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

Full text of "Fishery bulletin"

U.S. Department 
of Commerce 

Volume 104 
Number 1 
January 2006 




Fishery 
Bulletin 




U.S. Department 
of Commerce 

Carlos M. Gutierrez 

Secretary 



National Oceanic 
and Atmospheric 
Administration 

Vice Admiral 

Conrad C. Lautenbacher Jr., 

USN (ret.) 

Under Secretary for 
Oceans and Atmosphere 



National Marine 
Fisheries Service 

William T. Hogarth 

Assistant Administrator 
for Fisfieries 



^'i^^'°"=o^^ 



% 



^ATSS 0» *■ 



The Fishery Bulletin (ISSN 0090-0656) 
is published quarterly by the Scientific 
Publications Office, National Marine Fish- 
eries Service, NOAA, 7600 Sand Point Way 
NE, BIN 01,5700, Seattle, WA98115-007d. 
Periodicals postage is paid at Seattle, WA, 
and at additional mailing offices. POST- 
MASTER: Send address changes for sub- 
scriptions to Fishery Bulletin, Superin- 
tendent of Documents, Attn.: Chief, Mail 
List Branch, Mail Stop SSOM, Washing- 
ton, DC 20402-9373. 

Although the contents of this publica- 
tion have not been copyrighted and may 
be reprinted entirely, reference to source 
is appreciated. 

The Secretary of Commerce has deter- 
mined that the publication of this peri- 
odical is necessary according to law for 
the transaction of public business of this 
Department. Use of funds for printing of 
this periodical has been approved by the 
Director of the Office of Management and 
Budget. 

For sale by the Superintendent of 
Documents, U.S. Government Printing 
Office, Washington, DC 20402. Subscrip- 
tion price per year: $5.5.00 domestic and 
$68.75 foreign. Cost per single issue: 
$28.00 domestic and $35.00 foreign. See 
back for order form. 




Scientific Editor 

Adam Moles, Ph.D. 

Technical Editor 

Elizabeth Calvert 

11305 Glacier Highway 
Juneau, Alaska 99801-8626 



Managing Editor 

Sharyn Matriotti 

National Marine Fisheries Service 
Scientific Publications Office 
7600 Sand Point Way NE, BIN C15700 
Seattle, Washington 981 15-0070 



Editorial Committee 

Harlyn O. Halvorson, Ph.D. 
Ronald W. Hardy, Ph.D. 
Richard D. Methot, Ph.D. 
Theodore W. Pietsch, Ph.D. 
Joseph E. Powers, Ph.D. 
Harald Rosenthal, Ph.D. 
Fredric M. Serchuk, Ph.D. 
George Walters, Ph.D. 



University of Massachusetts, Boston 
University of Idaho, Hagerman 
National Marine Fishenes Service 
University of Washington, Seattle 
National Manne Fisheries Service 
Universitat Kiel, Germany 
National Marine Fisheries Service 
National Marine Fisheries Service 



Fishery Bulletin web site: www.fishbull.noaa.gov 



The Fishery Bulletin carries original research reports and technical notes on investigations in 
fishery science, engineering, and economics. It began as the Bulletin of the United States Fish 
Commission 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. 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 e.xchange 
for other scientific publications. 



U.S. Department 
of Commerce 

Seattle, Washington 

Volume 104 
Number 1 
January 2006 



Fishery 
Bulletin 



Contents 



The conclusions and opinions ex- 
pressed in Fishery Bulletin are 
solely those of the authors and do 
not represent the official position of 
the National Marine Fisher-ies Ser- 
vice (NOAA) or any other agency or 
institution. 

The National Marine Fisheries 
Service (NMFSl does not approve, 
recommend, or endorse any proprie- 
tary 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 approves, recom- 
mends, or endorses any proprietary 
product or proprietary material 
mentioned herein, or which has 
as its purpose an intent to cause 
directly or indirectly the advertised 
product to be used or purchased 
because of this NMFS publication. 



Articles 



MBLWHOI Library 
JAN 1 2006 

Mas: 



1 Rochet, Marie-Joelle, Jean-Francois Cadiou, and 
Verena M. Trenkel 

Precision and accuracy of fish length measurements obtained 
with two visual underwater methods 

10 Witteveen, Briana H., Robert J. Foy, and Kate M. Wynne 

The effect of predation (current and historical) 

by humpback whales (Megaptera novaeang/iae) on fish 

abundance near Kodiak Island, Alaska 



Companion papers 

21 Weinberg, Kenneth L., and David A Somerton 

Variation in trawl geometry due to unequal warp length 

35 Kotwicki, Stan, Kenneth L. Weinberg, and 
David A. Somerton 

The effect of autotrawl systems on the performance 
of a survey trawl 

46 Zeidberg, Louis D., William M. Hamner, Nikolay P. Nezlin, 
and Annette Henry 

The fishery for California market squid f.Loligo opa/escens) 
(Cephalopoda: Myopsida), from 1981 through 2003 

60 Criales, Maria M. John D. Wang, Joan A. Browder, 
Michael B. Robblee, Thomas L. Jackson, 
and Clinton Hittle 

Variability in supply and cross-shelf transport of pink shrimp 
(Farfantepenaeus duorarum) postlarvae into western Florida Bay 

75 Grandcourt, Edwin M., Thabit Z. Al Abdessalaam, 
Ahmed T. Al Shamsi, and Franklin Francis 

Biology and assessment of the painted sweetlips (Diagramma pictum 
(Thunberg, 1792)) and the spangled emperor (Lethrinus nebulosus 
(Forsskal, 1775)) in the southern Arabian Gulf 



Fishery Bulletin 104(1) 



89 Porch, Clay E., Anne-Marie Ekiund, and Gerald P. Scott 

A catch-free stoack assessment model with application to goliath grouper (Epinephelus itajara) 
off southern Florida 

102 Tuckey, Troy D., and Mark Dehaven 

Fish assemblages found in tidal-creek and seagrass habitats in 
the Suwannee River estuary 

118 Clarke, Lora M., Alistair D. M. Dove, and David O. Conover 

Prevalence, intensity, and effect of a nematode (Philometra saltatnx) 
in the ovaries of bluefish (Pomatomus saltatnx) 

125 Edwards, Elizabeth F. 

Duration of unassisted swimming activity for spotted dolphin (Stenella attenuata) calves; 
implications for mother-calf separation during tuna purse-seme sets 

136 Saillant, Eric, and John R. Gold 

Population structure and variance effective size of red snapper (Lutianus campechanus) 
in the northern Gulf of Mexico 



Note 

149 Friedland, Kevin D., Lora M. Clarke, Jean-Denis Dutil, and Matti Salminen 

The relationship between smolt and postsmolt growth for Atlantic salmon (Salmo salar) 
in the Gulf of St. Lawrence 

156 Erratum 

157 Guidelines for authors 



Abstract — During the VITAL cruise 
in the Bay of Biscay in summer 2002, 
two devices for measuring the length 
of swimming fish were tested: 1) a 
mechanical crown that emitted a pair 
of parallel laser beams and that was 
mounted on the main camera and 
2 1 an underwater auto-focus video 
camera. The precision and accuracy 
of these devices were compared and 
the various sources of measurement 
errors were estimated by repeatedly 
measuring fixed and mobile objects 
and live fish. It was found that fish 
mobility is the main source of error 
for these devices because they require 
that the objects to be measured are 
perpendicular to the field of vision. 
The best performance was obