(vy:

Greenlinq

p-.

/■■'■■

4-6

1 -

^[Vaquina

A^y BAY

:■ 1

6-10 mm Cabezon

English Sole

• • • • • •

50 40 30 20 15 10
km

•

4 4"

_ Rockfish SDD.

3 0'

1 1 1 1

Blue Lanternfish

Northern Anchovy

125°
00'

124°
30'

124°
00'

Figure 10. — Summary of the zonal distributions of larval and juvenile neustonic fishes in
the pre- and early upwelling period (April through mid-June) and the strong upwelling
period (July).

313

FISHERY BULLETIN: VOL. 86, NO. 2

barrier to settlement to neustonic juvenile red
hake, Urophycis chuss, (Steiner et al. 1982;
Steiner and 011a 1985).

In contrast, juvenile sablefish apparently re-
main in the water column off Oregon following
the transition to upwelling conditions. Small ju-
veniles (<50 mm) have been captured as far as
250 km offshore in the spring (Kendall and Clark
1982a), and specimens up to 250 mm have been
collected from the surface waters in late summer
with a purse seine (Brodeur and Pearcy 1986).
The very rapid growth rates of young juveniles,
reaching 2 mm/day (Boehlert and Yoklavich
1985; Shenker and 011a 1986), undoubtedly re-
sulted in their increasing ability to avoid the
towed neuston nets used in this study.

The most abundant larvae occurring after the
onset of upwelling were northern anchovy and
rockfish. As observed in previous studies
(Richardson 1973, 1980; Richardson et al. 1980),
these larvae were primarily found offshore. The
occurrence of these species, along with a high
abundance of vertically migrating blue lantern-
fish, on the periphery of the Columbia River
plume 60-90 km offshore on only one cruise, fur-
ther illustrates the patchy nature of neustonic
distributions.

Although the fish fauna in 1984 was generally
characterized by discrete temporal and/or spatial
limitations, the occurrence of Dungeness crab
megalopae transcended these limits through the
4-mo study. Megalopae were the most abundant
organisms collected throughout the survey, but
their abundance varied widely between adjacent
stations, with occasional very dense patches. Sim-
ilar patchy distributions of megalopae were ob-
served off British Columbia by Booth et al. (1985),
who measured horizontal patch dimensions of
2-4 km. Several dense swarms of megalopae were
observed in Bodega Bay, CA, during 1985
(Shenker and Botsford^). These patches were
long, sinuous aggregations extending 10-20 m
along the surface, approximately circular in cross
section and about 1 m in diameter. Densities were
visually estimated to be on the order of thousands
of megalopae per m^.

The temporal occcurrence of crab larvae in the
plankton spans two distinctly different oceanic
regimes, and the larvae are potentially trans-
ported long distances by the seasonal currents.

^Shenker, J. M., and L. W. Botsford, University of California,
Bodega Marine Laboratory, P.O. Box 247, Bodega Bay, CA
94923, unpubl. data.

Larvae typically hatch in mid-winter, pass
through 5 zoeal stages in approximately 90 days,
and then a month-long megalops stage before set-
tling to the bottom (Reilly 1983a). After hatching,
zoeae are released into the northerly flowing
Davidson Current. Despite a general onshore
component of flow of the Davidson Current, older
zoeal stages have been found progressively far-
ther offshore (Lough 1976; Reilly 1983a).

About the time of the spring transition in the
alongshore currents from a northerly to southerly
direction (Huyer et al. 1975), zoeae metamor-
phose into megalopae. To survive, these megalo-
pae must be transported back toward the shore,
and settle to the bottom in depths shallower than
25 m (Reilly 1983b). Larvae have been found at
least 100 km from shore (Lough 1976), although
it is unclear if these larvae make it back to shore,
or are simply lost from the population.

Again, an apparent anomaly exists between the
directions of movement of megalopae and surface
waters, where the Ekman layer moves offshore in
response to upwelling winds. These discrepancies
may be explained by mesoscale mixing processes,
as cited earlier. Alternatively, the diel vertical
movements of the larvae can move them into dif-
ferent water masses with different directions of
zonal movement.

The data from this study and previous research
on vertical migration indicate that at least the
early zoeal stages and megalopae move to the sur-
face during twilight, and below the surface dur-
ing the day (Reilly 1983a). Surface abundance
estimates of megalopae obtained in the 27-h sam-
pling on 8-9 June (Fig. 8), and by Booth et al.
(1985), decreased during the middle of the night.
This movement away from the surface is an ex-
ample of "midnight scattering", perhaps result-
ing from the lack of a light cue to orient plank-
tonic organisms to the surface (Owen 1981). If
midnight scattering is typical for megalopae,
abundance estimates made by sampling along a
transect throughout the night (Figs. 7, 9) proba-
bly underestimate the actual abundance of mega-
lopae utilizing the surface habitat. Megalopae
have been collected as deep as 50-70 m during
the day (Booth et al. 1985; Shenker and Botsford
fn. 6). These observations correlate with Jacoby's
(1982) laboratory demonstration that megalopae
are positively phototactic to dim light, but avoid
bright light.

The present study indicates that the phototac-
tic response of megalopae may assist their return
to shore in several ways. Downward movement

314

SHENKER: OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON

during sunny days, when upwelling wind stress
and offshore Ekman transport is generally
strongest, can move the larvae into the slow on-
shore flow below the surface Ekman layer (Peter-
son et al. 1979). In contrast to the usual northerly
upwelling winds, occasional storms blow from the
southwest, driving the surface layer onshore. The
occurrence of megalopae at the surface during the
day was most pronounced on these dark stormy
days (especially on the second May cruise), and
thus may facilitate their onshore transport.

Observations of megalopae entering into em-
bayments and nearshore areas from Washington
to northern California in 1984 and 1985 indicated
a dramatic increase in the abundance of megalo-
pae and an extended seasonal occurrence in the
plankton, as compared to previous years (Shenker
and Botsford fn. 6; Armstrong^). This high abun-
dance of crab larvae may presage an upswing in
the cyclical crab fishery along these coasts.

Numerous mechanisms have been proposed as
causes of the 10-yr cycles in crab abundance. Al-
though some hypotheses have been discounted,
several models have survived scrutiny as possible
causes of the cycles (see Botsford 1986 for review).
Potential mechanisms of environmental forcing
of the cycles focus on larval transport and sur-
vival. Johnson et al. (1986) detected periodic 10-
yr cycles in the occurrence and strength of south-
ward stress during the late larval period that
significantly correlated with commercial crab
catch 4 and 5 years later. This lag corresponds to
the time between larval settlement and growth
into the adult fishery (Botsford 1984).

Model simulations by Botsford (1986) indicated
that nonlinear effects of wind on larval transport
can produce the cyclical swings in crab abun-
dance. However, the models do not preclude the
possibility that density-dependent phenomena
(such as cannibalism by adult crabs on newly set-
tled juveniles, and predation by nemertean
worms on egg masses) may act in concert with the
environmental forcing to produce the observed
cycles.

The water's surface is the only oceanic habitat
that is easily accessible to observation using tech-
niques ranging from satellite and aerial scanning
to shipboard visual sighting of targets and contin-
uous monitoring of environmental parameters.
Because of this accessibility, micro- and meso-
scale patterns in distribution of neustonic taxa

'^D. A. Armstrong, School of Fisheries, WH-10, University of
Washington, Seattle, WA 98195, pers. commun. October 1984.

and their associations with hydrographic and bio-
logical characteristics of the surface zone can be
determined more easily than in other environ-
ments. The neustonic realm thus offers an excel-
lent opportunity to investigate the mechanisms of
transport of the early stages into appropriate
nursery habitats, and the availability of food for
growth that are required for successful recruit-
ment into adult stocks (Hjort 1914; Lasker 1975,
1981; Frank and Leggett 1982; Sinclair et al.
1984).

ACKNOWLEDGMENTS

This research was supported by the Northwest
and Alaska Fisheries Center of the National
Marine Fisheries Service. 1 thank Captain Leland
Oldenberg and his crew on the FV Cumberland
Trail, and my colleagues R. Brodeur, A. Chung,
J. Fisher, J. Hennessey, L. Krasnow, B. Mundy,
C. Paczkowski, and E. Rexstad for their assis-
tance on the sampling cruises. W. Laroche helped
identify juvenile Sebastes. W. G. Pearcy provided
valuable advise on the design and operation of
this project, and comments on the manuscript.
Helpful suggestions on this paper were also given
by R. Brodeur, A. W. Kendall, and two anony-
mous referees.

LITERATURE CITED

Abbott. M. R., and P. M. Zion.

1985. Satellite observations of phytoplankton variability
during an upwelling event. Cont. Shelf Res. 4:661-680.
Barkley, R. A.

1972. Selectivity of towed net samplers. Fish. Bull., U.S.
70:799-820.

Bates. R. D.

1984. Ichthyoplankton off Washington, Oregon and
northern California, October-November, 1981. Natl.
Mar. Fish. Serv., Northwest and Alaska Fish. Cent. Proc-
ess. Rep. 84-12, 41 p.

Boehlert, G. W., D M. Gadomski, and B C. Mundy.

1985. Vertical distribution of ichthyoplankton off the Or-
egon coast in spring and summer months. Fish. Bull.,
U.S. 83:611-622.

Boehlert, G. W., and M. M. Yoklavich.

1985. Larval and juvenile growth of sablefish,
Anoplopoma fimbria, as determined from otolith incre-
ments. Fish. Bull., U.S. 83:475-481.
Booth. J., A. Phillips, and G S Jamieson.

1985. Fine scale spatial distribution of Cancer magister
megalopae and its relevance to sampling methodol-
ogy. In B. R. Melteff (editor), Proceedings of the sympo-
sium on Dungeness crab biology and management,
p. 273-286. Univ. Alaska Sea Grant Coll. Program,
Rep. No. 85-3. Fairbanks, AK 99701.
Botsford. L. W.

1984. Effect of individual growth rate on expected behav-

315

FISHERY BULLETIN: VOL. 86, NO. 2

ior of the northern California Dungeness crab (Cancer
magister) fishery. Can. J. Fish. Aquat. Sci. 41:99-107.

1986. Population dynamics of the Dungeness crab (Cancer
magister). In G. S. Jamieson and N. Bourne (editors),
North Pacific workshop on stock assessment and man-
agement of invertebrates. Can. Spec. Publ. Fish. Aquat.
Sci. 92:140-153.

Brodeur, R D. D M Gadomski, W G Pearcy. H P.
Batchelder, and C B Miller

1985. Abundance and distribution of ichthyoplankton
in the upwelling zone off Oregon during anomalous el
Nino conditions. Estuarine Coastal Shelf Sci. 21:365-
378.

Brodeur, R. D. B C Mundy, W G Pearcy, and R W.

WiSSEMAN.

1987. The neustonic fauna in coastal waters of the north-
east Pacific: Abundance, distribution, and utilization by
juvenile salmonids. Oreg. State Univ. Sea Grant Coll.
Program, Rep. ORESU-T-87-001. Corvallis, OR 97331.

Brodeur, R D . and W. G. Pearcy.

1986. Distribution and relative abundance of pelagic non-
salmonid nekton off Oregon and Washington, 1979-
1984. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 46,
85 p.

Brown, D M , and L Cheng

1981. New net for sampling the ocean surface. Mar.
Ecol. Prog. Ser. 5:225-227.

Clark, J B

1984. Ichthyoplankton off Washington, Oregon and
northern California, May-June, 1981. Natl. Mar. Fish.
Serv., Northwest and Alaska Fish. Cent. Process. Rep.
84-11, 46 p.

Davis, R E

1985. Drifter observations of coastal surface currents dur-
ing CODE: the method and descriptive view. J.
Geophys. Res. 90:4741-4755.

Frank, W T , and W C. Leggett.

1982. Coastal water mass replacement: its effect on
zooplankton dynamics and the predator-prey complex as-
sociated with larval capelin (Mallotus villosus ). Can. J.
Fish. Aquat. Sci 39:991-1003.

Grover, J J , AND B L Olla.

1986. Morphological evidence for starvation and prey size
selection of sea-caught larval sablefish, Anoplopoma fim-
bria. Fish. Bull., U.S. 84:484-489.

HEMPEL, G , AND H Weikert

1972. The neuston of the subtropical and boreal North-
eastern Atlantic Ocean. A review. Mar. Biol. (Berl.)
13:70-88.
Hjort, J

1914. Fluctuations in the great fisheries of northern
Europe viewed in the light of biological research.
Rapp. P. -v. Reun. Comm. int. Explor. Sci. Mer 20:1-
228.
HUYER, A , R D PiLLSBURY, AND R L. SMITH

1975. Seasonal variation of the alongshore velocity field
over the continental shelf off Oregon. Limnol.
Oceanogr. 20:90-95.
Ikeda, M . AND W J Emery

1984. Satellite observations and modeling of meanders in
the California Current system off Oregon and Northern
California. J. Phys. Oceanogr. 14:1434-1450.
Jacoby, C a.

1982. Behavioral responses of the larvae oiCancer magis-
ter Dana (1852) to light, pressure, and gravity. Mar.
Behav. Physiol. 8:267-283.

Johnson, D F , L W Botsford, R D Methot, Jr , and T C.
Wainwright

1986. Wind stress and cycles in Dungeness crab (Cancer
magister) catch off California, Oregon and Washing-
ton. Can. J. Fish. Aquatic Sci. 43:838-845.
Kendall, A W . Jr , and J Clark

1982a. Ichthyoplankton off Washington, Oregon, and
Northern California. April-May, 1980. Natl, Mar.
Fish. Serv., Northwest Alaska Fish. Cent. Process. Rep.
82-11, 44 p.
1982b. Ichthyoplankton off Washington, Oregon and
Northern California. August, 1980. Natl. Mar. Fish.
Serv., Northwest and Alaska Fish. Cent. Process. Rep.
82-12, 43 p.
Kingsford, M J , AND J H Choat

1985. The fauna associated with drift algae captured with
a plankton-mesh purse seine net. Limnol. Oceanogr.
30:618-630.

1986. Influence of surface slicks on the distribution and
onshore movements of small fish. Mar. Biol. (Berl.)
91:161-171.

Laroche, J L, AND S. L Richardson

1979. Winter-spring abundance of larval English sole,
Parophrys vetulus , between the Columbia River and
Cape Blanco, Oregon, during 1972-1975, with notes on
occurrences of three other pleuronectids. Estuarine
Coastal Mar. Sci. 8:455-476.

Lasker, R

1975. Field criteria for survival of anchovy larvae: the
relation between the inshore chlorophyll layers and suc-
cessful first feeding. Fish. Bull., U.S. 71:453-462.

1981. The role of a stable ocean in larval fish survival and
subsequent recruitment. In R. Laisker (editor), Marine
fish larvae. Morphology, ecology and relation to fisheries,
p. 80-87. Wash. Sea Grant Program, Univ. Wash.
Press, Seattle.

Lough, R G

1976. Larval dynamics of the Dungeness crab, Cancer
magister, off the central Oregon coast, 1970-71. Fish.
Bull., U.S. 74:353-375.

MooERS, C. N K , and a. R. Robinson.

1984. Turbulent jets and eddies in the California Current
and inferred cross-shore transport. Science 223:51-53.
MUNDY, B. C.

1983. Yearly variation in the abundance and distribution
of fish larvae in the coastal zone off Yaquina Head, Ore-
gon, from June 1969 to August 1972. M.S. Thesis, Ore-
gon State Univ., Corvallis, OR, 158 p.
MuRPHY, G I , AND R I Clutter

1972. Sampling achovy larvae with a plankton purse
seine. Fish. Bull., U.S. 70:789-798.
Owen, R W

1982. Fronts and eddies in the sea: Mechanisms, interac-
tions and biological effects. In A. R. Longhurst (editor).
Analysis of marine ecosystems, p. 197-233. Acad.
Press, N.Y.

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

1981. Transport mechanisms and reproductive success of
fishes in the California Current. Biol. Oceanogr. 1:175-
203.
Peterson, W T , C B Miller, and A. Hutchinson.

1979. Zonation and maintenance of copepod jwpulations
in the Oregon upwelling zone. Deep-Sea Res. 26:467-
494.
Reilly, P N

1983a. Dynamics of Dungeness crab. Cancer magister.

316

SHENKER: OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON

larvae off central and northern California. In P. W.
Wild and R. N. Tasto (editors!, Life history, environment,
and mariculture studies of the Dungeness crab. Cancer
magister . with emphasis on the central California fish-
ery, p. 57-85. Calif Dep. Fish game, Fish Bull. 172.

1983b. Predation on Dungeness crabs, Cancer magister,
in central California. In P. W. Wild and R. N. Tasto
(editors), Life history, environment, and mariculture
studies of the Dungeness crab. Cancer magister, with
emphasis on the central California fishery, p. 155-
164. Calif Dep. Fish. Game, Fish Bull. 172.
Richardson. S L.

1973. Abundance and distribution of larval fishes in wa-
ters off Oregon, May-October 1969, with special empha-
sis on the northern anchovy, Engraulis mordax. Fish.
Bull., U.S. 71:697-711.

1980. Spawning biomass and early life of northern an-
chovy, Engraulis mordax, in the northern subpopulation
off Oregon and Washington. Fish. Bull., U.S. 78:855-
877.
Richardson, S L . J L Laroche, and M. D. Richardson.

1980. Larval fish assemblages and associations in the
northeast Pacific Ocean along the Oregon coast, winter-
spring, 1972-1975. Estuarine Coastal Mar. Sci.
11:671-699.
Richardson, S L , and W G Pearcy.

1977. Coastal and oceanic fish larvae in an area of up-
welling off Yaquina Bay, Oregon. Fish. Bull., U.S.
75:125-146.
Shanks, A L

1983. Surface slicks associated with tidally forced inter-

nal waves may transport pelagic larvae of benthic inver-
tebrates and fishes shoreward. Mar. Ecol. Prog. Ser.
13:311-315.

1985. Behavioural basis of intemal-wave-induced shore-
ward transport of megalopae of the crab Pachygrapsus
crassipes. Mar. Ecol. Prog. Ser. 24:289-295.

Shenker, J. M . AND B L Olla.

1986. Laboratory feeding and growth of juvenile
sablefish, Anoplopoma fimbria. Can. J. Fish. Aquat.
Sci. 43:930-937.

Sinclair, M. F.. M. J. Tremblay, and P. Bernal.

1984. El Nino events and variability in a Pacific mackerel
{Scomber japonicus ) survival index: supj)ort for Hjort's
second hypothesis. Can. J. Fish. Aquat. Sci. 42:602-
608.

Small. L. F . and D W Menzies.

1981. Patterns of primary productivity and biomass in a
coastal upwelling region. Deep-Sea Res. 28:123-149.

Smith, P E

1972. The increase in spawning biomass of northern an-
chovy, Engraulis mordax. Fish. Bull., U.S. 70:849-874.
Steiner, W W., J. J. LuzcKovicH. and B L. Olla.

1982. Activity, shelter usage, growth and recruitment of
juvenile red hake, Urophycis chuss. Mar. Ecol. Prog.
Ser 7:125-135.

Steiner, W W., and B L Olla

1985. Behavioral responses of prejuvenile red hake, Uro-
phycis chuss, to experimental thermoclines. Environ.
Biol. Fish 14:167-173.

317

WINTER-TIME DISTRIBUTION AND ABUNDANCE OF COPEPOD
NAUPLII IN THE NORTHERN GULF OF MEXICO

M J Dagg^, P B Ortner2, and F. Al-Yamani^

ABSTRACT

Copepod nauplii were collected from continental shelf waters in 3 regions of the northern Gulf of
Mexico during winters between 1981 and 1984, off Cape San Bias, Florida, off the Mississippi River
delta, and off of Galveston, Texas. Some statistically significant iP < 0.05) patterns in the abundance
and distribution of nauplii were observed: there was significant interannual variability in naupliar
concentrations within the region around the Mississippi River delta; naupliar concentrations in the
upper 10 m decreased in the onshore-offshore direction in 2 of 4 comparisons; naupliar concentrations
in the upper 10 m differed regionally in 2 of 3 comparisons; and naupliar concentration was correlated
with chlorophyll concentration in 4 of 5 comparisons.

Maximum concentrations of nauplii (number per m^) within a water column were 2-10 times
greater at stations influenced by the Mississippi River plumes than in the other 2 regions. This
condition in attributed to vertical stratification imparted to the water column by the inflowing low
salinity water from the Mississippi River. We conclude that the physical stratification provides a
mechanism for the establishment of high concentrations of nauplii that otherwise would not exist in
the winter months on the continental shelf.

Microzooplankton are important diet items for
larval fish (Arthur 1976; Gamble et al. 1981;
Checkley 1982; Govoni et al. 1983; Houde and
Lovdal 1984; Stoecker and Govoni 1984), and gut
analyses indicate that copepod nauplii are fre-
quently the dominant prey form found in the lar-
vae of many fish species (Duka and Gordina
1973). Available concentrations of micro-
zooplankton are considered an important deter-
minant of larval survival rates in the ocean be-
cause this relationship has been demonstrated in
the laboratory (Laurence 1974; Houde 1978) and
because field studies have demonstrated a rela-
tionship between regions or periods of high mi-
croplankton and high larval abundance (Arthur
1977; Lasker 1978). Because survival is enhanced
by increased food availability, oceanographic
processes that result in increased concentration
or production of prey items are important.

The gulf menhaden, Brevoortia patronus (Clu-
peiformes), supports the largest volume fishery in
the United States (U.S. Department of Commerce
1983). Spawning occurs in the wintertime, from
October to March, in continental shelf waters of
the northern Gulf of Mexcio (Fore 1970; Lewis
and Rothmayr 1981; Warlen 1988), primarily off

iLouisiana Universities Marine Consortium, Chauvin, LA
70344.

^Applied Oceanic Marine Laboratory, Ocean Coastal Divi-
sion, NOAA, 4301 Rickenbacker Causeway, Miami, FL 33149.

of Mississippi and Alabama to the east of the Mis-
sissippi River delta, and ofi" of Louisiana to the
west of the delta. During February and December
1982, concentrations of menhaden larvae in the
region near the Mississippi River delta were
greater in plume waters than outside and were
much higher at the plume front (Govoni and
Hoss"^). Furthermore, gut contents of these larvae
indicated that nauplii were the most abundant
prey items.

Much of the continental shelf water of the
northern and western Gulf of Mexico is vertically
unstratified during the winter, and winter is also
the season of minimum primary productivity in
these regions, as it is in shelf waters of the north-
east and southeast United States. However,
coastal waters and nearshore regions influenced
by freshwater inputs can be physically stratified
during winter when low salinity plumes disperse
over higher salinity shelf waters. We postulated
that shelf waters influenced by fresh water from
the Mississippi River plumes would be regions of
increased production and concentration of micro-
zooplankton. The purpose of this paper is to exam-
ine this hypothesis by describing the vertical and
horizontal distribution and abundance of copepod
nauplii in shelf waters of the northern Gulf of

Manuscript accepted December 1987.
FISHERY BULLETIN: VOL. 86, NO. 2, 1987.

^Govoni and Hoss. Unpubl data. Southeast Fisheries Center
Beaufort Laboratory, National Marine Fisheries Service,
NOAA, Beaufort, NC 28516.

319

FISHERY BULLETIN: VOL. 86, NO. 2

Mexico and by comparing these with oceano-
graphic and hydrographic indicators of water
column stratification and productivity.

METHODS

Samples were collected during 5 winter cruises
between November and March, during 1981-84,
(Table 1). Microplankton samples and hydro-
graphic data were collected from 3 general re-
gions of the northern Gulf of Mexico: off of Galve-
ston, TX, in the region of the Mississippi River
delta, and off of Cape San Bias, FL (Fig. 1). Sam-
ples were not collected from all 3 regions on each
cruise.

Water samples for chlorophyll and nutrient
analyses were collected with Niskin bottles. For
each chlorophyll analysis, between 25 and 150
mL of sea water was filtered onto a GF/F or GF/C
glass fiber filter and homogenized by grinding in
90% aqueous acetone. Fluorescence of the filtrate,
brought up to a volume of 10.0 mL, was deter-
mined before and after acidification with 2 drops
of 10% HCl using a Turner Designs'' Model 10
fluorometer. Chlorophyll and pheopigment con-
centrations as chlorophyll equivalents were de-
termined from

chlorophyll (fxg/L) =

Table 1.— Station information, 1981-84.

Kifo-

fa)

KiRfa

-fo)

pheopigment (|JLg/L)

where K is the machine calibration constant, [„
and fa are the fluorescence readings before and
after acidification, R is the acid ratio, and u is the
volume of seawater filtered, in mL (Strickland
and Parsons 1968).

Samples for nutrient analyses were frozen. Ni-
trate and nitrite were analyzed according to
method number 353.2 described by EPA publica-
tion number EPA 600/4-79-020 (Environmental
Protection Agency 1979).

Temperature and salinity were measured by
several methods. During cruise I, temperature
was measured by the temperature sensor on the
MOCNESS^ net, and salinity was measured with
a refractometer. During cruises II and III, tem-
perature was measured with expendable bathy-

Depth

Lat.

Long.

Station

Date

Time

(m)

(N)

(W)

1-1

2-13-81

0000

29 '05'

94°07'

1-2

2-13-81

1200

2906'

94' 07'

1-3

2-14-81

0001

28 04'

93"03'

1-4

2-14-81

1200

28 04'

93"03'

1-5

2-15-81

0000

180

27' 54'

92'^51 '

1-6

2-15-81

1200

180

25' 54'

92'50'

1-7

2-16-81

0001

180

28 34'

89"37'

1-8

2-17-81

1200

180

28 35'

89 38'

1-9

2-18-81

0001

2850'

89 "16'

1-10

2-18-81

1200

28 50'

89 "16'

1-11

2-19-81

0000

28^56'

89'29'

1-12

2-19-81

1200

28°56'

89'^29'

1-13

2-20-81

0000

29 09'

85' 56'

1-14

2-21-81

1200

29 08'

85"56'

1-15

2-22-81

0000

180

29°28'

8607'

1-16

2-22-81

1200

180

29°28'

86°07'

1-17

2-23-81

0001

29°36'

85°47'

1-18

2-23-81

1200

29^35'

85 47'

1-19

2-24-81

0001

30°12'

87"06'

11-1

12-05-82

0700

28°53'

89' 29'

11-2

12-05-82

1255

28°53'

89^32'

11-3

12-05-82

1945

28' 59'

89°34'

11-4

12-06-82

2120

887

28 20'

89°27'

11-5

12-07-82

0715

885

28°20'

89'27'

11-6

12-07-82

1220

834

28' 19'

89 25'

11-7

12-08-82

1820

443

28°32'

89°53'

11-8

12-09-82

0830

402

28°34'

89°53'

11-9

12-09-82

1215

461

28^32'

89 '53'

11-10

12-13-82

1245

29'26'

85°53'

11-11

12-15-82

1930

28°54'

89°29'

11-12

12-16-82

0900

28°51'

89°32'

III-1

11-19-83

1200

28°54'

89°29'

III-2

11-20-83

0800

28°47'

89°59'

III-3

11-20-83

1946

732

28°38'

89°00'

III-4

11-21-83

0555

28°54'

89°30'

III-5

1 1 -22-83

1100

28°51'

89°30'

III-6

1 1 -23-83

0500

194

28°31'

89'=37'

III-7

1 1 -24-83

1030

28=50'

89°31'

III-8

11-25-83

0820

28^55'

89°34'

III-9

11-25-83

1945

28°56'

89°36'

111-10

11-26-83

1200

28°54'

89°29'

111-11

11-27-83

0900

29'02'

89°30'

111-12

11-28-83

0840

28^48'

89^57'

111-13

11-28-83

1250

28°56'

89°51'

111-14

11-29-83

0950

29''03'

89°39'

111-15

1 1 -30-83

0930

28°47'

89°58'

111-16

12-01-83

0800

28°52'

89°29'

IV- 1

3-14-83

27' 58'

95°53'

IV-2

3-19-83

28=18'

90^41 '

V-1

2-21-84

0850

28 54'

90°25'

V-2

2-22-84

0830

28' '54'

90°25'

V-3

2-23-84

0910

28°54'

90°25'

V-4

3-1-84

0835

28°54'

90°25'

V-5

3-1-84

0830

28^54'

90°25'

V-6

3-3-84

0730

28'54'

90°25'

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

^Multiple opening-closing net and environmental sensing
system.

thermographs (Sippican Instruments). Salinity
was measured with a YSI Model 33-S-C-T sali-
nometer on cruise II, and with a Beckman Model
RS5-3 salinometer on cruise III. For both of these
cruises, bottle samples, analyzed in the labora-
tory with a Guildline Model 8400A Autosal, were
used to check the shipboard salinity measure-

320

DAGG ET AL.: DISTRIBUTION OF COPEPOD NAUPLII

c
o

C/2

D
O

321

FISHERY BULLETIN: VOL, 86, NO. 2

ments. On cruise IV, a Plessey Model 9040 CTD
was used for both temperature and salinity. On
cruise V, salinity samples were stored in bottles
and analyzed in the laboratory using a Guildline
Model 8400A Autosal, and temperature was
measured with reversing thermometers.

Microzooplankton were collected using 5 L
(cruises I, II, and V) or 30 L (cruises III and IV)
Niskin bottles. During cruise I, 1 L samples of
untreated water were collected from 3 depths and
preserved for later analysis. During subsequent
cruises, samples were collected from up to 8
depths; four liters were poured gently through a
20 |xm sieve, backwashed with filtered seawater
into a sample jar, and preserved in a 5% formalin-
seawater solution. In the laboratory, all organ-
isms in each preserved sample were identified
and counted. In selected samples, the length and
width of the first 50 copepod nauplii were mea-
sured using an ocular micrometer. In addition,
during cruise I larger nauplii were collected with
0.1 m^ 64 |jLm mesh nets nested inside the 1 m^
333 ^JLm mesh nets of a MOCNESS net (Wiebe et
al. 1976). Estimates of naupliar abundance from
these collections were used separately in the
analyses for patterns in distribution and abun-
dance. Lastly, from cruise I, the total number of
copepods in the shallow-water samples, total
number of copepods from the 333 jxm nets, and the
zooplankton displacement volume from the 333
(xm nets were analyzed for patterns in distribu-
tion and abundance.

RESULTS

Samples from the Mississippi River delta re-
gion, collected during 1981 (cruise I), 1982 (cruise
II), and 1983 (cruise III), were compared to deter-
mine if there was significant interannual vari-
ability within a region. Whole water samples and
samples retained on 20 ixm mesh sieves were
pooled for this analysis under the assumption
that nauplii in a whole water sample would be
retained on a 20 jjim mesh screen. The 64 [im net
samples were excluded from this analysis. This
test was made using average naupliar concentra-
tions in the upper 10 m only; naupliar concentra-
tions were typically lower in deeper waters, and
we did not have the same number of deep and
shallow stations for each of the 3 sampling years.
A 1-way analysis of variance indicated significant
variation between years within the Mississippi
River delta region {P <0.01). All 3 years were
statistically different. These results indicated to

us that regional or other spatial comparisons can
only be made using samples collected during the
same year. Interannual comparisons for other re-
gions were not possible.

Samples were collected from all 3 regions dur-
ing cruise I. Four tests for regional differences
during this cruise were made. First, regions were
compared based on naupliar concentrations in the
upper 10 m only (Table 2); based on whole water
collections F was significant at P < 0.05 and
based on samples collected with 64 \xm mesh nets
F was significant at P < 0.09. Second, regions
were compared based on naupliar concentrations
from all depths (Table 2); based on whole water
collections F was significant at P < 0.01 and
based on samples collected with 64 (xm mesh nets
F was significant at P < 0.22. Thus, whole water
samples collected on cruise I indicated significant
regional differences in naupliar concentration
but samples collected with 64 fxm mesh nets did
not. In addition, there were significant regional
differences in the concentrations of total copepods
(P < 0.01) and in zooplankton displacement vol-
ume (P < 0.05) during cruise I (data not shown).
Regional comparisons were not possible during
other cruises because all 3 regions were not sam-
pled in years other than 1981.

Where possible, we also tested for onshore-
offshore gradients in naupliar concentrations in
the upper 10 m. Stations were categorized accord-
ing to bottom depth as shallow, <18 m, interme-
diate, 18-90 m, and deep, >90 m, and compari-
sons were made between these 3 depth categories.
A highly significant onshore-offshore difference
existed during cruise I for nauplii collected with
the 64 |xm MOCNESS nets (Table 3). During
cruise III samples were only collected in the Mis-
sissippi River delta region but onshore-offshore
differences within this region were significant
(Table 3). No significant onshore-offshore gradi-
ents were observed in naupliar concentrations de-
termined from whole water samples collected
from all 3 regions during cruise I, or in nauplii
from only the Mississippi River delta region dur-
ing cruise II (Table 3).

Although regional differences in average con-
centrations of nauplii were not strong, the verti-
cal distributions were different. For example, sev-
eral stations at approximately the 50 m isobath
are compared in Figure 2. At station IV- 1 (Fig.
2a) ofT the coast of Texas, there was no marked
vertical heterogeneity in naupliar concentra-
tions, the maximum concentration was low (23
nauplii/L), and the average concentration was

322

DAGG ET AL.: DISTRIBUTION OF COPEPOD NAUPLII

Table 2. — Regional comparisons of naupliar concentrations during
cruise I. Samples were collected by 2 methods: whole water sam-
ples were collected with Niskin bottles, and 64 (i.m mesh nets were
used to collect the net caught samples.

Cruise

Depths

Sample
type

Probability
of F

upper 10 m

whole water

<0.05

upper 10 m

64 ^m nets

126

<0.09

all depths

whole water

<0.01

all depths

64 |jLm nets

222

<0.22

Table 3. — Onshore-offshore comparisons of naupliar concentra-
tions from the upper 10 m of each station. Tests for cruise I are
based on naupliar concentrations from all 3 regions pooled then
separated into 3 depth categories. During cruises II and III. nauplii
were only collected from the Mississippi River delta region.

Cruise

Regions

Sample
type

Probability
of F

all

whole water

<0.14

all

64 ^m nets

126

<0.01

Mississippi R.

20 jjim sieve

<0.25

Mississippi R.

20 M-m sieve

<0.01

NAUPLIUS CONCENTRATION (number/ liter)

UJ
Q

I
»-
Q.
UJ
Q

25 50 75
1, . 1 1

10-

^ /

If 1

1
1

30-

50-

112

Figure 2. — Naupliar concentrations (number/liter) from selected stations near the 50 m isobath on the Texas continental
shelf (a, station IV-1), the Louisiana Shelf (b, station IV-2, c, stations II-l, II-2, II-3), and the northwest Florida shelf (d,
station 11-10).

323

FISHERY BULLETIN: VOL. 86, NO. 2

low (11 nauplii/L). The same pattern was ob-
served off Cape San Bias (Fig. 2d); there was no
marked vertical heterogeneity, and the maxi-
mum and average concentrations were low, 14
and 12 nauplii/L. In contrast, nauplii at stations
near the 50 m isobath in the region near the Mis-
sissippi River delta typically showed marked ver-
tical heterogeneity in abundance. At station IV-2
(Fig. 2b), the maximum concentration was 187
nauplii/L and the minimum was 24 nauplii/L. At
station II- 1, II-2, and II-3 in the Mississippi River
delta region (Fig. 2c) naupliar abundances were
also vertically heterogenous, although concentra-

tions were lower than at Station IV-2. In general,
based on all our vertical profiles (most not
shown), regions influenced by the Mississippi
River can contain high concentrations of nauplii
in the surface layer, while subsurface concentra-
tions are similar to concentrations in the other 2
regions.

Vertical structure in the distribution of nauplii
appears to be related to physical structure. At the
stations off the coasts of Texas and Florida (Fig.
2a, 2d), temperture and salinity were essentially
vertically homogenous (Fig. 3a, 3d) whereas off
the Mississippi River delta temperature and

20%. 24 28

32 36

24%. 28
16*0 18

Figure 3. — The vertical distribution of temperature (•) and salinity (x) at the stations represented in Figure 2.

324

DAGG ET AL.: DISTRIBUTION OF COPEPOD NAUPLII

salinity were vertically stratified (Fig. 3b, 3c).
The same pattern was seen in the chlorophyll a
distributions (Fig. 4) and the nitrate distributions
(Fig. 5).

Correspondence between chlorophyll a concen-
tration and naupliar abundance was computed;
the correlation coefficient for a linear regression
of chlorophyll concentration on naupliar abun-
dance for all 5 cruises in = 269) was 0.39, signifi-
cant at the 1% level. However, analysis of covari-
ance showed that the relationship between

chlorophyll and naupliar concentration was sig-
nificantly different between cruises, and there-
fore the data from the 5 cruises should not be
pooled. Although a significant relationship be-
tween chlorophyll concentration and naupliar
concentration typically exists, it is not consistent
from cruise to cruise. Regression parameters and
correlation coefficients for each cruise are shown
in Table 4. Pairwise tests indicate that the slope
of the regression for cruise V was significantly
different from all the others, cruise IV was differ-

0 I 0

• CHLOROPHYLL a (ug/liter)

2.0 3 0 4.0 10 2.0

Figure 4. — The vertical distribution of chlorophyll a at the stations represented in Figure 2.

NITRATE (ug-at/liter)

Figure 5. — The vertical distribution of nitrate at the stations represented in Figure 2.

325

FISHERY BULLETIN: VOL. 86, NO. 2

Table 4. — Parameters and coefficients for linear regressions of chlorophyll
concentration on nauplius concentration for each cruise. Y = a + b(X) where
y = naupliar concentration as number per liter, and X = chlorophyll concentra-
tion as fjig per liter. Significance level tested is 0.01.

Significance

Cruise

Intercept

Slope

of slope

/-2

correlation

12.9

9.7

<0.01

0.40

<0.01

5.4

15.4

-0.01

0.53

<0.01

III

3.7

9.2

<0.01

0.34

<0.01

6.9

39.4

<0.01

0.94

<0.01

26.1

1.1

0.04

ent from all others, cruise II was different from
all the others, and cruises I and III were not dif-
ferent from each other but different from cruises
II, IV, and V. The slopes of all relationships ex-
cept cruise V were significantly different from
zero.

The size-frequency distribution of nauplii did
not vary in any systematic manner (Fig. 6). For
example, at station IV-1, the bulk of the nauplii
were between 20 and 100 \x.m at all depths sam-
pled during both the daytime and nighttime peri-
ods. The size-frequency distribution was essen-
tially identical at station IV-2, off Louisiana (data
not shown).

At some stations, copepodid stages <600 ixm in
length (50-200 \xm in width) were counted in ad-
dition to nauplii because they are potential prey
items for menhaden larvae. Copepodites were
usually not as abundant as nauplii, especially at

stations with high naupliar abundances. On occa-
sion, they were as abundant or, at specific depths,
more abundant than the nauplii (Table 5). The
ratio of nauplii to copepodites varied widely, be-
tween 16.3 and 0.3, so it is not possible to assign
a constant factor to naupliar abundances to esti-
mate the increase in available prey attributable
to copepodid stages.

Other microzooplankton were usually not as
abundant as nauplii or copepods. In our samples,
various forms of eggs reached a maximum density
of 6 eggs/L (cruise III station 6 at 3 m, data not
shown). Pelecypod larvae at one station (III-ll)
were abundant, reaching a maximum of 29 lar-
vae/L at 4 m, compared with 22 nauplii, 7 cope-
podites, and 2 eggs/L in the same sample. Only on
this one occasion were organisms other than cope-
podites or nauplii the dominant form of micro-
zooplankton.

Table 5. — Abundance (number/liter) of other microzooplankton in addition to nauplii at

selected stations during cruise III.

Station

Depth
(m)

Nauplii

Cope-

podids

<600^m

Cope-

podids

>600 ii.vn

Larva-
ceans

Eggs

Other

III-3

17.3

7.5

1.0

7.8

0.5

1.8

29.3

12.0

0.5

11.0

1.8

17.5

8.0

1.0

6.8

1.5

2.3

7.8

2.8

1 0

2.0

0.5

III-7

24.0

6.3

4.0

1.0

15.3

12.3

1.0

0.3

1.3

3.0

5.0

1.0

0.8

0.3

4.8

5.8

0.3

0.5

2.0

1.0

0.3

111-14

15.8

12.0

1.3

2.0

44.8

3.8

0.3

0.5

3.0

4.0

40.0

6.8

0.8

1.3

4.3

0.3

11.5

8.0

3.8

1.0

1.8

0.3

12.0

35.0

6.8

0.3

0.8

111-16

15.8

25.3

5.0

12.8

8.0

16.0

2.3

1.0

2.5

30.0

14.0

2.0

0.3

2.5

29.0

13.3

0.5

0.8

1.3

17.0

12.5

1.0

0.8

0.5

31.8

12.8

0.8

5.0

326

DAGG ET AL.: DISTRIBUTION OF COPEPOD NAUPLII

NIGHT

20-

^i=H

-T 1

DAY

0 m 20

-1 1 1 I

20-

0^— -^
20

-I 1 1 1

^ 0

-I 1 1

20-1

13
C

LlI

:z.
o
o

_J
CL

0
20-

0
20

0
20n

r~| .

somple lost

-1 1 1 1

20-1

-I 1 r

20-

:i-^

X4-

lOm

15m

20 m

25m

30m

20n

20-

0
20-

^aXL

5=L

35m

-I ' 1

-T r ■!

1 1 I

o o o o o o o

^ CD ^ (I> O ^

BODY WIDTH (mm)

40m

=F=L

o o o o o o o

•3- CD oj i£ o ^

BODY WIDTH (mm)

45 m

Figure 6.— The size-frequency distribution
of nauplii from station IV- 1.

327

FISHERY BULLETIN: VOL. 86, NO. 2

DISCUSSION

Menhaden larvae between 3 and 20 mm long
collected during wintertime from 3 regions in the
northern Gulf of Mexico contained a variety of
diet items, including dinoflagellates, tintinnids,
copepod eggs, nauplii, copepodids, juvenile pele-
cypods, and pteropods (Govoni et al. 1983, 1985).
Dinoflagellates and tintinnids constituted the
main diet items of larvae <5.0 mm long but were
replaced by copepod nauplii, then copepodids and
small adult copepods as larvae grew larger.

In another study, Stoecker and Govoni (1984)
found that copepod nauplii were the dominant
items in the diet of menhaden larvae 7.0-9.0 mm
long, although only 12 fish were examined. Later
studies, however, have verified that nauplii are
the dominant items in the diet of a wide size
range of larval menhaden collected in the
boundary between the plume of the Mississippi
River and oceanic waters (Govoni and Hoss fn. 3)

The maximum width of food items in menhaden
larvae <5 mm long was about 200 |xm (Govoni et
al. 1983). This width increased up to about 400
(xm for larvae 10 mm long. In another study
(Stoecker and Govoni 1984) maximum prey size
was calculated in a different manner; the average
width of the largest prey type was used. With this
index the estimated maximum width of prey
items was 50 ^JLm for larvae <5 mm long and
about 140 |jLm for larvae 9-11 mm long. In our
study, most of the nauplii were between 40 and 80
|jLm body width, and most of the copepodites were
between 50 and 200 p-m body width. These cope-
pod developmental stages were in the size range
of diet items typically found in guts of intermedi-
ate size (7-11 mm) larvae of the gulf menhaden
and we believe our samples are reasonably repre-
sentative of the prey concentrations available to
these larvae. Other prey items for larvae in this
size range were common only on rare occasions.

The importance of high prey concentrations to
successful feeding and survival offish larvae has
frequently been noted (Hunter 1981). Concentra-
tions required to give high survival in laboratory
experiments are seldom found in the ocean but
careful attention to culture techniques can result
in reasonably high survival at prey densities that
are close to or overlap the maximum natural con-
centrations (Houde 1978). At stations near the
river delta that are strongly affected by the river
plumes, maximum concentrations of nauplii were
typically in the range of 20-50 nauplii/L. At sta-
tions farther down plume, away from the delta.

maximum naupliar concentrations were higher,
up to 187 nauplii/L at station IV-2 for example. It
is not clear whether this is a seasonal pattern, a
pattern due to the down-plume development of
the food web, or a pattern attributable to some
other factors. However, the concentrations ob-
served in waters with salinities lowered by the
river plumes are mostly within the range of 10-
100 nauplii/L that is frequently reported for
oceanic and coastal waters for microzooplankton,
or the 50-100 nauplii/L range for coastal and es-
tuarine areas (see Houde 1978; Hunter 1981 for
summaries). The addition of copepodite stages
and other potential prey items to the naupliar
abundances in this study would increase the esti-
mates of available prey somewhat but usually not
more than 20%. At least during the wintertime,
the waters surrounding the Mississippi River
delta do not appear to contain exceptionally high
concentrations of copepod nauplii, compared to
other coastal and estuarine areas.

The vertical distribution of nauplii was fre-
quently similar to that of chlorophyll. Regres-
sions of naupliar concentration on chlorophyll
concentration were significant for 4 of the 5
cruises although not the same in each case. As-
suming chlorophyll is a reasonable indicator of
phytoplankton abundance and thus food abun-
dance for nauplii, then nauplii appear to be ag-
gregated at the depth of highest food availability.
Year to year variability in the relationship indi-
cates that other factors are also important in de-
termining the abundance and distribution of nau-
plii. Because eggs released from adult female
copepods would sink out of the surface water be-
fore hatching, it is probable that active swimming
by nauplii plays a part in the aggregation process,
perhaps enchanced by physical convergence proc-
esses.

In our study we noted that there were interan-
nual differences in naupliar concentrations, there
were sometimes significant regional differences,
and there were sometimes significant onshore-
offshore differences. Although we anticipated
finding higher concentrations of nauplii in the
region near the Mississippi River delta, this pat-
tern was not always observed. This finding might
indicate that there is not an important difference
between the 3 regions as far as larval food
availability is concerned. Alternatively, it might
indicate that there are subregions within the
larger Mississippi River delta region that contain
a food environment for fish larvae that allows for
enhanced survival and growi;h, but that we failed

328

DAGG ET AL.: DISTRIBUTION OF COPEPOD NAUPLII

to clearly identify and segregate these regions.
The overall region did not always appear signifi-
cantly better than the other regions because only
parts of it are better.

On the continental shelf in the Gulf of Mexico,
as in other shelf regions of the United States, the
typical pattern is for the water to be vertically
well-mixed during the wintertime (Parker 1968).
Autumn cooling breaks down the thermal strati-
fication that has existed throughout the summer
and allows the isothermal water column to be
easily mixed. The major exception to this general
pattern in the Gulf of Mexico is the shelf region
influenced by the large volume of freshwater
runoff from the Mississippi River. During this pe-
riod, the freshwater influx is of sufficiently large
volume (average flow for December to March in
1975 through 1979 was 17,290 m^/second (U.S.
Army Engineer District, New Orleans Corps of
Engineers, 1980)) to physically stratify the shelf
waters hundreds of km downstream from the
delta. We suggest that this salinity-induced strat-
ification is a vital component of the recruitment
success of Gulf menhaden because it provides an
environment in which prey aggregations can
occur. It has been suggested that vertical stratifi-
cation by other small-scale physical phenomena
(e.g., Langmuir circulation) allows significant
patchiness of prey items to exist (Lasker 1975).
This patchiness provides small regions of compar-
atively high food concentrations for fish larvae,
and results in improved feeding success. In this
study, copepod nauplii were aggregated vertically
at stations with physical stratification and were
nearly homogenously distributed at stations lack-
ing physical stratification. Maximum naupliar
concentrations (no. per m^) at stratified stations
were typically 2-10 times greater than at non-
stratified stations.

In conclusion, we believe that the large fresh-
water inflow of the Mississippi River during the
wintertime spawning period of the Gulf men-
haden contributes to the feeding success and
survival of larval fish by providing physical strat-
ification which in turn results in a vertical strat-
ification of phytoplankton and microzooplankton
in layers or patches of relatively high concentra-
tions.

ACKNOWLEDGMENTS

Part of this work was done in cooperation with
personnel from the National Marine Fisheries
Service at Beaufort, NC and at Pascagoula, MS.

In particular, we would like to thank D. Hoss for
providing time and space in his field programs,
and J. Govoni for discussion of several portions
of the manuscript. S. Cummings, L. Hill,
E. Walser, J. White, J. Turner, P. Morgan,
C. Neill, and E. Turner assisted in collecting and
analyzing samples. Part of this work was done
in cooperation with D. Checkley and A. Amos at
the University of Texas; their contributions are
greatly appreciated. Thanks also to the captains
and crews of the F.R.V. Oregon II, the RV
Longhorn, and the R. J. Russell. This work was
supported by the Louisiana Universities Marine
Consortium, by NSF Grant OCE-81 19848, and by
the Ocean Assessment Division, National Ocean
Services, NOAA.

LITERATURE CITED

Arthur, D. K

1976. Food and feeding of larvae of three fishes occurring
in the California Current, Sardinops sagas, Engraulis
mordax, and Trachurus symmetricus. Fish. Bull., U.S.
74:517-530.

1977. Distributions, size, and abundance of microcopte-
pods in the California Current system and their possible
influence on survival of marine teleost larvae. Fish.
Bull., U.S. 75:601-611.

Checkley, D M, Jr.

1982. Selective feeding by Atlantic herring (Clupea
harengus ) larvae on zooplankton in natural assem-
blages. Mar. Ecol. Prog. Ser. 9:245-253.

DUKA, I A , AND A. D GORDINA.

1973. Abundance of ichthyoplankton and feeding of fish
larvae in the western Mediterranean and adjacent areas
of the Atlantic Ocean. Hydrobiol. J. 9:54-59.
Environmental Protection Agency.

1979. Methods of chemical analysis of water and wastes.
Nitrogen, nitrate-nitrate. U.S. Environ. Prot. Agency,
Cincinnati, OH 45268.
Fore, P L

1970. Oceanic distribution of eggs and larvae of the Gulf
menhaden. Rep. Bur. Comm. Fish. Biol. Lab., Beaufort,
NC, Fish. Wildl. Serv. Circ. 341.
Gamble, J C , P Maclachlan, N T Nicholl. and I. G. Baxter.
1981. Growth and feeding of Atlantic herring larvae
reared in large plastic enclosures. Rapp. P. -v. R6un.
Cons. int. Explor. Mer 178:121-134.
Govoni, J J , A J Chester. D E Hoss, and P B Ortner

1985. An observation of episodic feeding and growth of
larvae Leiostomus xanthrus in the northern Gulf of Mex-
ico. J. Plankton Res. 7:137-146.
Govoni, J. J , D E Hoss, and A. J Chester

1983. Comparative feeding of three species of larval fishes
in the northern Gulf of Mexico: Brevoortia patronus,
Leiostomus xanthrus, and Micropogonias undula-
tus. Mar. Ecol. Prog. Ser. 13:189-199.

Houde, E D

1978. Critical food concentrations for larvae of three spe-
cies of subtropical marine fishes. Bull. Mar. Sci.
28:395-411.

329

FISHKRY BULLKTIN: VOL. 86, NO 2

HOUDE. E D . AND J. A LOVDAL.

1984. Seasonality of occurrence, foods and food prefer-
ences of ichthyoplankton in Biscayne Bay, Florida.
Estuarine Coastal Shelf Sci. 18:403-419.
Hunter. J R

1981. Feeding ecology and predation of marine fish lar-
vae. In R. Lasker (editor). Marine fish larvae.
Morphology, ecology, and relation to fisheries, p. 34-
77. Univ. Wash. Press, Seattle.
Laurence, G C

1974. Growth and survival of haddock (Melanogrammus
aeglefinus) larvae in relation to planktonic prey concen-
tration. J. Fish, Res. Board Can. 31:1414-1419.

Lasker, R.

1975. Field criteria for survival of anchovy larvae: The
relation between inshore chlorophyll maximum layers
and successful first feeding. Fish. Bull., U.S. 73:453-
462.

1978. The relation between oceanographic conditions and
larval anchovy food in the California Current: Identifica-
tion of factors contributing to recruitment failure.
Rapp. P.-v. Reun. Cons. int. Explor. Mer 173:212-230.
Lewis, R. M., and C M Rothmayr

1981. Spawning and sexual maturity of gulf menhaden,
Brevoortia patronus. Fish. Bull., U.S. 78:947-951.

Parker. C A

1968, The effect of a cold air outbreak on the continental
shelf waters of the northwestern Gulf of Mexico. Tech.
Rep. 68-3T, 89 p. Off. Nav. Res.
Stoecker. D K , and J J Govoni.

1984. Food selection by young larval gulf menhaden (Bre-
voortia patronus). Mar. Biol. 80:299-306.
Strickland, J D H , and T R. Parsons.

1968. A practical handbook of seawater analysis. Bull.
Fish. Res. Board Can. 167, p. 201-203.
US Department of Commerce

1983. Fisheries Statistics for the United States, 1982.
US Army Engineer District, New Orleans Corps of Engi-
neers.

1980, Stages and discharges of the Mississippi River and
its tributaries and other watersheds in the New Orleans
District, 1975-1979.
Warlen, S. M.

1988. Age and growth of larval menhaden, Brevoortia pa-
tronus, in the northern Gulf of Mexico. Fish. Bull., U.S.
86:77-90.
WiEBE, P H , K. H Burt, S H Boyd, and A W Morton.

1976. A multiple opening/closing net and environmental
sensing system for sampling zooplankton. J. Mar. Res.
34:313-326.

330

THE EFFECTS OF SILTATION ON RECRUITMENT OF SPINY LOBSTERS,

PANULIRUS ARGUS

William F Herrnkind.' Mark J. Butler IV,' and
Richard A. Tankersley^

ABSTRACT

Several surveys in the Florida Keys indicated fewer juvenile spiny lobsters, Panulirus argus, in an
area where their primary habitat, stands of benthic algae Laurencia spp., was heavily silted as
compared with similar, less silted habitat. We tested several hypotheses explaining this relationship:
1) planktonic postlarval lobster abundances are lower in the silted area, 2) siltation of algae impedes
postlarval settlement or subsequent juvenile habitat selection, or 3) siltation increases mortality at
the time of metamorphosis. We also compared the time-to-metamorphosis for settling pueruli within
silted and nonsilted algae, analyzed the physical character of algal silt in low-silt and high-silt
regions and measured the abundances of epifauna constituting prey of juvenile spiny lobsters. Plank-
tonic postlarval abundances were substantially higher in the high-silt area thus rejecting hypothe-
sis 1. Likewise, results from laboratory experiments testing the effect of algal siltation on postlarval
time-to-metamorphosis and early postsettlement survival showed no short-term increase in mortal-
ity. Limited postlarval settlement and avoidance of silted algal habitats by juveniles, as determined
in substrate choice experiments, probably accounts for the paucity of young spiny lobsters in heavily
silted localities. In addition, although juvenile spiny lobsters are nonselective predators, lower prey
availability in silted algae probably promotes transciency which, in turn, causes increased mortality
by predation while juveniles are exposed. Large-scale siltation exacerbated by human activity must
be viewed as potentially deleterious to spiny lobster recruitment.

The western Atlantic or Florida spiny lobster,
Panulirus argus, is the focus of an intense com-
mercial and recreational fishery in south Florida,
particularly the Florida Keys. Besides severe
fishing pressure, spiny lobster populations are
subject to a variety of other factors that poten-
tially limit population size. For example, habitat
degradation, like that resulting from chronic sil-
tation, may affect not only adult lobsters but the
postlarval settlement stage as well. During 1983
and 1984 we sampled numerous sites in a region
of about 40 km^ east of Big Pine Key which was
chronically heavily silted and held low numbers
of newly settled spiny lobsters despite extensive
benthic algal growth typical of settlement habi-
tat. We hypothesized that postlarval spiny lob-
sters either do not settle in silted habitat or settle
there but do not survive. In either case, we sup-
posed that the heavy siltation reduced the carry-
ing capacity of otherwise suitable habitat, poten-
tially reducing regional recruitment where
siltation is widespread.

•Department of Biological Science, The Florida State Univer-
sity, Tallahassee, FL 32306-3050.

^Department of Biological Science, The Florida State Univer-
sity, Tallahassee, FL; present address: Department of Biology,
Wake Forest University, Winston-Salem, NC 27109.

Sediment particle size, composition, and stabil-
ity influence larval settlement in a variety of
marine benthic invertebrates (Crisp 1974, 1976;
Gray 1974; Rhoads 1974; Pearson and Rosenberg
1978). For example, heavy siltation and sediment
instability, created by natural biogenic rework-
ing of the substrate and subsequent resuspension
of sediments by turbulence reduces the abun-
dance of suspension feeding infauna (Rhoads and
Young 1971; Aller and Dodge 1974). In addition,
siltation from human activities (e.g., dredging,
shoreline development, boat traffic, etc.) can de-
grade benthic community structure via anaero-
biosis, direct burial, toxic poisoning, or increased
turbidity (Morton 1977; Allen and Hardy 1980;
Jones and Candy 1981; Cortes and Risk 1985).
Most available information concerns sessile or in-
faunal species, but little information exists for
mobile, epibenthic forms (Pearson and Rosenberg
1978). Although the habitat selection and bur-
rowing behavior of some shrimps and juvenile
clawed-lobsters has been investigated in relation
to substrate character (Ruello 1973; Howard and
Bennett 1979; Aziz and Greenwood 1982; Botero
and Atema 1982; Pottle and Elner 1982; Roach
1983; Herrnkind and Butler 1986), we know of no
research describing the effect of siltation on deca-

Manuscnpt accepted February 1988.
FISHERY BULLETIN: VOL. 86, NO 2, 1988.

331

FISHERY BULLKTIN: VOL. 86, NO. 2

pod recruitment. Our research focuses on the im-
pact of siltation on spiny lobster postlarvae and
early benthic juveniles, stages that are morpho-
logically and behaviorally distinct from adults.

Late stage P. argus phyllosome larvae drift in
the oceanic plankton for 6-9 months after hatch-
ing and metamorphose offshore into nonfeeding
pueruli (postlarvae) that swim inshore and settle
in benthic vegetation (Marx 1986). Nev^ly settled
pueruli metamorphose into cryptically colored
benthic juvenile instars after about one week.
Pueruli preferentially settle in highly architec-
tured benthic algal assemblages where subse-
quent survival and growth depend upon available
prey and physical refuge from predators
(Herrnkind and Butler 1986). Ubiquitous, widely
distributed stands of bushy red algae, Laurencia
spp., provide these essential conditions and prob-
ably serve as the most important regional settle-
ment and nursery habitat for juvenile spiny lob-
sters (Marx and Herrnkind 1985a, b; Herrnkind
and Butler 1986; Marx 1986). The early instars
remain within the algae for several months until
attaining about 20 mm carapace length (CL)
(Andree 1981; Marx and Herrnkind 1985a) when
they begin to occupy crevices in rubble or under
sponges, coral, and exposed seagrass rhizome
mats. In Florida, postlarval settlement is year-
round with vernal, autumnal, newmoon, and oc-
casionally aperiodic peaks (Little 1977; Little and
Milano 1980; Marx 1986). The spatial pattern of
settlement is poorly known although new recruits
are widely dispersed within algal habitats; diver
surveys have yielded estimates of one juvenile per
36 m^ of profuse algal growth (Marx and Herrn-
kind 1985a). Yet because postlarvae settle con-
tinuously and juveniles grow rapidly, a single
hectare of the above habitat is estimated to nur-
ture about 1,000 spiny lobsters annually (Marx
and Herrnkind 1985a; Marx 1986). There is no
compelling evidence suggesting that benthic
stage lobsters immigrate into Florida waters from
other Caribbean areas, although their planktonic
larvae presumably do so (Lyons 1980; Marx
1986). Recruitment is thus primarily limited to
postlarval influx. Therefore, precise knowledge of
the factors influencing postlarval recruitment
and recruit mortality is essential to managing the
intensive Florida spiny lobster fishery.

Here we report on studies undertaken to inves-
tigate the impact of the observed algal siltation
on spiny lobster recruitment. We compared field
abundances of both pueruli and early juveniles in
a representative silted and unsilted area, exam-

ined the relationship between siltation and avail-
able epifaunal prey, determined the impact of silt
load on puerulus survival from settlement
through metamorphosis, and tested the prefer-
ence of settling pueruli and algal-dwelling juve-
niles for silted and unsilted algae.

METHODS

Postlarval-Juvenile Abundance in
Silted and Unsilted Habitats

During June through August 1985, we com-
pared the natural abundance of newly settled ju-
venile spiny lobsters (6-20 mm CL) in previously
sampled, chronically silted and unsilted areas.
The silted site (No Name Key) was located ap-
proximately 30 m off the western shore of No
Name Key (Monroe County, FL, U.S.A.) and the
unsilted site (Burnt Point) 30 m off the northwest
shore of Grassy Key at Burnt Point (Fig. 1). The
benthic habitat at both sites was similar and
characterized by nearly contiguous stands of
algae (Laurencia spp.) at depths of 1.5-3.0 m. In-
tensive visual search in algal clumps by divers
was used to estimate the relative number of ben-
thic juveniles. Because newly settled spiny lob-
sters are almost exclusively found associated with
algal clumps (Marx and Herrnkind 1985a), catch
per unit effort (CPUE) as search time within
algae, gives a more suitable estimate for our pur-
poses of comparing abundance than density/area
per se. Modified Witham-type postlarval collec-
tors (Witham et al. 1964, 1968; Little and Milano
1980; Marx and Herrnkind 1985a) were used to
compare postlarval abundance among sites.
Twelve collectors were initially deployed at both
sites and visited approximately every 2 weeks for
3 months. Collector results are reported in CPUE
to standardize catch records biased by the loss of
collectors and different sampling durations.

Algal Silt Content and Prey Content

To determine the amount of silt and macro-
fauna contained in algal clumps at the silted and
unsilted sites, we bagged individual clumps (ap-
proximately 25 cm diameter) of Laurencia in the
field for subsequent laboratory processing. Care
was taken to ensure that loose silt present on the
surface of the algae was not disturbed during col-
lection. Ten clumps, ranging in displacement vol-
ume from 55 to 300 mL, were collected at each
site. Algal samples were rinsed through a series

332

HERRNKIND ET AL RECRUITMENT OK Sl'lNY LOBSTERS

Florida Keys ^

-r«*.

^B''

LITTLE PINE
KEY

NAME
KEY
r BAHIA
' HONDA
KEY

^NEWFOUND
HARBOR KEYS

BURNT POINT

GRASSY
KEY

Figure l. — Map of field sites in the middle and lower Keys, Monroe County, FL, U.S.A. Insets provide more detail of the areas
surrounding our field sites (•). Diagonal lines depict approximate extent of the heavily silted area we surveyed.

of sieves (500 |x, 250 |jl, and 63 jx; U.S. Standard
Sieve Series), but only the two smallest size frac-
tions were retained because subsamples >500 |x
consisted entirely of shell and algae fragments.
Silt samples were dried for 48 hours at lOO'C and
then weighed. The amount of algal-entrained silt
at the two sites was compared using a two-sample
^-test. Organic weight of the silt was derived by
digesting three silt samples in SO^f hydrogen per-
oxide for 1 week, then oven drying the remaining
silt at 60°C for 48 hours (Cortes and Risk 1985).
The fraction of carbonates in the silts was deter-
mined by dissolving the three samples in 5% hy-

drochloric acid for 1 week, then drying the sam-
ples as above (Cortes and Risk 1985). Percent
organics and carbonates (by weight) in the silts at
the two sites were compared in two-sample ^-tests
on arcsin transformed data.

We counted the number of epifaunal prey in
silted and unsilted clumps to determine the possi-
ble influence of siltation on juvenile spiny lobster
food abundance. The reported estimates of prey
abundance are means of two separate counts per
clump; 5 clumps per treatment were processed.
The volume of each Laurencia clump was deter-
mined by water displacement, and all silt load

333

FISHERY BULLETIN; VOL. 86, NO. 2

and prey abundance estimates standardized by
clump volume. Prey abundance data were ana-
lyzed using a two-way fixed-effects ANOVA on
log transformed data and Bonferroni pairwise
multiple comparisons.

Habitat Selection/Settling Experiments

We tested postlarval settlement and juvenile
habitat selection in laboratory experiments using
clumps of Laurencia spp. with high- and low-silt
loads (referred to hereafter as silted and un-
silted); the null hypothesis being equal selection
of both habitats. Experiments were conducted in
fourteen 75.7 L aquaria with subgravel filters
and circulating current of 3 cm s ^ Light was
provided by skylights and fluorescent lights with
a photoperiod of approximately 14L:10D. Two 20
cm diameter algal clumps, one silted and one un-
silted, were situated 25 cm apart at opposite ends
of each aquarium and at least 5 cm from aquar-
ium walls. The number of natural prey in both
silted and unsilted clumps far exceeded the num-
ber eaten daily by a juvenile. To further control
food availability in experiments with juveniles
we added equal amounts (10 mg) of Tetramin^
fish food to each clump, providing an overabun-
dance of food available ad libitum. If juveniles
chose one type of algal clump over the other, then
their selection was most likely based on the pres-
ence or absence of silt, because food abundance
and quality were similar, if not strictly identical,
in both types of algal clumps. Pueruli neither feed
nor respond to the differential abundance of po-
tential prey (Herrnkind and Butler 1986). Silted
algae was collected from the No Name Key site
(see section on Algal Silt Content and Prey Con-
tent); unsilted algae was collected just offshore of
the Sea World Marine Science and Conservation
Center on Long Key. Fresh algal clumps were
used in each experimental replicate. An experi-
ment was initiated by introducing a single
puerulus or juvenile spiny lobster to the center of
an aquarium through a 5 cm diameter PVC pipe.
Once a spiny lobster settled to the substrate, the
pipe was slowly withdrawn allowing the lobster
to move freely about the aquarium. This tech-
nique prevented "tailflipping" by lobsters and fa-
cilitated active selection of habitats. Twenty-four
hours later we located the lobsters and recorded
their positions, as in previous experiments

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

(Herrnkind and Butler 1986). Fourteen spiny lob-
sters were tested on day 1, 14 more on day 2, and
so on until our stock of animals was depleted.
Each lobster was used only once. All pueruli were
collected on the incoming tide from the plankton
in interisland channels. Pueruli were either
tested immediately or allowed to metamorphose
for later use in experiments requiring juveniles.
Data were analyzed with log-linear Goodness-of-
fit tests.

Metamorphosis Experiment

The effect of siltation on the survival and time-
to-metamorphosis of pueruli was tested experi-
mentally in an outdoor, flow-through seawater
system. One freshly collected puerulus was placed
in each of 46 seawater-filled 1 L plastic beakers,
23 containing 5 cm diameter clumps of unsilted
algae and the other 23 an equal amount of silted
algae. Each container was independently sup-
plied with flowing, filtered seawater. Algal
clumps were replaced daily. Pueruli do not feed
and their habitat selection operates independent
of food availability (Herrnkind and Butler 1986),
thus no food was added to the containers. Sea-
water temperature in the beakers remained be-
tween 26° and 28°C; photoperiod was approxi-
mately 14L:10D. Pueruli were monitored daily
and their survival and time-to-metamorphosis
recorded.

Pueruli were collected as transparent postlar-
vae from the plankton in interisland channels
which concentrate oceanic postlarvae as they
move into Florida Bay nursery areas. Time-to-
metamorphosis values represent the elapsed time
(in days) from puerulus collection until metamor-
phosis into the first benthic stage. Values are
likely to differ among collections as different co-
horts of pueruli arrive inshore. There are cur-
rently no techniques available to determine the
actual age of pueruli (i.e., time since metamor-
phosis from the phyllosoma stage), but estimates
of duration of the puerulus stage range from
2 weeks to 1 month (Lyons 1980; Calinski and
Lyons 1983). Differences in time-to-
metamorphosis between the two treatments were
analyzed via a two-sample ^-test.

Juvenile Spiny Lobster Prey Selection
Experiments

Laboratory experiments were conducted to de-
termine juvenile spiny lobster prey preference

334

HERRNKIND ET AL.: RECRUITMENT OF SPINY LOBSTERS

and rate of consumption of algal epifauna. Prey
were obtained by rinsing large clumps oiLauren-
cia through a 100 |jl sieve. Prey included small
gastropods, amphipods, isopods, and ostracods
ranging in size from 1 to 9 mm (Marx and Herrn-
kind 1985a). Prey were individually counted and
placed in 1 L plastic beakers containing 750 mL of
seawater. One starved (24-h) juvenile lobster (6-
8 mm CL) was introduced to each container, al-
lowed to feed for 12 hours, and was then removed;
the remaining prey were counted. Fifteen spiny
lobsters were tested in each experiment; each lob-
ster was used once. Three experiments were con-
ducted using different prey combinations (Table
1), but the total number of prey available re-
mained similar and exceeded the amount a single
lobster could consume in 12 hours. Electivity in-
dices calculated for each experiment were used in
multiple comparison tests to determine whether
juvenile spiny lobsters fed preferentially or ran-
domly (Johnson 1980).

Table 1 . — Relative prey availabilities (A) and predator usage (U)
values (percentages) in the three juvenile spiny lobster prey selec-
tion experiments. N = ^5 lobsters per experiment. F-values calcu-
lated from Johnson (1980) indicate whether prey choice differed
significantly from random in each trial: none of the tests were signif-
icant at P = 0.05.

Experiment

Prey type

A U

Gastropoda

Tncolia sp.

Batlillaria spp.

Tegula spp.

Atys spp.
Amphipoda
Isopoda
Copepoda
Ostracoda
Decapoda

F-values
df

43.5 615 714 83.0 58.0 54.5

8.7 10 4 3

6.5 00 —

— — 14.3

32.6 33.9 10.0

0.2

7.2
9.6

8.3

0.46
4,11

0.40
3,14

7.0
17.4
11.6

6.0

0.09

4,11

7.1
20.0
13.1

5.3

RESULTS

Postlarvae were more abundant on collectors at
the silted site than at the unsilted site. Diver
surveys revealed that higher numbers of algal-
dwelling juveniles (<20 mm CL) resided at the
unsilted site despite greater influx of pueruli into
the silted area (Table 2). Only one juvenile spiny
lobster collected at the silted site was <20 mm
CL, most were considerably larger (25—35 mm
CL) than those at the unsilted site, and some were
possibly large enough to have immigrated there
from adjacent unsilted areas.

Silt recovered from the algal clumps was
largely calcareous and formed a cohesive cast
around the algal filaments. Classifying algal-
bound silts by particle size would yield irrelevant
values because sieving caused fragmentation of
aggregated particles. Therefore we report only
the total dry weight of the silts. Algae at No
Name Key carried a higher silt load than algae at
Burnt Point ix = 125 vs. 65 g/L algae, respec-
tively; t = 2.90, df = 18, P <0.01). Silt at No
Name Key was characterized as 12.5 ± 7.4% or-
ganic and 28.1 ± 27.6% carbonate by weight,
whereas Burnt Point silt was 16.9 ± 1.06% or-
ganic and 66.3 ± 14.5% carbonate (means ± 1 SD).
There was no significant difference in silt compo-
sition between sites (organics: t =0.83, df=4,
P > 0.05; carbonates: t = 1.97, df = 4, P > 0.05),
although sample sizes at each site were small and
sample variance substantial. Algae at both sites
contained primarily gastropods, amphipods, and
isopods, although significantly more gastropods
and echinoderms occupied unsilted clumps
(Fig. 2, Table 3; P < 0.05 in Bonferroni pairwise
multiple comparisons).

Significantly more pueruli settled in unsilted
algal clumps than in silted clumps during labora-
tory settlement choice experiments (38 vs. 11, re-
spectively; G = 15.72, P < 0.001). Juvenile spiny
lobsters responded similarly in the habitat selec-
tion experiment (54 vs. 24; G = 11.78, P < 0.001).
We excluded the open sand habitat in aquaria
from our analysis because 1) juvenile spiny lob-
sters are never found residing on open sand in the
field, presumably due to a lack of food and refuge

Table 2. — (A) Postlarval lobster abundances at silted and unsilted
Florida Bay study areas. Postlarval catch per unit effort (CPUE)
was estimated from Witham collector catches. (B) Juvenile lob-
ster (8-20 mm CL) abundances at the two study sites in 1985 and
adjacent areas sampled dunng 1983 and 1984. Juvenile CPUE
was estimated via diver surveys.

A. POSTLARVAL ABUNDANCE

CPUE

Location Description

June July

August

No Name Key silted
Burnt Point unsilted

0.08 0.32
0.01 0.05

0.15
0.10

B. JUVENILE ABUNDANCES

Location Description

Diving
hours

CPUE2

No Name Key (1983-85) silted
Burnt Point (1983-85) unsilted

28
20

0.04
0.50

ippiip - no ot postlarvae no. o( collectors
no. of days between collections

2CPUE

no. of lobsters collected
no. of diver hours

335

FISHERY BULLETIN: VOL. 86, NO. 2

LU
CC

u.
O

9
>

CC
Hi

1200 -

1000 -

800 -

600 -

- 400 -

200 -

GASTRO* OSTRA

AMPHI

ISO

DECA ECHINO*

Figure 2. — Abundances (x ± 1 SE) of the six most common prey found in algal clumps at
the silted No Name Key and unsilted Burnt Point sites. Five clumps were collected at each
site. Values are standardized by clump volume. Asterisks denote significant difference in
Bonferroni multiple comparison tests (P = 0.051. Abbreviation key: GASTROpod, OSTRAcod,
AMPHlpod, isopod, DECApod, ECHlNOderm.

Table 3. — Two-way fixed-effects ANOVA testing for differences in
the total number of individuals among six prey categones (see
Table 1) at two sites, one silted (No Name Key) and one unsilted
(Burnt Point). Data were log transformed.

Source

Site 1 0.841 18.78 0.001

Prey type 5 6 242 27.89 0.001

Site X prey type 5 0.847 3.78 0.006

Error 48 2.149

(Marx and Herrnkind 1985b; Herrnkind and But-
ler 1986), 2) most spiny lobsters recovered from
open sand were actually in corners, indicating
edge-seeking behavior rather than selection for
sand per se, and 3) only 10.6% of 142 spiny lob-
sters tested were found on sand even though it
constituted 68% of the exposed substrate in
aquaria. Silt had no effect on puerulus survival
through metamorphosis to the first benthic instar
(13% vs. 9% mortality in silted and unsilted
algae, respectively), or time-to-metamorphosis
(Fig. 3; t = 0.37, P > 0.05).

The total number of prey items consumed in the
juvenile prey selection experiments ranged from
19 to 57 prey per lobster per 12 hours. Juveniles
fed randomly from the three different prey combi-
nations and frequencies offered to them (Table 1).

C/) 20
O

I
Ql
CC

10 "

— O O

' //^

SILTED

" U

. // UNSILTED

•

0 1 2 3 4 5 6

DAYS

FlGt_iRE 3. — Cumulative number of spiny lobster postlarvae
metamorphosing as a function of time in one of two treatments:
silted algae (Laurencia spp.) or unsilted algae. Twenty-three
postlarvae were tested in each treatment; two postlarvae died in
the unsilted treatment and three in the silted treatment.

DISCUSSION

The relative paucity of newly settled spiny lob-
sters in the heavily silted region around No Name
Key over a 3-yr period indicates that low recruit-
ment to benthic habitat is typical there. The ab-
sence of juveniles was apparently not due to a

336

HERRNKIND ET AL : RECRUITMENT OF SPINY LOBSTERS

lack of postlarval influx, which was higher than
that at the unsilted Burnt Point site, but instead
to low rates of postlarval settlement. Results from
our habitat selection experiments support this
hypothesis because settlement was significantly
lower in silted algal clumps than in unsilted
clumps. Previous studies showed that postlarvae
selectively settled in highly architectured materi-
als, like algal clumps (Herrnkind and Butler
1986). Thus, heavy silt covering an otherwise pre-
ferred habitat either masks the stimuli triggering
settlement or contains stimuli that elicit rejection
by pueruli. This question remains for further
study. We cannot conclusively ascertain from our
laboratory experiments the mechanisms govern-
ing habitat choice in the field where silted and
unsilted habitats may not be adjacent, as they
were in our aquaria. Yet for many species with
planktonic larvae, these kinds of experiments,
coupled with field observations of more general
patterns of behavior and abundance, provide
valuable insights into natural processes (Sulkin
1986).

Twenty percent of the pueruli we tested in lab-
oratory tanks settled in silted clumps despite the
general rejection of this habitat. Pueruli settling
in silted algae probably metamorphose normally
into the first benthic instar, as indicated by the
equivalent time-to-metamorphosis and early post-
settlement survival in both silt levels tested.
However, subsequent residency by juveniles pre-
sumably is limited, given their great mobility and
preference for unsilted algae with high food levels
(Marx and Herrnkind 1985b; Herrnkind and But-
ler 1986). We found that prey abundances were
significantly lower in silted algae. Thus, even if
pueruli settle in the silted habitat the subsequent
juveniles may leave to obtain adequate food. Fre-
quent interclump movement by juvenile spiny
lobsters, searching either for food or unsilted
habitat, would predictably result in increased
predatory mortality. Susceptibility to predation
is much greater for juveniles in the open, than it
is for individuals amidst algal clumps or dense
seagrass (Herrnkind and Butler 1986). Thus, ju-
venile residency patterns and susceptibility to
predation may, in addition to locally low settle-
ment, contribute to the paucity of lobsters in the
silted habitat.

The algal-bound silt load at No Name Key was
roughly twice that at Burnt Point where spiny
lobster recruitment was considerable. Our cur-
sory surveys from Key Largo to Boca Chica Key
indicate that similar silt levels are common.

though geographically variable in Florida Bay.
Benthic algae, including La//re/icm spp., serve as
sediment traps (Scoffin 1970) and demonstrate a
remarkable resistance to siltation, growing pro-
fusely even in heavily silted areas. Silt in these
areas is primarily calcareous, most of it probably
a byproduct of sediment processing by deposit
feeding shrimp (particularly Callianassa), an-
nelids, and sea cucumbers.

We did not evaluate the geographic extent of
siltation relative to spiny lobster settlement in
Florida Bay. However, the demonstrated aver-
sion to settling in naturally silted algae, charac-
teristic of the region around No Name Key,
strongly suggests that low postlarval recruitment
and juvenile abundances would occur in similar
conditions elsewhere. The sparse juvenile popula-
tion at our silted site, one-tenth that of the un-
silted site, suggests deleterious impact of high
chronic silt levels in areas of potential recruit-
ment. Human activities also cause siltation (Mor-
ton 1977; Allen and Hardy 1980). We noted that
algal stands adjacent to heavily trafficked boat
channels typically were more heavily silted than
adjacent areas. We suspect the effect of manmade
siltation to be similar to that from natural causes.
Although it is now generally accepted that Flor-
ida Bay shallows serve as the main nursery
grounds for the south Florida spiny lobster popu-
lation (Marx 1986), the regional distribution of
settlement and early juvenile habitation remains
to be mapped. Future wide-area surveys by con-
cerned researchers and agencies should include
sampling of new spiny lobster recruits as well as
silt levels. Meanwhile, sizable human activities
such as channel construction, dredging, spoil
dumping, coastal development, and mineral min-
ing must be viewed as potentially deleterious to
spiny lobster recruitment.

ACKNOWLEDGMENTS

We thank the Sea World Center for Marine Sci-
ence and Conservation on Long Key, FL, for logis-
tical support and the use of their facilities. John
Hunt and Jim Marx of the Florida Department of
Natural Resources Marine Research Laboratory
in Marathon provided field assistance and helpful
advice throughout the project. Comments by
P. Greenwood, D. Wilber, and two anonymous re-
viewers substantially improved the manuscript.
This research was supported by a Sea Grant
award (R/LR-B-16) to W. F. Herrnkind. Addi-
tional support was provided M. J. Butler via

337

FISHERY BULLETIN: VOL. 86, NO. 2

Marine Science Fellowships from the Aylesworth
Foundation and the International Women's Fish-
ing Association (IWFA).

LITERATURE CITED

Allen, K D . and J W Hardy

1980. Impacts of navigational dredging on fish and
wildlife: a literature review. U.S. Fish Wild. Serv., Biol.
Serv. Program FWS/OBS-80/07, 81 p.

Aller. R C , AND R E Dodge

1974. Animal-sediment relations in a tropical lagoon,
Discovery Bay, Jamaica. J. Mar. Res. 32:209-232.
Andree. S W

1981. Locomotory activity patterns and food items of ben-
thic postlarval spiny lobsters, Panulirus argus. MS
Thesis, Florida State University, Tallahassee, FL, 50 p.

Aziz. K A . and J G Greenwood

1982. Response of juvenile Metapenaeus bennettae Racek
& Doll 1975 (Decapoda: Penaeidae) to sediments of differ-
ing particle size. Crustaceana 43:121-126.

BOTERO, L , AND J ATEMA

1982. Behavior and substrate selection during larval set-
tling in the lobster Homarus americanus. J. Crust.
Biol. 2:59-69.

Calinski, M D , AND W G Lyons

1983. Swimming behavior of the puerulus of the spiny
lobster Panulirus argus (Latrielle, 1804) (Crustacea: Pal-
inuridae). J. Crust. Biol. 3:329-335.

Cortes, J N . and M J Risk

1985. A reef under siltation stress: Cahuita, Costa
Rica. Bull. Mar. Sci. 36:339-356.

Crisp, D J

1974. Factors influencing the settlement of marine inver-
tebrate larvae. In P. T. Grant and A. M. Mackie (editor),
Chemoreception in marine organisms, p. 177-265.
Acad. Press, N.Y.

1976. Settlement responses in marine organisms. In
R. C. Newell (editor). Adaptation to environment: essays
on the physiology of marine animals, p. 83-124.
Butterworths, Lend.

Gray, J S

1974. Animal-sediment relationships. Oceanogr. Mar.
Biol. Annu. Rev. 12:223-261.

HERRNKIND, W F , AND M J BUTLER IV.

1986. Factors regulating postlarval settlement and juve-
nile microhabitat use by spiny lobsters Panulirus
argus. Mar. Ecol. Prog. Ser. 34:23-30.

Howard, A E., and D B Bennett

1979. The substrate preference and burrowing behavior of
juvenile lobsters (Homarus gammarus (L.)). J. Nat.
Hist. 13:433-438.

Johnson, D H

1980. The comparison of usage and availability measure-
ments for evaluating resource preference. Ecology
61:65-71.

Jones, G , and S Candy.

1981. Effects of dredging on the macrobenthic infauna of
Botany Bay. Aust. J. Mar. Freshwater Res. 32:379-398.

Little, E J

1977. Observations on recruitment of postlarval spiny
lobsters, Panulirus argus, to the South Florida coast.

Fla. Mar. Res. Publ. No. 29, 35 p.
Little. E J , Jr , and G R Milano

1980. Techniques to monitor recruitment of postlarval
spiny lobsters, Panulirus argus, to the Florida
Keys. Fla. Mar. Res. Publ. No. 37, 16 p.
Lyons, W. G

1980. Possible sources of K'lorida's spiny lobster popula-
tion. Proc. Gulf Caribb. Fish. Inst. 33:253-266.
Marx. J M

1986. Recruitment and settlement of spiny lobster pueruli
in south Florida. Can. J. Fish. Aquat. Sci. 43:2221-
2227.
Marx, J M , and W F Herrnkind

1985a. Macroalgae (Rhodophyta: Laurencia spp.) as a
habitat for young juvenile spiny lobsters, Panulirus
argus. Bull. Mar. Sci. 36:423-431.
1985b. Factors regulating microhabitat use by young ju-
venile spiny lobsters, Panulirus argus: food and shel-
ter. J. Crust. Biol. 5:650-657.
Morton, J. W.

1977. Ecological effects of dredging and dredge spoil dis-
posal: a literature review. U.S. Fish Wild. Serv. Tech.
Pap. No. 94, Wash., DC.

Pearson, T H , and R Rosenberg

1978. Macrobenthic succession in relation to organic en-
richment and pollution of the marine environ-
ment. Oceanogr. Mar. Biol. Annu. Rev. 16:229-311.

Pottle, R A . and R W Elner

1982. Substrate preference behavior of juvenile American
lobsters, Homarus americanus, in gravel and silt-clay
sediments. Can. J. Fish. Aquat. Sci. 39:928-932.

Rhoads, D C

1974. Organism-sediment relations on the muddy
seafloor. Oceanogr. Mar. Biol. Annu. Rev. 12:263-300.
Rhoads, D C , and D K Young

1971. Animal-sediment relations in Cape God Bay, Mas-
sachusetts. II. Reworking by Molpadia oolitica
(Holothuroidea). J. Mar. Biol. 11:266-261.
Roach, S G

1983. Survivorship, growth, and behavior of juvenile lob-
sters Homarus americanus Milne-Edwards in controlled
environments in nature. Nova Scotia Dep. Fish. Mar.
Res. Tech. Rep. Ser. 83-02, 60 p.

RUELLO. N V

1973. Burrowing, feeding, and spatial distribution of the
school prawn Metapenaeus macleayi (Haowell) in the
Hunter region (Australia). J. Exp. Mar. Biol. Ecol.
13:189-206.

SCOFFIN, T p.

1970. The trapping and binding of subtidal carbonate sed-
iments by marine vegetation in Bimini Lagoon, Ba-
hamas. J. Sediment Petrol. 40:249-273.
SULKIN, S D

1986. Application of laboratory studies of larval behavior
to fisheries problems. Can. J. Fish. Aquat. Sci. 11:2184—
2188.

WiTHAM, R , R M iNGLE, AND H W SiMS jR.

1964. Notes on postlarvae o{ Panulirus argus. Q. J. Fla.

Acad. Sci 27:289-297.
WiTHAM, R , R M iNGLE, AND E A JOYCE jR

1968. Physiological and ecological studies of Panulirus

argus from the St. Lucie estuary. FL Board Conserv.

Tech. Ser. 53:1-31.

338

A LIMITED INFORMATION APPROACH FOR
DETERMINING CAPITAL STOCK AND INVESTMENT IN A FISHERY^

James E Kirkley^ and Dale E. Squires^

ABSTRACT

There have been few empirical studies on the level of capitalization and investment in fisheries
because the necessary data are often inadequate. Specifically, data on capital stock and investment
in a fishery are not routinely collected and compiled or are limited in scope. In this study, a method
is provided for estimating the aggregate capital stock and investment in a fishery utilizing the
available information. Data on acquisition and list prices and vessel characteristics for a sample of
New England vessels are obtained. The data are then used to estimate an hedonic cost function which
specifies the acquisition price as a function of vessel characteristics. The resultant equations are
subsequently used, with information on vessel characteristics for all New England vessels, to esti-
mate aggregate capital stock and investment. The results indicate that substantial investment oc-
curred in the otter trawl and scallop dredge fisheries, particularly since the Magnuson Fisheries
Conservation and Management Act. Moreover, the results demonstrate that the number and change
in the number of vessels are inadequate indicators of the level of capital stock and investment in a
fleet comprised of vessels with heterogeneous characteristics.

The common property nature of fisheries is recog-
nized as causing excess capitalization and har-
vesting capacity (Gordon 1954). The theoretical
argument is that since fishermen do not have to
pay for the utilization of common property fish
stocks, new vessels enter a fishery until net rev-
enue is driven to zero. This common-property fea-
ture results in more capital and investment than
is economically optimum. Public regulation of
fishing industries is usually advocated to redress
the excess entry, economic inefficiency, and loss
of economic rents in a common property fishery
(Scott 1979; Sissenwine and Kirkley 1982).

In essence, overcapitalization and excess in-
vestment are perceived to be the reasons for many
of the major fisheries problems (Hilborn 1983).
Economists argue that management of overcapi-
talization is necessary to realize the benefits of
fisheries (Cunningham et al. 1985). Alterna-
tively, gains from fisheries management require
control of overcapitalization and excess invest-
ment (Charles 1983a, b).

Clark et al. (1979) and Charles (1983a, b) ex-
tended the static theory of optimal fisheries in-
vestment as developed by Gordon (1954). They

^Virginia Institute of Marine Science (VIMS) Contribution
No. 1467.

^College of William and Mary, Schools of Marine Science and
Business Administration, Gloucester Point, VA 23062.

3Southwest Fisheries Center La Jolla Laboratory, National
Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 92038.

also demonstrated the relationships between in-
vestment, capital stock, overexploitation, and
fisheries management. They concluded that in-
vestment should be more conservative to prevent
overexploitation.

The economic literature provides substantial
justification for the need to solve the problems of
overcapitalization and excess investment. Yet,
few empirical studies document the level of capi-
talization and investment."* Moreover, there ap-
pears to be no attempt by any U.S. agency to
routinely collect and compile statistics on either
the stock of capital or the level of investment in
U.S. fisheries. Overcapitalization and excess in-
vestment, though, continue to be suggested as the
reasons for many of the economic problems of
fisheries.

Since data are often inadequate, many empiri-
cal studies on fisheries consider capital stock and
investment in terms of the number of vessels.
That is, capital stock and investment in a fishery
are tyically measured in terms of number of ves-
sels. In the absence of appropriate information,
the number of vessels may be the only basis for
examining overcapitalization and excess invest-
ment. Alternatively, if a fleet has identical-sized
vessels, fixed inputs or vessel characteristics, and

Manuscript accepted February 1988.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.

'•Tettey et al. ( 1986) provided an exception, but their analysis
is restricted to the shrimp fishery and is based on cost data
obtained from boat builders.

339

FISHERY BULLETIN: VOL. 86, NO. 2

gear, the number of vessels may be used to indi-
cate the stock of capital and investment. How-
ever, Kendrick (1961) demonstrated that the total
number of operating units, such as the number of
plants, is an inadequate measure of capital stock.
Also, few fleets have identical-sized vessels, char-
acteristics, or gear. As a consequence, the number
of vessels does not provide adequate information
to indicate the level of capitalization and invest-
ment.

This study presents an approach for estimating
the aggregate or industry level of capital stock
and investment in a fishery comprised of hetero-
geneous vessels using available information.
Data available include acquisition and list prices
for some but not all vessels and boat characteris-
tics for all vessels in a fleet. The approach was
developed and used to obtain estimates of capital
stock in support of the United States and Cana-
dian maritime boundary dispute. It is viewed as
an initial step towards examining the problems of
excess capitalization by providing estimates of
aggregate capital stock and investment.

THE DEFINITION AND MEASUREMENT
OF CAPITAL AND INVESTMENT

Capital

The concept of capital has created problems for
economists for quite some time. The term has
been used in so many different contexts that it is
a source of enormous confusion (Hirshleifer
1970). Sloan and Zurcher (1968), in their
"Dictionary of Economics", define capital as "One
of the major factors of production consisting of
property from which an income is derived, ex-
pressed in terms of money. Popularly, the term is
frequently used interchangeably with capital
good. A distinction is sometimes made between
money capital, or that part of the capital held in
the form of money and bank deposits, and prop-
erty capital, or that part of the capital held in the
form of evidences of ownership such as stocks,
bonds, and mortages."

Hirshleifer (1970) presented three meanings of
capital: 1) real capital or capital stock, 2) capital
value, and 3) liquid capital. Real capital is de-
fined as a collection of capital goods or an aggre-
gate of heterogeneous capital inputs. It is one of
the major productive commodities or economic
factors of production. Capital value is the net dis-
counted value of expected future income streams
associated with a capital good. Liquid capital is

the level of current funds available or intended
for investment.

In the case of fisheries, real capital or capital
stock is the form of capital which should be exam-
ined with respect to the problems of overcapital-
ization and excess harvesting capacity. It is the
relevant measure of capital goods used or avail-
able for production (National Academy of Sci-
ences 1979). Moreover, capital stock is the con-
cept of capital used to define and measure the
services of capital inputs which are required to
harvest fish; that is, it is the concept of capital
required to define the economic production tech-
nology of a fishery. Thus, this study is concerned
with the concept of real capital stock. ^

The measurement of capital stock in a fishery,
however, presents several problems. First, capital
inputs are usually quite heterogeneous and can-
not be easily aggregated without restrictive as-
sumptions about the form of the catch equation or
fishermen's behavior. Conceptually, it should be
possible to combine all the different types of capi-
tal goods by weighting each type by its average
compensation (i.e., the rental price). However, in-
formation at this level of detail is typically not
available. Second, fisheries agencies, particularly
in the United States, generally do not collect and
compile information necessary to calculate capi-
tal stock and investment. Third, in order to com-
pare changes in capital stock over time, the meas-
ure of capital stock must be converted to some
base period value by deflation.

In empirical economic studies of traditional in-
dustries, the common practice is to measure the
capital input or stock by converting the purchase
or acquisition price or the book values of capital to
base period values by the use of a price index
(National Academy of Sciences 1979). Varian
(1984) noted: "The usual procedure is to measure
capital value and then deflate by a price index; in
some sense, this should measure the level of cap-
ital stock." In this study, the acquisition price is
used as the measure of the stock of capital or
capital value; the real stock of capital is obtained
by dividing the capital value by the producer
price index for heavy machinery.

Investment

The definition and measurement of investment
is more straightforward than is the definition and

5The market value of capital, acquisition price, and cost of
capital are used interchangeably in this study.

340

KIRKLEY AND SQUIRES: CAPITAL STOCK AND INVESTMENT IN A FISHERY

measurement of capital. In simple terms, invest-
ment is the exchange of money for some form of
property. Although there are several types of in-
vestment (Branson 1972), attention is restricted
in this study to an examination of net investment
(i„) which is defined as the difference between the
stock of capital in two periods of time (Baumol
1977). This is the same concept of investment ex-
amined by Tettey et al. (1986).

Replacement investment is another type of in-
vestment. The sum of replacement investment
and net investment equals total investment. Re-
placement investment is measured by multiply-
ing the rate of depreciation times the level of the
capital stock. Replacement investment, however,
is not estimated in this study since the method
and rate of depreciation must be arbitrarily as-
sumed.

METHOD OF ANALYSIS

The Hedonic Approach

While the definition and measurement of real
capital stock and investment are conceptually
straightforward, limited data and heterogeneous
capital inputs complicate the measurement of
capital stock and investment. In particular, data
are usually only available for a small number of
vessels which are often similar in their character-
istics; these data are inadequate for calculating
actual or observed capital stock and investment
in a fishery. However, data are available on the
acquisition price and characteristics of vessels
which permits estimation of the stock of capital
by an hedonic approach. In turn, estimates of cap-
ital stock from the hedonic approach can be used
to estimate net investment in a fishery.

The hedonic approach hypothesizes that the
price of a commodity is influenced by its charac-
teristics (Rosen 1974). The hedonic price or char-
acteristics function in market equilibrium re-
flects both the distribution of marginal rates of
substitution over households and the distribution
of marginal rates of transformation over firms. In
effect, the hedonic hypothesis states that goods
may be valued for their attributes and that im-
plicit or hedonic prices exist as a function of the
attributes.

Griliches (1971) estimated hedonic cost func-
tions of U.S. automobiles using advertised or list
prices and actual transaction prices. Griliches,
however, was concerned with explaining quality
differentials. Ladd and Martin (1976) examined

prices and demands for input characteristics
using a Neoclassical input characteristics model.
Triplett (1986) estimated quality adjusted price
indexes for computers. Thus, there is a history of
deriving prices of commodities as a function of
their characteristics.

The hedonic approach offers several attractive
properties for estimating capital stock in fishing
industries. First, the hedonic method incorpo-
rates changes in the quality of capital over time
(Triplett 1986). This is because changes in quality
should be reflected in vessel acquisition prices.
The hedonic approach, thus, permits a quality ad-
justed measure of capital stock to be obtained. In
contrast, vessel count ignores changes in the
quality of capital. Second, the hedonic approach
allows easy aggregation of the heterogeneous cap-
ital inputs frequently observed in a fishing fleet
because the heterogeneous capital is measured by
value rather than physical measures.

In the hedonic approach, cost (C) or input price
may be expressed as a function of the associated
characteristics of the commodity or input,

C = f(CHi, CH2, ...,CH„) ,

(1)

where C is cost, CH, is the tth characteristic
(Braeutigam et al. 1982). The characteristic cost
equation can be obtained either from a dual speci-
fication or by determining the reduced form equi-
librium equation.

The Dual and Reduced Form

The dual specifies cost as a function of input
prices and their characteristics. The partial
derivative of the dual cost function with respect to
input prices yields the demand for the inputs as a
function of prices and input characteristics. The
reduced form is obtained by solving structural
equations, equations which explain the behavior
and interrelations of endogeneous variables, of
demand and supply for the endogeneous variable,
price, in terms of exogeneous variables and dis-
turbance terms. ^

In this study, the reduced form is directly speci-
fied and estimated for two reasons: First, all of the
input price data necessary for estimating a dual
are not available. Second, estimation of the struc-
tural equations is complicated by the need to use
a limited dependent variable method of estima-

^See Rosen (1974) for additional information on obtaining
reduced form equations for hedonic prices.

341

FISHERY BULLETIN: VOL. 86, NO. 2

tion since demand is a discrete choice, and there
are individuals or firms which do not purchase a
vessel.

Rather than attempt to deal with these difTi-
culties which do not appear to have widely ac-
cepted solutions, it is assumed that there are
underlying structural equations of demand and
supply which permit the derivation of the reduced
form equation. Moreover, the direct specification
and estimation of the reduced form is the more
commonly used approach to estimate the value of
a commodity as a function of its characteristics
(Bockstael et al. 1986). Last, the use of the re-
duced form should not create a problem since the
hedonic function is a joint envelope in character-
istic space of demander's bid schedules and sellers
offer schedules, and thus, the observed implicit
price locus is a reduced form equilibrium vector
reflecting both supply and demand influences. As
a result, shifts in both market demand and supply
are incorporated into the reduced form equi-
librium vector.

The Capital Stock Model

As previously indicated, it is common practice
to use the acquisition price as a measure of capital
stock. By the hedonic approach, the capital stock
or acquisition price of a fishing vessel is postu-
lated to be a function of several vessel character-
istics. The following general specification is con-
sidered:

C = f(LN, CS, YEAR, FI, AGE, DNO, HPWR, GRT),

(2)

where C is the list or advertised price or the ac-
quisition price of a vessel, LN is length, CS is hull
type (1 = steel, 0 = other), FI is fishery or gear
tj^e,' AGE is age of vessel, DNO is a dummy
variable for new and used vessels (DNO = 1 for
new and 0 for used), HPWR is engine horsepower,
and GRT is gross registered tonnage. The vari-
able year equals the year of the observation. It is
included to incorporate technological changes

'^Gear type is an integer valued variable set equal to the gear
codes used by the Northeast Fisheries Center. This may influ-
ence the results similarly to the use of nonbinary dummy vari-
ables (Kmenta 1971). The use of this variable implies that the
cost difference between gear types is equal to a scalar multiple
of the lowest integer valued gear code. Thus, statistical rejection
of cost differences due to gear type is actually a rejection of
differences not being equal to scaleir multiples. However, the
large number of gear codes prohibits their separate treatment.

over time, such as the introduction of electronics
and stern trawling, and structural changes in the
industry such as changes in organization of capi-
tal and fish markets and changes in public regu-
lation.

Functional Form

Several functional forms could be specified for
Equation (2). Alternatively, generalized Box-Cox
(Box and Cox 1962) transformations could be used
to determine the functional form. However, the
selection of functional form is mostly concerned
with obtaining the relationship between the de-
pendent variable and individual independent
variables. This study is primarily concerned with
estimating the capital stock. Moreover, it may be
quite difficult to estimate capital stock from Box-
Cox transformations or other functional forms
since the conditional expectation is often of a com-
plex form (Smallwood and Blaylock 1986). Also,
Dadkhah (1984) noted that predictions based
upon some transformations and functional forms
result in biased predictions and asymmetric confi-
dence intervals.

Since this study is primarily concerned with
estimating capital stock and investment rather
than estimating the implicit prices for character-
istics, a linear functional form is proposed for
Equation (2). The capital stock equation is speci-
fied as a second-order polynominal in order to
provide an approximation to an unknown, under-
lying hedonic function. This is equivalent to
specifying a generalized quadratic flexible func-
tional form and imposing zero valued restrictions
on the cross product coefficients or interaction
terms (Lau 1978). Flexible functional forms are
widely used in economic studies concerned with
determining the underlying economic structure
because they impose very little structure on the
economic equations of concern (Blackorby et al.
1978).

Postulated Relations

The postulated relations between cost and each
of the explanatory variables is as follows: 1) the
coefficients for the first order terms of vessel ton-
nage, length, and engine horsepower should be
positive since costs should increase as the size
characteristics increase; 2) the coefficient for
steel-hulled vessels should be positive since steel
vessels are generally more expensive than wood
or fiberglass vessels; 3) the dummy variable coef-

342

KIRKLEY AND SQUIRES: CAPITAL STOCK AND INVESTMENT IN A FISHERY

ficient for new vessels should be positive since
new vessels tend to cost more because they are
newer and usually incorporate more recent tech-
nological advances; 4) the second-order coeffi-
cients for the size characteristics are expected to
be negative since cost is believed to increase at a
decreasing rate in response to increases in the
size of a vessel.

THE DATA

Data used to estimate the capital stock equa-
tion were obtained from three sources: First, in-
formation on vessel acquisition price and associ-
ated characteristics were obtained for 164 new
vessels from the Northeast Regional Office of the
National Marine Fisheries Service.^ Second, data
on list prices and vessel characteristics for 946
used vessels were obtained from various trade
magazines and vessel brokers. Third, detailed
vessel and fishery data were obtained from the
U.S. Coast Guard master vessel listing and the
Northeast Fisheries Center. Data were obtained
for the years 1965 through 1981.

The first two sets of data were used to estimate
the capital stock equations. The third data set
was used to estimate aggregate capital stock and
investment in selected New England fisheries
using the estimated capital stock equations; that
is, vessel characteristics for all of the vessels in
the New England fleet were inserted into the esti-
mated capital stock equations to obtain estimates
of capital stocks per vessel in a given year. The
estimated capital stock per vessel was summed
over all vessels in a year to obtain total or aggre-
gate capital stock. Net investment was then cal-
culated as the annual change in total capital
stock.

The use of both the list and acquisition prices,
however, presents a problem. The list price re-
flects the supply price or the price at which a
vessel is offered for sale. The acquisition price
reflects the price determined by the equilibrium
between demand and supply. As a consequence,
estimates of capital stock and investment may be
in error.

It is not known by how much the acquisition
price differs from the list price for used vessels.
Thus, the magnitude of the error cannot be calcu-
lated. Since the list price is not less than the ac-
quisition price, estimates of capital stock based on

Equation (2) are likely upwards biased. Counter
to this problem is the argument that by the time
a vessel is ready for fishing, required capital mod-
ifications or improvements may result in the cap-
ital stock being close in value to the list price.

There are no completely satisfactory economic
justifications for using both the list and acquisi-
tion prices. Two possible justifications are that
Griliches (1971) and others used both prices, and
capital improvements or modifications may be
nearly equal to the difference between the list
and acquisition prices. Another possible justifica-
tion is the accepted use of cost data obtained from
boat builders as in Griffin et al. (1978) which may
impose similar problems plus the statistical prob-
lem of measurement error. In the remainder of
this paper, the term "acquisition price" is used
although the estimation and calculation of capital
stock and investment are based on both price
series.

EMPIRICAL RESULTS

Equation (2) was estimated by ordinary-least-
squares with the dependent variable, the acquisi-
tion price, measured in both nominal and real
terms; the real price is the nominal price deflated
by the producer price index for heavy machinery.
As previously stated, deflation is necessary to es-
timate the real capital stock. The estimated coef-
ficients and statistical results are presented in
Tables 1 and 2.

Equation (2), however, was also estimated
using data for several groups of years between
1965 and 1981. The reason for considering differ-
ent time periods was that prior knowledge of New
England fisheries suggested that various changes
occurred in the fisheries during the selected time
periods.^ Moreover, economic studies of New Eng-
land fisheries have often been criticized for as-
suming stable relationships over time or over dif-
ferent cross sectional units. Failure to incorporate
changes in the estimated relationship between
cost and vessel characteristics over time may re-
sult in biased parameter estimates.

Changes which have possible ramifications for
the estimated relationships include 1) depressed
resource conditions and the presence of foreign
fishing between 1965 and 1971; 2) management
by the International Commission for Northwest

^Data are confidential and may not be available to other re-
searchers or the general public.

^Additional information on possible structural changes in
New England fisheries is available in Kirkley et al. (1982),
Dewar (1983), Doeringer et al. (1986), and Kirkley (1986).

343

FISHERY BULLETIN: VOL. 86. NO. 2

Tj-

r~.

o r;^

•* —

o> ^

r--

to ^

CM <0

■* CO

T-

CO ^

in in

tv. CO

^ —

- s

£2.

' 2!.

C3>

CD U)
> 05

>. >

E ?

C:in

cop

5 P

CJ>
00

CM
Op

CM
CM

e£

Cvj o

CM Tt

CM CO

T—

^—

y—

•>*'

"g.

« aJ

C\J —

T- --^

r^ ^-^

■*

,_.

C/D ®

CD
CO

CM
•<*

^ w

r- —

. Ol

S.9

CM ■•-

^ C

<^ C

C^J

T—

y—

1 ^^

""^

^~^

"■"'^

§<

C\J ^_^

t^ ^^

in ^__

•<r

(35

!= CO

CM 05

CO o

C» CM

T—

t-~

j_ (0 - —
o ■ - w

z
Q

05 00

r- o

CO CO

CJ>

in C.

5£i

O T-

CM
CM

5
CM

T-^

•<r

» S2

o ^^

1^ ^^

in ^^

S2 s

Q.
U

d CO

CM O

CO h-

T—

CM O

CM Tt

<o ^

> -g (0

1 d

^ei

gS2

<5
in

rii s

1
o

(NJ

2 —

CO ^^
CO ^

•,- CO

<75 r>J

(O o

■«T

C\)

<o to
c II a
n y- c

"" si

m ih

1 '*"'

1 "^^

c ^^ —

o c S2

y—

c o o

CD r: ^

tf)

o> —

■* .~.

C\J ^^

, ^

,,_^

,_^

T—

, ,

CJ c-

tn 2 '^

z
-I

§?

d°
CO '^

CD
O

CM O

CO CO

»- ■*

•*

Cvj

CD 0

1 "^

CO ^-'

CD "-^

" —

•<t

^ —

CVJ

,..

CD ,^

Y—

T •*

^ in

■<*

"D -;-.' i -^

o in

o r--:

O <£>

m nj © >

Ol => ss e

1 T—

(/)

^-^

""^

^-^

' — '

""^

ra CT-g ®

^ « o x:

,—

<y>

S§oo

CO —

CM O

o -~-

1-^

?;"

(5 CO

O
CD

■^

CtD

S =* 0 o

O -;

"^ ,^

to )-c

(J)

S « >, .

CM £.

1 2.

2= CD S

« t: .c (0

5i c — c

ffl ra .i" i;

Csj

® d

CM —

C35

S"
in

■*

« £ tr ^

[

« 5 < —
2 S UJ m

IE 3) ^ CS

■ — '

2. w ^ 01
mo.-®

CNj x: -o >
QC c ra »

■«r

Tj-

CM ^^

Cvj

•<t

™ £ i §

«D CO

d CO

in o

T-

iri

^ CO

•■- CM

CO CO

<J>

CO 2.

72i

^-S

"" q! X) 0

■>s^ _^

d ^

•<J-

0 ra ?>§
"5 - >,

GO - — -
5= CM

CM ®

CO
<D

<ji r-

in >«

c
o

■.- o

22 -*

iP CO

y-i

C*J

Tj-

t-- ~^

?% S E
1^ S E

t;: ^ — > ^

■V

T—

•S "^ m "C^

l~~

(DPS"

T—

y—

(0 ^CM -«

ll^i

</)

1--.

5 a'l^

— CO -

(0
0)

1
in

/>-

(j>

<7>

o
o

CO
CD

a.
to
o

"(0
0)

c
o

13
CT
(t>

'c
o

CD
(0

m
<

CO
CO

05
00

corr- y- ^^ eg—, h-^^ t^ —

in"' "-o cMin -"jco •<rin

COT Tj-o 'tCD y- y- -^ <M

I CO I cvj I ^ C^ 06 ^ CD

CM O

in CM

CO ,^

£2 =

CMd OOlh '-CO COCM

m ^ CD :;r

CD '^ 05 "^
••- -r- CM •.-

00 --.
CO

05 — .

■<3- —

^CM $2Jm SP 5^
CM--- CM'- incM ■>-o

I "^ I ~^ "^ I "^

CO — .

CO -^ CO — ~

" Eo «> ^

CO °^ CM ^

Tt -I- 1-

CO
t^ 00

••-t^ r^TT t^OO -^(D

■r-05 OOO) CMCD r^o

'- _ ■•- _ 00 _ o

00 C. CO C- CM —

- o

CM ^^ 1- ^^

t^ "tj- CD r^

^-05 in 00

in ^ CD .^

CD C CO C

oco r^co yf (£> coin

o r^ •.- -^ eg Tj- ,- <d

"^(S '-(-i CDfvi CD_.

I S. I 2. CO 2i eg C-

CO,^ y- ^^ O^^ ■^^^ y- ^^ CO— ^

opco c^co ^co Tin ^. <y> Pm

coeg cb^ CM'— 05® 'i'^ -^o

'-COCO-q^CO.^'-.^ \ y^ jC3

— I — I -^ ^ ^ ^

000
CD ~ CO — . CM . — .

§s gte «s

CO o o 10 in '*•

I ^^ •* -^ CO -^

o o

CO ^^ CO . —

05 ^^ CO

o S o

05 -^

CD "^ eg

CO o in eg T- CM

^-' eg "-' CO ^

^ CM

o r-

P ^

'. 00

O CO O '-

'- _ CM _ CO

^ CO o CO '"' ~

d> y- O CO

CM '-

P in

O CO

05 fs.

eg £. n

CM CO

CO ^ — ^ o ^ —
7^ in • o
Si o r: CO

CO ■ CO •

'- in CD

■^ 'X- f^ 'X- CD ;r-

■ in _; in ■ o

05 CD CO '^ ^ CO

in Z; CD ■ r- ■

I O '- '-

•t,^ CO. h^^ co,^ CM—, in,^

ooco '-r^ 'S-'- r^Tj- •^r^ cdcd

CO ,nCO rSI^ CO*" d"^ lri°

CO ^^

CM r^

O I '-

CO CD O CO

CD^ in^ Tt^ 05^

ihcD COCO -^in oocM

cocD cocg cor^ oin

'- y- f-T*: C*'d ■^ y^

y- CO

■* -^ d

o ;r o r~

CO o 1^ '^

00 ■ '- T

■* '- 00 .r-

CO ;;;-

r^ 00

eg CO

CO eg

f^ — -

r-~

r^ • —

CO eg

CO CN

h- CM

in
in

CD 00
•«1- CM

'^ eg

in ---

00 CM

CO
CO

»- eg '-

05
00

1^
eg

'- in r-~ 05

t^ t^ t^ t^

I I I I

in CM CD CO o '-

CD r^ r- h~ CO 00

05 05 05 05 05 05

-C5 _:

(0 Q)
1- O)

m tfi

> <D
>, >

™ §>

en to

CO ®
O S
— at
1^ —
..LU

^5.

S,|S

O _:_; 52

o S 5
5 i "P
eo e —

§5 £

>3 CO

I II S
3 — £

■^ £ c

^ .<» m

C ! CL

-O T- C

^ — ■-

o c <"

^ o £

™ B-g

M 2 ^

^^ o

10 o

n £

c o
o =:

£ 2 o o

■D 5

s —

03 to

o) 2

i? *

o o

(0 o

CM £

QC c
O cj)

<0 IS

I- re
QC t>j
CD CC

-§

. . Q.
«I

o *^

re $

T3 Q.
? I

£^

re 0)
> 2

05
0)

(T>

<i>

<1)

(ft

■n

(Tl

•V

■0

■0
c

^
0

344

KIRKLEY AND SQUIRES: CAPITAL STOCK AND INVESTMENT IN A FISHERY

Atlantic Fisheries and increased fuel prices be-
tween 1972 and 1975; 3) the formulation and im-
plementaiion of the Fisheries Conservation and
Management Act in 1976 and 1977; 4) extremely
restrictive regulations in 1978 and 1979; and 5) a
long dock strike in New Bedford, frozen harbors,
and foreign purchases from American fishermen
in 1981. In addition, there was substantial entry
of new steel-hulled vessels between 1975 and
1977 and 1979 and 1980.

Chow (1960) tests of the equality of regression
equations were conducted to further examine the
possibility that the statistical relationships be-
tween cost and vessel characteristics changed
over time. The Chow test is an F-test of the stabil-
ity of the coefficients (Maddala 1977). The results
of these tests are presented in Table 3.

The results of the tests indicate that the esti-
mated relationships between cost and character-
istics were not the same for the selected time peri-
ods. Alternatively, the hypothesis of the equality
of the regression equations for different periods of
time could not be accepted at any reasonable level
of significance. However, these results only verify
that different models should be estimated for dif-

Table 3. — Results of tests for the equality of regression
equations.

Periods

Critical values'

tested

F-statistic2

0.05

0.01

1965-71

1972-75

1976-77

1978-79

1980-81

39.38

'^12,1050 = ''•^5

f^12,1050 =

2.18

1965-71

1972-75

8.45

1^12.325 = ■'•'75

f^12,325 ^

2.18

1965-75

1976-81

20.55

f^12.1086 = ''•^5

f^12.1086 =

2.18

1965-75

1976-79

21.33

1^12.668 = ■'■''5

'^12.668 =

2.18

1976-77

1978-79

3.60

'^12,329 = l-^S

'^12,329 =

2.18

1976-77

1978-81

3.22

^12.737 = l-^S

f^1 2.737 =

2.18

1976-79

1980-81

3.35

f^12.737 = ■'•75

f^12.737 =

2.18

1980

31981

2.76

f^12,384 = "I -^5

f^l2,384 =

2.18

'F-statistIc for denominator degrees of freedom equal to infin-
ity.

^Restricted residual sum of squares obtained by pooling data
over all time periods and estimating Equation (2). Unrestncted
residual sum of squares obtained by estimating Equation (2) for
each penod being examined and then adding the residual sum
of squares for each equation.

^Residual sum of squares obtained from estimating Equation
(2) without the time vanable

ferent periods of time. They are not conclusive
proof of the selection of the years for a particular
group of years. A more accurate determination of
the years to be included in each group requires
considerably more estimation and hypothesis
testing which is beyond the intent of this study.

As indicated by the ^-statistics in Tables 1 and
2, the age of the vessel, whether or not the vessel
is new or used, and the size characteristics appear
to be the more statistically significant explana-
tory variables. These results are consistent with
the results of Griffin et al. (1978). The statisti-
cally significant negative coefficients for the size
characteristics during the years 1972-75 are par-
ticularly interesting. These years coincide with
the Arab oil embargo when fuel prices increased;
expected increases in future operational costs
may have deterred new entry. The results also
indicate that the value of a vessel declines as it
becomes older. The coefficient for hull construc-
tion, CS, suggest that steel-hulled vessels are
more expensive than are those of other materials
such as fiberglass, ferro cement, wood, and alu-
minum. There does not appear to be a difference
in the value of a vessel based on gear type. This
may be a result of specification problems with the
dummy variable for gear t)^e (see footnote 7).

It is of further interest to examine the elastic-
ities of cost with respect to the vessel characteris-
tics. ^° These are presented in Table 4. As indi-
cated, vessel tonnage, length, and age have the
greatest influence on the acquisition price. The
elasticity for length is consistent with the manner
in which vessels are sold; that is, vessel prices are
largely stated in terms of their length.

There are, however, inconsistences in the esti-
mated capital stock equations and elasticities. In

'0 Elasticities indicate the p)ercentage change in cost result-
ing from a 1% change in the value of a characteristic.

Table 4. — Estimated elasticities of cost with respect to
vessel characteristics.'

Elasticities

Year

GRT2

HPWR

AGE

1965-71

0.43

0.06

2.88

-0.32

1972-75

0.78

-0.04

-0.08

1976-77

1.14

0.09

0.11

-0.11

1978-79

-0.14

0.30

2.38

-0.40

1980

0.16

0.30

1.43

-0.30

1981

-0.64

0.56

2.38

-0.41

' Elasticities based on nominal estimates and observed mean
values of cost and vessel charactenstlcs.

2GRT is gross registered tonnage; HPWR is engine horse-
power; LN is lengtti; AGE is age ol vessel.

345

FISHERY BULLETIN: VOL. 86, NO. 2

particular, not all of the estimated coefficiencts
have the desired or expected sign. For example,
the coefficient for tonnage in the 1981 estimate is
negative and that for the square of tonnage is
positive. In contrast, a positive coefficient for ton-
nage was expected since vessel cost should in-
crease with vessel tonnage, and the coefficient for
tonnage squared was expected to be negative be-
cause vessel costs are believed to increase at a
decreasing rate.

The incorrect signs could be the result of multi-
collinearity between vessel characteristics. Aux-
iliary regressions, though, do not indicate severe
multicollinearity. The R^ values are <0.35. If
multicollinearity is a problem, its major effect is
on the estimated relationships between cost and
individual characteristics, the estimated elastic-
ities, and the previously discussed Chow tests.

Multicollinearity is not thought to be a problem
for several reasons. First, Chow tests in the pres-
ence of multicollinearity usually result in the ac-
ceptance of the null hypothesis of the equality of
regression equations or that differences in the es-
timated coefficients are statistically insignificant
(Maddala 1977). Second, multicollinearity does
not pose a problem for estimating capital stock
and investment; the estimated equations still per-
mit estimation of the conditional mean.^^

11 For additional information on the problems of multi-
collinearity with respect to analysis and prediction, see Kmenta
(1971).

The equations for the real or deflated value of
capital were used to estimate total capital stock
and investment for the New England fleet be-
tween 1965 and 1981. The real capital stock equa-
tions were applied to vessel data available from
the Northeast Fisheries Center, and total capital
stock was calculated as the sum of the stock over
all vessels. Aggregate net investment was calcu-
lated as the difference in total capital stock be-
tween consecutive years.

Prior to estimating total capital stock and in-
vestment, it was necessary to define vessel type
by gear to avoid double counting which might
occur since vessels frequently switch port and
gear. Three categories of gear types were estab-
lished: 1) otter trawl, 2) scallop dredge, and 3) all
others, which include lobster trawl and pots, her-
ring gear, harpoons, etc. Vessels were assigned a
gear type based on a plurality of days at sea. Esti-
mates of capital stock and investment are pre-
sented in Table 5.

Several limitations should be considered when
evaluating the estimates. First, there is the previ-
ously discussed problem that estimates of capital
stock may be biased or overestimated. Second,
estimates are only for New England vessels or
those which are believed to be homeported in New
England. Third, the capital stock and investment
series pertain to New England and not the United
States; that is, estimates should not be inter-
preted as net changes in the capital stock and
investment in U.S. fisheries. Fourth, under-

Table 5. — Estimates of real capital stock and investment in New England harvest-
ing sector, 1965-81.1

Number

of
vessels

Capital stock

Investment

Index

investment

Year

Trawl

Dredge

Other

Trawl

Dredge

Other

- $ Millie

inc --_--.

Mio ■

1965

594

25.96

16.29

1.98

1966

619

28.84

14.75

2.82

2.88

-1.54

0.84

100

1967

628

39.37

11.42

1.63

10.53

-3.33

-1,19

276

1968

610

28.65

15.43

1.68

-10.72

4.01

0.05

-301

1969

602

29.22

12.92

1.29

0.57

-2.51

-0.39

-107

1970

607

25.61

8.29

1.37

-3.61

-4.63

0,08

-374

1971

620

23.65

12.80

1.47

-1.96

4.51

0,10

122

1972

655

27.01

7.74

4.70

3.36

-5.06

3,23

1973

666

28.42

7.12

5.13

1.42

-0.62

0,43

1974

695

21.26

4.19

6.40

-7.17

-2.93

1,27

-405

1975

737

29.04

5.15

8.48

7.78

0.96

2.08

496

1976

783

20.73

7.72

8.16

-8.31

2.57

-0.32

-278

1977

836

28.86

14.19

7.86

8.13

6.47

-0.30

656

1978

881

37,45

12.85

8.65

8.59

-1.34

0,79

369

1979

1,107

45.41

27.30

11.25

7,96

14.45

2,60

1147

1980

1,260

52.26

34.30

11.69

6.85

7.00

0,44

656

1981

1,246

43.35

30.15

11.60

-8.91

-4.15

-.09

-603

'Real Investment and capital are deflated with producer pnce Index for tieavy machinery,
1967 = 100.

346

KIRKLEY AND SQUIRES: CAPITAL STOCK AND INVESTMENT IN A FISHERY

tonnage vessels or those which are <5 gross
registered tons are excluded since data are
unavailable; thus, total capital stock and invest-
ment in New England fisheries are underesti-
mated.

Despite these limitations, the results provide
information on investment and capital stock in
New England fisheries which has not been avail-
able. As is characteristic of open-access fisheries,
investment in New England fisheries appears to
have increased over time. This is quite evident
subsequent to the passage of the Magnuson Fish-
eries Conservation and Management Act of 1976.
While possibly true, the observation that invest-
ment increased when the act was passed should
not be construed to imply that the act was respon-
sible for increased investment. Positive expecta-
tions regarding fish stocks or economic conditions
could also be responsible for increases in capital
stock and investment.

The largest increase in investment occurred in

1979 which coincided with the greatest influx of
vessels. It also coincided with increased fish
stocks in 1978 and 1979 and high vessel profits of
the previous year. In 1979, 226 vessels entered
the New England fisheries; 92 vessels were newly
constructed.

Disinvestment also occurred, but the reasons
are unknown. ^^ Disinvestment may be associated
with vessel sinkings, foreclosures on vessel mort-
gages, and exit from the region. Alternatively, if
vessel prices reflect the expected net discounted
present value of earnings from a fishery, the dis-
investment or even changes in investment may
be due to changes in economic expectations. Dis-
investment occurred between 1968 and 1970,
1974, 1976, and in 1981.

Disinvestment between 1968 and 1970 may re-
flect expectations of declining future earnings.
Disinvestment in 1974 may reflect the effects of
high fuel prices. Disinvestment in 1976 and 1981
may be associated with fishermen's expectations
of declining fish stocks, the effects of manage-
ment, and high fuel prices.

The reasons for investment are also unknown.
Increasing levels of investment between 1977 and

1980 may be due to economic expectations associ-
ated with fisheries management, fish stocks, and
future earnings.

The real capital stock increased approximately
6% per year between 1965 and 1981. The largest

increase occurred in 1978 and 1979 when there
was a substantial increase in the number of new
steel-hulled vessels and entry of vessels from the
mid-Atlantic region. Capital stock declined in
1981 when some of the mid-Atlantic vessels left
New England and the stocks of fish, particularly
sea scallops, declined (Northeast Fisheries Cen-
ter 1985).

The results also indicate why a vessel count
should not be used to indicate capital stock or
investment. The number of vessels increased be-
tween 1969 and 1970 and 1975 and 1976, but
investment decreased in 1970 and 1976 (Table 6).
Alternatively, the number of vessels increased in
each year relative to 1965, but the real capital
stock between 1969 and 1976 declined with re-
spect to 1965.

Table 6. — Indices of capital stock based on constant dollar value
and vessel count.

Capital stock

Constant dollar value

Vessel count

Year

Trawler!

Dredge'

Total

Trawler!

Dredge!

Total

1965

100

1966

111

105

102

104

1967

152

119

109

106

1968

110

103

104

106

103

1969

113

104

106

101

1970

108

102

1971

108

104

1972

104

108

110

1973

109

105

112

1974

107

117

1975

112

105

124

1976

109

202

132

1977

111

115

107

202

141

1978

144

133

116

189

148

1979

175

168

190

137

313

186

1980

201

211

222

158

440

212

1981

167

185

192

158

413

210

i2Hirshleifer (1970) provided additional information about
disinvestment.

'Gear type assigned by plurality of days absent.

The inadequacy of using vessel count is more
pronounced when comparing the indices based on
gear type. As shown in Table 6, the number of
trawl vessels in every year exceeded the number
of vessels in 1965 while capital stock was less
than that of 1965 for four years. The two scallop
dredge indices also indicate dissimilar changes
between vessel count and real capital stock. It
may, though, be possible than an index based on
vessel count and weighted by vessel size would
more closely compare to the real capital stock
index. The construction of these indices would re-
quire considerable additional analysis and deter-
mination of the weights necessary for aggrega-
tion.

347

FISHERY BULLETIN; VOL. 86, NO. 2

SUMMARY AND CONCLUSIONS

This study developed procedures for estimating
the capital stock and net investment using only
the information generally available. That is, data
on acquisition and list prices for some but not all
vessels and characteristics for most vessels in a
fleet were used to estimate capital stock and in-
vestment in a fishery. An hedonic approach which
specified the acquisition price as a function of ves-
sel characteristics was suggested as a possible
method for obtaining information necessary for
estimating capital stock and net investment. It
was argued that the approach was consistent with
procedures used by economists to estimate the
capital stock and investment in traditional indus-
tries.

Estimates were based on 1,110 observations ob-
tained from NMFS and classified advertisements
for the period 1965-81. Vessel characteristics
pertaining to size and age were shown to be the
more significant characteristics for explaining
the value of a vessel. The corresponding informa-
tion was then used to estimate the capital
stock and investment in the New England fish-
eries.

An interesting result was that although invest-
ment increased over time, there also was disin-
vestment in the fisheries. It was suggested that
this was possibly the result of vessel sinkings,
exit of vessels from the region, foreclosures, and
expectations of declining future net returns.

It also was demonstrated that a vessel count
should not be used to indicate capital stock or
investment. The number of vessels generally in-
creased over time, but the level of capital stock
did not coincide with these changes. This suggests
that more attention should be given to developing
economic measures of capital stock and invest-
ment.

A question still in need of attention, however,
cannot be answered from this study. That is,
given the level of investment, can an optimal uti-
lization and allocation of resources be deter-
mined. The answer is clearly no. This requires
management authorities to specify the objectives
of fisheries management and a detailed bioeco-
nomic model.

There is a need for more research on invest-
ment, and in particular, the determination of the
optimum rate of investment. This includes re-
search on the social discount rate, reasons for in-
vestment, and the marginal productivity of capi-
tal. More important, additional research is

required to better address the issues of fisheries
management, particularly whether or not public
expenditures should be allocated to managing
fisheries.

ACKNOWLEDGMENTS

Comments by William DuPaul, Wade Griffin,
Ray Hilborn, Maurice Lynch, Ivar Strand, and
two anonymous reviewers are gratefully ac-
knowledged. Any errors remain the responsibil-
ity of the authors. The article does not necessarily
reflect the opinion of either the National Marine
Fisheries Service or the Virginia Institute of
Marine Science.

LITERATURE CITED

Baumol, W J

1977. Economic theory and operations analysis.
Prentice-Hall, Englewood Cliffs, N.J.

Blackorby, D., D Primont, and R R. Russel

1978. Duality, separability, and functional structure: the-
ory and economic applications. North-Holland Publish-
ing Company, Amsterdam.

BocKSTAEL, N. E., W. M Hannemann, and I. E. Strand.

1986. Measuring the benefits of water quality improve-
ments using recreation demand models. Volume II of ben-
efit analysis using indirect or imputed mtirket methods.
EPA contract No. CR-811043-01-0. U.S. Environmen-
tal Protection Agency, Washington, D.C.

Box, G E P , and D R Cox

1962. An analysis of transformations (with discussion).
J. R. Stat. Soc, Ser. B:21 1-243.

Braeutigam, R R., a. F Daughety, and M. A. Turnqutst.

1982. The estimation of a hybrid cost function for a rail-
road firm. Rev. Econ. Stat. 64:394^04.

Branson, W H

1972. Macroeconomic theory and policy. Harp)er & Row,
Publishers, N.Y.
Charles. A. T.

1983a. Optimal fisheries investmentxomparative dy-
namics for a deterministic seasonal fishery. Can. J.
Fish. Aquat. Sci. 40:2069-2079.
1983b. Optimal fisheries investment under uncertainty.
Can. J. Fish. Aquat. Sci. 40:2080-2091.
Chow, G C

1960. Tests of equality between subsets of coefficients in
two linear regressions. Econometrica 28:591-607.
Clark, C W., F. H. Clarke, and G R Munro.

1979. The optimal exploitation of renewable resource
stocks: problems of irreversible investment.
Econometrica 47:25-47.

Cunningham, S., M. R. Dunn, and D. Whitmarsh.

1985. Fisheries Economics: an introduction. St. Mar-
tin's Press, N.Y.
Dadkhah, K. M

1984. Confidence interval predictions from a logarithmic
model. Rev. Econ. Stat. 66:527-528.
Dewar, M. E.

1983. Industry in trouble: the Federal government and

348

KIRKLEY AND SQUIRES: CAPITAL STOCK AND INVESTMENT IN A FISHERY

New England fisheries. Temple University Press,
Philadelphia, PA.
DOERINGER, P B.. P I Moss, AND D. G. TERKLA.

1986. The New England fishing economy. The Univer-
sity of Massachusetts Press, Amherst, MA.
Gordon, H S.

1954. The economic theory of a common property re-
source: the fishery. J. Polit. Econ. 62:124-142.
Griffin, W. L.J P Nichols, R. G. Anderson. J E Buckner. and
C M Adams.

1978. Costs and returns data: Texas shrimp trawlers.
Gulf of Mexico 1974-75. Texas A&M Univ., Dep. Agric.
Econ. tech. Rep. TAMU-SG-76-601.
Griuches, Z

1971. Hedonic price indexes for automobiles: an econo-
metric analysis of quality changes. In Price indexes and
quality changes, p. 103-130. Harvard University
Press, Cambridge, MA.
HILBORN. R.

1985. Fleet dynamics and individual variation: why some
people catch more fish than others. Can. J. Fish. Aquat.
Sci. 32:2-13.

Hirshleifer. J.

1970. Investment, interest and capital. Prentice-Hall,
Englewood Cliffs, NJ.

Kendrick, J

1961. Some theoretical issues in capital measurement.
Am. Econ. Rev. 51(2):102-111.
KMENTA, J.

1971. Elements of econometrics. Macmilliam Publish-
ing Co., Inc., N.Y.

KiRKLEY, J E , M P Pennington, and B. E. Brown.

1982. A short-term approach for analyzing the effects of
harvesting quotiis. J. Cons. int. Explor. Mer 40:173-
175.
KiRKLEY, J. E

1986. The relationship between management and the
technology in a multi-species fishery. Ph.D. Thesis,
Univ. Maryland, Collage Park, MD.

Ladd, G W . AND M B Martin.

1976. Prices and demands for input characteristics. Am.
J. Agric. Econ. 58:21-30.

Lau, L J

1978. Applications of profit fiinctions. In M. Fuss and
D. McFadden (editor), Production economics: a dual ap-
proach to theory and applications, p. 133-216. North-
Holland Publishing Company, Amsterdam.

Maddala. G S.

1977. Econometrics. McGraw-Hill Book Company, N.Y.
National Academy of Sciences

1979. Measurement and interpretation of productivity.
National Research Council, Washington, D.C.

Northeast Fisheries Center.

1985. Status of the fishery resources off the northeastern
United States for 1985. U.S Dep. Commer., NCAA
Tech. Memo. NMFS-F/NEC-42, 137 p. Northeast Fish-
eries Center, National Marine Fisheries Service, NOAA,
Woods Hole, MA.

Rosen, S.

1974. Hedonic Prices and implicit markets:product differ-
entiation in pure competition. J. Polit. Econ. 82:34-55.
Scorr, A.

1979. Development of economic theory of fisheries regula-
tion. J. Fish. Res. Board Can. 36:725-741.
SISSENWINE, M P , and J E. KiRKLEY

1982. Fishery management techniques:practical aspects
and limitations. Mar. Policy 6:22-42.
Sloan, H S., and A. J. Zurcher.

1968. Dictionary of economics. Barnes & Noble, Inc.,
N.Y.
Smallwood, D M., and J. R. Blaylock.

1986. Forecasting performance of models using the Box-
Cox transformation. Agric. Econ. Res. 38:14-24.

Tettey, E O . W L. Griffin, J. B. Penson, and J. R Stole.

1986. Implications of tax policy on investment in a com-
mon property resource. N. Am. J. Fish. Manage. 6:100-
104.
Triplett, J

1986. The economic interpretation of hedonic methods.
Surv. Curr. Bus. 66:36-40.
Varian, H. R.

1984. Microeconomic analysis. 2d, ed. W. W. Norton &
Company, N.Y.

349

ANALYSES OF THE RELATIONSHIP BETWEEN

THE DISTRIBUTION OF SEARCHING EFFORT, TUNA CATCHES, AND

DOLPHIN SIGHTINGS WITHIN INDIVIDUAL PURSE SEINE CRUISES

Tom Polacheck^

ABSTRACT

The fine scale distribution of searching effort within individual purse seine cruises in the eastern
tropical Pacific is analyzed in relationship to sightings of spotted dolphin, Stenella attenuata, and
tuna catches. The data for these analyses were derived from detailed observations made by National
Marine Fisheries Service observers aboard U.S. purse seiners. A clustering algorithm is developed
which separates the activity of a vessel into areas where sets were common and areas where they are
infrequent. Within clusters of high set densities, vessels tend to concentrate their searching effort.
Vessels searched proportionately greater distances relative to the physical distances between sets
while within clusters than when outside clusters. Encounter rates with schools of spotted dolphins
tend to be either much higher or much lower within defined clusters than outside them. Clusters with
low encounter rates were clusters in which non-dolphin associated tuna sets predominated. Because
of this dichotomy in the magnitude of the dolphin encounter rates within clusters, overall encounter
rates appeared to have relatively small biases if the concentration of searching effort within clusters
is ignored. The average catch of tuna per set was higher within the defined clusters than between
them. The overall results suggest that fine scale geographic effects need to be considered when using
data from purse seiners to examine changes in relative abundances of either dolphins or tuna.

Catch and effort data underlie most indices of
abundances used for assessing the status of com-
mercially exploited fish stocks. The validity of
using catch and effort data from commercial har-
vests has long been questioned because of the
likelihood that fishermen concentrate their effort
in areas of high fish densities (Helland-Hansen
1909). Yet, almost nothing is known about the
allocation of effort by individual vessels. If data
are available on the activity of individual vessels,
these data have been mainly used for standard-
ization of effort. The catch and effort activities
within individual cruises have not been examined
in detail (in part because the data for such an
examination generally do not exist). The purpose
of the present paper is to examine the searching
behavior of tuna purse seiners in the eastern trop-
ical Pacific (ETP) based on detailed data compiled
by the National Marine Fisheries Service
(NMFS). The main questions addressed are
whether seiners concentrate their effort and what
is the relation between searching behavior, en-

iSouthwest Fisheries Center La Jolla Laboratory, National
Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA
92038 and Department of Biology, University of Oregon, Eu-
gene, OR 97403; present address: Northeast Fishery Center
Woods Hole Laboratory, National Meirine Fisheries Service,
NOAA, Woods Hole, MA 02543.

Manuscript accepted January 1988.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.

counter rates for dolphins, and tuna catches.

The maximization of profit is presumably a
strong influence on the behavior of commercial
fishermen. Upon leaving port, the catch rate real-
ized by a fisherman is probably the most impor-
tant factor affecting his profits. Given this orien-
tation, it is reasonable to assume that fishermen
have developed strategies of when and where to
fish that increase their catch rates beyond that
achieved by random search. There is little empir-
ical information to support this assumption other
than correlations, which have been noted for
some fisheries, between the spatial distribution of
catch rates and effort for the fleet as a whole (Gul-
land 1955; Calkins 1963).

In a fishery where the detection of fish depends
upon visual cues, searching would be expected to
be located in the vicinity of previously located
fish, if there is a tendency for the underlying pop-
ulation to be spatially clustered. The search path
for a vessel in such a fishery might be expected to
look something like the hypothetical one depicted
in Figure 1. The amount of crisscrossing or
zigzagging in the vicinity of a catch and the area
over which the search extends would be expected
to vary between fishermen. The solution to the
optimal searching strategy for such a situation is
nontrivial and depends upon information on the

351

FISHERY BULLETIN; VOL. 86, NO. 2

Figure l. — A hypothetical cruise track for a vessel searching
for clustered prey. The location of catches are indicated by
xs.

underlying distribution of the fish population
(Koopman 1980). The question of optimal search-
ing strategies for fishermen has been receiving
increased attention (Pazynich 1966; Salia and
Flowers 1979; Clark and Mangel 1983). These
studies are primarily theoretical at this time and
their application to actual fisheries requires
knowledge of the spatial distribution of the fish
population.

Locating schools of tuna (e.g., yellowfin, Thun-
nus albacares, and skipjack, Katsuwonus
pelamis) in the purse seine fishery in the ETP
depends on visual cues. Fishermen use a variety
of cues including birds which feed on the same
prey as tuna, disturbances on the surface of the
water, floating debris which frequently have as-
sociated tuna, and schools of dolphins which are
often associated with tuna (primarily yellowfin).
Fishermen have names for the different types of
sets depending upon what is associated with the
tuna school. They refer to sets associated with
floating debris as log sets, sets associated with
dolphins as porpoise sets, and sets not associated
with other animals (except possibly birds) as
school sets. Fishermen when not engaged in a set

usually spend their day actively searching for
signs of tuna. They use 25 x binoculars to scan the
water while the boat cruises at speeds generally
between 10 and 12 knots.

The distribution of schools of tuna, as well as
schools of the most commonly associated dolphins
(i.e., the spotted dolphin, Stenella attenuata), ap-
pears to be spatially and temporally clustered
within the ETP considered as a whole (Calkins
and Chatwin 1967, 1971; Blackburn and Wil-
liams 1975; Suzuki et al. 1978; Au et al. 1979^;
Polacheck 1983). At finer geographic scales, there
is little available information although Au et al.
(fn. 2) suggested that schools of spotted dolphin
tend to be locally concentrated in areas of conver-
gences and fronts.

Given the above observation, it is not surpris-
ing that the detection or encounter process for
tuna or dolphins does not, in general conform to a
Poisson process when the distribution of search-
ing times, searching distances, or physical dis-
tances between nearest encounters are analyzed
(Polacheck 1983; Allen and Punsely 1984). In
such analyses it is impossible to separate or dis-
tinguish the effects of nonrandom search from
nonrandom distributions of tuna or dolphins.

In the harvesting of tuna schools associated
with dolphins, fishermen chase and capture the
associated dolphins (Perrin 1968, 1969), and some
dolphins may be incidently killed. NMFS, as part
of its responsibility under the Marine Mammal
Protection Act of 1972 for managing and monitor-
ing the status of dolphin populations, placed
trained observers aboard tuna purse seiners.
From the data collected by these observers, ap-
proximate cruise tracks can be drawn by connect-
ing all positions that were recorded. Many of
these approximate cruise tracks (e.g.. Figure 2)
have superficially a strong similarity to the hypo-
thetical one depicted in Figure 1. It was this sim-
ilarity that provided the impetus for the analyses
presented below.

METHODS

NMFS observers aboard tuna purse seiners col-
lected a wide variety of information, both on the
sightings of marine mammals and fishing opera-

2Au, D. W. K., W. L. Peryman, and W. F. Per-
rin. 1979. Dolphin distribution and the relationship to envi-
ronmental features in the eastern tropical Pacific. Natl. Mar.
Fish. Serv., Southwest Fish. Cent., Adm. Rep. LJ-79-43, 59 p.

352

POLACHECK: DISTRIBUTION OF SEARCHING EFFORT

Figure 2. — Examples of the approximate cruise track for two tuna purse seiners. Diamonds represent the
location of a set or chase. No geographic coordinates are given, and the orientation of figures were rotated at
random so as not to compromise any proprietary fishing information. The distance between tick marks equals
300 nautical miles.

353

FISHERY BULLETIN: VOL. 86, NO. 2

tions. Included in the recorded data are informa-
tion on all changes in a vessel's activity; the loca-
tion, tjqae, and catch of all purse seine sets; and
the location and identity of all marine mammal
sightings. In these data, a vessel's activity is clas-
sified into one of five mutually exclusive cate-
gories: searching, running, setting, chasing, or
resting. Searching is defined to be whenever a
vessel is moving and the crew are actively search-
ing for signs of tuna; running, anytime the vessel
is moving but not actively searching for signs of
fish (e.g., moving locations at night); chasing,
anytime schools of dolphins are being pursued
before the net has begun to be set. More detailed
descriptions of the available data, collection pro-
cedures and their preparation for the analyses
below can be found in Polacheck (1983, 1984^).
The analyses in this paper were part of a larger
project on the use of these observer data for as-
sessing the relative abundances of dolphin stocks.
As such, the emphasis in this paper is on the
encounter rate for the most important dolphin
species for the fishery (spotted dolphin), although
catch rates for tuna are also considered. The re-
sults presented in this paper are based on two
different approaches for analyzing the data. The
first method is a set of nearest neighbor calcula-
tions, and the second is a cluster analysis.

The nearest neighbor calculations were per-
formed in order to get an indication whether ves-
sels tend to search in the vicinity of a previous
encounter (either a sighting of marine mammals
or a set on tuna). In these calculations, the phys-
ical distance between either the next or preceding
encounter is compared with the distance to the
nearest other encounter made within the entire
cruise. Also, the proportion of times in which the
nearest encounter is not either the next or preced-
ing one is calculated. For a vessel that never re-
turned to the area of an encounter, this propor-
tion would equal 1. Similarly, if a vessel never
returned to the area of an encounter, the ratio of
the distance between either the next or preceding
encounter and the distance to the nearest other
encounter within an entire cruise would also
equal 1. Note that the expected values for these
proportions with random search are not necessar-
ily 1. The expected value will be dependent both
on the distribution of potential encounters and

3Polacheck, T. 1984. Documentation of the time sequen-
tial files created ft-om the tuna boat observer data bases for
analyzing relative abundances. Natl. Mar. Fish. Serv., South-
west Fish. Cent., Adm. Rep. LJ-84-33, 26 p.

the definition of random search (see Discussion).
These calculations were performed separately for
sets and chases for tuna and for the sightings of
spotted dolphin. In performing these nearest
neighbor calculations, the first and last encounter
during a cruise were not included.

The other main approach used for examining
the data is a form of cluster analysis. When the
sequences of distances between sets and chases
within any cruise were examined, they appeared
to be spatially and temporally clustered in the
sense that sets and chases in which the distance
to the next set or chase was small tended to be
clumped sequentially. This observation led to the
development of an algorithm for clustering sets
and chases that were spatially and temporally
related. Standard clustering algorithms were not
appropriate in this situation because of the prob-
lem of scaling spatial and temporal distances
within a common metric (i.e., how much time
should be equal to a given distance).

Note that the term "clustered" or "clustered
distribution" is used in this paper to refer to any
distribution in which high- and low-density areas
are more frequent than would be expected if the
distribution was generated by a Poisson process.
The term is not meant to refer to any particular
nonhomogeneous process. A cluster is considered
as an area of high density and should not be con-
strued as referring to a discrete unit.

The primary purpose of the clustering al-
gorithm was to define areas which a fisherman
might have thought to have a high density of
potential fishing targets so that the searching be-
havior of a vessel could be compared between
these areas and outside them. This analysis ex-
ploits the fact that the physical distance between
events is partially independent of the distance
that a vessel travels to locate them. Since the
purpose of the algorithm was to define areas of
potentially good fishing, chases of dolphin, as well
as sets, have been included as events in the clus-
tering algorithm. (Sets made for the purpose of
washing the net were not used.) The clustering
algorithm began with consideration of the dis-
tance between the first and second set and/or
chase. If this distance was less than a specified
amount, then these two events were placed in the
same cluster, and the distance between these two
and the third event were examined. This specified
amount will be referred to as the clustering
parameter. If the distance between the third
event and either of the events within the cluster
was less than the value of the clustering parame-

354

POLACHECK: DISTRIBUTION OF SEARCHING EFFORT

ter, this third event was included in the cluster
and the fourth event was examined. Whenever a
set or chase was found for which the distance be-
tween it and all members of the last defined clus-
ter exceeded the clustering parameter, a new
cluster was formed. This process was repeated
until all sets and chases within a cruise were
placed in a cluster.

Using this algorithm, all the activities within a
cruise could be considered to occur either between
or within clusters. Isolated sets or chases (i.e.,
clusters containing only a single event) were con-
sidered as occurring between clusters. Distances
travelled (i.e., distance searched and distance
run) within a cluster were defined as the dis-
tances travelled after the first set or chase until
the last set or chase in that cluster. Distances
travelled between clusters were defined from the
last event of the previous cluster to the first event
of the subsequent cluster. The distances travelled
until the location of the first set or chase and after
the last one were not included because of the large
distances involved in reaching the fishing
grounds.

The location of a cluster was estimated by cal-
culating the centroid for all sets and chases
within it. The size of a cluster was estimated by
determining the radius of the smallest circle with
a center at the centroid that encompassed all sets
and chases within it.

The sensitivity of this algorithm to the value of
the clustering parameter was examined for val-
ues of 50, 75, 100, and 150 miles. For most of the
results, only clusters with at least three members
are considered as clusters. Clusters with only two
members have been excluded from most of the
summary results describing a cluster and also in
the comparisons of results between and within
clusters. This was done because two physically
close events did not seem to warrant being called
a cluster. Yet, given the relative difficulty in lo-
cating potential sets, two close events might be
considered as areas of potentially good fishing.
Clusters with only two members contained 18% of
all sets and chases when the cluster parameter
equalled 50 miles and 6% of all sets when the
value equalled 150 miles. The overall results and
conclusions are robust to whether or not clusters
with two members are included or excluded.

An average intercluster distance for a cruise
was calculated in order to get an indication of the
stability of the clustering algorithm to the value
of cluster parameters. The intercluster distance
was defined as the distance from the nearest

member of a cluster or isolated set to the next set
and represents the minimum value that the clus-
ter parameter would have to be for a cluster or
isolated set to be combined in a single cluster with
the next set.

Encounter rates for schools of spotted dolphin
for each cruise were calculated between and
within clusters as the total number of sightings
divided by the total distance searched. In these
rates, if the first chase or set within a cluster was
based on a sighting of spotted dolphin, this sight-
ing was included in the encounter rates between
and not within a cluster. For the analysis of these
encounter rates, clusters were classified accord-
ing to the percentage of the total number of sets
and chases within a cluster that involved schools
of dolphins. In calculating the number of events
that occurred within a cluster, sequential non-
dolphin sets in which no searching was done be-
tween them were counted as a single event. This
was done to reduce the effect of multiple sets on
the same floating object counting as a large clus-
ter.

In order to see whether the searching behavior
within the defined clusters resulted in biased esti-
mates of encounter rates for dolphins if the clus-
ters were ignored in the estimates, two different
estimates for the overall encounter rate for a
cruise were calculated and compared. The first
estimate, which will be referred to as the unad-
justed rate, was simply the total number of en-
counters divided by the total distance searched
for an entire cruise. This would be an unbiased
estimate if search was in fact random. The second
estimate, which will be referred to as the adjusted
encounter rate, was calculated as the weighted
average of the encounter rate within and between
clusters. The weights for the encounter rate
within clusters were equal to the diameter of the
cluster. The weights between clusters equalled
the total distance searched between clusters. In
effect, this adjusted encounter rate is an estimate
of what the encounter rate would have been if a
vessel had made a straight line crossing of each
cluster. (In the calculation of these adjusted en-
counter rates, clusters with two members were
treated the same as other clusters to simplify the
calculations.)

The analyses in this paper are based on 35
cruises. This represents a subset of the cruises in
1979 with NMFS observers, which in turn is a
subset of all purse seine cruises for tuna in the
ETP during 1979. Analyses were restricted to
1979 because only for this year have the data

355

FISHERY BULLETIN: VOL. 86, NO. 2

been carefully edited for positional errors.
Cruises from 1979 were excluded either because
of insufficient positional data or because the ves-
sel made port stops during the middle of the
cruise. Inclusion of the time spent searching on
the way in and out of port could distort the results
on searching and encounter rates between clus-
ters. Preliminary analyses suggested that there is
little difference in the results between vessels
that went into port and those that did not.
However, to avoid further complicating the
analyses, these cruises have not been included
(see Polacheck 1983 for more detail).

RESULTS

Is Effort Concentrated?

The percentage of sets or chases for which the
nearest one was not the preceding or next set or
chase ranged from 12 to 77 among cruises (Fig. 3).
The mean percentage was 41 (SE = 1.9, n = 35).
The average ratio of the physical distances be-
tween the next or preceding set or chase com-
pared to the distance to the nearest set or chase
within a cruise was 1.45 (SE = 0.044, n = 35) and
ranged from 1.00 to 2.24. Consideration of the
same statistics for the distances between sight-
ings of spotted dolphin indicates an even more
concentrated pattern. The mean percentage of
sightings for which their nearest neighbor was
not either the next or preceding one equalled 80
(SE = 2.8, n = 34, note one cruise recorded no
sightings of spotted dolphins) and ranged from 25
to 100 (Fig. 4). The average ratio of the distance
between the next or preceding sighting compared
with the nearest sighting was 4.05 (SE = 0.261,
n = 34) and ranged from 1.14 to 7.82. For these
sighting statistics, low percentages and ratios
near 1.00 are found in vessels with few sightings
(Fig. 4). These results suggest that in general ves-
sels return to the area of a previous sighting and/
or set and search in that area at least 41% of the
time.

There appears to be large differences among
vessels in their ability and success at locating
potential fishing targets. Thus, the average dis-
tance searched between sets or chases varies by
about a factor of 4 among cruises, while the aver-
age physical distance between sets or chases
varies by about a factor of 6 (Fig. 5). There is little
relationship between the average distance
searched between sets and chases and the aver-
age distance to the next one. However, vessels

>
u

U4
3
O
m
a
u.

0 I 1 1 1 1 1 1 1

10 20 30 40 50 60 70 80

PERCENT OF THE SETS AND CHASES

IN WHICH THE NEAREST ONE WAS

NEITHER THE PRECEDING OR NEXT ONE

Figure 3. — The frequency distribution for the percentage of the
sets and chases within a cruise for which the nearest other set
or chase was neither the preceding or next one.

111

cc
u.

5 -

— I —
60

100

30 40 50 60 70 80 90

PERCENT OF SIGHTINGS IN WHICH

THE NEAREST ONE WAS NEITHER

THE PRECEDING OR NEXT ONE

Figure 4. — The frequency distribution for the jjercentage of
sightings of schools of spwtted dolphins, Stennella attenuata ,
within a cruise for which the nearest other sighting was neither
the preceding or next one. Shaded portions represent cruises
with less than 15 sightings (note one cruise had no sightings of
spotted dolphins).

356

POLACHECK: DISTRIBUTION OF SEARCHING EFFORT

with the largest distances between sets also have
large average search distances, but the converse
is not true (Fig. 5). The importance of dolphin
fishing as compared to log and school fishing ap-
pears not to be related to these differences among
cruises (Fig 6). It should be noted that the points
in Figure 5 must lie above the straight line with
a slope of 1.0. The expected relationship between
the variables in Figure 5 depends upon both the
underlying searching process and the spatial dis-
tribution of potential sets. While a positive rela-
tionship would be expected if both of these are
random, a more precise definition is beyond the
scope of this paper. The purpose in presenting
Figures 5 and 6 is to display the range of differ-
ences in the success of vessels in locating poten-
tial fishing targets. The large variation among
cruises suggests that all vessels may not be using
the same searching strategy and is an important
factor to keep in mind when considering the re-
sults from the clustering algorithm.

^ 220

■5 200

S 180

oc
o

»-

Ui
(A

160

140

120

O 100

I-

D
UI

« 40

UI 20„
S 0

MEAN DISTANCE TO THE NEAREST
SET OR CHASE (nautical miles)

Description of Clusters

The clustering algorithm grouped between 30
and 100% of all sets and chases within a cruise
into clusters with at least three members (Table
1). The percentage of all sets and chases that are
included in clusters increases with the value of
the cluster parameter so that 60-100% of all sets
and chases within a cruise occur within 150 miles
of another one. These percentages can only in-
crease with increases in the value of the cluster
parameter (i.e., a set that is included within a
cluster for a lower value of the cluster parameter
will always be included in a cluster at a higher
value). Similarly, the average cluster size and the
total percentage of the distance searched that oc-
curs within clusters must be nondecreasing func-
tions of the clustering parameter (Table 1). How-
ever, even when the clustering parameter equals
150 miles so that most sets occur within clusters,
the percentage of the total distance searched
within clusters averages only 59% of the total
distance searched during the cruise. This indi-
cates that substantial searching activity occurs
far from any set or chase.

The average intercluster distance for a cruise
ranged from 134 to 425 miles for clusters defined

220
200

UJ -^

< w

Ui UJ

««

liJ ^

O X

^i
<S<

o«

Z H

< UJ
UJ V>

180

160

140

120

100

20
0

• .

• •

•• •

10 20 30 40 50 60 70 80

PERCENT OF THE SETS AND
CHASES INVOLVING DOLPHINS

Figure 5. — The relationship between the average distance
searched between sets and chases within a cruise and the aver-
age distance to the nearest one.

Figure 6. — The relationship between the average distance
searched between sets and chases and the proportion of sets and
chases in which schools of dolphins were associated for a cruise.

357

FISHERY BULLETIN: VOL. 86, NO. 2

Table 1. — Statistics summarizing the results of the clustering algorithm showing the
sensitivity to a range of values for the clustering parameter. Standard deviations are
given as opposed to standard errors to provide an indication of the variation between
cruises. All means are the simple average for 35 cruises.

Cluster parameter

100

150

Percent of sets within
clusters with at least
3 members

Mean

Range

14.1

30-90

13.5

31-97

13.3

31-97

9.2

60-100

Average cluster radius
for clusters with at least
3 members (nautical
miles)

Mean

Range

8.5

19-56

19.7

27-142

40.4

34-225

116

66.0

38-390

Percent of the search-
ing within clusters with
at least 3 members

Mean

Range

25.8
12.6

9-61

16.4

16-91

16.1
19-91

10.6

31-91

Average intercluster
distance (nautical
miles)

Mean

Range

210

65.9

134-425

261

75.9

155-465

297

82.2

173-513

356

100.9

196-691

30 r

N = 767

30r

20 -

N = 570

50
40 r

30 -

O 20

75 100 125 150 175 200 225 250 >250 75 100 125 150 175 200 225 250 275 >275

40r

N = 448

N = 309

_1_

100 125 150 175 200 225 250 275 300 >300

INTER-CLUSTER DISTANCE (nautical mNaa)

150 175 200 225 250 275 300 325 350 >350
MTEH-CLUSTCR MSTANCC (nauUcal nritM)

Figure 7. — The frequency distribution of intercluster distances. Distances from all cruises have been pooled. The values of the

clustering parameter are A) 50, B) 75, C) 100, and D) 150.

358

POLACHECK: DISTRIBUTION OF SEARCHING EFFORT

by a value of 50 for the clustering parameter and
from 196 to 691 miles for a value of 150 miles
(Table 1). The frequency distribution of these in-
tercluster distances (Fig. 7) is an indication of the
stability of the clusters to the value of the cluster-
ing parameter. Thus, for a value of 100, over 65%
of the clusters have an intercluster distance
exceeding 175 miles. This suggests that 65% of
clusters will be stable up to a value of 175 miles
for the clustering parameter. (Note, this is not
strictly true. If the set preceding the first mem-
ber of a cluster was less than 175 miles away and
this set was also less than 175 miles from the first
set of the next cluster, this set plus these two
clusters would be combined in a single cluster
for a value of the cluster parameter less than
175.)

These statistics describing the characteristics
of the defined clusters suggest that the algorithm
used to create them successfully separates the ac-
tivities of a cruise into areas where sets are com-
mon and areas where they are infrequent. The
major differences in the clusters with different
values for the clustering parameters result from
the merging of two relatively close clusters or the
inclusion of an isolated set or chase near the
boundary of a cluster (e.g., for 80% of the cruises,
the actual number of clusters decreases or re-
mains the same over a range of 50-150 miles

for the clustering parameter). However, the fact
that many of these descriptive statistics vary
continuously with the value of the clustering
parameter suggests that these defined clusters
do not represent distinct units, but areas of
high concentration in a continuously grading
system.

Cruises vary greatly with respect to the
amount of variability they exhibit in response to
changes in the value of the clustering parameter
(Table 2). Such variability is to be expected since
no single searching strategy is used by all vessels
and vessels may change their strategy during the
course of a cruise. In addition, the spatial distri-
bution of potential sets probably also varies with
time and space. These sources of variability
among cruises, combined with the relatively
small sample sizes within a cruise, may be part of
the reason that the descriptive statistics charac-
terizing clusters vary continuously with the value
of the clustering parameter.

The lack of any sharp demarcation in the clus-
ters as a function of the clustering parameter,
combined with the large amount of variability
exhibited among different cruises, creates a prob-
lem in presenting results based on the clustering
algorithm. Consequently, whenever summary
statistics are presented, results are given for a
range of values for the clustering parameter.

Table 2. — Examples of the effect of changes in the value of the clustering parameter for 5 arbitrarily selected

cruises.

Average

Percent of

cluster radius

Percent of

Number of

sets and

for clusters

the search-

Average

Value

clusters

chases with

with at least

ing within

Intercluster

of the

with

clusters with

3 members

clusters with

distance

cluster

at least

(nautical

at least

(nautical

Cruise

parameter

3 members

miles)

3 members

miles)

171

142

298

100

142

298

150

142

298

210

294

100

105

397

150

126

443

402

450

100

479

150

614

425

465

100

513

150

513

302

362

100

362

150

362

359

FISHERY BULLETIN: VOL. 86, NO. 2

Whenever it seemed important to display the re-
sults by cruise, a value of 100 miles for the clus-
tering parameter was used. It should be empha-
sized that the general patterns appear to be
independent of the exact value of the clustering
parameter.

Is Searching DiflFerent in Clusters?

To determine whether vessels tended to in-
crease their searching effort in the vicinity of
sets, the ratio of the total distance travelled (i.e.,
the distance searched plus the distance run) to
the actual physical distances between sets was
compared between and within clusters. This ratio
is about 1.7 times greater in clusters than be-
tween clusters (Table 3). Also for no more than 7

cruises is the ratio within clusters less than the
ratio between clusters (Table 3). By a sign test
(Snedecor and Cochran 1967), these results imply
that this ratio is significantly greater than 1.00
(P < 0.005).

In addition, the proportion of the total distance
travelled that is devoted to searching is much
greater when a vessel is between than when it is
within clusters (Table 4). The fact that vessels
run, and are not actively looking for tuna, propor-
tionately more between than within clusters is an
indication that clusters are areas in which a ves-
sel has decided to remain. Vessels tend only to
run at night. Usually, as long as the vessel is
moving during daylight, the crew will be search-
ing. Thus, large amounts of running tend to occur
when vessels are actively moving to new areas.

Table 3. — Comparison between and within clusters of the ratio of the total distance
travelled (i.e., distance searched plus distance run) to the actual physical distance
between sets and chases. Means are the average values of the ratio for the 35 cruises
being considered.

Value of the cluster parameter

100

150

Between

Mean

1.79

1.70

1.65

1.59

0.064

0.062

0.060

0.065

Range

1.24-2.70

1.13-2.71

1.12-2.68

1.11-2.81

Within

Mean

3.08

2.84

2.79

2.68

0.343

0.254

0.250

0.239

Range

1.13-11.03

1.16-8.67

1.09-8.67

1.12-8.67

Number of cruises out

128

131

132

131

of 35 in which the ratio

is greater within than

between clusters

ip < 0.05 of observing this many ratio greater within clusters than between if they were in fact
equal based on a sign test (Snedecor and Cochran 1967).

Table 4. — Comparison between and within clusters of the ratio of the distance run without
searching to the distance searched. Means are the average values of this proportion for the
35 cruises being considered.

Value of the cluster parameter

100

150

Between
clusters

Mean
ratio

Range

0.85

0.063

0.35-1.94

0.93

0.068

0.26-1.94

0.97

0.067

0.26-2.05

1.06

0.095

0.26-3.45

Within clusters

Mean
ratio

Range

0.29

0.040

0.01-0.88

0.33

0.036
0.04-0.98

0.38

0.042

0.04-1.13

0.46

0.048

0.05-1.06

Number of cruises out
of a total of 35
in which the propor-
tion is greater be-
tween clusters than
within them

134

ip < 0 001 of observing the portion being greater between clusters than within 34 out of 35 cruises, if
in fact the proportion was equal based on a sign test (Snedecor and Cochran 1 967).

360

POLACHECK: DISTRIBUTION OF SEARCHING EFFORT

However, some running can still be expected
within clusters because vessels mark favorable
logs with radio transmitter in order to return to
them at a latter time. It should be noted that
three or more sets or chases rarely occur on the
same day (Fig. 8). Therefore, the proportionately
smaller amount of running that occurs within
clusters is not an artifact resulting from clusters
that do not span more than a single day.

Are Clusters Areas of
High Spotted Dolphin Densities?

Encounter rates {total number of sightings of
spotted dolphins divided by the total distance
searched) tend to be much greater within clusters
dominated by sets in association with dolphins
than either within clusters dominated by non-
dolphin sets or while searching between clusters
(Table 5). The clustering algorithm would tend to
ensure that such differences are likely to occur,
but the magnitude of the differences is large and

900

800

700

600

500

LLI

lu 400
a.

300
200
100

° 1 2 3 4 5

NUMBER OF SETS OR CHASES

Figure 8. — The frequency distribution of the number of sets
or chases during a day for all days on which there was at
least one chase or set.

Table 5. — Comparison of encounter rates (number of schools of spotted dolphins, Stenella
attenuata, per 100 miles searched) while searching within clusters classified according to the
percentage of sets involving dolphins, and while searching between clusters. Unweighted
means are the average value of the encounter within each cruise. Weighted means are based
on the encounter rate weighted by the distance searched by a cruise and are equivalent to the
total number of sightings divided by the total distance searched pooled across all vessels.

Within

clusters in which the percent

of the cluster parameter

of sets on dolphins ranged

from

Between

Value

0-25

25-75

75-100

clusters

Unweighted mean

0.17

1.45

2.74

0.52

0.070

0.182

0.219

0.046

Weighted mean

0.10

1.04

2.46

0.50

0.047

0.120

0.205

0.046

Number of cruises

Total distance searched

17.9

10.0

23.1

148.2

(thousands of miles)

Unweighted mean

0.15

1.32

2.32

0.47

0.051

0.191

0.188

0.044

Weighted mean

0.12

0.88

2.16

0.46

0.039

0.111

0.166

0.044

Number of cruises

Total distance searched

28.2

15.1

33.0

125.4

(thousands of miles)

100

Unweighted mean

0.14

1.15

2.08

0.46

0.046

0.181

0.174

0.045

Weighted mean

0.15

0.81

1.82

0.45

0.043

0.117

0.148

0.044

Number of cruises

Total distance searched

35.1

16.0

45.0

106.7

(thousands of miles)

150

Unweighted mean

0.10

0.73

1.83

0.43

0.040

0.106

0.159

0.049

Weighted mean

0.11

0.68

1.63

0.40

0.037

0.096

0.114

0.041

Number of cruises

Total distance searched

33.9

27.2

57.1

85.7

(thousands of miles)

361

FISHERY BULLETIN: VOL. 86, NO. 2

relatively insensitive to the value of the cluster-
ing parameter. However, the fact that a large per-
centage of the clusters tend to be dominated by
either dolphin or non-dolphin sets is not a neces-
sary consequence of the clustering algorithm and
suggests that the two types of methods for locat-
ing and catching tuna tend to be spatially and
temporally distinct. Encounter rates are substan-
tially lower in clusters dominated by non-dolphin
sets than when a vessel is searching between
clusters (Table 5). This result is also not a neces-
sary consequence of the clustering procedure and
suggests that these clusters not only define areas
of high densities of spotted dolphin schools but
also areas of low densities.

Encounter rates could be lower in non-dolphin
areas because of differences in detectability not
related to the density of schools. For example,
Hammond'* suggested that the crew may scan
closer to the vessel when searching in non-
dolphin areas (see also Polacheck 1983). It seems
unlikely that such factors could account for all of
the differences between the encounter rates in
Table 5.

Differences in detectability due to differences
in weather conditions between and within clus-
ters could also affect the results in Table 5. En-
counter rates do decrease at higher Beaufort sea
states (Polacheck 1983). However, little search-
ing occurs above Beaufort state 4. The difference
in encounter rates at Beaufort 0-2 compared with
Beaufort 3-4 (about a factor of 1.28) is insuffi-
cient to explain the difference in Table 5
(Polacheck 1983). Moreover, areas of non-dolphin
sets tend to be in nearshore areas with calmer
seas and fishermen do not consider Beaufort 4
conditions as being too rough to fish.

Do Tuna Catches Differ Between
and Within Clusters?

The average tons of tuna caught per set tend to
be greater for sets which occur within clusters
than sets between clusters (Fig. 9). For all values
of the clustering parameter, the average catch per
set was greater within than outside of clusters in
approximately 70% of the cruises (Table 6). Para-
metric statistical comparisons of the average tons
per set are not appropriate because of the large
differences in the average catch per set among

vessels (Fig. 9). A nonparametric sign test (Sned-
cor and Cochran 1966) suggests that the differ-
ences in catch per set between and within clusters
are significant at least at the 0.05 probability
level for all values of the clustering parameters
that were considered.

60 t-

CO
QC
liJ

t-
V)

u
_l

UJ
UJ

I-

LLI

UJ
Ui

I-

<
O

</)

z
o

I-

z
<

•

/
/
/

/
/
/
/
• /

/
/
/
/

•

/
/•
• /

•

/
/

• •

/
• /
/

• •

•

/
/

•

•
•

•

/ •

, • •

•

(■ 1

- i

1 1 1 1 1

0 4 8 12 16 20 24 28 32

MEAN TONS CAUGHT PER SET WITHIN CLUSTERS

Figure 9. — The average tons of tuna caught per set while a
vessel was travelling between clusters versus the average tons
per set within clusters with at least three members. The dashed
line represents the expected value if the catch rates were equal.
The value of the cluster parameter equals 100.

Table 6. — The ratio of the tons of tuna caught per set within clus-
ters with at least 3 members compared to tons caught per set
between clusters. The means are the average values for the ratio
within a cruise. Cruises in which 100% of the sets were within
clusters are not inlcuded in the results.

■♦Hammond, P. S. 1981. Some problems in estimating the
density of dolphin populations in the eastern tropical Pacific
using data collected aboard tuna purse seiners. Inter-Am.
Trop. Tuna Comm. Intern. Rep.

Value of the cluster parameter

50 75

100 150

Mean

173 2.02

298 2.76

0.22 0.379

0.476 0.499

Range

0.41-6.83 0.23-11.18

0.39-11.68 0.43-8.43

35 34

34 25

Number of

125 125

125 119

cruises in

which ratio

was greater

than 1

' P < 0 05 of getting the observed number if the ratio equaled 1 by a sign test
(Snedecor and Cochran 1966).

362

POLACHECK: DISTRIBUTION OF SEARCHING EFFORT

Does Searching Behavior Bias

the Overall Estimate of

Dolphin Encounter Rates?

When the two different methods of estimating
the overall encounter rates within a cruise are
compared, the adjusted encounter rates tend to be
smaller (Table 7). However, the differences are
not large, not because the biases are necessarily
small, but because positive and negative effect of
concentrating searching effort tend to cancel each
other. Positive biases would be expected in the
unadjusted rate due to concentrating of searching
effort in clusters dominated by sets made in asso-
ciation with dolphins, since these appear to be
areas of high dolphin densities (Table 5). Simi-
larly, negative biases would be expected due to
searching in non-dolphin clusters. For the data
considered here, substantial and roughly equal
amounts of searching occurred in both types of
clusters. The effects of concentrating searching
effort within the two types of clusters tended to
cancel each other. This suggests that the major
effect of nonrandom searching on measures of
relative abundance of dolphins will change in re-
lationship to the importance of dolphin and non-
dolphin fishing. Such changes tend to occur with
changes in the relative abundance of large yel-
lowfin tuna compared to skipjack and small yel-
lowfin tuna.

DISCUSSION

The fundamental question in interpreting the
results of this paper is whether the clusters that
have been defined bear any relationship either to
the searching strategy of the vessels or to the
underlying distribution of dolphin and tuna
schools. Two factors hamper answering this ques-
tion: The first is the large variability among ves-

sels; the second is the lack of appropriate null
hjqjotheses by which to test the results. The large
variability is to be expected and is inescapable.
Not only is there a large stochastic element in the
catch and encounter process, but large differences
can be expected in searching strategies among
vessels. Thus, a large range exist in skill and
experience among fisherman. In addition, the
amount of information, which is shared among
vessels varies and some vessels may be acting as
scouts for other vessels (Orbach 1975).

The specification of null hypotheses is difficult
because an infinite number of searching models
are compatible with the definition of random
search (i.e., defining a random search as one in
which the search path is independent of the dis-
tribution of the objects being sought). In order to
actually model a random searching process, the
probability of changing the direction of the search
path needs to be specified. A random search could
encompass an)^hing fi-om Brownian motion to
random straight line crossings of an area. With a
finite amount of searching, these will not neces-
sarily yield the same results.

Most of the results from this paper are compat-
ible with a model of clustered searching for clus-
tered prey, and some of them seem hard to explain
unless the searching and the schools of dolphins
are nonrandomly distributed. That the distance
travelled relative to the actual distances between
sets tends to be greater in areas where the density
of sets is high, as is the proportion of this distance
which is spent searching, are unlikely results un-
less searching is concentrated in these areas.
Also, the higher encounter rates in clusters dom-
inated by dolphins, the comparisons of the dis-
tances and frequency of the nearest set with the
preceding or next set, and the high percentage of
sets which fall into clusters are results that would
be expected if searching, tuna, and dolphins were

Table 7. — Comparison between the unadjusted overall encounter rate (nunnber of
schools of Stenella attenuata per 1 00 miles of searching) for a cruise and the encounter
rate adjusted for possible bias due to the concentration of searching within clusters.

Number of cnjises

out of 34 in which

the corrected esti-

Cluster

mate was less than

parameter

Mean

Range

the uncorrected one^

Unadjusted

0.80

0.078

0.00-1.77

Adjusted

0.73

0.074

0.00-1.66

0.74

0.074

0.00-1.60

100

0.75

0.077

0.00-1.73

150

0.76

0.078

0.00-1.78

'One cruise had no recorded sighting of spotted dolphin schools.

363

FISHERY BULLETIN: VOL 86, NO. 2

spatially clustered. Perhaps the most surprising
result in this context is that the average catch per
set tends to be greater within the defined clusters
than outside of them. There is nothing in the clus-
tering algorithm that would tend to produce this
result. If these larger catches per set reflect large
schools of tuna, this argues that these defined
clusters are areas where tuna tend to concentrate
and not just areas of high densities of sets that
could be found in any random distribution.

Within many of the cruises, the locations of two
or more of the defined clusters overlap spatially
(Fig. 10), indicating that vessels often return to

an area after a period of searching elsewhere. In
addition, clusters from different cruises overlap
spatially and temporally. This overlapping of
clusters indicates that the overall searching be-
havior is even more nonrandom than the results
from this paper suggest. The overlapping both
within and among cruises is a dimension that
should be considered in future extensions to the
present work.

Orbach (1975), in a nonquantitative, anthropo-
logical study of the purse seine fishery, included
a general qualitative description of the searching
behavior of the fishermen which supports many of

Figure 10. — Examples of the spatial relation between clusters for the two cruises depicted in Figure 2. Open circles indicate the
position of all clusters with at least three members. The radius of each circle is scaled to the estimated radius of a cluster. The
associated numbers are the number of sets and chases within a cluster. Solid circles indicate the position of clusters with only two
members. Isolated sets are indicated by an x. The value of the clustering parameter equals 100. Arrows indicate the order of
movement between clusters. Note A and B are drawn to different scales. Distance between tick msirks equals 300 miles. No geographic
coordinates are given and the orientations were rotated at random so as not' to compromise any proprietary fishing information.

364

POLACHECK: DISTRIBUTION OF SEARCHING EFFORT

the quantitative results in this paper. He stated
that fishermen perceive two kinds of areas: The
first, in which fishermen refer to their activity as
"scratching", are regions where isolated schools of
fish are encountered; the second, which fishermen
refer to as an "area", are portions of the ocean
where schools of tuna are congregated. He also
reported that isolated tuna schools tend to be
small. Fishermen prefer to concentrate in an
"area" and actively search for them. I did not be-
come aware of this study by Orbach until the final
calculations of this paper were completed. Thus,
his description is an independent indication that
the results are not an artifact of the clustering
algorithm.

The method used for estimating the location
and size of a cluster could be refined. Such refine-
ments were beyond the scope of the present work
and would not affect any of the main results.
However, such refinements might be important if
the method used for calculating the adjusted en-
counter rates (Table 7) was used to develop rela-
tive abundance indices for dolphins in the ETP.
As pointed out by one reviewer, the problem of
estimating the shape, size, and location of a clus-

ter is analogous to the problem of determining the
home range for an animal from a set of observed
positions over time (Sanderson 1966; Cooper
1978; Schoener 1981; Swihart and Slade 1985)
and is part of the more general problem of how to
estimate the limits and size of clusters from any
clustering algorithm. The methods used for esti-
mating home ranges cannot be directly applied to
the tuna boat observer data but might provide a
basis for developing a better estimator for these
statistics relating to cluster size in any extension
to the present work.

To the extent that the results from this paper
indicate that searching is clustered, they suggest
that when these data are used for estimating rel-
ative densities of dolphin schools, or when catch
and effort data from purse seiners are used to
assess tuna stocks, rather fine geographic stratifi-
cations are needed to avoid biases from the non-
random searching within a cruise. The estimates
of the cluster radius could be considered as a
guide to appropriate levels of stratification. For
example, from 31 to 46% of all clusters with at
least three members had a radius of less than 60
miles (Fig. 11). This suggests that at a minimum,

50 r

40-

5 30
Ui

oc
111
Q. 20

10 -

N=230

t t

1 1 1 1 1

40 -

30 -

20 -

N=230

1 1 1 1 1

120

150

180 >180

120

150

180 >180

O 20
c

N=214

30 60 90 120 150 180 >180

CLUSTER RADIUS (nautical mllaa)

40r-

30-

20 -

10 ~

]

N=184

gj;;:

1 t

30 60 90 120 150 180 >180

CLUSTER RADIUS (nautical miles)

Figure 11. — The frequency distribution for estimated radii for all clusters with at least three members. Estimates from
all cruises have been pooled. The values of the clustering parameter are A) 50, B) 75, C) 100, and D) 150.

365

FISHERY BULLETIN. VOL. 86, NO. 2

a 2° stratification would be necessary to avoid
biases in these areas.

ACKNOWLEDGMENTS

I wish to thank Peter Frank and Tim Smith for
their generous support and useful comments dur-
ing the research presented here. This paper would
not exist without the data collected by the Na-
tional Marine Fisheries Service observers. I wish
to acknowledge these observers and the staff of
the Southwest Fisheries Center of the National
Marine Fisheries Service for collecting, editing,
and initially preparing these data. A number of
the staff of the Southwest Fisheries Center pro-
vided support and technical assistance at various
stages in this research. The number of people in-
volved is too large to acknowledge all of them
individually. Special thanks go to Phil Tarentino,
Michael Lichter, and Susan Boyer.

The art department of the Southwest Fisheries
Center were responsible for the final preparation
of some of the figures. Veronica van Kouwen
helped with the preparation of the final
manuscript.

The manuscript is a portion of a Ph.D. Thesis
submitted to the University of Oregon, Eugene,
OR.

LITERATURE CITED

Allen, R., and R. Punsley.

1984. Catch rates as indices of abundance of yellowfin
tuna, Thunnus albacares , in the eastern Pacific Ocean.
Inter-Am. Trop. Tuna Comm. 18:303-379.
Blackburn, M., and F. Williams.

1975. Distribution and ecology of skipjack tuna, Katsu-
wonus pelamis , in an offshore area of the eastern tropical
Pacific. Fish. Bull., U.S. 73:382-411.
Calkins, T. P.

1963. An examination of fluctuations in the "concen-
tration index" of purse seiners and bait boats in the fish-
ery for tropical tuna in the eastern Pacific. Inter-Am.
Trop. Tuna Comm. Bull. 8:255-316.
Calkins. T. P., and B M Chadwin.

1967. Geographical distribution of yellowfin tuna and
skipjack catches in the eastern Pacific Ocean, by quarters
of the year 1963-1966. Inter-Am. Trop. Tuna Comm.
Bull. 12:435-508.

1971. Geographical catch of yellowfin and skipjack tuna
in the eastern tropical Pacific Ocean, 1967-1970, and
fleet and total catch statistics. Inter-Am. Trop. Tuna
Comm. Bull. 15:285-377.

Clark. C. W., and M. Mangel.

1983. Search theory in ecology and resouce manage-
ment. In M. Freedman (editor). Proceedings of the In-
ternational Conference on Population Biology, p. 380-
388. Springer Lecture Notes in Biomathematics, Vol.
52. Springer Verlag, N.Y.

Cooper. E

1979. Home range criteria based on temporal stability of
areal occupation. J. Theor. Biol. 73:687-695.

GULLAND, J. A

1955. Estimation of growth and mortality in commerical
fish fxipulations. Fish. Invest. Minist. Agric. Fish. Food
(G.B.), Series 2. 18:1-46.
Helland Hansen, B

1909. Statistical research into the biology of the haddock
and cod in the North Sea. Rapp. P-v R6un 10:1-62.
Koopman, B O

1980. Search and screening. Pargamon Press, N.Y.,
509 p.

Orbach, M K

1975. The cultural system of the tuna seinermen of San
Diego, California. Ph.D Thesis, Univ. California, San
Diego, CA, 411 p.

Pazynich, G I

1966. Primery primeneniya teorii statisticheskiskh resh-
enii k zadacham taktiki promysla (Examples of the appli-
cation of statistical-solution theory to problems in the
fishing tactics). Noe Khoz. 6:80-81. (Translation
available in the library of National Marine Fisheries
Service Northeast Fishery Laboratory, Woods Hole, MA.)

Perrin, W. F.

1968. The porpoise and the tuna. Sea Frontiers 14:166-
174.

1969. Using porpoise to catch tuna. World Fishing
18:42-45.

POLACHECK, T

1983. The relative abundance of dolphins in the eastern
tropical Pacific based on encounter rates with tuna purse
seiners. Ph.D Thesis, Univ. Oregon, Eugene, OR, 444 p.

Sall\, S. B., AND J M. Flowers.

1979. Elementary applications of search theory to fishing
tactics as related to some aspects of fish behavior.
F.A.O. Conference on Fish Behavior in Relation to Fish-
ing Techniques and Tactics, p. 343-355.

Sanderson, G.

1966. The study of mammal movements - a review. J.
Wildl. Manage. 30:215-235.

SCHOENER, T W

1981. An empirically based estimate of home range.
Theor. Popul. Biol. 20:281-325.

Snedecor, G W , AND W C. Cochran

1967. Statistical methods. 6th ed. Iowa State Univ.
Press, Ames, 593 p.

Suzuki, Z , P K. Tomlinson, and M Honma.

1978. Population structure of Pacific yellowfin tuna.
Inter-Am. Trop. Tuna Comm. Bull. 17:277-411.
SwiHART, R. K . and N. a. Slade.

1985. Influence of sampling interval on estimates of home
range size. J. Wildl. Manage. 49:1019-1025.

366

A COMPREHENSIVE THEORY ON THE ETIOLOGY OF BURNT TUNA

Cheryl Watson,' Robert E. Bourke,^ and Richard W Brill^

ABSTRACT

Over the past 14 years, the Hawaii handline fishery has experienced phenomenal growth in the catch
for large yellowfin tuna, Thunnus albacares, and bigeye tuna, Thunnus obesus. These fish are
primarily caught for the sashimi (raw consumption) market but have been continually plagued with
a product quality problem known as "burnt tuna" or, in Japanese, yake niku. Not only does this
problem significantly reduce the value of the catch, it also limits export markets and expansion of this
low-capital, high-return fishery to other areas of the Pacific. Previous research and suggestions for
mitigating burnt tuna have centered on the hypothesis that it is caused by high muscle temperature
and low pH, which is the result of a violent struggle during capture.

A new, more comprehensive hypothesis is presented: Burnt tuna is actually caused by the p)ost-
mortem activation of enzymes known as calcium-activated proteases and by the enhancement of the
effect of these enzymes by high blood catecholamine levels. Previously unexplainable observations,
such as the propensity of female fish to become burnt more often during the summer months, the
efficacy of brain destruction in preventing burnt tuna, and the lack of effect of cooling on the incidence
of burnt tuna, are explainable in light of this new hypothesis.

One of the largest fisheries in Hawaii is the hand-
line fishery for large (>50 kg) yellowfin tuna,
Thunnus albacares, and bigeye tuna, T. obesus,
caught primarily for raw consumption as sashimi.
Yearly landings increased from 89 short tons (ex-
vessel value, $131,000) in 1973 to 615 short tons
(ex-vessel value, $2.1 million) in 1984 (Yuen
1979; Hudgins and Pooley 1987). The total eco-
nomic value of the fishery has been estimated as
high as $5 million yearly (Ikehara'*). In Hawaii,
the night handline fishery is known as ika shibi
from the Japanese words for squid and tuna, and
the daytime fishery is known as palu ahi from the
Hawaiian words for chum and yellowfin tuna.
There is also growing international interest in
this tjTDe of fishing because of its low initial capi-
tal investment, low operating and fixed expenses,
strong export markets, and high profitability
(Strong 1979; Gibson 1981; Jerrett 1984). Boats
can be as small as 6 m and require only one- or
two-man crew. Catch rates in Hawaii have
ranged from two fish per hook per night (Yuen

'Pacific Gamefish Research Foundation, 74-425 Kealakehe
Parkway, #15, Kailua-Kona, HI 96740; present address: De-
partment of Physiology, John A. Bums School of Medicine, Uni-
versity of Hawaii, Honolulu, HI 96822.

2465B Kawailoa Road, Kailua, HI 96734.

3Southwest Fisheries Center Honolulu Laboratory, National
Marine Fisheries Service, NOAA, 2570 Dole Street, Honolulu,
HI 96822-2396.

^Ikehara, W. N. 1981. A survey of the ika-shibi fishery in
the State of Hawaii, 1980. Southwest Fish. Cent. Adm. Rep.
H-82-4C, 12 p. Southwest Fisheries Center Honolulu Labora-
tory, National Marine Fisheries Service, NOAA, Honolulu, HI
96822-2396.

1979) to one-half fish per hook per night
(Bourke^), illustrating the profitability for the in-
dividual fisherman.

Unfortunately, handline (as well as primarily
recreational trolling) fishermen are plagued by a
product quality problem known as "burnt tuna"
or, in Japanese, as yake niku (literally translated
as "cooked meat"). When fish are intended for raw
consumption, product quality is of utmost impor-
tance. Prime quality tuna flesh should be red,
translucent, and firm and have a delicate flavor.
Burnt tuna is pale, exudes a clear fluid, and has
a soft texture and a slightly sour taste. Although
perfectly palatable when cooked or canned, burnt
tuna is considered unsuitable for raw consump-
tion and commands only a fraction of the price of
prime quality fish. Fish are usually exported
whole to preserve freshness, so burnt tuna often is
not detected until shipping costs have been in-
curred. This discourages exports of tuna caught
from areas or via fishing techniques with high
incidences of the problem.

Burnt tuna aflects from 5 to 100% of the tissue
from an individual fish and ranges from mild to
severe. Approximately 25% of the fish caught by
the Hawaii handline fishery are burnt, as are
50% of the large yellowfin tuna caught by com-
mercial and recreational trollers (Bourke fn. 5).

^Bourke, R. E. 1985. Hilo ikashibi fishery 1984 survey:
Problems of a maturing fishery and a potential solution to the
burnt tuna problem. Poster presentation at 1985 Tuna Confer-

Manuscript accepted January 1988.
fishery BULLETIN: VOL. 86, NO. 2, 1988.

367

FISHERY BULLETIN: VOL. 86, NO. 2

Annual losses to the handline fishery are an esti-
mated 169^ of the value of the total catch (Cramer
et al. 1981). The problem has also been reported
in fish intended for raw consumption caught by
purse seine (Nakamura et al. 1977). Surprisingly,
the problem rarely occurs in fish caught by
longlining (Williams 1986).

PAST RESEARCH ON BURNT TUNA

The Japanese were the first to investigate the
causes of burnt tuna and possible mitigating
strategies (Itokawa 1968, 1969). The first con-
trolled laboratory investigations were those of
Nakamura et al. (1977) and Konagaya and Kona-
gaya (1978, 1979). Nakamura et al. (1977) con-
cluded that high muscle temperature and low
muscle pH caused myofibrillar protein denatura-
tion and also noticed that, once denaturation
began, it continued even if the tissue was kept at
0°C. Because of the relatively high thermostabil-
ity of tuna myofibrillar protein and because yake
niku occurs in species (e.g., frigate mackerel and
sardine) that do not generate high muscle tem-
peratures during struggling, Konagaya and Kon-
agaya (1978) concluded that acid denaturation of
myofibrillar proteins at moderate temperatures
was the underlying cause.

Cramer et al. (1981) studied handline-caught
yellowfin tuna in Hawaii and found that the oc-
currence of burnt tuna did not correlate with
muscle temperature at time of landing and corre-
lated only loosely with extracellular muscle pH.
Ikehara^ conducted an engineering study to de-
velop methods to cool large yellowfin tuna more
rapidly, the presumption being that rapid cooling
would prevent muscle degradation. Although suc-
cessful in developing a technique to increase cool-
ing of deep muscle temperature, as shown in Fig-
ure 1, there was no apparent correlation between
rate of cooling and incidence of burnt flesh. In
spite of this lack of directly observed correlation,
some publications designed for fishermen still
stress that high muscle temperatures and low
muscle pH are the prime causes of burnt tuna
(Gibson 1981). Others have expressed doubt as to
the validity of this hypothesis (Jerrett 1984).

The high muscle temperature-low pH hypothe-
sis appears to fit with what is known about tuna
physiology, in that these fishes are capable of pro-

ducing muscle temperatures significantly above
ambient (Carey et al. 1971; Carey 1973) and ex-
hibit some of the highest rates of muscle glycoly-
sis (production of muscle lactate and concomitant
production of acidity) observed in nature
(Hochachka et al. 1978; Hochachka and Momm-
sen 1983). Yet some observations do not fit this
hypothesis. For example, burnt tuna occurs more
frequently in summer, more frequently in female
fish, and more frequently in fish fought for short
periods of time than in fish fought for long periods
(>7 minutes or <2 hours) (Davie and Sparksman
1986; Nakamura^). Furthermore, burnt tuna oc-
curs rarely in longline-caught fish and in fish
subjected to brain or spinal column destruction
immediately following capture (Nakamura et al.
1977; [Suisan Sekai] 1977; Cramer et al. 1981;
Davie and Sparksman 1986; Nakamura fn. 7).
The hypothesis that burnt tuna is caused by high
muscle temperature and low muscle pH does not
seem to directly fit with any of these observa-
tions.

A NEW ANALYSIS OF THE BURNT
TUNA PROBLEM

At the biochemical level, the high muscle
temperature-low pH h3rpothesis would predict
that the observed drop in extracellular pH would
be accompanied by a similar drop in intracellular
pH and activation of lysosomal proteases. These
proteases would then degrade actin and myosin,
the dominant muscular proteins, resulting in the

'^R. Nakamura, Department of Animal Sciences, University of
Hawaii, Honolulu, HI 96822, pers. commun. October 1987.

I—
<

o
I/)

Q.
LiJ

□ UKN 1

30.0

^^=!5--

NOT BURNT

^^>

"~^^^^^

•-«

.^^_^^6.2)

20.0

s
s

~~^~~^ (64.4)

(M.9)

\s._^ N.

(103.0) ^ - ^

~(62.1)

"(61.7)

10.0-

■

X,,^^

9.0
8.0

(99.8 )

7.0-

— 1 1-

— 1 1

H 1 1 1 1

10 11

ELAPSED TIME (h)

^Ikehara, W. N. 1981. Development of a small-boat chill-
ing system for the reduction of burnt tuna. Final Report for a
Pacific Tuna Development Foundation contract, var. pag.

Figure 1. — Data are from Ikehara (text fn. 6). Initial and final
deep muscle temperatures are plotted on semilogarithmic axes
to linearize the rate of temperature change. Numbers in paren-
theses are the body weight (kg) for each fish.

368

WATSON ET AL : ETIOLOGY OF BURNT TUNA

undesirable changes seen in burnt tuna. The lyso-
somal proteases, specifically cathepsins B, D, and
L, have an optimum pH of about 5 (Dahlmann et
al. 1984). However, recent work by Abe et al.
(1985, 1986) has shown that, intracellularly, tuna
muscle is well buffered and that a fall in extracel-
lular pH is not necessarily accompanied by an
equivalent fall in intracellular pH. Using is-
chemic rat gastrocnemius muscle, Hagberg
(1985) found that a 1 pH unit drop (7.30-6.36) in
extracellular pH was accompanied by a drop in
intracellular pH of only 0.4 unit (7.00-6.60).
Thus, the acidic intracellular environment that
would favor the action of the lysosomal proteases
probably is not present in burnt tuna. Also, with
the exception of calcium-activated neutral
protease, all other muscle proteases (cathepsins
B, D, L; alkaline serine protease; neutral trypsin-
like protease; and alkaline cysteinyl protease) de-
grade myosin (Dahlmann et al. 1984). Yet
Hochachka and Brill (1987) found burnt muscle
had no increase in 3-methyl-histidine, a specific
marker for myosin breakdown. Decomposition of
other myofibrillar proteins is, therefore, impli-
cated.

Electronmicrographs of postmortem burnt tuna
muscle (Davie and Sparksman 1986) showed a
consistent, rapid disintegration of Z-discs and ir-
regularities in the sarcoplasmic reticulum (SR).
The changes in burnt muscle were not different in
kind from postmortem changes seen in unburnt
tissue, but were a result of a significant increase
in the rate of disintegration. Selected destruction
of the Z-discs, troponin and tropomyosin, and the
SR is characteristic of a pair of proteases known
as calcium-activated neutral proteases (CANP's)
(Sugita et al. 1984; Suzuki et al. 1984). These
proteases are cytoplasmic, ubiquitous, activated
by increased intracellular calcium levels, and ac-
tive at pH 5.5-8.0 (Sakamoto and Seki 1985;
Koohmaraie et al. 1986; Seki and Kimura 1986;
Zeece et al. 1986b). The intracellular pH most
likely found in burnt tuna muscle is, therefore,
more consistent with CANP action than with
lysosomal proteases whose activity requires a pH
closer to 4.5 (Hochachka and Brill 1987). In addi-
tion, while cathepsin D's activity is greatly re-
stricted at 15°C, CANP is still active at 5°C
(Koohmaraie et al. 1986; Zeece et al. 1986a).

A new etiology of burnt tuna proposed by
Hochachka and Brill (1987) is summarized in Fig-
ure 2. Their hypothesis predicts that low intracel-
lular ATP concentrations lead to the leaking of
Ca"^^ into the cell and increases in intracellular

INTENSE MUSCLE ACTIVITY

LACK OF 02 + ATP

METABOUC COLLAPSE OF MEMBRANE

RISE IN INTRACELLULAR CALCIUM

PSE

LLU

IN 0

ACTIVATION OF CANP

UNDESIRABLE MUSCLE TEXTURE

Figure 2. — Biochemical reac-
tions involved in the development
of burnt tuna as proposed by
Hochachka and Brill (1987).

Ca"^^ concentrations. These increases, in turn, ac-
tivate CANP, which specifically attacks troponin,
tropomyosin, SR, and mitochondria. The break-
down of the latter two intracellular organelles
releases more calcium into the cytoplasm, thus
further increasing the activity of CANP.

The effect of brain and spinal cord destruction
on reducing the incidence of burnt tuna and of
similar muscle degradation seen in other fish spe-
cies can be explained by this new hypothesis,
which assumes the initial drop in ATP is the root
cause of the elevated intracellular calcium
(Amano et al. 1953; Fujimaki and Kojo 1953;
Konagaya and Konagaya 1978; Ikehara fn. 6).
Brain destruction maintains elevated muscle
ATP levels after capture (Boyd et al. 1984). Data
recently collected on the use of brain and spinal
cord destruction in large yellowfin tuna (Fig. 3)

I-

I 35
m

g 30

I 20
o

Sl5

z
o

UJ
Q.

10--

1 4 FISHERMEN NOT USING BRAIN DESTRUCTION

^ ^\

SINGLE FISHERMAN

■^

/ \

\ r~^^^^ \

\ / \

\ 1 \

1 l\

\ J \

1 / \

\ 1 \

/ 1 ^

\ 1 \

--^ \

1
/ 1

"■--. ^' \

1 I

BRAIN DESTRUCTION USED \

1 /

(■ 1 1—

1 1 1 1 1 \

JAN

MAR

MAY JUL

MONTH

SEP NOV

Figure 3. — Incidence of burnt tuna occurring in large yellowfin
tuna caught by 14 commercial handline fishermen operating
out of Hilo, HI, January-September 1984, and by 1 fisherman
who began using brain destruction in May. His incidence offish
becoming burnt dropped dramatically in spite of the normal
summer increased incidence seen in the catch of 14 other fisher-
men.

369

FISHERY BULLETIN: VOL. 86, NO. 2

clearly show the effectiveness of this technique in
preventing burnt tuna. A drop from 30 to < 10% of
the tuna becoming burnt was noted when the
fisherman changed his killing technique from
shooting to the use of a brain spike.

A MORE COMPREHENSIVE

HYPOTHESIS ON THE ETIOLOGY

OF BURNT TUNA

The hypothesis presented by Hochachka and
Brill (1987) does not explain all the observed fac-
tors leading to tunas' different propensity to be-
come burnt. It cannot explain why tunas strug-
gling for <7 minutes rarely become burnt
(Nakamura fn. 7), nor can it account for the obser-
vations of Davie and Sparksman (1986) and
Bourke (unpubl. data), who found that tuna
struggling for extended (>1 hour) periods of time
have a lower probability of becoming burnt. It
also cannot explain why female fish become burnt
more often than males (Nakamura et al. 1977), or
why fish caught in summer become burnt more
often (Fig. 3).

Our revised hypothesis is presented in Figure
4. We have incorporated into the Hochachka and
Brill (1987) hypothesis, the action of the neuro-
transmitters-hormones norepinephrine (NE) and
epinephrine (E) (collectively referred to as cate-
cholamines). Periods of intense physical activity
or capture stress increase the levels of circulating
catecholamines by about 10-200 times in fish
other than tunas (Nakano and Tomlinson 1967;
Ling and Wells 1985). Most likely a similar situa-
tion occurs in tuna during hooking, fighting, and
capture. The importance of elevated circulating
catecholamines in this particular schema comes

INTENSE MUSCLE ACTIVITY

LACK OF 02 + ATP

METABOUC COLLAPSE OF MEMBRANE

RISE IN INTRACELLULAR CALCIUM

ACTIVATION OF CANP

STRESS OF CAPTURE

NE & E RELEASE

■INCREASE OF CANP ACTION ■

from their potentiating effect on the action of
CANP. Catecholamines cause the phosphoryla-
tion of troponin, resulting in a more rapid and
prolonged proteolysis of this muscle structure
(Toyo-oka 1982); in other words, high circulating
levels of catecholamines greatly increase the ef-
fectiveness of the enzymes responsible for a tuna
becoming burnt. ^

The gills are the organ primarily responsible
for the degradation of catecholamines, and circu-
lating catecholamines are rapidly cleared from
the blood (Nekvasil and Olson 1986). We estimate
circulating catecholamines in tuna have a half-
life of 30 minutes or less. Therefore, we hypothe-
size that, in a longline-caught fish remaining in
the water for several hours after hooking (Davie
and Sparksman 1986), plasma catecholamines
are reduced to low levels before the fish is landed
and killed, thus resulting in a low percentage of
burnt fish. Similarly, fish caught by rod and reel
or by handline and fought for several hours would
have the low circulating catecholamine levels and
low propensity to become burnt. Indeed, Gibson
(1981) recommends that handline-caught fish be
attached to a buoy and left for an hour prior to
being brought on board and killed, as a measure
to prevent burnt tuna. That this technique pre-
sumably lowers blood catecholamine levels prior
to death could explain its efficacy. On the other
hand, we hypothesize that a fish landed during
the peak of its blood catecholamine concentra-
tions would have the greatest propensity for be-
coming burnt. This indeed appears to be the case.

The mechanism that ties catecholamines to the
observed seasonality and increased number of fe-
male fish becoming burnt lies in the biochemical
structural similarity of catecholamines and re-
productive steroids, particularly estrogen. These
steroids reach a peak during spawning season,
which is May through October for yellowfin tuna
in Hawaii (June 1960). A corresponding increase
occurs in the percentage of the tuna catch that
becomes burnt during this season (Fig. 3). In the
presence of tyrosine hydroxylase, estrogen is con-
verted to catecholestrogen and consequently com-
petes for the same degradative enzymes in the
gills, therefore slowing the clearance of cate-
cholamines from the blood (Nekvasil and Olson
1986). Female fish, with high circulating estro-
gen levels, would then be expected either to reach

UNDESIRABLE MUSCLE TEXTURE

Figure 4. — Comprehensive theory of the etiology of burnt tuna
proposed in this paper.

^Although Cortisol is also known to be released during stress,
its mechanism of action is much slower than NE or E, making
it unlikely that it could exert catabolic effects within 15 min-
utes.

370

WATSON ET AL.: ETIOLOGY OF BURNT TUNA

higher maximum circulating catecholamine lev-
els or to maintain high levels for longer periods,
thus explaining the greater number of female fish
that become burnt and the seasonality of the oc-
currence of burnt tuna.

As pointed out by Davie and Sparksman (1986)
and Hochachka and Brill (1987), burnt tuna is a
quantitative change in tissue decomposition, not
a qualitative one. High blood catecholamine lev-
els at the time of death act as accelerators in this
inevitable metabolic cascade. However, because
we are dealing with a change in rate, the action of
catalysts can make all the difference as to
whether or not a fish becomes burnt due to CANP.

FUTURE RESEARCH

Under normal conditions, CANP is inhibited by
calpastatins. This class of proteins is distinct from
the other cysteine protease inhibitors, the cys-
tatins. Although the cystatins can inhibit cathep-
sins B and H, they are unable to alter CANP
activity. Conversely, calpastatin is only effective
against calcium-activated neutral protease (Bar-
rett et al. 1986; Parkes 1986). Unfortunately, al-
most nothing else is known about the relationship
of CANP to its endogenous inhibitor, either struc-
turally or physiologically. Certainly investiga-
tions into the role of calpastatin would prove
valuable for this research.

Immediate future research will concentrate on
tracking the specific action of CANP in burnt
tuna and attempting to stop this action with Ca^ ^
chelating agents such as EGTA (ethyleneglycol-
bis-(aminoethyl ether )-N,N,N',N'-tetraacetic
acid) or by use of its intracellular inhibitor, cal-
pastatin. Also, blood catecholamine levels of
stressed and unstressed fish will be measured,
along with metabolic clearance rates of nore-
pinephrine and epinephrine.

When viewed as a process of metabolic deregu-
lation of CANP, the rapid deterioration of tuna
muscle ceases to be an isolated muscular phe-
nomenon. The Z-disc disintegration characteris-
tic of burnt tuna is also present in cardiac muscle
injury due to ischemia and muscular dystrophy
(Sugita et al. 1984). Given the highly conserved
nature of muscle tissue, an understanding of
burnt tuna may also provide insights into the
metabolic processes of human disease.

ACKNOWLEDGMENTS

This research was supported by a grant to

Pacific Gamefish Research Foundation from the
Department of Land and Natural Resources, Di-
vision of Aquatic Resources, State of Hawaii.

LITERATURE CITED

Abe, H , R W Brill, and P W Hochachka.

1986. Metabolism of L-histidine, camosine, and anserine
in skipjack tuna. Physiol. Zool. 59:439-450.
Abe, H., G P Dobson, U Hoeger, and W. S. Parkhouse.

1985. Role of histidine-related compounds to intracellular
buffering in fish skeletal muscle. Am. J. Physiol.
249:R449-R454.

Amano, K., M Bito, and T Kawabata

1953. Handling effect upon biochemical change in the fish
muscle immediately after catch — I. Difference of gly-
colysis in the frigate mackerel killed by various meth-
ods. Bull. Jpn. Soc. Sci. Fish. 19:487-498.
Barrett, A J . N D Rawlings, M E. Davies. W. Machleidt,
G Salvesen, and V Turk

1986. Cysteine proteinase inhibitors of the cystatin super-
family. In A. Barrett and G. Salvesen (editors),
Proteinase inhibitors, p. 515-569. Elsevier Sci. Publ.
BV, North Holland.

Boyd, N S , N D Wilson, A R Jerrett, and B I Hall

1984. Effects of brain destruction on post harvest muscle
metabolism in the fish kahawai (Arripis trutta ). J. Food
Sci. 49:177-179.
Carey, F. G

1973. Fishes with warm bodies. Sci. Am. 228:36-44.
Carey, F G , J M Teal, J W Kanwisher, and K D Lawson.

1971. Warm-bodied fish. Am. Zool. 11:137-145.
Cramer, J L, R. M Nakamura, A E Dizon, and W N. Ikehara.
1981. Burnt tuna: Conditions leading to rapid deteriora-
tion in the quality of raw tuna. Mar. Fish. Rev.
43(6):12-16.
Dahlmann, B , L Kuehn, and H. Reinauer

1984. Proteolytic enzymes and enhanced muscle protein
breakdown. In W, H. Horl and A. Heidland (editors),
Proteases, p. 505-517. Plenum Press, N.Y.

Davie, P S , and R. I. Sparksman.

1986. Burnt tuna: An ultrastructural study of post-
mortem changes in muscle of yellowiin tuna iThunnus
albacares) caught by rod and reel and southern bluefin
tuna (Thunnus maccoyii) caught on handline or long-
line. J. Food Sci. 51:1122-1128, 1168.

FUJIMAKI, M , AND K KOJO

1953. Handling effect upon biochemical change in the fish
muscle immediately after catch — II. Changes of acid-
soluble phosphorus compounds of frigate mackerel mus-
cle. Bull. Jpn. Soc. Sci. Fish. 19:499-504.

Gibson, D J M

1981. A handbook on processing southern bluefin tuna for
the fresh chilled sashimi market in Japan. Minist.
Agric. Fish., Wellington, var. pag.

Hagberg, H

1985, Intracellular pH during ischemia in skeletal mus-
cle: Relationship to membrane {wtential, extracellular
pH, tissue lactic acid and ATP. Pfluegers Arch.
404:342-347.

Hochachka, P. W., and R W Brill

In press. Autocatalytic pathways to cell death: A new
analysis of the tuna bum problem. Fish Physiol.
Biochem. 3.

371

FISHERY BULLETIN: VOL. 86, NO. 2

HOCHACHKA, P W . W C HULBERT, AND M. GUPPY.

1978. The tuna power plant and furnace. In G. D. Sharp

and A. E. Dizon (editors), The physiological ecology of

tunas, p. 153-174. Acad. Press, N.Y.
HOCHACHKA, P. W., AND T. P MOMMSEN.

1983. Protons and anaerobiosis. Science (Wash. D.C.)
219:1391-1397.

HUDGINS. L L.. AND S. G. POOLEY.

1987. Growth and contraction of domestic fisheries:
Hawaii's tuna industry in the 1980s. In D. J. Doulman
(editor). Tuna issues and perspectives in the Pacific is-
lands region, p. 225-241. East-West Center Press, Hon-
olulu.

ITOKAWA, S

1968. On maintaining the freshness of tuna seine
catches. Reports of the Mie Prefecture Owase Fishery
Experiment Station, p. 1-15.

1969. On experiments in prevention of burnt flesh in yel-
lowfin tuna. Rep. Mie Pref Owase Fish. Exp. Stn. p.
1-9.

Jerrett. a R.

1984. Detection and prevention of burning in southern
bluefin tuna. Part A: Research results and discus-
sion. Fish Processing Bull. 4, 22 p. + app.

June, F C

1960. Spawning of yellowfin tuna in Hawaiian
waters. Fish. Bull., U.S. 54:47-64.
KONAGAYA, S., AND T KONAGAYA

1978. Denaturation at moderate temperatures of myo-
fibrillar protein of red-meat fish: A possible cause of
yake-niku. Bull. Tokai Fish, Res. Lab. 96:67-74.

1979. Acid denaturation of myofibrillar protein as the
main cause of formation of "yake-niku", a spontaneously
done meat, in red meat fish. Bull. Jpn. Soc. Sci. Fish.
45:245.

KOOHMARAIE. M . J. E SCHOLLMEYER, AND T R DUTSON.

1986. Effect of low-calcium-requiring calcium activated
factor on myofibrils under varying pH and temperature
conditions. J. Food Sci. 51:28-32, 65.
Ling, N., and R. M. G Wells

1985. Plasma catecholamines and erythrocyte swelling
following capture stress in a marine teleost fish. Comp.
Biochem. Physiol. 820:231-234.

NaKAMURA, K , Y FUJII. AND S ISHIKAWA

1977. Experiments on the prevention of "burning" of
tunas — L An examination of causes of occurrence. Bull.
Tokai Reg. Fish. Res. Lab. 90:39-43. (Engl, transl. by
W. G. Van Campen, 1980, 8 p., Transl. No. 46; available
Southwest Fish. Cent., Natl. Mar. Fish. Serv., NOAA,
Honolulu, HI 96822-2396.)

NaKANO. T . AND N TOMLINSON

1967. Catecholamine and carbohydrate concentrations in
rainbow trout iSalmo gairdneri ) in relation to physical
disturbance. J. Fish. Res. Board Can. 24:1701-1715.
Nekvasil, N P , AND K. R. Olson

1986. Extraction and metabolism of circulating cate-

cholamines by the trout gill. Am. J. Physiol. 250:R526-
R531.

Parkes, C.

1986. Calpastatins. In G. Barrett and A. Salvesen (edi-
tors). Proteinase inhibitors, p. 571-587. Elsevier Sci.
Publ. BV, North Holland.

Sakamoto, S . and N Seki

1985. Limited hydrolysis of carp myosin by cal-
pains. Bull. Jpn. Soc. Sci. Fish. 51:1551-1557.

Seki, N , and Y Kimura

1986. Degradation of troponin and tropomyosin by cal-
pain. Bull. Jpn. Soc. Sci. Fish. 52:1673-1680.

Strong. R D

1979. 1978-79 PTDF funded feasibility study to establish
an ika-shibi fishery on Guam — Final Report. Pac. Tuna
Dev. Found., May 1979, var. pag.
SuGiTA. H., S Ishiura. K Kamakura, H. Nakase, K. Hagiwara.
I Nonaka, and K Tomomatsu.

1984. Ca-activated neutral protease in physiological and
pathological conditions. In S. Ebashi, M. Endo, K. Ima-
hori, S. Kakiuchi, and Y. Nishizuka (editors). Calcium
regulation in biological systems, p. 243-256. Acad.
Press, Tokyo.
[Suisan Sekai]

1977. How to effectively kill tunas in order to maintain
quality and higher prices. Suisan Sekai 9:52-
57. (Engl, transl. by T. Otsu, 1978, 4 p., Transl. No.
25; available Southwest Fish. Cent. Honolulu Lab.,
Natl. Mar. Fish. Serv., NOAA, Honolulu, HI 96822-
2396).
Suzuki. K., S. Kawashima, and K. Imahori.

1984. Structure and function of Ca2+-activated
protease. In S. Ebashi, M. Endo, K. Imahori, S. Kaki-
uchi, and Y. Nishizuka (editors). Calcium regulation in
biological systems, p. 213-226. Acad. Press, Tokyo.
Toyo-OKA, T

1982. Phosphorylation with cyclic adenosine 3:5'
monophosphate-dependent protein kinase renders bovine
cardiac troponin sensitive to the degradation by calcium-
activated neutral protease. Biochem. Biophys. Res.
Commun. 107:44-50.
Williams, S C

1986. Marketing tuna in Japan. Queensl. Fish. Ind.
Train. Comm., Brisbane, Queenl., Aust., 60 p.
Yuen, H S H

1979. A night handline fishery for tunas in
Hawaii. Mar. Fish. Rev. 41(8):7-14.
Zeece, M G , K. Katoh, R M Robson, and F. C Parrish, Jr.
1986a. Effect of cathepsin D on bovine myofibrils under
different conditions of pH and temperature. J. Food Sci.
51:769-772, 780.
Zeece, M G , R M Robson. M L Lusby, and F C Parrish, Jr.
1986b. Effect of calcium activated protease (CAF) on
bovine myofibrils under different conditions of pH and
temperature. J. Food Sci. 51:797-803.

372

REPRODUCTIVE BIOLOGY OF
THE SPOTTED SEATROUT, CYNOSCION NEBULOSUS , IN SOUTH TEXAS^

Nancy Brown-Peterson,^ Peter Thomas,^ and Connie R. Arnold'

ABSTRACT

The spotted seatrout, Cynoscion nebulosus, spawns from April through the end of September in
shallow bays in South Texas. All females were sexually mature at 300 mm SL and males at 200 mm
SL. Histological examination of the testes showed that spermatogenesis began in January and contin-
ued until September, although spermatozoa remained in the testes until November. Spermatogenesis
was more active in the peripheral lobules of the testes than in the central lobules during the latter
half of the spawning season. The sequence and timing of final oocyte maturation (FOM) was investi-
gated for the first time in this species. Lipid coalescence began at dawn and germinal vesicle migra-
tion started by midmoming. By early afternoon, oocj^tes were hydrated and spawning occurred at
dusk.

Evidence of multiple spawning was examined. Morphological and histological data showed that
oocytes were continually recruited from March through the end of September, and the percentages of
vitellogenic and fully yolked oocytes did not decline during the spawning season. An average of only
15.5% of the vitellogenic oocytes underwent FOM and hydration during a single spawn. Postovulatory
follicles were found in many fish with mature ovaries throughout the reproductive season. Laboratory
studies showed that this species is capable of repeated spawns. Batch fecundity was best predicted by
ovary-free body weight of the fish and averaged 451 ±43 eggs/g ovary-free body weight. Estimates
of spawning frequency ranged from every other day to once every three weeks.

The spotted seatrout, Cynoscion nebulosus , sup-
ports important commercial and recreational
fisheries throughout its range (Chesapeake Bay,
Virginia to Tampico, Mexico; Tabb 1966).
Whereas aspects of the reproductive biology of
this sciaenid species have been documented
throughout its range (i.e., Chesapeake Bay:
Brown 1981; Georgia: Mahood 1975; Florida:
Moody 1950, Klima and Tabb 1959, Tabb 1961;
Mississippi: Overstreet 1983; Louisiana: Hein
and Shepard 1979; Texas: Pearson 1929, Miles
1950, 1951), a comprehensive study of reproduc-
tion in spotted seatrout does not exist for any
area. An extensive knowledge of the reproductive
life history of a species is necessary to understand
certain aspects of its reproductive physiology and
endocrinology. Most previous studies have con-
centrated on the size at sexual maturity and the
extended spawning season of this species (Pear-
son 1929; Moody 1950; Tabb 1961; Mahood 1975).
A protracted spawning season is generally char-

iContribution No. 696, University of Texas at Austin, Marine
Science Institute, Port Aransas, TX 78373.

2Marine Science Institute, University of Texas at Austin,
Port Aransas, TX; present address; Florida Department of Nat-
ural Resources, Indian River Lagoon Aquatic Preserves, 4842 S.
U.S. Highway 1, Fort Pierce, FL 34982.

3Marine Science Institute, University of Texas at Austin,
Port Aransas, TX 78373.

Manuscript accepted February 1988.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.

acteristic of multiple spawners (Nikolskii 1969).
In addition, Overstreet's (1983) histological data
suggest that C. nebulosus may be a multiple
spawner. However, the possibility that spotted
seatrout are multiple spawners has not been thor-
oughly discussed in the literature, and previous
estimates of fecundity (Sundararaj and Suttkus
1962; Overstreet 1983) have not considered the
multiple spawning nature of C. nebulosus or have
included estimates of batch fecundity.

In the present study, the reproductive biology
of C. nebulosus in South Texas was investigated,
and particular attention was given to evidence for
multiple spawning and estimates of batch fecun-
dity and spawning frequency. The spawning sea-
son, spawning sites, time of spawning, and per-
centage of running ripe fish were documented. In
addition, the temporal pattern of final oocyte
maturation was determined. The histological ap-
pearance of the gonads was examined, and partic-
ular attention was given to the presence of post-
ovulatory follicles in ovarian tissue, an indicator
of multiple spawning in the northern anchovy,
Engraulis mordax (Hunter and Goldberg 1980).
The size-frequency distribution of vitellogenic
oocytes was examined, since this can indicate
multiple spawning (deVlaming 1983). Batch fe-
cundity was determined and spawning frequency

373

FISHERY BULLETIN: VOL. 86, NO. 2

was estimated from both field-caught and labora-
tory fish.

MATERIALS AND METHODS
Collection of Samples

Spotted seatrout were collected in Redfish Bay
or Lydia Ann Channel near Port Aransas, TX,
U.SA. at depths of 1-3 m (Fig. 1). Fish were cap-
tured using a 300 m gill net (82 mm stretch) or a
300 m trammel net (outer panels, 178 mm
stretch; inner panel, 89 mm stretch) over shallow
beds of turtle grass, Thalassia testudinum or
shoal grass, Halodule wrightii, bordered by a 1-2
m drop-off into a channel. Fish <300 mm SL were
captured with hook and line in 1-2 m of water.

Samples were collected weekly or twice
monthly from March 1982 through early May
1985. No samples were taken in November and
December 1983, October and December 1984, and
February 1985. During 1982, samples were col-
lected at dusk only; in 1983, 1984, and 1985, sam-
ples from dawn, midday, and midnight were also
taken. During each sampling period, salinity and

temperature were recorded as well as the time of
capture.

Analysis of Fish and Gonads

Total length (TL) and standard length (SL)
were measured to the nearest mm for each speci-
men and total body weight (WT) was determined
to the nearest 10 g. Gonads were removed and
weighed to the nearest 0.1 g (gonad weight, GW)
and the gonadosomatic index (GSI) was calcu-
lated, using the formula: GSI = (GWAVT) x 100.

Reproductive stage of the gonads was assessed
macroscopically using the criteria in Table 1. The
macroscopic criteria used were similar to those
used by Overstreet (1983) and Macer (1974). A
small portion of tissue was removed from the
anterior or midsection of one gonad from each fish
and preserved in Davidson's fixative for histolog-
ical analysis (Jones 1966). Tissues were dehy-
drated and embedded in paraffin. Seven micron
sections were cut and stained with Harris' hema-
toxylin and counterstained with eosin. Reproduc-
tive stage of each sample was assessed microscop-
ically using the criteria in Table 1. Fish with

^Area enlarged

Gulf of Mexico

Figure 1. — Location of spotted seatrout sampling sites in South Texas. Asterisks denote sampling sites.

374

BROWN-PETERSON ET AL.: REPRODUCTIVE BIOLOGY OF THE SPOTTED SEATROUT

mature gonads were considered sexually mature.
The microscopic criteria used were similar to
those described by Yamamoto (1956), Hyder
(1969), and Macer (1974). The percentages of each
stage of oocyte, atretic structures, and postovula-
tory follicles were determined from histological
slides of the ovaries of females >305 mm SL by

lected fields of view. An oocyte was counted if
>50% of the cell was in the field of view. Two
hundred and fifty to 500 oocytes were counted for
each histological specimen.

A small portion of tissue (<0.5 g) was removed
from the center of each ovary of females >305 mm
SL collected in 1984 and 1985, placed in clearing

slides of the ovaries of females >305 mm SL by SL collected in 1984 and 1985, placed in clearing
counting all the oocytes in three randomly se- solution (6 parts ethanol, 3 parts formalin, 1 part

Table 1 . — Criteria used to describe gonadal reproductive stages of male and female spotted seatrout collected in South Texas.

Stage

Sex

Macroscopic appearance

Microscopic appearance

Immature

Regressed

Early
developing

Developing

Mature

Ripe

Running
ripe or
spawning

Partially
spent

Spent

F Ovary small, thin, light pink, slight vasculariza-

tion. GSI range: 0.39-0.55.

M Testes small, thin, light grey, appearance similar to
mesentery.

F Ovary small, light pink, vascularization more obvious

than in immature fish. GSI range: 0.39-1.16.

Testes small, thin, white, slightly larger than imma-
ture fish. GSI range: 0.04-0.20.

Ovary similar to regressed fish but slightly
larger. GSI range: 0.80-1.26.

Testes similar to regressed fish but slightly
larger. GSI range: 0.16-0.48.

Ovary visably enlarged, light yellow, highly vascular,
approximately 60% of the length of the body cav-
ity. GSI range: 0.95-2.6.

M Testes thickened, creamy white, no free milt ex-
pelled when cut. GSI range: 0.21-0.68.

F Ovary large, brilliant yellow orange, oocytes visible

to naked eye, vascularization prominent. Ovaries
80-90% length of body cavity. GSI range: 2.5-
9.6.

M Testes creamy white, thicker, more firm and elon-
gated than developing testes. A small amount of
milt expelled when cut. GSI range: 0.54-1.42.

F Oocytes hydrated, ovary looks clear, takes up al-
most entire body cavity, highly vascular-
ized. GSI range: 12.5-19.9.

M Testes white, swollen, milt does not flow with light
pressure, but flows freely when cut. GSI range:
1.07-1.62.

F Hydrated oocytes expelled with little to no pressure,

ovary fluid, fills almost entire body cavity. GSI
range: 7.7-17.6.

M Testes creamy white, milt freely flowing with slight
pressure. GSI range: 1.41-3.84.

F Ovary looks similar to mature condition but more

flacid; occupies smaller percentage of body cav-
ity. GSI range: 2.0-5.3.

M Tests looks the same as running ripe but
smaller. GSI range: 0.85-1.77.

F Ovary flacid but still highly vasculahzed, no longer

than 50% of body cavity, pinkish in color. GSI
range: 1.33-2.63.

M Testes flacid, width reduced, white. GSI range:
0.13-0.80.

Only primary oocytes present; no atretic oocytes.

Lamellar margin thin.
Only primary spermatogonia present.

Primary chromatin nucleolar and early perinuceolar
stage oocytes. Some atretic oocytes. Lamellar
margin thicker than immature, more convoluted.

Primary and secondary spermatogonia present.

Appearance of late perinucleolar oocytes and ph-
mary cortical alveoli stage oocytes. No atretic
oocytes present.

Many secondary spermatogonia and primary sper-
matocytes.

Oocytes in secondary cortical alveolar stage and
yolk granule stage. Many oocytes still in primary
cortical alveolar stage. Early yolk globular stage
present.

All stages of spermatogenesis present, with few pri-
mary spermatogonia and free spermatozoa. Pri-
mary and secondary spermatocytes predominate.

Oocytes in yolk globular stage most common. Yolk
and oil globules begin to encroach on nucleus.
Largest oocytes range from 300 to 375 y.m.

No primary spermatogonia present, some second-
ary spermatogonia. Secondary spermatocytes
and spermatids most numerous. Spermatozoa in
central lobules.

Many hydrated oocytes, irregularly shaped,
eosinophilic. Other oocytes in yolk granular and
yolk globular stages.

Spermatozoa, spermatids and secondary spermato-
cytes predominate. Few to no spermatogonia.
Central lobules filled with spermatozoa.

Same as ripe ovaries but fewer hydrated oocytes.
Atretic and postovulatory follicles may be present.

Same as ripe fish, but spermatozoa evident in sperm
ducts.

Similar to mature ovary, except atretic oocytes in
several stages of degeneration always evident.
Postovulatory follicles occasionally present.

No spermatogonia, few primary spermatoocytes.
Most active spermatogenesis in peripheral lob-
ules. Spermatozoa partially filling lobules and
abundant in sperm ducts.

Massive atresia of all remaining vitellogenic oocytes.
Many primary and chromatin nucleolar oocytes
present. Lamellar membrane highly convoluted.

Spermatozoa present in some lobules. Most lobules
small, with only primary and secondary sper-
matogonia.

375

FISHERY BULLETIN: VOL, 86, NO. 2

glacial acetic acid), and vigorously shaken for 30
seconds. Within a few minutes the cytoplasm
cleared and the germinal vesicle could be easily
observed microscopically. Ovarian fragments
were taken from females collected at the spawn-
ing site over a 24-h period, placed in clearing solu-
tion and then examined under low-power magni-
fication to determine the stage of final oocyte
maturation.

Oocyte Size-Frequency Distributions
and Estimates of Batch Fecundity

To determine fecundity and the frequency dis-
tribution of oocyte diameters, a 2-15 g piece of
tissue was removed from the midsection of the
ovaries of 57 fish and weighed to the nearest
0.01 g. The tissues were placed in a modified
Gilson's solution (Bagenal 1966) for 3-12 months
and periodically shaken to separate the oocytes
from connective tissues. Ovaries containing hy-
drated oocytes were examined after three months
since hydrated oocytes of spotted seatrout began
to disintegrate when left in Gilson's solution for a
longer period of time.

The volumetric method was used to estimate
fecundity (Bagenal and Braum 1971). The oocyte
samples were suspended in 500-1,500 mL of
water and three replicate 0.5 or 1 mL subsamples
were taken. All the oocytes >30 ^JLm were
counted, and those >80 |jim in diameter (the
growing oocytes) in each sample were measured
to the nearest 15 ^m using an ocular micrometer.
A total of 556-1,110 growing oocytes were mea-
sured in each sample. The number of resting
oocytes (oocyte diameter 30-80 ^JLm) was deter-
mined by diluting the original oocyte suspension
1:10, and counting three replicate subsamples.
Altogether, the frequency distributions of oocytes
from 48 fish were analyzed (3 in developing stage;
9 in mature, spawning not imminent stage; 14 in
mature, just prior to spawning stage; and 22 in
running ripe stage). Fecundity was calculated fol-
lowing Macer's (1974) formula and expressed as
relative fecundity of number of eggs per gram
ovary-free body weight. Batch fecundity (BF) is
defined as all oocytes >350 ixm which were
undergoing final oocyte maturation that formed a
distinct batch, and all hydrated oocytes. This defi-
nition of batch fecundity is in agreement with
Hunter and Macewicz's (1985) statement that
oocytes undergoing final oocyte maturation may
be included as hydrated oocytes when hydration
occurs very rapidly.

Spawning of Fish in the Laboratory

Four female and two male spotted seatrout
were maintained in a 30,000 L recirculating sys-
tem. The tank, filtration system and feeding
regime of the fish has been described previously
(Arnold et al. 1976). The salinity ranged from 25
to 30%(. Spawning was induced by increasing the
temperature and photoperiod from wintertime
settings of 13°C, 9L:15D to 26°C and 15L:9D
(Arnold et al. 1976). The filter boxes were checked
daily for the presence of buoyant, newly fertilized
eggs.

Statistical Analysis

Simple linear regression, oneway analysis of
variance, and analysis of covariance were com-
puted for the data using SPSS packaged programs
(SPSS 1981).

RESULTS
Size at Maturity

Some female spotted seatrout were sexually
mature after they reached 231 mm SL and >90%
of the females had reached sexual maturity at 271
mm SL (Table 2). By 300 mm SL, all female spot-
ted seatrout were sexually mature. Fish 300 mm
SL or larger made up 85.4%, and immature fish
comprised 6.5%, of all the females sampled.

Male spotted seatrout reached sexual maturity
at a much smaller size than females. The size at

Table 2. — Number and percentage of mature female and male
spotted seatrout by 10 mm size categories collected in South)
Texas, April 1982-N^ay 1985. Maturity was judged by histological
and macroscopic inspection.

Standard

Female

Male

length

(mm)

% mature

201-210

100

211-220

100

221-230

100

231-240

100

241-250

251-260

100

261-270

100

271-280

100

281-290

100

291-300

100

301-310

100

311-320

100

321-330

115

100

>330

945

100

507

100

Total

1,297

764

376

BROWN-PETERSON ET AL.: REPRODUCTIVE BIOLOGY OF THE SPOTTED SEATROUT

which most male spotted seatrout attain sexual
maturity could not be determined, since the
smallest fish collected by the sampling methods
(200 mm SL) were all sexually mature (Table 2).

Season and Time of Spawning

Histological and macroscopic examination of
the gonads (see Table 3 and Figure 3) and mean
GSI values (Fig. 2) show that spotted seatrout
have an extended reproductive season in South
Texas. Mean GSI values of males and females
increased by April 1982-85 and remained ele-
vated through the end of September. A 5°C in-
crease in water temperature at the sampling sites
to 23°C during the first week of April 1982-85
was paralleled by an increase in GSI to 2.1 in
males and 4.5 or greater in females. The pattern
of seasonal changes in mean GSI values of males
was relatively consistent during the three-and-a-
half years of sampling. Mean GSI values began to
increase in mid-February, reached a maximum of
1.9-2.4 by April, and slowly declined during the
spawning season until they dropped rapidly to
regressed levels of 0.2 by the first half of October
(Fig. 2). The seasonal patterns in mean GSI val-
ues of females were also similar from 1982 to

1985. In all four years, mean GSI increased in
April and subsequently declined in May. In 1982-
84, mean GSI increased again later in the season
prior to the final decrease to regressed levels in
October (Fig. 2). Thus, mean GSI values appear to
be bimodal, with one period of peak spawning
activity in April, and the second period of peak
spawning activity varying between August 1982
and July 1983 and 1984.

Male and female spotted seatrout in spawning
condition (males with freely flowing milt, females
with ovulated oocytes) were consistently captured
during a 2-h period around dusk over shallow (1
m) beds of Thalassia testudinum or Halodule
wrightii bordered by a channel 2 m deep. The
salinity at the spawning sites ranged from 20 to
37%c. Although actual spawning was not ob-
served, collection of newly fertilized eggs from the
spawning area at dusk confirmed that spawning
was taking place (Holt and Holf*). Spawning fish
were not captured over beds of scattered H.
wrightii or T. testudinum that were not immedi-
ately adjacent to a channel. Females with freely

^S. A. Holt and G. J. Holt, University of Texas at Austin,
Marine Science Institute, Port Aransas, TX 78373, pers. com-
mun. 1983.

1-25
17
1-9

I I I I I I I I I I I I |'i''rV;'i^|"r|T| I M I M ' M 1 M ' I M I I ly'V' I I I I M M I I I Ml I I I I I I I I I ' i ' i ' i

°C

MM J SNJMMJ
1982 1983

N J M M' J S N
1984

J M M
1985

Month

Figure 2. — Bi-monthly mean water temperature and mean gonadosomatic index (GSI) of male and female spotted seatrout collected
in South Texas from March 1982 to May 1985, including ± 1 SE of the mean. Sample size for each data point: 3-61 for females (sample
size <15 in only 9 cases) and 2-52 for males (sample size <9 in only 9 cases).

377

FISHERY BULLETIN: VOL. 86, NO. 2

flowing oocytes were only captured at dusk and
milt appeared to flow more freely in males at dusk
than at other times of the day.

Gonadal Development in Males

Testicular recrudescence began in January
with the appearance of primary spermatocytes
(Fig. 3), and by February the majority of the
males were in the early developing or developing
reproductive stage. In March, 89% of the males
had testes containing free spermatozoa (Fig. 3),
although only 27% were running ripe. Over 94%
of all males captured from April until the end of
August were running ripe. Spent males first ap-
peared in August and all the males during Octo-
ber and November were either in the spent or
regressed condition. In December, all males cap-
tured had regressed testes containing only pri-
mary and secondary spermatogonia (Fig. 3).

Histological observations revealed that the
testes are the common unrestricted spermatogo-
nia type, as described by Grier (1981). Spermato-
genesis occurred throughout the entire testis dur-
ing the majority of the year. However, from late
February through early August, spermatogenesis
was more advanced in the central lobules then in
the peripheral lobules. By August there was no
spermatogenic activity in the central lobules, al-
though spermatogenesis continued in the periph-
eral lobules until mid-September. The only period
of the year when there was no active spermato-
genesis was from October through late January,
although primary and secondary spermatogonia
were present from October through the end of

April (Fig. 3). Primary spermatocytes were com-
mon from late January until March, although
they did not disappear from the testes entirely
until the end of July. Secondary spermatocytes
first appeared in mid-February, were common
through the end of July and did not disappear
until early September (Fig. 3). Spermatids and
spermatozoa were present in the central lobules
by late February, and from March through mid-
August the central lobules were swollen with
spermatozoa (Fig. 3). In late August and Septem-
ber many of the central lobules appeared to be
partially empty of spermatozoa, although some
spermatozoa were still present up to mid-Novem-
ber (Fig. 3). The peripheral lobules contained ac-
tive spermatogenic cysts of primary and second-
ary spermatocytes during June and July and the
lumens of these lobules began to fill with sperma-
tozoa. The peripheral lobules were swollen with
spermatozoa during August and September, but
by mid-October no spermatozoa remained.

Gonadal Development in Females

Gonadal recrudescence was observed in a small
percentage of the females captured in January
and by March 94% of the females were undergo-
ing ovarian development (Table 3). Fish with ma-
ture ovaries were found from March until the end
of September, while running ripe fish were cap-
tured fi-om April through the end of September
(Table 3). Partially spent fish (females that ap-
peared to have spawned at least once but still
contained vitellogenic oocytes) were captured
from May through the end of September. No com-

1° Spermatogonia
2° Spermatogonia
1° Spermatocytes
2° Spermatocytes
Spermatids
Spermatozoa

-| r

J J

Month

— r
N

Figure 3. — Seasonal cycle of spermatogenesis in spotted seatrout collected in South Texas, as
determined by histological and macroscopic observations. Data from April 1982 through
mid-May 1985 are combined. 1° - primary, 2° - secondary.

378

BROWN-PETERSON ET AL : REPRODUCTIVE BIOLOGY OF THE SPOTTED SEATROUT

Table 3. — Percentage of female spotted seatrout In seven repro-
ductive stages by month, as assessed by histological and macro-
scopic examination of the ovaries. Data from April 1982 to May
1985 are combined. REG ^ Regressed, E DEV = Early Develop-
ing, DEV = Developing, MAT = Mature, RR = Ripe and Running
Ripe, P SP = Partially Spent, SP = Spent.

Percent in

each

reproductive

stage

Month

REG

EDEV

DEV

MAT

PSP

January

February

March

124

April

372

May

220

June

104

July

114

August

100

September

October

November

100

December

Total

1,278

tions: 150-190 (xm) varied from a lov^^ of 7.3% in
June to a late season high of 16.1%, and averaged
13% of the total number of oocytes in the ovary.
Yolk globular oocytes, the largest oocytes present
(diameter: 200-375 p.m), were most common. The
percentage of oocytes in the yolk globular stage
ranged from a high of 30.8% in April to an end-of-
season low of 22.6% in September and averaged
26.9% during the reproductive season. In addition
to the actively growing oocytes in the ovary, a
small percentage of atretic oocytes was always
found in fish with mature ovaries. Atretic oocytes
are defined here as vitellogenic oocytes undergo-
ing alpha stage atresia. The percentage of atretic
oocytes ranged from a low of 2.2% in March to a
midsummer high of 8% in July and averaged
5.4%. Additionally, postovulatory follicles (POF)
were observed in the ovaries offish captured from

pletely spent fish were captured before July and
few were captured during the remainder of the
reproductive season.

Histological observations offish with regressed
ovaries collected from late September to mid-
February showed only primary chromatin nucle-
olar and early perinucleolar oocytes. Atretic
oocytes were present from September until mid-
December; no atretic oocytes were observed from
mid-December through the end of February. The
appearance of late perinucleolar and primary cor-
tical alveoli stage oocytes in late January, Febru-
ary, and early March represented the initial
stages of ovarian recrudescence. Ovarian devel-
opment was proceeding rapidly by early March
and oocytes in the secondary cortical alveoli and
yolk granule stages were common.

During the reproductive season, histological
observations of the ovary showed a heterogeneous
morphology. Oocytes in all stages of growth (from
resting to the fully grown yolk globular stage)
were distributed throughout the ovaries of all ma-
ture, running ripe, and partially spent fish col-
lected from March through the end of September.
Vitellogenesis was probably continuous from
February until the end of September, as shown by
the continual presence of large numbers of
oocytes in the cortical alveoli and yolk granule
stages. The relative percentages of three types of
post cortical alveolar stage oocytes in fish in the
mature reproductive stage did not change
markedly from March through the end of Septem-
ber (Fig. 4). The percentage of oocytes in the yolk
granule stage (diameter in histological prepara-

Q
o
<

1-
z
m
o

LU
Q.

LU
>

I-
<

_l
LU
IT

LU
DC

P771 Yolk granule frr^ Yolk globule ^g Atretic

N I 5 I 58 I 21 I 17 I 27 I 29 I 9 I
36-1

32
28

24H
20

12-
8-
4-

M J J

MONTH

Figure 4. — Frequency of three different types of oocytes in fe-
male spotted seatrout ovaries in the mature reproductive stage
(MAT, Table 3) as assessed by histological observation. Fish
were collected from South Texas during the reproductive sea-
son. Frequency refers to the percentage of each type of oocyte
relative to the total number of all types of oocytes counted. Data
from April 1982 through the end of September 1983 are com-
bined and the number (A^) of individuals examined each month
is indicated. Bars indicate mean percentage of oocytes ±1 SE of
the mean. Yolk granule oocyte diameters ranged from 150 to
190 p-m, and yolk globule oocyte diameters ranged from 200 to
375 fim in histological preparations. Atretic oocytes are defined
here as oocytes in alpha stage atresia.

379

FISHERY BULLETIN: VOL. 86, NO. 2

April through September, although the percent-
age of POF relative to the total number of oocytes
was always small (<5%).

The histological appearance of the zona radiata
and follicle layers changed as the oocytes devel-
oped. The zona radiata (chorion) was thin and
nonstriated in perinucleolar and cortical alveoli
oocytes. The zona radiata started to take on its
characteristic striated appearance when oocytes
reached the yolk granular stage and became no-
ticeably thicker and more striated as oocytes
grew into the yolk globular stage. In contrast, the
granulosa and thecal layers appeared to decrease
in thickness as oocyte development proceeded.
Oocytes in the perinucleolar and cortical alveoli
stages had thick, well-developed granulosa and
thecal layers. In yolk granular and yolk globular
oocytes, both follicle layers were noticeably thin-
ner and the thecal layer was not always continu-
ous around the oocyte.

Frequency distributions of growing oocytes (di-
ameters in Gilson's solution >80 fxm; perinucle-
olar to hydrated stage oocytes) from ovaries of 48
fish were analyzed, and Figure 5 shows typical
patterns in fish from four reproductive stages. In
a fish captured in March with developing ovaries,
growing and vitellogenic oocytes (oocytes during
the phase of active vitellogenin uptake) ranging
from 80 to 185 jxm in diameter were present (Fig.
5A; 625 oocytes counted). Fish in the mature re-
productive stage in which spawning did not ap-
pear to be imminent, as indicated by GSI values
<4, had growing and vitellogenic oocyte diame-
ters ranging from 100 to 320 M-m (Fig. 5B; 556
oocytes counted), comprising 21% of the total
number of oocytes >30 fxm in the ovary. Fish just
prior to spawning, as indicated by GSI values >7,
had a distinct batch of oocytes with diameters

Figure 5. — Frequency distributions of growing oocyte diame-
ters O80 p.m) in female spotted seatrout collected from South
Texas. Growing oocytes comprised 21% of the total number of
oocytes >30 ^.m in the ovary. Oocyte diameter refers to the
diameters of oocytes after preservation in Gilson's solution. Fre-
quency refers to the percentage of each size of oocyte relative to
the total number of oocytes in the subsample. Each graph repre-
sents data from a single fish at a different reproductive stage.
The arrow at 350 jim indicates the minimum size necessary for
final maturation to occur. A. Developing stage in March,
GSI = 1.6. B. Mature stage in which spawning is not immi-
nent, GSI = 3.7. Fish in this stage were collected from April
through the end of September. C. Mature stage just prior to
spawning, GSI = 7.4. Fish in this stage were collected from
April through the end of September. D. Running ripe stage,
GSI = 17.6. Fish in this stage were collected from April through
the end of September.

>350 fxm that were undergoing final oocyte mat-
uration (Fig. 5C; 585 oocytes counted). However,
there were no other distinct modes of vitellogenic
oocytes. In running ripe fish, the batch of large
oocytes hydrated to a diameter of 520 ixm or
greater (Fig. 5D; 1,110 oocytes counted). The re-
sults (Figs. 5A-D) clearly demonstrate that C.
nebulosus has a continuous distribution of grow-
ing and vitellogenic oocytes. The oocyte frequency
distribution of the vitellogenic oocytes <320 \i.vc\
in diameter remaining in running ripe fish ap-

I- 2-

DC 2-J

80 200 320 440 560 680
OOCYTE DIAMETER (nm)

380

BROWN-PETERSON ET AL.: REPRODUCTIVE BIOLOGY OF THE SPOTTED SEATROUT

peared similar to the oocyte distribution in non-
spawning fish (Fig. 5B, D). The percentage (21%)
of growing and vitellogenic oocytes in the ovary
remained constant throughout the reproductive
season in fish in the mature and running ripe
stages.

Final Oocyte Maturation

Final oocyte maturation (FOM) was highly syn-
chronized in spotted seatrout and occurred only in
oocytes >400 ixm ^. Figure 6A shows a photomi-
crograph of a histological section of a spotted
seatrout ovary in the mature reproductive stage
that was not undergoing final oocyte maturation.
Many oocytes were in the yolk globular stage and
appeared to be fully grown. The first readily ob-
servable stage of FOM in "cleared" oocytes was
lipid coalescence (Fig. 6B). The oil droplets in the
oocytes began to coalesce around the germinal
vesicle (nucleus) and subsequently formed one to
three large oil droplets. This stage was not always
observed in histological preparations since many
of the oocytes were not sectioned through their
centers. The yolk globules remained discrete dur-
ing lipid coalescence. After the lipids had coa-
lesced, the germinal vesicle (GV) began to mi-
grate to the periphery of the oocyte (germinal
vesicle migration, or GVM). GVM could be seen
in both histological sections (Fig. 6C) and in
"cleared" oocytes (Fig. 6D). The oil droplet occu-
pies the center of the oocytes shown in Figure 6C,
D. Histological observation of this stage (Fig. 6C)
showed that the yolk globules were not coalesced,
the oil droplets had coalesced to form one or two
large droplets and the GV had begun to lose its
integrity and often appeared semicircular. At the
completion of GVM, the nuclear (germinal vesi-
cle) membrane broke down (GVBD) and the nu-
clear material intermingled with the cytoplasm of
the oocyte. Hydration occurred shortly thereafter,
followed by ovulation and spawning of the fully
mature oocyte.

Final oocyte maturation occurred within 10
hours in spotted seatrout in the natural environ-
ment (Fig. 7). A total of 209 fish were collected
over eight 24-h periods from April through Au-
gust in 1984 and 1985. Forty-three percent of the
fish collected between the hours of 0500 and 1500
were undergoing FOM. Lipid coalescence was

first observed at dawn (0545), and GVM started
at 0900. By 1430, all fish undergoing final matu-
ration had hydrated oocytes and ovulation and
spawning commenced at dusk (1830) and contin-
ued until 2100. None of the fish collected from
2100 to 0500 were undergoing FOM.

Batch Fecundity

The significant positive relationship {P <
0.001) between BF and ovary-free body weight
can be best described by the following equation:
BF = 459WT - 56,066, r^ = 0.56 (Fig. 8), while
curvilinear equations best described the relation-
ship between BF and SL and TL. The coefiicients
of determination in all cases were <0.56.

A one-way analysis of variance showed that
mean BF (number of eggs per gram ovary-free
weight) did not vary significantly during the
April through September spawning season. Mean
relative batch fecundity was highest in Septem-
ber, lowest in May, and varied little during April,
June, and July (Table 4).

A prominent batch of oocytes was present only
in females that were in all stages of final oocyte
maturation or were running ripe (Figs. 5C, D; 7).
The average batch size calculated from 14 fish
containing hydrated oocj^es and no postovulatory
follicles was 451 ± 43 eggs/g ovary-free body
weight. This number averaged 15.5 ± 2.5% of the
number of growing and vitellogenic oocytes in the
ovary.

Table 4. — Monthly mean batch fecundity
expressed as number eggs/g ovary-free
body weight of spotted seatrout in South
Texas. All means were not statistically dif-
ferent.

Month

Mean fecundity ±1 se

April

477 ± 42

May

320 ± 72

June

435 ± 109

July

409 ± 76

August

361 ± 60

September

560 ± 79

5A 400 (xm live oocyte equals a 350 ^.m oocyte preserved in
Gilson's solution. Both measurements represent oocytes begin-
ning FOM.

Spawning Frequency

To estimate the spawning frequency of spotted
seatrout in South Texas, the percentage of run-
ning ripe females captured monthly from April
through the end of September in 1982 through
1985 was examined. Only fish captured at dusk
and >305 mm SL were included in this analysis.
The percentage of spawning females ranged fi-om

381

FISHERY BULLETIN: VOL. 86, NO. 2

«Q

•#«,

«l

^IIf

^ Qy

»:

382

BROWN-PETERSON ET AL.: REPRODUCTIVE BIOLOGY OF THE SPOTTED SEATROUT

Figure 6. — Photomicrographs of oocytes from spotted seatrout
in the mature reproductive stage collected from South
Texas. A. Histological section of an ovary in the mature stage
that is not undergoing final oocyte maturation (FOM). Oocyte
development is continuous, with fully grown yolk globular
oocytes co-occurring with oocytes in earlier stages of develop-
ment (Magnification 120x). B. "Cleared" oocytes in the lipid
coalescence stage, the first stage of FOM (magnification 40 x).
C. Histological section of an oocyte undergoing germinal vesicle
migration (GVM). (Magnification 160 x ). D. "Cleared" oocytes
in the GVM stage (magnification 40 X). Key: GV = germinal
vesicle or nucleus, L = Lipid droplets, YG = yolk globules,
YGO = yolk globular oocyte, ZR = Zona radiata.

Lipid Coalescence

KV\\\\\\\N

GVM

14L:10D

Hydration

K\\\\\\1

ion and Spawning

kWWN

^ ^^^^^M

0000

04'00

III

0800 1200

Time

I — r^

1600 2000

Figure 7. — Time-course of final oocyte maturation in spotted seatrout collected from
South Texas. Lipid coalescence is the initial stage in final oocyte maturation. Data
obtained from 46 fish undergoing final oocyte maturation in April and May
1985. GVM = germinal vesicle migration. 14L: lOD = hours of light and dark.

Figure 8.— Relation between batch fecundity (BF)
and ovary-free body weight (WT) of spotted seatrout
from April 1984 to May 1985. Thirty-three fish with
oocytes >350 \i.m that were undergoing final oocyte
maturation and formed a distinct batch of oocytes or
had hydrated oocytes are included.

Weight (g)

383

FISHERY BULLETIN: VOL. 86, NO. 2

a low of 8% in May to a high of 45% in September
(Table 5). An average of 27.5% of the females
captured during the spawning season were run-
ning ripe, corresponding to an average spawning
frequency of once every 3.6 days or 50 times dur-
ing the 6-mo spawning period.

The spawning frequency was also estimated
from females that were undergoing final oocyte
maturation (FOM) between the hours of 0600 and
1400 in April, May, July, and August 1984 and
1985 (Table 5). For comparison, the actual per-
centage of running ripe females captured during
the same months is also presented. An average of
42.8% of the females examined for final oocyte
maturation from April through August were
undergoing FOM. Therefore, average spawning
frequency was once every 2.3 days, or 80 times
during the 6-mo spawning season.

Spawning frequency was also estimated from
the percentage of females captured with ovaries
containing postovulatory follicles (POF). An av-
erage of 13.1% of the females captured had
ovaries which contained POF ranging from 12
hours to 2 days old. This would correspond to a
spawning frequency of once every 7.6 days or 24
times during the spawning season.

Finally, the spawning frequency of four female
C. nebulosus in the laboratory under conditions of
controlled temperature and photoperiod were ex-
amined (Table 6). The fish spawned from 1 to 10
times each month for 17 months, an average of 17
spawns per individual over a 12-mo period (Table
6). This would correspond to a spawning fre-
quency of once every 21 days.

DISCUSSION
Sexual Maturity and Spawning Season

Male spotted seatrout reached sexual maturity
at a smaller size than females which appears to be
a fairly common phenomenon in spotted seatrout
throughout its range (see Mercer 1984 for a re-
view). Both male and female spotted seatrout in
South Texas reached sexual maturity at a size
similar to that reported for other groups of C.
nebulosus along the Gulf Coast (Moody 1950;
Klima and Tabb 1959; Overstreet 1983) and at a
smaller size than along the East Coast (Tabb
1961; Brown 1981).

Running ripe females were captured only be-
tween one hour before and two hours after sunset
in South Texas, suggesting a high degree of syn-
chrony in spawning fish. Collections of newly fer-

Table 5. — Percentage of spotted seatrout spawning (RR) or under-
going final oocyte maturation (FOf^) in South Texas. Data on
spawning fish collected April 1982-f^ay 1985. Data on fish under-
going FOM collected April 1984-IVIay 1985.

Spawning

fish

Fish undergoing FOM

1982-85

1984-85

Month

# RR

% RR

# FOM

% FOM

% RR

April

256

May

140

June

101

—

July

August

101

—

September

—

Mean

27.5

42.8

25.5

Table 6. — Number of spawns and average
water temperature by month for four female
spotted seatrout contained in a 30,000 L tank in
Porl Aransas, TX, under conditions of con-
trolled temperature and photoperiod, July
1974-November 1975. Expenmental proce-
dures descnbed in Arnold et al. (1976).

Month

Number of spawns

°C

July

26.75

August

26.75

September

26.0

October

25.5

November

24.5

December

24.0

January

23.5

February

24.0

March

24.0

April

25.75

May

25.0

June

22.0

July

25.75

August

25.0

September

23.0

October

23.5

November

24.0

Total

Average: 1 7

spawns/ 1 2 months/ female

tilized eggs during a 25-h period after sunset (Holt
et al. 1985) provides supporting evidence for this
spawning synchrony.

Both histological data and GSI values showed
spotted seatrout have an extended spawning sea-
son in South Texas. Gronadal recrudescence began
in January in male and in February in female
spotted seatrout. Spawning commenced in April
and continued until the end of September. Other
studies of C. nebulosus along the Gulf and East
coasts of the United States have also reported
long spawning seasons (Mercer 1984). Addition-
ally, other sciaenids have extended spawning sea-
sons (Merriner 1976; White and Chittenden 1977;
DeVries and Chittenden 1982; Love et al. 1984),

384

BROWN-PETERSON ET AL.: REPRODUCTIVE BIOLOGY OF THE SPOTTED SEATROUT

suggesting a prolonged spawning season is a com-
mon reproductive strategy among sciaenids liv-
ing in temperate and subtropical waters.

The duration of the spawning season may be
related to water temperature. Perhaps this appar-
ent association with temperature is related to the
viability and development of spawned spotted
seatrout eggs. Twenty-three degrees may be the
minimum temperature necessary for successful
spawning, as indicated by both the failure to cap-
ture running ripe fish at lower temperatures and
by data from spotted seatrout induced to spawn
under laboratory conditions (Arnold et al. 1976;
Table 6). However, since spawning ceased in Sep-
tember when water temperatures were well
above 23°C, possibly a decrease in photoperiod in
combination with a decrease in temperature pro-
vides the necessary cue for termination of spawn-
ing. Hein and Shepard (1979) suggested photope-
riod may be an important regulating factor in
C. nebulosus spawning. Data from laboratory-
spawned spotted seatrout (Arnold et al. 1976) also
support this speculation.

The seasonal pattern of mean GSI values dur-
ing the spawning season was relatively consistent
for both males and females over the Sj-yr period,
1982-85 (Fig. 2). The bimodality of the female
GSI data suggests the possibility of two peaks in
spawning activity, although the timing of the sec-
ond peak varies from year to year, thus demon-
strating the need for several consecutive years of
data. Bimodal spawning peaks have been previ-
ously reported for the species by Hein and Shep-
ard (1979) in Louisiana, Stewart (1961) in
Florida, and Brown (1981) in Chesapeake Bay,
VA. However, mean GSI values should be used
with caution when attempting to predict actual
peaks in spawning activity (deVlaming et al.
1982).

Fecundity

Accurate annual fecundity measurements are
difficult to determine for multiple spawning
fishes with an extended spawning season. Meth-
ods of calculating annual fecundity from mea-
surements of the total number of growing oocytes
at the beginning of the spawning season (Bagenal
1966), or other approaches based on the total
number of oocytes at the beginning of the spawn-
ing season minus egg retention at the end of the
spawning season (Conover 1985), are inappropri-
ate for multiple spawning species such as spotted
seatrout which show continuous recruitment of

oocytes during the reproductive season. Thus, the
previous estimates of annual fecundity in spotted
seatrout, which did not take continuous recruit-
ment of oocytes into consideration and measured
either total fecundity (Overstreet 1983) or fecun-
dity of growing and vitellogenic oocytes (Sun-
dararaj and Suttkus 1962), probably underesti-
mated annual fecundity. Furthermore, BF has
not previously been calculated for this species.

Although no monthly differences in BF were
apparent (Table 4), sample sizes were too small to
draw any definite conclusions from these data.
The relatively low coefficient of determination
(0.56) is similar to values reported by Conover
(1985) for Atlantic silversides, Menidia menidia,
another multiple spawning species. Perhaps a
more accurate estimate of annual fecundity than
previously reported for spotted seatrout can be
obtained by multiplying BF by the number of
spawns during the reproductive season. Unfortu-
nately, as discussed later, estimates of spawning
frequency vary considerably, so it is not possible
to make an accurate estimation of the annual fe-
cundity. However, available data indicate that
average annual fecundity may be greater than 10
million eggs.

Multiple Spawning

Histological examination of the testes revealed
that spermatogenesis ceased earlier in the central
lobules, although they contained spermatozoa
one-and-one-half months longer than the periph-
eral lobules. It is possible that the same central
lobules act as storage areas for spermatozoa pro-
duced by the more spermatogenically active pe-
ripheral lobules during the second half of the
spawning season, as suggested by Hyder (1969)
for Tilapia . This may represent a strategy in mul-
tiple spawning fish with a prolonged spawning
season that allows for a constant supply of sper-
matozoa while investing a minimal amount of en-
ergy into spermatogenesis.

Several lines of evidence indicate that female
C. nebulosus also spavni several times during the
reproductive season. The relatively high percent-
age of running ripe females, fish undergoing
FOM and partially spent fish captured through-
out the spawning season, and the absence of com-
pletely spent fish until the last third of the spawn-
ing season, suggest that an individual does not
spawn all the vitellogenic eggs in the ovary at one
time. Indeed, oocyte size-frequency analysis
shows a continuous distribution of growing and

385

FISHERY BULLETIN: VOL. 86, NO, 2

vitellogenic oocytes (Fig. 5) and fecundity esti-
mates show that only about 15% of the growing
oocytes undergo FOM prior to a spawn. Histolog-
ical data shows that the percentage of vitellogenic
oocytes in the ovary remains constant throughout
the spawning season (Fig. 4), which suggests that
new oocytes may be recruited into the vitellogenic
phase as rapidly as mature oocytes are released.
Convincing histological evidence of multiple
spawning is the presence of postovulatory follicles
(POF) from May through the end of September in
ovaries containing many vitellogenic oocytes.
Hunter and Goldberg (1980) characterized pos-
tovulatory follicles in laboratory-spawned En-
graulis mordax, a multiple spawning fish, and
found POF in all females that had spawned in the
laboratory one or two days previously. Finally,
laboratory studies also show that spotted seatrout
are capable of multiple spawning under relatively
constant environmental conditions (Table 6).
Tucker and Faulkner (1987) also found that six
female fish kept in raceways outdoors at the am-
bient summer temperature and photoperiod
spawned repeatedly.

Spawning Frequency

It is especially difficult to determine the
spawning frequency of wide-ranging, multiple-
spawning marine fishes such as C. nebulosus that
are not group-synchronous spawners. One
method to estimate spawning frequency is to
count the number of distinct batches of vitel-
logenic oocytes in the ovary (Shackley and King
1977). However, only one distinct batch of vitel-
logenic or hydrated oocytes can be distinguished
in spotted seatrout ovaries at any one time (Fig.
5) and the reliability of this method has been
questioned (deVlaming 1983). Therefore, three
techniques were used to estimate spawning fre-
quency in spotted seatrout.

Spawning frequency was estimated to be once
every 3.6 days from the percentage of running
ripe fish caught on the spawning grounds. Al-
though this is probably an overestimate owing to
sampling bias, the error may not be susbtantial,
since the spawning grounds are also the feeding
grounds for this species (Moody 1950) and many
nonspawning individuals were captured. The
time of sample collection did not significantly in-
fluence the estimate of spawning frequency. High
spawning frequencies were also obtained (every
2.3 days) when fish were captured 6-12 hours
prior to spawning and examined for signs of final

oocyte maturation (Table 5). The spawning fre-
quency of other sciaenids fishes has been esti-
mated by this technique (DeMartini and Foun-
tain 1981; Love et al. 1984). Additionally, Hunter
and Macewicz (1985) suggested that this method
produces a useful first approximation of spawning
frequency.

The proportion of fish having POF in the ovary
has also been used to determine spawning fre-
quency (Hunter and Goldberg 1980; Albeit et al.
1984; Hunter and Macewicz 1985). Spotted
seatrout were found to have POF throughout the
spawning season, although the age of the POF
was often difficult to determine. Furthermore, de-
tailed laboratory studies have not been under-
taken to accurately age POF in spotted seatrout.
However, the once a week estimate of spawning
frequency obtained using this method is similar
to spawning frequencies reported for two other
sciaenids, the queenfish, Seriphus politus, (De-
Martini and Fountain 1981) and the white
croaker, Genyonemus lineatus, (Love et al. 1984).
Spawing frequency estimates from POF are prob-
ably more reliable than estimates based on the
number of spawning fish since sampling bias
is less likely to occur when capturing fish with
POF.

Another method used to quantify spawning fre-
quencies in various species is direct observation
of spawning in the laboratory or in "controlled"
field situations, such as impoundments (Gale and
Deutsch 1985; Hubbs 1985; Heins and Rabito
1986). Spotted seatrout spawned an average of
once every three weeks per individual under con-
trolled temperature and photoperiod in the labo-
ratory (Table 6). Tucker and Faulkner (1987)
found that six female spotted seatrout kept out-
doors at ambient temperature and photoperiod
averaged one spawn per individual every 2.3
weeks. The same spawning frequency was noted
for an individual female, although that same in-
dividual later spawned three times in four days
(Tucker and Faulkner 1987). Thus, spotted
seatrout appear to be capable of the high spawn-
ing fi"equencies estimated from field-caught fish
with hydrated ovaries, although this frequency is
probably not sustained throughout the entire
spawning season. In general, the spawning fre-
quencies in both laboratory studies are lower
than those estimated from actual spawning fish
in the field. However, it is unclear whether this is
due to an overestimation of spawning frequency
in the field or to a decline in spawning frequency
owing to confinement in the laboratory.

386

BROWN-PETERSON ET AL.: REPRODUCTIVE BIOLOGY OF THE SPOTTED SEATROUT

ACKNOWLEDGMENTS

We thank John Trant, John Gourley, John
Smith, and Wayne Wofford for their help with the
field work and other aspects of the research.
Funding was provided by grants from the Sid
Richardson Foundation and the Caesar Kleberg
Foundation.

LITERATURE CITED

Alheit. J . V H. Alaron, and B. J. Macewicz.

1984. Spawning frequency and sex ratio in the peruvian
anchovy, Engraulis ringens. CalCOFI. Rep. 25:43-52.

Arnold. C R . T D Williams, W. A. Fable, Jr., J. L Lasswell,
AND W H Bailey.

1976. Methods and techniques for spawning and rearing
spotted seatrout (Cynoscion nebulosus ) in the laboratory.
Proc. 30th Conf Southeast. Assoc. Game Fish Comm.
30:167-178.
Bagenal. T B.

1966. The ecological and geographical aspects of the fe-
cundity of the plaice. J. Mar. Biol. Assoc. U.K. 46:161-
186.
Bagenal, T. B., and E. Braum.

1971. Eggs and early life history. In W. E. Ricker (edi-
tor), Methods for assessment of fish production in fresh
waters, p 159-181. IBP (Int. Biol. Prgramme). Handb.
3, 2d ed., Blackwell Sci. Publ. Oxford, England.
Brown, N J.

1981. Reproductive biology and recreational fishery for
spotted seatrout, Cynoscion nebulosus, in the Chesa-
peake Bay area. MA Thesis, College of William and
Mary, Gloucester Point, VA, 120 p.
Conover, D O.

1985. Field and laboratory assessment of patterns in fe-
cundity of a multiple spawning fish: The Atlantic silver-
side, Menidia menidia. Fish. Bull., US. 83:331-341.

DeMartini, E. E . AND R K. Fountain.

1981. Ovarian cycling frequency and batch fecundity in
the queenfish, Seriphus politus: attributes representa-
tive of serial spawning fishes. Fish. Bull., U.S. 79:547-
560.

deVlaming, V.

1983. Oocyte development patterns and hormonal in-
volvements among teleosts. In J. C. Rankin, T. J.
Pitcher, and R. T. Duggan (editors). Control processes in
fish physiology, p. 176-199. Wiley & Sons, N.Y.

deVlaming, V , G Grossman, and F Chapman.

1982. On the use of the gonosomatic index. Comp.
Biochem. Physiol. 73A:31-39.

DeVries, D a., and M. E Chittenden, Jr.

1982. Spawning, age determination, longevity and mor-
tality of the silver seatrout, Cynoscion nothus , in the Gulf
of Mexico. Fish. Bull., U.S. 80:487-500.
Gale, W. F., and W G Deutsch.

1985. Fecundity and spawning fi-equency of captive tesse-
lated darters - fractional spawners. Trans. Am. Fish.
Soc. 114:220-229.
Grier, H J.

1981. Cellular organization of the testis and spermatoge-
nesis in fishes. Am. Zool. 21:345-357.
HEIN, S , AND J SHEPARD.

1979. Spawning of spotted seatrout in a Louisiana estuar-

ine ecosystem. Proc. Annu. Conf. Southeast. Assoc.
Fish. Wildl. Agencies 33:451-465.
Heins, D. L , AND F. G. Rabito, Jr.

1986. Spawning performance in North American min-
nows: Direct evidence of the occurrence of multiple
clutches in the genus Notropis. J. Fish Biol. 28:343-
357.
Holt, G. J., S. A. Holt, and C. R Arnold.

1985. Diel periodicity of spawning in sciaenids. Mar.
Ecol. Prog. Ser. 27:1-7.
HUBBS, C.

1985. Darter reproductive season. Copeia 1985:56-68.
Hunter, J R , and S. R Goldberg.

1980. Spawning incidence and batch fecundity in north-
em anchovy, Engraulis mordax. Fish. Bull., U.S.
77:641-652.
Hunter J. R., and B. J. Macewicz.

1985. Measurement of spawning frequency in multiple
spawning fishes. In R. Lasker (editor). An egg produc-
tion method for estimation spawning biomass of pelagic
fish: application to the northern anchovy, Engraulis mor-
dax, p. 79-94. U.S. Dep. Commer., NOAA Tech. Rep.
NMFS 36.
Hyder, M.

1969. Histological studies on the testis of Tilapia leucos-
ticta and other species of the genus Tilapia (Pisces:
Teleostei). Trans. Am. Microsc. Soc. 88:211-231.
Jones, R. McC.

1966. Basic microscopic techniques. 5th ed. Chicago
Univ. Press. (See p. 249).
Klima, E F , AND D. C. Tabb.

1959. A contribution to the biology of the spotted weak-
fish Cynoscion nebulosus from northwest Florida, with a
description of the fishery. Fla. State Board Conserv.
Tech. Ser. 30, 24 p.
Love, M S , G E McGowen, W Westphal, R J Lavenberg, and
L. Martin

1984. Aspects of the life history and fishery of the white
croaker, Genyonemus lineatus (Sciaenidae), off Califor-
nia. Fish. Bull., U.S. 82:179-198.
Macer, C. T.

1974. The reproductive biology of the horse mackerel,
Trachurus trachurus (L.) in the North Sea and English
Channel. J. Fish Biol. 6:415-438.

Mahood, R K.

1975. Spotted seatrout in coastal waters of Geor-
gia. Proc. Annu. Conf Southeast. Assoc. Game Fish
Comm. 29:195-207.

Mercer, L. P

1984. A biological and fisheries profile of spotted seatrout,

Cynoscion nebulosus. N.C. Dep. Nat. Res. Comm. Dev.

Div. Mar. Fish., Spec. Sci. Rep. No. 40, 87 p.
Merriner, J V.

1976. Aspects of the reproductive biology of the weakfish,
Cynoscion regalis (Sciaenidae), in North Carolina. Fish.
Bull., U.S. 74:18-26.

Miles, D W

1950. The life histories of the spwtted seatrout, Cynoscion
nebulosus and the redfish, Sciaenops ocellatus. Tex.
Game Fish Comm. Mar. Lab. Annu. Rep. 1949-1950, 38

P-

1951. The life histories of the seatrout, Cynoscion nebulo-
sus, and the redfish, Sciaenops ocellatus: Sexual develop-
ment. Tex. Game Fish Comm. Mar. Lab. Annu. Rep.
1950-1951, 11 p.

387

FISHERY BULLETIN: VOL. 86, NO. 2

Moody, W. D.

1950. A study of the natural history of the spotted
seatrout, Cynoscion nebulosus, in the Cedar Key, Florida
area. Q. J. Fla. Acad. Sci. 12:147-172.
NiKOLSKn, G. V

1969. Theory offish population dynamics as the biological
backgtound for rational exploitation and management of
fishery resources. Oliver & Boyd, Edinb., 323 p.
OVERSTREET, R. M.

1983. Aspects of the biology of the spotted seatrout,
Cynoscion nebulosus, in Mississippi. Gulf Res. Rep.
Suppl. 1:1-43.
Pearson, J. C

1929. Natural history and conservation of redfish and
other commercial Sciaenids on the Texas coast. Bull.
U.S. Bur. Fish. 44:129-214.
Shackley, S E , AND P E King

1977. The reproductive cycle and its control; frequency of
spawning and fecundity in Blennius pholis L. J. Exp.
Mar. Biol. Ecol. 30:73-83.
SPSS Inc.

1981. SPSS user's guide. 2d ed. SPSS, Inc., 988 p.
Stewart, K W

1961. Contributions to the biology of the spotted seatrout
(Cynoscion nebulosus) in the Everglades National Park,

Florida. M.S. Thesis, Univ. Miami, Coral Gables, FL,

103 p.
Sundararaj, B I , and R D Suttkus

1962. Fecundity of the spotted seatrout Cynoscion nebulo-
sus (Cuvier) from Lake Borgne area, LA. Trans. Am.

Fish. Soc. 91:84-88.
Tabb, D C

1961. A contribution to the biology of the spotted seatrout,

Cynoscion nebulosus of East Central Florida. Fla. State

Board Conserv. Tech. Ser. 35, 21 p.
1966. The estuary as a habitat for spotted seatrout,

Cynoscion nebulosus. Am. Fish. Soc. Spec. Publ. 3:59-

67.
Tucker, J W , Jr , and B E Faulkner.

1987. Voluntary spawning patterns of captive spotted

seatrout. Northeast Gulf Sci. 9:59-63.
White. M L . and M E Chittenden, Jr

1977. Age determination, reproduction and population

dynamics of the Atlantic croaker, Micropogonias undula-

tus. Fish. Bull., U.S. 75:109-123.
Yamamoto, N.

1956. Studies on the formation of fish eggs. I. Annual

cycle in the development of ovarian eggs in the flounder,

Liopsetta obscura. J. Fac. Sci. Hokkaido Univ. Ser. VI,

Zool. 12:362-373.

388

NOTES

NOTES ON REPRODUCTION IN

THE SCALLOPED HAMMERHEAD, SPHYRNA

LEWINI, IN NORTHEASTERN TAIWAN WATERS

The reproductive mode of hammerhead sharks
(family Sphyrnidae) is placental viviparity, fol-
lowing Teshima's (1981) designations. Three spe-
cies of hammerhead sharks are known from
northeastern Taiwanese waters: Sphyrna lewini
(Griffith and Smith), S. mokarran (Riippel), and
S. zygaena (Linnaeus). The scalloped hammer-
head, S. lewini, commonly found from Pung-Chia
Island to Guei-Shan Island (Fig. 1) is one of the
most abundant shark species in this area. Based
on data from the Nan Fan Ao Fish Market (lo-

cated in northeastern Taiwan), 460-510 t/year
from 1982 through 1984 were landed, represent-
ing one-fourth of the total catches of sharks in
this area (the total landing being 1,760-2,240 t/
year). Catches peaked in spring and winter and
were lower in summer and autumn.

This species is also distributed in western and
southern waters of Taiwan; however, the catch
from those areas were smaller. The employees of
the southern and western fish markets explained
that the scalloped hammerhead is occasionally
seen in small numbers in those areas, but there
are no landing data. It is also common in coastal
warm temperate and tropical seas throughout
much of the world (Compagno 1984). Although

26°N

Figure 1. — Sampling area of
Sphyrna lewini.

FISHERY BULLETIN; VOL. 86, NO. 2, 1988

25°N

the scalloped hammerhead is one of the most
valuable food resources in Taiwan, many facets of
its life history, particularly reproduction, are not
well known. This study provides information
about certain aspects of reproduction in the scal-
loped hammerhead in northeastern Taiwan
waters.

Materials and Methods

From September 1982 to June 1983 and from
December 1983 to September 1985, shark speci-
mens were examined and material collected
monthly at Nan Fan Ao Fish Market. These
sharks had been caught by drift longlines set near
the surface to around 100 m, or by surface har-
poon. A total of 674 scalloped hammerhead
sharks were examined at the fish market (Table
1). Data recorded included total length measured
from the tip of the snout to the tip of the upper
lobe of the caudal fin (straight line measure),
body weight, clasper length (measured from
cloaca to the tip of claspers), ovarian egg diameter
and number, and condition of the uteri. In addi-
tion, uterine embryos were counted, sexed, and
measured for total length. Counts of litter size
included any uterine eggs as well as embryos.

Stages of maturity for both females and males
were categorized simply as "mature" or "imma-
ture". Females having threadlike uteri and tiny
ovarian eggs were called immature, while those
with eggs larger than 25 mm in diameter, with
uteri containing embryos or eggs or with empty
but expanded and flaccid uteri, were designated
as mature. Males with rigid claspers were classi-
fied as being mature. Clasper length relative to
total length also gave an indication of maturity.

Table 1 . — The number of specimens examined in
this study.

Number of

specimens

Immature

Mature

Month

Male

Female

Male

Female

Jan.

Feb.

Mar,

Apr.

May

June

July

Aug.

Sept.

Oct.

Nov.

Dec.

Total

446

Results

The reproductive organs of scalloped hammer-
head closely resemble those of the bonnethead
shark, S. tiburo (described by Schlernitzauer and
Gilbert 1966).

As with the bonnethead shark, only the right
ovary of the scalloped hammerhead is functional,
supplying both oviducts. As the ovarian eggs ma-
ture (>25 mm diameter), they pass through the
common ostium into the oviducts, where they are
fertilized. The eggs then become encased in the
embryonic membrane as they pass through the
nidamental gland and descend into the uterus. In
the uterus, embryonic development proceeds,
nourished by a yolk sac. After a period, the uter-
ine compartments develop, which enclose the em-
bryo, and a yolk-sac placenta is implanted. After
the yolk is exhausted, the embryo is nourished
until birth by the placenta through the umbilical
stalk.

Based on the condition of the uterus and ovary,
female scalloped hammerheads became mature
at a larger size, around 210 cm, than males. All
females over 230 cm were mature.

Based on the rigidity of the claspers, male scal-
loped hammerheads reached their first maturity
at a total length of 198 cm, while all those over
210 cm were mature.

The clasper length of males increases rapidly
relative to total length, until the sharks reach
around 200 cm, at which size the clasper length/
total length relationship plateaus, suggesting
that sexual maturity has been attained (Fig.
2). Also, this size marked the approximate transi-
tion point from fiaccid to rigid claspers (Fig.
2).

It takes roughly 10 months of development
from egg formation to ovulation (Fig. 3). In Octo-
ber and November, eggs were very small, measur-
ing about 2 mm in diameter. By July they had
increased in size to about 30-38 mm and num-
bered 40-50 per female, and by August and Sep-
tember they had grown to about 40-45 mm. In
one mature female, in September we counted 28
uterine eggs and 4 ovarian ones measuring 45
mm in diameter. This suggested that the 4
ovarian eggs were ready for ovulation. Because
ovarian eggs larger than 30 mm in diameter
seemed near ovulation, we concluded that ovula-
tion occurred between July and October.

The parturition season lasts from May to July,
and the gestation period of this species was esti-
mated to be roughly 10 months (Fig. 4). No em-

390

X
I 18

5 14
<

•• •

• • •• • •

• •• •

CO
CDO

ooo
o o

6 -

• : rigid
O flaccid
■ ■

120 140 160 180 200 220 240

TOTAL LE^4CTH (cm)

260

280

Figure 2. — Relationship between total length and clasper length in Sphyrna lewini.

Jan heh Mtr Apf M«y Jun lul Au* Sep < >cl No* Dec

Figure 3. — Monthly increase in the ovarian egg diameter of Sphyrna lewini. Error
bars represent 2 SD; numerals, sample size.

bryos were found in July and August, though
some mature sharks had large ovarian eggs
measuring about 30-40 mm in diameter during
those months. The first uterine eggs, and embryos
measuring around 2 cm were found in September.
Uterine eggs were observed until November.

After nine more months of rapid growth, the em-
bryos attained a total length of around 45 cm in
the period between May and June. These embryos
were regarded as full term because they were eas-
ily separated from placenta implying that partu-
rition would occur soon.

391

-i_

-L.

M J J
MONTH

N [)

Figure 4. — Monthly growth of embryos of Sphyrna lewini.
Solid triangles indicate the uterine eggs. Error represent 2 SD;
numerals, sample size.

Counts of the number of uterine embryos as
well as eggs of 110 gravid females (230-320 cm
TL) ranged between 12 and 38 (mean 25.8). The
relationship between the total number of uterine
embryos or eggs and the size can be described by
the regression equations (Fig. 5):

N = -26.105 + 0.179L,

where N is litter size, and L is total length in cm
of the female. There is considerable variation in
number of embryos with length, and the correla-
tion coefficient r is low (0.567), but obviously fe-
cundity is related to the size of the parent.

Examination of all females carrying developing
embryos showed that occasionally some uterine
eggs failed to develop. The number of nondevelop-
ing uterine eggs carried by a single female was
1-4.

As with litter size, the number of ovarian eggs
increased with the length of the adult female.

The embryos from 51 gravid females were
sexed; of a total of 1,281 embryos, 637 were fe-
males. The sex ratios of embryos differed by litter;
for instance, some individuals had predominantly
male (13:5) or predominantly female (23:10) lit-
ters. But, as a whole, sex ratio was about 1:1.

The ratio of males to females for immature
sharks in northeastern Taiwan waters was about
1:2 (42:91), but decreased to about 1:5 (95:446) for
mature individuals (Table 1). There was no ap-
parent increase in the relative number of males
in the catch during the parturition season.

Discussion

Scalloped hammerhead sharks are abundant in
the coastal seas around northeastern Taiwan, es-
pecially during the spring and winter, and are
captured by harpoon or drift longlines at or near
the surface. Rarely individuals smaller than 120
cm TL are captured because fishing gear and
strategy are likely selective for larger fish.
Clarke (1971) reported that in Hawaii, scalloped
hammerhead pups usually stay close to the bot-
tom.

Approximately equal numbers of male and fe-
male scalloped hammerhead sharks are born, al-
though a much higher proportion of females than
males are caught in the studied area. We are un-
certain whether scalloped hammerhead females
are more vulnerable to the fishing gear or are
simply more numerous in this area. Similar pre-
dominances in females in the catch have been
found in the scalloped hammerhead in the Gulf of
California (Klimley 1981; Klimley and Nelson
1981). During the summer in 1979 off Baja Cali-
fornia, females outnumbered males by 1.6 x at
Isla Cerrolvo, 3.8 x at the El Bajo Seamount, and

38-

28 -

o
>

S 20

z
18

•t

•
•

•

•
• • •

•

• •

•

• • •• • X

• • •/ •

•

■

•

• •/ • • •••

•

• • \ •

•

•• • \

•

••

N = -26.105 + 0.179L

(r=0.5665)

- y

•

, (n=110)

•

•
•

J l_

•

12-

240 280 280 300 320

TOTAL LENGTH (cm)

Figure 5. — Relationship between fecundity (number of
uterine eggs or embryos) and total length of Sphyrna
lewini females.

392

3.1 X farther north at Isla Las Animas (Klimley
1981; Klimley and Nelson 1981). Similar dispari-
ties in sex ratios were also observed in sandbar
sharks by Springer ( 1960). He felt that adult male
sandbar sharks live over a large geographical and
depth range, perhaps in deep cool oceanic waters
inaccessible to the fishermen's gear, while fe-
males occur in warmer inshore water where they
are more accessible to fishermen. He suggested
that the males move inshore only to mate. We had
insufficient data to test this suggestion in our
sample area.

Castro (1983) reported that S. lewini from
North American waters probably mature at about
180 cm; he did not mention whether this length
referred to male, female, or both. Bass et al.
(1975) reported that male scalloped hammerhead
at Mozambique matured between 140 and 165
cm, reaching a maximum length of at least 295
cm; females matured at about 212 cm, reaching at
least 309 cm. Compagno (1984) reported "maxi-
mum" sizes ranging from 370 to 420 cm. Our
largest males were 305 cm TL, and the smallest
mature male was 198 cm TL. The largest female
observed was 324 cm, and the smallest mature
female was 210 cm. It seems that in our sample
area, males reached maturity at a somewhat
larger size than in the other studied areas while
females were about the same.

The close relationship of growth pattern of
uterine embryos and ovarian eggs implies that
eggs are transferred into the uterus and fertilized
immediately after parturition. If the estimate of
10 month gestation period is correct, adult fe-
males give birth once each year.

The length at birth of scalloped hammerheads
has been reported to be around 50 cm from Natal
and southern Mozambique coastal waters (Bass et
al. 1975), 43 cm from northeastern United States
to Chesapeake Bay waters (Casey 1964), 38-45
cm in North American waters (Castro 1983), and
42-55 cm combined from all oceans (Compagno
1984). Our largest uterine embryos measured
about 47 cm TL.

Pupping season appears to be during the sum-
mer in Taiwan as well as in Mozambique (Bass et
al. 1975) and North America (Castro 1983). In
Hawaii, Clarke (1971) found newborn pups
throughout the year but with increased numbers
in summer.

Fisheries Center, National Marine Fisheries Ser-
vice, NOAA, for reading the first draft of this
manuscript and offering useful comments. We
also thank Jin-Jehn Wu, Department of Fish-
eries, National Taiwan College of Marine Science
and Technology, for his help in collecting speci-
mens. Financial support was received from Na-
tional Science Council (contract: NSC 72-0409-
B019091 and NSC 73-0409-B019-02).

Literature Cited

Bass, A. J . J D. D' Aubrey, and N Kistnasamy.

1975. Sharks of the east coast of southern Africa. III. The
families Carcharinidae and Sphymidae. Oceanogr.
Res. Inst., Durban, Invest. Rep. No. 38, 100 p.
Casey, J G

1964. Angler's guide to sharks of the northeastern United
States, Maine to Chesapeake Bay. Bur. Sport Fish.
Wildl., Circ. No. 179, 32 p.
Castro, J I.

1983. The sharks of North American waters. Texas
A&M Univ. Press, 180 p.

Clarke, T A

1971. The ecology of the scalloped hammerhead shark,
Sphyrna lewini, in Hawaii. Pac. Sci. 25:133-144.
Compagno, L J. V.

1984. FAO species catalogue. Vol. 4. Sharks of the world.
An annotated and illustrated catalogue of shark species
known to date. Psirt 2. Carchariniformes. FAO Fish.
Synop. No. 125, 4(2):251-655.

Klimley, A. P

1981. Grouping behavior in the scalloped hammer-
head. Oceans 24(4):65-71.
Klimley. A P and D R. Nelson

1981. Schooling of scalloped hammerhead shark, Sphyrna
lewini, in the Gulf of California. Fish. Bull., U.S.
79:356-360.
Schlermitzauer, D a , and P W. Gilbert.

1966. Placentation and associated aspects of gestation in
bonnethead shark, Shyrna tiburo. J. Morph. 120:219-
231.
Springer, S.

1960. Natural history of the sandbar shark, Eulamia mil-
berti. U.S. Fish. Wildl. Serv., Bull. 61:1-38.
Teshima, K.

1981. Studies on the reproduction of Japanese smooth
dogfish, Mustelus manazo, and M. griseus. J. Shi-
monoseki Univ. Fish. 29:113-119.

Che-Tsung Chen

Tzyh-Chang Leu

Shoou-Jeng Joung

Graduate School of Fisheries

National Taiwan College of Marine Science

and Technology
Keelung, Taiwan
Republic of China

Acknowledgments

We sincerely thank Susumu Kato, Southwest

393

OCCURRENCE OF YOUNG-OF-THE-YEAR KING,
SCOMBEROMORUS CAVALLA , AND SPANISH,

5. MACULATUS, MACKERELS IN

COMMERCIAL-TYPE SHRIMP TRAWLS ALONG

THE ATLANTIC COAST OF

THE SOUTHEAST UNITED STATES^

King mackerel, Scomberomorus cavalla, and
Spanish mackerel, S. maculatus , are migratory
scombrids that support large recreational and
commercial fisheries along the southeast coast of
the United States (Manooch 1979). Recent evi-
dence indicates that both species may be overex-
ploited in portions of their range, prompting the
South Atlantic Fishery Management Council to
impose catch limits and landing quotas^. Many
aspects of the biology and ecology of adult mack-
erels in this region have been studied (Manooch
et al. 1978; Collette and Russo 1984), and the
larval stages have also received attention (Fahay
1975; Collins and Stender 1987). However, little
is known concerning the distribution and occur-
rence of juvenile (young-of-the-year) mackerels
along the Atlantic coast of the southeastern
United States, nor does it seem to be widely
known that large numbers of these young fishes
may be included in the bycatch of a major fishery.
This report provides preliminary information on
both of these topics.

Methods

During 1980-82 and 1985-86 the Marine Re-
sources Monitoring and Assessment Program
(MARMAP) at the South Carolina Marine Re-
sources Research Institute conducted trawl sur-
veys of the nearshore fish fauna in the South At-
lantic Bight (Cape Hatteras, NC, to Cape
Canaveral, FL). Before 1986, trawl gear consisted
of two semiballoon shrimp trawls with an 18.3 m
footrope, a 12.2 m headrope, 4.1 cm stretch mesh
in the cod end, and 1.5 x 0.9 m doors towed at 2.5
knots (4.6 m/second). In 1986, paired "tongue"
trawls with a 22.9 m footrope, 4.1 cm mesh, and
3.0 X 1.0 m doors were towed at approximately
2.5 knots. Sampling strategy and length of tow
(20 minutes to 1 hour) varied between cruises.
Station depths were 3-18 m in 1980-81 and 3-9
m in subsequent years. In each sample, all mack-

erels were identified to species and measured
(fork length), and number and total weight were
recorded for each species. We conducted two addi-
tional cruises in 1986 to test tongue trawl nets
equipped with trawl efficiency devices (TED's). A
National Marine Fisheries Service (NMFS) TED
equipped with finfish deflector"^ was installed in
one net, a Georgia TED^ in a second net, and a
third net acted as a control. The TED's were
fished against each other and against the control
for a total of 30 tows on the first cruise (July-Au-
gust 1986) and 15 tows on the second (September-
October 1986).

Tongue nets, also known as bib, falcon, cobra,
or mongoose trawl nets, have become widely used
by commercial shrimpers in some areas during
recent years (Edwards 1987). The major differ-
ence between these and other common towed
gears used in the penaeid shrimp fishery is a mod-
ified and elongated headrope that is held well
above the footrope by attachment to the trawl
warp with a third bridle. A large float, usually
attached to the center of the headrope, produces a
high, vertical mouth opening. The result is a net
that fishes a larger portion of the water column
than other common nets with similar footrope
and door configurations.

Results

During years when semiballoon nets were used,
catch per unit effort of mackerels was relatively
low (king mackerel: 0.2-0.4 individuals/net-hour;
Spanish mackerel: 0.3-2.2 individuals/net-hour).
Tongue trawl nets on four 1986 cruises gave
cruise-specific average catches of 2.5-8.7 individ-
uals/net-hour for king mackerel and 12.4-115.2
individuals/net-hour for Spanish mackerel, using
catches of control [unmodified] nets for TED
cruises (Table 1). Both species were taken as late
as 30 October, except for two king mackerel
caught in December 1982. Approximately 79% of
king and 91% of Spanish mackerel were caught in
depths <9 m during the two years in which sam-
ple depths extended to 18 m.

Mackerels taken in shrimp trawls were almost
entirely juvenile fishes. Rather than pool catches
between cruises that often differed in time of
year, geographic area, and sampling strategy, ex-
amples of length frequencies of Spanish and king

iContribution No. 244 of the South Carolina Marine Re-
sources Center.

^South Atlantic Fishery Management Council, Council Meet-
ing Summary, 27-29 April 1987, Charleston, SC.

3Described and illustrated in the Federal Register, vol. 52, no.
124; Monday, June 29, 1987 - Rules and Regulations; p. 24244-
24262.

394

FISHERY BULLETIN: VOL. 86, NO 2, 1988.

Table 1. — Results of nearshore cruises using two types of shrimp trawls.

Cruise dates

Net type

Sampling!
strategy

Sample area2

Total
net-hour

No.
Spanish

No.
kings

July-Sept. 1980

semiballoon

SR + NR

NC,

SC, GA, FL

114

250

Apr.-June 1981

semiballoon

NC.

SC. GA. FL

Sept. 1982-
Jan. 1983

semiballoon

NC,

SC, GA, FL

Aug. -Sept. 1985

semiballoon

SR + NR

NC,

SC, GA, FL

120

Aug. -Sept. 1986

tongue

NC,

SC, GA, FL

1,421

104

Oct. 1986

tongue

NC,

SC, GA

681

481

July-Aug. 1986

tongue

2,303

Sept.-Oct. 1986

tongue

327

^SR = stratified random sampling; NR = nonrandom sampling.

2At least some portions of the waters of these states were sampled.

mackerels taken in tongue nets on two cruises are
presented in Figures 1 and 2, respectively. Com-
parisons of the catch rates of Spanish mackerel
during the TED evaluations in July- August 1986
showed that significantly fewer were taken in the
tongue net equipped with a NMFS TED than the

control net (Mann-Whitney: P < 0.001) or the
Georgia TED-net (P < 0.05) (Table 2). There were
no significant differences in catches of king mack-
erel (Table 3), or in catches of either species be-
tween nets on the September-October cruise, per-
haps due to the smaller sample sizes.

Figure 1. — Length-frequency distribution of
Spanish mackerel taken in shrimp tongue trawl nets
during July-August 1986 along the southern At-
lantic coast of the United States in = 2,303).

FORK LENGTH (cm)

395

130 -I

24 28 32 36

FORK LENGTH (cm)

Figure 2. — Length-frequency distribution of king mackerel taken in shrimp tongue trawl nets
during October 1986 along the southern Atlantic coast of the United States (n = 481).

Table 2. — Catches of Spanish) mackerel during evaluations of tongue nets equipped with trawl efficiency devices
(TED'S). C = 22.9 m footrope length tongue net; NMFS = 22.9 m footrope length tongue net with a NMFS TED;
GA = 22.9 m footrope length tongue net with a Georgia TED. Tow times = 1 hour.

No. of

Net

No. of Spanish mackerel

Net

No. of Spanish mackerel

Cruise

Comparison

tows

type

Total

X ±SD

Range

type

Total

X ±SD

Range

July-Aug.

C-NMFS

1,104

110 ±27

68-144

NMFS

519

52 ±23

12-73

C-GA

1,199

120 ±65

40-276

1,219

122 ± 74

44-300

NIVIFS-GA

NMFS

880

88 ±53

35-176

1,650

165 ±142

67-533

Sept.-Oct.

C-NMFS

104

21 ± 15

2-41

NMFS

11 ± 11

2-27

C-GA

206

41 ±38

1-103

155

31 ±28

0-76

NI^FS-GA

NMFS

108

22 ±23

4-58

149

30 ±25

3-59

Table 3. — Catches of king mackerel during evaluations of tongue nets equipped with trawl efficiency devices
(TED'S). C = 22.9 m footrope length tongue net; NMFS = 22.9 m footrope length tongue net with a NMFS TED;
GA = 22.9 m footrope length tongue net with a Georgia TED. Tow times = 1 hour.

Comparison

No. of
tows

Net
type

No.

of king

mackerel

Net
type

No.

of king

mackerel

Cruise

Total

X ±

Range

Total

X ±

Range

July-Aug.

C-NMFS

4±

1-13

NMFS

5±

-30

C-GA

4±

0-15

4 ±

-13

NMFS-GA

NMFS

2±

0-8

3±

-17

Sept.-Oct.

C-NMFS

5±

0-15

NMFS

3±

-5

C-GA

8±

4-22

4 ±

-14

NMFS-GA

NMFS

5±

2-8

4±

-11

Discussion

Although it is possible that juvenile mackerels
were more abundant in 1986 than in previous
years, the increased catches of these fishes in
tongue nets over semiballoon nets suggests that
the former are much more efficient in capturing
these fishes. Preliminary data from a gear com-

parison cruise in 1987 indicate that tongue nets
do catch more pelagic fishes than semiballoon
nets even after adjusting for differences in
footrope lengths (G. Sedberry"*). Unfortunately,

■*G. Sedberry, Marine Resources Research Institute, South
Carolina Wildlife and Marine Resources Department, P.O. Box

396

these tests were conducted during March-April
when juvenile mackerels are rare in the coastal
waters of South Carolina.

In 1986 we collected juvenile mackerels during
July through October. Because of incomplete tem-
poral sampling, we do not know if they were
present earlier and later in the year in this re-
gion. Based on the occurrence of early larval
stages, spawning of both mackerels in the South
Atlantic Bight extends from May through at least
September (Collins and Stender 1987). If growth
rate estimates of ca. 3 mm/day for juveniles are
correct (M. R. Collins, unpubl. data), king mack-
erel spawned in early May could be recruited into
the bycatch of the shrimp fishery in June. Late-
spawned fish from the previous year may also be
present at this time. In South Carolina, the open
season for commercial trawling for penaeid
shrimps in state waters usually extends from
June through December, which coincides with the
presence of juvenile mackerels in the heavily
fished nearshore waters. In addition, mackerels
were much more abundant in tows made in
depths <9 m, which includes the preferred
shrimping areas, than in deeper waters. This may
be due either to greater abundance in these
depths or to greater catchability in response to
the fact that the trawl nets fish a larger portion of
the water column in shallower areas.

It is difficult to accurately estimate the bycatch
of mackerels in the commercial shrimp fishery
owing to lack of current, detailed information
from throughout the region on number of vessels,
effort expended, gears used, and areas fished.
However, our catch rates suggest that the impact
of tongue nets on mackerel stocks may be signifi-
cant. As the current status of these stocks is such
that strong restrictions have been imposed on
both the recreational and commercial fisheries, it
is unfortunate that the situation may be exacer-
bated by a potentially large bycatch of juvenile
mackerels in the shrimp fishery. More informa-
tion is needed on the ecology and behavior of
young mackerels, and their vulnerability to vari-
ous gears, in order to resolve this conflict.

Acknowledgments

We wish to express our appreciation to the
many people who assisted in the tedious tasks of
collecting and processing trawl samples over the

years, including the captains and crews of the
vessels involved. This work was sponsored by the
National Marine Fisheries Service (Southeast
Fisheries Center), the South Carolina Wildlife &
Marine Resources Department, and the Sport
Fishery Research Foundation.

Literature Cited

COLLETTE. B B . AND J L RUSSO

1984. Morphology, systematics, and biology of the

Spanish mackerels (Scomberomorus , Scombridae).

Fish. Bull., U.S. 82:545-692.
Collins, M R . and B W Stender.

1987. Larval king mackerel (Scomberomorus cavalla),
Spanish mackerel (S. maculatus), and bluefish iPo-
matomus saltatrix ) off the southeast coast of the United
States, 1973-1980. Bull. Mar. Sci. 41:822-834.
Edwards. M L.

1987. The shrimp trawl. Natl. Fisherman 67(13):89-92.
Fahay, M P

1975. An annotated list of larval and juvenile fishes cap-
tured with surface-towed meter net in the South Atlantic
Bight during four RV Dolphin cruises between May 1967
and February 1968. U.S. Dep. Commer., NOAA Tech.
Rep. NMFS SSRF-685, 39 p.
Manooch. C S , III

1979. Recreational and commercial fisheries for king
mackerel, Scomberomorus cavalla , in the South Atlantic
Bight and Gulf of Mexico, U.S.A. In E. L. Nakamura
and H. R. Bullis (editor), Proceedings of the mackerel
colloquium, p. 33-41. Gulf States Mar. Fish. Comm.,
Brownsville, TX.
Manooch. C S , III. E L. Nakamura, and A B Hall

1978. Annotated bibliography of four Atlantic scombrids:
Scomberomorus brasiliensis , S. cavalla, S. maculatus,
and S. regalis. U.S. Dep. Commer., NOAA Tech. Rep.
NMFS Circ. 418, 166 p.

Mark R Collins
Charles A. Wenner

Marine Resources Research Institute
South Carolina Wildlife and Marine

Resources Department
P.O. Box 12559
Charleston, SC 29412

12559, Charleston, SC 29412, pers. commun. May 1987.

STOMACH CONTENTS OF COMMERCIALLY

CAUGHT HUDSON RIVER STRIPED BASS,

MORONE SAXATILIS , 1973-75

The Hudson River estuary is a detritus-driven
ecosystem. Only a few of the 100 or more reported
fish species function as tertiary piscivores more
typical of a grazing food chain. Of these few spe-
cies, which include the American eel, Anguilla
rostrata, and the summer-transient juvenile

FISHERY BULLETIN: VOL. 86, NO. 2. 1988.

397

bluefish, Pomatomus saltatrix, striped bass, Mo-
rone soxatilis, are probably the most important in
terms of biomass and commercial value. How-
ever, commercial fishing for the American eel and
striped bass has been banned in the Hudson since
1976 because both species are contaminated with
higher levels of toxic polychlorinated biphenyls
(PCB's) than most other Hudson River fish spe-
cies.

To date, only one published paper has touched
upon the diet of prespawning Hudson River
striped bass larger than 400 mm TL (Gardinier
and Hoff 1982), and their results for fish of that
size are based on less than 10 samples for which
fish remains were identified to species or family
level. The present paper describes the findings of
stomach content analysis for 510 striped bass,
most of which were prespawning adults larger
than 400 mm TL (Fig. 1) collected during the
spring months of 1973-75 by commercial gill nets
in the Tappan Zee region of the Hudson River.

Description of Sampling Area and Methods

The Tappan Zee gill-net fishery is located ap-
proximately 90 m south of the Tappan Zee Bridge
at river km 43.5 (measured from the Manhattan
Battery) in a relatively shallow section that is 4.5
km wide, a location through which many of the
spawners move on the way to upstream spawning
grounds. The staked gill nets were set perpendic-
ularly to the north-south current in water depths

of 2.4-9.8 m and spanned a distance of approxi-
mately 1.6 km out to the dredged channel on the
east side of the river. Stretched mesh sizes were
11.4, 11.8, 12.1, 12.7, 13.3, and 14.0 cm. Water
temperatures during the sampling period ranged
from approximately 6° to 19.5°C, and the salinity
was essentially that of freshwater, rarely rising
above 300 mg/liter. Sampling dates included the
following periods in each year: 28 March-9 May
1973, 8 April-20 May 1974, and 6 April-19 May
1975. Fish were collected from the nets at 8-12 h
intervals, placed on ice, and taken directly to a
laboratory for total length (TL) measurements
and stomach content analysis. Prey items were
identified with the aid of a binocular dissecting
scope and frequency of occurrence was noted.

Results

Stomach Content Analysis

A summary table of stomach content data,
pooled for the three years of analysis (Table 1),
shows that adult, prespawning striped bass were
highly piscivorous and that 89% of the stomachs
containing identified food items contained fish
and 21% contained invertebrate remains, primar-
ily those ofCrangon species. Clupeid species were
the most prevalent fish, with blueback herring,
Alosa aestivalis, predominating. Most of the clu-
peids were adult, prespawning herring, approxi-
mately 200 mm TL.

NUMBER OF FISH = 510

140

400 450 500 550 600 650 700 750 800 850 900 950 1000 1050

TOTAL LENGTH (mm)

Figure 1. — Length distribution of Hudson River striped bass examined for stomach con-
tents.

398

Table 1— Stomach contents of adult prespawning striped bass collected by commercial gill net during

spring 1973, 1974, and 1975.

% Frequency of

% Frequency of occurrence

Number of

occurrence based on

based on number of

stomachs

total stomachs

stomachs containing

in which

analyzed

identified food items

Food item

item occurred

(n = 510)

(n =201)

Fish

179

35.1

89.1

Clupeidae

14.3

36.3

Alosa pseudoharengus

2.5

6.5

Alosa aestivalis

5.7

14.4

Brevoortia tyrannus

0.6

1.5

Clupeidae

5.5

13.9

Morone spp.

1.8

4.5

Morone americana

1.2

3.0

Morone saxatilis'^

0.2

0.5

Morone spp.

0.4

1.0

Other fish

5.7

14.4

Ammodytes americanus

2.7

7.0

Microgadus tomcod

1.4

3.5

Osmerus mordax

0.4

1.0

Syngnathus fuscus

0.4

1.0

Anchoa mitciiilii

0.4

1.0

Ictalurus catus

0.2

0.5

Scomber scombrus

0.2

0.5

Unidentified fish remains

13.3

33.8

Invertebrates

8.4

21.4

Crangon spp.

7.5

18.9

Polyctiaeta

1.0

2.5

Unidentified contents

2.2

Not applicable

Total stomachs analyzed

510

Number (percent) containing identified food items

201 (39.4)

Number (percent) containing food material (identified and unidentified)

212 (41.6)

Number (percent) empty

296 (58.4)

'Two yearling striped bass in female collected 15 April 1974.

The next most abundant fish prey species in the
composite 1973-75 sample was the American
sand lance, Ammodytes americanus. Sand lance
were present in striped bass stomachs only during
1975, when the frequency of occurrence was about
ll9c (Fig. 2). Atlantic tomcod, Microgadus tom-
cod, were found in striped bass stomachs only
during 1973 when the frequency of occurrence
was about 20%, somewhat higher than the fre-
quency of clupeid species for that year (Fig. 2).
These fish were yearling, postspawning Atlantic
tomcod averaging about 125 mm TL. The Morone
species, white perch, M. americana, and striped
bass, had a frequency of occurrence of 4.5% for the
composite sample and were present in striped
bass stomachs every year (Fig. 2).

Discussion

Similar to findings in this study, Trent and
Hassler (1966) found that blueback herring were
predominant in the diet of adult, prespawning
striped bass from the Roanoke River, NC, and
Manooch (1973) found that herring ranked second
after Atlantic menhaden, Brevoortia tyrannus , in

the spring diet of adult striped bass from Albe-
marle Sound, NC. Data from Gardinier and HofF
(1982) do not indicate that blueback herring or
clupeids were important in the diet of prespawn-
ing Hudson River striped bass collected during
1976 and 1977, but fish remains in bass stomachs
were generally not identified to the family or spe-
cies level in their study.

American sand lance are rarely found over
muddy bottoms (Bigelow and Schroeder 1953;
Leim and Scott 1966) or within estuaries (Mass-
man 1960; Norcross et al. 1961), but they ranked
second only to blueback herring in the diet of
Hudson River striped bass. Sand lance were not
reported from the Hudson River prior to 1975
when adults and larvae were relatively abundant
in February-April ichthyoplankton collections
south of the Tappan Zee Bridge (Dew and Hecht
1976). Data from the present study (Fig. 2) and
from early spring ichthyoplankton sampling dur-
ing 1976 (Dew, unpub. data), indicate that sand
lance are not abundant every year in the Hudson,
but it is probable that during some years adult
sand lance serve as important alternate prey for
prespawning striped bass.

399

o
o

>-
o

100

.. 60

^ CLUPEIDS

[^ MORONE SPP.

^ ATLANTIC TOMCOD

□ SAND LANCE

□ UNIDENTIFIED FISH REMAINS

N = 36

-iii

1973

N = 39

N =126

1974

1975

N = 201

^±.„L^

1973- 1975

Figure 2. — Percent frequency of occurrence offish prey categories based on number of stomachs containing identified food.

The fact that Atlantic tomcod were found only
during 1973 may be due to a sampling artifact
because 1973 was the only year in which March
samples were taken, and all tomcod prey were
found during March 1973. In March, gravid
alewives, Alosa pseudoharengus , and blueback
herring have not yet moved into the estuary and,
prior to the arrival of these apparently preferred
prey species, Atlantic tomcod may serve as an
alternate forage fish. Thus the winter-spawning
Atlantic tomcod may be a regular diet item at a
time when alternate prey species are not avail-
able.

It is evident from Table 1 that there are several
alternate fish prey for striped bass in the Hudson
River (e.g., blueback herring, sand lance, alewife,
Atlantic tomcod, white perch, and striped bass),
and it is expected that striped bass predation
upon these species would be of a compensatory
nature. In a simple predator-prey situation where
there are no alternate prey, predation is often
depensatory (Neave 1953) because a population of
predators would tend to seek out and consume a
high proportion of small year classes and a lower
proportion of strong year classes. In a more com-
plex community where there are several species
of prey, predation may be compensatory if preda-
tors change their feeding habits in response to the
availability of food (Ivlev 1961; Forney 1971).
That is, if several equally suitable prey species
were available within a system, the predator spe-

cies would tend to feed more heavily on the most
abundant species, thus acting in a compensatory
manner.

The ratio of white perch to striped bass in stom-
ach samples identified to the species level is 6:1
(Table 1). This ratio is based on a small sample
size, but it closely approximates the long-term
average ratio of 5.9:1 for annual CPUE values for
white perch and striped bass collected in several
hundred bottom trawls from the Haverstraw Bay
region (Milepoint 36) during 1971-77 (Lawler,
Matusky and Skelly Engineers unpubl. data). In
other words, yearling and older white perch and
yearling striped bass may be equally attrac-
tive prey species, and their frequency of oc-
currence in adult striped bass stomachs may
depend primarily upon the frequency of ran-
dom encounters rather than active prey selec-
tion. If this were true, the frequency of can-
nibalism should be greatest during those years
when the abundance of yearling striped bass rela-
tive to white perch (and other species) was great-
est.

Acknowledgments

Financial support for this project came from
Lawler, Matusky and Skelly Engineers; Con-
solidated Edison of New York; Central Hudson
Gas and Electric; and Orange and Rockland
Utilities.

400

Literature Cited

BiGELOW. H. 3., AND W. C. SCHROEDER

1953. Fishes of the Gulf of Maine. Fish. Bull. 53:1-
577.
Dew. C B . and J H Hecht

1976. Observations on the papulation dynamics of At-
lantic tomcod Microgadus tomcod in the Hudson River
estuary. Proc. Fourth Symposium On Hudson River
Ecology, Paper 25. Hudson River Environmental Soci-
ety, Bronx, N.Y.
Forney, J L

1971. Development of dominant year classes in a yel-
low perch population. Trans. Am. Fish. Soc. 100:739-
749.

GaRDINIER. M N , AND T B HOFF

1982. Diet of striped bass in the Hudson River estuary.
New York Fish Game J. 29(2):152-165.
IVLEV, V S.

1961. Experimental ecology of the feeding of fishes.
Yale Univ. Press, New Haven, 302 p.
Leim. a H . AND W B Scott

1966. Fishes of the Atlantic coast of Canada. Fish. Res.
Board Can. Bull. 155, 485 p.
Manooch, C S.

1973. Food habits of yearling and adult striped bass, Mo-
rone saxatilis (Walbaum), from Albemarle Sound, North
Carolina. Chesapeake Sci. 14(2):73-86.
Massmann, W H

1960. Additional records for new fishes in Chesapeake
Bay. Copeia 1960:1-70.

Neave, F.

1953. Principles affecting the size of pink and chum

salmon populations in British Columbia. J. Fish. Res.

Board Can. 9:450-491.
Norcross, J J , W H Massmann, and E. B. Joseph

1961. Investigations of inner continental shelf waters off
lower Chesapeake Bay. Part II. Sand lance larvae,
Ammodytes americanus. Chesapeake Sci. 2(l-2):49-
59.

Trent, L., and W. W. Hassler.

1966. Feeding behavior of adult striped bass, Roccus sax-
atilis , in relation to stages of sexual maturity. Chesa-
peake Sci. 7(4):189-192.

C. Braxton Dew

Lawler, Matusky, and Skelly Engineers

One Blue Hill Plaza

Pearl River, NY 10965

Present address;

Northwest and Alaska Fisheries Center Kodiak Laboratory

National Marine Fisheries Service, NOAA

P.O. Box 1638

Kodiak, AK 99615

ACCUMULATION OF AGE PIGMENTS

(LIPOFUSCIN) IN

TWO COLD-WATER FISHES

In fisheries management, age structured models
are the preferred method for meeting the key ob-
jectives of estimating optimal yields and deter-
mining the effect of fishing on stock structure
(Gulland 1978). However, fev^^ species of commer-
cial marine fishes exist in which age can be deter-
mined with certainty (Boehlert 1985). The con-
centration of age pigments (lipofuscin)
(Ettershank 1984) in fish tissues (Agius and
Agbede 1984) may be a measure of fish age that
could be used to validate other ageing techniques
and might also improve estimates of age of long-
lived species where other techniques are difficult
to apply.

Lipofuscin originates in biological membranes
through lipid peroxidation (Tappel 1975). Lipo-
fuscin accumulation has been documented for a
wide variety of animals, from mammals to the
bread-mold Neurospora (Ettershank et al. 1983
and references therein). The rate of accumulation
has been shown to be constant during the lifetime
of laboratory-raised mice (Reichel 1968; Miquel et
al. 1978), dogs (Munnell and Getty 1968), fiesh-
flies (Ettershank et al. 1983); man (Strehler et al.
1959); and also wild populations of mice (Dapson
et al. 1980). On the other hand, the rate of lipofus-
cin accumulation has been shown to vary with
level of activity and lifespan (Sohal and Donato
1978). It is expected that the rate in which lipo-
fuscin accumulates with age in a natural popula-
tion of fishes, or stock, with its free genetic inter-
change and likely common habitat, would be
fairly uniform (but see Smith 1987). In addition,
if measurements are made in nonmitotic and
constantly metabolizing tissues such as brain
or myocardium, the variation in concentra-
tion due to environmental effects is least like-

ly.

In this paper we present the results of a prelim-
inary study designed to assess the usefulness
of extracted lipofuscin as a method of ageing
fishes. Two species of cold-water fishes are in-
cluded: the rainbow trout, Salmo gairdneri,
reared in captivity, and of known age; and the
Dover sole. Microstomas pacificus, a long-lived
fish, in which age is not known with certain-
ty. We present spectral characteristics of the ex-
tracted lipofuscin from several tissues and the
change in concentration of lipofuscin with fish
age.

FISHERY BULLETIN: VOL. 86, NO. 2

401

Materials and Methods

Specimens of Dover sole were collected on De-
cember 1986 ofTPoint Conception. Ages were esti-
mated by J. Butler, E. Lynn, and M. Drawbridge
of the Southwest Fisheries Center using otolith
sections. Ages ranged from 2.7 to 44.7 years (av-
erage of three readings). Specimens of the species
Salmo gairdneri were collected in July 1986 at
Hot Creek Hatchery. Ages ranged from 3 months
to 3 years.

Samples were frozen after collection and kept
at -80°C until analysis. Studies by Nicol (1987)
showed that this form of preservation yielded the
lowest fluorescence when compared with ethanol
and formalin-preserved samples. There was no
indication that the level of fluorescence changed
as a function of time of preservation or by inter-
action with the extracted lipofuscin, as in the
case for formalin. To excise the brain, the top of
the skull was opened and the four brain lobes
were removed as a unit, without the optic nerve.
To excise the heart, we separated the muscle
from connective tissue, blood vessels, and fat de-
posits.

Three methods were compared for maximum
lipofuscin extraction. The first two methods (Tap-
pel 1975; MacArthur and Sohal 1982) were specif-
ically developed for lipofuscin extraction and
employed chloroformrmethanol (2:1) as the ex-
tractive solvent; they differ in the optimal
volume-to-weight ratio (30:1 and 20:1, respec-
tively), temperature of extraction, and number of
times the chloroform phase is washed with water.
The third method was developed for lipid extrac-
tion in fishes (Bligh and Dyer 1959) and uses chlo-
roform:methanol:water (1:2:0.8) as the extractive
solvent. Subsamples (n = 3) of cerebellum of
Stenella sp. collected at the eastern tropical
Pacific and kept at -30°C were defrosted, dried
with lint-free paper, weighed, and extracted fol-
lowing the three methods as described originally.

Tissues from Dover sole were extracted basi-
cally following the MacArthur and Sohal (1982)
technique, with two additional steps to ensure a
complete washing out of flavoproteins and pho-
tooxidation of retinol when present (Fletcher et
al. 1973). Tissue was freeze dried prior to analy-
sis. Ten mL of chloroform: methanol (2:1, v/v) was
added to a homogenizer containing the sample for
a final solvent:sample ratio of about 120:1 (vol/
dry weight). The sample was ground with a teflon
pestle attached to an electric drill and later sub-
merged in a 40°C water bath for 1 minute. A 2 mL

subsample was taken from the homogenate.
Three mL of deionized water was added, the sam-
ple was shaken, and the emulsion centrifuged 10
minutes at 1,912 g and 0°C. After centrifugation,
the hyperphase was decanted and a second rinse
performed. After decanting the hyperphase a sec-
ond time, 1 mL of the hypophase (chloroform con-
taining lipofuscin) was sampled and transferred
to a polypropylene tube. Three mL of chloroform
were added to the sample (for a total of 4 mL)
which was then exposed to UV irradiation (254
nm) for 1-2 minutes to photooxidize retinol. This
last step was routinely performed for liver tissue.

Samples were then transferred to glass tubes,
sealed, and kept in the refrigerator until analysis.
Sample fluorescence was measured at the emis-
sion peak (430-440 nm) in a quartz cuvette with
a Perkin Elmer Fluorescence Spectrophotometer^
Model MFP-44A. The sample was excited at the
peak of fluorescence excitation (—360 nm). The
intensity of the fluorescent emission (at 430 nm)
was normalized to the intensity of the quinine
solution standard (1 mg L"^ in IN sulphuric acid)
and expressed in fluorescence units.

Lipofuscin in Dover sole is expressed as 1) total
lipofuscin content per organ and 2) weight-
specific lipofuscin concentration, calculated by di-
viding the total lipofuscin content by the dry
weight of the entire organ.

Lipofuscin ft-om rainbow trout tissues was ex-
tracted following the Bligh and Dyer (1959) tech-
nique, without modifications. Whole tissues were
ground in water with a tissue homogenizer to give
a final concentration of 100 mg of tissue (wet
weight) in 0.8 mL of homogenate. A sample of 0.8
mL was taken and solvents were added to give a
final ratio of 1:2:0.8 (chloroform:methanol:water)
with a solvent to sample ratio of 20:1 (vol/w). The
sample was then filtered through a 2.4 cm glass
fiber filter (Whatman GF/C), the tissue re-
extracted with 1 mL of chloroform and refiltered.
The extract was then washed with 3 mL of water,
shaken, and centrifuged for 10 minutes at 1,912 g
and 0°C. One mL of the chloroform hypophase
was subsampled and fluorescence estimated as
described above.

Lipofuscin in rainbow trout is expressed as

1) total lipofuscin content per organ and

2) weight-specific lipofuscin concentration (total
lipofuscin content of the organ divided by its wet
weight).

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

402

Results

EfiBciency of Extraction

Lipofuscin was successfully extracted and
quantified from three different fish tissues: brain,
heart, and liver. Wavelengths of fluorescence ex-
citation and emission maxima of extracted lipo-
fuscin in chloroform are presented in Figure 1.
All maxima are within the range cited by Shel-
dahl and Tappel (1973); fluorescence excitation
maxima were between 340 and 370 nm and fluo-
rescence emission maxima between 420 and 470
nm.

Retinol (wavelengths of maximum fluorescence
excitation at 325-340 nm and maximum fluores-
cence emission at 475 nm) was photooxidized by
exposing the chloroform extract to UV irradiation
(254 nm). As expected, retinol was found in liver
and sometimes it was present in brain and heart
tissues. For example, wavelengths of fluorescence
excitation and emission maxima in liver of Dover
sole shifted from 352 to 365 nm and from 470 to
440 nm, respectively, after UV irradiation (see
Figure IC). Mullin and Brooks (1988) also found
that this UV irradiation is effective in oxidizing
retinol although they did not find significant
retinol interference in fish tissue. It seems liver
tissues may require UV irradiation to oxidize
retinol while brain and heart tissue should be
checked for retinol presence before irradiation.

The extractive efficiency of the three methods
tested are compared in Table 1. All methods ex-
tracted similar fluorescent compounds from the
brain oi Stenella sp., as the fluorescence excita-
tion and emission spectra were similar. The
MacArthur and Sohal (1982) method extracted

Table 1. — Comparison of the three methods of lipofuscin extrac-
tion in brain tissue Fluorescence excitation at 360 nm and fluores-
cence emission at 440 nm. Results are presented as fluorescence
units per mg of wet tissue, where the fluorescence emission signal
is normalized to the intensity of emission of a standard quinine
sulfate solution (1 mg L Hn IN sulphuric acid). The three methods
were significantly different from each other, P < 0.05, Newman-
Keuls range test (Zar 1974). ANOVA: F ratio = 21.81, 2 df,
P < 0.05.

Fluorescence
units

Fluorescence
maxima

Average
10-3

SD
10-3

Excitation
(nm)

Emission
(nm)

f^acArthur and
Sohal (1982)

Tappel (1975)

Bligh and Dyer
(1959)

4.30
2.83

1.98

0.39
0.37

0.45

3
3

362
362

360

445
440

445

3E0 360 400 440 480 520

(/)

Z
LiJ

UJ
U

(r
o

365

435

1 1 1

AZ)

320 360 400 440 480 520

365

430

450

t(l) //

\(3) N

\ r

\ \

\ N

\ \
\ S

\ X

X X

\^ X

1 1

1 1 1

320 360 400 440 480 520

WAVELENGTH (nm)

Figure 1. — Spectral characteristics of lipofuscin extracted from
Dover sole, Microstomus pacificus, tissues (uncorrected spec-
tra). In chloroform. (A) Brain: (1) Lipofuscin fluorescence exci-
tation spectrum (emission wavelength 440 nm); (2) lipofuscin
fluorescence emission spectrum (excitation wavelength 365
nm). (B) Heart: (1) Lipofuscin excitation spectrum (emission
wavelength 430 nm); (2) lipofuscin fluorescence emission spec-
trum (excitation wavelength 365 nm ). (C ) Liver: ( 1 ) Lipofuscin
fluorescence excitation spectrum (emission wavelength 430
rmi); (2) lipofuscin emission spectrum (excitation wavelength
365 nm) before UV radiation; (3) Same as (2) after UV radiation
(254 nm).

significantly more fluorescent pigment, than did
either the Tappel (1975) or the Bligh and Dyer
(1959) method which extracted 66% and 46% of
maximum extraction, respectively). Thus all
three methods are useful for quantitative estima-
tion of extractable lipofuscin in fish tissue but

403

comparisons of amount of concentrations between
tissues or species cannot be made if different ex-
traction methods are used.

Lipofuscin Concentration

The total concentration of extracted lipofus-
cin in Dover sole brain tissue was positively
correlated with fish length (Fig. 2A)
{Y = -15.4 + 0.516X,r2 = 0.43;6 ^0,P<0.01,
t - 5.1, df=34). Furthermore, total extracted
lipofuscin content in the brain was positively cor-
related with age in organisms estimated by annu-

5 300

>-
cr

LlI

cc
o

200

100

-A

o
o

CTTL

o
o

o O

Y = -l54+05 X

1 1

0 08

o
o

[ct

X 006

oo yU

^ 004
Q

m 002

^ o

CD
O y

/ o

= -19 1

10"' +0.2

10"^ X

lar increments in otoliths to be from 2 to 15 years
old (Fig. 3A) (Y = -2.5 + 1.0 X, r'^ = 0.75; 6^0,
P < 0.01, t - 5.5, df = 10). On the other hand, in
older Dover sole (estimated age from 20 to 45
years; n = 26) the total extracted lipofuscin in the
brain did not increase with their estimated age.
The concentration of lipofuscin in the brain (ex-
tracted lipofuscin concentration per unit of g dry
weight) followed a similar pattern (Fig. 3B). The
concentration in the brain increased linearly with
age for fish 2-15 years old (7 = 30.39 + 14.85 X,
r2 = 0.79, b i=0, P < 0.01, t = 6.66, df = 10) but
fishes older than 15 years did not show an in-
crease in pigment concentration with age.

The total extracted lipofuscin content extracted
ft-om rainbow trout brains increased with age
(Fig. 4A) (Y = 0.14 + 0.18 X, r^ = 0.37; 6^0,
P < 0.01, t = 3.75, df = 23), as did the content of
the heart (Fig. 4C) (Y = -0.54 + 0.82 X,
r2 = 0.67; 6 ^ 0, P < 0.01, t = 7.36, df = 23), and
liver (Fig. 4E) (Y = -2.86 + 5.81 X, r^ - 0.62;
b i=0,P < 0.01, t = 7.54, df = 34) from 3 months
to 3 years. The concentration of extracted lipo-
fuscin per unit wet weight of heart tissue did
not change with age (Fig. 4D) whereas that of
brain and liver tissue decreased with age (Fig.
4B, F).

CC
UJ
CL

UJ
O

111
o

cc
o

200 400

LENGTH (mm)

600

oo o

Figure 2. — Dover sole, Microstomas paciftcus, brain:
(A) Extracted lijwfuscin per unit of dry weight as a function of
fish length (mm) (n = 36); (B) brain dry weight (g) as a func-
tion of fish length (mm). Fluorescence units: the intensity of
fluorescence normalized to a standard solution of quinine sul-
fate (1 mg L-i in IN sulphuric acid).

Figure 3. — Lipofuscin in Dover sole. Microstomas
pacificas, brain tissue as a function of age in = 36).
(A) Total lipofuscin content per brain; (B) weight-
specific (g dry weight) lipofuscin in brain.

I
a

>-
cc

I-

cc
o

300

200

100

o B

o o°

Yn.,<s = 30 4 +I4 85X

-J I I I L

10 20 30

AGE (years)

404

<
o
cc
o

o
CO

on
o

(

: J

}

1 1

AGE (yeors)

Figure 4. — Lipofuscin in rainbow trout, Salmo gairdneri, as a function of age (n = 21). (A) Total lijxjfuscm in
brain; (B) weight-specific (mg wet weight) lipofuscin in brain; (C) total lipofuscin in heart; (D) weight-specific
(mg wet weight) lipofuscin in heart; (E) total lipofuscin in liver; (F) weight-specific (mg wet weight) lipofuscin in
liver.

Discussion

Our results indicate that lipofuscin accumu-
lated in the brain of Dover sole with time (Fig.
2A). The concentration of lipofuscin increased
over a wide range of fish lengths and with esti-
mated age of 15 years but did not increase with
older fish. Several explanations exist for the lack
of change in concentration in older fish: 1) older
fish were incorrectly aged; 2) growth of brain
tissue masked the actual rate of accumulation;
and 3) the rate of accumulation changed during
the lifespan of this species due to changes in
metabolic activity. Either one or several of these

factors may cause the lack of lipofuscin accumula-
tion in older fish. We consider each of these issues
below.

We do not know the accuracy of the age deter-
mination in Dover sole but believe it is unlikely
that 3 readers would confuse fish aged 30-40
years with those of 20 years. Although future re-
search will shed more light on this controversial
subject, we think that grossly inaccurate age de-
termination is the least likely an explanation.

A key difference between fishes and other or-
ganisms in which lipofuscin accumulation has
been clearly documented as a function of chrono-
logical age (mammals and invertebrates) is that

405

fishes have indeterminate growth. For example,
Figure 3B indicates that the brain of Dover sole
continues to grow long after this fish begins re-
production (450 mm). Thus, unlike mammals and
invertebrates which have determinate growth,
the total content of extractable lipofuscin in any
fish organ cannot be used as a measure of age but
rather concentration (content per unit of weight)
must be used. Furthermore, even the concentra-
tion of lipofuscin in the brain, a slow growing
tissue, is a function of both the rate of lipofuscin
accumulation and the rate of tissue growth. It
follows then, that if the rate of brain growth is
such that it masks the actual rate of lipofuscin
accumulation in older organisms, lipofuscin con-
centration will not increase with time.

A critical assumption underlying the use of
lipofuscin as a determinant of age is that
metabolic rate, and hence lipofuscin accumula-
tion rate, remains fairly constant over the portion
of the life history of interest. The rainbow trout
were reared in a hatchery where no net change in
the environment occurred from 1983 to 1985.
Dover sole, on the other hand, gradually migrate
into deeper, colder, and less oxygenated waters as
they age. Presumably the metabolism of fish
under these conditions would be lower as might
be the rate of lipofuscin accumulation in the
brain.

The literature on fishes provide no conclusive
evidence that lipofuscin is an accurate index of
age over the entire lifespan. Hill and Radtke (in
press) reported that the extracted lipofuscin per
unit dry weight in the brain of the tropical fish
Dascillus albisella accumulates exponentially
with age. The relationship is driven by a single
point for an 11-year-old individual and it would
probably be linear without that single point, if
only fish 1-7 years old were considered. In
hatchery-reared Cyprinus carpio the total ex-
tracted lipofuscin per unit of dry weight in the
brain of fish of the same age (6 years) increased
with weight over 8-fold range in weight (Griven
et al.^). About 45% of the lipofuscin content could
be explained by difference in weight among fish.
Thus, in fish of the same age there was a strong
size effect on lipofuscin concentration. These re-
sults are very similar to those found in this study.
Aloj Totaro et al. (1985) found that lipofuscin in-
creased over a range of 0-2 years. But the method

2Girven, R. J,, R. W. Gauldie, Z. Czochanska, A. D. Wool-
house. Manuscr. in prep. A critical tests of the lipwfuscin
technique of age estimation in fish.

was different as they measured lipofuscin gran-
ules present in the electric lobe of Torpedo mar-
morata brains rather than using extractable lipo-
fuscin of the entire brain. Thus, all studies to date
seem to indicate that accumulation occurs in the
brains of fishes, but results are not conclusive
owing to small sample sizes, limited age ranges,
and failure to identify the effects of brain growth
on rates of accumulation.

Additional research is required to evaluate
lipofuscin as a method of age determination in
fishes. The effect of brain growth on lipofuscin
accumulation rates must be considered in such
studies. A promising approach in this regard may
be to estimate lipofuscin on a per cell basis in-
stead of on a weight basis. This could be accom-
plished by either expressing extracted lipofuscin
relative to DNA concentration or by histological
techniques.

Acknowledgments

This work was funded by Sea Grant contract
RyNP-l-15C and by NOAA contract 43ABNF6
1987. The authors would like to thank A. Dizon,
M. M. Mullin, E. Brooks, and J. Butler for helpful
discussions; J. Butler, E. Lynn, and M. Draw-
bridge for age determination by otoliths of Dover
sole specimens; and M. Rowan of California Fish
and Game for providing the rainbow trout sam-
ples.

Literature Cited

Agius, C and S. a. Agbede.

1984. An electron microscopical study on the genesis of
lipofuscin, melanin, and haemosiderin in the haemopoi-
etic tissues offish. J. Fish Biol. 24:471-488.

Aloj Totaro. E., F. A Pisanti, P Russo, and P Brunetti.

1985. Evaluation of aging parameters in Torpedo mar-
morata. Ann. See. R. Zool. Belg. 115(2):203-209.

BuGH, E G . AND W J Dyer.

1959. A rapid method of total lipid extraction and purifi-
cation. Can. J. Biochem. Physiol. 37:911-917.
BOEHLERT, G. W.

1985. Using objective criteria and multiple regression
models for age determination in fishes. Fish. Bull, U.S.
83:103-107.
Dapson, R W , A T. Feldman. and G Pane.

1980. Differential rates of ageing in natural populations
of old-field mice (Peromyscus polionotus). J. Gerontol.
35:39-44.
Ettershank. G.

1984. A new approach to the assessment of longevity in
the Antarctic krill Euphausia superba. J. Crustacean
Biol. 4:295-305.
Ettershank, G , I MacDonnell, and R Croft.

1983. The accumulation of the age pigment by the fleshfiy

406

Sarcophaga bullatta Parker (Diptera: Sarcophagidae).
Aust. J. Zool. 31:131-138.
Fletcher. B. L . C J Dillard. and A L Tappel,

1973. Measurement of fluorescent lipid peroxidation
products in biological systems and tissues. Anal.
Biochem. 52:1-9.
Gulland, J. A.

1978. Analysis of data and development of models. In
J. A. Gulland (editor), Fish population dynamics, Ch.4.
Wiley.
Hill, K. T., and R L Radtke.

In press. Gerontological studies of the damselfish, Dascyl-
lus albisella. Bull. Mar. Sci.
MacArthur, M. C, and R. S Sohal.

1982. Relationship between metabolic rate, aging, lipid
peroxidation, and fluorescent age pigment in milkweed
bug, Oncopeltus fasciatus (Hemiptera). J. Gerontol.
37:268-274.
MlQUEL, J., P. LUNDGREN, AND J E. JOHNSON, JR.

1978. Spectrofluorometric and electron microscopic study
of lipofuscin accumulation in the testis of ageing mice.
J. Gerontol. 33:5-19.
MuLLiN, M. M., AND E Brooks.

1988. Extractable lipofuscin in larval marine fish. Fish.
Bull., U.S. 86:407-415.
Munnell, J. F., AND R. Getty.

1968. Rate of accumulation of cardiac lipofuscin in the
ageing canine. J. Gerontol. 23:154-158.
NiCOL, S.

1987. Some limitations on the use of the lipofuscin ageing
technique. Mar. Biol. (Berl.) 93:609-614.
Reichel, W.

1968. Lipofuscin pigment accumulation and distribution
in five rat organs as a function of age. J. Gerontol.
23:145-153.
Sheldahl, J A , AND A. L. Tappel.

1973. Fluorescent products from aging Drosophila
melanogaster: an indicator of free radical lipid peroxida-
tion damage. Exp. Gerontol. 9:33-41.

Smith. P J

1987. Homozygous excess in sand flounder, Rhombosolea
plebeia, produced by assortive mating. Mar. Biol.
(Berl.) 95:489-492.
Sohal, R. S., and H. Donato.

1978. Effects of experimentally altered life spans on the
accumulation of fluorescent age pigment in the housefly,
Musca domestica. Exp. Gerontol. 13:335-341.
Strehler. B L , D D Mark, A S Mildvan, and M V Gee.

1959. Rate and magnitude of age pigment accumulation
in the human myocardium. J. Gerontol. 14:430-
439.
Tappel, A L.

1975. Lipid peroxidation and fluorescent molecular dam-
age to membranes. In B. F. Trump and A. V. Arstila
(editors), Pathobiology of cell membranes. Vol. 1. Acad.
Press, N.Y.
Zar.J H

1974. Biostatistical analyses. Prentice-Hall, Englewood
Clifis, NJ, 620 p.

Maria Vernet

John R. Hunter
Southwest Fisheries Center La Jolla Laboratory
National Marine Fisheries Service, NOAA
P.O. Box 271
La Jolla, CA 92037

Russell D. Vetter

Marine Biology Research Division
Scripps Institution of Oceanography
University of California San Diego
La Jolla, CA 92093-0202

Institute of Marine Resources
Scripps Institution of Oceanography
University of California San Diego
La Jolla, CA 92093-0202

EXTRACTABLE LIPOFUSCIN IN LARVAL
MARINE FISH

The "age pigment", lipofuscin, is thought to be a
biochemically heterogeneous byproduct of the
peroxidation of polyunsaturated lipids which ac-
cumulates in dense, intracellular inclusions
called ceroid bodies. The material has been stud-
ied both microscopically and biochemically in tis-
sues of several species (Miquel et al. 1977; Shi-
masaki et al. 1980; Dowson 1982), and some
portion of it is quantitatively extractable with
organic solvents (Fletcher et al. 1973).

Flies prevented from flying by putting baffles
in the bottles in which they were raised, accumu-
lated lipofuscin (as assessed by solvent extrac-
tion) more slowly than did free-flying flies but
had a longer lifespan, so that at the ends of the
respective lifespans the body contents of lipofus-
cin were similar in the two groups (Sohal and
Donato 1978). Extractable lipofuscin thus ap-
pears to accumulate as a function of cumulative
oxidative metabolism; it could be an indicator of
physiological (rather than strictly chronological)
age.

Additionally, if lipofuscin represents an inte-
gral of oxidative metabolism since birth and
weight represents an integral of growth over the
same period, the ratio of lipofuscin to organic
weight should be proportional to the reciprocal of
cumulative net growth efficiency [K2 = growth/
assimilation = growth/( growth + respiration),
hence I/K2 = 1 + respiration/growth].

Ettershank (1984a) introduced the fluoromet-
ric measurement of extractable lipofuscin as a
measure of physiological age in growing marine
crustaceans, based on the work with insects, and
(1984b) recommended a simple method for rou-
tine use in marine work. He also argued (without
presenting extensive evidence) that preservation

FISHERY BULLETIN: VOL. 86, NO. 2, 1988.

407

of tissue in formalin-seawater did not invalidate
the analysis; thus, historical samples of zooplank-
ton (and, presumably, larval fish) appeared to be
usable for study of intraspecific geographic or in-
terannual variability.

Although most studies of lipofuscin (especially
by histology) have concerned postmitotic cells,
such as nervous tissue, we evaluated the assay as
an estimate of the relative efficiency of growth of
larvae analyzed whole. That is, we were less in-
terested in lipofuscin as an indicator of age (for
which, as we show, other measures of mass are
useful) than as an indicator of health whose rela-
tion to mass would reflect environmental condi-
tions over time. We therefore investigated the im-
portance of interfering fluorescing pigments
(Csallany and Ayaz 1976), tested the effect of
preservation in formalin (Nicol 1987), and mea-
sured the accumulation of extractable lipofuscin
in three species of larval fish reared in the labora-
tory— California grunion, Leuresthes tenuis;
white seabass, Atractoscion nobilis; and Califor-
nia halibut, Paralichthys californicus (hereafter
referred as grunion, seabass, and halibut respec-
tively).

Analytical Considerations

We analyzed lipofuscin by a method first de-
scribed by Fletcher et al. (1973), as modified by
Ettershank (1984b). The tissue to be analyzed
(usually whole larvae) was frozen (-15° or
-70°C) and later freeze-dried, and a 1-5 mg sam-
ple was homogenized in at least 2 mL of 2:1
chloroform: methanol in a Wheaton glass tissue
homogenizer. After extracting for 3-4 hours at
4°C, 100 mM MgCl2 (25% of the solvent volume)
was added, and the solutions were thoroughly
mixed and then centrifuged for 20 minutes at
3000 rpm at -4°C. The lower, chloroform layer
was withdrawn for fluorometric analysis after
reaching 20°C in a water bath, and the fluores-
cence was measured on a Turner 111^ fluorometer
using a CS 7-60 filter (approximately 360 nm) for
excitation and a CS 47B filter (approximately 430
nm) for emission. A known concentration of
quinine sulfate was the standard, and results are
therefore reported as "fluorescence units" or FU/
mg.

A stock solution of quinine sulfate (2 mg/L in 1
N H2SO4) was stored in a light-tight reagent bot-

tle. To compare analyses done at different times,
sets of standards were prepared from this stock in
distilled water (0.02, 0.04, 0.06, 0.08, and 0.10
|jLg/mL). Some of the sets prepared over a year's
time are shown in Figure 1; the overall reproduci-
bility is good.

60 1-

UJ 30

tr
o

iReference to trade names does not imply endorsement by the
National Marine Fisheries Service, NCAA

0 .03 .06 .09 .12

QUININE SULPHATE (/ig-mr')

Figure 1. — Reproducibility of different sets of quinine sulfate
standards prepared from a stock solution over a year.

Precision of the method, including extraction
from tissue, was estimated by comparing repli-
cate samples of liver and white muscle from adult
halibut. The coefficients of variation for 5 deter-
minations on each of 3 sets of tissue were for mus-
cle, 0.2, 0.35, and 0.57, and for liver, 0.1, 0.15, and
0.18. The difference in variability between the
two kinds of tissues is probably due to the greater
difficulty in homogenizing muscle tissue.

Quinine sulfate standards were compared on a
Farrand Spectrofluorometer, with excitation at
350 nm and emission at 420 nm, and on the
Turner 111 fluorometer. The correlation coeffi-
cient for measurements on the two instruments
was 0.97. We therefore used the Turner 111 rou-
tinely, so that analyses could be done easily in a
hood.

Csallany and Ayaz (1976) described interfer-
ence by retinol in the analysis of organic-solvent-
soluble lipofuscin in mammalian tissues, and rec-
ommended a chromatographic step to remove this
contaminant. We extracted a variety of fish tis-
sues in 2:1 chloroform: methanol, and after addi-
tion of MgCl2 and centrifugation, dried the chlo-

408

roform layer gently (35°C) in an open vial
overnight in a hood on a sand bath. The dried
extract was reconstituted with 1:9 chloro-
form: methanol and chromatographed on a 30 cm
Sephadex LH20 column at a flow rate of 3-4 mL/
hour. A fraction was collected by hand every 20
minutes, its volume measured, and its fluores-
cence determined on the Turner 111 at the same
wavelengths used for routine analysis of lipofus-
cin. To determine where a retinol peak would ap-
pear, we chromatographed a tissue extract to
which a commercial preparation (Sigma) of
retinol had been added. Also, extracts of larval
fish in the pre-eyed stage were compared with
extracts of larvae that had pigmented eyes, as-

suming that the eyed larvae would have more
retinol.

A UV irradiation step had been proposed to
degrade retinol where it may interfere with mea-
surement of extracted lipofuscin, but Csallany
and Ayaz (1976) reported that this procedure was
ineffective. We therefore determined the time
course of degradation of commercial retinol in 1:9
chloroform: methanol by UV irradiation in quartz
tubes.

At the wavelengths we used, retinol was not an
important interfering substance in a variety of
larval fish and in the adult grunion muscle and
adult halibut liver tissues (Fig. 2). There also was
little difference between extracts of eyed and pre-

FlGURE 2. — Chromatograms (fluorescence vs. fraction number)
for various fish, compared with fish tissue plus retinol.

'50- HALIBUT LIVER

100

LARVAL GRUNION (whole)

LARVAL SEABASS (whole)

20 5

FRACTION (ml)

409

eyed larvae. We therefore did not use UV irradia-
tion routinely. If UV irradiation is used to de-
grade retinol, the exposure must be kept short (1
or 2 minutes), or else another compound fluoresc-
ing at these wavelengths appears (Fig. 3), giving
the spurious impression that UV did not affect
retinol; this may explain the negative results of
Csallany and Ayaz (1976).

To evaluate the effect of preservation in forma-
lin, we froze subsamples from stocks of larvae,
and preserved other subsamples in 10% formalin
in a glass container, using formalin from a glass
reagent bottle. We analyzed frozen and preserved

80,-

4 6

MINUTES

Figure 3. — Time course of fluorescence resulting from UV
irradiation of duplicate batches (two symbols) of retinol.

larvae after 3 months, and after an additional
year we analyzed more preserved larvae.

Nicol (1987) recently demonstrated that forma-
lin preservation significantly increases the fluo-
rescence of lipofuscin extracted in chloroform-
methanol, and our results show similar analytical
problems (Table 1). There was significant en-
hancement in fluorescence in the larvae pre-
served for over a year, when compared with those
preserved for 3 months. Quinine sulfate stan-
dards run at the same time as the samples for
different periods were very similar (Fig. 4), indi-
cating real changes with time of preservation.
Thus, this method will not permit using historical
collections of formalin-preserved animals to de-
termine their relative physiological states until
more is known about the time course and nature
of the effect of formalin on the extract of pre-
served tissue.

Larval Growth Experiments

Seabass larvae were obtained from Hubbs Re-
search Institute and halibut larvae from the Los
Angeles County Natural History Museum; larval
grunion were obtained by stripping adults and
bringing fertilized eggs into the laboratory for
hatching in an aerated, 10 L container of sea-
water at room temperature.

Larval grunion were reared in 40 L Nalgene
tubs with spigots at the bottom (which facilitated
emptying and cleaning), initially stocked with
100—120 newly hatched grunion per tub, with
replicated "high" and "low" food concentrations.
Larvae were fed newly hatched Artemia nauplii
and the rotifer, Brachionus , which was cultured
on Dunaliella tertiolecta in 100 L, lighted

Table 1 . — Lipofuscin (as fluorescence units, FU) per animal and per unit dry weight (DW) of frozen vs. formalin-
preserved larval fish. N = number of larvae per analysis.

Frozen (3 months)

Preserved (3 months)

Preserved (1 year)

FU (larva) -1

(mg DW)-i

FU (larva) -1

FU (mg DW)-1

FU (lar^a)i

(mg DW)-'

Grunion

16.0

47.4

24.3

60.1

44.3

98.1

A/ = 15

16.7

47.3

31.0

76.7

45.0

93.1

17.3

47.7

23.6

60.4

48.0

100.1

17.0

49.7

24.0
26.0

63.0
65.3

54.3
51.0

112.7
106.2

X =

16.8

48.0

25.7

65.3

48.5

102.0

Seabass

0.68

18.6

0.82

13.6

1.72

22.3

A/ = 30

0.47

13.0

0.71

12.0

2.05

26.3

0.50

14.5

0.46

8.0

2.03

27.2

0.43

12.7

0.52

8.7

1.63

22.2

0.33

9.4

0.60

10.5

2.05

28.9

X =

0.48

13.6

0.62

10.6

1.89

25.4

410

CO
UJ

en
o

0 .02 .04 .06 .08 .10
QUININE SULPHATE {/xg -mr')

Figure 4. — Fluorescence of various concentrations of
quinine sulfate standards read at 10 x gain (lower pair
of lines) and 30x gain (upper pair) in 1985 (lower line of
each pair) and 1986 (upper of each pair).

polyethylene tubs augmented with f/2 phyto-
plankton nutrients (May 1971; Theilacker and
McMaster 1971). One spring and one summer ex-
periment were completed at ambient tempera-
tures (spring = 17.5°-19°C, summer = 21°-23°C),
with one of the "high food" containers in the sec-
ond experiment kept at 26°C with an aquarium
immersion heater.

We were unable to maintain absolute high and
low food concentrations because of oscillations in
the supply of food organisms. Thus, the high con-
centration was kept at 3 times the low concentra-
tion, although absolute amounts varied. The
mean initial food concentrations in the spring ex-
periment were 50 jxg C/L (= "low") and 154 (xg
C/L (= "high"); in summer, 122 and 394 ^JLg C/L,
respectively. Tubs were censused every 2 days
through each experiment to determine how much
food was uneaten and how many larval fish had
died (estimated by counting and removing
corpses), and to add fresh food. In the low food
containers, it was not unusual to find little un-
eaten food, particularly as the larvae grew. Once
a week the tubs were emptied and the remaining
larvae counted directly. It was clear from this
direct census that all dead larvae were not ac-

counted for by searching for corpses every 2 days
because of cannibalism, necrophagy, or decay.
Every 4 days a known number of larvae was re-
moved; their lengths were measured and they
were frozen for future analyses.

We could not estimate larval ingestion pre-
cisely because of the uncertainty in how many
fish were alive through a 2-d interval. This prob-
lem was exacerbated as the larvae within each
tub diverged in size, so that variance in individ-
ual ingestion increased. Although 3 times more
food was offered in the high food containers, these
larvae actually ingested about twice the amount
of food as did those in the low food containers
(Table 2). This difference was due to better sur-
vival in the high food containers, which affected
the ratio between available ration and number of
larvae. To compensate for this, we routinely har-
vested more animals from the high food contain-
ers than from the low food containers.

Freeze-dried animals or tissues were weighed
on a Cahn electrobalance. Protein was deter-
mined by a method of Dorsey et al. (1977) on an
aliquot of tissue homogenized in cold 1 M NaCl.
DNA was measured by an ethidium bromide tech-
nique (Bentle et al. 1981, as modified by M. S.
Lowrey). Basic measures of size — dry weight,
protein, and DNA — were strongly and linearly
correlated (Fig. 5), so that comparing lipofuscin
with any of these measures would give similar
patterns.

Lipofuscin accumulated as the larval fish grew
(Fig. 6), but at quite different rates for the 3 spe-
cies, grunion accumulating most rapidly (relative
to gain in weight) and seabass much the slowest.

Table 2. — Estimated average ingestion, size, composition, and
growth efficiency for 20-day-old, laboratory-reared larval California
grunion. Compare with Figure 5.

Sphng

Summer

experiment

Low

High

Low

High

food

M-g C ingested'

per individual

701

1,723

1,379

2,792

p.g dry weight per

individual

755

1,305

901

1,895

^.g DNA per mg

protein

84.8

59.4

48.1

31.3

\xg C2 per

individual

245

603

482

977

Growth from

hatching (|j.g C)

104

462

341

836

Gross growth

efficiency

15%

27%

25%

30%

'Calculated from measured carbon In Brachionus and Anemia.
^Estimated from literature values.

411

o
>

cr>

100

LARVAL

GRUNION

•

. /Y=0.43X-0.06

1 1 1

LARVAL SEABASS

4 6 8 10 0 2 4

DRY WEIGHT (mg- larva"')

Figure 5. — Relations of protein (upper) and DNA (lower) to dry weight for laboratory-reared larval grunion (left)

and white seabass (right).

This may reflect differences in the Hfespans and
metabolic rates; grunion are small fish, reaching
maturity in a year and living less than 5 years
(Frey 1971), while seabass and halibut grow to
very much larger size and can live more than 20
years (Frey 1971; Thomas 1968).

Figure 7 shows the variability in the amount of
lipofuscin in larvae of the same age. At time "0",
all fish were newly hatched. The variance was
low in the first few days, but increased dramati-
cally with time because larvae within each tub
grew (and, presumably, respired) at very different
rates. There were differences in averages between
high food and low food conditions, and between
experiments (Table 2), but the variance in dry
weight, protein, DNA, or lipofuscin was so large
that the overlap obscured any differences be-
tween conditions of rearing.

Grunion and halibut larvae start life with
greater concentrations of lipofuscin than do sea-

bass, and though the concentration decreases
rapidly as grunion and halibut age (Fig. 8), they
still have almost a 10-fold greater concentration
than do larval seabass when all are 20 days old.
All three species increased in weight faster than
they increased in lipofuscin, so the concentration
of lipofuscin was "diluted" by growth. Because
protein was a constant fraction of dry weight
(Fig. 5, upper), this dilution was not due to skele-
tal growth alone. If the rate of weight-specific
growth exceeds the rate of weight-specific respi-
ration during early life, the concentration of lipo-
fuscin should decrease, as observed, and only
when growth ceases or slows considerably should
lipofuscin accumulate relative to weight. Alter-
natively, lipofuscin could change chemically with
time, becoming more difficult to extract (Vernet
et al. 1988), so that the rate of accumulation of
lipofuscin would be underestimated.
We attempted to stop growth by starving the

412

200

160

r^ 120 -

GRUNION

4 -

HALIBUT

Y=I4.5X +0.8

.15

.30

.45

.60

.75

6 8 10

DRY WEIGHT ( mg • larva"')

Figure 6. — Relation of lipofuscin fluorescence per larva to dry weight for larval grunion (upper left) white seabass (lower left), and

California halibut (right). Note different axes.

larvae to see if we could detect an increase in
lipofuscin. Table 3 shows results for seabass
starved for the final 20% of the rearing period;
there was no significant change in the concentra-
tion of lipofuscin. This is not what one would ex-

pect if accumulation of lipofuscin is proportional
to physiological age, unless no lipid is metabo-
lized during starvation. However, this result also
could reflect slow transformation of lipofuscin
from a more to a less soluble pool (Vernet et al.
1988).

Table 3. — Lipofuscin fluorescence (FU) per
unit dry weigfif of larval white seabass starved
for various periods after age 29 days.
N = number of analyses; ranges in parenthe-
ses.

FU (mg dry weight) -1 N

Initial

29 days old
Starved

2.4
(1.6-3.0)

2 days

3.1
(2.3-4.0)

4 days

1.3
(0.54-1.7)

6 days

1.8
(1.4-2.3)

8 days

2.5

Conclusions

Our intent was to evaluate the utility of mea-
suring extracted lipofuscin fluorometrically as an
indicator of the integrated metabolic health of
fish, especially preserved ones, and of relative net
efficiency of growth. We conclude that this tech-
nique is unlikely to be useful in these ways, at
least within the larval period. Although the accu-
mulation of total body burden of lipofuscin was
demonstrated, the variability among individuals
grown under the same conditions became so large
over time that we were unable to calibrate the
method in an ecologically meaningful sense. The
variability was evident in all measures of growth,

413

lUU

"GRUNION

•

120

•
•

•

• •

• •
•

• -1 • 1

•

•
•
•

•

• •

!:•

! 1 1

LU
O

LU
CJ
LD
Ld

a:
o

15 -

5 -

SEABASS

(5).

'OfHALIBUT

8 -

6 -

2 -

0 10 20 30

DAYS AFTER HATCHING

Figure 7. — Lipofuscin fluorescence per larva vs. age since
hatching for laboratory-raised larval grunion (top), white sea-
bass (center), and California halibut (bottom).

200

160

120

100

GRUNION

Y=-0.99X + 55

SEABASS

. Y=-0.48X + 13

HALIBUT

Y = -I.7X + 67

0 10 20 30

DAYS AFTER HATCHING

Figure 8. — Lipofuscin fluorescence per unit dry weight vs. age
since hatching for laboratory-reared larval grunion (top), white
seabsiss (center), and California halibut (bottom). Linear fits Eire
for comparative purposes only; relations for seabass and halibut
look curvilinear.

414

and is common in culture of larval fishes. Such
variability may be reduced in nature, where
runts may be subject to intense predation.

Lipofuscin analysis may be usefijl only when
applied to postmitotic tissue, such as nervous tis-
sue in mature fish, or to whole organisms whose
mitotic growth has essentially ceased, such as
adult copepods or insects (since the methods ap-
pear usefiil in arthropods — Ettershank et al.
1983; Sohal and Donato 1978). In these organ-
isms, the vagaries of growth are reduced, and the
accumulation of lipofuscin during starvation or
exercise might show that lipofuscin concentration
is interpretable as a measure of physiological age,
habitat quality, and net growth efficiency.

Lipofuscin is known to be a polytypic sub-
stance, probably variable in composition among
different organisms. We have assumed that a con-
stant proportion of the same substance is ex-
tracted. This may not be true (Vernet et al. 1988),
and considerable work remains to be done on the
basic biochemistry of the component substance(s).
Though the extraction and fluorometric measure-
ment is tantalizingly simple, it may well be that
the microscopical method used to quantify "ceroid
bodies" is the best approach. Fluorescent tech-
niques used in histochemical research (Brizzee
and Jirge 1981), combined with automatic imag-
ing procedures, might decrease the tedium of
staining and visual microscopy.

Acknowledgments

We thank Refik Orhun and Steve Cadell for the
larval white seabass and halibut, and Maria Ver-
net and two referees for helpful comments. Re-
search was supported by California Department
of Fish and Game project C-921 and National Sci-
ence Foundation grant OCE86-00742.

Literature Cited

Bentle, L a.. S Dutta, and J Metcoff.

1981. The sequential enzymatic determination of DNA
and RNA. Anal. Biochem. 116:5-16.
Brizzee, K R , and S K Jirge.

1981. Fluorescent microscope techniques for the visual-
ization and histological quantification of autofluorescent
lipofuscin bodies in brain tissues. In J. E. Johnson (edi-
tor). Current trends in morphological techniques, Vol.
III. CRC Press.

CSALLANY, A. S , AND K L AYAZ

1976. Quantitative determination of organic solvent solu-
ble lipofuscin pigments in tissue. Lipids 11:412-417.

DoRSEY, T E , P W McDonald, and O A Roels

1977. A heated Biuret-Folin protein assay which gives
equal absorbance with different proteins. Anal. Bio-

chem. 78:156-164.
DOWSON. J H

1982. The evaluation of autofluorescence emission spec-
tra derived from neuronal lipopigment. J. Microsc.
128:261-262.

Ettershank, G.

1984a. A new approach to the assessment of longevity in
the Antarctic krill Euphausia superba. J. Crust. Biol.
4(Spec. No. l):295-305.
1984b. Methodology for age determination of Antarctic
krill using the age pigment lipofuscin. Biomass Handb.
No. 26. SCAR/SCOR/LABO/ACMRR.
Ettershank, G., I MacDonnell, and R Croft.

1983. The accumulation of age pigment by the fleshfly
Sarcophaga bullata Parker (Diptera: Sarcophagiidae).
Aust. J. Zool. 31:131-138.

Fletcher, B L . C J Dillard. and A L Tappel.

1973. Measurement of fluorescent lipid peroxidation
products in biological systems and tissues. Anal. Bio-
chem. 52:1-9.
Frey. H W

1971. California's living marine resources and their uti-
lization. Calif. Dep. Fish Game, Resour. Agency, 148 p.
May, R C

1971. Effects of delayed initial feeding on larvae of the
grunion Leuresthes tenuis (Ayres). Fish. Bull., U.S.
69:411-425.
MiQUEL, J , J Org, K G. Bensch, and J E Johnson, Jr.

1977. Lipofuscin: fine-structural and biochemical studies.
In W. A. Pryor (editor). Free radicals in biology, p. 133—
182. Acad. Press, N.Y.

NICOL, S

1987. Some limitations on the use of the lipofuscin ageing
technique. Mar. Biol. (Berl.) 93:609-614.

Shimasaki, H., N Veta, and O. S. Privett

1980. Isolation and analysis of age-related fluorescent
substances in rat testes. Lipids 15:236-241.
Sohal, R. S , and H Donato.

1978. Effects of experimentally altered life spans on the
accumulation of fluorescent age pigment in the housefly,
Musca domestica. Exp. Gerontol. 13:335-341.

Theilacker, G H , AND M F. McMaster.

1971. Mass culture of the rotifer Brachionus plicatilis and
its evaluation as food for larval anchovies. Mar. Biol.
(Berl.) 10:183-188.
Thomas, J C

1968. Management of the white seabass (Cynoscion no-
bilis ) in California waters. Calif Dep. Fish Game, Fish.
Bull. 142, p. 1-34.
Vernet, M , J R Hunter, and R. D. Vetter

1988. Accumulation of age pigments in two cold-water
fishes. Fish. Bull., U.S. 86:401-407.

M M. Mullin
E. R. Brooks

Institute of Marine Resources
Scripps Institution of Oceanography
University of California, San Diego
La Jolla, CA 92093-0218

415

NOTICES

NOAA Technical Reports NMFS published during last 6 months of 1987

55. Proximate composition, energy, fatty acid, sodium,
and cholesterol content of fmfish, shellfish, and their
products. By Judith Krzynowek and Jenny Mur-
phy. July 1987, iii + 53 p., 2 tables.

56. Some aspects of the ecology of the leatherback turtle,
Dermochelys coriacea, at Laguna Jalova, Costa
Rica. By Harold F. Hirth and Larry H.
Ogren. July 1987, iii + 14 p., 12 tables, 13 figures.

57. Food habits and dietary variability of pelagic nekton
off Oregon and Washington, 1979-1984. By
Richard D. Brodeur, Harriet V. Lorz, and William G.
Pearcy. July 1987, iii + 32 p., 32 tables, 1 figure.

58. Stock assessment of the gulf menhaden, Brevoortia
patronus, fishery. By Douglas S. Vaughan.
September 1987, iii + 18 p., 19 figures, 14 tables.

59. Atlantic menhaden, Brevoortia tyrannus , purse seine
fishery, 1972-84, with a brief discussion of age and
size composition of the landings. By Joseph W.
Smith, William R. Nicholson, Douglas S. Vaughan,
Donnie L. Dudley, and Ethel A. Hall. September
1987, iii + 23 p., 3 figures, 12 tables, 14 appendix
tables.

60. Gulf menhaden, Brevoortia patronus, purse seine
fishery, 1974-85, with a brief discussion of age and
size composition of the landings. By Joseph W.
Smith, Eldon J. Levi, Douglas S. Vaughan, and Ethel
A. Hall. December 1987, iii + 8 p., 1 figure, 8 ta-
bles, 2 appendix tables.

61. Manual for starch gel electrophoresis: a method for
the detection of genetic variation. By Paul B.
Aebersold, Gary A. Winans, David J. Teel, George B.
Milner, and Fred M. Utter. December 1987, iii + 19
p., 8 figures, 1 table, appendices.

62. Fishery publication index, 1980-85; Technical Memo-
randum index, 1972-85. By Cynthia S. Martin,
Shelley E. Arenas, Jacki A. Guffey, and Joni M.
Packard. December 1987, iii + 149 p.

Some NOAA publications are available by purchaise from the Superinten-
dent of Documents, U.S. Government Printing Office, Washington, DC
20402.

416

ERRATA

Fishery Bulletin: Vol. 86, NO. 1

Butler, John L., and Darlene Pickett, "Age-specific vulnerability of Pacific sardine,
Sardinops sagax, larvae to predation by northern anchovy, Engraulis mor-
dax," p. 163-167.

Pages 165 and 166 have incorrect figures for the legends, please replace with the
following:

o
z
o
a.

t-
z

LLI

o
cc

LJ
Q.

—

..- X

ANC

HOVY ..'

- V.--

'■''

—

' J

t ,

- ■

uf'

-l

^^SARDINE

~ <

■' «

--^

1 1

1 1 1 1 1 1 1 1

8 12 16

LENGTH (mm)

100

O 80

z
a

O 60

ANCHOVY

v.-

SARDINE

8 12 16

LENGTH (mm)

Figure 3. — Increase by size of the percentage of Pacific sardine
larvae, Sardinops sagax, and northern anchovy larvae, En-
graulis mordax, escaping attack by adult northern anchovy and
95% confidence intervals. Data on anchovy larvae from
Folkvord and Hunter (1986).

100

z
o

lij

ANCHOVY If

SARDINE

20
AGE (days)

100

80 -

Q.
<

O 60
V)

40 -

ANCHOVY , -'-

..'

r' J

SARDINE

•••"f*Y -T-

, — t

, t-^

,1 F= 1 f^ 1 ] 1 1

20
AGE (days)

Figure 4. — Increase by age of the percentage of Pacific sardine
larvae, Sardinops sagax, and northern anchovy larvae, En-
graulis mordax , escaping attack by adult northern anchovy and
95% confidence intervals. Size categories of reared larvae have
been converted to ages using growth rates estimated from the
field. Data on anchovy larvae from Folkvord and Hunter (1986).

303

j 2 .q

INFORMATION FOR CONTRIBUTORS TO THE FISHERY BULLETIN

Manuscripts submitted to the Fishery Bulletin will reach print faster if they conform to the following instructions.
These are not absolute requirements, of course, but desiderata.

CONTENT OF MANUSCRIPT

The title page should give only the title of the paper,
the author's name, his/her affiliation, and mailing ad-
dress, including ZIP code.

The abstract should not exceed one double-spaced
page.

Text footnotes should be typed separately from the
text.

the literature cited. The accuracy of the literature

Common abbreviations and symbols, such as mm,
m, g, L, mL, mg, °C (for Celsius), %, %c, and so forth,
should be used. Abbreviate units of measure only when
used with numerals. Periods are only rarely used with
abbreviations.

We prefer that measurements be given in metric
units; other equivalent units may be given in parenthe-
ses.

FORM OF THE MANUSCRIPT

Figures and tables, with their legends and headings,
should be self-explanatory, not requiring reference to
the text. Their placement should be indicated in the
right-hand margin of the manuscript.

Do not send original drawings to the Scientific Editor;
if they, rather than the photographic reductions, are
needed by the printer, the Scientific Publications Office
will request them.

Acknowledgments, if included, are placed at the
end of the text.

TITLE PAGE

ABSTRACT

TEXT

LITERATURE CITED

TEXT FOOTNOTES

APPENDIX

TABLES (Each table should be numbered with an

arable numeral and heading provided).
LIST OF FIGURES (Entire figure legends)
FIGURES (Each figure should be numbered with

an arable numeral; legends are desired)

ADDITIONAL INFORMATION

Send the ribbon copy and two duplicated or carbon
copies of the manuscript to:

Dr. Andrew E. Dizon, Scientific Editor

Fishery Bulletin

Southwest Fisheries Center La Jolla Laboratory

National Marine Fisheries Service, NOAA

P.O. Box 271

La Jolla, CA 92038

Fifty separates will be supplied to an author free of
charge and 50 supplied to his organization. No covers
will be supplied.

Contents — Continued

WATSON, CHERYL, ROBERT E. BOURKE, and RICHARD W. BRILL. A compre-
hensive theory on the etiology of burnt tuna 367

BROWN-PETERSON, NANCY, PETER THOMAS, and CONNIE R. ARNOLD.
Reproductive biology of the spotted seatrout, Cynoscion nebulosus, in South
Texas 373

Notes

CHEN, CHE-TSUNG, TZYH-CHANG LEU, and SHOOU-JENG JOUNG. Notes
on reproduction in the scalloped hammerhead, Sphyrna lewini, in northeastern
Taiwan waters 389

DEW, C. BRAXTON. Stomach contents of commercially caught Hudson River
striped bass, Morone saxatilis , 1973-1975 397

VERNET, MARIA, JOHN R. HUNTER, and RUSSELL D. VETTER. Accumula-
tion of age pigments (lipofuscin) in two cold-water fishes 401

MULLIN, M. M., and E. R. BROOKS. Extractable lipofuscin in larval marine
fish 407

Notices: NOAA Technical Reports published during the last 6 months of 1987 . . . 416

GPO 791-008

.<*<■': ">».

BuUetin

^^AT£S O^

Marine Biological Laboratory
LIBRARY

JAN 2 0 1989

Vol. 86, No. 3

Woods Hole, Mass.

July 1988

BARLOW, JAY. Harbor porpoise, Phocoena phocoena, abundance estimation for Califor-
nia, Oregon, and Washington: I. Ship surveys 417^

BARLOW, JAY, CHARLES W. OLIVER, TERRY D. JACKSON, and BARBARA L.
TAYLOR. Harbor porpoise, Phocoena phocoena, abundance estimation for California,
Oregon, and Washington: IL Aerial surveys 433-'

SIGLER, MICHAEL R, and JEFFREY T FUJIOKA. Evaluation of variability in sable-
fish, Anoplopoina fimbria, abundance indices in the Gulf of Alaska using the bootstrap
method 445

BOEHLERT, GEORGE W, and TAKASHI SASAKI. Pelagic biogeography of the armor-
head, Pseudopentaceros wheeleri, and recruitment to isolated seamounts in the North
Pacific Ocean 453

SAVOY, THOMAS F, and VICTOR A. CRECCO. The timing and significance of density-
dependent and density-independent mortality of American shad, Alosa sapidissima. . 467

HOSS, DONALD E., LINDA COSTON-CLEMENTS, DAVID S. PETERS, and PATRICK.
A TESTER. Metabolic responses of spot, LeiosUymus xanthurus, and Atlantic croaker,
Micropogonias undulatiis, larvae to cold temperatures encountered following recruitment
to estuaries 483

WATANABE, YOSHIRO, JOHN L. BUTLER, and TSUKASA MORI. Growth of Pacific
saury, CololaMs saira, in the northeastern Pacific Ocean 489

DAVIES, N. M., R. W GAULDIE, S. A. CRANE, and R. K THOMPSON. Otolith
ultrastructure of smooth oreo, Pseudocyttits maculatus, and black oreo, Allocyttus sp.,
species 499

BECKER, D. SCOTT. Relationship between sediment character and sex segregation in
English sole, Parophrys vetulus 517

JAMIESON, GLEN S., and ANTAN C. PHILLIPS. Occurrence of Cancer crab (C.
magister and C. oregonensis) megalopae off the west coast of Vancouver Island, British
Columbia 525

STONER, ALLAN W, and ROGER J. ZIMMERMAN. Food pathways associated with
penaeid shrimps in a mangrove-fringed estuary 543^

WILSON, ELIZABETH A., ERIC N. POWELL, and SAMMY M. RAY The effect of the
ectoparasitic pyramidellid snail, Boonea impressa, on the growth and health of oysters,
Crassostrea virginica, under field conditions 553

BROUSSE AU, DIANE J., and JENNY A BAGLIVO. Life tables for two field populations
of soft-shell clam, Mya arenaria, (Mollusca: Pelecypoda) from Long Island Sound . . . 567

(Contimied on back cover)

Seattle, Washington

U.S. DEPARTMENT OF COMMERCE
C. William Verity, Secretary

NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION
William E. Evans, Under Secretary for Oceans and Atmosphere

NATIONAL MARINE FISHERIES SERVICE
James W. Brennan, Assistant Administrator for Fisheries

Fishery Bulletin

The Fishery Bulletin carries original research reports and technical notes on investigations in fishery science^ engineering, and economics.
The Bulletin of the United States Fish Commission was begun in 1881; it became the Bulletin of the Bureau of Fisheries in 1904 and
the Fishery Bulletin of the Fish and Wildlife Service in 1941. Separates were issued as documents through volume 46; the last document
was Na 1103. Beginning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate appeared as a numbered bulletin.
A new system began in 1963 with volume 63 in which papers are bound together in a single issue of the bulletin instead of being issued
individually. Beginning with volume 70, number 1, January 1972, the Fishery Bulletin became a periodical, issued quarterly. In this form,
it is available by subscription from the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402. It is also
available free in limited ntunbers to libraries, research institutions, State and Federal agencies, and in exchange for other scientific
publications.

SCIENTIFIC EDITOR, Fishery Bulletin

Dr. Andrew E. Dizon

Southwest Fisheries Center La Jolla Laboratory

National Marine Fisheries Service, NOAA

P.O. Box 271

La Jolla, CA 92038

Editorial Committee

Dr. Jay Barlow

National Marine Fisheries Service

Dr. William H. Bayliff

Inter-American Tropical Tftma Commission

Dr. George W. Boehlert
National Marine Fisheries Service

Dr. Bruce B. CoUette

National Marine Fisheries Service

Dr. Robert C. Francis
University of Washington

Dr. James R. Kitchell
University of Wisconsin

Dr. William J. Richards
National Marine Fisheries Service

Dr. Tim D. Smith

National Marine Fisheries Service

Mary S. Fukuyama, Managing Editor

The Fishery BvMetin (ISSN 0090-0656) is published quarterly by the Scientific Publications Office, National Marine Fisheries Service,
NOAA, 7600 Sand Point Way NE, BIN C15700, Seattle, WA 98115. Second class postage is paid in Seattle; Wash., and additional offices.
POSTMASTER send address changes for subscriptions to Fishery Bulletin, Superintendent of Documents, U.S. Government Printing
Officci Washington, DC 20402.

Although the contents have not been copyrighted and may be reprinted entirely, reference to source is appreciated.

The Secretary of Commerce has determined that the publication of this periodical is necessary in the transaction of the public business
required by law of this Department Use of funds for printing of this periodical has been approved by the Director of the Office of Manage-
ment and Budget

Fishery Bulletin

CONTENTS

Vol. 86, No. 3 July 1988

BARLOW, JAY. Harbor porpoise, Phocoena phocoena, abundance estimation for Califor-
nia, Oregon, and Washington: I. Ship surveys 417

BARLOW, JAY, CHARLES W OLIVER, TERRY D. JACKSON, and BARBARA L.
TAYLOR. Harbor porpoise, Phocoena phocoena, abundance estimation for California,
Oregon, and Washington: IL Aerial surveys 433

SIGLER, MICHAEL P., and JEFFREY T FUJIOKA. Evaluation of variability in sable-
fish, Ancyplopoma fimbria, abundance indices in the Gulf of Alaska using the bootstrap
method 445

BOEHLERT, GEORGE W, and TAKASHI SASAKI. Pelagic biogeography of the armor-
head, Pseudopentaceros wheeleri, and recruitment to isolated seamounts in the North
Pacific Ocean 453

SAVOY, THOMAS F, and VICTOR A. CRECCO. The timing and significance of density-
dependent and density-independent mortality of American shad, Alosa sapidissima. . 467

HOSS, DONALD E., LINDA COSTON-CLEMENTS, DAVID S. PETERS, and PATRICIA
A. TESTER. Metabolic responses of spot, Leiostomus xanthurus, and Atlantic croaker,
Micr&pogonias undidatvs, larvae to cold temperatures encountered following recruitment
to estuaries 483

WATANABE, YOSHIRO, JOHN L. BUTLER, and TSUKASA MORI. Growth of Pacific
saury, Cololabis saira, in the northeastern Pacific Ocean 489

DAVIES, N. M., R. W GAULDIE, S. A. CRANE, and R. K. THOMPSON. Otolith
ultrastructure of smooth oreo, Pseudocyttits maculatus, and black oreo, Allocyttus sp.,
species 499

BECKER, D. SCOTT. Relationship between sediment character and sex segregation in
English sole, Parophrys vetulus 517

JAMIE SON, GLEN S., and ANTAN C. PHILLIPS. Occurrence of Cancer crab (C.
magister and C. oreganensis) megalopae off the west coast of Vancouver Island, British
Columbia 525

STONER, ALLAN W, and ROGER J. ZIMMERMAN. Food pathways associated with
penaeid shrimps in a mangrove-fringed estuary 543

WILSON, ELIZABETH A., ERIC N. POWELL, and SAMMY M. RAY. The effect of the
ectoparasitic pyramidellid snail, Boonea impressa, on the growth and health of oysters,
Crassostrea virginica, under field conditions 553

BROUSSE AU, DIANE J., and JENNY A. BAGLIVO Life tables for two field populations
of soft-shell clam, Mya arenaria, (MoUusca: Pelecypoda) from Long Island Sound . . . 567

{Contimied on next page)

Seattle, Washington
1988

For sale by the Superintendent of Documents, U.S. Government Printing Offi
DC 20402— Subscription price per year; $16.00 domestic and $20.00 foreign
issue; $9.00 domestic and $11.25 foreign.

Marine Biological Laboratory
LIBRARY

e, Washingti
'ost per si

t^m 2 0 1989

Woods Hole, Mass. j

Contents— ConiiriMed

FARLEY, C. AUSTIN, PETER H. WOLF, and RALPH A. ELSTON. A long-term study
of "microcell" disease in oysters witii a description of a new genus, Mikrocytos (g. n.),
and two new species, Mikrocytos mackini (sp. n.) and Mikroq^tos roughleyi (sp. n.) . 581

Notes

WILLIAMS, AUSTIN B. Cojoined twin adult shrimp (Decapoda: Penaeidae) 595

GOULD, EDITH, DIANE RUSANOWSKY, and DONNA A. LUEDKE. Note on
muscle glycogen as an indicator of spawning potential in the sea scallop, Placopecten
magellanixrus 597

JAMIESON, GLEN S, and ELLEN K PIKITCH. Vertical distribution and mass mortality
of prawns, Pandalvs platyceros, in Saanich Inlet, British Columbia 601

YANG, MEI-SUN. Morphological differences between two congeneric species of pleuro-
nectid flatfishes: Arrowtooth flounder, Atheresthes stomias, and Kamchatka flounder,
A. evermanni 608

McINTYRE, JOHN D, REGINALD R. REISENBICHLER, JOHN M. EMLEN, RICHARD
L. WILMOT, and JAMES E. FINN. Predation of Karluk River sockeye salmon by coho
salmon and char 611

The National Marine Fisheries Service (NMFS) does not approve, recommend or en-
dorse any proprietary product or proprietary material mentioned in this publication.
No reference shall be made to NMFS, or to this publication furnished by NMFS, in
any advertising or sales promotion which would indicate or imply that NMFS ap-
proves, recommends or endorses any proprietary product or proprietary material
mentioned herein, or which has as its purpose an intent to cause directly or indirect-
ly the advertised product to be used or purchased because of this NMFS publication.

HARBOR PORPOISE, PHOCOENA PHOCOENA,
ABUNDANCE ESTIMATION FOR CALIFORNIA, OREGON, AND

WASHINGTON: I. SHIP SURVEYS

Jay Barlowi

ABSTRACT

The density and total population size of harbor porpoise along the coasts of California, Oregon, and
Washington are estimated from ship surveys using line transect methods. Surveys were completed
between September 1984 and May 1986 using teams of 3-5 observers. Data include 852 porpoise groups
sighted during 6,590 km of transects. Sighting rates varied more due to effects of sea state than due
to the presence of rain, fog, or sun glare. Experiments using additional observers indicate that approx-
imately 22% of trackline groups were missed by a team of 5 observers. Harbor porpoise density is
calculated from transects along the 18 m isobath and is extrapolated to other depth zones based on a
model of porpoise abundance as a function of depth. Total population size is estimated as 45,713 (SE
= 7,865) animals.

Approximately 200-300 harbor porpoise are taken
annually in central California set net fisheries (Dia-
mond and Hanan^; Hanan, et al.^). Little is known
about porpoise abundance in this area. Dohl et al.*
estimated that 1,600-3,000 porpoise reside in cen-
tral and northern California based on their aerial
surveys of coastal cetaceans. However, because har-
bor porpoise are frequently missed in aerial surveys
(Kraus et al. 1983), this estimate is probably low.
More information is needed on abundance, distribu-
tion, and population structure to determine the
significance of harbor porpoise mortality in set
nets.

Beginning in 1984, the National Marine Fisheries
Service (NMFS) has conducted ship and aerial
surveys of harbor porpoise abundance in California,
Oregon, and Washington. This report presents
results from four ship surveys. Results of the aerial
surveys are presented by Barlow et al. (1988).

'Southwest Fisheries Center La JoUa Laboratory, National
Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA
92038.

^Diamond, S. L., and D. A. Hanan. 1986. An estimate of har-
bor porpoise mortality in California set net fisheries: April 1, 1983
through May 31, 1984. Adm. Rep. SWR-86-15, 40 p. Available
from National Marine Fisheries Service, Southwest Region, 300
S. Ferry Street, Terminal Island, CA 90731.

^Hanan, D. A., S. L. Diamond, and J. P. Scholl. 1986. An esti-
mate of harbor porpoise mortality in California net fisheries April
1, 1984 through March 31, 1985. Adm. Rep. SWR-86-16, 38 p.
Available from National Marine Fisheries Service, Southwest
Region, 300 S. Ferry Street, Terminal Island, CA 90731.

^Dohl, T. P., R. C. Guess, M. L. Duman, R. C. Helm. 1983.
Cetaceans of central and northern California, 1980-83: status,
abundance, and distribution. Report prepared for U.S. Minerals
Management Service, Contract #14-12-0001-29090, 284 p.

Porpoise density is estimated from survey data
using line transect models (Burnham et al. 1980).
Total abundance is estimated by extrapolating from
density observed along transect lines to the entire
area inhabited. Abundance in offshore regions is
based on a model of porpoise density as a function
of water depth. In addition to abundance estimation,
survey data are used to examine the effect of envi-
ronmental conditions on sighting efficiency and the
possibility of temporal changes in harbor porpoise
distribution.

METHODS

Ship Survey Methods

Surveys were conducted from two National
Oceanic and Atmospheric Administration (NOAA)
research vessels, the 52 m RV David Starr Jordan
(Surveys 1, 3, and 4) and the 54 m RV McArthur
(Survey 2)^. Both vessels were of similar design with
viewing stations located on top of the pilothouse
(viewing height was approximately 10 m above sea
surface). Transect lines followed as close as possi-
ble to the 18 m isobath (roughly 2-4 km from the
coast), although the actual depth along the transect
varied from approximately 15-45 m, depending on
the presence of local navigational hazards. The areas

Manuscript accepted May 1988.

FISHERY BULLETIN: VOL. 86, NO. 3, 1988.

^Cruise reports available from the Southwest Fisheries Center,
P.O. Box 271, La Jolla. CA 92038.

417

FISHERY BULLETIN: VOL. 86, NO. 3

surveyed are shown in Figure 1. Survey design
va^ried among the four trips:

Surveys 1 and 3 were in September 1984 and 1985,
respectively. Both were designed to survey harbor
porpoise density and abundance from Point Concep-
tion, CA to Cape Flattery, WA. An attempt was
made to survey the entire coastline on each of these
cruises, but several sections of the coast were missed

(Fig. 1) because of fog and heavy weather. Five
observation positions were used on these two
surveys.

Survey 2 was from 24 January to 9 February 1985
and was primarily designed to examine seasonal
changes in harbor porpoise distribution between
Point Conception, CA and Cape Flattery, WA. Data
from this cruise were not used for density or abun-
dance estimation.

1 \ \ \ 1 n 49°N

48°

- 47°

10 PORPOISE/KM

I I

SURVEY 1 SURVEY 2 SURVEY 3

SEPTEMBER JANUARY/FEBRUARY SEPTEMBER

1984 1985 1985

SURVEY 4

APRIL/MAY

1986

- 46°

- 45°

- 44°

41°

- 38°

37°

36°

- 35°

34°

- 33°

32°

123°

121°

119°

117°W

Figure 1.— Relative sightings per kilometer based on 30-minute latitudinal strata. Lines parallel to the coast indicate areas that were
surveyed. Histograms indicate relative numbers of harbor porpoise seen per kilometer of transect, with bars to the left indicating rela-
tive numbers in calm seas (Beaufort 0, 1, and 2) and bars to the right indicating relative number in rough seas (Beaufort 3, 4, and
5).

418

BARLOW: SHIP SURVEYS OF HARBOR PORPOISE

Survey ^ was from 24 April to 5 May 1986 and
was designed to investigate factors which affect har-
bor porpoise density estimation. The surveys con-
centrated on several areas of high porpoise density
in central California. The vessel's activities were
coordinated with a helicopter to gather information
on the avoidance of the ship by harbor porpoise. Ex-
periments were also conducted on survey 4 to deter-
mine whether an independent team of 3 observers
would sight any porpoise that were missed by the
primary team of 5 observers. Data from this survey
were not used for density or abundance estimation.

Typically, 8-10 observers were used on each
survey, with a rested observer starting every half
hour and rotating through 5 primary observation
positions at half-hour intervals. The 5 positions con-
sisted of port and starboard inboard observers, port
and starboard outboard observers, and a recorder
positioned amidship. The inboard observers
searched with 7 power (7 x ) binoculars from straight
ahead to 90° (survey 1) or to 45° (surveys 2, 3, and
4) on their respective sides of the vessel. On survey
1 the outboard observers searched with 25 x,
pedestal-mounted binoculars. Although sightings
could be made at great distances from the vessel
using the 25 x binoculars, these distant sightings
contributed little to the estimation of trackline den-
sity, and use of 25 x binoculars was discontinued.
On subsequent surveys, both the inboard and out-
board observers used 7 x binoculars. The outboard
observers searched from straight ahead to 90° on
their respective sides of the vessel. The recorder
searched in the immediate vicinity of the ship using
unaided eyes and (intermittently) 7x binoculars.

On survey 2, only 3 observation positions were
used from Point Conception to Point Sur, CA and
from Point Reyes, CA to Cape Flattery, WA. When
effort was reduced to 3 observers, the inside obser-
vation positions were eliminated.

On survey 4, a second team of 3 observers was
added to monitor the effectiveness of the principal
team. This monitor team searched using unaided
eyes and (intermittently) 7x binoculars from the
pilothouse deck (viewing height approximately 7 m
from sea surface). The principal team and the moni-
tor team did not communicate sighting information,
and independent records were kept.

Data were noted by the recorder on data coding
forms. Data on search effort included the beginning
and ending times and positions for continuous legs
of effort, the ship's heading and speed, personal
identification codes for the observers, sea surface
temperature, water depth, Beaufort sea state, sun

position relative to the ship, and codes indicating
the presence of rain or fog within 5 km. The ship
position was determined from a Loran navigational
system or by triangulation using coastal landmarks
and dead reckoning. Ship speed was recorded direct-
ly from the Omega/Loran system or was calculated
based on time and distance traveled between suc-
cessive position fixes along straight transect lines.
Water depth was measured using a 38 kHz acoustic
depth sounder.

Data for sightings consisted of the above ele-
ments, plus estimated group size, distance to shore,
an estimate of the angle between the trackline of
the ship and the group, and an estimate of the
distance from the ship to the group. Group size
refers to all the individuals associated with a sight-
ing event. In most cases, groups were closely asso-
ciated individuals that surfaced together (mean =
2.92, median = 2.0). In two cases (Point Arena and
Monterey Bay, CA), groups consisted of 50-80 loose-
ly associated individuals that were organized in
subgroups of 4-10). Group size was estimated and
recorded independently by each observer; the mean
of these estimates was used in subsequent analyses.
The angle from the trackline to the porpoise was
estimated visually with the aid of a pelorus mounted
in front of the observer stations, or, when 25 x
binoculars were used, from a calibrated collar on the
pedestal mount. On surveys 1 and 2, distances to
harbor porpoise were estimated visually using the
radar distance-to-shore as a reference, or, when 25 x
binoculars were used, distances were estimated
using calibrated reticles in the oculars. On surveys
3 and 4, distances were estimated using calibrated
reticles in the oculars of 7 x binoculars. Data were
also collected on the porpoises' direction of travel
relative to the ship.

The length of a transect was estimated as the
product of ship speed and elapsed time. To stratify
density estimates by sea state, rain, and fog, the
effort record was divided into segments during
which the sea state, rain, and fog codes did not
change.

In five areas, information was collected on varia-
tion in harbor porpoise density with water depth.
During survey 3, three sections of the coast were
surveyed intensively (Fort Bragg to Cape Vizcaino,
CA; Cape Blanco to Coquille Point, OR; and Cape
Lookout to Tillamook Head, OR), with transect lines
following the 18, 56, 92, and 185 m isobaths. On
survey 4, the 18 and 46 m isobaths were surveyed
in Monterey Bay, CA and in the vicinity of the Rus-
sian River, CA. These data formed the basis of a
model (below) to extrapolate porpoise density from

419

FISHERY BULLETIN: VOL. 86, NO. 3

the usual transect lines (along the 18 m isobath) to
deeper waters.

Helicopter Observations

During survey 4, a Hughes 500-D helicopter was
used to collect information on harbor porpoise
behavior in response to the survey ship. The heli-
copter flew approximately 10 km ahead of the
vessel, and 3 observers in the helicopter looked for
harbor porpoise. Once a group of harbor porpoise
was sighted, the helicopter hovered at 200-300 m
while observers made behavioral observations and
periodically recorded the helicopter's position using
an on-board Loran system. Fluorescein dye pack-
ages were dropped in the water to allow the heli-
copter to maintain its position when harbor porpoise
were diving. Radio communication was maintained
with personnel on the ship who also kept records
of the helicopter position using radar distances and
bearings based on returns from an X-band radar
transponder in the helicopter. The ship changed
course, when necessary, to ensure that it passed in
close proximity to the porpoise that were being
observed. Porpoise observers on the ship were not
aware of the helicopter's activities and were not told
of sightings made by the helicopter observers (al-
though they were able to see dye patches in some
cases). Behavioral observations from the helicopter
included time spent at the surface, time spent div-
ing, and direction of porpoise movement.

Density Estimation

Line transect methods were used to estimate the
density of harbor porpoise from sightings. The
assumptions of these methods are considered in
detail in the discussion. The usual formula for
estimating density (D) based on line transect
surveys of small cetaceans is given by

D =

m ■ n- G
2 ■ L

(1)

where /(O)

n
G

L =

the probability density function for
sightings evaluated at zero perpen-
dicular distance,
number of sightings of groups,
average group size calculated as the
total number of individuals in all
groups divided by the number of
groups (iN/n), and
length of the transect.

(Holt and Powers 1982; Hammond and Laake 1983;
Holt in press). I did not use mean group size explicit-
ly in abundance estimation, and density of harbor
porpoise individuals, D, was estimated as

D = /(O) ■ iRI2)

where R = the number of individuals
length of transect (XN/L).

(2)

seen per

Equation (2) is functionally equivalent to Equation
(1), but it simplifies variance estimation. Typically
when using Equation (1), variances (and possibly
covariances) must be estimated for/(0), G, and n.
Using Equation (2), variances are needed only for
/(O) and R, and covariance between mean group size
and number of groups is handled implicitly. Sight-
ing distributions appear to be independent of group
size, G, (Results section), hence no adjustments were
made to /(O) for group size bias.

The parameter /(O) is, in effect, a measure of sight-
ing efficiency and should not vary with porpoise
abundance. Sighting efficiency is, however, likely
to change with sighting conditions, such as Beau-
fort sea state. Given these expectations and because
relatively large sample sizes are needed to estimate
/(O) accurately, values for/(0) were estimated for
each survey by pooling all sightings within defined
sea state categories. In order to estimate density
on a finer scale, estimates of R were stratified by
geographic region and multiplied by the pooled esti-
mate of /(O).

The sighting probability density function evalu-
ated at zero distance, /(O), was determined
empirically by fitting curves to the frequency
distribution of sightings as a function of perpen-
dicular distance from the trackline (Burnham and
Anderson 1976). Differences in distributions of
perpendicular distance were tested using the
Kolmogorov-Smirnov 2-sample test. To avoid bias
due to rounding error, angle and radial distance data
were "smeared" (Butterworth 1982; Hammond and
Laake 1983). Angles were smeared by adding a
uniformly distributed random number between - 5°
and -1-5° to angle estimates. Radial distances were
smeared by adding a uniformly distributed random
number between 0.2 and -(-0.2 times the estimated
distance. These smearing levels were based on the
degree of rounding that was apparent from the data
(Barlow^).

^Barlow, J. 1987. Abundance estimation for harbor porpoise
(Phocoena phocoena) based on ship surveys along the coasts of
Cahfornia, Oregon, and Washington. Adm. Rep. LJ-87-05.

420

BARLOW: SHIP SURVEYS OF HARBOR PORPOISE

Several models were investigated for estimating
/(O) from sighting distributions. The FORTRAN pro-
gram Transect (Laake et al. 1979) was used to fit
2-, 3-, 4-, and 5-parameter Fourier series and
2-parameter exponential power series models. The
FORTRAN programs Hazard and Hermite (S. Buck-
land'') were used to fit the 2-parameter hazard rate
model (constrained such that parameter P > 2,
Buckland 1985) and the 1-, 2-, 3-, and 4-parameter
Hermite polynomial model (Buckland 1985). Of these
models, the 2-parameter hazard rate model was
selected based on its ability to fit the observed
distributions and its lack of dependence on group-
ing criteria (Buckland 1985).

Perpendicular distances were grouped into strata,
the size of which increased with perpendicular
distance: 0-25 m, 25-50 m, 50-100 m, 100-200 m,
200-400 m, 400-800 m, 800-1,600 m, and 1,600-
3,200 m. Several alternative groupings were inves-
tigated, and the choice of outpoints made very little
difference in estimates of /(O). The above strata
(increasing with distance) gave lower variances in
/(O) than when each stratum was of equal size
(possibly because the hazard rate model assumes a
distinct shoulder in the sighting distribution, and
that shoulder is lost if the first distance strata are
large).

No established criteria exist for choosing an appro-
priate perpendicular distance at which to truncate
sighting distributions. Burnham et al. (1980) recom-
mend that no more than 1-3% of sightings be
eliminated by truncation. Using this recommenda-
tion, models were not able to adequately fit the
observed sighting distributions. In this report, trun-
cation distance was chosen in four ad hoc steps:
1) The hazard rate model was fit to perpendicular
distance data truncated at distances of 400, 800,
1,600, and 3,200 m. 2) Truncation distances were
identified which gave acceptable x^ values (P >
0.1). 3) Of the acceptable truncation distances, the
standard error in/(0) was estimated empirically by
randomly drawing 10 samples (of the same size as
the original sample) from the observed distribution
of perpendicular distances and by calculating the
standard deviation of/(0) estimated from each ran-
dom sample. 4) Truncation distances were chosen
as those which gave the lowest coefficient of varia-
tion in/(0).

Variance in R, the number of porpoise seen per
kilometer, was estimated using jackknife statistics

(Efron 1982). Jackknife estimates were calculated
by first estimating the value of i? using all data. The
value, Rj^, was again estimated excluding the A;th
segment of search effort. This process was repeated
for each effort segment. To ensure that each kth seg-
ment was of equivalent length, effort segments with
the same sea state, rain, and fog codes were com-
bined in a linear array and were then divided into
10 segments of approximately equal length. The
variance in the estimate of R was calculated as

s2 =

~ • 1 (i?, - Rf

10 ^=1

(3)

Avail, from Southwest Fisheries Center, P.O. Box 271, La Joila,
CA 92038.

^S. Buckland, Inter-American Tropical Tuna Commission, P.O.
Box 271, La Jolla, CA 92038, pers. commun. July 1986.

The variance of D was estimated using the Good-
man (1960) product variance formula (assuming no
covariance) using this jackknife variance for R and
the above Monte Carlo variance for/(0).

Fraction of Missed Animals

On survey 4, a second, independent team of 3
observers were used to estimate the fraction of
harbor porpoise that are missed by the primary
team of 5 observers. The fraction of missed animals
in a sighting survey is analogous to the fraction of
unmarked animals in a mark/recapture experi-
ment (Pollock and Kendall 1987). This fraction
was estimated using the Chapman (1951) modifica-
tion of the Petersen (or Lincoln) index method
(Pollock and Kendall 1987). Confidence limits were
estimated using Adams' (1951) method, which
assumes a binomial sampling distribution. Standard
error was estimated using standard binomial
formulas.

Abundance Estimation

A model was used to estimate the number of har-
bor porpoise along the entire coastline based on the
density that was observed along the 18 m isobath.
In shallow areas, such as the Bering Sea and
Georges Bank, harbor porpoise are found a con-
siderable distance from land (Gaskin 1984), hence
offshore distribution is better modelled as a func-
tion of depth than as a function of distance from
shore. (Although harbor porpoise are also found in
very deep water in fjords and inland waterways of
Alaska [Taylor and Dawson 1984], this represents
a special case that is not applicable to coastal waters
considered here.) The model used to estimate abun-
dance was based on data collected on surveys 3 and
4 and on data from a ship surveys by La Barr and

421

FISHERY BULLETIN: VOL. 86, NO. 3

Ainley^ and Szczepaniak^ in central California. The
number of harbor porpoise seen per kilometer of
transect was taken as an index of relative density
along each isobath. A simple descriptive model was
then constructed to give relative density as a func-
tion of water depth.

Fifteen depth strata were used in abundance
estimation: 0-10, 10-20, 20-30, . . . , and 140-150 m.
The surface area within the strata was calculated
from digitized bathymetric data. Kelp beds were
assumed to be unsuitable as harbor porpoise habitat;
hence, kelp bed area was subtracted from the total
area within the 0-10 m stratum. Kelp bed areas for
the entire west coast were taken from Crandall
(1915). More recent estimates for limited areas in
central California are in good agreement with these
previous values (G. Van Blaricom^").

For each of 15 depth strata, the abundance of har-
bor porpoise was estimated as the product of their
density along the survey line (the 18 m isobath), the
density in that depth strata relative to that along
the survey line, the surface area included within that
depth strata, and the inverse of the estimated frac-
tion of trackline animals that were seen. Since
survey effort and harbor porpoise density both
varied geographically, abundance estimates were
made for each of 8 geographic regions (Fig. 2).
Areas within the depth strata were estimated from
NOAA bathymetric data. The estimate of total abun-
dance along the coast, A/'y, is therefore given by

^- = ^ I «^ .?. <^' ■ ^^.'>

(4)

where Dj = density of individuals observed on the

transect line in the jth geographic

strata,
4 = ratio of density in depth strata k to

that on transect line (see Figure 4),
Aji^ = area in geographic region j' and depth

strata k, and
F = the estimated fraction of trackline

animals seen by the usual team of 5

observers.

'LaBarr, M. S., and D. G. Ainley. 1985. Depth distribution of
harbor porpoise off central California: A report of cruises in April
and May-June 1985. Report to U.S. National Marine Fisheries
Service, Northwest and Alaska Fishery Center, 7600 Sand Point
Way N.E., Seattle, WA. Contract No. 41-USC252.

'Szczepaniak, I. D. 1987. Abundance and distribution of har-
bor porpoise (Phocoena phocoena) in the Gulf of the Farallones
National Marine Sanctuary. Contract report prepared for
National Park Service, Point Reyes National Seashore, Point
Reyes, CA 94956.

'"G. Van Blaricom, U.S. Fish and Wildlife Service, University
of California, Santa Cruz, CA 93106, pers. commun. August 1986.

Equation (4) was applied independently to the dif-
ferent surveys and, within surveys, to different sea
state strata. When combining estimates from differ-
ent sea states or different cruises, abundance was
calculated as the mean of the densities in each of
the stratum, weighted by the length of the transect
line within that stratum.

In estimating standard error for total abundance,
variances of products were calculated using the
Goodman (1960) product variance formula, and vari-
ances of ratios were estimated using a Taylor
approximation (Yates 1953, p. 198). Area was
assumed to be known without error. Statistical error
in the indices of abundance for the depth strata could
not be estimated given the paucity of available in-
formation. To account for uncertainty in the model
of depth distribution, three versions of the model
are proposed to span a range of possibilities.

RESULTS

On the four surveys, 852 groups of harbor por-
poise were sighted (an estimated 1,818 individuals).
A distance of 6,590 km was surveyed during 56
days. The number of sightings per kilometer sur-
veyed varied geographically and these geographic
patterns appeared to change appreciably between
cruises (Fig. 1).

Sighting Distributions

The number of sightings on the inshore and off-
shore sides of the vessels were approximately equiv-
alent (383 and 392, respectively). The cumulative
distributions of perpendicular sighting distances
were not significantly different for these two sides
(P = 0.06). Therefore, sighting distributions were
assumed to be symmetrically distributed about the
trackline, and the distributions of perpendicular
sighting distances from both sides of the vessel were
pooled for subsequent analyses.

The distributions of perpendicular sighting dis-
tances for the first three surveys were significant-
ly different from one another (P < 0.01 for all). This
was probably the result of the modifications in
survey methods between these cruises. Surveys 3
and 4 used the same methods, and sighting distribu-
tions were not significantly different (P = 0.39).
Given that changes in methods result in differences
in sighting distributions, all surveys were treated
separately in subsequent analyses.

Distributions of perpendicular distance were not
significantly different between individuals sighted

422

BARLOW: SHIP SURVEYS OF HARBOR PORPOISE

48°

46°

45»

44*

43"

420 .

41* -

40°

39°

38*

37°

36°

35°

34° -

33° -

32°N

132°W 131° 130° 129° 128" 127" 126" 125° 124° 123° 122° 121" 120° 119° 118° 117»W

Figure 2.— Geographic regions used as strata in abundance estimation. Broken line indicates the 200 m

isobath and delineates likely harbor porpoise habitat.

423

FISHERY BULLETIN: VOL. 86, NO. 3

alone (group size of 1) and larger groups (group size
of 3 and greater) (P = 0.56).

Environmental Conditions Affecting
Sightings

Sighting efficiency was not significantly affected
by rain, fog, or sun glare. Rain/fog conditions were
considered "poor" if rain or fog were present within
5 km of the vessel and "good" if neither were pres-
ent. The distributions of perpendicular sighting
distances were not significantly different between
these two strata (P = 0.32, 0.44, 0.78, and 0.64,
respectively, for surveys 1, 2, 3, and 4), and the
number of porpoise per kilometer surveyed was
higher in the "poor" category for two of the surveys.
Sun glare from the water's surface was considered
to contribute to "poor" sighting conditions if the sun
was within 45° of the trackline in front of the ship.
Conditions were considered "good" when the sun
was in other positions or was obscured by clouds.
As with rain/fog conditions, the distributions of
perpendicular sighting distances were not signifi-
cantly different between these "good" and "poor"
sun glare categories (P = 0.87, 0.47, 0.30, and 0.55,
respectively, for surveys 1, 2, 3, and 4). The number
of harbor porpoise per kilometer surveyed were
slightly higher in the poor category for three of the
surveys. In paired comparisons when glare was pres-
ent on only one side of the bow, approximately equal
numbers of sightings were made on the sides with
and without glare (60 vs. 59, respectively). All
categories of rain, fog, and glare are included in
subsequent analyses.

Sea state did have a significant effect on porpoise
sightings. Sea state was categorized as calm
(without white-caps, Beaufort sea states 0, 1, and
2) or rough (with white-caps, Beaufort sea states 3,
4, and 5) following the classification used by Holt
and Cologne (1987). Distributions of perpendicular
distances were not significantly different between
these categories for any of the surveys (P > 0.05);
however for all surveys combined, the number of
harbor porpoise detected per kilometer was much
lower during rough seas (0.32 km"^) than during
calm seas (1.22 km"^). There were insufficient
sightings to estimate density for rough seas separ-
ately; therefore, rough sea data were excluded in
subsequent analyses. For all three surveys, the
numbers of harbor porpoise detected per kilometer
was higher at Beaufort 0 & 1 than at Beaufort 2,
and for survey 3, the distributions of perpendicular
sighting distance were significantly different be-
tween these categories {P = 0.03). Porpoise density

is, therefore, estimated separately for Beaufort 0
& 1 and for Beaufort 2 conditions. (For comparison,
harbor porpoise abundance was also estimated pool-
ing Beaufort sea states 0, 1, and 2. Estimated abun-
dance was approximately the same by both methods,
but the variance was slightly lower using the
stratified sea state categories. For this reason, only
the stratified estimates are presented here.)

Helicopter Observations

Helicopter observation of the behavior of harbor
porpoise in response to the survey ship were made
on only 6 groups of animals. Plots of vessel tracks
and movements of the groups are given in the cruise
report (see footnote 5). Only in one case was a
distinct behavioral change noted in response to the
ship. In that case, when the vessel was within 800
m, the group moved rapidly, perpendicular to the
path of the vessel and then parallel to and in the op-
posite direction of the vessel. Observers on the ship
saw this harbor porpoise group as they moved rapid-
ly out of the path of the vessel. Observers on the
ship also saw 2 of the other 5 groups. Although this
sample of behavior is small, movement in response
to the survey vessel appeared limited to within 1 km
of the vessel and, when it occurred, animals did not
travel far from their original positions.

Porpoise Density

The probability density distributions of perpen-
dicular sighting distances are shown in Figure 3 for
surveys 1 and 3 and for Beaufort sea states 0 & 1
and 2. The hazard rate model gave acceptable fits
for all sighting distributions {P > 0.1) when the trun-
cation criteria was set at 400 m (Table 1). For survey
1, the optimum truncation points were chosen as 400
m for Beaufort 0 & 1 and 800 m for Beaufort 2; for
survey 3, this distance was 400 m for both Beaufort
sea state categories. The fits of these models are
shown in Figure 3. Estimates of density and stand-
ard errors are given in Table 2.

Depth Distribution Model

The model of harbor porpoise depth distribution
was based on the relative densities of harbor por-
poise at different water depths. Ship survey data
were pooled into five depth ranges: 18-37 m (10-20
fathoms), 37-55 m (20-30 fathoms), 55-73 m (30-40
fathoms), 73-91 m (40-50 fathoms), and 91-110 m
(50-60 fathoms). Ship surveys are generally not
practical inshore of the 18 m isobath, but estimates

424

BARLOW: SHIP SURVEYS OF HARBOR PORPOISE
10

0.0

0.1 0.2

PERPENDICULAR DISTANCE (km)

SURVEY 1, BEAUFORT 2

6 -

4 -

2 -

0 -

■

0.4 00
10

0.4 0.0

0.2

0.4

0.6

0.8

SURVEY 3, BEAUFORT 2

-I r

0.1

0.2

1 r-

0.3

0.4

PERPENDICULAR DISTANCE (km)

Figure 3.— Probability density distributions for perpendicular sighting distances. Histograms indicate observed distributions, and solid

lines indicate the best fit of the hazard rate model to these data.

Table 1 .—Estimated values of the probability density functions
evaluated at zero perpendicular distance, ^(0). Estimates are based
on the hazard rate model and were made for truncation distances
of 400, 800, 1,600, and 3,200 m. Estimates are given only if the
model gave an acceptable fit to the data (P > 0.1). Asterisks in-
dicate f(0) values with the lowest coefficient of variation (paren-
theses).

Survey

Beaufort
sea state

Truncation distance

400 m 800 m 1 ,600 m 3,200 m

1
1
3
3

0 & 1
2

7.85 •

(0.23)

10.48
(0.59)

4.51 •
(0.22)

6.97 •
(0.19)

5.31
(0.29)

8.15 *
(0.21)

3.10
(0.66)

4.31
(0.24)

7.09
(0.33)

2.69
(0.31)

5.78
(0.51)

from aerial surveys (Barlow et al. 1988) show
roughly equal density at 0.61 and 1.85 km from the
shore (the latter corresponding approximately to the
18 m isobath). Relative density from ship surveys
was measured in the number of sightings per kilo-
meter of searching effort. Relative densities at

Table 2.— Density estimates, D, for harbor porpoise (km"^) along
the 18 m isobath in each of eight geographic strata. Density was
calculated per Equation (2), using estimates of f(0) (Table 1) which
had the lowest coefficients of variation. Values are not adjusted
for missed animals. Standard errors are in parentheses.

Geo-
graphic
region

1
2
3
4
5
6
7
8
Total

Survey 1

Survey 3

Beaufort
0 & 1

Beaufort
2

Beaufort
0 & 1

Beaufort
2

0.0
(0.0)

1.1
(1.0)

1.4
(0.8)

7.9
(3.4)

1.8
(0.8)

2.5
(1.1)

0.0
(0.0)

2.5
(1.0)

2.1
(1.3)

2.0

(0.6)

0.8
(0.6)

0.0
(0.0)

1.3
(0.6)

0.0
(0.0)

0.6
(0.5)

0.1
(0.1)

0.0
(0.1)

0.8
(0.6)

2.5
(0.8)

2.5
(0.6)

0.8
(0.4)

0.1
(0.1)

0.0
(0.0)

6.7
(7.2)

0.6
(0.3)

2.6
(0.9)

1.3
(0.7)

1.2
(0.5)

Pooled
estimates

0.04
(0.02)

0.51
(0.30)

0.03
(0.03)

2.83
(1.69)

0.91
(0.32)

2.22
(0.40)

2.64
(0.78)

1.09
(0.45)

1.33
(0.30)

425

FISHERY BULLETIN: VOL. 86. NO. 3

18-37 m show no consistent relationship to those
at 37-55 m or 55-73 m (Table 3), but on average
these appear to be approximately equal. Relative
densities at 18-37 m are, however, consistently
higher than densities at 73-110 m in all areas
(Table 3). A total of 236 km were searched in waters
deeper than 110 m and no harbor porpoise were
seen.

Despite high variability in patterns of depth
distribution and lack of ship coverage in shallow
waters, some generalizations can be made about the
depth distribution of harbor porpoise along the west
coast. The relative abundance of harbor porpoise ap-
pears to be roughly constant from shore to 55 m,
to be markedly lower at 73-110 m, and to be very
low in waters deeper than 110 m.

Based on the above relationships, I propose the
following preliminary model for the depth distribu-
tion of harbor porpoise along the coasts of Califor-
nia, Oregon, and Washington: constant abundance
from the coast to the 80 m isobath, linearly decreas-
ing abundance from the 80-120 m isobaths, and zero
abundance in waters deeper than 120 m (Fig. 4a).
Because considerable uncertainty exists in this
model, I propose two alternative models (Fig. 4b,
c). Alternative models b and c are less likely than
the primary model given because both conflict with
some of the available data. The alternative models
do, however, encompass the likely range of relative
density values and provide a means to evaluate the

sensitivity of the abundance estimate to different
models of depth distribution.

Fraction of Missed Animals

The experiment on survey 4 indicates that some
trackline groups were seen by 1 group of observers
and were missed by the other. A total of 103 sight-
ings was made by both teams, 33 of which were
estimated to be within 100 m perpendicular distance
from the transect line. Of the 103 total sightings,
85 were detected only by the 5 principal observers,
6 were detected only by the 3 monitor observers,
and 12 were detected by both teams. Of the 33
trackline sightings, 20 were detected only by the
principal observers, 3 were detected only by the
monitor observers, and 10 were detected by both
teams. The Petersen estimate of the fraction of
trackline porpoise seen by the primary team of 5
observers is thus 0.780 (SE = 0.117, 95% C.L. =
0.45-0.95). This indicates that approximately 22%
of trackline sightings are missed by the principal
teams of 5 observers.

Porpoise Abundance

Estimates of porpoise abundance in each of the
eight geographic strata are given in Table 4 for the
primary model of offshore distribution. Independent
estimates are given for survey 1 and for survey 3

Table 3. — Relative harbor porpoise abundance observed within the specified depth
ranges at a variety of study sites. Relative abundance is measured as number
of porpoise sightings made per kilometer. Numbers in parentheses indicate
kilometers surveyed.

Depth range

Location

18-37 m

37-55 m

55-73 m

73-91 m

91-110 m

Central California'

0.02
(89)

0.08
(172)

0.03
(403)

0.03
(279)

0.01
(166)

Gulf of the
Farallones, CA^

0.29
(181)

0.00
(159)

0.30
(133)

0.00
(7)

Fort Bragg, CA

0.05
(41)

—

0.05
(43)

—

0.03
(35)

Coquille Pt, OR

0.24
(43)

—

0.35
(52)

—

0.08
(50)

Tillamook
Head, OR

0.50
(57)

—

0.12
(52)

—

0.00
(17)

Monterey Bay, CA

0.46
(220)

0.29
(76)

—

Russian River, CA

0.00
(26)

0.30
(33)

—

'Data taken from LaBarr and Ainley (see text footnote 8) assuming an average survey speed
of 9 knots.

^Data taken from Szczepaniak (see text footnote 9) assuming an average survey speed of
9 knots.

426

BARLOW: SHIP SURVEYS OF HARBOR PORPOISE

DEPTH DISTRIBUTION MODELS

2
18
16
14
12

1
0.8
0.6
0 4
0.2

2
1.8
16
1.4
1.2

1
0.8
0 6
0.4
0.2

Primary Model

5 15 25 35 45 55 65 75 85 95 105 115 125 135 145

LU
>

LU
CC

lU
Q

LJJ
>
I-
<

_J
UJ
CC

LU
CC

5 15 25 35 45 55 65 75 85 95 105 115 125 135 145

WATER DEPTH (meters)

Figure 4.— Proposed models for the depth distribution of harbor porpoise expressed as relative den-
sities, 4, within 10 m depth intervals. Density at 20-30 m is assumed to be knowm and is given a
relative value of 1. Figures represent a) a primary model of offshore distribution, b) a high estimate
of offshore range, and c) a low estimate.

—

High Model

—

' F " '

5 15 25 35 45 55 65 75 85 95 105 115 125 135 145

in each area. Both surveys show similar patterns,
with higher abundances in the northern strata (4-8)
and very low abundance in strata 1 and 3. Despite
similar patterns, differences between the paired
estimates are in some cases, large and statistically
significant (^tests, P < 0.05). Because region 8 was
not covered on the third survey, it is not possible
to compare estimates of total abundance for the en-
tire coast between surveys. The total abundances
for regions 1-7 (Point Conception to the Columbia

River) are 46,550 (SE = 10,932) animals and 32,029
(SE = 10,906) animals for surveys 1 and 3, respec-
tively. The difference between these estimates is not
statistically significant (^test, P > 0.05). Pooling the
results of the two surveys, the estimate of harbor
porpoise abundance between Point Conception and
Cape Flattery in September of 1984 and 1985 is
45,713 (SE = 7,865) animals (Table 4). The same
estimate using the alternate models of offshore
distribution ranges from 28,769 to 78,019 (Table 5).

427

FISHERY BULLETIN: VOL. 86, NO, 3

Table 4.— Estimated abundance of harbor porpoise in each of the
eight geographic strata based on the primary model of offshore
distribution. Estimates for Beaufort 0 & 1 and for Beaufort 2 were
computed separately and then averaged, weighting by transect
length. Pooled estimates for the eight strata were obtained as an
average of the two surveys, weighting by transect length. All esti-
mates are adjusted for missed animals. Standard errors are in
parentheses.

Geographic
region

Survey 1

Survey 3

Pooled
estimates

Totals
Regions 1-3

Regions 1-7

Regions 1-8

2,401
(2,180)

6,909
(2,959)

11,245
(6,943)

9,061
(2,724)

16,934
(7,097)

9,808
(4,311)

2,401
(2,180)

46,550
(10,932)

56,358
(11,751)

126
(68)

932
(646)

153
(158)

9,096
(9,855)

3,296
(1,410)

12,786
(3,676)

5,641
(2,424)

1,210
(669)

32,029
(10,906)

(52)

1,459
(885)

112
(116)

7,909
(4,784)

4,806
(1,745)

11,107
(2,363)

10,416
(3,311)

9,808
(4,311)

1,667
(895)

35,904
(6,578)

45,713
(7,865)

DISCUSSION

Distribution

Harbor porpoise are not uniformly distributed
between Cape Flattery and Point Conception. Al-
though there are no obvious discontinuities within
this range, density varies geographically and tem-
porally. The most dramatic temporal changes are
between the two September surveys and the Janu-
ary-February survey (Fig. 1). The coasts of Wash-
ington and northern Oregon were found to have
relatively high densities of harbor porpoise in Sep-
tember, but, despite excellent sighting conditions,
very few porpoise were seen there in January. High
densities of harbor porpoise were also seen in
Monterey Bay on both September cruises and on
survey 4 in May. This area was intensively surveyed
in February, and few harbor porpoise were seen. As
can be seen in Figure 1, adjacent areas tended to
have similar densities within a survey. Less consis-
tency is found when the same areas are compared
between different surveys.

Table 5.— Estimated abundance of harbor porpoise in central
California (regions 1-3) and along the entire coast (regions 1-8)
based on two alternate models of offshore distribution. All
estimates are adjusted for missed animals. Standard errors are
in parentheses.

Survey 1 Survey 3

Pooled
estimates

Alternate Model b
Regions 1-3

Regions 1-8

Alternate Model c
Regions 1-3

Regions 1-8

3,966
(3,602)

95,132
(19,515)

1,505
(1,367)

35,736
(7,550)

1,986
(1,104)

770
(421)

2,744
(1,478)

78,019
(13,356)

1,054
(561)

28,769
(4,995)

The apparent changes in distribution could be
caused by small changes in depth distributions. The
majority of survey effort was along the 18 m iso-
bath. A large fraction of animals could be missed
if their depth distribution changed by 10 m or less.
More information on depth distributions is needed
before the apparent temporal changes in geographic
distribution can be interpreted.

Porpoise Density

Estimates of harbor porpoise density ranged from
0.03 to 2.8 animals/km^ along transect lines in the
eight geographic regions (pooled estimates, Table
2). In another study, Szczepaniak (fn. 9) estimated
0-1.9 porpoise/km^ in four study areas in the Gulf
of the Farallones, CA. Taylor and Dawson (1984)
found 1.2-5.9 porpoise/km- at study sites in Glacier
Bay, AK. Flaherty and Stark^^ estimated 0.8-1.6
porpoise/km^ in Washington Sound. Densities in
the present study are therefore within the range of
densities found in other areas along the same coast.

Harbor porpoise density was estimated for
California, Oregon, and Washington based on aerial
surveys that were concurrent with the present study
(Barlow et al. 1988). The overall estimate of harbor
porpoise density from that study (corrected for
missed animals) was 1.79 porpoise/km^. The overall
estimate from the ship survey (1.33 porpoise/km^)
can be corrected for missed animals to yield an
estimate of 1.73 porpoise/km^. Given that the coef-

I'Flaherty, C, and S. Stark. 1982. Harbor porpoise {Phocoena
phocoena) assessment in "Washington Sound". Final Report
#80-ABA-3584 submitted to National Marine Mammal Laboratory,
National Marine Fisheries Service, NOAA, 7600 Sand Point Way,
NE, Seattle, WA 98115. 84 p.

428

BARLOW: SHIP SURVEYS OF HARBOR PORPOISE

ficient of variation in the pooled ship estimates is
nearly 25%, these estimates are in very close agree-
ment. However, because the aerial estimates are
based only on the small fraction of the coastline that
was surveyed under optimal conditions, the ship
estimates are probably a better representation of
porpoise density for the entire coast.

Of the areas surveyed, harbor porpoise density is
highest in northern California and Oregon. The
highest density was seen in northern Oregon (region
7) during survey 1. The second highest density was
observed in northern California between Bodega
Head and Cape Mendocino (region 4) on survey 3.

Two areas in central California (regions 1 and 3)
were found to have very low densities. Region 1 in-
cludes the Big Sur coastline from Point Conception
to Point Sur. This area is characterized by steep
depth gradients and hence has little habitat that is
suitable for harbor porpoise. Region 1 was relatively
well covered, with 378 km of trackline surveyed at
Beaufort sea states 0-2. In contrast, region 3 in-
cludes the Gulf of the Farallons with its broad
coastal shelf within the 100 m isobath. Based on
surveys of 764 km, Szczepaniak (fn. 9) estimated
1,033 harbor porpoise are found in the Gulf of the
Farallones alone. This is much greater than my
estimate of 112 animals in region 3 based on only
175 km of survey effort. Because of his greater
amount of search effort in this area, I believe that
Szczepaniak' s estimates for region 3 are more ac-
curate than mine. Although regions 1 and 3 were
both identified as low density areas, more confidence
can be placed on this conclusion for region 1 than
for region 3.

Abundance

The size and behavioral characteristics of harbor
porpoise make estimating their abundance difficult.
Harbor porpoise are small, occur in groups of only
a few individuals, and surface without conspicuous
splashes; their distribution is extremely patchy.
Even v^th 5 observers, the effective path width that
can be searched from a ship is <1 km, and that path
width decreases very rapidly in rougher sea states.
All of these factors contribute to high variability in
the abundance estimates presented here. Seasonal
and year-to-year changes in the distribution of har-
bor porpoise may also contribute to the variability
seen within geographic strata. These are, however,
the best (and, for some regions, the only) estimates
of harbor porpoise abundance for the study area.

Although there are no prior estimates for Oregon
or Washington coasts, Dohl et al. (fn. 4) estimated

harbor porpoise abundance in central and northern
California. Their estimates range from 3,000 har-
bor porpoise in autumn to 1,600 in summer, which
correspond (approximately) to the pooled estimate
of 11,457 for regions 1-4 based on the present study.
There are, however, several problems with the ap-
plication of their methods to the estimation of har-
bor porpoise abundance. In a direct comparison with
shore counts, Kraus et al. (1983) showed that
observers on aircraft saw only 10-20% of harbor
porpoise groups. Dohl et al. (fn. 4) did not apply a
correction to account for harbor porpoise groups
that are submerged at the time the aircraft passed.
Also, Dohl et al. did not stratify estimates by
distance from shore or depth. Although most of their
harbor porpoise sightings were within 0.5 km (0.25
nmi) of shore, their density estimates were extrap-
olated to an area extending 166 km from the coast.
Estimates from the current study are based on
better methodology than previous estimates.

In addition to exposed coastal habitats, harbor
porpoise are also found in bays along the coasts of
California, Oregon, and Washington. Goetz (1983)
reported that harbor porpoise are found throughout
the year in Humboldt Bay, CA. Harbor porpoise
have been seen in San Francisco Bay, but are
described as rarely present^^. Abundance of harbor
porpoise in inland waters may, however, vary
seasonally (Taylor and Dawson 1984). No estimates
exist for the total number of harbor porpoise in-
habiting bays. Survey effort in the present study
was limited to exposed coastal areas (including
Monterey Bay, but excluding San Francisco Bay,
Humboldt Bay, Coos Bay, Yaquina Bay, the mouth
of the Columbia River, Willapa Bay, and Grays Har-
bor). If harbor porpoise density in bays were the
same as that which was observed along the 18 m
isobath, population sizes presented here could be in-
creased by approximately 3.1% to account for por-
poise inhabiting 900 km"^ (the approximate com-
bined area of Humboldt Bay, Coos Bay, Yaquina
Bay, the mouth of the Columbia River, Willapa Bay,
and Grays Harbor).

Line Transect Assumptions

Biases in abundance estimates can be an even
greater problem than high variability. In the case
of estimates presented here, biases could be intro-
duced if the assumptions of line transect sampling

'^Szczepaniak, I. D., and M. A. VV^ebber. 1985. Status of the
harbor porpoise {Phocoena phocoena) in the eastern North Pacific,
with an emphasis on California. Contract report to the Center
for Environmental Education, Washington, D.C., 52 p.

429

FISHERY BULLETIN: VOL. 86, NO. 3

are not met (Burnham et al. 1980; Hammond and
Laake 1983). Of these assumptions, the most rele-
vant to this study are 1) the area must be sampled
randomly or the animals must be randomly distrib-
uted within the area; 2) all groups on the trackline
must be detected; and 3) group size must be esti-
mated without error. These assumptions will be
addressed below.

To address the first assumption (random distribu-
tion), cruise tracks were chosen to systematically
cover the coast from Point Conception to Cape Flat-
tery. Because the surveys were designed to cover
the entire longshore range of harbor porpoise in this
area, randomly placed survey tracks were deemed
unnecessary. vVlthough some areas of the coast were
missed, these locations were determined by weather
and were presumably not correlated with porpoise
abundance. Surveys were, however, limited to a
very narrow strip along the 18 m isobath. Initially,
the choice of this survey track was based on the
observation that, in aerial surveys, harbor porpoise
were usually found within 0.5 km (0.25 nmi) of the
shoreline in California (Dohl et al. fn. 4). The 18 m
isobath was simply the shallowest reasonable work-
ing depth for the NOAA survey ships. In the course
of these surveys, it was found that harbor porpoise
are commonly distributed much further from the
coast than 0.5 km and that one survey track could
not adequately cover their habitat. The offshore
distribution of harbor porpoise is not random, but
is related to water depth, distance from shore, or
both. The model from which I extrapolated density
at 18 m to density at other depths was based on a
rather limited sample at a few locations along the
coast. The assumption of random search in offshore
areas was not met. Additional work is required to
evaluate the effect of this.

The second assumption is that 100% of the
animals in the immediate vicinity of the trackline
were detected. Animals near the trackline can be
missed because they move away from the path of
the ship, because they do not surface within the
visual range of the observers, or because the
observers fail to detect animals that do surface. Any
of these would result in a negative bias and an
underestimation of porpoise abundance using line
transect methods. These three problems are con-
sidered in more detail.

West-coast harbor porpoise are commonly said to
avoid vessels (Flaherty and Stark fn. 11; Szczepa-
niak and Webber fn. 12) and may be missed or not
counted in the proper perpendicular distance
category for this reason. On the surveys, the major-
ity of harbor porpoise were oriented roughly parallel

to the ship at the time they were sighted and were
swimming parallel to the ship and in the opposite
direction (see footnote 5). This was also observed
in one instance from the helicopter; however, in that
case the group first moved perpendicular to the path
of the ship. These observations indicate that harbor
porpoise are reacting to the ship before they are
seen by observers. Reaction to and avoidance of the
ship does not necessarily mean that estimates of
trackline density are biased if animals are detected
before they travel an appreciable distance from the
trackline. In several instances, harbor porpoise sur-
faced within 50 m of the ship and directly in its path.
These animals appeared startled and quickly moved
to avoid the ship. In these cases, the rapid move-
ment of the animals and splashes associated with
that movement made the animals more visible to
observers. Because avoidance behavior may make
harbor porpoise more visible and because the
distributions of perpendicular distance show only a
single mode (at the origin), vessel avoidance prob-
ably does not introduce a large bias in harbor por-
poise abundance estimation. More work is needed
in this area.

Harbor porpoise near the trackline may also be
missed if they either inadvertently or intentionally
do not surface within the visual range of the ob-
servers. Typical mean dive times for harbor porpoise
have been measured as 1.5-2.3 minutes (Glacier
Bay, AK; Taylor and Dawson 1984), 1.8 minutes
(northern Oregon; B. Taylor^^), and 0.4-1.4 minutes
(Bay of Fundy; Watson and Gaskin 1983). The ships'
speed during surveys was approximately 18.5 km/h
or 310 m/min; thus, in 2 minutes the ship would
travel 620 m. The average distance at which animals
were first seen was 704 m from the ship. If in-
dividual dive times were appreciably longer than 2
minutes, some trackline individuals would not be
detected by observers. In data collected in north-
ern Oregon, 16% of dive times were greater than
2.5 minutes (B. Taylor fn. 13). In addition, harbor
porpoise have been reported to increase dive times
up to 7 minutes in the presence of boat traffic
(Flaherty and Stark fn. 9). (This latter estimate is
considerably longer than any other published esti-
mate, and it is possible that those researchers missed
one or more surfacings). Helicopter observations in
Monterey Bay indicated that porpoise groups did not
extend dive times in the presence of the survey
vessel (see footnote 5). This area might not be repre-
sentative, however, because harbor porpoise may

''B. Taylor, Department of Biology, University of California, San
Diego, CA 92093, pers. commun. August 1986.

430

BARLOW: SHIP SURVEYS OF HARBOR PORPOISE

be more accustomed to vessel traffic there than
along the majority of the coast. It is likely that some
harbor porpoise are missed because they do not sur-
face near the vessel; however, it is not possible to
quantify this source of bias without additional study.

Trackline animals may be missed even if they do
not avoid the ship and do surface within visual range
of the observers if their surfacing is not detected.
In another study comparing ship surveys to aerial
and shore surveys, Kraus et al. (1983) found that
observers on ships saw only about 50% of the har-
bor porpoise in an area. In that study, however, ship
observers stood only 2.5 m above the sea surface
(versus 10 m in this study), and the estimate of 50%
was based on all groups, not just on trackline
animals. Based on the experiment using monitor
observers in the present study, an estimated 22%
of harbor porpoise that surface on the trackline are
missed by the usual team of 5 observers. If this is
underestimated by some percentage, population size
would be underestimated by the same percentage.

The third critical assumption is that group size is
estimated without error. In the case of harbor por-
poise, group size is small and estimates are typical-
ly based on actual counts. For tropical dolphins,
which school in groups of several hundreds, the
problem of group size estimation is more acute (Holt
and Powers 1982; Hammond and Laake 1983). Only
in two instances did harbor porpoise group size
exceed 20: in Monterey Bay and near Point Arena,
both in California. Excluding these two sightings,
mean group sizes are 2.05, 2.33, 2.03, and 1.59 for
surveys 1, 2, 3, and 4 (respectively); including the
two sightings, means are 2.30 and 2.26 for surveys
1 and 3. These values are comparable to other esti-
mates of mean group size for coastal populations of
harbor porpoise: 2.2 based on aerial surveys in
California (Dohl et al. fn. 4), 2.6 based on ship
surveys in the Gulf of the Farallons (Szczepaniak
and Webber fn. 12), 2.3 based on shore surveys in
northern Oregon (see footnote 5), and 2.75-3.23
based on aerial surveys along California, Oregon,
and Washington (Barlow et al. 1988). The consis-
tency of all these estimates from different platforms
indicates that group size estimation from ships is
not likely to be a major source of bias in abundance
estimation.

Variance Estimation

Although the estimates of standard error for
abundance and density are very high, these may still
be underestimates because the choice of a trunca-
tion criterion was based on minimizing variance and

because all possible sources of sampling errors were
not considered. The model upon which relative abun-
dance in the various depth strata was based is too
crude to allow reasonable estimates of its variabil-
ity. Estimates based on alternate models of depth
distribution indicate that abundance estimation is
relatively sensitive to the choice of models. Addi-
tional field work may help refine this model and
allow estimation of variance for the parameters 4
in Equation (4).

ACKNOWLEDGMENTS

Surveys of this magnitude could not be executed
without the help of many people. I thank the survey
crews for their many hours of labor: S. Beavers,
P. Boveng, S. Bragg, S. Chivers, S. Diamond,
V. DoUarhide, J. Flanders, B. Goetz, S. Hawes,
S. Heimhch-Boran, A. Hohn, S. Kruse, S. Mizroch,

F. Mann, M. Newcomer, R. Rasmussen, A. Read,

A. Robles, M. Scott, K. Sechiguchi, S. Sexton,

G. Silber, I. Szczepaniak, B. Taylor, B. Troutman,
M. Webber, J. Wexler, and K. Wynne. The heli-
copter crew included R. Holt, C. Oliver, and

B. Taylor. The initial survey was planned in col-
laboration with D. DeMaster and T. Jackson. I thank
S. Buckland, R. Holt, and S. Sexton for aid in
analysis and interpretation of line transect data.
B. Taylor provided unpublished data from shore
observations of harbor porpoise in northern Oregon.
Geographic areas were computed by K. Forney.
R. Allen prepared the figures presented here. This
manuscript was improved by the critical reviews of
P. Boveng, R. Brownell, D. Chapman, D. DeMaster,
S. Diamond, D. Goodman, D. Hanan, A. Hohn,
R. Holt, J. Lecky, S. Sexton, G. Smith, and
B. Taylor.

LITERATURE CITED

Adams, L.

1951. Confidence limits for the Petersen or Lincoln index
used in animal population studies. J. Wildl. Manage. 15:
13-19.
Barlow, J., C. Oliver, T. D. Jackson, and B. L. Taylor.
1988. Harbor porpoise, Phocoena phocoena, abundance esti-
mation for California, Oregon, and Washington: II. Aerial
surveys. Fish. Bull., U.S. 86:433-444.
Buckland, S. T.

1985. Perpendicular distance models for line transect sam-
pling. Biometrics 41:177-195.

BURNHAM, K. P., AND D. R. ANDERSON.

1976. Mathematical models for nonparametric inferences
from line transect data. Biometrics 32:325-336.
BuRNHAM, K. P., D. R. Anderson, and J. L. Laake.

1980. Estimation of density from line transect sampling of
biological populations. Wildl. Monogr. No. 72, 202 p.

431

FISHERY BULLETIN: VOL. 86, NO. 3

BUTTERWORTH, D. S.

1982. On the functional form used for g{y) for Minke whale
sightings, and bias in its estimation due to measurement in-
accuracies. Rep. int. Whal. Comm. 32:883-888.
Chapman, D. G.

1951. Some properties of the hypergeometric distribution
with applications to zoological censuses. Univ. Calif. Publ.
Stat. 1:131-160.
Crandall, W. C.

1915. II. The kelp beds from lower California to Puget Sound.
In F. W. Cameron (editor), Potash from kelp, 122 p. U.S.
Dep. Agric. Rep. No. 100.
Efron, B.

1982. The jackknife, the bootstrap and other resampling
plans. CBMS Regional Conference Series in Applied
Mathemetics 38, 92 p. Society for Industrial and Applied
Mathematics, Phila.

Gaskin, D. E.

1984. The harbour porpoise (Phocoena phocoena L.): regional
populations, status, and information on direct and indirect
catches. Rep. int. Whal. Comm. 34:569-586.

GOETZ, B. J.

1983. Harbor porpoise (Phocoena phocoena L.) movements
in Humboldt Bay, California and adjacent ocean waters.
MS Thesis, Humboldt State University, Humboldt, CA, 118

P-
Goodman, L. A.

1960. On the exact variance of products. J. Am. Stat. Assoc.
55:708-713.
Hammond, P. S., and J. L. Laake.

1983. Trends in estimates of abundance of dolphins (Stenella
spp. and Delphinics delphis) involved in the purse-seine
fishery for tunas in the eastern tropical Pacific Ocean,
1977-81. Rep. int. Whal. Comm. 33:565-588.

Holt, R. S.

In press. Estimation of abundance of dolphin stocks taken
incidentally in the eastern tropical Pacific yellowfin tuna
fishery. Mar. Mamm. Sci.
Holt, R. S., and J. Cologne.

1987. Factors affecting line transect estimates of dolphin
school density. J. Wildl. Manage. 51:836-843.
Holt, R. S., and J. E. Powers.

1982. Abundance estimation of dolphin stocks involved in the
eastern tropical Pacific yellowfin tuna fishery determined
from aerial and ship surveys to 1979. U.S. Dep. Commer.,
NOAA Tech. Memo. NOAA-TM-NMFS-SWFC-23, 95 p.

Kraus, S. D., J. R. Gilbert, and J. H. Prescott.

1983. A comparison of aerial, shipboard, and land-based
survey methodology for the harbor porpoise, Phocoena pho-
coena. Fish. Bull., U.S. 81:910-913.

Laake, J. L., K. P. Burnham, and D. R. Anderson.

1979. User's manual for program TRANSECT. Utah State
University Press, Logan, Utah, 26 p.
Pollock, K. H., and W. L. Kendall.

1987. Visibility bias in aerial surveys: A review of estimation
procedures. J. Wildl. Manage. 51:502-510.
Taylor, B. L., and P. K. Dawson.

1984. Seasonal changes in density and behavior of harbor por-
poise (Phocoena phocoena) affecting census methodology in
Glacier Bay National Park, Alaska. Rep. int. Whal. Comm.
34:479-483.

Watson, A. P., and D. E. Gaskin.

1983. Observations on the ventilation cycle of the harbour
porpoise (Phocoena phocoena L.) in the coastal waters of the
Bay of Fundy. Can. J. Zool. 61(1):126-132.
Yates, F.

1953. Sampling methods for censuses and surveys. Hafner
Publishing Co. Inc., N.Y., 401 p.

432

HARBOR PORPOISE, PHOCOENA PHOCOENA,

ABUNDANCE ESTIMATION FOR CALIFORNIA, OREGON, AND

WASHINGTON: II. AERIAL SURVEYS

Jay Barlow,! Charles W. Oliver,^ Terry D. Jackson,^ and
Barbara L. Taylor^

ABSTRACT

We conducted aerial surveys in September 1984 and September and October 1985 to determine the abun-
dance of harbor porpoise along the coasts of California, Oregon, and Washington. Two observers and
a recorder searched along predetermined transect lines at 0.61 and 1.85 km offshore. Strip transect
methods were used. A total of 366 groups of harbor porpoise were seen in the 9,500 linear kilometers
that were surveyed. Apparent density was significantly affected by sea state and cloud cover. Using
observations made during optimal conditions (clear skies and calm seas), apparent harbor porpoise den-
sity averaged 0.56 animals km'". Behavioral observations from shore and from a helicopter indicated
that porpoise are near the surface only 23.9% of the time. To account for this, porpoise density was
multiplied by a factor of 3.2, resulting in an adjusted estimate of 1.79 animals km"". Only a small percent-
age of the total area inhabited was surveyed under optimal sighting conditions, hence density estimates
were not extrapolated to estimate total porpoise abundance. Harbor porpoise density showed similar
patterns to those measured from ship surveys, and adjusted aerial estimates are approximately equal
to ship estimates.

Harbor porpoise, Phocoena phocoena, are subject to
mortality in the halibut set net fishery in central
California (NMFS^; Diamond and Hanan^). To evalu-
ate the significance of this mortality, an estimate
of population size is needed. Two aerial surveys and
three ship surveys were conducted from 1984 to
1986 to gather information on harbor porpoise abun-
dance along the coasts of California, Oregon, and
Washington. Observations were also made from
shore-based stations and from a helicopter to pro-
vide ancillary information needed for population
estimation. Results and population estimates from

'Southwest Fisheries Center La Jolla Laboratory, National
Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA
92038.

^National Marine Fisheries Service, Southwest Region, 300
South Ferry Street, Terminal Island, CA 90731; present address:
Pacific Marine Center, NOAA, 1801 Fairview Avenue East,
Seattle, WA 98102.

^National Marine Fisheries Service, National Marine Mammal
Laboratory, 7600 Sand Point Way, N.E., Seattle, WA 98115;
present address: University of California, San Diego, Department
of Biologj', La Jolla, CA 92093.

^NMFS. 1980. A report based on the workshop on stock
assessment and incidental take of marine mammals involved in
commercial fishing operations. January 1980. Available from
National Marine Fisheries Service, National Marine Mammal
Laboratory, 7600 Sand Point Way NE, Seattle, WA 98115.

^Diamond, S. L., and D. A. Hanan. 1986. An estimate of har-
bor porpoise mortality in California set net fisheries: April 1, 1983
through March 31, 1984. Adm. Rep. SWR-86-15, 40 p. Available
from National Marine Fisheries Service, Southwest Region, 300
S. Ferry Street, Terminal Island, CA 90731.

the ship surveys are reported by Barlow (1988).
Preliminary results from the 1984 and 1985 aerial
surveys were presented by Oliver and Jackson^ and
Oliver^ respectively. Here we present population
density estimates based on the aerial surveys and
on shore and helicopter observations.

The aerial surveys were flown in September of
1984 and in September and October of 1985 from
Point Conception, CA to Cape Flattery, WA. Sur-
veys were coordinated by the National Marine Fish-
eries Service (NMFS) in collaboration with the
California Department of Fish and Game, the Ore-
gon Department of Fish and Wildlife, and the
Washington Department of Wildlife. Survey design
was based on information given by Dohl et al.^
regarding harbor porpoise distribution in California.
They reported that harbor porpoise were usually

^Oliver, C. W., and T. D. Jackson. 1987. Occurrence and
distribution of marine mammals at sea from aerial surveys con-
ducted along the U.S. west coast between December 15, 1980 and
December 17, 1985. Adm. Rep. LJ-87-19, 189 p. Available from
National Marine Fisheries Service, Southwest Fisheries Center,
P.O. Box 271, La Jolla, CA 92038.

^Oliver, C. W. 1986. Trip report: 1985 harbor porpoise aerial
survey, September 9 to October 15, 1985. Adm. Rep. LJ-86-21,
29 p. Available from National Marine Fisheries Service, South-
west Fisheries Center, P.O. Box 271, La Jolla, CA 92038.

»Dohl, T. P., R. C. Guess, M. L. Duman, R. C. Helm. 1983.
Cetaceans of central and northern California, 1980-83: status,
abundance, and distribution. Report prepared for U.S. Minerals
Management Service, contract #14-12-0001-29090.

Manuscript accepted May 1988. ,

fishery BULLETIN: VOL. 86, NO. 3, 1988. V//' ^Yj

433

FISHERY BULLETIN: VOL. 86, NO. 3

found within 0.25 nautical miles (nmi) of the shore-
line. We therefore designed our aerial surveys to
cover a very narrow coastal band. Subsequent in-
formation from ship surveys (Barlow 1988) has
shown their distribution to extend considerably far-
ther from the coast. Therefore, estimates of porpoise
density from aerial surveys apply to a relatively
small portion of harbor porpoise habitat. For this
reason we do not estimate population size by ex-
trapolating aerial density estimates to the entire
area inhabited. The density estimates presented
here are used to corroborate estimates based on ship
surveys and to estimate density for areas that were
too shallow to be surveyed by ship.

Based on previous studies of dive times (Watson
and Gaskin 1983; Taylor and Dawson 1984) we ex-
pected a proportion of the harbor porpoise to be
diving and therefore missed by aerial observers.
Shore-based studies were conducted in September

1985 from cliffs in northern Oregon to determine
average dive times for west-coast harbor porpoise.
Helicopter observations were made in April and May

1986 in Monterey Bay and near Bodega Head, CA
to gather dive time information and obtain a direct
measure of the fraction of time that harbor porpoise
groups are visible from the air. These two samples
did not differ significantly from previous samples
of harbor porpoise in Alaska, so all samples were
pooled to adjust estimates of porpoise density from
aerial surveys to account for the probability of miss-
ing submerged animals.

METHODS

Aerial Survey Methods

Strip transect methodology (Seber 1973) was used
during the aerial surveys. This method assumes that
all individuals within a transect strip are detected.
Transect lines were flown parallel to the coast line
at distances of 0.61 and 1.85 km (0.33 and 1.0 nmi)
offshore. Transect strips of equal width were sur-
veyed on both sides of the aircraft. The margins of
the strips were denoted by tape marks or streamers
on the wing struts. Strips were divided into inside
and outside swaths of unequal width (Fig. 1) by a
third tape mark or streamer between the other two.
When porpoise were sighted within the transect
strip, the pilot was directed to leave the transect
line and circle over the porpoise to obtain an ac-
curate count of the number within the original group
that was sighted. If additional groups or individuals
were sighted during this circling, they were ex-
cluded from density estimates. Porpoise density, x,

was calculated as the number of individuals sighted
within a transect, n, divided by the product of the
transect width, w, times the distance, d, that was
flown:

X = n/iW ■ d)

(1)

We used both single and twin-propeller, high-
wing, 4-passenger aircraft in our surveys. The
search team consisted of two observers seated in the
right and left passenger seats. A data recorder sat
in the copilot's seat and did not search. If the
recorder sighted animals that were missed by the
observers, these were noted but were not included
in density estimates. The planes were flown at an
altitude of 213 m (700 feet) and at an airspeed of
158-167 km/h (85-90 knots). The original survey
plan called for all sections of the coast to be covered
twice on each survey. This was accomplished in
1984, but poor weather in 1985 resulted in the
Washington coast and part of the Oregon coast be-
ing covered only once. The dates flown and areas
covered are given in Table 1.

Table 1.— Dates, areas covered, and observer teams during
aerial surveys for harbor porpoise. Geographic regions refer
to those shown in Figure 2. Observer team refers to a pair of
individuals.

Regions

Observer

Distance

Date

covered

team

to shore (km)

9/09/84

7,8

0.61

9/10/84

5,7

0.61

9/11/84

5,6

0.61

9/1 3/84

1,2,3,4,5

0.61

9/14/84

1,2,3,4,5

0.61

9/17/84

1,2,3,4

0.61

9/18/84

1,2,3,4

0.61

9/11/85

6,7

0.61

9/16/85

1,2,3

0.61

9/17/85

0.61

9/18/85

4,5

0.61,1.85

9/1 9/85

4,5

0.61,1.85

9/20/85

2,3

1.85

10/04/85

7,8

0.61

10/14/85

1,2,3

1.85

10/15/85

4,5

1.85

Data gathered on both the 1984 and 1985 surveys
were similar in format. Recorded data on sighting
conditions included Beaufort sea state, a measure
of cloud cover, a code indicating the presence of haze
or fog, sun position relative to the aircraft, and a
subjective measure of the observers' ability to see
into the water through turbidity, surface reflection
and diffraction. The latter was called surface pene-
tration and was recorded separately for each of the

434

BARLOW ET AL.: AERIAL SURVEYS OF HARBOR PORPOISE

STRIP TRANSECT

300 400
DISTANCE (meters)

500

600

700

\ Inside
\ Swatt

\. c

1 ^

)utsi(

vath

--.

1985

\ '-y/yy/y//.

•I-;-;-;-!-;-!";'!

|------^-N ^ 1

100 200 300 400

DISTANCE (meters)

500

600

700

Figure 1.— Configuration of transect strip widths and distributions of perpendicular sighting distances for aerial
surveys in 1984 and 1985. Angles are given as declinations from horizontal. Histograms indicate the relative number
of porpoise seen in the given distance interval.

four swaths (inside and outside on both sides of the
aircraft). In severe sun glare conditions, searching
was discontinued for one or both swaths on one side
of the aircraft. Additional data included individual
observer numbers, date, time, and position (meas-
ured to lOths of minutes of latitude and longitude).
All of these data were recorded at the beginning and
end of continuous transects and whenever condi-
tions changed or a sighting was made. Additional
data were recorded for marine mammal sightings,
including a code for the species of animal seen, an
estimate of the number of individuals, a code in-
dicating on which side of the aircraft the animals
were seen, and a code indicating in which of the
swaths the animals were found (inside, outside, or,
if the animals were not within the designated strips,
neither).

An attempt was made to gather the above infor-
mation in a similar manner for both surveys. Some

differences in subjective measures of sighting con-
ditions could, however, be expected because there
was no overlap in observers between years. In addi-
tion, there were some differences in design between
the 1984 and 1985 surveys. In 1984, the inside and
outside swaths were from 123 to 305 m and from
305 to 620 m, respectively (as measured from the
midline of the transect). The margins of these
swaths corresponded to declinations angles of 19°,
35°, and 60° (Fig. 1). In 1985, the swaths were
91-294 m and 294-503 m and corresponded to
declination angles of 23°, 36°, and 67° (Fig. 1). In
1984, effort was concentrated on the inside swath
and the outside swath was not intended for abun-
dance estimation. The change in swath size was in-
tended to reduce the total area being searched, thus
potentially allowing the outer swath to be used for
density estimation. In 1985, effort was divided
equally between the two swaths. During the 1985

435

FISHERY BULLETIN: VOL. 86, NO. 3

surveys the declination angle to marine mammals
was measured when the animals were perpendicular
to the aircraft using hand-held inclinometers.
Declination angles were not measured during the
1984 survey.

Distance to the coast was monitored using
declination angles (19° for 0.61 km and 6° for 1.85
km). On the 1984 survey, this distance was 0.61
km for the entire survey. During the 1985 survey,
we surveyed at both 0.61 and 1.85 km from the
coast. In 1984, the coast was taken to be the outer
limit of the surf zone. In 1985, the coast was taken
to be the outer limit of the surf zone or, if kelp
beds were present, the outer margin of those
beds.

Shore Observation Methods

Observations of harbor porpoise diving behavior
were made from rocky headlands in northern Ore-
gon (Tillamook Head, Neahkahnie Mountain, Cape
Meares, and Cape Lookout) immediately before the
second aerial survey (7-11 September 1985). Ob-
servers were equipped with 7 x 50 binoculars with
compasses and ocular reticles and a single 20 x 120
binocular. Ventilation data were collected whenever
possible and included the number of animals at the
surface and the length of time spent at the surface.
Observations were recited aloud by the observer and
were written down by a second person or were
recorded onto magnetic tape. The ventilation cycle
typically consisted of a period with several surfacing
rolls and breaths (which we call a surfacing series)
followed by a much longer period of submergence
(which we call a dive). This dive cycle corresponds
to ventilation pattern B as described by Watson and
Gaskin (1983) for harbor porpoise in the Bay of
Fundy and the pattern described by Taylor and
Dawson (1984) for porpoise in Glacier Bay.

Helicopter Observation Methods

Behavioral observations were also made by three
observers in a 4-passenger, jet-turbine helicopter.
Upon locating a group of harbor porpoise, a fluores-
cein dye marker was dropped and the helicopter
hovered or circled slowly above the group at an
altitude of approximately 300 m. The number of
animals, the time they were visible at the surface,
and the dive times were recorded, along with infor-
mation on cloud cover, sea state, and water turbid-
ity. Each behavioral session was given a subjective
rating based on how well the observers could follow
the group and obtain accurate dive times. Only ses-

sions with good or excellent ratings were included
in analyses.

Probability of
Missing Submerged Animals

Given that a porpoise would be within the visual
range of an observer, the probability that it will be
at the surface during the passage of the aircraft is
related to the average time it spends at the surface,
s, the average time spent below the surface, d, and
the window of time during which it is within the
visual range of an observer, t. This probability was
calculated as

Pr (being visible)

s -I- t
s + d

(2)

The probability of missing a submerged animal is
equal to the complement of this value.

Density Estimation

Density of harbor porpoise was estimated as the
number of animals seen divided by the area searched
(Equation (1)). This raw density estimate was ad-
justed by dividing by the probability that an animal
would be visible from the air at any given instant
(Equation (2)). The area searched was estimated as
the swath widths times the lengths of the transects.
Transect lengths were calculated as the sum of the
great circle distances between successive position
fixes. Densities were calculated for each of the eight
statistical regions used by Barlow (1988) (Fig. 2).

The statistical difference in harbor porpoise
density between different sighting conditions or
different areas was tested using the raw density
estimates. Density estimates for short transects
were frequently zero, thus violating the parametric
assumptions of normally distributed, homoscedastic
error. Nonparametric tests were therefore chosen
for density comparisons. In discussing statistical
tests, a transect segment refers to the length of
transect line between two successive position fixes
and are typically <20 km. The measured variables
relating to sighting conditions are constant within
a segment, and because each sighting is accom-
panied by a new position fix, a segment will contain
at most, one sighting.

Whenever applicable, the Wilcoxon paired-sample
test (Wilcoxon 1945) was used to test one factor
while controlling for as many other factors as possi-
ble. Ten paired measures of density were created

436

BARLOW ET AL.: AERIAL SURVEYS OF HARBOR PORPOISE

49°N

48°

47°

46°

43°

-T — \ 1 ^-y

40°

38°

37°

36°

35°

32^

-l_

REGION

REGION 4

REGION 3

_i_

UNITED STATES

Bodega Head

San Francisco

REGION 2

REGION 1

Point Sur

\ i Point Conception

\?^pa

_i_

132°W 131° 130° 129° 128° 127° 126° 125° 124° 123° 122° 121° 120° 119° 118° 117°W

Figure 2.— Geographic regions used as strata in density estimation.

by adding all appropriate transect segments to a
linear array and dividing that array into 10 equal
parts.

When paired tests were not applicable, nonpara-
metric ANOVA models were used. For simple com-
parisons, the Kruskal-Wallis single-factor analysis

of variance was used (Kruskal and Wallis 1952). For
two-way comparisons, we used a two-factor exten-
sion of the Kruskal-Wallis test (Scheirer et al. 1976).
For both tests, three replicate measures of density
were created for each cell by randomly assigning
transect segments as replicate 1, 2, or 3.

437

FISHERY BULLETIN: VOL. 86, NO. 3

A principal assumption of strip transect methods
is that all individuals within the designated strip are
counted. Assuming that the fraction of diving
animals does not vary with sighting conditions, any
effect of sighting conditions on apparent density is
likely due to missed animals. In any instance where
we show that poorer sighting conditions result in
a significant decrease in apparent density (one-tailed
for paired tests, tw^o-tailed for ANOVA tests), we
eliminate the category of sighting conditions which
resulted in that lower estimate. If it is not possible
to predict which category would be worse a priori
in paired tests (e.g., right swath vs. left), two-tailed
probabilities are used. Although we cannot be sure
of eliminating all biases using these methods, this
pattern of data paring should avoid much of the bias
due to missed animals.

RESULTS

In 1984, 247 groups of harbor porpoise (680 in-
dividuals) were seen within transect strips which
covered a linear distance of 5,763 km. In 1985, we
saw 119 groups (384 individuals) in surveys of 3,715
km. Mean group sizes were 2.75 and 3.23 in-
dividuals, respectively, for 1984 and 1985. For 1984,
the relative frequencies of individuals seen within
the inside and outside swaths are illustrated in
Figure 1. For 1985, the perpendicular distances
from the trackline to the animals were calculated
from declination angles, and the relative distribu-
tion of sightings is shown in Figure 1 as a function
of perpendicular distance.

Inside vs. Outside Swath

For 1984 data, only the inside swaths were used,
but for 1985 both inside and outside swaths were
considered for density estimation. For 1985, we
tested whether the density in the inside swaths was
greater than the density of the outside swaths. We
only considered cases when the water surface pene-
tration codes were equal in both the inside and out-
side swaths. Data for inside and outside were thus
paired, with all other sighting factors equal. For
1985, the density in the inside was greater (0.09 vs.
0.06 porpoise/km^), but this difference was not
significant (P > 0.10).

Surface Penetration

Observers used a subjective coding system to
describe their ability to see through the sea surface.
Cloud cover, haze, and water turbidity contributed

to poor surface penetration. In 1984, observers used
codes to indicate good and poor conditions; in 1985,
observers used codes for excellent, good, and poor.
There were frequent cases when observers recorded
different codes for the inside swaths on opposite
sides of the plane; hence, paired tests were again
appropriate. For 1984, we tested whether density
in the "good" category was higher than density in
the "poor" category. When surface penetration was
different in the inside swaths on opposite sides of
the plane, mean density in the "good" category was
higher than in the "poor" category (0.16 vs. 0.12
porpoise/km^), but this difference was not signifi-
cant (P > 0.25). For 1985, no tests were necessary
because the mean density in "excellent" category
was lower than in the "good" category, and like-
wise, density in the "good" was lower than in the
"poor" category. All categories of water surface
penetration were included in subsequent analyses.

Effects Due to Observers and
Side of the Plane

Sightings were classified based on which observer
made the sighting and on whether the sighting was
on the inshore or offshore side of the aircraft. In
fact, these two classifications were confounded in
1985 because the two principal observers were sit-
uated on the same sides of the aircraft for most of
this survey. Effects of these classifications on den-
sity estimation were considered together. Survey
teams were defined as pairs of observers who
worked together. There were two such teams for
1984 and three for 1985 (Table 1). Only one of the
teams in 1985 had sufficient numbers of sightings
to be considered here. Statistical tests were based
on paired cases during which both members of the
sighting team were searching.

For 1984, porpoise density on the offshore side
of the airplanes was greater than on the inshore side
for both team A and team B (Table 2). The differ-
ence in density estimates between observers was
less than the difference between inshore and off-

Table 2.— Relative harbor porpoise densities (km"^) for teams
of observers. Density estimates are stratified by inshore and
offshore sides of the aircraft and by individual observers.

Observer
team

Inshore Offshore

Observer
1

Observer
2

A
B
D

0.275
0.168
0.281

0.425
0.244
0.164

0.320
0.232
0.300

0.380
0.179
0.146

438

BARLOW ET AL.: AERIAL SURVEYS OF HARBOR PORPOISE

shore (Table 2), and neither was statistically signif-
icant. For 1985, the opposite was seen; the density
on the inshore side was greater than on the offshore
side and this difference was less than the difference
between observers (Table 2), but again, neither dif-
ference was statistically significant. In 1984, both
observers had previous experience doing aerial
surveys, but in 1985, the observer with the lower
density estimates had no previous experience in
cetacean surveys. Experience may be a factor in
density estimates from strip transects, but since the
inexperienced observer was always on the same side
of the plane, it was not possible to test this with a
factorial design.

Area and Sea State

Analysis of research vessel data (Barlow 1988) in-
dicated two geographic regions in California with
low porpoise density (regions 1 and 3 in Figure 2).
We tested whether density observed from aircraft
are also lower in these areas. Because observation
conditions may have differed in the two areas, we
included sea state as a second factor in a two-way
ANOVA. Mean values are presented in Table 3. For
1984, three categories of sea state were used:
Beaufort 0 & 1, Beaufort 2, and Beaufort 3 and
greater. The effect due to area was significant (P
< 0.001), with the area that showed low density in
the ship surveys also showing lower density in the
aerial survey. For 1985, only the first two categories
of sea state were used due to insufficient data at
Beaufort 3 and greater. Again the effect of area was
significant (P < 0.025), and the same trends were
seen. In neither case were the effects of sea state

Table 3. — Uncorrected harbor porpoise densities (km' ) for
the two-way comparison of area and sea state. Low-density
areas refer to two regions in California that were found to have
much lower than average density in previous ship surveys (see
text). High-density areas include all other regions. Only inside
swaths were included. Numbers in parentheses refer to area
(km^) surveyed under the given condition. Densities in
brackets were excluded from two-way comparisons due to the
small area covered in one cell.

1984

survey

1985

survey

Beaufort
sea state

Low-
density
area

High-
density
area

Low-
density
area

High-
density
area

0 & 1

0.027
(183)

0.579
(264)

0.089
(124)

0.762
(126)

0.081
(308)

0.312
(1012)

0.089
(214)

0.183
(425)

3-1-

0.102
(79)

0.191
(252)

[0.000]
(28)

[0.049]
(566)

or the interaction effects significant. To eliminate
area effects from confounding statistical results,
only data for the larger, high-density area were in-
cluded in subsequent tests. The low-density area was
included in later estimates of overall harbor porpoise
density.

Sea State and Cloud Cover

Both sea state and cloud cover can affect sighting
conditions. Because both are affected by local
weather, the effects of these are likely to be con-
founded. These two factors were therefore tested
simultaneously in a two-way ANOVA. We con-
sidered only the inside swath and excluded the
two low-density regions. We used the same sea
state categories as above. The sky was categorized
as clear if cloud cover was <25% and cloudy if
>25%. Mean porpoise densities for each category
are given in Table 4. It was necessary to exclude
the Beaufort 0 & 1 category for 1985 because only
52 km were surveyed in cloudy conditions for these
sea states. The effect due to cloud cover was signifi-
cant for 1984 (P < 0.025) and 1985 (P < 0.05). The
effect of sea state and the interaction effect of
sea state and cloud cover were not significant for
either survey (P > 0.10). Transect segments with
>25% cloud cover were excluded from subsequent
analyses.

Table 4. — Uncorrected harbor porpoise densities (km"^)
for the two-way comparison of sea state and cloud cover.
Clear refers to <25% cloud cover, and cloudy refers to
>25%. Data include only inside swaths in high-density
areas. Numbers in parentheses refer to area (km^) sur-
veyed under the given condition. Densities in brackets were
excluded from two-way comparisons due to the small area
covered in one cell.

Beaufort
sea state

1984

survey

1985

survey

Clear

Cloudy

Clear

Cloudy

0& 1

1.340
(81)

0.240

(183)

[0.832]

(115)

[0.000]
(11)

0.371
(572)

0.236
(440)

0.271
(232)

0.078
(193)

3 +

0.266
(162)

0.056
(90)

0.088
(272)

0.014
(294)

Sea State

The effect of sea state was tested alone using only
the transect segments which occurred under clear
skies (<25% cloud cover) and within the high-density
areas. For 1984, only the inside swath was included.

439

FISHERY BULLETIN: VOL. 86, NO. 3

For 1985, both inside and outside swaths were in-
cluded. As in previous stratifications, apparent den-
sity clearly decreased with increasing sea state
(Table 5). This effect was significant for surveys in
both 1984 (P < 0.05) and 1985 (P < 0.005). Transect
segments surveyed during sea states of 2 or greater
were excluded in subsequent analyses.

•Uncorrected harbor porpoise den-
^) for the stratification based on sea

Table 5.-
sities (km

state. Data include only high-density areas sur-
veyed when cloud cover was <25%. Data for
1984 include inside swaths only; data for 1985
include inside and outside swaths. Numbers in
parentheses refer to area (km^) surveyed
under the given condition.

Beaufort

sea state

1984 survey

1985 survey

0 & 1

1.340

0.807

(81)

(234)

0.371

0.193

(572)

(471)

3 +

0.266

0.058

(162)

(553)

Between Survey Differences

We considered the 1984 survey and the 1985
surveys at 0.61 km and 1.85 km from shore as three
independent estimates of harbor porpoise density.
Because apparent density was shown to vary greatly
with sighting conditions and because sighting con-
ditions varied between surveys, it was necessary to
compare these three under similar conditions.

The highest (and presumably least biased) den-
sities were obtained when sea state was Beaufort
0 & 1 and when cloud cover was <25%. Between
survey comparisons imder these conditions are given
in Table 6 for the eight geographic regions given
in Figure 2. For 1985, there were no transect
segments at 0.61 km from shore under the condi-
tions Beaufort 0 & 1 and clear skies. For 1984, only
three regions contained more than 10 km- of
searching effort at 0.61 km from shore. For 1985,
only four areas had any searching effort at 1.85 km
from shore. The only direct density comparisons
with reasonable sample sizes are for region 1 (0.000
vs. 0.048 porpoise/km^) and region 3 (0.111 vs.
0.110 porpoise/km^) (respectively for 1984 and
1985). The densities for all regions pooled (0.671 and
0.510 porpoise/km^) are similar, but because of the
small sample size and geographic variation in sam-
pling, a statistical test of this difference is mean-
ingless.

In comparing surveys, sample size and regional
coverage improved slightly when Beaufort 2 was
considered (still allowing a maximum of 25% cloud
cover) (Table 7). For 1984, coverage was relatively
complete in all regions. For 1985, coverage at 0.61
km from shore was limited to regions 2-4, and
coverage at 1.85 km was limited to regions 1-5.
Comparing the two surveys in 1985, distance from
shore made little difference in overall density for
all regions combined, and neither survey had con-
sistently higher values than the other. Comparing
the 1984 survey to the two 1985 surveys, the former
had a higher overall density for all regions combined,
but again this difference was not consistent among
regions. Sample size and regional coverage were
again too poor for meaningful statistical tests.

Table 6.— Uncorrected harbor porpoise densities (km"^) in
eight geographic regions surveyed during Beaufort 0 & 1 con-
ditions. Data for 1984 are based on inside swaths of transects
flown 0.61 km from the coast. Data for 1985 are based on in-
side and outside swaths of transects flown at 0.61 and 1.85
km from the coast. Only those segments surveyed when cloud
cover was <25% are included. Numbers in parentheses refer
to area (km^) surveyed under the given condition.

Geographic
region

1984 survey
density
0.61 km

1985 survey
density
0.61 km

1985 survey
density
1.85 km

0.000

0.048

(43)

(0)

(103)

0.000

—

0.609

(1)

(0)

(51)

0.111

—

0.110

(45)

(0)

(55)

0.953

—

0.862

(6)

(0)

(183)

1.120

—

(9)

(0)

1.374

—

(64)

(0)

5.745

—

(1)

(0)

All regions

0.671

—

0.510

(170)

(0)

(392)

Ventilation Patterns

Harbor porpoise did not appear to react to the
helicopter during aerial observations; they were
visible throughout a surfacing series and were not
visible during dives. Knowing this, we were able to
use data on surfacing series and dive times to deter-
mine the fraction of time harbor porpoise would be
visible from the air. Data on ventilation patterns

440

BARLOW ET AL.: AERIAL SURVEYS OF HARBOR PORPOISE

were available from the helicopter study in Califor-
nia (13 groups were observed, mean group size was
2.7), the Oregon shore observations (11 groups,
mean group size was 4.2), and a previous shore study
in Glacier Bay, AK (28 solitary individuals). The
mean times spent in surfacing series were 34.2, 24.6,
and 30.0 seconds (respectively for the three studies).
The corresponding mean dive times did not differ
significantly between study sites (P > 0.05 using
pairwise ^tests). Using the pooled data set (n = 52),
the mean time spent in a surfacing series was 30.02
seconds (SE = 1.95) and the mean time spent in a
dive was 95.81 seconds (SE = 5.32). The percent-
age of time spent at the surface is 23.9%.

Average and Adjusted
Density Estimates

Harbor porpoise densities under optimal condi-
tions (Beaufort 0 & 1 and <25% cloud cover) were
averaged for the two surveys, weighting by transect
length (Table 8). Given an average survey speed of
160 km/h and assuming that the window for harbor
porpoise observation is 400 m long, the time win-
dow during which a point would be visible is 9.0

Table 7— Uncorrected harbor porpoise densities (km"^) in
eight geographic regions surveyed during Beaufort 2 condi-
tions. Data for 1984 are based on inside swaths of transects
flown 0.61 km from the coast. Data for 1985 are based on in-
side and outside swaths of transects flown at 0.61 and 1.85
km from the coast. Only those segments surveyed when cloud
cover was <25% are included. Numbers in parentheses refer
to area (km^) surveyed under the given condition.

Geographic
region

1984 survey
density
0.61 km

1985 survey
density
0.61 km

1985 survey
density
1.85 km

1
2
3
4
5
6
7
8
All regions

0.000
(107)

0.092
(109)

0.196
(117)

0.562
(180)

0.489
(119)

0.439
(57)

0.114
(70)

0.268
(37)

0.295
(796)

(0)

0.102
(59)

0.062
(48)

0.282
(110)

0.459
(6)

(0)

0.192
(224)

0.184
(103)

0.103
(68)

0.287
(52)

0.070
(57)

0.234
(171)

(0)
(0)

(0)

0.188
(452)

seconds. Using this estimate and the surface and
dive times estimated above, the probability that a
porpoise will be seen is estimated as 0.310 from
Equation (2). An instantaneous count would there-
fore underestimate porpoise abundance by a factor
of 3.2. Average values were therefore multiplied by
this factor (Table 8).

DISCUSSION

Results indicate that sighting conditions must be
very good in order to estimate harbor porpoise abun-
dance from aerial strip transects. Both sea state and
cloud cover had very large and significant effects
on apparent density. Limiting observations to the
best categories of sea state (Beaufort 0 & 1) and
cloud cover (<25%) can be used to minimize the bias
due to missed animals. These conditions are, how-
ever, rare and only occurred during 5.3% and 10.3%
of the transects in 1984 and 1985. The actual occur-
rence of these conditions is even more rare if one
considers flights that were cancelled due to bad
weather.

The effects on sea state and cloud cover on sight-
ing conditions were predicted by observers before
analysis of survey data was begun. Most harbor por-
poise were first seen when submerged a small
distance below the surface. Surfacings were rela-
tively inconspicuous to aerial observers and were not
an important cue in sighting porpoise. Both sea state
and cloud cover affect the ability of observers to see
through the water's surface and to spot submerged
animals. Increasing sea state causes more refrac-
tion of light at the water's surface, increases glit-

Table 8.— Harbor porpoise densities (km"^) in eight geographic
regions from a) a weighted average of uncorrected estimates from

1984 and 1985 aerial surveys, b) the same average adjusted by
a factor of 3.2 to account for submerged porpoise that were missed
by aerial observers, and c) a weighted average for the 1 984 and

1985 ship surveys. Aerial estimates are based on observations
made under conditions of Beaufort 0 & 1 and with <25% cloud
cover. Ship estimates include Beaufort sea states of 0, 1, and 2.

Geographic
region

Area

surveyed

(km*)

Aerial estimates

Uncorrected
a

Research
vessel
Corrected estimates
b c

146

0.03

—

100

0.11

189

0.86

1.12

1.37

—

All regions

562

0.55!

0.10

0.05

—

0.66

0.35

0.04

2.75

3.68

3.58

1.18

4.38

2.88

—

3.43

—

1.42

1.79

1.73

441

FISHERY BULLETIN: VOL. 86, NO. 3

tery reflection of sunlight, and causes whitecaps
which obscure subsurface observation. Cloud cover
decreases penetration of sunlight into the sea and
causes its surface to appear dark and glazed. It is
therefore not surprising that calm seas and clear
skies result in higher apparent densities.

It is more surprising that apparent density did not
vary with observer's subjective appraisal of surface
penetration. This may have been because surface
penetration was only tested in paired cases for which
sea state and cloud cover were identical. In these
cases, differences in surface penetration may have
been due primarily to subjective differences in the
way individual observers were coding it. If all cases
are considered, surface penetration is very highly
correlated with sea state and cloud cover and prob-
ably could be used as an alternative measure of
sighting conditions. We prefer, however, to use sea
state and cloud cover because their measure is less
subjective than surface penetration and could more
easily be used by other researchers.

Missed Animals

A principal assumption of strip transect methods
is that all individuals within the strip are counted.
We cannot necessarily meet this assumption just by
eliminating the categories of sighting conditions
with significantly lower density. In fact, it is possi-
ble that this method of selectively eliminating data
could overestimate density by eliminating a category
of sighting conditions which (by random chance) had
a significantly lower density. We do not believe that
this is likely in the cases of sea state or cloud cover
because the trends were the same for both surveys
and because the categories that were eliminated
were judged a priori as being poorer sighting con-
ditions. We believe that porpoise density is more
likely underestimated due to missed animals. Kraus
et al. (1983) found that observers in aircraft saw only
14% of the harbor porpoise groups known to be pres-
ent based on shore-based observations. Missed
animals may include some individuals that were near
the surface and visible but were not seen, as well
as others that were diving and were too deep to be
seen.

We infer that some near-surface animals were
missed based on three reasons. First, apparent den-
sity decreased with increasing sea state and cloud
cover, hence near-surface animals must be missed
in (at least) the poorer conditions. (An alternative
explanation is that porpoise spend less time near the
surface when sighting conditions are poor.) Second,
of the two principal observers in 1985, the less ex-

perienced observer may have missed more porpoise.
Third, in five instances in 1984 and one instance in
1985, the data recorder saw harbor porpoise in the
inside swath that were missed by the observer. In
three of these cases, conditions were Beaufort 1 with
<25% cloud cover. Recorders searched only occa-
sionally as conditions permitted, so it is not possi-
ble to use these data to quantify how many near-
surface animals were missed under good conditions.

Based on behavioral observations, we can also be
certain that some harbor porpoise are missed be-
cause they are too deep to be seen. Water visibility
was typically only 2-5 m during the surveys. Aerial
observers have seen harbor porpoise dive out of view
during the passing of the plane or while circling.
Harbor porpoise were not visible from the helicopter
during dives, even in very calm water. Some frac-
tion of porpoise must be missed because they are
too deep to be seen.

We have tried to account for the fraction of diving
animals by dividing density estimates by the frac-
tion of time harbor porpoise are known to be near
the surface (i.e., within surfacing series). Uncor-
rected estimates of harbor porpoise density are
based on the assumption that porpoise are never too
deep to be seen; this undoubtedly results in an
underestimate of porpoise density. Our method of
adjusted estimates of porpoise density assumes that,
when diving, porpoise are always too deep to be
seen; this has been corroborated by helicopter ob-
servations. The latter estimate should therefore be
closer to the true value of porpoise density (if biases
due to other factors have been eliminated).

Offshore Distribution of
Harbor Porpoise

In 1985, aerial transects were flown at 0.61 and
1.85 km from the shore. The latter value was chosen
to correspond approximately to the ship transects
along the 10-fathom (18.3 m) isobath. The intent was
to directly compare aerial estimates of density to
estimates made from ships at the same distance
from shore (considered below) and to provide a
means to extrapolate ship estimates to regions that
were too shallow to survey by ship. In regard to the
latter, we wish to know whether harbor porpoise
density at 0.61 km from shore is different from that
at 1.85 km.

In 1985, surveys at 0.61 km were never flown
under good sighting conditions (Beaufort 0 & 1 and
<25% cloud cover), so direct comparisons between
0.61 and 1.85 km are not possible. Considering only
the best category of sighting conditions, it is possi-

442

BARLOW ET AL.: AERIAL SURVEYS OF HARBOR PORPOISE

ble to compare 1985 transects at 1.85 km with 1984
transects at 0.61 km in geographic regions 1 and
3 (Table 6). Sample sizes are very small, but den-
sities are roughly comparable (0.070 vs. 0.057,
respectively for 1.85 and 0.61 km from shore). Sam-
ple sizes in 1985 can be increased if Beaufort 2 is
considered instead of Beaufort 0 & 1 (still with <25%
cloud cover). Based on geographic regions 1 through
5, porpoise densities are virtually identical at 1.85
and 0.61 km from shore (0.188 vs. 0.192, respective-
ly, Table 7).

In a preliminary model of harbor porpoise depth
distribution. Barlow (1988) hypothesized that har-
bor porpoise density is constant from the shore to
the 75 m isobath. Ship data did not, however, in-
clude any transects inshore of 18.3 m depth. In this
shallow area, the model was not based on any data.
Although sparse, data from aerial surveys show that
densities are similar in areas that were surveyed by
ship (1.85 km from shore) and in areas that were
too shallow to be surveyed by ship (0.61 km from
shore). These data are consistent with the model pro-
posed by Barlow.

Comparison of Ship
and Aerial Density Estimates

Estimates of harbor porpoise density from 1984
to 1985 ship surveys (Barlow 1988) can be compared
with adjusted and unadjusted estimates from the
aerial surveys (Table 8). The overall density for all
regions is higher for the ship surveys than for the
unadjusted aerial. Adjusted estimates from the
aerial surveys are very close to the overall estimates
from ship surveys.

Previous comparisons have been made of sighting
efficiency from aerial and surface vessel platforms
(Kraus et al. 1983). They found that observers on
boats saw 52% of the harbor porpoise groups seen
by shore-based observers, whereas aerial observers
saw only 14%. Based on this, density from aerial
surveys might be expected to be only 27% of that
from ship surveys. In the present study, unadjusted
density based on aerial surveys is 32% of the den-
sity from ship surveys. The two studies are not
directly comparable, however. Weather conditions
are not reported by Kraus et al. (1983) and may have
included less than optimal sighting conditions. It
should be noted that the fraction of harbor porpoise
groups seen by aerial observers in their study (14%)
is even lower than the fraction of harbor porpoise
we assumed would be in surfacing series and hence
near the surface (23.9%). It is possible that
behavioral differences between harbor porpoise

from the two coasts (such as travelling or behavior
mode "A" noted by Watson and Gaskin (1983)) could
account for some of the differences noted above.
Estimates based on ship surveys cannot, of course,
be considered the true density of harbor porpoise.
The overall estimate based on ship surveys is rela-
tively imprecise (C.V. = 49%) and may be biased
(Barlow 1988). The ship survey estimate is, however,
superior to current estimates from aircraft for sev-
eral reasons. Line transect methods were used on
the ship surveys, and the principal assumption of
this method (that 100% of the animals in the imme-
diate vicinity of the trackline are seen) is more easily
met than the comparable strip transect assumption
(that 100% of the animals within a strip are seen).
Acceptable sighting conditions for ship surveys in-
cluded Beaufort 0 & 1 , and 2 and were not restricted
by cloud cover (Barlow 1988). This allowed more
complete geographic coverage than did aerial
surveys. Also, the ship travelled much slower than
the aircraft (10 knots vs. 80-90 knots), thus the prob-
ability of missing a diving individual was much less.
Barlow (1988) calculated that diving animals located
near the trackline would be missed by observers on
ships only if dive times exceeded 2 minutes. Final-
ly, estimates of the correction factor to account for
submerged animals is relatively imprecise. Addi-
tional observations on ventilation patterns may
allow further refinements in density estimates based
on aerial surveys.

RECOMMENDATIONS

The design of future surveys for harbor porpoise
could be improved based on the results obtained
from our aerial surveys. We found that sighting con-
ditions deteriorated rapidly with both increasing
cloud cover and rougher sea states. To the extent
that is possible, aerial surveys for harbor porpoise
should only occur on clear days with little wind. Ob-
servations made by the data recorders indicate that
some harbor porpoise will be missed even in good
sighting conditions. If strip transects are used, ex-
periments with two independent teams of observers
searching at the same time could be used to quan-
tify the fraction of animals that are missed by using
just one team. Given that fewer harbor porpoise
were seen between 400 and 500 m of the track line,
we also suggest that, when surveying at 213 m
altitude, the strip widths should be decreased to only
include the area between 100 and 400 m.

We believe, however, that the problem of miss-
ing harbor porpoise could be reduced if line transect
methods were used in place of strip transects. Line

443

FISHERY BULLETIN: VOL. 86, NO. 3

transect methods assume that 100% of all animals
are seen directly along the trackline and use
statistical techniques to estimate the number of
animals that are missed as a function of the distance
from this tracWine. Line transects would require use
of an aircraft with unobstructed downward visibil-
ity through a belly window and use of a third ob-
server who could view animals directly under the
aircraft.

Harbor porpoise are now known to occur further
from the shoreline than was believed at the begin-
ning of this study (Barlow 1988). In future surveys,
transect lines should be placed so as to cover a
greater fraction of the harbor porpoise habitat.

ACKNOWLEDGMENTS

Support, personnel, and flight time were provided
by the Southwest Fisheries Center (NMFS), the
Southwest Regional Office (NMFS), the California
Department of Fish and Game, the National Marine
Mammal Laboratory (NMFS), the Washington State
Department of Wildlife, and the Oregon Department
of Fish and Wildlife. The observers and data
recorders were R. Brown, S. Diamond, D. Hanan,
R. Holt, S. Jeffries, W. Ferryman, and B. Troutman
(also Jay Barlow, Terry D. Jackson, and Charles W.
Oliver). The pilots were R. Anthes, J. Cain, L. Heitz,
J. York, and B. Zorich. Shore observations were
made by M. Herter, T. LoughHn, D. Merrick,
R. Rowlett, D. Rugh, and B. Yerman (also Barbara
L. Taylor). Unpublished porpoise data from Alaska
were collected by P. Dawson (and Barbara L.

Taylor). The method of adjusting for submerged
animals was suggested by D. Chapman based on an
unpublished report by R. A. Davis. Drafts of this
paper were reviewed by R. Brown, D. DeMaster,
K. Forney, D. Hanan, R. Holt, S. Jeffries, and two
anonymous reviewers.

LITERATURE CITED

Barlow, J.

1988. Harbor porpoise Phocoena phocoena, abundance estima-
tion for California, Oregon and Washington; \. Ship
surveys. Fish. Bull., U.S. 86:417-432.

Kraus, S. D., J. R. Gilbert, and J. H. Prescott.

1983. A comparison of aerial, shipboard, and land-based
survey methodology for the harbor porpoise, Phocoena pho-
coena. Fish. Bull., U.S. 81:910-913.

Kruskal, W. H., and W. a. Wallis.

1952. Use of ranks in one-criterion analysis of variance. J.
Am. Stat. Assoc. 47:583-621.
Scheirer, C. J., W. S. Ray, and N. Hare.

1976. The analysis of ranked data derived from completely
randomized factorial designs. Biometrics 32:429-434.
Seber, G. A. F.

1973. The estimation of animal abundance and related param-
eters. Hafner Press, N.Y., 506 p.
Taylor, B. L., and P. K. Dawson.

1984. Seasonal changes in density and behavior of harbor por-
poise (Phocoena phocoena) affecting census methodology in
Glacier Bay National Park, Alaska. Rep. int. Whaling
Comm. 34:479-483.

Watson, A. P., and D. E. Gaskin.

1983. Observations on the ventilation cycle of the harbour
porpoise Phocoena phocoena (L.) in the coastal waters of the
Bay of Fundy. Can. J. Zool. 61(1):126-132.
Wilcoxon, F.

1945. Individual comparisons by ranking methods. Biom.
Bull. 1:80-83.

444

EVALUATION OF VARIABILITY IN

SABLEFISH, ANOPLOPOMA FIMBRIA, ABUNDANCE INDICES IN

THE GULF OF ALASKA USING THE BOOTSTRAP METHOD

Michael F. Sigler and Jeffrey T. Fujioka^

ABSTRACT

Relative population numbers (RPN's) and length compositions were computed for sablefish, Anoplopoma
fimbria, in the Gulf of Alaska from the results of the Japan-United States cooperative longline survey
from 1979 to 1986. A statistical evaluation of annual changes in the RPN's using the bootstrap method
is demonstrated and showed that sablefish abundance increased significantly from 1979 to 1986, an in-
crease likely due to recruitment of two strong year classes. The effect of missing data on the bootstrap
calculations was examined and found to be negligible.

Early in this century, United States and Canadian
fishermen began harvesting sablefish, Anoplopoma
fimbria, in nearshore waters from California north
to southeastern Alaska, but sablefish were not
heavily exploited until Japanese longline vessels
began fishing in the Bering Sea in 1958. Japanese
catches off both the U.S. and Canadian coasts rose
dramatically in the following two decades. After
passage of the Fishery Conservation and Manage-
ment Act, foreign catches were reduced and the
domestic allocation was increased. The domestic
catch eventually increased nearly fivefold, thereby
replacing the foreign fishery, and in 1985, sablefish
in the Gulf of Alaska were harvested entirely by
domestic fishermen.

Before the reduction of the foreign fishery, infor-
mation on sablefish abundance consisted of statistics
on catch per unit effort from the Japanese longline
fishery. In 1978, the Fisheries Agency of Japan and
the U.S. National Marine Fisheries Service (NMFS)
began a cooperative longline survey along the con-
tinental slope of Alaska to assess the abundance of
sablefish and Pacific cod, Gadus macrocephalus. The
survey, conducted annually, has provided eight con-
secutive years (1979-86) of data for the Gulf of
Alaska, seven years (1980-86) of data for the Aleu-
tian region, and five years (1982-86) of data for the
eastern Bering Sea. The first year of the survey,
1978, was experimental.

Relative population numbers (RPN's) and length
compositions from the longline survey results from

'Northwest and Alaska Fisheries Center Auke Bay Laboratory,
National Marine Fisheries Service, NOAA, P.O. Box 210155, Auke
Bay, AK 99821.

1979 to 1985 have been estimated previously by
Sasaki (1986). In this study, the observed increase
in the RPN's for sablefish in the Gulf of Alaska is
evaluated statistically and an explanation of the
probable source of the increase is discussed.

Statistical analysis of the survey results is based
on the bootstrap method (Efron 1982; Efron and
Gong 1983). This method is a relatively new statis-
tical procedure that has been little used in fisheries
analysis. Thus, this paper also demonstrates an
application of the bootstrap method to statistical
evaluation of fishery survey data.

SURVEY METHODS

The Gulf of Alaska portion of the Japan-U.S.
cooperative longline survey, conducted annually
each summer from 1979 to 1986, covered five
International North Pacific Fisheries Commission
(INPFC) statistical areas: Shumagin, Chirikof,
Kodiak, Yakutat, and Southeastern (Fig. 1). One of
a total of 47 stations each ranging in depth from
about 100 to 1,000 m was sampled daily by longline.
The longline was 16 km long and consisted of 160
hachis (the Japanese word for "skate" or length of
longline); each hachi was 100 m long and consisted
of 45 hooks baited with squid. Soak time, the time
between setting and retrieval, varied from 3 hours
at the beginning of the longline gear to 7 or 8 hours
at its end. The depth at which the fish were caught
was estimated by measuring the depth of water
under the vessel with an echo sounder every fifth
hachi. The fish caught were tallied by species and
hachi as the longline was brought aboard, then they
were weighed and their length was measured. Most

Manuscript accepted May 1988.

FISHERY BULLETIN: VOL. 86, NO. 3, 1988.

445

FISHERY BULLETIN: VOL. 86, NO. 3

Figure 1.— International North Pacific Fisheries Commission statistical areas sampled during the Gulf of Alaska
portion of the Japan-U.S. cooperative longline survey, 1979-86.

sablefish were sexed, some were tagged and re-
leased, and others were sampled for otoliths and
scales. Detailed survey methods are described in
Sasaki et al. (1983).

STATISTICAL METHODS

The catch data were stratified because of differ-
ences in the catch rate by depth. Assignment of the

150 -

z
o

100-

0.5

5.5

10.5 15.5

CATCH PER HACHI

20.5

25.5

446

Figure 2.— Frequency distribution of sablefish catch per hachi (numbers) by year, station, and strata,

1979-86.

SIGLER and FUJIOKA: VARIABILITY IN SABLEFISH

catch data to a stratum was based on the recorded
depth of every fifth hachi or the interpolated depth
of each of the intervening four hachis. The number
of strata totaled nine, with the first stratum repre-
senting 101-200 m, the second representing 201-
300 m, and so on. A catch per hachi value was
calculated for each stratum of a station (Fig. 2), and
the resultant values within each statistical area were
averaged.

Not all strata were sampled at each station and
as a result, sometimes no stations were sampled
within certain strata (Table 1). Although not sam-
pled one year, these strata generally were sampled
the previous year, because of slight annual differ-
ences in the depths sampled at each station. Catch
per hachi values CPUE,j were estimated for sta-
tions within the unsampled strata from the catch per
hachi values for sampled stations from the previous
year and the annual change in the catch per hachi
values in the adjacent stratum. For each station:

CPUE, J = CPUE,_i J * CPUE ,,j /CPUEi_i__,.

where i = yed.r,j = stratum, a.ndj' = the adjacent
stratum. Catch per hachi values were estimated only
for stations within unsampled strata. If one or more
stations within a stratum were sampled, then no
catch per hachi values were estimated for the un-
sampled stations. The effect of these missing values
on the survey results vdll be discussed later.

Relative population numbers were calculated from
the catch per hachi values to index annual changes
in sablefish abundance. As in Sasaki (1985), the aver-
age catch per hachi was multiplied by the areal size
of each stratum and statistical area to calculate an
RPN for each stratum and statistical area. The areal
sizes used to calculate the RPN's are for the con-
tinental slope only (Table 2), which corresponds to
the area covered by the survey. The resultant RPN's
were summed across strata to calculate an RPN for
each statistical area, and these RPN's were summed

Table 1.— Number of stations sampled by year, INPFC area, and depth during the Japan-U.S. cooperative longline survey, 1979-86.

Depth

INPFC Statistical area

Depth

INPFC statistical area

Shu-

Chiri-

South-

Shu-

Chiri-

South-

Year

(m)

magin

kof

Kodiak

Yakutat

east

Year

(m)

magin

kof

Kodiak

Yakutat

east

1979

101-200

1983

101-200

201-300

301-400

401-500

501-600

601-700

701-800

801-900

901-1,000

1980

101-200

1984

101-200

201-300

301-400

401-500

501-600

601-700

701-800

801-900

901-1,000

1981

101-200

1985

101-200

201-300

301-400

401-500

501-600

601-700

701-800

801-900

901-1,000

1982

101-200

1986

101-200

201-300

301-400

401-500

501-600

601-700

701-800

801-900

901-1,000

447

FISHERY BULLETIN: VOL. 86, NO. 3

Table 2.— Slope area sizes (km^.

2\1

Depth
(m)

Slope area (km'')

Shumagin Chlrikof Kodiak Yakutat Southeast

201-400

4,001

2,350

3,106

2,988

1,781

401-

-600

2,269

1,766

2,255

1,666

822

601-

-800

1,629

1,955

1,923

1,470

1,006

801-

-1,000

1,248

2,012

2,296

1,489

1,165

201-

-1.000

9,147

8,083

9,580

7,613

4,774

'Shumagin, Chirikof, and Kodiak areas and Yakutat area from long.
147-I440W; data from E. Brown (NWAFC Seattle Laboratory, NMFS. NOAA,
7600 Sand Point Way NE. Seattle, WA 98115, pers. commun. December
1985). Yakutat area from long. 144-137°W. and Souttieastern area; data from
R. Haight (NWAFC Auke Bay Laboratory, NMFS, NOAA, P.O. Box 210155,
Auke Bay, AK 99821, pers. commun. 1986).

across statistical areas to calculate an RPN for the
Gulf of Alaska. The bootstrap method (Efron 1982;
Efron and Gong 1983) then was applied to the re-
sultant RPN's to test the statistical significance of
annual changes in the RPN's.

The bootstrap method is a nonparametric statis-
tical procedure based on Monte Carlo methods (see
Shreider [1966] for a description of Monte Carlo
methods). The bootstrap method is a new technique
not common in the fisheries and ecology literature,
but examples of its application to survey design and
biomass estimation can be found in Kimura and
Balsiger (1985) and Haslett and Wear (1985), re-
spectively. In addition, Rao and Wu (1984) proved
the applicability of the bootstrap method to strati-
fied sampling, which is the sampling method used
in the longline survey. The bootstrap method is
useful when parametric assumptions are difficult to
justify; no parametric estimate is readily available
for the accuracy of a statistic, e.g., a sample median;
or the procedure to compute the statistic of interest
is complicated. Simply described, the bootstrap
method works as follows: Given the observed data
set <Xi,X2,. . .,X^>, the sample <Xi*,X2*,. . .,
Xn*> is drawn by independent random sampling
with replacement from the observed data set and
the desired statistic (e.g., a median) is computed
from the sample. The resultant statistic is termed
the bootstrap replicate. In the next step, the sam-
ple is drawn and the bootstrap replicate is computed
some large number B times. The resultant B boot-
strap replicates form the bootstrap distribution. An
estimate of the accuracy of the median, a standard
error, then can be calculated from the bootstrap
distribution by standard methods.

In this study, we used the bootstrap method to test
the null hypothesis that the difference RPN^j;. -
RPN^jr. = 0, for any RPNjjt , where i = year
(1979-86), i' = any subsequent year, and k = sta-

tistical area (Shumagin, Chirikof, Kodiak, Yakutat,
and Southeastern). The bootstrap method was used
because parametric assumptions are difficult to
justify for the longline survey data and the pro-
cedure to compute the statistic of interest, the
variance of the RPN, is tedious and error prone. In
our application of the bootstrap method, stations
were randomly sampled with replacement within
each area. A value denoted RPN^ ;^* was computed
from the catch per hachi values of the sampled sta-
tions by the method of RPN calculation described
previously. Stations then were sampled with re-
placement from year i' within each area, a second
value denoted RPNj ;^* was computed, and the
difference RPNj j^* - RPN^j^* was found. Sampl-
ing with replacement from the 2 years and the com-
putation of the difference were repeated 1,000 times
producing a bootstrap distribution of 1,000
differences.

Efron and Tibshirani (1986) outlined three
methods for setting an approximate confidence
interval from a bootstrap distribution for a statistic
of interest, here the difference RPNj /^ - RPN^ /£.
Use of the simplest method, the percentile method,
is considered correct if the bootstrap distribution of
the statistic of interest is described by a normal
distribution (Efron and Tibshirani 1986). The nor-
mality of the bootstrap distribution for the dif-
ference was tested using the D'Agostino D Test
(D' Agostino and Stephens 1986) and found to be nor-
mal, thus justifying the use of the percentile method.
The statistical significance of the difference
RPNj /£ - RPNj /£ then was evaluated by the
following criteria. If the 95% confidence interval for
the difference did not include zero, then the null
hypothesis was rejected, the annual change in the
RPN was considered statistically significant, and the
change in sablefish abundance was considered real.
Conversely, if the 95% confidence interval for the
difference included zero, the null hypothesis was ac-
cepted and the change in sablefish abundance was
not considered significant.

RESULTS

The RPN for the Gulf of Alaska increased 111%
from 1979 to 1986 (Fig. 3). The 95% confidence
interval for this increase did not include zero and
therefore was judged statistically significant (alpha
= 0.05; Table 3), showing that the difference was
not due to random error in the survey and that
sablefish abundance in the Gulf of Alaska has in-
creased markedly since 1979. The difference con-
sists of significant increases from 1980 to 1981, 1981

448

- KODIAK

/
/

I 1

1 1 1

30
1979 80

81 82 83 84

85 86 1979 80 81 82 83
YEAR

84 85 86

Figure 3.— Relative population number (RPN) for the Gulf of Alaska and the Shumagin, Chirikof, Kodiak, Yakutat,
and Southeastern areas, 1979-86. Dashed lines (— -) signify that the annual change was statistically significant.

to 1982, and 1984 to 1985 (Fig. 3). Differences be-
tween other years were not significant.

The RPN for each of the statistical areas of the
Gulf of Alaska generally increased from 1979 to
1986 (Fig. 3); the differences between 1979 and 1986
were statistically significant for all areas (Table 3),
showing that sablefish abundance has increased
throughout the Gulf of Alaska. The sharp jump in
the RPN for the Gulf of Alaska from 1980 to 1982
was caused by significant increases in four areas,
Shumagin, Chirikof, Kodiak, and Yakutat. The
sharp jump from 1984 to 1985 was caused by signif-
icant increases in two areas, Kodiak and Yakutat.

DISCUSSION

As noted earlier, sometimes not all strata were
sampled at a station (Table 1). This shortcoming was
a consequence of the length of the sampling gear
and the topography at a station. At stations where
the bottom gradient was slight for all or part of the
station, the 16 km of longline gear sometimes was
not long enough to sample all strata. As a result,
sometimes no stations were sampled within certain
strata. For these strata, catch per hachi values were
estimated using combined data from the current and
previous year's data. The resultant values are esti-

449

FISHERY BULLETIN: VOL. 86, NO. 3

Table 3.— Statistical significance of annual changes in
relative population number of sablefish, Gulf of Alaska,
Japan-U.S. cooperative longline survey, 1979-86. The sym-
bols used are defined as follows; + signifies a significant
increase in RPN; - signifies a significant decrease in
RPN; O signifies no significant change.

Area

Year

Gulf of Alaska

1979

1980

1981

1982

1983

1984

1985

Shumagin

1979

1980

1981

1982

1983

1984

1985

Chirikof

1979

1980

1981

1982

1983

1984

1985

Kodiak

1979

1980

1981

1982

1983

1984

1985

Yakutat

1979

1980

1981

1982

1983

1984

1985

Southeastern

1979

1980

1981

1982

1983

1984

1985

mates of catch per hachi as if the strata had been
sampled. Because the estimated values mimic the
observed data, the probability of rejecting the null
hypothesis is somewhat higher than the nominal
value. In general an inaccurate estimate should have
a small total effect because catch per hachi values
were missing for only 9 of the 320 strata available
during the survey (8 years x 5 areas x 9 strata).
The only comparison where the estimated catch per
hachi values might have caused an incorrect conclu-
sion is the 1984-85 comparison where 7 strata were
unsampled. This comparison was retested using only

the observed catch per hachi values. The annual
change, like the comparison using the estimated
catch per hachi values, was found to be statistically
significant.

Another potential effect of incomplete sampling
at a station is bias in the bootstrap calculation. Each
bootstrap replicate RPN^j^* was computed from a
bootstrap sample selected by sampling with replace-
ment from the stations within an area. If one or
more stations were missing an observation for a
strata, then it was possible for the bootstrap sam-
ple to have all missing values for the strata. In this
case, no catch per hachi values were available to
calculate a value for the strata and the value for the
strata was treated as zero. This treatment may
negatively bias the resultant RPN, ;,*. The extent
of the negative bias was tested by comparing the
RPN to a bootstrap estimate of the RPN, denoted
RPNfc, where RPN^ is the mean of the 1,000
RPNijt* from an area and year. If RPN^ generally
was smaller than RPN, then the bootstrap replicates
were negatively biased. The percentage difference
between the resultant RPNj's compared to the
RPN's ranged from - 0.7 to + 0.5 and averaged only
-1-0.03, showing that the negative bias had little
effect on the RPN,^*.

Other fishery and survey data substantiate the
significance of the overall increase in the RPN and
also the marked increases from 1981 to 1982 and
from 1984 to 1985. The CPUE in the Japanese long-
line fishery showed a similar pattern to the RPN's
for the years the data overlapped, from 1979 to
1983; the fishery CPUE increased from the late
1970's, with the largest increase from 1981 to 1982,
and decreased from 1982 to 1983 (Fig. 4).

Examination of length compositions for depths
101-200 m indicate that the strong 1977 year class
(Sasaki 1982; McFarlane and Beamish 1983; Funk
and Bracken 1984) was responsible for the RPN in-
crease from 1980 to 1982 and that the RPN increase
from 1984 to 1985 was due to a strong 1980 year
class. Sablefish recruiting to the survey area first
appear at depths 101-200 m, and strong year classes
are more distinguishable in the length compositions
at these depths. The length compositions for depths
101-200 m show the initial appearance and subse-
quent increase in length of fish of the strong 1977
year class (Fig. 5). The mode at 47 cm FL in 1979
indicates the first year that the 1977 year class was
available to the survey gear. The rightward move-
ment of the mode in succeeding years illustrates the
increase in fish length for the 1977 year class. The
movement of the mode slowed in 1982 which we in-
terpret to be due mainly to the movement of larger

450

SIGLER and FUJIOKA: VARIABILITY IN SABLEFISH

460
420

■o 380

§340

300
260
220

180

- \

FISHERY CPUE /

v..

1977 78

Figure 4.— Sablefish catch per hachi (kg/10 hachis) for the Japanese longline fishery, 1977-83 (Fujioka
1986) and relative population number (RPN), 1979-86.

Figure 5.— Length-frequency distributions of sablefish
in the Gulf of Alaska for depths 101-200 m, 1979-86.
The shaded bands illustrate the progression of the modes
of two strong year classes, 1977 and 1980.

T r

55 63

LENGTH (cm)

T
79

451

FISHERY BULLETIN: VOL. 86, NO. 3

fish to depths >200 m. The timing of this movement
corresponds to the year that the RPN significantly
increased.

The pattern of length compositions from 1983 to
1985 parallels those due to the strong 1977 year
class and suggests that the 1980 year class is also
strong (Fig. 5). The shoulder at 49-51 cm FL in
1983, not present in 1982, indicates the first year
that the 1980 year class was available to the survey
gear. This shoulder is due to distinct modes at 47-51
cm FL in the Chirikof, Yakutat, and Southeastern
areas. The rightward movement of the mode in suc-
ceeding years illustrates the increase in fish length
for the 1980 year class. The mode at 53 cm FL in
1984 is similar to the modes at 50-57 cm FL found
in the Gulf of Alaska for depths 101-200 m during
a trawl survey conducted in 1984 by the Northwest
and Alaska Fisheries Center (Brown 1986) and is
further evidence for a strong 1980 year class. The
rightward movement of the mode slowed in 1985,
presumably due to movement of larger fish to depths
>200 m, and again corresponds to the year of a
significant increase in RPN.

ACKNOWLEDGMENTS

Thanks to Evan Haynes and Leslie Williams for
thoroughly editing the manuscript, Jerry Fella for
statistical advice, and Takashi Sasaki for courteously
providing cooperative survey data.

LITERATURE CITED

Brown, E.

1986. Preliminary results of the 1984 U.S. -Japan cooperative
bottom trawl survey of the central and western Gulf of
Alaska. In R. L. Major (editor), Condition of groundfish
resources of the Gulf of Alaska region as assessed in 1984,
p. 259-296. U.S. Dep. Commer., NOAA Tech. Memo.
NMFS F/NWC-106.
D'Agostino, R. B., and M. A. Stephens.

1986. Goodness-of-fit techniques. Marcel Dekker, N.Y.
Efron, B.

1982. The jackknLfe, the bootstrap and other resampling

plans. Appl. Math. 38, 92 p. Soc. Ind. Appl. Math., Phila.
Efron, B., and G. Gong.

1983. A leisurely look at the bootstrap, the jackknife, and
cross-validation. Am. Stat. 37:36-48.

Efron, B., and R. Tibshirani.

1986. Bootstrap methods for standard errors, confidence in-
tervals, and other measures of statistical accuracy. Stat.
Sci. 1:54-77.

FUJIOKA, J.

1986. Sablefish. In R. L. Major (editor). Condition of
groundfish resources of the Gulf of Alaska region as assessed
in 1985, p. 79-108. U.S. Dep. Commer., NOAA Tech.
Memo. NMFS F/NWC-106.

Funk, F., and B. E. Bracken.

1984. Status of the Gulf of Alaska sablefish (Ancyplopoma fim-
bria) resource in 1983. Alaska Dep. Fish Game, Info. Leafl.
235, 55 p.

Haslett, S. J., AND R. G. Wear.

1985. Biomass estimation of Artemia at Lake Grassmere,
Marlborough, New Zealand. Aust. J. Mar. Freshwater Res.
36:537-557.

KiMURA, D. K., AND J. W. BALSIGER.

1985. Bootstrap methods for evaluating sablefish pot index
surveys. N. Am. J. Fish. Manage. 5:47-56.
McFarlane, G. a., AND R. J. Beamish.

1983. Overview of the fishery and management strategy for
sablefish (Ancrpkrpoma fimbria) off the west coast of Canada.
In Proceedings of the International Sablefish Symposium,
p. 13-35. Alaska Sea Grant Rep. 83-8, Univ. Alaska,
Fairbanks.

Rao, J. N. K., and C. F. J. Wu.

1984. Bootstrap inference for sample surveys, /n J. E.
Gentle (editor). Proceedings of the section on survey
research methods, p. 106-112. Am. Stat. Assoc.

Sasaki, T.

1982. Condition of sablefish stocks in the North Pacific. Far
Seas Fish. Lab. Fish. Agency Jpn., Shimizu, 17 p. (Document
submitted to the International North Pacific Fisheries
Commission.)

1985. Studies on the sablefish resources in the North Pacific
Ocean. Far Seas Fish. Res. Lab., Bull. 22, 107 p. Fisheries
Agency of Japan, Shimizu.

1986. Stock assessment of sablefish in the eastern Bering
Sea, Aleutian Islands region, and the Gulf of Alaska. Far
Seas Fish. Res. Lab., Fish. Agency Jpn., Shimizu, 33 p.

Sasaki, T., D. Rodman, and K. Funato.

1983. Preliminary report on Japan-U.S. joint longline survey
by Ryusho maru No. 15 in the eastern Bering Sea, Aleutian
region and Gulf of Alaska, 1982. Far Seas Fish. Res. Lab.,
Fish. Agency Jpn., Shimizu, 116 p.

Shreider, Y. a.

1966. The Monte Carlo method. Pergammon Press, N.Y.

452

PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD, PSEUDOPENTACEROS
WHEELERI, AND RECRUITMENT TO ISOLATED SEAMOUNTS

IN THE NORTH PACIFIC OCEAN

George W. Boehlerti and Takashi Sasaki^

ABSTRACT

The pelagic armorhead, Pseudopentaceros wheeleri, occurs widely in the North Pacific Ocean. Benthic
specimens have been taken from Japan, the Hawaiian Archipelago, and the west coast of North America,
but the main reproductive populations are located on southern Emperor-northern Hawaiian Ridge sea-
mounts between lat. 29° and 35°N. The period between spawning and recruitment to the seamounts
is apparently between 1.5 and 2.5 years, suggesting an extended pelagic existence. We describe the
distributional patterns in the North Pacific based upon over 30 years of published and unpublished records.
The majority of pelagic specimens are captured in the subarctic water mass in the northeast Pacific.
Based upon the distributional patterns and the oceanography of the North Pacific, we propose migratory
paths for both the main population and for the individuals that occur rarely in other locations. The long
pelagic period and variability in ocean conditions may play an important role in recruitment to seamounts
and the variability in year-class strength for this species.

The pelagic armorhead, Pseudopentaceros wheeleri,
is a member of the boarfish family Pentacerotidae.
Until recently, considerable confusion existed about
its taxonomy and distribution. Originally described
as Pentaceros richardsoni Smith 1844, it was
thought to be distributed virtually worldwide, with
centers of abundance in the North Pacific and in the
South Atlantic near South Africa (Fujii 1986). Oc-
currences were typically sporadic; most records
were from pelagic captures, hence the name pelagic
armorhead. Several authors, including Welander et
al. (1957), Clemens and Wilby (1961), and Wagner
and Bond (1961), used the name Pseudopentaceros
richardsoni, but more recent studies used the
original binomial following Follett and Dempster
(1963).

The 1967 discovery of large concentrations of
armorhead in the mid-Pacific, over southern
Emperor-northern Hawaiian Ridge (SE-NHR)
seamounts (from lat. 29° to 35°N) (Sasaki 1974;
Uchida and Tagami 1984) stimulated increased
interest in this species. Over the next 10 years,
nearly 1 million metric tons (t) were taken by Soviet
and Japanese trawlers (Borets 1975; Takahashi and
Sasaki 1977). The increased availability of specimens
allowed Hardy (1983) to revise the family. He first

'Southwest Fisheries Center Honolulu Laboratory, National
Marine Fisheries Service, NOAA, 2570 Dole Street, Honolulu, HI
96822-2396.

^Far Seas Fisheries Research Laboratory, 5-7-1 Orido, Shimizu,
Shizuoka 424, Japan.

separated the Pentaceros richardsoni complex
from other members of the genus by referring it to
Pseudopentaceros Bleeker, based on morphological
considerations. Second, he separated the complex
into three species: P. richardsoni confined to the
Southern Hemisphere, and P. wheeleri and P.
pectoralis in the North Pacific. The features used
to distinguish P. wheeleri and P. pectoralis were
largely morphological. Pseudopentaceros pectoralis
is more robust and deeper bodied and its dis-
tribution is typically pelagic; P. wheeleri was
known only from benthic specimens on sea-
mounts. Hardy (1983), however, lacked transitional
specimens; subsequent work based on morpho-
logical and electrophoretic grounds revealed that
P. wheeleri was a morphological derivative of
P. pectoralis, with changes occurring after settle-
ment. The deeper bodied pelagic form ceases growth
in length after recruiting to the seamounts and
instead transforms body shape to a leaner form.
"Fat" and "lean" specimens of the same length thus
differ significantly in appearance (Humphreys et al.
in press).

Pseudopentaceros wheeleri has an unusual life
history. Spawning occurs at the benthic population
centers at the SE-NHR seamounts between Novem-
ber and March (Sasaki 1974; Borets 1976; Bilim et
al. 1978). Larvae and juveniles are pelagic (Honma
and Mizusawa 1969; Fedosova and Komrokov 1975;
Borets and Sokolovsky 1978) and widely distributed
in the North Pacific. Recruitment to the seamounts

Manuscript accepted April 1988.
Fishery Bulletin; Vol. 86, No. 3, 1988.

453

FISHERY BULLETIN: VOL. 86, NO. 3

occurs between ages 1.5 and 2.5 years, but in the
major spawning population centers, over 95% of the
fish are <2 years (Uchiyama and Sampaga^). This
lack of older specimens and presence of fish in poor
physiological condition has led to suggestions that
this species is semelparous, with death after a single
spawning (Humphreys and Tagami 1986; Hum-
phreys et al. in press). The long interval between
spawning and recruitment for this species suggests
that oceanographic conditions in the North Pacific
regulate its distribution (Boehlert 1986).

Published records of the pelagic distribution of
P. wheeleri are too few and varied to understand its
biogeography. In different years, specimens have
been caught in the Gulf of Alaska, near the Aleu-
tian Islands, and off the coasts of Japan, Oregon,
California, and British Columbia. In this paper, we
document the pelagic and benthic occurrences from

'Uchiyama, J. H., and J. Sampaga. In review. Age and growth
of the pelagic armorhead, Pseudopentaceros wheeleri, at Hancock
Seamounts. Southwest Fisheries Center Honolulu Laboratory,
National Marine Fisheries Service, NOAA, 2570 Dole Street,
Honolulu, HI 96822-2396.

both published and unpublished accounts, consider
this distributional pattern in light of general ocean-
ographic conditions in the North Pacific, propose
hypotheses about the migration and recruitment of
armorhead to the seamounts of the central North
Pacific, and assess interannual variation in abun-
dance and how it may relate to oceanographic
variability.

MATERIALS

The data sources used include published and un-
published collection records from throughout the
North Pacific Ocean. Published records on occur-
rences in Japan (Abe 1957, 1969; Zama et al. 1977;
Okamura et al. 1982), the eastern North Pacific
(Welander et al. 1957; Larkins 1964; Honma and
Mizusawa 1969; Chikuni 1970; Ignell et al. 1986),
and the west coast of North America (Wagner and
Bond 1961; Follett and Dempster 1963; Smith 1965)
were typically of single or a few specimens. Larger
numbers of unpublished collection records were in
data from large-scale surveys (Table 1). These data

Table 1 .—Data bases searched for pelagic armorhead, Pseudopentaceros wheeleri, captures excluding single
collections and literature reports. Areas and seasons are averages over the years covered.

Years, season effective

Area

Data source

Lat. N

Long.

Salmon drift gill-net surveys^

1972-86, spring-fall

38°-65<'

140°E-166°W

University of Washington (FRI)

purse seining, longlining^
Canada salmon longlining^

1956-82, spring-summer

42°-60°

165°E-125°W

1962-67, spring-summer

42°-54°

120°-145°W

Bureau of Commercial Fisheries

salmon gill net^'^

1955-71, spring-summer

40°-60°

165'»E-125°W

Oshoro Maru salmon

1963-66, 1968, 1969, 1971,

gill-net-longline surveys'*

1974-86, spring-summer

36°-66°

163°E-124°W

Hokusei Maru squid and salmon

gill-net surveys'*

1973-86, summer

18°-58°

141°E-156°W

Japan Marine Fishery Resource

Research Center pomfret surveys*

1979-82, all months

22°-49°

143°E-124°W

Japanese commercial whaling reports*

1952-79, summer

40°-58°

Pacific Biological Station

1983, 1985-86,

gill-net surveys'^

spring-summer

47°-54°

138«'-130<'W

Taiwan Fisheries Research Institute

squid surveys®

1986, July-August

38°-46°

153°E-175°W

National Marine Fisheries Service

Auke Bay cooperative

squid gill-net surveys^

1985-86

43°-55°

145°-175°W

Japan Fisheries Agency

squid gill-net surveys

1986-87, July-August

36°-47°

150°-165°W

'Salmon research data file. North Pacific Ocean, 1972-86. Salmon Division, Far Seas Fisheries Research Latxjratory, Shimizu,
Japan.

2t^acy et al 1978

sLarklns 1964.

♦Data record of oceanographic observations and exploratory fishing. No. 1 (1957)-No. 28 (1985) Faculty of Fisheries, Hokkaido
University.

sjapan Marine Fishery Resourse Research Center 1980, 1983a, 1983b, 1985

^Biological data file of whales In the North Pacific Ocean, 1952-79. Whale Section, Far Seas Fisheries Research Laboratory,
Shimizu. Japan. See also ChikunI (1970).

'Sloan 1983; Robinson and Jamieson 1984.

'Z. Shyu, Taiwan Fisheries Research Institute, Keelung 20220, Taiwan, pers. commun. May 1987.

^Ignell et al. 1986.

454

BOEHLERT and SASAKI: PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD

are seasonally limited to primarily spring, summer,
and early fall months, but cover 1952-87 and pro-
vide broad geographic coverage between lat. 18°
and 66 °N. Greater detail and a summary of many
of these surveys are provided by Macy et al. (1978).
Since armorhead were an incidental catch, it is dif-
ficult to assess the relative value of the different
surveys with respect to sampling effort.

The most typical sampling gears used were long-
lines, gill nets, and purse seines targeting salmon,
squid, or pomfret; rarer collections were made with
plankton nets, dip nets, or hook and line. Thus,
sampling was typically in surface waters, most likely
in the upper 50 m; the true depth distribution of the
pelagic animals, however, is unknown. The only
other "gear" was the sei whale, Balaenoptera bore-
alis, from which stomach contents were studied for
about 27 years; Chikuni (1970) reported armorhead
in the stomachs.

RESULTS AND DISCUSSION

Distributional Patterns

The known center of distribution and spawning
for benthic P. wheeleri is in the SE-NHR seamount

region, bounded by lat. 29°-35°N and long. 171°E-
179°W (Takahashi and Sasaki 1977; Humphreys and
Tagami 1986). These reproductive fish are typical-
ly between 23.0 and 28.5 standard length (SL)
(Sasaki 1986) and range in age from 1.5 to 2.5 years
(Uchiyama and Sampaga fn. 3). They are found at
depths between 200 and 500 m (Takahashi and
Sasaki 1977). Other benthic occurrences have been
recorded but are sporadic (Fig. 1). Several speci-
mens were captured off Japan in 1957 (Abe 1957),
1969, 1971-73, 1976 (Zama et al. 1977), and 1979
(Okamura et al. 1982). These fish were typically the
same size as those captured on the seamounts, but
no mention of reproductive status was made. Fish
of similar length occur rarely on the west coast of
North America, but only between lat. 37°00' and
44°25'N (Wagner and Bond 1961; Follett and Demp-
ster 1963; Smith 1965). Again, reproductive status
is unknown for these specimens. No benthic occur-
rences off British Columbia or Alaska have been
reported, despite extensive trawling surveys which
included seamounts (Hughes 1981; Alton 1986).
Large specimens have been taken rarely in the
Hawaiian Archipelago (Fig. 1; Randall 1980;
Humphreys et al. in press) and are typically 4-5
years old and appear to be in reproductive condi-

120'

150

Figure 1.— Known distribution of benthic occurrences of pelagic armorhead, Pseudopentaceros wheeleri. Squares represent the main
population centers at central North Pacific seamounts. All others are rare or sporadic occurrences.

455

FISHERY BULLETIN: VOL. 86, NO. 3

tion. Their relationship with the SE-NHR seamount
populations is unknown.

The most common pelagic occurrences (1955-87)
were fish ranging from 18 to 26 cm SL; smaller sizes
are poorly represented, probably because of the
nature of the sampling gear. Smaller specimens
were generally captured with plankton nets (Fedo-
sova and Komrakov 1975; Borets and Sokolovsky
1978), dip nets (Honma and Mizusawa 1969; Ran-
dall 1980; Fujii'*), and in one case, in a whale stomach
(Kawamura 1982). The sea-surface temperatures at
which the specimens were captured ranged from
8.6° to 15.0°C (average 12.1°C). For all years,
pelagic captures were restricted to the eastern
North Pacific, with the exceptions of larvae or early

juveniles captured near the spawning centers at the
seamounts (Komrakov 1970; Table 2), over 20
specimens in a Bryde's whale stomach in 1979
(Kato^, Fig. 2B), and two individuals captured in
1986 at long. 155°E (Fig. 2). This is despite exten-
sive sampling in the western North Pacific, par-
ticularly by the Japan Marine Fishery Resource
Research Center (JAMARC) and the Hokkaido
University, and in the Bering or Okhotsk Seas (Table

1).
Pelagic captures in the 1950's were sporadic, but

several fish were taken north of lat. 50°N, especially

in 1958 (Fig. 2 A). Relatively few were captured in

the 1960's, with the exception of July through early

August 1969, when the armorhead was an impor-

*E. Fujii, Nippon Luther Shingaku Daigaku, Tokyo, Japan, pers.
commun. July 1987.

^H. Kato, Whales Research Institute, Tokyo, Japan, pers.
commun. December 1987.

Table 2. — Pelagic specimens of armorhead, Pseudopentaceros wheeleri, for which sizes were
available. The growth curve from Uchiyama and Sampaga (text footnote 3), based upon
enumeration of daily growth increments, was used to deduce age from length. For the youngest
fish (bottom), average readings from daily growth increments, rather than the growth curve,
were used. Data are from various published sources and collections listed in Table 1 . (SST
= sea-surface temperature.)

Standard

Lat.

length

Year

Date

Long.

SST

(mm)

Age (yr)

No.

Gear

1955

9/27

45°41'

165°5'W

108-122

0.38-0.42

Gill net

1955

9/27

45°35'

165°5'W

—

200

0.79

Dip net

1956

7/31

45°49'

160°3'W

12.2

245

1.29

Gill net

1956

9/15

45°00'

155°00'W

12.1

245

1.29

Gill net

1956

9/17

49°00'

150°00'W

13.2

245

1.29

Gill net

1956

9/17

51°00'

150°00'W

11.0

245

1.29

Gill net

1958

8/14

44°39'

174°48'W

—

64-91

0.26-0.33

Dip net

1958

8/25

49°43'

146°10'W

11.8

256

1.50

Gill net

1963

8/24

49°0'

162°0W

12.9

249

1.40

Gill net

1967

6/15

54°38'

150°1'W

8.8

121-134

0.42-0.46

Dip net

1968

41°30'

165°30'W

—

285

>2.00

Handline

1969

4/1

29°48'

179°5'E

—

8-23

0.08-0.25

Plankton

1969

4/1

31 °6'

176°0'E

—

8-23

0.08-0.25

—

Plankton

1969

4/1

32°6'

173°0'E

—

8-23

0.08-0.25

—

Plankton

1972

5/21

38°30'

175°3W

14.3

42-50

0.21-0.23

Whale stomach

1984

7/25

45°30'

155°0'W

13.0

262-265

1.81-2.02

Gill net

1984

7/25

45°26'

154°58'W

12.0

278

>2.00

Longline

1985

7/9

46°57'

129°4W

—

272

>2.00

Gill net

1985

7/10

47°30'

128°36W

—

272-298

>2.00

Gill net

1985

7/11

47°59'

129°19'W

—

277

>2.00

Gill net

1985

7/13

46°23'

131°29W

—

188-219

0.71-0.94

Gill net

1985

7/14

46°40'

131°8W

—

183-233

0.68-1.1

Gill net

1985

7/15

46°35'

131°5W

—

206

0.83

Gill net

1985

7/19

46°54'

131°28'W

—

190-217

0.72-0.92

Gill net

1985

7/21

47°18'

130°23'W

—

191-220

0.73-0.95

Gill net

1985

7/22

47°35'

130°48'W

—

190-201

0.72-0.79

Gill net

1985

7/23

47°37'

130°3rW

—

198-200

0.77-0.79

Gill net

1985

7/12

47°0'

155°1'W

11.6

263-265

1.87-2.02

Longline

1985

7/13

45°28'

155°5W

11.8

238-262

1.17-1.81

Longline

1985

7/11

47'='0'

155°0'W

11.6

248-262

1.35-1.81

Gill net

1985

7/12

45°30'

155°0W

11.8

235-240

1.12-1.20

Gill net

1985

7/13

44°0'

155°0'W

12.8

248-258

1.35-1.63

Gill net

1985

2/23

30° 16'

181°18W

—

8-15

0.08-0.16

Neuston

1985

2/24

29°27'

180°56'W

—

5-15

0.04-0.16

Neuston

1986

8/3

43°51'

164°54'W

13.0

50-100

0.23-0.35

Dip net

456

BOEHLERT and SASAKI: PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD

150°

120'

Figure 2.— Records of pelagic occurrences of armorhead, Pseudopentaceros wheeleri, in the North Pacific. Most data were obtained
fronn either published records or unpublished data described in Table 1: Each symbol represents a collection taking armorhead,
rather than single specimens. (A) 1955-69. Here, note that the area in the box contains the locations of 195 sei whales captured
in July-August 1969 that had fed extensively upon armorhead. The samples from the central Pacific were larvae and early juveniles
(Fedosova and Komrakov 1975). (B) 1970-87.

457

FISHERY BULLETIN: VOL. 86, NO. 3

tant diet component of sei whales taken by the
Japanese commercial fleet. This species normally
feeds on zooplankton and micronekton (Kawamura
1982), but in this year, 195 whales were recorded
to have fed extensively on armorhead. Although
data records for whale stomach contents extend
from 1952 to 1979, 1969 was the only year when
large numbers fed on armorhead, with the excep-
tion of two occurrences each in 1971 and 1972.
Based on the known feeding mode of sei whales,
Chikuni (1970) suggested that the armorhead ap-
parently schools in surface waters. In the 1970's,
catches were relatively small, but in the 1980's,
specimens have been captured in each year, with the
greatest catches in 1985. In that year, fish occurred
in more easterly areas as compared to other years.

Relationship of Distribution Patterns
to North Pacific Oceanography

The SE-NHR seamounts are located in the mid-
Pacific transition zone, between subarctic and sub-
tropical water masses (Roden 1970). The spawning
season of armorhead typically lasts from late
November through March (Sasaki 1974; Bilim et al.
1978), and larvae are neustonic, at least through the

first few months of life (G. W. Boehlert unpubl.
data). Near surface drift in this region is largely
wind driven (McNally 1981); as shown by Lagran-
gian drifters, mean surface flow in the SE-NHR sea-
mount region during winter months is typically
eastward (Kirwan et al. 1978; Emery et al. 1985)
or southeastward (McNally et al. 1983), and long-
term mean Ekman transport has a southeastward
component (Favorite et al. 1976). Based on mean
ship-drift data, progressive vector diagrams over a
3-mo period, with starting points at four SE-NHR
seamounts (Bakun^), show this typical eastward flow
with a southward component (Fig. 3). If pelagic
armorhead followed such a pattern, individuals
would remain in the subtropical gyre, continuing
eastward and then turning south, entering the
California Current system. The scarcity of speci-
mens in this region (Figs. 1, 2) suggests instead that
most fish move northeastward in some manner,
entering subarctic waters and residing within the
Alaska gyre; surface drift in winter is northeastward
in anomalous years (McNally 1981), and average

'Compilation by the NMFS Pacific Fisheries Environmental
Group from historic ship drift files assembled by the U.S. Naval
Oceanographic Office (A. Bakun, Pacific Fisheries Environmen-
tal Group, National Marine Fisheries Service, NOAA, P.O. Box
831, Monterey, CA 93942, pers. commun. April 1987.

36°N

35°N

30*N

25°N

17rE

175°E

180'

178°W

Figure 3.— Progressive vector diagrams of mean surface currents (starting date, 15
December; duration, 90 days), from long-term mean ship drift data from 1° squares,
with distances calculated on a daily basis. Triangles indicate starting locations at four
seamounts with armorhead spawning populations. Each mark on the vector between
the origin and arrow represents a 5-d period. See text footnote 6.

458

BOEHLERT and SASAKI; PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD

surface flow is in this direction in summer months
(McNally et al. 1983).

The two available studies on larval distribution of
armorhead are conflicting and suggest that patterns
of movement may differ from year to year. Komra-
kov (1970) conducted surveys in March-April 1969
and observed larvae only at northern seamounts and
in the open ocean several hundred kilometers north-
east of the SE-NHR seamounts. Borets (1979), how-
ever, observed 5-20 mm larvae remaining mostly
around the seamounts, with highest abundance
south of lat. 33°N in 1976. Because spawning oc-
curs at seamounts northwest of this area, this pat-
tern is consistent with southeastward drift during
the first few months of life (Fig. 3). If larvae remain
in the seamount region, they could be transported
to the northeast in summer months, or they may
actively migrate northeastward after reaching the
juvenile stage. A similar migration apparently
occurs for the pomfret, Bramajaponica, which prior
to spawning moves into the North Pacific current
region in a broad latitudinal band between the sub-
tropical convergence and the subarctic boundary;
immature fish and spent adults then move north-
ward into the subarctic region to feed during April-
May (Shimazaki and Nakamura 1981). The pomfret
differs from the armorhead, however, in that it
remains pelagic throughout its life and also is abun-
dant in the western North Pacific. Larval and
juvenile Brama sp. do occur in the same area as lar-
val armorhead (Borets and Sokolovsky 1978; G. W.
Boehlert unpubl. data).

A possible scenario of movements after the lar-
val stage can be deduced from the ages of individuals
captured in different regions. Based upon the
growth of pelagic armorhead (Uchiyama and Sam-
paga fn. 3), most of the larger specimens indicated
in Figure 2 are from 1 to 2 years old, and the pelagic
duration for this species is most often from 1.5 to
2.5 years. Using daily growth increments for lar-
val and small juveniles and fitted growth curves for
larger juveniles and adults, one can convert lengths
to ages. The pelagic specimens of known length
range in estimated age from 0.04 to >2 years (Table
2). The youngest animals are typically found in the
region of the seamounts (Komrakov 1970; Fedosova
and Komrakov 1975; Borets and Sokolovsky 1978),
but intermediate-sized fish are found north and east
of the seamounts. Different age groups seem to be
distributed in different areas in the eastern North
Pacific (Fig. 4). Armorhead spawn in the seamount
region (as area A in Figure 4) during November to
March, and juveniles 5-25 mm long are found in the
SE-NHR seamounts area from February to early

April (Fedosova and Komrakov 1975; Borets 1979).
Subsequent occurrences of fish older than 0.25 year
(Table 2) are found only in the region northeast of
the seamounts and in the Gulf of Alaska (area B in
Figure 4). Northeastward movements are contrary
to mean ocean currents, but Roden et al. (1982)
described intense northward flow along the axis of
the southern Emperor Seamounts that could con-
ceivably play a role in transport of fish at the young-
est stages.

Recruitment of later stage armorhead to the sea-
mounts requires movement to the west. The only
significant westward circulation in the North Pacific
is found in the Alaskan Stream (Favorite 1967). This
narrow current occurs north of lat. 51 °N and ex-
tends westerly to near long. 170 °E, where it splits
to two branches, one turning northerly into the
Bering Sea and the other southerly under the influ-
ence of the northern Emperor Seamounts or Kor-
mandorskie Ridge (Favorite 1967; dashed line in
Figure 4). Damitsky et al. (1984) described southerly
movement of this water to at least lat. 40 °N and
suggested that it played a role in transport of
planktonic food for armorhead on the seamounts.
Unless returning armorhead are deep in the water
column and therefore not sampled, distributional
data (Fig. 2) do not support this route for westerly
movement. Fish near recruitment size and age are
instead captured in the region marked C on Figure
4, south of the area where younger fish are captured.
These fish apparently migrate back to the region of
the seamounts by some unknown, but probably
active, mechanism.

Two less important, but plausible, movement
patterns may explain the rarer occurrences in south-
ern Japan, off Oregon and California, and in the
Hawaiian Archipelago (Fig. 4). Chelton (1984) dis-
cussed possible interannual changes in the lati-
tudinal position of the West Wind Drift which may
lead to differing magnitudes of transport in the
Alaska and California Currents. The rare benthic
occurrences of the pelagic armorhead off the west
coast of North America (area D in Figure 4) may
come from specimens that drift southward in the
California Current in years of greater southern
transport. Some of these may not settle out and re-
main within the subtropical gyre, possibly recruit-
ing to the Hawaiian chain far to the south of the
normal reproductive population. This route, because
of its distance, apparently takes considerably longer
than that in the subarctic gyre; based upon satellite
drifters, McNally et al. (1983) suggested that a full
circuit of the subtropical gyre takes 4.5 years. The
armorhead collected in the Northwestern Hawaiian

459

FISHERY BULLETIN: VOL. 86, NO. 3

Figure 4.— Distributions of benthic and pelagic specimens of armorhead, Pseudopentaceros wheeleri, by age group with possible
patterns of movement. Area A represents the major spawning area, and area E the region inhabited by the large, older specimens.
Area B is the region in which specimens younger than 1 year have been captured, and area C where fish from 1 to 2 years old
have been captured. The arrow through this region represents the main population movement pattern. Other arrows represent
possible movements for stray fish. Areas marked D and F represent locations of benthic collections from California and Japan,
respectively.

Islands (area E in Figure 4) are typically larger than
those in the normal part of the range and are corre-
spondingly older, with ages estimated at 4 or 5 years
(Humphreys et al. in press), similar to the time sug-
gested for a circuit of the gyre by McNally et al.
(1983).

The other plausible movement pattern can be used
to explain the existence of specimens in Japan. Ben-
thic specimens of armorhead are rarely captured in
Japan (Fig. 1); a report of 3,000 t landed in 1969
by Soviet fishermen in this region (Abe 1969) may
be in error, since the Soviet fishery on the SE-NHR
seamounts developed heavily that year (Borets 1975)
and SE-NHR fish were likely sold in Japan. The
reproductive condition of these animals is not
known, so the source of the spawning population
supplying them is at question. The large, reproduc-
tive individuals in area E (Fig. 4) have been found
as far south as lat. 23°43'N (Humphreys et al. in
press). It is possible that larvae and juveniles are
transported to Japan in the northern part of the
North Equatorial Current (Uda and Hasunuma

1969). This is consistent with the capture of over
20 fish in a Bryde's whale stomach in September
1979 at lat. 26°N (H. Kato fn. 5; Fig. 2B). Subse-
quent captures of this species have not been made
in Japanese waters, where their presence may re-
quire a rare recruitment event from an upstream
population source, as has been suggested for sea-
mount populations of lobsters (Lutjeharms and
Heydorn 1981). These animals may also remain
within a gyral circulation, however, as suggested by
the presence of two specimens captured in the
Kuroshio in 1986 (Fig. 2B).

Recruitment to the Seamounts

The seasonal timing of the recruitment of the
pelagic armorhead to the SE-NHR seamounts is
unknown. That no records of larger pelagic fish cap-
tured near the seamounts (Fig. 2) exist may be an
artifact of sampling in inappropriate seasons or lack
of sampling in deeper water. The youngest benthic
specimens from the seamounts were slightly in ex-

460

BOEHLERT and SASAKI: PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD

cess of 1.5 years in early summer (Uchiyama and
Sampaga fn. 3) and 2 years in winter, suggesting
that some recruitment occurs in spring.

In most fisheries, temporal recruitment patterns
can be discerned from length-frequency analyses,
because the smaller recruits represent a larger pro-
portion of the population during the season of
recruitment. For the armorhead, no increase in
smaller size classes is apparent in monthly length-
frequency samples. This may be due to cessation of
somatic growth after recruitment to the seamount
(Humphreys and Tagami 1986; Uchiyama and Sam-
paga fn. 3). Early recruits, with their deeper bodies
and greater fat content, differ morphologically from
longer seamount residents. The transition from
so-called "fat" to "lean" morphotypes occurs in
association with development of gonads (Humphreys
et al. in press). Thus, an index of morphological
change can be used to detect recruitment patterns,
much the same as length frequencies can be used
for other species. Condition factor is frequently used
to assess "fatness" of fish; it is normally expressed
as weight divided by length to a power (typically 3)
multiplied by some scaling factor (Ricker 1975). For
armorhead, there is a marked change in condition
factor with morphological change; newly recruiting
fish have high condition factors relative to inter-

mediate or lean fish. By considering the monthly
proportion of fish above an arbitrary value of con-
dition factor, we can estimate the influx of new
recruits to the seamounts. During May 1972-
December 1973, the major influx of new recruits
with a high condition factor occurred in April-May
1973 (Fig. 5). The data for 1972 are incomplete, but
some recruitment apparently occurred in August
and September (Fig. 5). Differences in seasonality
of recruitment between 1972 and 1973 may be
indicative of interannual variability in temporal
recruitment patterns.

An influx of fish to the seamounts may also be
reflected in the catch per unit effort (CPUE), an
index of stock abundance. Monthly averages of
CPUE (in metric tons per hour) by Japanese
trawlers for armorhead at the SE-NHR seamounts
were highest in March and April (Sasaki 1986).
Overall armorhead CPUE decreased continuously
from 54.1 t/h in 1972 to only 0.3 t/h in 1982 (Sasaki
1986) and further decreased to 0.06-0.07 t/h in
1983-85 (T. Sasaki unpubl. data). The estimated
CPUE in 1986, however, was 0.31 t/h, an increase
of four or five times that in the previous year. The
CPUE of trawlers operating in May-Septem.ber
1986 was highest in May, decreased in June and
July, but increased slightly in August and Septem-

35-

30-

H 25

O 20-

°- 15H

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

1972

MONTH

1973

Figure 5.— Seasonality of the proportion of armorhead, Pseudopentaceros wheeleri, with
condition factors >2.0 from four seamounts of the northern Hawaiian Ridge (Southeast
Hancock, Northwest Hancock, C-H, and Colahan) for 1972-73. Squares, males; triangles,
females. Data are based upon 2,104 males and 1,704 females.

461

FISHERY BULLETIN: VOL. 86, NO. 3

ber. Based on all the information, the major recruit-
ment may occur in spring with some additional
recruitment in August and September (Fig. 5); data
are not sufficient to determine the recruitment
pattern in detail.

A broad geographic separation exists between the
location of pelagic captures of larger armorhead and
the location of the spawning populations (Figs. 2,
4). If recruitment to the seamounts occurs predom-
inantly in spring (as suggested for 1973 in Figure
5), then temporal sampling patterns may have
missed these fish, although JAMARC pomfret
surveys covered this area in some seasons (Table 1).
Locating the seamounts, which have small (2-5 km)
summits must be a formidable task given the wide
ocean areas over which armorhead are distributed.
A similar situation exists for rock lobster, Jasus
tristani, in the South Atlantic, that recruits from
an upstream population some 2,000 km away and
in sufficient numbers to support a fishery on Vema
Seamount in some years (Lutjeharms and Heydorn
1981).

Open-ocean migrations of fishes may depend
upon many potential cues, including electric fields
(McCleave and Power 1978), magnetic fields
(Walker 1984), gyres (Williams 1972), and phero-
mones (Nordeng 1977). Certain characteristics of
these isolated, open-ocean seamounts may promote
their detection by armorhead. First, current-topog-
raphy interactions may create significant signals in
physical and biological features. The region of the
SE-NHR seamounts is active in front development
(Roden and Paskausky 1978); upwelling, eddies, and
other aspects of flow complexity also occur around
these seamounts (Roden et al. 1982) and down-
stream from them (Royer 1978). The biological sig-
nals may include increased chlorophyll in response
to upwelling or doming of isotherms (Genin and
Boehlert 1985), or aggregations of various organ-
isms and the larger animals which prey upon them
around seamounts (see review in Boehlert and Genin
1987). Gravity anomalies associated with seamounts
may also play a role; positive gravity anomalies exist
at the summit and slopes, and negative anomalies
are seen in the surrounding "moat" regions (Wedge-
worth and Kellogg 1987). Seamounts often have
strong magnetic dipoles associated with them, and
the dipole might serve as a landmark for magnetic
orientation by fish (Klimley^). While fish have been
shown to have magnetoreceptors (Walker et al.
1985), their use of magnetic maps remains specula-

T. Klimley, Scripps Institution of Oceanography, La JoUa, CA
92038, pers. commun. 30 June 1987.

tive but possible (Gould 1985). Although we cannot
postulate the mechanism that armorhead use for
recruitment, it is clear that the effects of seamounts
may be detected at distances greater than their area
alone would suggest.

Interannual Variations in
Recruitment Strength

The year-class strength of armorhead recruiting
to seamounts appears to be independent of the
parent stock size (Wetherall and Yong 1986). As an
example, Borets (1975) estimated that on the SE-
NHR seamounts from 1968 to 1973, the stock size
varied by a factor of <1.8 while recruitment varied
by >5.5 times. The relative abundance of armorhead
at the SE-NHR seamounts area increased in 1986
after a long period at a very low level. This increase
probably corresponds to the high abundance of
pelagic specimens captured in the northeastern
Pacific in 1985 (Fig. 2B; Table 2) that consisted of
two age groups. The increased recruitment at the
SE-NHR seamounts in 1986 suggests that environ-
mental conditions were favorable to the survival of
young armorhead in the 1984 and 1985 winter
seasons.

A wide variety of factors, both biotic and physical,
can affect survival and ultimate year-class strength
in fishes (Lasker 1978). During the 2 years between
spawning and recruitment for armorhead, an ex-
tended migration through varied pelagic environ-
ments occurs (Fig. 4). Feeding conditions for larval
and juvenile stages are characterized by interannual
variability; Fedosova (1980) suggested that warm
years were more productive for zooplankton prey
and, thus, favorable to the survival of young armor-
head. Interannual variation in atmospheric systems
(Seckel 1988) or large-scale ocean currents of the
kind described by Mysak et al. (1982) may also play
a role in armorhead recruitment strength. Changes
in the position of the Alaska gyre by up to 700 km
southwest of its normal position may have occurred
from 1981 to 1985, with an associated increase in
seawater temperature (Royer and Emery 1987).
Large-scale atmospheric phenomena, such as the
longitudinal position of the Aleutian Low, may
create definite interannual variations in winter wind
systems that may be seen in surface current pat-
terns (Seckel 1988). These patterns may, in turn,
be related to the latitudinal position of the subtrop-
ical front, which varies interannually between lat.
28° and 32°N (Roden 1970). Variability in these
features influences surface drift (McNally 1981),
which in turn affects the neustonic young of armor-

462

BOEHLERT and SASAKI: PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD

head. If flow in certain years results in transport
of armorhead to other regions (such as the western
Pacific or the southern part of the northeast Pacific;
Fig. 4), it is possible that large-scale mortalities of
armorhead occur later in life, resulting in weak year
classes at the SE-NHR seamounts.

ACKNOWLEDGMENTS

We appreciate assistance from many individuals
who provided access to different data sources used
in this paper, including Akito Kawamura, Colin Har-
ris, Sigeiti Hayasi, Genji Kobayashi, and Glen Jamie-
son. We also thank Ted Pietsch, Gunter Seckel, and
Peter Klimley and an anonymous reviewer for com-
ments on an earlier draft of the manuscript.

LITERATURE CITED

Abe, T.

1957. New, rare or uncommon fishes from Japanese [waters].
VI. Notes on the rare fishes of the family Histiopteridae.
Jpn. J. Ichthyol. 6:35-39, 71-74.
1969. Notes on some edible marine fishes collected between
the Bonin Islands and the mouth of Sagami Bay— III. Bull.
Tokai Reg. Fish. Res. Lab. 60:5-8.

Alton, M. S.

1986. Fish and crab populations of Gulf of Alaska seamounts.
In R. N. Uchida, S. Hayasi, and G. W. Boehlert (editors),
Environment and resources of seamounts in the North
Pacific, p. 45-51. U.S. Dep. Commer., NOAA Tech. Rep.
NMFS 43.

BiLIM, L. A., L. A. BORETS, AND L. K. Platoshina.

1978. Characteristics of ovogenesis and spawning of the boar-
fish in the region of the Hawaiian Islands. In Fisheries
oceanography, hydrobiology, biology of fishes and other
denizens of the Pacific Ocean, p. 51-57. Izv. Pac. Ocean
Sci. Res. Inst. Fish. Oceanogr. (TINRO), Vladivostok, 102.
(Engl, transl. by W. G. Van Campen, 1986, 9 p., Transl. No.
106; available Southwest Fish. Cent. Honolulu Lab., Natl.
Mar. Fish. Serv., NOAA, Honolulu, HI 96822-2396.)

Boehlert, G. W.

1986. Productivity and population maintenance of seamount
resources and future research directions. In R. N. Uchida,
S. Hayasi, and G. W. Boehlert (editors). Environment and
resources of seamounts in the North Pacific, p. 95-101.
U.S. Dep. Commer., NOAA Tech. Rep. NMFS 43.

Boehlert, G. W., and A. Genin.

1987. A review of the effects of seamounts on biological
processes. In B. Keating, P. Fryer, R. Batiza, and G. W.
Boehlert (editors), Seamounts, islands, and atolls, p. 319-
334. Geophys. Monogr. 43.

BORETS, L. A.

1975. Some results of studies on the biology of the boarfish
{Pentaceros richardsoni Smith). Invest. Biol. Fishes Fish.
Oceanogr., TINRO, Vladivostok 6:82-90. (Engl, transl. by
W. G. Van Campen, 1984, 9 p., Transl. No. 97; available
Southwest Fish. Cent. Honolulu Lab., Natl. Mar. Fish. Serv.,
NOAA, Honolulu, HI 96822-2396.)

1976. The state of the boarfish, Pentaceros richardsoni
(Smith) stock in the Hawaii submersive range areas. Invest.
Biol. Fishes Fish. Oceanogr., TINRO, Vladivostok 7:58-62.

(Engl, transl. by W. G. Van Campen, 1987, 5 p., Transl.
No. 112; available Southwest Fish. Cent. Honolulu Lab.,
Natl. Mar. Fish. Serv., NOAA, Honolulu, HI 96822-
2396.)

1979. The population structure of the boarfish, Pentaceros
richardsoni, from the Emperor Seamounts and the
Hawaiian Ridge. J. Ichthyol. 19:15-20.

BORETS, L. A., AND A. S. SOKOLOVSKY.

1978. Species composition of the ichthyoplankton of the
Hawaiian submarine ridge and the Emperor Seamounts.
In Fisheries oceanography, hydrobiology, biology of fishes
and other denizens of the Pacific Ocean, p. 43-50. Izv. Pac.
Ocean Sci. Res. Inst. Fish. Oceanogr. (TINRO), Vladivostok,
Vol. 102. (Engl, transl. by W. G. Van Campen, 1986, 10
p., Transl. No. 105; available Southwest Fish. Cent. Hono-
lulu Lab., Natl. Mar. Fish. Serv., NOAA, Honolulu, HI
96822-2396.)

Chelton, D. B.

1984. Commentary: Short-term climatic variability in the
northeast Pacific Ocean. In W. G. Pearcy (editor). The in-
fluence of ocean conditions on the production of salmonids
in the North Pacific, p. 87-99. Oreg. State Univ. Sea Grant,
ORESU-W-83-001.

Chikuni, S.

1970. On gregarious fish, Pseudopentaceros richardsoni
(Histiophoridae). Enyo (Far Seas) Fish. Res. Lab. News
3:1-4. (Engl, transl., The "phantom fish," "kusakari tsu-
bodai"— an outline, by J. H. Shohara, 7 p.; available Natl.
Mar. Fish. Serv., NOAA, Terminal Island, CA 90731.)

Clemens, W. A., and G. V. Wilby.

1961. Fishes of the Pacific coast of Canada. 2d ed. Bull.
Fish. Res. Board Can. 68, 443 p.

Darnitsky, V. B., B. L. Boldyrev, and A. F. Volkov.

1984. Environmental conditions and some ecological char-
acteristics of fishes from the central North Pacific sea-
mounts. In P. A. Moiseev (editor). Proceedings, conditions
of formation of commercial fish concentrations, p. 64-77.
Minist. Fish., U.S.S.R., All-Union Res. Inst. Mar. Fish.
Oceanogr. VNIRO. (Engl, transl. by W. G. Van Campen,
1987, 12 p., Transl. No. 114; available Southwest Fish. Cent.
Honolulu Lab., Natl. Mar. Fish. Serv., NOAA, Honolulu, HI
96822-2396.)

Emery, W. J., T. C. Royer, and R. W. Reynolds.

1985. The anomalous tracks of North Pacific drifting buoys
1981 to 1983. Deep-Sea Res. 32:315-347.

Favorite, F.

1967. The Alaskan Stream. Int. N. Pac. Fish. Comm. Bull.
21, 20 p.
Favorite, F., A. J. Dodimead, and K. Nasu.

1976. Oceanography of the subarctic Pacific region, 1960-
1971. Int. N. Pac. Fish. Comm. Bull. 33, 187 p.
Fedosova, R. a.

1980. Inter-year changes in the mesoplankton of the waters
of the submarine banks of the Hawaiian Ridge (Northwest-
ern Pacific Ocean). [In Russ.] Izv. Pac. Ocean Sci. Res.
Inst. Fish. Oceanogr. (TINRO), Vladivostok 104:84-87.

Fedosova, R. A., and 0. E. Komrakov.

1975. Feeding of Pentaceros richardsoni frys in the Hawaiian
region. [In Russ.] Invest. Biol. Fishes Fish. Oceanogr.,
TINRO, Vladivostok 6:52-55. (Engl, transl. by W. G. Van
Campen, 1985, 4 p., Transl. No. 100; available Southwest
Fish. Cent. Honolulu Lab., Natl. Mar. Fish. Serv.. NOAA,
Honolulu, HI 96822-2396.)

FOLLETT, W. I., AND L. J. DEMPSTER.

1963. Relationships of the percoid fish Pentaceros richard-
soni Smith, with description of a specimen from the coast

463

FISHERY BULLETIN: VOL. 86, NO. 3

of California. Proc. Calif. Acad. Sci., 4th Ser. 32(10):315-
338.

FUJII, E.

1986. Zoogeographical features of fishes in the vicinity of sea-
mounts. In R. N. Uchida, S. Hayasi, and G. W. Boehlert
(editors). Environment and resources of seamounts in the
North Pacific, p. 67-70. U.S. Dep. Commer., NOAA Tech.
Rep. NMFS 43.

Genin, a., and G. W. Boehlert.

1985. Dynamics of temperature and chlorophyll structures
above a seamount: An oceanic experiment. J. Mar. Res.
43:907-924.

Gould, J. L.

1985. Are animal maps magnetic? In J. L. Kirschvink, D. S.
Jones, and B. J. MacFadden (editors), Magnetite biomineral-
ization and magnetoreception in organisms, p. 257-268.
Plenum Press, N.Y.

Hardy. G. S.

1983. A revision of the fishes of the family Pentacerotidae
(Perciformes). N.Z. J. Zool. 10:177-220.

HONMA, Y., AND R. MiZUSAWA.

1969. A record of the young of boar fish, Pseudopentaceros
richardsoni, from North Pacific. Jpn. J. Ichthyol. 15:134-
136.

Hughes, S. E.

1981. Initial U.S. exploration of nine Gulf of Alaska sea-
mounts and their associated fish and shellfish resources.
Mar. Fish. Rev. 43(l):26-33.

Humphreys, R. L., Jr., and D. T. Tagaml

1986. Review and current status of research on the biology
and ecology of the genus Pseudopentaceros. In R. N.
Uchida, S. Hayasi, and G. W. Boehlert (editors). Environ-
ment and resources of seamounts in the North Pacific, p.
55-62. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 43.

Humphreys, R. L., Jr., G. Winans, and D. T. Tagaml

In press. Synonomy and proposed life history of the North
Pacific pelagic armorhead, Pseudopentaceros wheekri Hardy
(Pisces: Pentacerotidae). Copeia 1988.
Ignell, S., J. Bailey, and J. Joyce.

1986. Observations on high-seas squid gill-net fisheries, North
Pacific Ocean, 1985. U.S. Dep. Commer., NOAA Tech.
Memo. NMFS F/NWC-105, 52 p.
Japan Marine Fishery Resource Research Center.

1980. Report of exploratory fishing of Pacific pomfret. Jpn.

Mar. Fish. Resour. Res. Cent. Rep. 53-16.
1983a. Report of exploratory fishing of Pacific pomfret.

Jpn. Mar. Fish. Resour. Res. Cent. Rep. 54-18.
1983b. Report of exploratory fishing of Pacific pomfret.

Jpn. Mar. Fish. Resour. Res. Cent. Rep. 55-16.
1985. Report of exploratory fishing of Pacific pomfret. Jpn.
Mar. Fish. Resour. Res. Cent. Rep. 57-12.
Kawamura, a.

1982. Food habits and prey distributions of three rorqual
species in the North Pacific Ocean. Sci. Rep. Whales Res.
Inst. 34:59-91.

KiRWAN, A. D., Jr., G. H. McNally, E. Reyna, and W. J.
Merrell, Jr.
1978. The near-surface circulation of the eastern North
Pacific. J. Phys. Oceanogr. 8:937-945.
KOMRAKOV, 0. E.

1970. Distribution and fishery of the boarfish (Pentaceros
richardsoni Smith) in the Hawaiian region. In the collection:
The present state of biological productivity and the volume
of biological resources of the world ocean and prospects for
their utilization. Kaliningrad, p. 155-163. (Engl, transl.
by W. G. Van Campen, 1987, 9 p., Transl. No. 117; available

Southwest Fish. Cent. Honolulu Lab., Natl. Mar. Fish. Serv.,
NOAA, Honolulu, HI 96822-2396.)
Larkins, H. a.

1964. Some epipelagic fishes of the North Pacific Ocean,
Bering Sea, and Gulf of Alaska. Trans. Am. Fish. Soc.
93:286-290.
Lasker, R.

1978. The relation between oceanographic conditions and lar-
val anchovy food in the California Current: Identification
of factors contributing to recruitment failure. Rapp. P. -v.
R6un. Cons. int. Explor. Mer 173:212-230.
Lutjeharms, J. R. E., and A. E. F. Heydorn.

1981. Recruitment of rock lobster on Vema Seamount from
the islands of Tristan da Cunha. Deep-Sea Res. 28A: 1 237.
Macy, p. Y., J. M. Wall, N. D. Lampsakis, and J. E. Mason.
1978. Resources of non-salmonid pelagic fishes of the Gulf
of Alaska and eastern Bering Sea. Parts 1-3, Final Rep.
BLM/OCSEAP Contracts R7120811 and R7120812, North-
west and Alaska Fish. Cent., NMFS, NOAA, Seattle, Wash.
McCleave, J. D., AND J. H. Power.

1978. Influence of weak electric and magnetic fields on turn-
ing behavior in elvers of American eel Anguilla rostrata.
Mar. Biol. (Berl.) 46:29-34.
McNally, G. J.

1981. Satellite-tracked drift buoy observations of the near-
surface flow in the eastern mid-latitude North Pacific. J.
Geophys. Res. 86:8022-8030.

McNally, G. J., W. C. Patzert, A. D. Kirwan, Jr., and A. C.
Vastano.

1983. The near-surface circulation of the North Pacific using
satellite tracked drifting buoys. J. Geophys. Res. 88:7505-
7518.

Mysak, L. a., W. W. Hsieh, and T. R. Parsons.

1982. On the relationship between interannual baroclinic
waves and fish populations in the northeast Pacific. Biol.
Oceanogr. 2:63-103.

Nordeng, H.

1977. A pheromone hypothesis for homeward migration in
anadromous salmonids. Oikos 28:155-159.

Okamura, I., K. Amaoka, and F. Mitani (editors).

1982. Fishes of the Kyushu-Palau Ridge and Tosa Bay. Jpn.
Fish. Res. Conserv. Assoc, Tokyo, 435 p.
Randall, J. E.

1980. New records of fishes from the Hawaiian Islands. Pac.
Sci. 34:211-232.
Ricker, W. E.

1975. Computation and interpretation of biological statistics
of fish populations. Fish. Res. Board Can. Bull. 191, 382 p.
Robinson, S. M. C, and G. S. Jamieson.

1984. Report on a Canadian commercial fishery for flying
squid using drifting gill nets off the coast of British Colum-
bia. Can. Ind. Rep. Fish. Aquat. Sci. 150, 25 p.

Roden, G. I.

1970. Aspects of the mid-Pacific transition zone. J. Geophys.
Res. 75:1097-1109.
Roden, G. I., and D. F. Paskausky.

1978. Estimation of rates of frontogenesis and frontolysis in
the North Pacific Ocean using satellite and surface meteor-
ological data from January 1977. J. Geophys. Res. 83:4545-
4550.

Roden, G. I., B. A. Taft, and C. C. Ebbesmeyer.

1982. Oceanographic aspects of the Emperor Seamounts
region. J. Geophys. Res. 87:9537-9552.
ROYER, T. C.

1978. Ocean eddies generated by seamounts in the North
Pacific. Science (Wash., D.C.) 199:1063-1064.

464

BOEHLERT and SASAKI: PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD

RoYER, T. C. AND W. J. Emery.

1987. Circulation in the Gulf of Alaska, 1981. Deep-Sea Res.
34:1361-1377.

Sasaki, T.

1974. Kita Taiheiyo no kusakari tsubodai (The pelagic armor-
head, Pentaceros richardsoni Smith, in the North Pacific).
Bull. Jpn. Soc. Fish. Oceanogr. 24:156-165. (Engl, transl.
by T. Otsu, 1977, 13 p., Transl. No. 16; available Southwest
Fish. Cent. Honolulu Lab., Natl. Mar. Fish. Serv., NOAA,
Honolulu, HI 96822-2396.)

1986. Development and present status of Japanese trawl
fisheries in the vicinity of seamounts. In R. N. Uchida,
S. Hayasi, and G. W. Boehlert (editors). Environment and
resources of seamounts in the North Pacific, p. 21-30. U.S.
Dep. Commer., NOAA Tech. Rep. NMFS 43.
Seckel, G. R.

1988. Indices for mid-latitude North Pacific winter wind
systems; an exploratory investigation. GeoJournal 16:97-
111.

Shimazaki, K., and S. Nakamura.

1981. Ecological studies of the pomfret (Brama japonica
Hilgendorf). I. The seasonal distribution pattern and eco-
logical considerations. Res. Inst. N. Pac. Fish., Spec. Vol.,
p. 91-103.

Sloan, N. A.

1983. Canadian-Japanese experimental fishery for oceanic
squid off British Columbia, summer 1983. Can. Ind. Rep.
Fish. Aquat. Sci. 152, 42 p.

Smith, J. G.

1965. A second record of the pelagic armorhead, Pentaceros

richardsoni Smith, 1849, in California waters. Calif. Fish

Game 51:213-214.
Takahashi, Y., and T. Sasaki.

1977. Kita Taiheiyo chubu kaizan ni okeru tororu gyogyo

(Trawl fishery in the central North Pacific seamounts).

Northern Waters Groundfish Resour., Far Seas Fish. Res.

Lab., 45 p. (Engl. Transl. by T. Otsu, 1977, 49 p.; Transl.

No. 22; available Southwest Fish. Cent. Honolulu Lab., Natl.

Mar. Fish. Serv., NOAA, Honolulu, HI 96822-2396.)
Uchida, R. N., and D. T. Tagami.

1984. Groundfish fisheries and research in the vicinity of sea-
mounts in the North Pacific Ocean. Mar. Fish. Rev. 46(2):
1-17.

Uda, M., and K. Hasunuma.

1969. The eastward subtropical countercurrent in the west-
ern North Pacific Ocean. J. Oceanogr. Soc. Jpn. 25:201-
210.
Wagner, E. J., and C. E. Bond.

1961. The percoid fish Pseudopentaceros richardsoni from
Oregon waters. Fish. Comm. Oreg., Res. Briefs 8(l):71-73.
Walker, M. M.

1984. Magnetic sensitivity and its possible physical basis in
the yellowfin tuna, Thunniis albacares. In J. D. McCleave,
G. P. Arnold, J. J. Dodson, and W. H. Neill (editors).
Mechanisms of migration in fishes, p. 125-142. Plenum
Press, N.Y.

Walker, M. M., J. L. Kirschvink, and A. E. Dizon.

1985. Magnetoreception and biomineralization of magnetite.
Fish. In J. L. Kirschvink, D. S. Jones, and B. J. MacFad-
den (editors). Magnetite biomineralization and magneto-
reception in organisms, p. 417-437. Plenum Press, N.Y.

Wedgeworth, B., and J. Kellogg.

1987. A 3-D gravity-tectonic study of Ita Mai Tai Guyot: an
uncompensated seamount in the East Mariana Basin. In
B. Keating, P. Frj-er, R. Batiza, and G. W. Boehlert (editors),
Seamounts, islands, and atolls, p. 73-84. Geophys. Monogr.
43.

Welander, a. D., R. C. Johnson, and R. A. Hajny.

1957. Occurrence of the boarfish, Pseudopentaceros richard-
soni, and the zeid, Allocyttus verrucosus, in the North
Pacific. Copeia 1957:244-246.

Wetherall, J. A., and M. Y. Y. Yong.

1986. Problems in assessing the pelagic armorhead stock on
the central North Pacific Seamounts. In R. N. Uchida,
S. Hayasi, and G. W. Boehlert (editors), Environment and
resources of seamounts in the North Pacific, p. 73-85. U.S.
Dep. Commer., NOAA Tech. Rep. NMFS 43.

Williams, F.

1972. Consideration of three proposed models of the migra-
tion of young skipjack tuna {Katsuwoniis pelamis) into the
eastern Pacific Ocean. Fish. Bull., U.S. 70:741-762.
Zama, a., M. Asai, and F. Yasuda.

1977. Records of the pelagic armorhead, Pentaceros richard-
soni from Hachijo Island and the Ogasawara Islands. Jpn.
J. Ichthyol. 24:57-60.

465

THE TIMING AND SIGNIFICANCE OF DENSITY-DEPENDENT AND
DENSITY-INDEPENDENT MORTALITY OF AMERICAN SHAD,

ALOSA SAPIDISSIMA

Thomas F. Savoy and Victor A. Creccoi

ABSTRACT

We used stock-recruitment, pre- and postrecruitment mortality data for American shad, Alosa
sapidissima, in the Connecticut River to estimate density-dependent and density-independent mortality
rates at the prejuvenile (age 1-100 days) and postjuvenile (age 101 days to 5 years) phases. Total post-
juvenile mortality rates from 1967 through 1982 were dominated by density-independent mortality, with
only 6% (^opost = 0.30) of the mean total (Z = 4.85) being ascribed to density-dependent processes. By
contrast, 23% {Z^ = 1.13) of the total prejuvenile mortality was compensatory, of which nearly all
took place during the embryonic and early larval periods. Egg and early larval mortality rates from 1979
through 1987 were positively correlated to June river flows, and inversely related to mean June
temperature, whereas mortality rates of all other life stages showed no such relationships. Daily egg
and early larval mortality rates from 1979 through 1987 were usually higher and more variable than
mortality rates during later stages. There were significant negative correlations between egg and early
larval mortality rates and the strength of 1979-87 year classes in the adult stock, whereas mortality
rates of late larvae and juveniles were independent of year-class strength. Density-dependent mortality
during the egg and early larval stages comprised over 40% of the total mortality at those stages, resulting
in the number of midlarvae and juveniles being positively correlated to adult recruitment. These data
suggest that year-class strength of American shad in the Connecticut River is established after the egg
and larval stages.

It is generally accepted that year-class strength of
most fishes is established prior to the juvenile stage
from density-independent (climatic factors) and
density-dependent (competition, predation, cannibal-
ism) processes (Gushing 1974; May 1974; Goodyear
1980). Although density-dependent mortality is
believed to be the mechanism that keeps fish popu-
lations stable under low to moderate exploitation
(McFadden 1977; Gushing 1980), density-dependent
processes such as inter- and intraspecific competi-
tion, predation, and cannibalism are difficult to
measure. This is particularly so for many highly
fecund fishes, whose population sizes undergo wide
fluctuations that often cannot be predicted by con-
ventional stock-recruitment models (Parrish and
MacGall 1978; Bakun 1984). Despite the acknowl-
edged importance of density-dependent mortality to
stock-recruitment theory (Ricker 1975; Ware 1980),
few studies have been able to quantify density-
dependent mortality, or determine the life stages
at which compensation occurs.

Density-dependent mortality may be confined to
the prejuvenile stage (Gushing 1974, 1980) or the

'Connecticut Department of Environmental Protection, Marine
Fisheries Office, Waterford, CT 06385.

postjuvenile oceanic stage (Peterman 1978, 1982),
or may occur throughout the prerecruitment period
(Gulland 1965). The testing of these hypotheses has
proceeded slowly because of environmental noise
surrounding many stock-recruitment relationships,
measurement errors associated with recruitment
estimates (Ludwig and Walters 1981), and the lack
of long-term mortality data on eggs and larvae.

One species, for which accurate and long-term
stock-recruitment and prerecruitment mortality
data exist, is the American shad, Alosa sapidissima,
an anadromous clupeid that spawns in many Atlan-
tic coast rivers (Walburg and Nichols 1967). Pre-
vious studies on American shad in the Gonnecticut
River have demonstrated that larval and juvenile
mortality rates decline with age (Grecco et al. 1983,
1986) and that growth and survival rates among
discrete larval cohorts are significantly affected by
short-term hydrographic and meteorological events
(Grecco and Savoy 1985b, 1987a). Although pre-
vious stock-recruitment studies of Gonnecticut River
shad (Leggett 1977; Lorda and Grecco 1987) found
evidence of density-dependent mortality before the
juvenile stage, no studies have attempted to esti-
mate density-dependent mortality during the egg
and larval stages, or quantify density-dependent and

Manuscript accepted March 1988.

FISHERY BULLETIN: VOL. 86, NO. 3, 1988.

467

FISHERY BULLETIN: VOL. 86, NO. 3

density-independent mortality after the juvenile
stage.

In this study we used stock-recruitment, pre- and
postrecruitment mortality data for American shad
in the Connecticut River from 1967 through 1987
to 1) estimate egg mortality rates for the 1979-87
year classes; 2) measure the contribution of density-
dependent and density-independent mortality before
and after the juvenile stage; and 3) determine the
life stage(s) at which most of the density-dependent
mortality takes place.

METHODS

Data Source

Estimates of adult recruitment and parent stock
size (± SE) from 1967 to 1982 (Table 1) were based
on annual population estimates derived by earlier
mark-recapture studies (Leggett 1976; Crecco and
Savoy 1987^) and by the annual mean number of
American shad lifted over the Holyoke Dam (Crecco
and Savoy 1985a). The parent stock size (PAR J
from 1967 through 1987 was the annual population
estimate of female shad minus that year's commer-

cial catch of female shad. Parent stock (PAR^) was
separated into the parent stock that spawned above
the Holyoke Dam (PJ (Moffitt et al. 1982; O'Leary
and Booke 1986^, 1987^) and the spawning stock
that spawned below the dam (P,, = PAR, - P„).
Since female American shad mature between ages
four and six, female shad recruitment from the
1967-82 year classes was the sum of virgin 4-, 5-,
and 6-yr-old female shad in the 1970-87 runs based
on the age-class structure from previous studies
(Jones et al. 1976; Leggett 1976; Crecco et al.
1984^). Since the sex ratio of mature progeny from
each year class is close to 1:1 (Leggett 1976), total
recruitment (Rf) from the 1967-82 year classes was
estimated by doubling the female shad recruitment
estimates. Direct estimates of male shad recruit-
ment are biased by gill net selectivity and differen-
tial culling practices of commercial fishermen
(Crecco et al. 1984^).

^Crecco, V. A., and T. F. Savoy. 1987b. Fishery Management
Plan for American shad in the Connecticut River. Unpubl.
manuscr., 117 p. Connecticut Department of Environmental Pro-
tection, Hartford, CT 06106.

'O'Leary, J., and H. E. Booke. 1986. Connecticut River anad-
romous fish investigations. Mass. Coop. Fish. Res. Unit Proj. Per-
formance Rep., F-45-R-2, 37 p.

"O'Leary, J., and H. E. Booke. 1987. Connecticut River anad-
romous fish investigations. Mass. Coop. Fish. Res. Unit Proj. Per-
formance Rep., 32 p.

^Crecco, V. A., T. Savoy, and L. Gunn. 1984. Population
dynamics studies of American shad in the Connecticut River. CT
Dep. Environ. Prot. Final Rep. AFC 13, 76 p.

•^Crecco, V. A., L. Gunn, and T. Savoy. 1981. Connecticut
River shad study, a progress report. Unpubl. manuscr., 87 p.
Connecticut Department of Environmental Protection, Hartford,
CT 06106.

Table 1.— Estimates of total female American shad parent stock, female spawn-
ing stock above and below the Holyoke Dam, mean June flows from 1967 to 1987,
and total number of adult recruits from the 1967-82 year classes. SE = stand-
ard errors about the estimates x 10^.

Total

Spawning

spawning

stock

Recruit-

June

stock

above

below

ment

flow

Year

X 10^

m^/s

1967

167

0.8

163

444

112

437

1968

202

0.9

197

236

603

1969

384

1.8

374

490

116

375

1970

413

2.8

401

550

126

243

1971

424

111

3.7

410

107

982

205

203

1972

167

1.9

159

430

616

1973

111

0.8

108

308

534

1974

306

2.0

296

652

139

334

1975

247

4.2

231

560

101

379

1976

435

166

31.0

269

650

105

286

1977

207

112

21.0

1,240

191

250

1978

210

12.0

165

714

110

488

1979

248

18.0

161

882

127

445

1980

341

196

33.0

145

1,256

196

201

1981

293

143

24.0

150

758

132

316

1982

501

109

15.0

392

282

643

1983

423

185

34.0

238

386

1984

610

245

27.0

365

661

1985

555

219

38.0

336

311

1986

392

175

32.0

217

439

1987

235

136

14.0

146

250

468

SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD

American shad larvae (10-30 mm TL) were sam-
pled during daylight hours in the Connecticut River
with a 6.1 m plankton bag seine (2.4 m deep, wing
and bag mesh of 0.505 mm) and 30 m lead ropes
from 15 May to 20 July 1979-84, 1986, and 1987.
No larval sampling was conducted in 1985. One seine
haul was taken weekly at 8-12 fixed stations located
throughout the major spawning areas (Fig. 1). Fur-
ther details on sampling and methods of estimating
larval abundance from net samples are contained
in Crecco et al. (1983).

Juvenile American shad (40-90 mm TL) were col-
lected weekly from 20 July through 15 October
1967-87 at 7-14 fixed stations located above and
below the Holyoke Dam (Fig. 1) based on weekly and
biweekly seine surveys (Scherer 1974; Foote 1976;
Marcy 1976; Crecco and Savoy 1984). The annual

juvenile index of relative abundance (IND() from
1967 to 1987 was the mean juvenile catch per seine
haul ( ± SE) from all stations and collection dates
(Table 2).

The Timing of
Density-Dependent Mortality

Analysis

One of the primary objectives of this study was
to determine the magnitude and timing of density-
dependent mortality for American shad. Peterman
(1978, 1982) found that density-dependent mortality
for some stocks of the anadromous sockeye salmon,
Oiworhynchus nerka, was confined mainly to the 4-5
yr oceanic postjuvenile phase. Given that the Ameri-

STATIONS
X - L A RVA L

0-JUVENILE

km 1 90

SUN D ERLU N D

HATFIELD

HO LYOK E

km 140

X AG AW AM
ENFIELD

Figure 1.— Location of larval and juvenile
American shad sampling stations on the Con-
necticut River.

MASS
CONN

WILSON O

GLASTON BURY

km 25
O SALMON

RIVER

DEEP
RIVER

ESSEX

k m 1 2

469

FISHERY BULLETIN: VOL. 86, NO, 3

Table 2.— Mean juvenile indices, scaled juvenile abundance from
Equation (2) and total egg production for American shad in the Con-
necticut River from 1967 through 1987. SE = standard errors
about the estimates.

Juvenile

Total egg

Juvenile

abundance

production

Year

index
20.2

SE
9.4

X 10^

X 10«

X 10^

1967

71,070

33,072

334

1968

11.1

1.7

39,053

5,981

404

1969

19.0

3.8

68,848

13,770

768

140

1970

27.8

8.2

97,809

28,850

826

190

1971

65.7

14.1

231,151

49,608

848

222

1972

15.3

2.9

53,830

10,203

334

1973

12.7

3.4

44,682

11,962

222

1974

21.4

6.3

75,292

22,165

612

124

1975

23.7

5.7

83,384

20,054

494

130

1976

22.4

5.9

78,810

20,760

870

160

1977

--NOT AVAILABLE- -

414

1978

27.2

5.9

95,698

20,758

420

108

1979

19.6

3.2

68,959

11,259

496

102

1980

42.7

11.0

150,231

38,701

682

116

1981

16.0

2.7

56,293

9,499

586

100

1982

4.7

1.1

16,536

3,870

1,002

134

1983

26.3

6.6

92,531

23,221

846

154

1984

13.0

2.1

45.738

7,388

1,220

134

1985

17.8

2.7

62,626

9,499

1,110

192

1986

17.0

2.8

59,811

9,851

784

142

1987

44.7

15.1

157,268

53,126

470

can shad has a similar 4-6 yr postjuvenile phase as
sockeye salmon, we tested Peterman's hypothesis
for American shad with key factor analysis (Bellows
1981; Rosenberg and Doyle 1986). We fixed the
postjuvenile period for American shad in the Con-
necticut River between 101 days and 5 years, corre-
sponding to the average age (101 days) at which
juvenile shad leave the river (Crecco and Savoy
1985b) and the average age (5 years) when they
return to the river as mature adults. The total post-
juvenile (age 101 days to 5 years) mortality rates
(ZA() for the 1967-82 year classes were related to
the scaled juvenile indices (J<) for those year
classes (Table 2) in a linear model:

ZA, =

a +

b{Jt),

(1)

where ZAf = -logg(i2(/J<). If significant density-
dependent mortality is present during the post-
juvenile stage, the slope (6) of Equation (1) would
be positive and differ significantly (P < 0.05) from
zero.

The juvenile indices were scaled to thousands of
fish (Jf) so that the y-axis intercept in Equation (1)
directly estimates the mean density-independent
mortality (Zj^g^i), and the slope (b) times the geo-
metric mean juvenile abundance (GM) from 1967
through 1987 is the mean density-dependent mor-
tality rate (Z^post). Total juvenile abundance (J()

was estimated by multiplying the juvenile indices
(IND,) by a scalar (SC):

SC = [GR exp{-EZA) ■ GM] = 3,518.3, (2)

where GR is the geometric mean total adult
recruitment for 1967 through 1982; GM is the geo-
metric mean juvenile index from 1967 through 1987;
EZA (4.85) is the mean total instantaneous mortality
among postjuveniles from 1967 through 1982. We
estimated EZA as the sum of mortality during the
late juvenile period (age 101-365 days), and the
subadult stage (age 1-5 years). The mean total in-
stantaneous mortality rate during the late juvenile
stage was estimated as 2.65 (0.01 • 265 days), using
a mean daily mortality rate of 0.01 (SE = 0.002)
extrapolated from the 1979-84 larval and juvenile
survivorship curves (Crecco and Savoy 1985b). The
mean total instantaneous mortality rate of subadult
shad was 2.2, based on an annual instantaneous
natural mortality rate of 0.45 from the method of
Pauly (1980) plus 0.10 to reflect oceanic fishing mor-
tality (2.2 = 4(0.45 -I- 0.10)). We estimated the an-
nual natural mortality rate (0.45) by substituting the
;^(0.25, SE = 0.03) and L (55 cm FL, SE = 3 cm)
parameters of the von Bertalanffy equation for male
and female shad combined and preferred ocean tem-
perature (14 °C) of American shad (Leggett and
Whitney 1972) into Pauly' s multiple regression
model. The oceanic fishing mortality estimate (0.10)
was based on tagging studies in Delaware Bay
(White et al. 1969; Zarbock 1969) and off the New
York-New Jersey coast (Nichols 1958).

To determine if density-dependent mortality takes
place during the egg and larval stages as Cushing
(1980) hypothesized, we related total prejuvenile
mortality rates {ZEJt) to annual egg production
(EggSj ), and to both egg production (Table 2) and
mean June river flow (JFLOW) from 1967 to 1987
(Table 1) in linear regression models:

ZEJt = a -H 6 (Eggs,)

(3a)

and

ZEJt = a + ^(Eggs,) -H c(JFLOW,), (3b)

where ZEJt = -log^(J( /Eggs,). Mean June river
flows (m^/s) were included in Equation (3b) because
previous studies (Crecco and Savoy 1985b, 1987a)
have shown that high June flows reduced prejuvenile
survival rates, leading to a significant inverse cor-
relation (r = - 0.74, P < 0.01) between the juvenile
indices of abundance from 1967 through 1980 and

470

SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD

mean June river flow (Crecco and Savoy 1984,
1985b). Mean June flows (m^/s) were measured
within the major spawning areas (Leggett 1977) by
the United States Geological Survey (U.S. Geo-
logical Survey 1967-84). We estimated total egg
production (Eggs,) in Equations (3a) and (3b) as the
product of the mean fecundity of a female American
shad times that year's parent stock (PARf) of
female shad (Table 1). The average fecundity was
reduced from 269,000 ova (Leggett 1969) to 200,000
to reflect the average rates of egg retention and in-
complete fertilization (Watson 1970; Reed and Russo
1976^). Since Leggett (1969) showed that the aver-
age fecundity of American shad varied by less than
10% from 1966 through 1973, we were justified in
using an average fecundity for all years.

The mean density-independent mortality rate
(Z/pre) present during the prejuvenile stage was the
y-axis intercept (a) in Equation (3b) plus the slope
(c) times the overall geometric mean June flow

'Reed, R. J., and A. Russo. 1976. American shad research
Connecticut River, Massachusetts, 1976. 1. Fecundity, egg reten-
tion, sex ratio, and age class composition. Unpubl. manuscr., 16
p. Massachusetts Cooperative Fisheries Research Unit, Univer-
sity of Massachusetts, Amherst, MA 01003.

(GJFLOW) from 1967 through 1987:
Zipre = a + c(GJFLOW),

(4)

The mean density-dependent mortality rate (Zq^^^)
was the slope (6) of Equations (3a) and (3b) times
the geometric mean egg production (GEgg) from
1967 through 1987. A positive and statistically sig-
nificant slope (6) of prejuvenile mortality on egg
abundance would support the Gushing hypothesis
that density-dependent mortality takes place before
the juvenile stage. If the b estimates of Equations
(1), (3a), and (3b) were all positive and significant,
our results would support Gulland's (1965) hypoth-
esis that density-dependent mortality occurs over
the entire prerecruitment phase.

Results

Although total juvenile abundance varied 14-fold
from 1967 through 1982 (Table 2), the total post-
juvenile mortality rates {ZA() exhibited relatively
low variability (95% G.I.: 4.70-5.00) about the esti-
mated mean {EZA = 4.85) (Fig. 2). The slope of the
linear regression between total postjuvenile mortal-

6.6

5.4

5.2

5.0

N 4.8

4.6

4.4

4.2

4.0

1 1 1

100 160 200

Juvenie Abundance (thousands)

260

300

Figure 2.— Relationship between American shad postjuvenile mortality (ZA,) and relative abundance of juveniles

(J() from 1967 through 1982.

471

FISHERY BULLETIN: VOL. 86, NO. 3

ity rates (ZAt) and scaled juvenile abundance (J,)
was positive and statistically significant (P < 0.04)
(Table 3). Since the average density-dependent
mortality (Z^post = 0.30) present during the post-
juvenile stage comprised only 6% of the mean total
postjuvenile mortality rate {ZA = 4.85), density-
dependent mortality is not large enough to alter the
significant linear correlation (r = 0.78, P < 0.01)

betv^^een the juvenile indices (J^ ) and adult recruit-
ment (Rf) for the 1967 through 1982 year classes
(Fig. 3).

The prejuvenile mortality rates (ZEJ,) from 1967
through 1987 v^ere positively correlated (r = 0.70,
P < 0.0004) to total egg production (EggS() (Fig. 4).
Egg production alone explained 49% of the varia-
tion in prejuvenile mortality (Table 3) and the slope

Table 3.— Estimates of density-dependent (Zp), density-independent mortality (Z,) and the fraction of density-
dependence (ZqIZi^Zq) during the postjuvenile (1967-82) and prejuvenile (1967-87) mortality phases for
American shad in the Connecticut River. SE = standard error, and numbers in parentheses = 95% C.I. for
Zq,ZEJ, = prejuvenile mortality, Z4, = postjuvenile mortality.

Model

Parameters

Z, Zol{Z,,Z^)

Prejuvenile
ZEJ, = a + b(Eggs,)

ZEJt = a + b(Eggs,)
+ c (J Flow,)

Postjuvenile
ZA, = a + b(J,)

a = 5.466

b = 1.88 X 10"^

r^ = 0.49

a = 4.287

b = 1.97 X 10"^

c = 0.0028

r^ = 0.80

a = 4.53

b = 3.49 X 10"

r^ = 0.275

0.313

4.43 X 10"^

0.303

2.87 X 10"^

0.0005

0.15

1.51 X 10"^

1.13 5.47 0.17

(0.862-1.392) (0.136-0.203)

1.18 5.34 0.18

(1.007-1.351) (0.158-0.202)

0.30 4.53 0.062

(0.035-0.560) (0.008-0.110)

1300

« 1025

c
<0
(0

I 750

< 475

200

— n
10

—J-
20

— I—
30

— r-
40

— I—
50

— I—
60

Juvenle Index

Figure 3.— Relationship between total adult recruitment (R,) and the juvenile abundance indices (/ATZ), ) of American

shad from 1967 through 1982.

472

SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD

of the regression differed significantly from zero,
indicating the presence of significant density-
dependent mortaHty during the prejuvenile stage.
When prejuvenile mortality rates were related to
both egg production (Eggs,) and mean June river
flow, (JFLOW,), the multiple regression model ac-
counted for 80% of the variability in prejuvenile
mortality (Table 3), and the slope estimates for egg
production (6) and June flow (c) were positive and
highly significant. Note that the standard error (SE)
about the slope estimate (6) was reduced by 60%
when June flow effects were considered. These
results suggest that prejuvenile mortality rates are
affected by a combination of density-dependent (egg
production) and density-independent (June river
flow) factors.

The mean density-dependent mortality rate dur-
ing the prejuvenile phase (Z^pj.^ = 1.18) from the
multiple regression model was four times greater
than the mean Z^post value (0.30) for postjuveniles
(Table 3), suggesting that 80% (1.18/1.18 + 0.30)
of the total compensatory reserve for American shad
occurs before the juvenile stage. Whereas compen-
satory density-dependent mortality may play a sig-
nificant role in regulating egg and larval abundance
during years of high egg production, it is clear that

most of the variability (82%) in egg and larval abun-
dance is ascribed to density-independent factors.

Egg, Larval, and Juvenile
Mortality, 1979-87

Analysis

To examine how larval and juvenile mortality
rates varied with year-class strength, survivorship
curves for 19,000-180,000 American shad larvae
and 800-3,500 juveniles were developed annually
from 1979 to 1987 following the techniques of Lough
(1976) and Hewitt et al. (1985). Mortality rates could
not be estimated directly for prolarvae because
American shad yolk-sac larvae (7-9 mm TL) remain
in deep water (Marcy 1976) and were only partially
susceptible to the plankton seine.

All larvae and juveniles collected annually were
separated into four length intervals: 1) 10-13 mm
TL, reflecting first- feeding larvae with undeveloped
pelvic fins (Wiggins et al. 1984); 2) 14-19 mm TL,
associated with the onset of pelvic fin development
(Lippson and Moran 1974); 3) 20-28 mm TL, repre-
senting larvae approaching metamorphosis, char-
acterized by invagination of the gut (Maxfield 1953);

9.0

W 7.0

6.0

6.0 i

T-
20

— r-
30

— r-
40

I
60

— I—
60

— r-
70

I
80

— I—
90

100

120

130

Total Eggs (bMons)

Figure 4.— Relationship between American shad prejuvenile mortality (ZEJ^) and the number of eggs produced

(Eggs,) from 1967 through 1987.

473

FISHERY BULLETIN: VOL. 86, NO. 3

and 4) 29-80 mm TL, corresponding to the juvenile
stage and the onset of scale development (Marcy
1976). These four groups are hereafter referred to
as early larval, midlarval, late larval, and juvenile
stages, respectively.

We used otolith increment counts (Savoy and
Crecco 1987) to age a subsample of 100-400 Ameri-
can shad larvae and 80-200 juveniles annually from
1979 to 1987. The age-length relationship for lar-
vae and juveniles in all years was well described by
the Gompertz growth equation (Crecco et al. 1983;
Crecco and Savoy 1985b) so we used a pooled
Gompertz equation to describe larval and juvenile
growth. The average age (t) of each of the four
length groups was estimated by rearranging the
Gompertz equation:

t = logAKIK - log,(L,(Lo))]/a

(5)

where L( = total length (mm); Lq = length at
hatching (8.0 mm); K = growth rate at the inflec-
tion point; and a = rate of exponential decay.

The daily instantaneous mortality rates (d) for
each stage and their standard errors from 1979
through 1987 were estimated by an exponential
model,

Nt = A exp{-dt),

(6)

that related abundance (A^^) and age (t) by non-
linear regression methods (SAS 1982). The total
instantaneous mortality rates of early (ZELf), mid
(ZMLf), and late (ZLLt) larvae and juveniles (ZJf)
were estimated by multiplying the corresponding
daily mortality rate (d) by the average duration (Ds
= days) within each stage. Previous studies (Crecco
and Savoy 1985a; Savoy and Crecco 1987) found
that the growth rates of early (10-13 mm) and mid-
larvae (14-19 mm) were positively correlated with
June water temperatures, whereas the growth of
late larvae and juveniles were independent of river
temperatures. As a result, we adjusted the stage
duration {Ds ) of early and midlarvae with the mean
June temperatures (U.S. Geological Survey 1979-
87) from 1979 through 1987 and the larval growth-

subtracting the sum of larval {ZEL, , ZMLf , ZLL,),
juvenile (ZJ^) and postjuvenile (ZA^) mortality rates
from the total prerecruitment mortality rate
(ZTotal,):

ZEt = ZTotalt - [ZELi + ZML, + ZLLt

+ ZJt + ZAt), (7)

where ZTotal^ = -logp(/2,/Eggs,). However, since
the 1983 through 1987 year classes have not been
fully recruited to the spawning population, we esti-
mated total adult recruitment {Rp^ ) for those year
classes by the following environment-dependent
stock-recruitment model:

Rpt = 24.29 (PARt) exp(- 0.0052 • P,,)

exp(- 0.0032 • PJ
exp(- 0.0025 • JFLOW()

(8)

Adult recruitment was estimated independently
of the juvenile indices by substituting each
year's parent stock size (PAR^), mean June flow
(e/FLOW), female parent stock lifted over the Hol-
yoke Dam (P„ ), and female parent stock below the
Holyoke Dam (P^) (Table 1) into the model. This
nonlinear model was shown (Lorda and Crecco 1987)
to be a good predictor (r^ = 0.81, P < 0.001) of
adult recruitment (P^) for the 1966 through 1982
year classes. Moreover, the predicted recruitment
levels (P() of the 1983 through 1987 year classes
from Equation (8) were closely correlated (r = 0.92,
P < 0.01) with the corresponding juvenile abundance
(J() for those years (Table 4) which is consistent
with the positive correlation (r = 0.78, P < 0.01)
between adult recruitment and the 1967-82 juvenile
indices (Fig. 2). This justifies the use of Equation
(8) to predict adult recruitment, total mortality, and
postjuvenile mortality rates for the 1983-87 year
classes.

The standard errors about the egg mortality rates
(ZEt) were derived as the sum of the variances of
all other terms (Cochran 1965):

SEze' = ^SEzel' + SE^ml' + S^zll' + ZJ^ + SE^/ + SE^xotai' 0)

temperature equations (Crecco and Savoy 1985b:
table 7).

We estimated total egg and prolarval mortality
rates (ZE^) indirectly from 1979 through 1987 by

The standard errors about the total prerecruitment
(ZTotal,) and postjuvenile (ZA^) mortality rates
(Table 5) were based on the same principle as Equa-
tion (9) (App. 1).

474

SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD

Table 4— Predicted total adult recruitment of
American shad from the environmental dependent
stock-recruitment model (Equation (8)), 95% con-
fidence limits about the recruitment values and
juvenile abundance for the 1983-87 year classes.

Predicted

adult

Juvenile

recruitment

abundance

Year

X 10^

95% C.I.

X 10^

1983

634

362-906

92,531

1984

196

112-280

45.738

1985

555

317-793

62,626

1986

587

335-839

59,811

1987

955

547-1,363

157,268

Table 5.— Estimates of the total (ZTotal,) and
postjuvenile (Z4,) instantaneous mortality rates
and their standard errors (SE) for American
shad from 1979 through 1987.

Year

ZTotal,

ZA,

1979

10.94

0.26

4.36

0.23

1980

10.90

0.24

4.78

0.32

1981

11.26

0.25

4.31

0.25

1982

13.00

0.27

4.77

0.34

1983

11.80

0.30

4.98

0.36

1984

13.34

0.26

5.45

0.28

1985

12.21

0.29

4.72

0.28

1986

11.80

0.30

4.62

0.29

1987

10.80

0.28

5.10

0.47

Given that ZE represents total mortality through-
out the egg and prolarval period, daily egg mortal-
ity rates were determined by dividing ZE by the
average duration (Ds) of the egg and prolarval
stages of American shad in the Connecticut River.
Watson (1968) reported an inverse relationship be-
tv^^een the incubation period (D ) of shad eggs and
water temperature (T) by the expression:

D = 120.95 exp(-0.154 • T).

(10)

To determine the duration {Ds) of the egg stage
from 1979 through 1987 we substituted the mean
June temperatures for the Connecticut River
(U.S.G.S. Annual Water Year Reports 1979-87) into
Equation (10).

The total number of prolarvae (age 2 days) and
early larvae (age 10 days) for the 1979 through 1987
year classes was estimated by

NE^ = Eggs, • exp(-Z£;,) (11)

and

NFt = Eggs, • exipi-ZEt - ZELi) (12)

respectively, where NEt is the estimated number
of prolarvae larvae and NFt is the number of early
larvae. The abundance of older larvae and juveniles
from each year class was estimated by adding their
respective total mortality rates to Equation (12). To
determine the life stage(s) at which year-class
strength is established, we related the stage-specific
total mortality rates and abundance estimates to the
number of adult recruits (i?, or Rpt) from the
1979-87 year classes in several linear models. If
year-class strength is established early, there should
be a significant positive correlation between the
number of prolarvae (NE) and early larvae (NF) and
adult recruitment, and total mortality rates during
these early stages {ZE and ZEL ) should be inverse-
ly related to adult recruitment.

To determine the extent of density-independent
mortality at the egg, larval, and juvenile stages, we
correlated the stage-specific total mortality rates
from 1979 through 1987 to mean May and June river
flows (m^/s) and water temperatures (°C) in sev-
eral linear models. May and June hydrographic
and meteorological parameters were used because
they coincide with egg, larval, and juvenile develop-
ment in the Connecticut River (Leggett 1977) and
were the only monthly abiotic variables that were
significantly linked to adult shad recruitment from
1966 to 1980 (Crecco and Savoy 1984, 1987c). May
and June water flows and temperatures were re-
corded by the U.S. Geological survey (U.S. Geo-
logical Survey 1966-1980) within the major spawn-
ing areas (river km 89) of American shad (Leggett
1977).

Results

The mean daily mortality rates from 1979 through
1987 declined by an order of magnitude from the
egg through the juvenile stages (Table 6). Total egg
mortality rates were relatively high (mean ZE =
2.584, cv = 18.5%) and were inversely correlated
(r = - 0.76, P < 0.03) with adult recruitment (Rt or
Rpt) from those year classes (Table 7). These data
indicate that shad eggs and prolarvae (age 1-2 days)
from 1979 through 1987 experienced high (85-96%)
mortality that was directly linked to year-class
success.

The total mortality rates (ZEL) among early lar-
vae were slightly lower and more variable (mean
ZEL = 1.608, cv = 31.5%) than the egg mortality
rates (Table 6), and were also inversely related (r
= - 0.83, P < 0.01) to adult recruitment from 1979
through 1987. By contrast, the total mortality rates
of older larvae {ZML and ZLL) and juveniles (ZJ)

475

FISHERY BULLETIN: VOL. 86, NO. 3

Table 6.— Stage-specific total instantaneous mortality rates of egg, larval, and juvenile
American sfiad from 1979 to 1987. SE = stage-specific standard errors; cv = coeffi-
cient of variation, Days = duration of each stage.

Instantaneous total mortality

Egg + pro-

Early

Mid-

Late

larvae

Juvenile

Year
or

(day 2)

(day

3-9)

(day 10-18)

(day 19-29)

(day 30

1-100)

statistic

mean

1979

2.481

0.554

1.323

0.245

1.116

0.189

0.330

0.110

1.330

0.280

1980

1.853

0.819

1.218

0.186

1.096

0.336

0.693

0.099

1.260

0.280

1981

2.580

0.684

1.836

0.108

0.408

0.328

0.726

0.154

1.400

0.210

1982

2.940

0.501

2.144

0.152

1.160

0.210

0.726

0.187

1.260

0.210

1983

3.012

0.785

1.400

0.546

0.522

0.234

0.836

0.143

1.050

0.140

1984

2.706

0.601

2.520

0.352

0.640

0.330

0.484

0.242

1.540

0.280

1985
1986

N 0
0.196

DATA
1.206

1.330
1.120

0 pao

3.158

0.652

1.323

0.261

0.374

0.066

0.280

1987

1.944

0.852

1.098

0.180

0.864

0.128

0.464

0.088

1.330

0.210

2.584

1.608

0.877

0.579

1.286

0.169

0.179

0.111

0.066

0.051

0.185

0.315

0.359

0.325

0.120

Days

7.2

6.9

8.9

11.0

70.0

Daily Z

0.359

0.233

0.099

0.053

0.018

Table 7.— Correlations between stage-specific total mortalities and
adult recruitment and several abiotic factors from 1979 to 1987 for
American shad. P = probability levels.

Stage

Adult
recruit-
ment

May
flow

May
tempera-
ture

June
flow

June
tempera-
ture

Egg
p

-0.76
0.03

0.38
0.35

-0.08
0.85

0.66
0.08

-0.59
0.13

Early
P

-0.83
0.01

0.42
0.29

-0.21
0.62

0.81
0.01

-0.64
0.09

Mid
P

0.16
0.71

-0.41
0.32

0.49
0.22

0.09
0.83

-0.34
0.42

Late
P

-0.02
0.96

0.20
0.64

-0.51
0.20

-0.13
0.75

0.18
0.68

Juv

-0.19
0.65

0.05
0.91

0.02
0.97

0.26
0.53

-0.08
0.86

were highly variable, but showed no relationship to
adult recruitment.

Given that high (94-99%) egg and early larval
mortality rates were inversely related to year-class
strength, the abundance of first-feeding larvae (age
10 days) was closely correlated (r = 0.84, P < 0.01)
with adult recruitment (Rpt) from 1979 through
1987 (Table 8). Because year-class strength from
1967 through 1987 was independent (r = - 0.15, P
< 0.46) of egg production (EggS(), our results sug-
gest that year-class strength of American shad is
determined by the number of shad that survive the
embryonic and early larval stages.

Early larval mortality rates from 1979 to 1987
were positively correlated with mean June river

flows. No such relationships were evident for any
other stage-specific mortalities (Table 7). The high
egg and early larval mortality rates (Table 6) and
relative failure of the 1982 and 1984 year classes
coincided vdth major storm events in mid- June 1982
and late May 1984 (Crecco and Savoy 1984) which
increased river flows and kept water temperatures
below 17°C until late June. The high river flows in
1982 and 1984 were also coupled with the highest
parent stocks on record since 1967 (Table 1). By con-
trast, the relatively low egg and early larval mor-
tality rates for the dominant 1980 and 1987 year
classes were associated with low June river flows,
a steady rise in June water temperatures (Crecco
and Savoy 1984) and low to moderate size parent
stocks (Table 1). These results suggest that domi-
nant year classes of American shad are most likely
to occur when lower than normal June flows are
coupled with relatively small spawning stocks.

Prejuvenile Density-Dependent
Mortality: A Closer Look

Analysis

Having estimated that 82% of the density-
dependent mortality for American shad takes place
before the juvenile stage (Table 3), we attempted
to estimate the relative contribution of density-
dependent mortality during the egg, early, mid-, and
late larval and juvenile stages from 1979 through
1987. We related egg mortality rates (ZEi) from

476

SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD

Table 8.— Relationship between the number (x 10®) of American shad at each life history
stage and adult recruitment from each year class, r = correlation coefficient and P =
probability levels from 1979 to 1987.

Stage

Year-

Pro-

Early

Mid-

Late

class

Egg

larvae

Juvenile

Adult^

1979

496

41.5

11.0

3.6

2.6

0.7

882

1980

682

106.9

31.6

10.6

5.3

1.5

1,256

1981

586

44.4

7.1

4.7

2.3

0.6

758

1982

1,002

53.0

6.2

1.9

0.9

0.3

282

1983

846

41.6

10.3

6.1

2.6

0.9

634

1984

1,220

81.5

6.6

3.5

2.1

0.5

198

1985
1986

1,110
784

NO DATA-

555
587

33.3

8.9

2.7

1.8

0.6

1987

470

67.3

22.4

9.5

5.9

1.6

955

r =

-0.15^

0.33

0.84

0.80

0.79

0.82

P =

0.46

0.43

0.01

0.02

0.01

'Predicted or observed recruitment in thousands.

^Correlation coefficient for egg abundance and recruitment was based on the 1967-87 data (Tables
1.2).

1979 to 1987 to total egg production (Eggs^) in a
linear model:

ZEt = a + 6(Eggs,).

(13)

As before, significant density-dependence would be
shown if the slope (6) in Equation (13) was positive
and differed significantly from zero.

The cumulative amount of density-dependent mor-
tality during the egg and early larval stages com-
bined was estimated by summing the instantaneous
total mortality rates of early larvae (ZELf) and
eggs (ZEt) and then regressing the total {ZEf +
ZELf) against egg production as in Equation (13).
This procedure was repeated for each subsequent
stage by adding their respective instantaneous total
mortality rates. We then estimated the relative
magnitude of density-dependent mortality (Z^) at
each stage by multiplying the slope (6) of each
regression equation by the geometric mean egg pro-

duction (GEgg) from 1967 through 1987. The
density-independent mortality rate {Zj) at each
stage was expressed by the y-a.xis intercept (a) of
each regression equation. The percentage contribu-
tion of density-dependent mortality {%Zf)) for each
period was the ratio of Z^ to total mortality {Z^ +
Zj) times 100.

Results

Our results showed that significant density-
dependent mortality first occurs during the early
larval stage and persists for all stages thereafter
(Table 9). The percentage contribution of density-
dependent mortality i%Z^) rose from 24% of the
total during the egg stage to 41% during the early
larval stage, and then declined during the mid- and
late larval stages as the magnitude of density-
independent mortality increased. The mean density-
dependent mortality rate for the egg and early larval

Table 9.— Relationship between the cumulative mortality rates and parent stock sizes (Eggs,).
The cumulative mortality rate was egg mortality (ZE) plus early larval mortality (Z2), midlarval
mortality (Z3), late larval mortality (Z4), and juvenile mortality (Z5). Zg = density-dependent

mortality rate, Z, = density-independent mortality rate

o/oZo =

percentage density-dependent

mortality, %Z, = percentage density-independent mortality, t = student /-statistic.

Model

b
parameter

SE(b)

o/oZo

o/oZ,

= a

+ b(Eggs,)

9.89

10-3

6.40 X

10"

-9

1.55

0.59

1.83

= a

+ fa (Eggs,)

2.57

io-»

7.76 X

10"

-a

^3.31

1.54

2.24

= a

+ b(Eggs,)

2.44

io-«

8.16 X

10"

-y

'2.99

1.46

3.21

= a

+ 6(Eggs,)

2.60

io-«

7.74 X

10"

-y

'3.36

1.55

3.67

= a

+ b(Eggs,)

2.69

10"®

7.50 X

10"

-y

'3.59

1.61

4.88

'Significant student f-statistic at the P < 0.01 level and SE(b) = standard error of b.

477

stages combined (Zq = 1.54) between 1979 and
1987 was not significantly different (Table 9) from
the total density-dependent mortality rate (Zq^j.^ =
1.18) estimated during the prejuvenile period from
1967 to 1982 (Table 3), suggesting that nearly all
of the density-dependent mortality for prejuvenile
American shad occurs before the midlarval stage.

DISCUSSION

Although American shad eggs and early larvae
experience high mortality (15-40%/day) in the Con-
necticut River, the average density-dependent mor-
tality rate (Z^p^g = 1.18) during those stages com-
prised a relatively small percentage (18%) of the
total prerecruitment mortality. This suggests that
most of the annual variability in American shad
recruitment is explained by density-independent fac-
tors, which is consistent with the significant positive
correlation between mean June flow and egg and
early larval mortality rates, and with the significant
inverse correlation (r = - 0.74, P < 0.001) between
mean June river flow and adult recruitment from
the 1967 through 1982 year classes (Crecco and
Savoy 1984, 1987b). Whereas these data illustrate
that density-dependent mortality plays a minor role
in governing annual variability in American shad
recruitment, this does not mean that compensatory
processes are trivial. As pointed out by Ricker (1954)
and others (Gushing 1974; Garrod and Horwood
1984), only a small amount of density-dependent
mortality is required to stabilize the growth poten-
tial of fish populations because density-dependent
effects become progressively more effective at
higher egg and larval densities (Shepherd and
Gushing 1980; Murray 1982).

Although our results on American shad support
the Gulland (1965) hypothesis that density-depend-
ent mortality persists throughout the prerecruit-
ment period, most (82%) of the density-dependent
mortality occurs during the egg and early larval
stages. Our average estimate of density-dependent
mortality {Zq = 1.54) during the early larval
periods does not differ significantly from the mean
Zd value (1.34) estimated among the 5-d larval
cohorts in 1983 and 1984 (Grecco and Savoy 1987a),
or from the mean density-dependent mortality rate
Zj) (1.21) estimated by stock-recruitment methods
for the entire prerecruitment period (Lorda and
Grecco 1987). That year-class strength is established
early in the ontogeny of American shad is supported
further by the significant positive correlation (r =
0.84, P < 0.009) between the relative abundance of
early larvae and all subsequent stages from 1979

FISHERY BULLETIN: VOL. 86, NO. 3

through 1987 and adult recruitment from those year
classes (Table 8).

The main causes of density-dependent mortality
are thought to be predation, competition, and canni-
balism (Ricker 1954). Since adult American shad are
not thought to feed in freshwater (Walburg and
Nichols 1967), we can probably eliminate cannibal-
ism as a mechanism for significant density-depend-
ent mortality during the egg and early larval stages.
Therefore, density-dependent mortality among early
American shad larvae is most likely caused by intra-
specific competition for food or space and predation.
The exact underlying density-dependent mortality
mechanisms are difficult to quantify because the out-
come of competition and predation may depend on
June flow effects shown here (Table 3) and else-
where (Grecco and Savoy 1987a, b) to be the prin-
cipal density-independent factor. High June river
flows have been shown to reduce June river tem-
peratures (Grecco and Savoy 1984) and the growth
rates of shad eggs (Watson 1968) and early larvae
(Grecco and Savoy 1985b). Because slower growing
larvae may be susceptible to predation for a longer
period of time (Hunter 1976), periods of high flow
may indirectly enhance egg and larval predation.
Additionally, since high flows reduce the spatial
patchiness (Grecco and Savoy 1987a), abundance
and availability of river zooplankton (Whitton 1980;
Threlkeld 1986), high flows may result in increased
levels of competition among American shad larvae
for available prey, especially if shad larvae are
capable of depleting local aggregations of edible
zooplankton. Periods of high runoff that coincide
with peak larval production, such as in June 1982
and 1984, may advect larvae and their zooplankton
prey from eddies and backwaters where they are
normally found (Gave 1978) to areas of high pred-
ator abundance. In light of the many ways in which
June flows potentially mediate larval mortality, it
is unlikely that a single compensatory mechanism
is responsible for the relatively high density-depend-
ent mortality rate (Z^pre) among early larvae.

Since egg and prolarval shad have endogenous
food reserves, density-dependent mortality of these
stages is likely due to predation and competition for
sites among spawning adults. As spawning stocks
reach high densities, such as in 1982-84, crowding
of adult fish on the spawning grounds may result
in a reduction in the number of eggs released ((jood-
year 1980). In addition, since the amount of spawn-
ing habitat in the Connecticut River can be con-
sidered fixed from year to year, larger spawning
stocks are more apt to deposit an increasing per-
centage of eggs in unfavorable areas. Layzer (1974)

478

SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD

noted that the highest survival rates of American
shad eggs occurred over gravel and rubble sub-
strates. Since these areas generally have higher
water velocities, eggs lodged within the interstices
of the gravel and rubble are less likely to be eaten
by predators or covered by silt. Dense aggregations
of eggs may also be highly susceptible to fungal
agents (Leach 1925), which could then act as effi-
cient density-dependent predators. However, since
predators of American shad eggs and larvae in the
Connecticut River have not yet been identified,
density-dependent losses are difficult to quantify.
Also, predation mortalities in general may be either
density-dependent or density-independent, depend-
ing on the functional response between the predator
and the prey (Hassell 1978), and whether predators
actively search for discrete aggregations of eggs and
larvae or prey on randomly encountered eggs and
larvae (Gulland 1987).

In the absence of direct egg mortality estimates
of American shad, the accuracy of our egg mortal-
ity estimates {ZEt) from Equation (7) are difficult
to evaluate, particularly since the standard errors
about the estimates are highly variable (Table 6). To
provide an independent estimate of daily egg mor-
tality, we used the multiple regression method of
McGurk (1987) to estimate the average daily egg
mortality rate. This method requires an estimate of
the patchiness and the mean dry weight of a shad
egg. We derived Lloyd's patchiness index (x =
7.26, SE = 0.376) for American shad eggs in the
Connecticut River from the 1974-75 egg surveys in
the Holyoke Dam impoundment (NUSCo 1977). Sub-
stituting the mean weight of a shad egg (0.00025
g, Leggett 1969) and the mean patchiness index into
McGurk's equation 4, yields an instantaneous daily
egg mortality rate of 0.34 which closely approx-
imates our average estimate (0.36) from 1979 to
1987 (Table 6).

Lastly, since it is presently unknown if younger
(smaller) virgin female American shad produce
smaller, less viable eggs than older shad as was
reported for cod, Gadus morhua, (Knutsen and
Tilseth 1985), the contribution of the age structure
to the parent progeny relationship (Rosenberg and
Doyle 1986) should be considered. This issue and an
examination of whether egg deposition rates are
density-dependent should be addressed by monitor-
ing egg retention rates among postspawning Ameri-
can shad.

ACKNOWLEDGMENTS

We would like to thank Rita Lorenzetti-Langan,

Mary Payette, Thomas Stanford, and all the other
people who have contributed field and laboratory
time or helpful advice. We also thank the two
anonymous reviewers for their comments on the
manuscript.

LITERATURE CITED

Bakun, a.

1984. Report of the working group on environmental studies
and monitoring. In J. Csirlce and G. Sharp (editors),
Reports of the expert consultation to examine changes in
abundance and species composition of neritic fish resources,
p. 41-54. FAO Fish. Rep. 291 Vol. 1.
Bellows, T. S., Jr.

1981. The descriptive properties of some models for density-
dependence. J. Anim. Ecol. 50:139-156.
Cave, J. R.

1978. American shad, Alosa sapidissima, larval distribution,
relative abundance and movement in the Holyoke Pool, Con-
necticut River, Mass. M.S. Thesis, Univ. Massachusetts,
Amherst, MA, 65 p.
Cochran, W. G.

1965. Sampling Techniques. 2d ed. John Wiley & Sons,
Inc., N.Y., 413 p.
Crecco, V. A., AND T. Savoy.

1984. Effects of fluctuations in hydrographic conditions on
year-class strength of American shad {Alosa sapidissima)
in the Connecticut River. Can. J. Fish. Aquat. Sci. 41:
1216-1223.

1985a. Density-dependent catchability and its potential
causes and consequences on Connecticut River shad, Alosa
sapidissima. Can. J. Fish. Aquat. Sci. 42:1649-1657.

1985b. Effects of biotic and abiotic factors on growth and
relative survival of young American shad in the Connecticut
River. Can. J. Fish. Aquat. Sci. 42:1640-1648.

1987a. Effects of climatic and density-dependent factors on
intra-annual mortality of larval American shad. Am. Fish.
Soc. Symp. 2:69-81.

1987b. Recruitment mechanisms of the American shad, Alosa
sapidissima: test of the critical period and match-mismatch
hypothesis. Am. Fish. Soc. Symp. 1:455-468.
Crecco, V. A., T. Savoy, and L. Gunn.

1983. Daily mortality rates of larval and juvenile American
shad {Alosa sapidissima) in the Connecticut River with
changes in year-class strength. Can. J. Fish. Aquat. Sci.
40:1719-1728.

Crecco, V., T. Savoy, and W. Whitworth.

1986. Effects of density-dependent and climatic factors on
American shad, Alosa sapidissima, recruitment: a predic-
tive approach. Can. J. Fish. Aquat. Sci. 43:457-463.
Gushing, D. H.

1974. The possible density dependence of larval mortality and
adult mortality in fishes. In J. H. S. Blaxter (editor), The
early life history of fish, p. 21-38. Springer- Verlag, Berlin.
1980. The decline of the herring stocks and the gadoid out-
burst. J. Cons. Int. Explor. Mer 39:74-85.
FOOTE, P. S.

1976. Blood lactic acid levels and age structure of American
shad {Alosa sapidissima, Wilson) utilizing the Holyoke Dam
fish lift, MA, 1974 and 1975. M.S. Thesis, Univ.
Massachusetts, Amherst, MA, 97 p.
GaRROD, D. J., AND J. W. Horwood.

1984. Reproductive strategies and the response to exploita-

479

FISHERY BULLETIN: VOL. 86, NO. 3

TION. In G. W. Potts and R. J. Wooton (editors), Fish
reproduction: strategies and tactics, p. 367-384. Acad.
Press, London.
Goodyear, P. C.

1980. Compensation in fish populations. In C. H. Howatt
and J. R. Stauffer, Jr. (editors). Biological monitoring of fish,
p. 253-280. Lexington Books, Lexington, MA.
GULLAND, J. A.

1965. Survival of the youngest stages of fish, and its rela-
tion to year-class strength. Spec. Publ. ICNAF 6, p.
363-373.
1987. Natural mortality and size. Mar. Ecol. Prog. Ser. 39,
p. 197-199.
Hassell, M. p.

1978. The dynamics of arthropod predator-prey systems.
Princeton Univ. Press, Princeton, NJ, 237 p.
Hewitt, R. P., G. H. Theilacker, and N. C. H. Lo.

1985. Causes of mortality in young jack mackerel. Mar.
Ecol. Prog. Ser. 26, p. 1-10.
Hunter, J. R.

1976. Culture and growth of northern anchovy, Engraulis
mordax, larvae. Fish. Bull., U.S. 74:81-88.
Jones, R. A., P. Minta, and V. A. Crecco.

1976. A review of American shad studies in the Connecticut
River. In Proceedings Workshop on American Shad, p.
135-164. U.S. Natl. Mar. Fish. Ser.
Knutsen, G. M., and S. Tilseth.

1985. Growth, development, and feeding success of Atlantic
cod larvae Gadiis rrwrhua related to egg size. Trans. Am.
Fish. Soc. 114:507-511.
Layzer, J. B.

1974. Spawning sites and behavior of American shad, Alosa
sapidissima, (Wilson), in the Connecticut River between
Holyoke and Turners Falls, MA, 1972. M.S. Thesis, Univ.
Massachusetts, Amherst, MA, 46 p.
Leach, G. C.

1925. Artificial propagation of shad. Rep. U.S. Comm.
Fish., 1925, app. VnL459-486. (Doc. 981.)
Leggett, W. C.

1969. Studies on the reproductive biology of the American
shad (Alosa sapidissima). A comparison of populations
from four rivers of the Atlantic seaboard. Ph.D. Thesis,
McGill Univ., 125 p.

1976. The American shad, Alosa sapidissima, with special
reference to its migrations and population dynamics in the
Connecticut River. In D. Merriman and L. M. Thorpe
(editors). The Connecticut River ecological study: the impact
of a nuclear power plant, p. 169-225. Am. Fish. Soc.
Monogr. 1, 252 p.

1977. Density-dependent, density-independent, and recruit-
ment in the American shad (Alosa sapidissima) population
of the Connecticut River. In W. Van Winkle (editor). Pro-
ceedings of the conference on assessing the effects of jx)wer-
plant-induced mortality on fish populations, p. 3-17.
Pergamon Press, N.Y.

Leggett, W. C, and R. R. Whitney.

1972. Water temperature and the migrations of American
shad. Fish. Bull., U.S. 70:659-670.
Lippson, A. J., and R. L. Moran.

1974. Manual for the identification of early developmental
stages of fishes of the Potomac River estuary. MD Dep.
Natl. Resour., PPSR-MP-13, 282 p.
LoRDA, E., AND V. A. Crecco.

1987. Stock-recruitment relationship and compensatory mor-
tality of American shad in the Connecticut River. Am. Fish.
Soc. Symp. 1:469-482.

Lough, R. G.

1976. Mortality and growth of Georges Bank - Nantucket

Shoals herring larvae during the 1975-1976 winter period.

Int. Comm. Northwest At!. Fish. Resident Doc. 76/6/123,

Ser. No. 4004, 30 p.
LuDWiG, D., AND C. Walters.

1981. Measurement errors and uncertainty in parameters
estimates for stock and recruitment. Can. J. Fish. Aquat.
Sci. 38:711-720.

Marcy, B. C, Jr.

1976. Early life history studies of American shad in the lower
Connecticut River and the effects of the Connecticut Yankee
plant. In D. Merriman and L. M. Thorpe (editors). The Con-
necticut River ecological study: the impact of a nuclear power
plant, p. 141-168. Am. Fish. Soc. Monogr. I. 252 p.

May, R. C.

1974. Larval mortality in marine fishes and critical period
concept. In J. H. S. Blaxter (editor). The early life history
of fish, p. 3-19. Springer- Veriag, N.Y.

Maxfield, G. H.

1953. The food habits of hatchery-produced pond-cultured
shad Alosa sapidissima, reared to a length of two inches.
Chesapeake Biol. Lab. Publ. No. 98, 38 p.

McFadden, J. T.

1977. An argument supporting the reality of compensation
in fish populations and a plea to let them exercise it. In
W. Van Winkle (editor), Proceedings of the conference on
assessing the effects of power-plant-induced mortality on fish
populations, p. 153-183. Pergamon Press, N.Y.

McGuRK, M. D.

1987. Natural mortality and spatial patchiness: reply to
Gulland. Mar. Ecol. Prog. Ser. 39:201-206.
Moffitt, C. M., B. Kynard, and S. G. Rideout.

1982. Fish passage facilities and anadromous fish restoration
in the Connecticut River. Fisheries 7:2-11.

Murray, B. G., Jr.

1982. On the meaning of density dependence. Oecologia
(Beri.) 53:370-373.
Nichols, P. R.

1958. Effect of New Jersey-New York pound-net catches on
shad runs of Hudson and Connecticut Rivers. U.S. Fish
Wildl. Serv., Fish Bull. 58:491-500.
NUSCo (Northeast Utilities Service Company).

1977. Entrainment analyses of shad eggs in the Connecticut
River. Northeast Util. Serv. Co., 167 p.

Parrish, R. H., and a. D. MacCall.

1978. Climatic variation and exploitation in the Pacific
mackerel fishery. Calif. Dep. Fish. Game Fish Bull. 167,
110 p.

Pauly, D.

1980. On the relationships between natural mortality, growth
parameters, and mean environmental temperature in 175
fish stocks. J. Cons. Explor. Mer 39:175-192.
Peterman, R. M.

1978. Testing for density-dependent marine survival in
Pacific salmonids. J. Fish. Res. Board Can. 35:1434-1450.
1982. Nonlinear relation between smolts and adults in Babine
Lake sockeye salmon (Onx^orhynchus nerka) and implications
for other salmon populations. Can. J. Fish. Aquat. Sci.
39:904-913.
RiCKER, W. E.

1954. Stock and recruitment. J. Fish. Res. Board Can. 11:
559-623.

1975. Handbook of computations for biological statistics
of fish populations. Bull. Fish. Res. Board Can. 119, 382
P-

480

SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD
ROSENBURG, A. A., AND R. W. DOYLE.

1.986. Analysing the effect of age structure on stock-recruit-
ment relationships in herring (Clupea harengus). Can. J.
Fish. Aquat. Sci. 43:674-679.
Savoy, T. F., and V. A. Crecco.

1987. Daily increments on the otoliths of larval American
shad and their potential use in population dynamics studies,
/n R. C. Summerfelt (editor), Age and growth of fish, p.
413-431. Iowa State Univ. Press, Ames.
SAS.

1982. Statistical analysis system user guide. SAS Institute
Inc. Gary, NC, 584 p.
SCHERER, M. D.

1974. Analysis of factors affecting passage of American shad
{Alosa sapidissima; Wilson) at Holyoke Dam, MA, and
assessment of juvenile growth and distribution above the
dam. Ph.D. Thesis, Univ. Massachusetts, Amherst, MA,
224 p.
Shepherd, J. G., and D. H. Gushing.

1980. A mechanism for density-dependent survival of larval
fish as the basis of a stock-recruitment relationship. J.
Gons. Int. Explor. Mer 39:160-167.
Threlkeld, S. T.

1986. Lifetable reponses and population dynamics of four
cladoceran zooplankton during a reservoir flood. J. Plank-
ton Res. 8:639-647.
U.S. Geological Survey.

1966-87. Water resources data, Connecticut, water year
reports. U.S. Geol. Surv., var. pagination.
Walburg, G. H., and P. R. Nichols.

1967. Biology and management of the American shad and

status of the fisheries. Atlantic coast of the United States,
1960. U.S. Fish Wildl. Serv. Spec. Sci. Rep. 550, 105

P-
Ware, D. M.

1980. Bioenergetics of stock and recruitment. Gan. J. Fish.

Aquat. Sci. 37:1012-1024.
Watson, J. F.

1968. The early life history of the American shad, Alosa
sapidissima (Wilson), in the Connecticut River above Hol-
yoke, Massachusetts. M.S. Thesis, Univ. Massachusetts,
Amherst, MA, 55 p.

1970. Distribution and population dynamics of American
shad, Alosa sapidissima (Wilson), in the Connecticut River
above Holyoke Dam, Massachusetts. Ph.D. Thesis, Univ.
Massachusetts, Amherst, MA, 105 p.
White, R. L., J. T. Lane, and P. E. Hamer.

1969. Population and migration study of major anadromous
fish. Final Project Report AFSC-1-1, 1-2. NJ Dep. Gonserv.
Econ. Dev. Misc. Rep. No. 3M, 6 p.

Whitton, B. a.

1980. River Ecology. Univ. Calif. Press, Berkeley, 711 p.

Wiggins, T. A., T. R. Bender, V. A. Mudrak, and J. A. Coll.

1984. The development, feeding, growth and survival of

cultured American shad larvae through the transition from

endogenous to exogenous nutrition. Spec. Publ. Penn. Fish.

Comm., Benner Spring Fish Research Station, 34 p.

Zarbock, W. M.

1969. Annual progress report, Delaware River Basin Anad-
romous Fishery Project, AFS-2-2. U.S. Bur. Sport Fish.
Wildl., 20 p.

481

FISHERY BULLETIN; VOL. 86, NO. 3

APPENDIX 1

The total prerecruitment (ZTotal,) and postjuvenile (ZAi) mortality
rates from 1979 through 1987 were estimated by

ZTotal^ = -\og,iPart * 200,000) - \og,{Rt)

and

ZAt = log,(J,) - \og,iR,),

respectively. The standard errors about ZTotal^ and ZAi were estimated
as the sum of the variances (Cochran 1965):

■'ZTotal, -

C.2

SE^Tntal. = V +

nt n^

and

respectively. The variance estimates (5|^, Spavf) of adult recruitment
and spawning stock size were calculated from the log transformed
estimates from the daily lift rates at the Holyoke Dam, where Uf is the
total number of days in which 99% of the American shad were lifted.
The variance (Sj) about the juvenile abundance estimates (J^) was based
on the log transformed catches per seine haul from all stations and col-
lection dates, where rij is the total number of seine hauls made in that
year.

482

METABOLIC RESPONSES OF SPOT, LEIOSTOMUS XANTHURUS, AND

ATLANTIC CROAKER, MICROPOGONIAS UNDULATUS, LARVAE

TO COLD TEMPERATURES ENCOUNTERED FOLLOWING

RECRUITMENT TO ESTUARIES

Donald E. Hoss, Linda Coston-Clements, David S. Peters,
AND Patricia A. Tester^

ABSTRACT

The larvae of marine fishes that spawn during fall-winter in coastal North Carolina waters experience
a decrease in temperature as they enter estuarine nursery areas. To determine the effect of changes
in temperature on larval metabolism, the oxygen consumption of spot, Leiostomus xanthurus, and Atlantic
croaker, Micropogonias undulatus, was measured and their QO2 and Q^g values were determined. Atlan-
tic croaker respiration decreased with temperature at rates that would be expected if no compensation
or stress were involved. Spot showed unexpectedly high respiration rates at low temperature. The in-
creased respiration is apparently due to stress. Based on laboratory feeding and growth data, we con-
cluded that spot are subject to an energy deficit at <10°C. We infer the timing of larval immigration
corresponds with environmental temperatures reaching tolerable levels. Atlantic croaker larvae immigrate
earlier in the winter and are exposed to cold water for longer periods than spot larvae. Our conclusion
is that stress and energy loss experienced by early immigrating spot larvae may result in increased
mortality.

The larvae of fishes that spawn during fall and
winter in offshore North Carolina waters experience
a decrease in both temperature and salinity as they
enter estuarine nursery areas (Fig. 1). The spot,
Leiostomus xanthurus, and Atlantic croaker, Micro-
pogonias undulatus, two sympatric species of Sciae-
nidae, are representative of winter spawning species
off the North Carolina coast.

Previously, we have examined the effects of in-
creased temperature on the oxygen and food con-
sumption of the postlarval stages of these two
species (Hoss et al. 1971, 1974; Peters and Kjelson
1975). In this paper we continue our research on the
early life history of these species and evaluate how
decreasing water temperature, encountered follow-
ing recruitment into estuarine waters, might affect
oxygen consumption, food consumption, and ulti-
mately survival.

Crawshaw et al. (1981) stated that young fish
typically select warm shallow water because 1) it
permits more rapid growth owing to higher metab-
olism, given an adequate food supply, and 2) the
predation by larger fish is less in shallow water. For
some fish this explanation is plausible, but this does
not apply to the larvae of many winter-spawning

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

marine fishes which begin life in relatively warm
coastal waters and then enter colder estuarine
waters. For these species we expect that metabolism
and growth of the estuarine immigrants should be
reduced (Brett 1956). The specific objective of this
paper is to describe how decreasing temperature
affects the metabolism of larval fish as they are
moved from warm to cold water by a combination
of passive and active transport mechanisms that are
not, as yet, completely understood. Oxygen con-
sumption is a common method of estimating meta-
bolic activity, which frequently changes in response
to environmental conditions (O'Hara 1968). In this
study we measured routine oxygen consumption
which is the amount used by fish whose only move-
ments are spontaneous.

STUDY AREA AND METHODS

Spot and Atlantic croaker spawn off the North
Carolina coast inshore of the Gulf Stream over the
continental shelf (Hildebrand and Cable 1930;
Dawson 1958; Powles and Stender 1978). Here, spot
spawn from October to February, but principally
from December to January while Atlantic croaker
spawn from September to May but principally be-
tween October and December (Lewis and Judy
1983). After between 30 and 60 days in coastal

Manuscript accepted May 1988.

FISHERY BULLETIN: VOL. 86. NO. 3, 1988.

483

North

Carol

^^6%o

10 km

21°C

HOSS ET AL.: METABOLIC RESPONSES OF SPOT AND ATLANTIC CROAKER

waters (Warlen 1982; Warlen and Chester 1985), the
larvae enter estuaries where they develop into
juveniles. In the spawning area, water temperatures
are usually between 18° and 25°C (Fahay 1975;
Hettler and Powell 1981). The fish encounter de-
creasing temperatures as they move inshore to the
estuarine nursery areas. In the lower Newport River
estuary, for example, mean water temperatures be-
tween November and March may range from 14°
to 6°C with the highest temperatures during this
period occurring in November and the lowest in
January (Hoss 1974).

Larvae of Atlantic croaker and spot were obtained
from both field collections and eggs spawned in the
laboratory. Older larvae were captured in a bridge
net (Hettler 1979) and held in the laboratory for no
more than a week prior to use. First feeding larvae
were obtained from spawned fish, reproduced by the
methods of Hettler and Powell (1981), and then were
reared at experimental temperatures.

Oxygen consumption was measured with a differ-
ential respirometer (Umbreit et al. 1964), following
procedures used by Hoss, Hettler, and Coston
(1974). Fish were transferred to 15 mL respiration
flasks and, following a 2-h acclimation period, their
oxygen consumption was measured. Numbers of lar-
vae per flask varied between 1 and 30, depending
on the size of the larvae. Acclimation temperatures
were 10°, 15°, and 20°C. Notochord or standard
lengths and dry weights were obtained for individual
fish.

The metabolic equation Q = aW^, was used to
describe the relation between oxygen consumption
and dry weight of fish acclimated at 10°, 15°, and
20°C. In this equation, W is the weight of the fish,
and a and k are constants for the species obtained
from least-squares regression of the log of oxygen
consumption on the log of weight (Winberg 1956).
A k value of 0.67 implies that oxygen consumption
varies in proportion to surface area, whereas a value
of 1 indicates that respiration varies in proportion
to weight.

We used the metabolic equation to estimate oxy-
gen consumption of larvae of equal weight at differ-
ent temperatures. We compared larvae of 4 mg dry
weight because this is the realistic estimate of their
weight as they are transported from coastal to estu-
arine waters (Warlen 1982; Warlen and Chester
1985).

Growth and feeding rates of small spot (^20 mm
SL) collected from the Newport River were calcu-
lated from data collected in the laboratory at several
temperatures. Wet weights (^15-30 mg) were re-
corded to the nearest milligram, and 10 fish were

randomly assigned to 4 L test and control con-
tainers. Control fish were dried to determine the
dry/wet weight ratio, which was then used to esti-
mate initial dry weight of experimental fish. One
experiment was conducted at 6°, 8°, 10°, 12°, and
16°C, and two experiments were conducted at 18°C.
In all cases fish were fed newly hatched brine shrimp
several times a day to assure an ad libitum food
supply. After 4-6 days all food was removed; lar-
vae were allowed time to clear their guts and then
were dried and weighed.

Growth and feeding rates were expressed as per-
cent of body weight per day. Growth rate was
calculated from the expression:

Growth rate = 100 [(WwAVo)

l/n

where Wn = dry weight of all fish in a tank at day
n
Wo = estimated original dry weight of fish
n = number of days fed.

Calculation of feeding rates required the assump-
tion of constant growth rates. Using original dry
weights and calculated growth rates we determined
the total dry weight of fish in each container at the
beginning of each day. Dry weights of brine shrimp
eaten each day divided by the calculated dry weights
of fish gives proportion of body weight ingested.
These proportions were then summarized as aver-
age daily percent of dry body weight ingested.

In order to compare metabolic parameters, i.e.,
oxygen consumption, feeding, and growth rates, the
following conversion factors were used: 1.0 mg dry
wt = 5.0 cal (Thayer et al. 1973; Paffenhoffer 1967);
1.0 mg O2 = 3.38 cal (Phillipson 1966) and 0.7 mg
O2 = 1 mL O2 at STP. One tenth calorie per fish
per day was added to all the rates so that measured
zeros could be shown on a log scale.

RESULTS AND DISCUSSION

The regression equations relating oxygen con-
sumption to weight at several temperatures are
shown in Table 1. Higher coefficient of determina-
tion (R^) values were found at higher temperatures,
a trend best explained by differences in the size
range of fish measured at different temperatures.
Values for k, were generally comparable to values
reported by other investigators for fish of a similar
size and at comparable temperatures— Hoss (1974),
pinfish; Houde and Schekter (1983), bay anchovy,
sea bream, and lined sole; Almatar (1984), herring;
and Laurence (1978), cod and haddock.

485

FISHERY BULLETIN: VOL. 86, NO. 3

Table 1.— Metabolic equations relating oxygen consumption to
body size. Q = aW* where Q = oxygen consumption (piL Oj ■
h"'), a and k are regression coefficients, W is dry wt in mg, N is
tfie number of observations, T the temperature, and Rj the coef-
ficient of determination.

Species

Atlantic

1.50 - 6.43

1.80

0.66

0.42

croaker

1.90 -11.71

2.00

1.02

0.65

0.013- 9.31

3.59

0.86

0.94

Spot

1.82 - 8.01

3.63

0.70

0.54

2.18 -10.27

3.15

0.64

0.63

0.014- 8.91

4.07

0.92

0.97

Comparing between measured oxygen consump-
tion rates at 10° and 15°C and those predicted from
Van't Hoff's equation (Vernberg and Vernberg
1972), we conclude that Atlantic croaker show no
sign of regulating their oxygen consumption as
water temperature is decreased. The difference be-
tween oxygen consumption rates based on Q^q
values of 2 and 3 (Fig. 2) is an expected range. For
every 10°C change in temperature, the rate of a
chemical reaction typically changes by a factor of
2 to 3. A Qio value of appreciably <2 or more than

z
O

h-
o.

ID
CO

z
o
o

<iU

Atlantic croaker

.-.-5

^ — ^ ^ -^

.-''5'' ,,'-'

_^--'' ,'''

-'''' ,--'

2 ^''''

i 1 1

20
15

10
8

Spot

-'i

10 15

TEMPERATURE (°C)

Figure 2.— Oxygen consumption rates for 4 mg Atlantic croaker
(A) and spot (B) at three temperatures estimated from equations
in Table 1. The bars indicate standard errors from regressions in
Table 1. Broken lines are estimates of the rates expected based
on Van't Hoff's equation, the rate measured at 20°C, a Q^ of 2
(upper line) and a Qjq of 3 (lower line).

3, indicates that some process other than a chemical
one is involved (e.g., a change in cell membrane
permeability). A Qjq of one indicates temperature
independence (Vernberg and Vernberg 1972). Our
conclusion that Atlantic croaker did not display
thermal stress is based on the fact that measured
respiration rates at reduced temperatures (10° and
15°C) were within the range expected (Fig. 2A).

For spot a decrease in temperature from 20° to
15°C resulted in a decrease in oxygen consumption
of approximately the amount expected for a Qjq of
3. A further decrease in temperature to 10°C, how-
ever, caused an increase in the respiration rate. The
changes in oxygen consumption at low temperatures
could be interpreted either as adaptive, i.e., main-
taining a high metabolic rate even at the lower tem-
perature, or inadaptive, i.e., a metabolic breakdown.

Based on feeding, growth, and survival data, how-
ever, we think the increase in respiration by spot
at 10° C is a result of cold stress, not adaptation.
In Figure 3 we present three measures of metabolic
rate, ad libitum feeding rate, maximum growth rate,
and routine oxygen consumption for spot, all as a
function of temperature. Feeding, growth, and oxy-
gen consumption rates decrease with decreasing
temperature and the rates are similar over a limited
range of the conditions tested (Fig. 3). The rates of
decline in feeding and growth from 18° to 12°C ap-
proximates that of oxygen consumption from 20°
to 15° C. At lower temperatures stress appears to
become important. For example, it was not possi-
ble to measure growth at 10° C or below because
only a fraction of the larvae survived. At 8° and
10° C some of the larvae did not eat and at 6°C none
of them did. This agrees with Dawson (1958) who
concluded that the lethal minimum temperature for
spot is in the 4.0°-5.0°C range and probably varies
with size. The intersection of ad libitum feeding rate
and routine oxygen consumption occurs at approx-
imately 10°C (Fig. 3). At this temperature there is
just enough energy available for routine metabolism.
Below this temperature there is not enough energy
available even at the ad libitum feeding rate to
maintain the larvae, and spot held at this
temperature for any length of time would be unlikely
to survive.

We conclude from our data on metabolic responses
to temperature that spot and Atlantic croaker lar-
vae differ in their response to cold temperatures
which prevail at the time of their recruitment to the
estuary and that this difference may have impor-
tant implications for their survival. Both species
spawn in warm waters of the continental shelf
where the eggs hatch. As the larvae grow they are

486

HOSS ET AL.: METABOLIC RESPONSES OF SPOT AND ATLANTIC CROAKER

I
W
Ll.

cr
o

Figure 3.— Ad libitum feeding rate, growth rate
at ad libitum feeding rate, and routine oxygen con-
sumption for 4 mg dry wt spot at various
temperatures between 6° and 20°C.

10. Op
80-

60-
4 0

20-

10
0 8
0 6

0.4

0.2

Spot

• = AD LIBITUM FEEDING RATE

A = GROWTH RATE AT AD LIBITUM FEEDING

O - ROUTINE OXYGEN CONSUMPTION

n 1' ^ ' 'I

10 14 18 22

TEMPERATURE (°C)

transported from warm coastal waters into cold
estuarine waters. Atlantic croaker are capable of
enduring low winter temperatures with decreased
metabolic rates that allow for balanced energy in-
take. Spot, in contrast, show signs of thermal stress
manifested as increased respiration rate (at 10°C).
This increased metabolism along with no attendant
increase in feeding results in an energy deficit and
in eventual mortality of the larvae. Species specific
differences in the time of entry to the estuary serves
as ecological evidence supporting our contention
that spot are more susceptible to cold weather. Most
spot enter the estuary after the peak in Atlantic
croaker immigration and generally after the coldest
weather.

Our findings have important implications with
respect to recruitment of estuarine-dependent fish
which spawn in the ocean during winter. It may be
that during severe winters, many of the larvae of
cold sensitive species (e.g., spot) that reach the
estuary early are killed by cold water temperatures
(10°C or less). Thus, only the late arriving larvae
survive to recruit into the fishery. The difference
in survival between severe and normal winters may
help to explain in part the difference between good
and poor year classes of certain fish.

Acknowledgments

We thank W. Hettler for providing eggs of labor-
atory spawned fish, and W. Hettler and J. Govoni

for providing critical reviews of the manuscript. This
research was funded in part by a cooperative agree-
ment between the National Marine Fisheries Service
and the Department of Energy E(49-7)5.

LITERATURE CITED

Almatar, S. M.

1984. Effects of acute changes in temperature and salinity
on the oxygen uptake of larvae of herring (Clupea harengus)
and plaice (Pleuronectes platessa). Mar. Biol. (Berl.) 80:117-
124.
Brett, J. R.

1956. Some principles in the thermal requirements of fishes.
Q. Rev. Biol. 31:75-87.
Crawshaw, L. I., B. P. MoFFiTT, D. E. Lemons, and J. A.
Downey.
1981. The evolutionary development of vertebrate thermo-
regulation. Am. Sci. 69:543-550.
Dawson, C. E.

1958. A study of the biology and life historj' of the spot,
Leiostomus xanthurus Lacepede, with special reference to
South Carolina. Bears Bluff Lab. Contrib. No. 28, 48 p.
Fahay, M. p.

1975. An annotated list of larval and juvenile fishes captured
with surface-towed meter net in the South Atlantic Bight
during four RV Dolphin cruises between May 1967 and
February 1968. U.S. Dep. Commer., NOAA Tech. Rep.
NMFS SSRF-685, 39 p.
Hettler, W. F.

1979. Modified neuston net for collecting live larval and
juvenile fish. Prog. Fish.-Cult. 41:32-33.
Hettler, W. F., and A. B. Powell.

1981. Egg and larval fish production at the NMFS Beaufort
Laboratory, Beaufort, N.C., U.S.A. Rapp. P.-v. R6un Cons,
int. Explor. Mer 178:501-503.

487

FISHERY BULLETIN: VOL. 86, NO. 3

HiLDEBRAND, S. E., AND L. E. CABLE.

1930. Development and life history of fourteen teleostean
fishes at Beaufort, N.C. Bull. U.S. Bur. Fish. 46:383-488.
Hoss, D. E.

1974. Energy requirements of a population of pinfish, Lago-
don rhomboides (Linnaeus). Ecology 55:848-855.
Hoss, D. E., L. C. CosTON, and W. F. Hettler, Jr.

1971. Effects of increased temperature on postlarval and
juvenile estuarine fish. Proc. Annu. Conf. Southeast. Assoc.
Game Fish. Comm. 25:635-642.
Hoss, D. E., W. F. Hettler, Jr., and L. C. Coston.

1974. Effects of thermal shock on larval estuarine fish - eco-
logical implications with respect to entrainment in power
plant cooling systems. In J. H. S. Blaxter (editor), The early
life history of fish, p. 357-371. Springer- Verlag, N.Y.
HouDE, E. D., and R. C. Schekter.

1983. Oxygen uptake and comparative energetics among
eggs and larvae of three subtropical marine fishes. Mar.
Biol. (Berl.) 72:283-293.
Laurence, G. C.

1978. Comparative growth, respiration and delayed feeding
abilities of larval cod (Gadv^ morhua) and haddock (Melano-
grammus aeglefinus) as influenced by temperature during
laboratory studies. Mar. Biol. (Berl.) 50:1-7.
Lewis, R. M., and M. H. Judy.

1983. The occurrence of spot, Leiostomus xanthunis, and
Atlantic croaker, Micropogcmias undulatus, larvae in Onslow
Bay and Newport River Estuary, North Carolina. Fish.
Bull., U.S. 81:405-412.
LiPPSON, A. J., AND R. L. MORAN (editors).

1974. Manual for the identification of early developmental
stages of fishes of the Potomac River estuary. Power Plant
Siting Program, Md. Dep. Nat. Res. PPSP-MP-13, p. 220-
222.
O'Hara, J.

1968. The influence of weight and temperature on the meta-
bolic rate of sunfish. Ecology 49:159-161.
Paffenhofer, G.-A.

1967. Calorie content of larvae of the brine shrimp Artemia

salina. Helgol. Wiss. Meeresunters. 16:130-135.
Peters, D. S., and M. A. Kjelson.

1975. Consumption and utilization of food by various postlar-
val and juvenile fishes of North Carolina estuaries. In L. E.
Cronin (editor), Estuarine research, vol. 1, p. 448-472.
Acad. Press, N.Y.
Phillipson, J.

1966. Ecological energetics. E. Arnold, Lond., 57 p.
Powell, A. B., and H. R. Gordy.

1980. Egg and larval development of the spot Leiosto-
mus xanthunis (Sciaenidae). Fish. Bull., U.S. 78:701-
714.

POWLES, H., AND B. W. STENDER.

1978. Observations on composition, seasonality and distribu-
tion of ichthyoplankton from MARMAP cruises in the South
Atlantic Bight in 1973. S.C. Mar. Res. Cent. Tech. Rep.
Ser. 11, 47 p.
Thayer, G. W., W. E. Schaaf, J. W. Angelovic, and M. W.
LaCroix.
1973. Caloric measurements of some estuarine organisms.
Fish. Bull., U.S. 71:289-296.
Umbreit, W. W., R. H. Burris, and J. F. Stauffer.

1964. Manometric techniques. Burgess Publ. Co., Minneap.,
305 p.
Vernberg, W. B., and F. J. Vernberg.

1972. Environmental physiology of marine animals.
Springer-Verlag, N.Y., 346 p.
Warlen, S. M.

1982. Age and growth of larvae and spawning time of Atlan-
tic croaker. Proc. Annu. Conf. Southeast. Assoc. Fish
Wildl. Agencies 34:204-214.
Warlen, S. M., and A. J. Chester.

1985. Age, growth, and distribution of larval spot, Leio-
stomus xanthums, off North Carolina. Fish. Bull., U.S.
83:587-599.
WiNBERG, G. G.

1956. Rate of metabolism and food requirements of fishes.
Can. Fish. Res. Board, Transl. Ser. 194. 202 p.

488

GROWTH OF PACIFIC SAURY, COLOLABIS SAIRA, IN
THE NORTHEASTERN AND NORTHWESTERN PACIFIC OCEAN'

YosHiRO Waia.nabe,2 John L. Builer,^ and Tsukasa Mori^

ABSTRACT

Growth of the Pacific saury, Cololabis saira, from the northeastern and northwestern Pacific Ocean was
studied using otolith growth increments. We found that growth of Pacific sauries from the western Pacific
was higher than that from the eastern Pacific. Assuming that otolith growth increments are deposited
daily, average growth rates from hatching up to 1 year old were 0.62 mm/d in the eastern and 0.85 mm/d
in the western Pacific. Because the growth rate changes at around 100 mm, two curves were used to
model the growth of Pacific saury in the western Pacific: one for fish up to 100 mm and the other for
fish larger than 100 mm. Based on counts of daily increments. Pacific sauries may be short lived. The
oldest specimen examined was only 14 months old.

The Pacific saury, Cololabis saira (Brevoort), is
distributed throughout the North Pacific Ocean and
is one of the most important commercial fishes in
the northwestern Pacific. The average annual catch
of Pacific saury in Japan has been approximately
200,000 1 (metric tons) in the last 20 years (Statistics
and Information Department, Japan 1985). The
catch has varied by an order of magnitude in the
last 20 years from a minimum of 63,000 t in 1969
to a maximum of 406,000 t in 1973. Fluctuation in
stock size is a major factor in catch variability
although economic factors such as fish price may
also affect total landings. However, the causes of
stock fluctuation in the western Pacific remain
unknown. In the eastern Pacific, the Pacific saury
has not been exploited but is recognized as a poten-
tial fishery resource (Ahlstrom 1968; Smith et al.
1970).

Investigations of the Pacific saury have mainly
been devoted to such subjects as systematics, abun-
dance, distribution, migration, and formation of
fishing ground in relation to oceanographic condi-
tions (e.g., Hubbs and Wisner 1980; Smith et al.
1970; Odate 1977; Fukushima 1979; Sablin and
Pavlychev 1982; Gong 1984). Age determination and
growth, however, remain controversial (Hatanaka
1955; Hotta 1960; Novikov 1960; Sunada 1974; Kim

"Contribution No. 429 from Tohoku Regional Fisheries Research
Laboratory.

^Tohoku Regional Fisheries Research Laboratory, Fisheries
Agency, Shiogama, Miyagi 985, Japan.

^Southwest Fisheries Center La Jolla Laboratory, National
Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA
92038.

*Facu]ty of Fisheries, Hokkaido University, Hakodate, Hokkaido
041, Japan.

and Park 1981), notwithstanding their critical im-
portance for fish stock assessment.

The discovery of daily increments in the otoliths
of fishes (Pannella 1971) has made it possible to
estimate age and growth of larval and juvenile fishes
accurately. Daily increments have been used to age
many species of fishes (Jones 1986). Nishimura et al.
(1985) reported the presence of growth increments
in Pacific saury otoliths observed by scanning elec-
tron microscopy and suggested that it is possible to
estimate age and growth of Pacific saury by using
daily increments in the otolith. The purpose of this
paper is to determine the age of Pacific sauries from
the eastern and western North Pacific using daily
increments and to compare the growth rates in these
areas.

MATERIALS AND METHODS

We read otoliths of 75 Pacific sauries from the
northeastern and 172 from the northwestern Pacific
Ocean. Details of sampling and methods of reading
otoliths are summarized in Table 1 and Figure 1.
Additional samples from the western Pacific were
used to determine the relation between otolith size
and fish length. Fish from the eastern Pacific were
fixed and preserved in 80% alcohol after capture,
and those from the western Pacific were stored
frozen and thawed when processed. Because speci-
mens frequently have damaged upper jaws, knob
length (the distance from the tip of the lower jaw
to the posterior end of the muscular knob at the base
of a caudal peduncle) is the standard measure of
body size in Pacific saury. All body lengths in this
paper are knob length.

Manuscript accepted March 1988.

FISHERY BULLETIN: VOL. 86, NO. 3, 1988.

489

FISHERY BULLETIN: VOL. 86, NO. 3

Table 1.— Collection records of saury samples from the eastern (# 1-10) and the western (# 11-22)
North Pacific. N = Neuston net; G = Gill net; D = Dip net.

Location

Size range

Micro-

Sample

Date

Lat.

Long.

No.

(KnL mm)

Gear

scope

80 06 29

36°02'N

124°04'W

19.8-109.0

80 06 29

36°07'N

123°55'W

19.8- 66.0

80 06 29

36°07'N

123°45'W

15.3- 96.0

81 10 26

46°36'N

127°49'W

47,8-235.0

81 10 27

47°19'N

126°09'W

38.0-142.0

81 10 28

47°20'N

124°30'W

23.9-216.0

81 10 29

46°21'N

127°38'W

209.0

81 10 30

46°38N

125°54'W

20.6- 85.0

81 11 01

45°21'N

127°38'W

27.0- 71.0

81 11 02

45°38'N

124°51W

70.0-206.0

84 05 26

37°01'N

164°02'E

145.0-230.0

SEM

84 06 02

37°00'N

158°00'E

114

38.0-125.0

SEM

84 07 16

42°00'N

172°00'E

213.0-282.0

SEM

84 10 06

43°09'N

153°14'E

300.0-330.0

SEM

85 05 20

36°00N

150°00'E

21.0- 69.5

85 05 21

38°15'N

149°59'E

8.3- 12.5

85 05 23

38°30'N

152°00'E

26.5- 45.5

85 05 24

38°00'N

152°00'E

29.0- 85.0

85 05 29

38°45'N

156°00'E

16.0- 33.5

85 05 29

38°30N

ise-'oo'E

27.0- 33.0

85 05 29

38015N

156°00'E

24.6- 65.0

85 05 30

38°00'N

156°00'E

31.5- 68.5

140°E

150°E 160°E 170°E

Figure 1.— Locations of Pacific saury collection in the North Pacific. Figures by

490

WATANABE ET AL.: GROWTH OF SAURY

Sagittae were dissected out from fish and left to
dry after removing tissues and membranes. We used
a dissecting microscope with a polarizing filter to
dissect otoliths from small larvae and juveniles. The
otoliths were read either by light microscopy (LM)
or by scanning electron microscopy (SEM). Otoliths
that were to be read by LM were mounted in
EUKITT^ after dissection. Otolith radius was
measured from the focus to posterior margin and
the increments were counted along the same tran-
sect using the otolith reading system, which was
developed by the Southwest Fisheries Center of
the National Marine Fisheries Service, NOAA, and
which consists of a light microscope, a video
monitor, a micro-computer, and a digitizer (Methot
1981).

For SEM, otoliths were mounted in epoxy or
methacrylate resin. The otolith radius was measured
from the focus to the posterior margin with an

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

optical comparator. The otoliths were ground
oblique to the sagittal plane parallel to the long
axis of the otolith in order to have a flat plane
through the otolith nucleus. The polished surface
was washed in xylene, using an ultrasonic washer,
then dried and etched for 50 seconds with 0.2M
EDTA-2Na (disodium ethylenediaminetetraacetic
acid). The etched surface was coated with palladium
platinum and observed under an SEM (JSM-25) at
15 kV.

The three authors of this paper read saury oto-
liths independently: the senior author read fish from
the western Pacific up to 85 mm by LM, the second
author read otoliths from the eastern Pacific by LM,
and the third read otoliths from the western Pacific
larger than 38 mm by SEM. To confirm that we
were all interpreting the same structure as growth
rings by SEM and LM, we compared 50 data points
read by SEM and 14 points read by LM for west-
ern Pacific sauries between 38 and 85 mm. The
distribution of increment number versus knob length
was the same. We also checked for possible biases
for the two readers using LM by having each read
the same set of otoliths independently.

130°W 120°W

dots indicate sample numbers in Table 1.

RESULTS

The nucleus of a Pacific saury sagitta is approx-
imately 20 ^m in diameter and is composed of four
to six small dense bodies which appear to be assem-
blages of calcareous spherules (Fig. 2a). These dense
bodies are separated from one another and each is
surrounded by a small concentric ring.

We observed the nucleus areas of otoliths from
Pacific sauries collected in the western Pacific in
1985 (sample # 15-22) and found that most of
them had a distinct ring of about 27 fim in radius.
Between the nucleus and this distinct ring, four
(or five) indistinct growth rings were detected (Fig.
2b).

We measured knob lengths of 27 larval and juven-
ile Pacific sauries before freezing and after thaw-
ing, and found that the ratio of these two measure-
ments was 0.997. There was virtually no shrinkage
by freezing and thawing. Theilacker (1980) found
that preservation of larval northern anchovy,
Engraulis mordax, in 80% alcohol did not cause
additional shrinkage of the body after net treatment.
Thus, knob lengths after both 80% alcohol preser-
vation and freezing are comparable to each other,
and this measurement corresponds to the size after
net treatment. Since shrinkage factors by net treat-
ment are not known for the saury, lengths are un-
corrected for net shrinkage.

491

FISHERY BULLETIN: VOL. 86, NO. 3

.^1

Figure 2.— Light micrographs of Pacific saury sagittae. a) OtoHth nucleus composed of 5 or 6 separate dense bodies with surrounding

cores, b) Assumed 4 embryonic and 1 hatching (arrow) rings.

The daily periodicity of growth increment forma-
tion in the Pacific saury has not been verified. For
that reason we plotted the number of increments
versus knob length instead of age versus length. We
used the Laird-Gompertz equation to describe the
relations of increment number and length as growth
curves for both the eastern and western Pacific
saury. Hatching size of artificially fertilized and in-
cubated Pacific saury from the western Pacific was
reported to be 7.19 mm in average live total length
(Yusa 1960). From the drawing of a newly hatched
larva in Yusa's paper, we estimated live knob length
to be 6.60 mm. Shrinkage factors of northern an-
chovy in the size range from 6.00 to 7.99 mm were
0.90 for a 5-min net treatment and 0.85 for a 10-min
net treatment (Theilacker 1980). Using these values,
the capture size of a newly hatched larva of Pacific
saury after a 5-min net treatment was estimated to
be 5.95 mm and a 10-min treatment to be 5.61 mm.
We fixed the hatching size from 5.85 to 5.95 mm
in the growth curve, because the Pacific saury lar-
vae at this size are in a more advanced develop-

mental stage and shrank less by net treatment than
northern anchovy.

The resulting growth equation for the eastern
Pacific saury was

KnL = 5.85 exp((0.0427/0.115)(l - e(-o.oii5(/-5))))

and the equation for the western Pacific saury was
KnL = 5.95 exp((0.0504/0.0128)(l - e(-ooi28(/-5))))

where KnL is a knob length in mm and / is the total
number of increments observed in an otolith. The
term, 7-5, indicates that five increments were pre-
sumed to have been present at hatching. Data from
the western Pacific saury appear to consist of two
curves separated around 100 mm in KnL. Two
Laird-Gompertz curves fit much better than one
curve. The intersection of the two curves was at 1 14
increments and 100 mm. The growth equation for
fish smaller than 100 mm was

492

WATANABE ET AL.: GROWTH OF SAURY

KnL = 5.90 exp((0.0865/0.0293Xl - e(-oo293(/-5))))
and for fish larger than 100 mm KnL was

KnL = 3.01 exp((0.0592/0.0126)(l - e(-o.oi26(/-5))))

The estimated mean square error, 215.7, of the two
curves was smaller than that for a single curve,
351.7. The two-curve model fits much better for the
smaller size range up to 100 mm. The estimated
mean square error of the two-curve model for this
size range, 75.7, was much smaller than that of the
one-curve model, 240.8.

The growth rate of Pacific saury in the eastern
Pacific was slower than that in the western Pacific
(Figs. 3, 4). The knob length of saury in the eastern
Pacific would be about 75 mm at 100 rings, 170 mm
at 200 rings, and 220 mm at 300 rings, whereas in
the western Pacific knob length would be about 100
mm at 100 rings, 230 mm at 200 rings, and 300 mm
at 300 rings. Assuming that the rings are formed
daily, overall growth rates of the first one year
of their life were 0.62 mm/d and 0.85 mm/d for

the eastern and western Pacific Ocean, respectively.

The largest specimen examined was 330 mm from
the western Pacific, which had 328 increments, and
the oldest fish, also from the western Pacific, which
measured 320 mm, had only 418 increments. These
fish would be classified as very large or large by
Novikov's categories offish size composition (Novi-
kov 1960, 1973). The largest fish examined from the
eastern Pacific was 235 mm and had 241 incre-
ments. Hughes (1974), however, reported larger fish
from the eastern Pacific.

In Pacific saury from the western Pacific, the rela-
tion between otolith radius in f^m (OR) and knob
length in mm (KnL) was linear on logarithm-
logarithm coordinates (Fig. 5). The equation com-
puted by the geometric mean regression (Ricker
1973) was

In (OR) = 2.33 + 0.749 In (KnL) (r = 0.979).

The otolith radius at hatching (5.9 mm KnL) cal-
culated by this formula was 38.9 ^im, which was 12
/im larger than the radius of the presumed hatch-
ing ring.

200 -

E
E

100 -

00427 f ^(-0.0115(l-5))\
0.0115 ^ ^® '

KnL: KNOB LENGTH

I: NUMBER OF INCREMENTS

100 200

NUMBER OF INCREMENTS

Figure 3.— Growth curve of the eastern Pacific saury.

300

493

FISHERY BULLETIN; VOL. 86, NO. 3

300

£ 200

X
I-

z
111

m
O

^ 100

KnL = 5.95e

0.0504 / ^(-0.0128(l-5))\
0.0128^^"® '

I <114

0.0865 / {-0.0293{l-5))\

KnL = 5.90e 0.0293 ^^'^ '

0.0592 / (-0.0126(1-5)))

KnL = 3.01e 0.0126^^"®

KnL: KNOB LENGTH

I: NUMBER OF INCREMENTS

100

200

300

400

NUMBER OF INCREMENTS

Figure 4.— Growth curve of the western Pacific saury.

DISCUSSION

The microstructure of otolith growth increments
of the Pacific saury is similar to that of daily incre-
ments in some other fishes (Nishimura et al. 1985).
Thus, the following discussion is based on the
assumption that the increments are daily growth
rings. A rearing experiment of larval sauries is
under way in the senior author's laboratory to verify
daily periodicity of the increment formation.

Formation of a few embryonic growth rings or a
lamellar structure has been reported in California
grunion, Leuresthes tenuis, (Brothers et al. 1976);
mummichog, Fundulus heteroclitus, (Radtke and
Dean 1982); and walleye pollock, Theragra chalco-
gramma, (Nishimura and Yamada 1984). Radtke
and Dean (1982) mentioned that deposition of
growth rings in the embryonic stage might be

related to a long incubation period. Pacific saury has
a long incubation period— about 17 days under
13.5°-15.7°C (Yusa 1960). At this temperature, eye
pigmentation begins 7 or 8 days before hatching,
and pectoral fins show constant movement from 5
or 6 days before hatching (Yusa 1960). Notochord
flexion occurs about midway through embryonic
development at 14°-22°C (Uchida et al. 1958). Thus
saury is more advanced at hatching than killifish
based on the embryonic development of killifish
reported by Armstrong and Child (1965). Observ-
ing the central area of otoliths, we found four faint
rings and a dark ring immediately outside of those
rings. We assumed therefore that the four faint
rings are embryonic rings and the dark ring is the
hatching ring. This assumption may be confirmed
by examining otoliths of late embryos and newly
hatched larvae of saury.

494

WATANABE ET AL.: GROWTH OF SAURY
1000 I-

500

Q
<
CC

I-

100

In OR =2.33 +0.749 In KnL
r = 0.979

OR: OTOLITH RADIUS
KnL: KNOB LENGTH

J \ \ \

50 100

KNOB LENGTH (mm)

200 300

Figure 5.— Knob length and otolith radius relationship of the western Pacific saury.

Previous studies on age and growth of the Pacific
saury have based on annuli on scales and/or otoliths.
Sunada (1974) found five age groups in Pacific
sauries off southern Oregon, California, and Baja
California. Mean fork lengths of age groups were
171, 220, 246, 270, and 268 mm for age 0, 1, 2, 3,
and 4, respectively. Hughes (1974) examined age
composition of 5,248 sauries collected in waters off
California up to Vancouver Island. He found spring-
and autumn-bom fish in his samples, but little differ-
ence was noted in growth rates between two groups.
Approximate knob lengths of 1.0- to 5.0-year-old fish
were 180, 230, 255, 290, 310 mm. The growth rates
given in these two papers are not very much dif-
ferent. The saury grows at 0.5-0.6 mm/d up to 1
year old, which is almost equal to our growth rate
in the eastern Pacific, 0.62 mm/d. Hatanaka (1955)
found five age groups in the western population of
the saury, 0-4 years old, and estimated mean body
length of age groups to be 80, 160, 230, 265 mm for
1- to 4-year-old fish, respectively. Novikov (1960,
1973) divided sauries captured in autumn into five
size groups, very small ( - 200 mm), small (201-240),
medium (241-290), large (291-320), and very large
(321-(-), and assigned the small, medium, and large

to 1-, 2-, and 3-year-old with maximum 5-year-old
fish.

A different model of Pacific saury growth in the
western Pacific was proposed by Hotta (1960) based
upon a hypothesis of two subpopulations. He separ-
ated the saury into spring-spawning and autumn-
spawning populations based upon the observations
of fish size composition, scales, otoliths, and num-
bers of vertebrae. He assigned four ages of half year
intervals, 1.0, 1.5, 2.0, 2.5 years old, to fish 210-240,
260-280, 290-300, and 310-330 mm, respectively.
The growth rate up to 1 year was 0.6-0.7 mm/d. Kim
and Park (1981) examined Pacific sauries from
Korean waters and found four size groups of four
different ages of half-year intervals as well. They
presented two growth models for each of two sub-
populations, spring and autumn spawning, based
upon the von Bertalanffy equation. The sizes at ages
were almost identical to those of Hotta (1960). How-
ever, the hypothesis of two saury subpopulations in
the western Pacific is not supported by electro-
phoretic analyses of genetic separation (Numachi
1971; Hara et al. 1982).

The average growth rate of the western Pacific
saury in this paper was 1.1 mm/d from 0 to 8 or

495

FISHERY BULLETIN: VOL. 86, NO. 3

9 months old. It was still faster than Novikov's
growth rate of the corresponding age period (0.83
mm/d), which was the highest rate of all the previous
reports. The fish would become 316 mm in one year
according to our model.

Support for the fast growth rate of Pacific saury
presented in this paper in the western Pacific can
be found in rearing experiments. Hotta (1958)
reared young sauries caught by a set net. He reared
them in a crawl and fed them minced anchovy and
mackerel twice a day. Young sauries 116 mm in
mean length became 172 mm in the rearing period
after 72 days. The growth rate was 0.78 mm/d. The
sauries fed three times a day grew 130-143% faster
than the group fed twice a day. Thus growth rates
of young sauries may be higher than 1.1 mm/d (0.78
X 1.4 mm/d) when food is readily available. Our
growth rate of sauries in this size range was approx-
imately 1.5 mm/d in the western Pacific. For Atlan-
tic saury, Scomberesox saurus scombroides, reared
by Brownell (1983), the average growth rate of the
larvae was 0.62 mm/d from hatching (7.5 mm SL)
to 47-day old (36.8 mm SL). The growth rates of
Cololabis saira in a corresponding period were 0.48
mm/d in the eastern Pacific and 1.0 mm/d in the
western Pacific.

Our results indicate that the growth rate of Pacific
saury in the western Pacific is much higher than in
the eastern Pacific. This could be due to a differ-
ence in food availability between the two areas.
However, mean zooplankton standing stock in
1951-66 was 34.8 g/m^ in the California Current
region in June (Smith and Eppley 1982), whereas
that of Kuroshio water off southern Japan was 4.7
g/m^ and of Oyashio area off northern Japan was
25.7 g/m2 in May to July (Odate 1986). Thus, differ-
ences in zooplankton standing stock do not explain
the difference in growth rates.

On the other hand, there seems to be a reasonable
explanation for the reacceleration of growth rate at
around 100 mm in the western Pacific saury. The
western sauries hatch out mainly in offshore water
of the Kuroshio Current (lat. 31-33°N) off Japan
in winter. They migrate north to the Oyashio area
(up to 46-50 °N) where copepods are highly avail-
able. Young and adult sauries feed actively and gain
fat. They are in the northward migration stage in
early summer when they are about 100 mm, and are
moving from poor Kuroshio water to rich Oyashio
water (Fukushima 1979). High zooplankton stand-
ing stock in the Oyashio water and its derivatives
might be responsible for the reacceleration of
growth rate in fish older than 100 days.

The growth rates of Pacific saury in the western

Pacific may differ from year-to-year due to environ-
mental factors and may result in changes in size
composition of the fish. Between 1968 and 1972,
mean knob length of exploited sauries in the west-
ern Pacific was 170-250 mm, whereas in the 1980's
the major mode in the size composition was 290-310
mm (S. Kosaka''). This increase could have been
due to an acceleration of growth rate or a shift of
spawning season to early months or both in recent
years. The high growth rate of western Pacific saury
presented in this paper has come from specimens
collected in 1984 and 1985. The growth rate in the
late 1960's and early 1970's may have been lower
than that presented in this paper. Investigation of
the interannual variation in growth rates using daily
increments would distinguish between these two
hypotheses.

We used three different gears to collect Pacific
saury samples in the western Pacific. Knob lengths
of sauries collected were from 8.3 to 125 mm by ring
net, 145 to 282 mm by gill net, and 300 to 330 mm
by stick-held dip net. Sauries of 125-145 mm might
not be available either to the ring net or to the gill
net. Further, the ring net may select small juveniles
of a cohort in the size range over 100 mm, and this
may have produced the two growth curves. This
problem needs to be examined further with data on
gear selectivity.

We do not know how long Pacific sauries survive
after becoming adult. The oldest specimen in our
sample was about 14 months old after hatching (418
increments). The largest saury aged (330 mm),
which had 328 growth rings, is close to the max-
imum size. Although the maximum known length
of the Pacific saury was reported to be about 400
mm (Hubbs and Wisner 1980), the largest fish ex-
ploited in Japan is about 340 mm. Therefore, the
lifespan of the Pacific saury is about one year in the
western Pacific. Our results are more consistent
with those of Kosaka (1979) who found two age
groups (0 and 1 year) than those (Sablin 1979) who
found three age groups (0, 1, and 2 years).

In Japan, fishing efforts of the Pacific saury is
regulated by fishing season as well as by the num-
ber of fishing boats. The fishing season starts in mid-
August. Pacific sauries hatched in the main spawn-
ing season are about 250 days old at this time of year
(approximately 270 mm) and are grovdng at the rate
of 0.8 mm/d. Thus a 2-wk postponement at the
beginning of fishing season would result in an 11
mm (10-15 g body weight) increase of average fish

*S. Kosaka, Tohoku Regional Fisheries Research Laboratory,
Fisheries Agency, Shiogama, Miyagi 985, Japan, pers. commun.

496

WATANABE ET AL.; GROWTH OF SAURY

length. The biomass and yield of Pacific saury need
to be reestimated based on faster growth rates
presented here.

ACKNOWLEDGMENT

We thank A. Nishimura of Hokkaido Salmon
Hatchery; J. Yamada of Hokkaido University; R.
Lasker, S. Tsuji, and N. C. H. Lo of Southwest
Fisheries Center of National Marine Fisheries Ser-
vice (NMFS), NOAA; and S. Kosaka and T. Wata-
nabe of the Tohoku Regional Fisheries Research
Laboratory for valuable suggestions and reading the
manuscript. We also thank A. W. Kendell Jr. of the
Northwest and Alaska Fisheries Center, NMFS,
NOAA, for providing us with some specimens from
the eastern Pacific, and D. Abramenkoff of the
Southwest Fisheries Center for help in reading Rus-
sian papers. Y. Watanabe would like to thank the
Science and Technology Agency of Japan for fund-
ing his stay at the Southwest Fisheries Center.

LITERATURE CITED

Ahlstrom, E. H.

1968. An evaluation of the fishery resources available to
California fishermen. In The future of the fishing industry
in the United States, p. 65-80. Univ. Wash. Publ. Fish.,
New Ser. IV.
Armstrong, P. R, and J. S. Child.

1965. Stages in the normal development of Fundulxis hetero-
clitus. Biol. Bull. 128:143-168.
Brothers, E. B., C. P. Mathews, and R. Lasker.

1976. Daily growth increments in otoliths from larval and
adult fishes. Fish. Bull., U.S. 74:1-8.
Brownell, C. L.

1983. Early growth rate and feeding of a small group of
laboratory-reared saury, Scomberesox saurus scombroides
(Pisces: Scomberesocidae). S. Afr. J. Mar. Sci. 1:245-248.

FUKUSHIMA, S.

1979. Synoptic analysis of migration and fishing conditions

of saury in the Northwest Pacific Ocean. (In Jpn.) Bull.

Tohoku Reg. Fish. Res. Lab. 41:1-70.
Gong, Y.

1984. Distribution and movements of Pacific saury, ColoJubis
saira (Brevoort), in relation to oceanographic conditions in
waters off Korea. Bull. Fish. Res. Dev. Agency 33:1-156.

Hara, M., a. Kijima, and Y. Fujino.

1982. Genetic study on population structure of Pacific saury.
(In Jpn.) Bull. Tohoku Reg. Fish. Res. Lab. 45:1-18.
Hatanaka, M.

1955. Biological studies on the population of the saury, Colo-
labis saira (Brevoort). Part I. Reproduction and growth.
Tohoku J. Agric. Res. 6:227-269.
Hotta, H.

1958. On the growth of the young saury, Cololabis saira, in
the rearing experiment. (In Jpn.) Bull. Tohoku Reg. Fish.
Res. Lab. 11:47-64.
1960. On the analysis of the population of the saury (Colo-
labis saira) based on the scale and the otolith characters.

and their growth. (In Jpn.) Bull. Tohoku, Reg. Fish. Res.
Lab. 16:41-64.
HUBBS, C. L., AND R. L. Wisner.

1980. Revision of the sauries (Pisces, Scomberesocidae) with
description of two new genera and one new species. Fish.
Bull., U.S. 77:521-566.

Hughes, S. E.

1974. Stock composition, growth, mortality, and availability
of Pacific saury, Cololahis saira, of the northeastern Pacific
Ocean. Fish. Bull., U.S. 72:121-131.
Jones, C.

1986. Determining age of larval fish with the otolith incre-
ment technique. Fish. Bull., U.S. 84:91-103.
Kim, Y. M., and Y. J. Park.

1981. A study on the growth of saury, Cololabis saira (Bre-
voort), based on the length composition in the Korean
waters. (In Korean) Bull. Fish. Res. Dev. Agency 27:
59-70.

Kosaka, S.

1979. Sanma no nenrei seicho ni tsuite (Age and growth of
saury). (In Jpn.) Rep. Japan-Soviet Coop. Res. Meet.
Saury Scomber 11:118-122.

Methot, R. D.

1981. Growth rates and age distributions of larval and juven-
ile northern anchovy, Engraulis mordax, with inferences on
larval survival. Ph.D. Thesis, Univ. California, San Diego,
209 p.

Nishimura, A., Y. Watanabe, and J. Yamada.

1985. Daily growth increment-like microstructure in otoliths
of the Pacific saury Cololabis saira. (In Jpn.) Bull. Tohoku
Reg. Fish. Res. Lab. 47:33-35.

Nishimura, A., and J. Yamada.

1984. Age and growth of larval and juvenile walleye pollock
Theragra chalcogramma (Pallas), as determined by otolith
daily growth increments. J. Exp. Mar. Biol. Ecol. 82:191-
205.

NoviKOV, Yu.V.

1960. Oprjedjenije vozrasta po chjeshchuje i vozrastnoi sostav
sairy (Cololabis saira (Brevoort)) v raionje ziuzhnykh
kuril'skikh ostrovov (Determination of age by scales and age
composition of saury (Cololabis saira (Brevoort)) in region
of southern Kuril Islands). (In Russ.) Izv. TINRO 46:233-
241.
1973. Izuchjenije populiatsionnoi struktury i vozrastnogo
sostava sairy sjevjero-zapadnoi chasti tikhogo okjeana (The
study of population structure and age composition of saury
of north-western part of the Pacific Ocean). (In Russ.) Izv.
TINRO 87:149-154.

NUMACHI, K.

1971. Kouso no takei ni yoru sanma keitougun ni kansuru
kenkyu (Study on the population of the Pacific saury based
upon analyses of enzymatic polymorphism). (In Jpn.)
Ocean Res. Inst., Univ. Tokyo, Tokyo, 57 p.

Odate, S.

1977. On the distribution of Pacific saury in the North Pacific
Ocean. (In Jpn.) Res. Inst. North Pac. Fish. Sp. Vol.,
p. 353-382. Fac. Fish., Hokkaido Univ., Hakodate, Ja-
pan.

Odate, K.

1986. Automatic data processing for estimation of abundance
of zooplankton in the waters off northeastern Honshu,
Japan, 1951-1976. an Jpn.) Bull. Tohoku Reg. Fish. Res.
Lab. 48:31-47.

Pannella, G.

1971. Fish otoliths: daily growth layers and periodical pat-
terns. Sci. 173:1124-1127.

497

Radtke, R. L., and J. M. Dean.

1982. Increment formation in the otoliths of embryos, larvae,
and juveniles of the mummichog, Fundulus heteroclitus.
Fish. Bull., U.S. 80:201-215.
RiCKER, W. E.

1973. Linear regressions in fishery research. J. Fish. Res.
Board Can. 30:409-434.
Sablin, V. V.

1979. Sanma no nenrei ni kansuru mondai ni tsuite (Problems
on age composition of saury). (In Jpn.) Rep. Japan-Soviet
Coop. Res. Meet. Saury Scomber 11:114-117.
Sablin, V. V., and V. P. Pavlychev.

1982. Dependence of migration and catch of Pacific saury
upon thermal conditions. Bull. Tohoku Reg. Fish. Res. Lab.
44:109-117.
Smith, P. E., E. H. Ahlstrom, and H. D. Casey.

1970. The saury as a latent resources of the California Cur-
rent. Calif Coop. Oceanic Invest. Rep. 14:88-130.
Smith, P. E., and R. W. Eppley.

1982. Primary production and the anchovy population in the

FISHERY BULLETIN: VOL. 86, NO, 3

Southern California Bight: Comparison of time series.
Limnol. Oceanogr. 27:1-17.
Statistics and Information Department, Japan.

1985. Fisheries statistics of Japan 1984. Minist. Agric, For.
Fish., Jpn., 288 p.
Sunada, J. S.

1974. Age and growth of the Pacific saury, Cololabis saira.
Calif Fish Game 60:64-73.
Theilacker, G. H.

1980. Changes in body measurements of larval northern
anchovy, Engraulis mordax, and other fishes due to handling
and preservation. Fish. Bull., U.S. 78:685-692.
UCHIDA, K., S. IMAI, S. Mito, S. Fujita, M. Ueno, Y. Shojima,
T. Senta, M. Tahuku, and Y. Dotsu.
1958. Studies on the eggs, larvae and juvenile of Japanese
fishes. (In Jpn.) Kyushu Univ., Fukuoka, Japan, 89 p.
YusA, T.

1960. Embryonic development of the saury Cololabis saira
(Brevoort). Bull. Tohoku Reg. Fish. Res. Lab. 17:1-14.

498

OTOLITH ULTRASTRUCTURE OF SMOOTH OREO, PSEUDOCYTTUS
MACULATUS, AND BLACK OREO, ALLOCYTTUS SR, SPECIES

N. M. Davies,' R. W. Gauldie/2 S. A. Crane,' and
R. K. Thompson^

ABSTRACT

The ultrastructure of sagittal otoliths from 14 Pseudocyttus maculatiis and 25 Allocyttus sp. individuals
were examined to determine their suitability for estimating age in these two species. Scanning electron
microscopy revealed high levels of complexity in both external surface topography and internal struc-
tural organization in the sagittae of both species. Many different crystal forms were found, including
calcite-like prisms. A close similarity in otolith structure exists between the two species. Deposition of
check rings analagous to annual and daily growth increments was found to be irregular with the underlying
complexity of crystalline growth obscuring the finer (analogous to daily) growth rings, making their
periodicity difficult to validate and implying that with present techniques the sagittal otoliths of the oreo
species Pseudocyttus maculatus and Allocyttus sp. are not suitable for age estimation.

The smooth oreo, Pseudocyttus maculatus, and
the black oreo, Allocyttus sp., are two related
species of the family Oreosomatidae. They are both
important commercial species in New Zealand. The
black oreo is the most commonly caught oreo in
New Zealand waters, while the smooth oreo is the
second-most commonly caught oreo. Little is known
about the biology of these fish. The black oreo is
endemic to New Zealand while the smooth oreo
occurs in New Zealand, South Australian, South
African, and South American waters (Last et al.
1983). In the waters south of New Zealand, the
distributions of the two species overlap (McMillan
1985). The habitat range of the smooth oreo is be-
tween 650 and 1,200 m, and that of the black oreo is
between 600 and 1,200 m (McMillan 1985).

A preliminary examination of the ultrastructure
of otoliths [sagittae] of these fish was undertaken
as part of a study to establish an ultrastructural
basis for a suitable ageing technique. This study
describes the external and internal structure and
organization of the otoliths in terms of the suitabil-
ity of the various check rings of the sagittae for age
estimation.

MATERIALS AND METHODS

Three otoliths (the sagitta, astericus, and lapillus)

'Fisheries Research Centre, Ministry of Agriculture and Fish-
eries, Greta Point, Evans Bay Parade, P.O. Box 297, Wellington,
New Zealand.

^To whom reprint requests should be sent.

'Electron Microscope Unit, Kirk Building, Victoria University
of Wellington, Private Bag, Wellington, New Zealand.

are contained in the endolymphatic sac (Fig. la). The
sagitta is the largest otolith and is located in the
most ventral position in the sac. The arrangement
of the three otoliths in the endolymphatic sac ranges
between the primitive where a large astericus and
sagitta with no lapillus is present (Gauldie et al.
1986) and the typical teleost arrangement where a
small astericus is located close to the sagitta, and
an even smaller lapillus is displaced into the atrium
of the semi-circular canal. The oreosomatids are
primitive fishes taxonomically, lying in the order
Beryciformes (Nelson 1976), and the arrangement
of otoliths reflects the taxonomic position of the fish.
The orientation of otoliths described here refers to
the orientation in situ. The lateral face is the out-
ward (antisulcal) surface; the medial face is the in-
ward (sulcal) surface. Investigation was restricted
to the sagitta primarily because of the difficulties
in establishing homologies for daily and annual type
check rings in the astericus and lapillus.

Sagittae were dissected from 14 smooth and 25
black oreo individuals caught in bottom trawls off
the east coast of New Zealand. These individuals
ranged in length from 24.5 to 40.1 cm (black oreo)
and 35.1 to 51.2 cm (smooth oreo).

Whole otoliths were photographed at 6 x to 20 x
using a WILD'* photomicroscope. The sagittae were
embedded on glass slides in epoxy resin with the
antisulcus surface uppermost and finely ground on
a Struers Planapol-2 petrographic grinder. The

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

Manuscript accepted April 1988.

FISHERY BULLETIN: VOL. 86, NO. 3, 1988.

499

FISHERY BULLETIN: VOL. 86, NO. 3

500

DAVIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO

Figure la. —Relative positions of the otoliths in the right endolymphatic sac (ENS) of the smooth oreo in lateral view: astericus (AS),
lapillus (LA), and sagitta (SA). Anterior (A), dorsal (D), posterior (P), and ventral (V) surfaces are indicated. Magnifi-
cation = 6.3 X.

b. —Dorsal (regular) and ventral (irregular) lobes of the otolith in lateral view, nucleus (N), sulcus (S), and rest as in Figure

la. Magnification = 12.5 x.

c. — SEM of the lateral surface topography of the otolith. Scale bar = 1 mm.

ground surface was polished with 2000 grit wet and
dry paper to yield a smooth surface for etching. The
most successful results were obtained by etching
with a 0.1 M solution of disodium salt of EDTA. The
otoliths were immersed in this solution for 15 to 20
minutes. Other suitable etching solutions employed
were 1) a 1% solution of HCl for 20 to 30 seconds
and 2) a 2% solution of Histolab RDO (a commer-
cial etching solution comprising a mixture of HCl
and EDTA) for 5 minutes. A cellulose acetate peel
was made of the etched surface to obtain an exact
replica of the surface features. The peel was placed
on a microscope slide under a cover slip, cleared with
ethanol or distilled water, viewed, and photographed
using a Zeiss photomicroscope. Direct observations

of thin sections (about 20 ^m) of otoliths did not
show any more information than that observed in
acetate peels. Acetate peels had the advantage of
allowing successive grinds to be examined thereby
avoiding the problem of losing information that
might be located only in very narrow layers within
the otolith.

Scanning electron microscope (SEM) photographs
were taken of otoliths using a Phillips 505 SEM.
Whole otoliths were glued on to SEM pin type
mounts, cemented in position with contact cement,
and sputter-coated with gold at approximately 5
Torr. The external surface topography of both the
medial and lateral faces of the otoliths was photo-
graphed. Selected pieces of otoliths broken by thumb

501

FISHERY BULLETIN: VOL. 86, NO. 3

pressure were examined and photographed to ob-
tain internal structural information. Finely ground
cross-section surfaces were also polished and etched
for examination with the SEM.

RESULTS

Smooth Oreo Otolith

The sagitta is clearly divided into two distinct
structures: a small, smooth, regular dorsal lobe and
a larger, irregular ventral lobe (Fig. lb). The irreg-
ular lobe has branched crystal formations and clefts
at the ventral edge. The lateral face topography is
complex (Fig. Ic). The central bumpy area contains
the nucleus between the two lobes. Radiating out-
wards from the nucleus are concentric ridges on the
antisulcul surface.

The crystal morphology of the lateral surface of
the sagitta is variable. Over much of the central
parts of the sagitta, large and variably oriented

crystals give a coarse appearance to the surface of
the otolith (Fig. 2a) which presumably obstructs, by
diffraction, potentially clear zones in the whole
otolith viewed by transmitted light. At the edge of
the regular lobe, the crystal type alters to form slabs
of crystal layers (Fig. 2b). The analogous area on
the irregular lobe yields variable crystalline struc-
tures with complex alignments (Fig. 2c). Deep
troughs and branching furrows break up the crystal
forms at the edge of the irregular lobe.

The medial surface has three distinct parts: the
central sulcal area, the edge of the irregular lobe,
and the edge of the regular lobe (Fig. 3a). Raised
ridges and two prominent knobs are found in the
central sulcal area. Crystals compacted into a petal-
like growth pattern are found in this part of the
otolith (Fig. 3b). Contrasting to this, more porous
crystal structures occur in the edge areas of the
regular lobe (Fig. 3c). Further variety is found on
the irregular lobe, where a very porous, honeycomb-
like crystal arrangement exists (Fig. 3d). At the

Figure 2a. —Coarse crystal structure of the central lateral surface of the smooth oreo otolith. Scale bar = 0.1 mm.

b. —Split-screen SEM of the transition of the crystal-type at the edge of the lateral surface of the regular

lobe. Magnification = 163 x and 652 x.

c. —Haphazard crystal alignments at the edge of the lateral face of the irregular lobe. Scale bar = 0.1 mm.

502

DAVIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO

50iim30.1kU 1.63E2 6648/99 SS07A

503

FISHERY BULLETIN: VOL. 86, NO. 3

SejinSe.lkU 1.63E2 6719/99 88676

Figure 3a. —Medial view of the smooth oreo otolith showing the irregular lobe (I), regular lobe (R), and sulcus
(S). Scale bar = 1 mm.

b. —Split-screen SEM of petallike crystal growth in the sulcus on the medial surface. Magnification
= 163 X and 1141 x.

504

DA VIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO

* -'-^

r*^. >\. ^ *^^ -^r ^9^ K^"^ yur

,?■ -■

18jiii3e.lkU 1.31E3 6717/99 88676

d a»

10jiii30.1kU 1.62E3 6749/99 SS014B

Figure 3c. —Porous crystals on the medial surface of the regular lobe. Scale bar = 10 tixn.

d.— Honeycomblike crystal structure on the medial surface of the irregular lobe. Scale bar = 10 fim.

505

FISHERY BULLETIN: VOL. 86, NO. 3

Figure 3e.— A laminar pattern of crystalline growth on the medial face of the irregular lobe. Scale bar = 10 fxm.

edge of this lobe, a laminar crystalline growth pat-
tern develops, adding to the overall variation of
crystal structures on the medial surface (Fig. 3e).
Other studies of otolith ultrastructure present the
otolith as homogeneous in crystal form, composed
almost entirely of monoclinal aragonite crystals
(Degens et al. 1969). The complex crystallinity of
the oreo otoliths resembles that of the mollusc shell,
v^hich, although aragonitic, often has a pattern of
complex variation in crystal habit (Carriker et al.
1980).

Within the broken otolith, the nucleus lies at the
center of a spherical primordium (Fig. 4a). Crystals
grow outwards from the primordium and epitaxial
(Degens 1976) growth patterns exist (Fig. 4b). Com-
plex leaf-shaped crystals occur in areas directly
beneath the lateral surface of the fractured irregular
lobe (Fig. 4c). Beneath the medial surface, a remark-
able series of hexogonal crystals of calcite occur as
large rectangular blocks embedded within the
otolith (Fig. 4d). Calcitic prisms have been described
in molluscs as resulting from the regeneration of
broken shells (Watabe 1983). It is difficult to imagine
otoliths being broken and regenerated in situ.

Major and minor check rings similar to those

described (Gaul die 1987) for otoliths from the orange
roughy, a deepwater species from the same habitat,
occur (Fig. 4e). When polished and etched, the check
rings become clearly visible with deep etched checks
occurring between the less deeply etched checks
analagous to microscopic growth increments in
other species (Fig. 4f). At a higher magnification the
finer increments become obscured by the coarseness
of the underlying crystal type (Fig. 4g). Those fine
increments that are visible occur irregularly and
have varying widths.

Large-scale rings analogous to opaque/hyaline an-
nual zones were observed in the regular lobe. The
mean width of these zones, measured using
transmitted light (Fig. lb), was 0.34 mm (±0.06).
The concentric ridges observed by SEM on the
lateral surface of the regular lobe (Fig. Ic) have a
mean width of 0.27 mm (±0.06). These two struc-
tures, opaque/hyaline zones and surface ridges, have
about the same width with no statistically signifi-
cant differences between them.

Examination of otolith cross sections reveals
widely spaced, large rings with many finer rings in
between (Fig. 4h). When examined in greater detail,
the demarcation between the large and finer rings

506

DA VIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO

Figure 4a. —Broken smooth oreo otolith reveals the spherical primordium (PM) and nucleus (N) and the surround-
ing radial crystal growth. Scale bar = 0.1 mm.
b.— Epitaxial crystal development in the broken otolith. Scale bar = 0.1 mm.

507

FISHERY BULLETIN: VOL. 86, NO. 3

10iim30.lkU 1.49E3 6828/99 SSOl

Figure 4c.— Leaf-shaped crystals beneath the lateral surface of the broken otolith. Scale bar = 10 ^m.

d.— Calcitic prism crystals beneath the medial surface of the broken irregular lobe. Scale bar = lOji.

508

DAVIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO

e -

Figure 4e.— Discontinuous uniform crystal growth forming rings in the broken otolith. Scale bar = 0.1 mm.
f. —Fine and deep (arrowed) increments on the polished and etched otolith surface. Scale bar = 0.1 mm.

509

FISHERY BULLETIN: VOL. 86, NO. 3

g '

i . V pit

10iiin30.1kU 1.31E3 6725/99 SSeSB

Figure 4g.— Fine increments obscured by coarse crystallinity on the polished and etched otolith surface. Scale bar = 10 jim.

h.— Fine increments (arrowed) found between widely spaced large rings in the otolith cross-section. Magnification = 72x .

510

DAVIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO

j r , - ' • . • . v-

.4,

o:^^"^^i*i ^.^

X \,* •^-.^r ^ ^:j-j

5^-

,-* «

■^-

Figure 4i. —Variation of increment widths (arrowed) in the otolith cross-section. Magnification = 160 x.

j.— Cellulose-acetate peel showing the intermittency of fine increment sequences (arrowed), often obscured by a coarse underly-
ing crystallinity. Magnification = 1 60 x .

511

FISHERY BULLETIN: VOL. 86, NO. 3

becomes difficult to determine because of the inter-
mittent nature of the growth increments and the
variety of width-sizes (Fig. 4i). In large areas of the
otolith, increments appear to be absent or indeter-
minate, with a coarse underlying crystal structure
(Fig. 4j) making accurate determination of incre-
ment sequences difficult. However, the fine incre-
ments of the oreo otolith are 3 to 5 ^m wide, which
is within the range of daily growth increments
described for other species (Jones 1986; Gauldie in
press).

Black Oreo Otolith

The black oreo otolith is almost identical to that
found in the smooth oreo in overall shape, propor-
tion, structure, topography, surface and internal
crystallinity, and increment pattern. Some minor
differences do, however, exist.

In the medial sulcus, the prominent knobs found
in the smooth oreo otolith are smaller than in the
black oreo. On the surface of the sulcus, large leaf-
like crystals having various orientations occur (Fig.
5a). Also present in the sulcus are porous, sponge-
like crystals adjacent to membranous structures
(Fig. 5b). Smooth patches, where crystals appear ab-
sent, occur on the lateral surface of the otherwise
coarsely crystalline irregular lobe (Fig. 5c). At
higher magnification the smooth patches are seen
to be smaller growth forms of the larger adjacent
crystals.

DISCUSSION

Despite some minor differences in topography and
crystallinity, the sagittae of both species are essen-
tially identical. The otoliths are structurally complex
v^ath a great variety of crystalline forms. The coarse

Figure 5a. —Split-screen SEM of large leaflike crystals in the black oreo otolith sulcus. Magnification = 356 x and 979 x .
b.— Porous, sponge-like crystals in the otolith sulcus. Scale bar = 10 ^^m.
c. —Smooth patches on the central lateral surface of the otolith. Scale bar = 0.1 mm.

512

DA VIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO

513

FISHERY BULLETIN: VOL. 86, NO. 3

crystals on the central lateral surface are com-
parable with those in the oyster shell described as
individual laths (Carriker et al. 1980). The leaflike
crystals in the sulcus of the black oreo otolith are
similar to the chalky crystal forms in the oyster shell
(Carriker et al. 1980). Such a variety of crystalline
forms is uncommon in teleost otoliths. The low legi-
bility of structures of various kinds in the otolith may
reflect this complex crystallinity. However, the com-
plex crystallinity of the mollusc shell is thought to
reflect changes in both the external and internal
milieu of the organism (Wilbur and Saleuddin 1983).
Thus the difficulties of reading the oreo otolith in
the conventional sense may be offset by the life
history record (albeit difficult to translate) provided
by its complex crystallinity.

The broken sections of the otolith reveal the in-
ternal structure organization, and development of
crystals. Epitaxial crystal grov^h in the oreo otolith
results in columnar, monoclinal crystals of ara-
gonite. However, the presence of calcite-like prisms
has not been described for other otoliths. Calcite
occurs on the antisulcal surface of some otoliths
apparently by simple crystallization out of the fluid
of the endolymphatic sac (Morales-Nin 1985), but
calcite has never been described from within an
aragonite otolith (Carlstrom 1963). In molluscs,
calcite replacement of aragonite results in an orderly
alignment of calcite crystals following the alignment
of the original aragonite crystals. The disorderly
appearance of the calcite-like hexagons in the
smooth oreo otolith may be due to a diagenetic
transformation of aragonite to calcite with depth.
The compensation depth for the aragonite/calcite
transformation is about 3,000 m (Fyfe and Bischoff
1965) well beyond the known range of the smooth
oreo which has a maximum recorded depth of 1,300
m. However, there may be enough variation in
either the kind or amount of stabilizing protein in
the smooth oreo otolith to allow crystal changes to
occur at shallower depths than 3,000 m.

The complex, and often coarse, crystal structure
of the oreo otolith obscures the sequences of incre-
ments when they do occur. As a result, large rings
observed at low magnification become indistinct at
higher magnification when many finer increments
appear. The large rings could be assumed to be
annual check rings, but the difficulties in differen-
tiating between the fine and large rings create am-
bivalence in one's interpretation. The finer micro-
scopic growth increments, analagous to daily growth
rings, have no uniform width and occur intermit-
tently making accurate counting impossible.

The suitability of an otolith for determining the

age of a fish depends on the pattern of both annual
and daily check rings inferred from the structure
of the otolith. The hyaline/opaque zones observed
in the regular lobe (using transmitted light) had a
similar mean width to the concentric ridges found
by SEM on the lateral surface. The ^statistic we
obtained accepts the null hypothesis that no sig-
nificant difference exists between the two means.
Mel'nikov (1981) regarded these opaque/hyaline
zones as annual check rings in the otolith of Allo-
cyttus verrucosus. However, because no evidence ex-
ists for a relationship between the surface ridges and
fish age, it is possible that Mel'nikov's (1981) ages
are incorrect. Furthermore, the width of the opaque/
hyaline zones (0.34 mm) would indicate daily growth
increments less than 1 \xm wide. There are no
reports in the literature of validated daily growth
rings of such small size. In addition, the microscopic
growth increments which we have observed in the
oreo otolith are 3 to 5 ^m wide, which is a size range
commonly observed in other species.

With the techniques available we have been unable
to use either annual- or daily-type structures to
develop a technique for age estimation for Pseudo-
cyttus maculatus and Allocyttu^ sp. The reasons for
these difficulties may lie in the crystal morphology
of the otoliths which are more complex than any so
far described in the literature.

ACKNOWLEDGMENTS

All otoliths were supplied by Peter McMillan (Fish-
eries Research Centre). All 35 mm photographs
were processed by Alan Blacklock (Fisheries Re-
search Centre). SEM photographs were taken at the
SEM Unit, Zoology Department, Victoria Univer-
sity of Wellington.

LITERATURE CITED

Carlstrom, D.

1963. A crystallographic study of vertebrate otoliths. Biol.
Bull. (Woods Hole) 125:441-463.
Carriker, M. R., R. E. Palmer, and R. S. Prezant.

1980. Functional ultra-morphology of the dissoconch values
of the oyster Crassostrea virginica. Proc. Natl. Shellfish.
Assoc. 70:139-183.
Degens, E. T.

1976. Molecular mechanisms of carbonate, phosphate and
silica deposition in the living cell. Top. Curr. Chem. 64:
1-112.
Degens, E. T., W. G. Deuser, and R. L. Haedrich.

1969. Molecular structure and composition of fish otoliths.
Mar. Biol. 2:105-113.
Fyfe, W. S., and J. L. Bischoff.

1965. The caJcite-aragonite problem. In L. C. Pray and R. C.
Murray (editors), Dolomitization and limestone diagenesis,

514

DA VIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO

p. 3-13. Symp. Soc. Econ. Minist. Spec. Publ. 13.
Gauldie, R. W.

1987. The fine structure of check rings in the otolith of the
New Zealand orange roughy. NZ J. Mar. Freshwater Res.
21:267-274.
In press. A study of otolith daily growth rings in the orange
roughy {Hoplostethus atlanticus) aimed at resolving prob-
lems in age, growth, recruitment and otolith architecture.
NZ Fish. Res. Bull.
Gauldie, R. W., D. Dunlop, and J. Tse.

1986. The remarkable lungfish otolith. NZ J. Mar. Fresh-
water Res. 20:81-92.
Jones, C.

1986. Determining age of larval fish with the otolith incre-
ment technique. Fish. Bull., U.S. 84:91-103.
Last, P. R., E. 0. G. Scott, and F. H. Talbot.

1983. Fishes of Tasmania. Hobart, Tasmanian Fisheries
Development Authority.
Mel'nikov, Y. S.

1981. Size-age composition and growth pattern oiAUocyttus
verrucosus (Oreosomatidae). Ichthyol. (Engl, transl. Vopr.

Ikhtiol.) 21:178-184.
Morales-Nin, B.

1985. Caracteristicas de los otolitos cristalinos de Genypterus
capensis, (Smith, 1847) (Pisces: Ophidiidae). Invest. Pesq.
49:379-386.
McMillan, P. J.

1985. Black and smooth oreo dories, /n J. A. Colman, J. L.
McKoy, and G. G. Baird (editors), Background papers for
the 1985 total allowable catch recommendations. (Report
held in the Fisheries Research Centre Library, P.O. Box 297,
Wellington, New Zealand.)
Nelson, J. S.

1976. Fishes of the world. John Wiley and Sons, N.Y., 416 p.
Watabe, N.

1983. Shell repair. In A. S. M. Saleuddin and K. M. Wilbur
(editors). The mollusca, No/4, p 289-316. Acad. Press, N.Y.
Wilbur, K. M., and A. S. M. Saleuddin.

1983. Shell formation. In A. S. M. Saleuddin and K. M.
Wilbur (editors), The mollusca, No/4, p 236-288. Acad.
Press, N.Y.

515

RELATIONSHIP BETWEEN SEDIMENT CHARACTER AND
SEX SEGREGATION IN ENGLISH SOLE, PAROPHRYS VETULUS'

D. Scott Becker^

ABSTRACT

English sole, Parophrys vetulus, were collected by otter trawl in the nearshore zone of Puget Sound,
Washington, during two surveys conducted in 1981-82 and 1984. Stations in both surveys were distributed
across a broad range of sedimentary environments. Sex segregation by English sole was strongly
associated with the grain-size characteristics of bottom sediments. This association was persistent across
a variety of sampling conditions, including different years, seasons, embayments, and depths. Fish age
did not appear to influence the observed association. Results of this study provide the first documenta-
tion of the influence of sediment character on sex segregation by a pleuronectid, and suggest that this
relationship should be considered in future studies of English sole and, perhaps, other pleuronectids as well.

Sex segregation by fishes of the family Pleuronec-
tidae (i.e., righteye flounders) has been documented
for numerous species from the United States (Alver-
son and Chatwin 1957; Fadeev 1970; Alton 1972),
Europe (Rae 1965), and Asia (Moiseev 1953;
Kovtsova 1982). These patterns have been related
to such factors as geographic location, depth,
season, and age of fish. However, because most of
these observations have been made while fishes
were being collected for other purposes, the under-
lying reasons for the observed patterns are poorly
understood.

Several studies have documented sex segregation
in English sole, Parophrys vetulus, a pleuronectid
found in nearshore areas along the west coasts of
the United States and Canada (Hart 1973). In de-
scribing the catch of the English sole fishery in
Hecate Strait, British Columbia, Ketchen (1956)
noted that considerable sex segregation takes place,
particularly during the summer months. In a study
of the population characteristics of this species in
Puget Sound, Holland (1969) found a difference of
18% between the percentages of males in the pop-
ulations from two locations. Because both of these
accounts were largely anecdotal, no evaluations of
the observed patterns were made.

One factor that has yet to be evaluated with
respect to sex segregation by pleuronectids is the
character of bottom sediments. Because pleuronec-
tids live in close association with the sea floor, it

'Contribution No. 759, School of Fisheries, University of Wash-
ington, Seattle, WA 98195.

^School of Fisheries, University of Washington, Seattle, WA
98195; present address: PTI Environmental Services, 3625 132nd
Ave. S.E., Bellevue, WA 98006.

might be surmised that the characteristics of that
environment exert a differential influence on the
distributions of the two sexes. Additional support
for this hypothesis comes from the fact that the
influence of sediment character on the overall
distribution (i.e., both sexes pooled) of several
pleuronectids has been well documented (Ketchen
1956; Alverson et al. 1964; Feder et al. 1974; Pearcy
1978; Scott 1982; Becker 1984).

In this paper, I evaluate patterns of sex segrega-
tion by English sole in relation to the character of
bottom sediments in nearshore areas of Puget
Sound, WA. This study provides the first determina-
tion of how sediment character influences sex segre-
gation in a pleuronectid. In relating sediment
character to sex segregation in English sole, I also
evaluate the potential influences of depth, season,
and fish age on the observed relationships.

METHODS

English sole were sampled during two independ-
ent surveys conducted in several embayments of
Puget Sound in 1981-82 and 1984. The relationship
between sex segregation and sediment character
was evaluated separately for each survey. Although
the two surveys differed with respect to such vari-
ables as station locations, sampling times, and sam-
ple sizes, the basic collection and analytical methods
used in both surveys were similar.

In both surveys, the sex ratio of the English sole
captured at each station was represented by the pro-
portion of the sample comprised of males (i.e., male
proportion). Sediment character at each station was
represented by the percentage (by weight) of the

Manuscript accepted April 1988.

FISHERY BULLETIN: VOL. 86, NO. 3, 1988.

517

FISHERY BULLETIN: VOL. 86, NO. 3

grain-size distribution comprised of fine-grained
material (i.e., <0.063 mm in size). This variable was
used as an index of sediment character because
many physical, chemical, and biological character-
istics of sedimentary environments covary with the
grain-size distribution (Buchanan 1984).

During 1981-82, English sole were collected from
15 May to 6 July 1981 at 12 transects in Commence-
ment Bay, Elliott Bay, and Sinclair Inlet (Fig. 1).
Fishes also were collected during three additional
time periods (23-24 November 1981; 8-9 March
1982; 30-31 August 1982) to evaluate seasonal
patterns at three stations in Commencement Bay
(CB-1, CB-4, and CB-5; Fig. 1). These stations were
selected to represent a gradient of sediment grain
size. Although the seasonal sampling occurred over
a 2-yr period, it was assumed that interannual varia-
tion was not sufficiently large to obscure major
seasonal patterns.

During the May- July sampling, English sole were
collected at each transect during morning (07:00-
09:30), midday (10:00-13:30), evening (20:30-23:30),
and midnight (00:30-04:00). During the remaining
sampling periods, fish were collected only during
midday and evening. For each transect, results ob-
tained for different periods of the diel cycle were
pooled before male proportions were calculated. Sex
determinations and evaluations of spawning condi-
tion were made for all English sole larger than 160
mm total length (TL). Age was not determined.

During 1984, English sole were collected from 4
to 9 June 1984 at 15 transects in Commencement
Bay and at 2 transects in Carr Inlet (Fig. 1). Sam-
pling was conducted during daylight hours (06:30-
17:00) at all transects. Sex determinations and
evaluations of spawning condition were made for all
individuals larger than 225 mm TL. Age was deter-
mined by otolith analysis for all fish.

During both surveys, English sole were collected
using a 7.6 m (headrope) Marinovich otter trawl
having a body mesh of 3.2 cm (stretched) and a cod
end liner mesh size of 0.8 cm (stretched). Trawling
was conducted along isobaths at a constant vessel
speed of approximately 2.5 knots.

During both surveys, bottom sediments were sam-
pled using a modified van Veen bottom grab and the
grain-size distribution of the top 2 cm of sediment
was determined using standard sieve and pipette
techniques (Folk 1968). During 1981-82, sediments
were collected at five sampling points located at
approximately equal distances along each transect.
During 1984, sediments were collected at a variable
number of stations located within 300 m from each
transect.

The association between percent fine-grained
sediment and male proportion of English sole was
tested using Spearman's rank correlation coefficient
{Vg). The association between station depth and
male proportion was tested in the same manner.
Differences in male proportion among age-classes
of fish and among the seasonal sampling transects
were tested using the G-test (Sokal and Rohlf
1981).

RESULTS

Station depth, percent fine-grained sediment,
sample size, and male proportion at each transect
from the 1981-82 and 1984 surveys are presented
in Tables 1 and 2, respectively. During 1981-82,
4,430 English sole were sampled from stations
ranging in depth from 8 to 36 m. Percent fine-
grained sediment ranged from 5.8 to 92.1% and
male proportion ranged from 0.20 to 0.91. During
1984, 1,007 English sole were collected from
stations ranging in depth from 5 to 20 m. Per-
cent fine-grained sediment ranged from 8.6 to
87.5% and male proportion ranged from 0.13 to
0.98.

The correlation between percent fine-grained sedi-
ment and male proportion of English sole was
significant (P < 0.01; r, = 0.73) for the May- July
sampling of the 1981-82 survey (Fig. 2). This cor-
relation was also significant (P < 0.001; r^ = 0.80)
for the 1984 survey (Fig. 2). Correlations between
depth and male proportion were not significant {P
> 0.05) for either the 1981-82 (r, = -0.10) or 1984
(Vg = -0.04) survey.

For the stations sampled seasonally in 1981-82
(CB-1, CB-4, and CB-5), heterogeneity of male pro-
portion among stations was significant {P < 0.05)
during all four sampling periods (Fig. 3). In all cases,
male proportion was lowest at CB-5, highest at
CB-4, and intermediate in magnitude at CB-1. This
gradient of increasing male proportion was consis-
tent with the gradient of increasing percent fine-
grained sediment at the three stations (Table 1). The
mean abundance of English sole at the three sta-
tions rose from a minimum level in March to a peak
in May-July, and then declined through August to
a level in November that was similar to that in
March (Fig. 3).

For both surveys combined, 64 (1.2%) English sole
were found to be in spawning condition. Twenty-
four of those individuals were collected during
March, to represent a spawning component of 14.8%
during that month. The remaining 42 individuals
were sampled between May and July to represent

518

BECKER: SEDIMENT AND SEX SEGREGATION IN ENGLISH SOLE

^^ Elliott Bay

^^EB-1

'^^EB-2

SEATTLE

I-

EB-4

LEGEND

CR-Carr Inlet
E B - Elliott Bay
SN - Sinclair Inlet

ii NALmCAL MLES

POINT DEFIANCE

LEGEND

Hyielxjs Waterway

Blair Waterway

Silcum Waterway

Milwaukee Waterway

St Paul Waterway

Middle Waterway

City Waterway

Souttiwest Shoreline

Commencement Bay

KILOMETERS

Figure 1.— Locations of transects sampled in 1981-82 (denoted by CB, EB, and SN) and 1984 (all remaining

transects).

519

FISHERY BULLETIN: VOL. 86, NO. 3

Table 1. — Characteristics of transects and samples of English sole during 1981-82.'

Mean percent
fine-grained

SampI

e size^

Male proportion

Depth

May-

Novem-

May-

Novem-

Station^

(m)

sediment^

July

August

ber

March

July

August ber

March

CB-1

27.1

655

249

0.83

0.74 0.58

0.57

CB-2

47.0

554

0.78

CB-3

44.2

160

0.79

CB-4

71.6

589

146

174

0.91

0.82 0.77

0,87

CB-5

7.6

181

0.55

0.64 0.44

0.20

CB-6

11.2

128

0.66

EB-1

56.0

0.68

EB-2

76.2

0.76

EB-3

5.8

0.53

EB-4

8.8

0.49

SN-1

92.1

683

0.84

SN-2

13.7

310

0.30

'Sampling in May-July and November was conducted in 1981 , whereas sampling in August and March was conducted in 1982.
^Locations of transects are presented in Figure 1 .
'Each value Is based on five replicate samples.
<AII English sole were larger than 160 mm TL.

Table 2.— Characteristics of transects and samples of English sole
during 1984.

Mean percent

Male

Depth

fine-grained

Sample

propor-

Station'

(m)

sediment^

size^

tion

CR-1

14.5 (2)

0.28

CR-2

8.6 (2)

0.13

SS-1

18.2 (2)

0.29

SS-2

22.9 (6)

0.48

SS-3

8.7 (6)

0.35

CI-1

68.8 (4)

0.75

CI-2

58.2 (4)

0.73

MD-1

40.4 (2)

0.42

SP-1

48.9 (3)

0.77

MI-1

87.5 (5)

0.87

SI-1

73.8 (5)

0.98

BL-1

55.2 (1)

0.77

BL-2

74.3 (9)

0.88

BL-3

55.4 (6)

0.88

HY-1

68.1 (5)

0.45

HY-2

79.8 (7)

0.80

HY-3

61.1 (12)

0.73

'Locations of transects are presented in Figure 1.
^Number of stations used to calculate mean percent fine-grained sediment
at each transect is given in parentheses.
'All English sole were larger than 225 mm TL.

I-

DC
Q.

Figure 2.— Comparison of percent fine-grained sediment and male
proportion of English sole populations using Spearman's rank
correlation coefficient (rj. *• = P < 0.01, *** = P < 0.001.

i.u -

•

0.8 -

•

••

•
•

•

0.6 -

•:

0.4 -

•

n? -

MAY- JULY 1981

rs=0.73**

1 1

100

i.u -

•

• • •

0.8-

•

0.6-

0.4-

•
•

•

••

n? -

JUNE 1984

•

rs = 0.80***

1 1

20 40

60 80 100

PERCENT FINE-GRAINED
SEDIMENT

520

BECKER: SEDIMENT AND SEX SEGREGATION IN ENGLISH SOLE

CB-5 illij CB-1

CB-4

***

o
o

o
oc
a.

MARCH MAY-JULY AUGUST NOVEMBER

1982 1981 1982 1981

SAMPLING PERIOD

Figure 3.— Comparisons of male proportion of English sole populations at three Commencement
Bay stations during four sampling periods using the G-test. Mean abundance of English sole at
all three stations is plotted for each sampling period. The values of percent fine-grained sedi-
ment at Stations CB-5, CB-1, and CB-4 were 7.6%, 27.1%, and 71.6%, respectively. ' = P <
0.05, ••• = P< 0.001.

a spawning component of 0.9% during that period.
No fish were found to be in spawning condition dur-
ing August and November.

For English sole aged 3-10 years old during the
1984 study, male proportion for each age-class

ranged from 0.50 to 0.66 (Fig. 4). However, the
heterogeneity among age-classes was not significant
(P > 0.05). Nineteen fish aged 2 or >10 years old
were not included in this analysis because sample
sizes by age-class were too small.

DC
Q.

0.8 -1

0.6-

0.4

0.2-

126 217 261 209 97 40 24 14

Figure 4.— Comparison of male proportion among age-
classes of English sole collected in 1984. Sample size
is presented above each bar.

T 1 \ \ \ 1 I r

3456789 10

FISH AGE (yr)

521

FISHERY BULLETIN: VOL. 86, NO. 3

DISCUSSION

Results of this study demonstrate that sex segre-
gation by English sole in nearshore areas of Puget
Sound is strongly associated with the grain-size
distribution of bottom sediments. In addition to the
results of statistical analyses, the strength of this
association is demonstrated by its persistence across
a variety of sampling conditions. For example, the
association was found in two independent surveys
conducted 2-3 years apart. Furthermore, the 1981-
82 survey represented an analysis of large-scale
patterns across three embayments, whereas the
1984 survey focused on small-scale patterns within
a single embayment. The association between sex
segregation and sediment grain size also persisted
across seasons and across a range of depth.

Although a significant association was found be-
tween sediment grain size and male proportion of
English sole, a variety of factors that covary with
sediment grain size may have been partly or totally
responsible for this association (Lagler et al. 1977;
Moyle and Cech 1982). Examples of physical/chem-
ical variables that often covary with grain size and
may influence fish distributions include depth, cur-
rent speed, turbidity, and dissolved oxygen. Ex-
amples of biological factors include the composition
of prey assemblages and the distribution of pred-
ators. Although results of this study demonstrated
that depth did not influence the observed associa-
tion between sediment grain size and male propor-
tion, further experimentation is required to identify
the factor or suite of factors that directly influences
sex segregation in this species.

Despite the fact that depth did not influence the
patterns of sex segregation identified for English
sole, the range of depth considered in this study
(5-36 m) is a small fraction of the total range
occupied by this species (0-550 m; Hart 1973). It
therefore is possible that depth-related sex segrega-
tion in this species may be observed if a wider range
of depths were examined. For example, considerable
sex segregation by depth was found for Dover sole,
Microstomus pacificus, off the west coast of the
United States when a large depth range (110-440
m) was evaluated (Hagerman 1952; Alton 1972).

Based on results of the 1984 survey, no relation-
ship between age and male proportion in English
sole was evident. It therefore is unlikely that age
influenced the observed relationship between male
proportion and sediment grain size. A decline in
male proportion with increasing age has been found
for several pleuronectids (Rae 1965; Kovtsova 1982).
Because most of the fish examined in this study were

relatively young (<8 years old) compared with the
maximum age of 15-17 years sometimes reached by
individuals of this species (Ketchen 1956; Holland
1969; Van Cleve and El Sayed 1969), consideration
of older fish may alter the patterns observed in this
study. Although sample sizes for fish aged 9-10
years were relatively small in this study, male pro-
portion in those age groups (0.50) was lower than
that observed in all younger age groups (0.57-0.66).

The season of sampling did not alter the observed
relationship between male proportion and sediment
grain size. This relationship remained intact despite
the fact that the mean abundance of English sole
changed substantially throughout the seasonal cycle.
The observed seasonal changes in mean abundance
are consistent with the seasonal pattern of migra-
tion exhibited by adults of this species and most
other pleuronectids. That pattern includes a migra-
tion to deeper water in the fall for overwintering
and spawning, and a return migration to shallower
water in spring for feeding throughout the summer
(Ketchen 1956; Alverson et al. 1964; Roff 1982). It
is unlikely that spawning substantially influenced
the observed seasonal patterns of sex segregation
in English sole, because the largest percentage of
fish in spawning condition during any season was
<15%. However, sex segregation in relation to
spawning in pleuronectids has been documented for
several species (Moiseev 1953; Alverson and Chat-
win 1957), and may influence sex segregation in
English sole when a substantial percentage of the
population is in spawning condition.

Results of this study suggest that the influence
of sex segregation should be considered when
characteristics of English sole populations are com-
pared among habitats that differ with respect to
sediment character. Patterns based on population
characteristics capable of exhibiting a dependence
on sex (growth, condition, disease prevalence, tissue
contamination) could be strongly biased if differ-
ences in the sex ratio among stations are not con-
sidered. This concern should probably be extended
to studies of other pleuronectids, because the rela-
tionship between sex segregation and sediment
character is unknown for most species.

Although the reasons for the observed patterns
of sex segregation in English sole are unknown, a
potential explanation is that different energetic
needs induce male and female fish to utilize differ-
ent habitats. Female English sole grow to a larger
size than males and mature at an older age (Ketchen
1956; Holland 1969; Van Cleve and El-Sayed 1969;
Hart 1973). In addition, fecundity in females is pro-
portionate to individual size (Ketchen 1956; Hart

522

BECKER: SEDIMENT AND SEX SEGREGATION IN ENGLISH SOLE

1973). These patterns are found for numerous
pleuronectids and are thought to allow females to
outgrow predators and maximize fecundity (Roff
1982). Female English sole may therefore select
different habitats than males to support their differ-
ential reproductive requirements. Because many
characteristics of benthic macroinvertebrate assem-
blages (i.e., the primary prey of English sole) are
related to sediment character (Gray 1974; Rhoads
1974; Pearson and Rosenberg 1978), selection of
habitats that differ in sediment grain size provides
females with different prey spectra and energetic
potentials than are available to males. In addition,
differential habitat selection would reduce the poten-
tial of competition for food between sexes.

In summary, sex segregation by English sole was
strongly associated with the grain-size character-
istics of bottom sediments. This association was
persistent across a variety of sampling conditions,
including different years, seasons, embayments, and
depths. Fish age did not appear to influence the
observed association. Results of this study provide
the first documentation of the influence of sediment
character on sex segregation in a pleuronectid, and
suggest that this relationship should be considered
in future studies of English sole and, perhaps, other
pleuronectids as well.

ACKNOWLEDGMENTS

The 1981-82 survey was supported by the
National Oceanic and Atmospheric Administration
(Contract NA80RAD00050), and was part of a
dissertation submitted to the School of Fisheries of
the University of Washington (Seattle) in partial
fulfillment of a Ph.D. degree for the author; A. J.
Mearns was the Project Officer. The 1984 survey
was supported under Cooperative Agreement No.
CX810926-01-0 between the U.S. Environmental
Protection Agency and the State of Washington
Department of Ecology; J. D. Krull of Ecology was
the Project Manager. I thank R. A. Pastorok and
two anonymous reviewers for their helpful sugges-
tions on the manuscript.

LITERATURE CITED

Alton, M. S.

1972. Characteristics of the demersal fish fauna inhabiting
the outer continental shelf and slope off the northern Oregon
coast. In A. T. Pruter and D. L. Alverson (editors), The
Columbia River estuary and adjacent ocean waters, p. 586-
634. Univ. Wash. Press, Seattle.
Alverson, D. L., and B. M. Chatwin.

1957. Results from tagging experiments on a spawning stock

of petrale sole, Eopsetta jordani (Lockington). J. Fish. Res.
Board Can. 14:953-974.
Alverson, D. L., A. T. Pruter, and L. L. Ronholt.

1964. A study of demersal fishes and fisheries of the north-
western Pacific Ocean. H. R. MacMillan Lect. Fish., Inst.
Fish., Univ. B.C., Vancouver, 190 p.

Becker, D. S.

1984. Resource partitioning by small-mouthed pleuronectids
in Puget Sound, Washington. Ph.D. Thesis, Univ. Wash-
ington, Seattle, 120 p. + appendices.
Buchanan, J. B.

1984. Sediment analysis. In N. A. Holme and A. D.
Mclntyre (editors), Methods for the study of marine benthos,
p. 41-65. Blackwell Scientific Publications, Boston, MA,
387 p.
Fadeev, N. S.

1970. The fishery and biological characteristics of yellowfin
soles in the eastern part of the Bering Sea. Transl. in Soviet
Fisheries Investigations in the northeastern Pacific (Israel
Program Sci. Transl), Part V, p. 332-396.
Feder, H. M., C. H. Turner, and C. Limbaugh.

1974. Observations on fishes associated with kelp beds in
southern California. Calif. Fish Game, Fish Bull. 160, 138 p.
Folk, R. L.

1968. Petrology of sedimentary rocks. Univ. Texas, Austin,
170 p.

Gray, J. S.

1974. Animal-sediment relationships. Oceanogr. Mar. Biol.
Annu. Rev. 12:223-261.
Hagerman, F. B.

1952. The biology of the Dover sole, Microstomus pacificvs
(Lockington). Calif. Fish Game, Fish Bull. 85, 48 p.

Hart, J. L.

1973. Pacific fishes of Canada. Fish. Res. Board Can., Bull.
180, 740 p.
Holland, G. A.

1969. Age, growth, and mortality of races of English sole
{Parophrys vetulus) in Puget Sound, Washington. Pac.
Mar. Fish. Comm., Bull. 7:36-50.

Ketchen, K. S.

1956. Factors influencing the survival of the lemon sole
(Parophrys vetulus) in Hecate Strait, British Columbia. J.
Fish. Res. Board Can. 13:647-694.
KOVTSOVA, M. V.

1982. Size-age structure and sex ratio of the Barents Sea
population of the plaice, Pleuronectes platessa (Pleuronec-
tidae). J. Ichthyol. 22:68-79.
Lagler, K. F., J. E. Bardach, R. R. Miller, and D. R. M.
Passino.

1977. Ichthyology. John Wiley and Sons, N.Y., 506 p.
MOISEEV, P. A.

1953. Cod and flounder of the Far Eastern Seas. Fish. Res.
Board Can. Transl. Ser., No. 119, 576 p.

MOYLE, P. B., AND J. J. CeCH.

1982. Fishes: an introduction to ichthyology. Prentice-Hall,
Englewood Cliffs, NJ, 593 p.
Pearcy, W. G.

1978. Distribution and abundance of small flatfishes and other
demersal fishes in a region of diverse sediments and bathym-
etry off Oregon. Fish. Bull., U.S. 76:629-640.

Pearson, T. H., and R. Rosenberg.

1978. Macrobenthic succession in relation to organic enrich-
ment and pollution of the marine environment. Oceanogr.
Mar. Biol. Annu. Rev. 16:229-311.

Rae, B. B.

1965. The lemon sole. Fishing News Ltd., Lond., 106 p.

523

FISHERY BULLETIN: VOL. 86, NO, 3

Rhoads, D. C. shelf. Can. J. Fish. Aquat. Sci. 39:943-947.

1974. Organism-sediment relations on the muddy sea floor. SoKAL, R. R., AND F. J. Rohlf.

Oceanogr. Mar. Biol. Annu. Rev. 12:263-300. 1981. Biometry. W. H. Freeman and Co., San Francisco,

ROFF, D. A. CA, 859 p.

1982. Reproductive strategies in flatfish: a first synthesis. Van Cleve, R., and S. Z. El Saved.

Can. J. Fish. Aquat. Sci. 39:1686-1698. 1969. Age, growth, and productivity of an English sole

Scott, J. S. {Parophrys vetuliis) population in Puget Sound, Washing-

1982. Selection of bottom type of groundfishes of the Scotian ton. Pac. Mar. Fish. Comm., Bull. 7:51-71.

524

OCCURRENCE OF CANCER CRAB (C MAGISTER AND

C. OREGONENSIS) MEGALOPAE OFF THE WEST COAST OF

VANCOUVER ISLAND, BRITISH COLUMBIA

Glen S. Jamieson and Antan C. Phillips^

ABSTRACT

The seasonal and cross-shelf occurrences of Cancer crab (C. magister and C. oregonensis) megalopae
in 1985 along a transect line perpendicular to the coast off Tofino, British Columbia, are presented.
Megalopae of both species were generally absent from surface waters during daylight hours. The two
species may have slight temporal differences in nocturnal surface abundance, with C. magister occur-
ring later in the evening and earlier in the morning than C. oregonensis. Their relative occurrence at
the surface during the night was used to calibrate cross-shelf megalopal abundance data. Cross-shelf
megalopal intermolt stage proportions were calculated, relating degree of megalopal development to
proximity of nearshore habitat required for successful settlement of larvae.

Cancer magister megalopae were present from April to August, with peak abundance in May and
June. Megalopae were abundant in a broad band 37-148 km from shore, with peak abundance (2,871
10 m"^) 56 km offshore in June. In May, some late stage megalopae were collected in coastal inlets but
settlement appeared low in coastal study areas. Megalopal abundance decreased abruptly shoreward of
28 km from the coast.

Cancer oregonensis megalopae were also present from April to August, with their pattern of cross-
shelf abundance basically similar to that of C. magister. However, in contrast to C. magister, abundance
of late stage megalopae in coastal inlets was relatively high (313 10 m"^) in June, indicating that a signifi-
cant settlement of megalopae of this species could have occurred.

Evidence for cross-shelf movement of Cancer megalopae is discussed.

Dungeness crab, Cancer magister, range from the
Aleutian Islands to northern Mexico in the eastern
Pacific (Hart 1982) and are commercially exploited
from northern California to Kodiak Island, AK. As
part of an ongoing study of Dungeness crab recruit-
ment off the west coast of Vancouver Island, the
abundance and distribution of larvae off Tofino,
British Columbia, are being studied to determine
how variability in annual recruitment is affected by
larval settlement.

Dungeness crab larvae are planktonic and pass
through five zoeal stages and one megalopal stage
before settling to the sea bottom. Studies of larvae
prior to the 1970s primarily involved descriptions
of larval morphology (Mir 1961; Poole 1966), and it
was not until Reed (1969) developed laboratory
culture methods that larval environmental require-
ments were first described. Optimal ranges of tem-
perature and salinity for laboratory-cultured zoeae
were 10.0°-13.9°C and 25-30%o respectively, but
their survival was not significantly affected by the
temperature and salinity ranges occurring in the

waters off Oregon, where Reed's study was con-
ducted, at the time of year when larvae are com-
monly found. Lough (1976) suggested offshore lar-
val movement would allow larvae to avoid lower
nearshore salinities, and that normal oceanic salinity
levels probably favor survival over the long term.

The temporal occurrence of larvae in open coast
oceanic waters varies somewhat according to lati-
tude, with larvae present earliest in the season in
the southern part of the species' range. Seasonal
occurrence has not been well documented in the in-
shore waters of Georgia Strait, Puget Sound, and
Juan de Fuca Strait, but appears to differ signifi-
cantly from that in open coast waters. Larval settle-
ment, which typically occurs in May and June off
the outer coast of Washington, can occur as late as
mid-September in northern Puget Sound (D. Arm-
strong2).

Temporal and spatial distributions of crab larvae
have been documented for the years 1975-80 in the
Gulf of Farallones and the San Francisco-San Pablo-
Suisun Bay complex in central California (Reilly

'Department of Fisheries and Oceans, Fisheries Research
Branch, Pacific Biological Station, Nanaimo, B.C., Canada V9R
5K6.

^D. Armstrong, University of Washington, Seattle, WA 98195,
pers. commun. December 1986.

Manuscript accepted March 1988.

FISHERY BULLETIN: VOL. 86, NO. 3, 1988.

525

FISHERY BULLETIN: VOL. 86, NO. 3

1983) and for the years 1969-70 along a transect
off Newport, OR (Lough 1975, 1976). These studies
suggest that while larvae are hatched in the near-
shore, shallow-water environment preferred by
adult Dungeness crab, they subsequently move
alongshore and offshore, and then inshore, often
over considerable distances. Larvae hatched off
California and Oregon in late winter have been
hypothesized to move northward with the Davidson
Current before its reversal in March- April, and off-
shore as a result of upwelling, Ekman transport,
estuarine runoff, and geostrophic flow, depending
on location and season (Lough 1976; Wild and Tasto
1983). Later stage zoeae are typically found pro-
gressively further offshore, and it seems to be the
megalopal stage that returns inshore (Lough 1976;
Reilly 1983). The megalopal stage is the strongest
swimming stage (Jacoby 1982), but there is no direct
evidence to indicate that their inshore movement is
active.

There are five species of crabs of the genus Cancer
in British Columbia (Hart 1982), and larvae of C.
oregonensis also occur in abundance in offshore
waters along with C. magister (Lough 1975). Cancer

oregonensis has no commercial importance, but since
it occurs with C. magister, data on both species of
crabs are included in this study. Comparisons of
occurrence between the two species may provide
insight into possible environmental mechanisms or
processes that influence transport from offshore
areas to the inshore juvenile habitats, which the lar-
vae of both species must reach.

This study establishes the seasonal and spatial
occurrences of Cancer megalopae along a transect
over and beyond the continental shelf off the west
coast of Vancouver Island (Fig. 1). The geograph-
ical area is of particular interest because of the loca-
tion of a major regional crab fishery near Tofino
(Jamieson 1985; Noakes and Jamieson 1986) and the
resulting importance of understanding factors influ-
encing the magnitude of local larval crab settlement.
The oceanography off Vancouver Island has been
relatively well studied (Freeland et al. 1984; Thom-
son 1981), and because of the intrusion of Juan de
Fuca Strait waters into the general longshore ocean-
ographic regime and the relative increase in topo-
graphical complexity of the continental shelf, it is
considerably more complicated than that found

Figure 1.— The survey transect line, with stations, over which this study was conducted.

526

JAMIESON and PHILLIPS: OCCURRENCE OF CANCER CRAB MEGALOPAE

south of Cape Flattery (Hickey 1979). Along the
eastern Pacific coast, Cape Flattery marks a sig-
nificant change in nearshore oceanography, and
this study to investigate larval crab distribution in
open coast waters is the first one north of this
landmark.

GENERAL OCEANOGRAPHY OFF
VANCOUVER ISLAND

A seasonal surface current moving northward
from about lat. 32°N to 51 °N (northern Vancouver
Island) typically exists in nearshore waters from
October to March (Fig. 2). South of Cape Flattery,
this is called the Davidson Current (Hickey 1979)
whereas off British Columbia, it appears to consist
of two components. Nearshore, there is the year-
round, northward flowing Vancouver Island Coastal
Current, about 20-30 km wide, and further offshore,
there is the seasonal Shelf-Break Current, perhaps
a continuation of the Davidson Current, which
reverses direction in response to changes in the
large-scale wind field (Freeland et al. 1984). The
Davidson Current off Oregon has an average
northward flow of 50 cm/s for 30% of the time
(Boisvert 1969), sufficient to transport larvae sig-
nificant distances northward during their develop-
mental period if they remained continuously in the
current. The California Current is a seasonal, south-
ward moving, surface current of similar magnitude
and location to the Davidson Current. It occurs off
Vancouver Island (Hickey 1979) in the spring and
summer, along with the seasonal, southward flow-
ing, Shelf-Break Current on the outer continental
shelf (Freeland et al. 1984). These currents could
transport larvae located off Vancouver Island dur-
ing the spring and summer southwards (Fig. 2).
North of Vancouver Island, outer continental shelf
currents are poorly described, but off Cape St.
James (southern tip of the Queen Charlotte Islands),
the surface flow is strongly southward almost year-
round, turning northward only in March and April
(Freeland et al. 1984). The Vancouver Island Coastal
Current originates at the mouth of Juan de Fuca
Strait, and so could transport larvae out of the Puget
Sound-Georgia Strait complex.

MATERIALS AND METHODS

General Methodology

Sampling was largely done along an offshore
transect line perpendicular to the coast off Tofino,
British Columbia (Fig. 1), with stations located at

0, 9.3, 18.5, 27.8, 37, 46, 55.6, 74.1, 92.3, 111.2,
148.2, and 185.3 km from shore. Stations were more
closely spaced nearshore where larval abundance
was expected to be greatest, but the transect ex-
tended well beyond the shelf break to determine
what the seaward distribution of megalopae might
be.

Six offshore samplings were made along the tran-
sect from February through August 1985, but only
three extended the full 185 km. Dates of sampling
were 27 February-4 March, 17 and 18 April, 30 and
31 May, 14 and 15 June, 6 and 7 July, and 24 and
25 August. February-March, June, and July sam-
plings were terminated 74 km, 130 km, and 111 km
offshore, respectively, because of poor weather and
time constraints.

At each station, a neuston tow and an oblique
bongo tow to 250 m, or to within 5 m of bottom,
were made, and a temperature profile was recorded
by means of an expendable bathythermograph. On
all cruises, bongo tows were done during daylight
hours and neuston tows at night; late-stage crab
larvae congregate near the water surface at night
(Booth et al. 1985). On 17 June 1985, an hourly, noc-
turnal series of neuston tows was made at one loca-
tion, 9.3 km offshore, to identify patterns of night-
time abundance of megalopae in surface waters.

To supplement offshore sampling along the tran-
sect, neuston tows were made after dusk in the in-
lets around Tofino biweekly from early May through
August 1985 to monitor presence and movement of
larvae. In addition, a beam trawl and epibenthic sled
were used in inshore waters in July 1985 to sample
for newly settled larvae in an effort to establish time
and magnitude of larval settlement.

The RV G. B. Reed was used for all offshore
sampling except for the June sample, which was
collected from a 27 m charter vessel. Inshore sam-
pling was conducted from a 7 m aluminum herring
skiff except for mid-June, when the charter vessel
was again used.

Gear

The neuston sampler was a modified otter surface
sampler (Mason and Phillips 1986) with a square
mouth opening 45 cm on each side; under calm sea
conditions, it sampled the top 35 cm of the water
column. A General Oceanics flowmeter^ in the
mouth of the net was used to establish volume of
water filtered. Netting was black, 500 pt Nitex, and

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

527

FISHERY BULLETIN: VOL. 86, NO. 3

WINTER

54°

51'

48°

CONFUSED AND
^;i^;S^ VARIABLE
CURRENTS

54°

51°

48°

Figure 2.— Prevailing surface circulation off the British Columbia-Washington coast in winter and
summer. Broken arrows indicate uncertain currents. Numbers give speeds (cm s"^) (modified from
Thomson 1981).

528

JAMIESON and PHILLIPS: OCCURRENCE OF CANCER CRAB MEGALOPAE

was the same size as that on the bongo gear. The
neuston sampler was towed at 4 kn approximately
10 m from the side of the vessel to reduce hull
avoidance by the larvae. Tow duration was usually
15 minutes but was shortened when crab larvae
occasionally became very abundant at dawn and
dusk.

Bongo gear was a modified SCOR design (Mason
et al. 1984), having a mouth opening of 25 cm (each
side) with the outboard (left) net of 500 ^a Nitex and
the inboard (right) net of 230 ^ Nitex. A General
Oceanics flowmeter was mounted in the mouth of
each net. Sampling procedure followed that de-
scribed by Smith and Richardson (1977*). At stations
<100 m water depth, the bongo was fished in an un-
dulating fashion from the bottom to the surface in
order to filter a standard volume of water (about
300 m^).

The epibenthic sled (Phillips and Mason 1986) used
in inshore waters had a 60 cm mouth opening and
1 mm Nitex netting; it was towed for 10 minutes
at 2-3 kn. Nine tows were made over bottoms of
unknown characteristics. The modified beam trawl
(Gunderson and Ellis 1986), also used in inshore
waters, had a 3 m mouth opening, a 7 mm mesh net,
and a 3 mm mesh cod end; it was towed for 10
minutes at about the same speed as the sled. For
both gear types, distance towed was calculated by
radar triangulation to reference points on land.

Data Analysis

All plankton samples were preserved initially in
a 4% formaldehyde solution of saltwater. In the
laboratory, settled volume was determined, and
general composition of the plankton noted. Samples
were then sieved, and Cancer larvae picked by hand
and returned to a 2% formaldehyde solution before
their identification. Cancer magister megalops are
readily identifiable by their larger size (Trask 1970;
Lough 1975), but the currently used key (Lough
1975) was not always effective in separating C. ore-
gonensis and C. jyroductus. Presence of lateral spines
is a subjective criterion, and morphological dimen-
sions and counts of setae were two variable to
distinguish species. These smaller megalops were
finally concluded to be C. oregonensis after rearing
hundreds of larvae to the juvenile stage and finding
no C. productus.

With beam trawl samples, the catch was sorted

*Smith, P. R., and S. L. Richardson. 1977. Manual of methods
for fisheries resource survey and appraisal. Part 4. Standard tech-
niques for pelagic fish egg and larval surveys. Southwest Fish.
Cent., Natl. Mar. Fish. Serv., NOAA, Adm. Rep. No. 77-11.

on deck to remove juvenile crabs, and the megalopae
were preserved as above. Species composition of the
catch was noted.

For the bongo tows, the volume of water filtered
was used to calculate a haul factor (Mason et al.
1984), which accounted for tow depth and allowed
expression of the data as the integrated number of
organisms beneath 10 m^ of sea surface.

For neuston tows, relative abundance was ex-
pressed as area swept rather than as volume filtered
and again expressed as number of animals under 10
m^ of sea surface.

Calculations of larval abundance are necessarily
conservative, and direct comparisons between dif-
ferent gear types are not presently possible, given
our current understanding of gear efficiency and lar-
val catchability, which varies with both depth and
time of day. Numbers of larvae reported here are
thus directly comparable only within each gear type
used. For bottom gear, numbers of crabs were cal-
culated with no consideration of gear efficiency.

Molt Staging

State of development within the megalopal stage
was determined for all Cancer megalopae collected
at a station, or 25 randomly selected individuals of
each species if the number collected exceeded 25.
The sequence of epidermal changes occurring dur-
ing this intermolt period has been described for
Dungeness crab by Hatfield (1983), and her proce-
dures and staging criteria were applied for both
species. Whole megalopae were stained with 0.25%
Fast Green dye in either water or polyvinyl lacto-
phenol for 2-18 hours, and then the second maxil-
lipeds were removed and mounted. Megalopae were
identified to 1 of the 13 intermolt stages recognized
by Hatfield (1983); these were then combined to
form 3 groups: early (stages 1-4), middle (stages
5-8), and late (stages 9-13) megalopae. Temporal
durations of these three groups in laboratory studies
were 5.8, 15.4, and 6.3 days, respectively (Hatfield
1983). Corresponding durations for C. oregonensis
are unknown. Stages were grouped into the three
categories described for simplification of analysis;
many of Hatfield's stages were of <48-h duration,
and this was considered too fine a resolution for our
purposes.

RESULTS

Water Temperature

Water temperature at the surface ranged from

529

FISHERY BULLETIN: VOL. 86, NO. 3

8.4° to 14.4°C, and from 7.6° to 8.8°C at 50 m, dur-
ing the period 17 April to 7 July (Table 1). Warmest
surface temperatures were in May and July, and at
50 m, were warmest in June.

Species Occurrence

Three species of crab larvae predominated in the
plankton collected. The porcellanid crab, Petrolisthes
cinctipes, predominated in waters within 8 km of
shore, while C. oregonensis and C. magister domi-
nated in more offshore waters. Other crab species
were present but at much lower abundance than
these three species. Megalopae of C. oregonensis
were of comparable abundance to C. magister at
most stations sampled.

Surface Abundance

Presence of crab megalopae at the surface was
dependent on time of day, with slight differences
in timing of maximal abundance occurring between
the two Cancer species (Fig. 3). At 9.3 km offshore
in June, C. oregonensis was several orders of mag-
nitude more abundant and also seemed to peak in
abundance a little earlier in the evening and later
in the morning than did C. magister. Relative abun-
dance of both species declined in the middle of the
night.

Data in Figure 3 was smoothed by eye (Fig. 4)
to allow megalopal abundance data from the
transect to be weighted by time of capture at
night. Multiplier values were determined, and these
were used to adjust actual abundance data by
sampling time for the May and June sampling
periods (Tables 2, 3). Only these data were weighted,
and since day length changes with time of year, ad-

justment of data collected more than two weeks on
either side of 14 and 15 June was considered
inappropriate.

In hindsight, some data (indicated by asterisks in
Tables 2, 3) was found to have been collected at a
time when megalopae were just beginning to reach,
or had just left, the surface. The calibration slope
(Fig. 4) is very steep at both these times and it was
impossible to estimate accurately a meaningful scal-
ing multiplier for these data. Our estimated mega-
lopal abundance at these stations should be viewed
with particular caution; a zero value may not in-
dicate that megalopae were absent, but only that
they were not at the surface.

Bongo tows during the day poorly sampled the
abundance of C. magister megalopae (Table 4)
shown to be present by nighttime neuston tows
(Tables 2, 3). This was possibly due to sampling the
total column below only a few square meters of sea
surface, because of integration of the results, or to
avoiding the net by the megalopae. It was unlikely
because of failure to sample deeply enough, since
on the continental shelf, tow depth approached bot-
tom depth. Megalopae (maximum per tow was 4)
were primarily collected by bongo gear at stations
where they were abundant in neuston samples at
night, but relative abundance estimates obtained
from bongo nets are considered only useful in a
general qualitative sense because of the low absolute
numbers caught.

Temporal Occurrence
(Time of Year)

No Cancer megalopae were found in samples col-
lected shortly after dusk on 27 February-4 March.
Cancer megalopae were first observed on the next

Table 1 .-

-Watei

' temperatures (°C) by depth (m) at the stations sampled off Tofino, B.C. on
17 April, 31 May, 14 June, and 6 and 7 July 1985.

Depth (m)

Distance

offshore

(km)

April May June July

20 50 0 20 50 0 20 50 0 20 50

111

130

148

185

8.8
8.8
8.7
8.8
8.6
8.7
8.7
8.5
8.5

8.5
8.4

8.8
8.7
8.1
8.8
8.6
8.7
8.6
8.5
8.4

8.5
8.6

7.9
8.2
8.3
8.2
8.2
8.2
8.0

8.0
8.4

12.2
13.4
13.4
13.1
13.8
13.3
11.8
11.9
11.5

11.0
11.4

9.2

9.5

8.6

9.6

9.3

9.8

10.0

10.3

10.5

10.0
10.0

7.6
7.9
7.9
8.2
8.2
8.2
8.1

8.7
8.4

12.2
12.2
12.3
12.8
12.7
12.9
12.9
12.8
12.7
12.7

11.4
9.0
10.4
12.3
9.4
10.5
12.0
12.2
12.2
12.4

8.0
8.0
8.3
8.1
8.7
8.5
8.5
8.6

13.5
14.0
13.3
13.3
14.4
14.4

11.0
9.0
10.0
10.7
13.3
13.3

7.6
7.7
7.7
8.6
8.6

530

JAMIESON and PHILLIPS: OCCURRENCE OF CANCER CRAB MEGALOPAE

gsijung-

jq6i|iM J I lAio - —

)L|6l|IM| |D3||nDU

in
O

g ID
<1> 00

mB|||Mj |DO!jnDu -

jgBjiiMj |iA|0 -

jasuns-

p— I — I — r
O
O

[III — I — I r

O
O

-r-[ — I — I — r

•q

()

-4->

C\J

"3.

f\j

bserved du
.C.

O pQ

2 -
5 o

9- c

° iG

0! O

8SI Jung —

mbi |iMt I IMO

-mbl|IM| |D3UnDU

(q6i|iM( I Dounou

|L|&||IM4 |IAI3

i8suns-

I I I
o

T — I — r

o
o
to

i §

o ^

( )

z
<

. 33

O <=>

s^ «^

»5

CVI

c«

-13

OS
D

(aoD^jns oas j.ojoI-N) 33NVaNnaV

531

Figure 4.— Smoothed abundance of Cancer
magister and C. oregonensis megalopae ob-
served during sampling through one night at
a station 9 km off the coast off Tofino, B.C.

1000

o
o

Z!
c/)

O
0)

'E
O

LlI
O

z
<

3
QD
<

I
t-

FISHERY BULLETIN: VOL. 86, NO. 3

10
JUNE 1985

in
O

100

10-

'E
O

<
Q

CD
<

51
^•|

Q
LU

I
f-

o
o

1800

2000

2200

2400

0200

0400

0600

TIME (POST)

Table 2.— Scaling values used to standardize the actual number of megalopae caught
(unsealed) to the estimated number which would have been caught at the times of peak
nocturnal abundance, 22:40 and 23:40 hours, for Cancer oregonensis and C. magister. respec-
tively, on 30 and 31 May 1985. * = Value uncertain because of time of sampling. Values
rounded to one decimal place.

magister

C. oregonensis

Distance
(km)

(A/

10 m-2)

(N 10 m-2)

Time

Multiplier

Unsealed

Scaled

Multiplier Unsealed

Scaled

18:12

2.3

19:02

36.7

0.2

7.2*

19:43

26.8

0.0

1.1*

20:26

20.0

21:09

11.0

0.4

4.7

21:52

3.9

14.4

56.4

5.2

2.2

11.5

22:35

1.6

29.9

47.0

1.0

0.1

23:37

1.0

17.0

—

00:47

3.3

36.0

119.9

24.8

1.0

25.7

01:55

111

5.0

2.6

13.3

40.0

1.8

72.1

03:56

148

110.0

1.1

118.4*

2.5

0.1

0.2

05:57

185

532

JAMIESON and PHILLIPS: OCCURRENCE OF CANCER CRAB MEGALOPAE

sampling date, 16 and 17 April, but maximum abun-
dance (unweighted as to diel time of sampling)
occurred in late May and the second week of June
(Fig. 5). Cancer magister megalopae were collected
as late as 24 August, the last sampling date, but

relative abundance along the transect was low after
the beginning of July.

Cancer oregonensis megalopae were caught in
April but occurred in abundance only in the late May
and mid- June surveys, with peak abundance in June.

1000

100 120 140 160 180 200 220 240 260

CALENDAR DAYS

Figure 5.— Neustonic abundance of megalopae observed along the transect line during April to August 1985 surveys.

Table 3.— Scaling values used to standardize the actual number of megalopae caught
(unsealed) to the estimated number which would have been caught at the times of peak
nocturnal abundance, 22:40 and 23:40 hours, for Cancer oregonensis and C. magister, respec-
tively, on 14 and 15 June 1985. * = Value uncertain because of time of sampling. Values
rounded to one decimal place.

magister

C. oregonensis

Distance

10m-2)

(N 10 m"^)

Time

(km)

Multiplier

Unsealed

Sealed

Multiplier Unsealed

Sealed

01:42

37.1

8.8

327

00:58

3.6

0.1

0.4

27.4

170.5

4,671

00:09

1.2

0.1

16.3

0.1

2.3

23:19

1.0

0.2

2.7

7.6

20.5

05:36

0.1

03:52

55.0

52.2

2,871*

2.6

15.7

40.9

02:02

5.2

3.7

19.2

40.0

4.8

192

00:25

2.4

8.9

21.6

20.8

1.5

30.4

22:45

111

1.2

8.4

10.3

1.0

12.5

21:10

130

—

533

FISHERY BULLETIN: VOL. 86, NO. 3

Table 4.— Tow characteristics and number of Cancer magister megalopae (10 m^^ of sea surface) caught by 0.25 m

bongo tows off Tofino, B.C., on 31 May and 14 June 1985.

May

June

Distance

Water

Max.

Water

Max.

offshore

Time

volume

depth

No.

Megalopae

Time

volume

depth

No.

Megalopae

(km)

(h)

(m3)

(m)
33

megalops

(A/

10 m-'')

(h)

(m)

megalops

(A/

10 m"^)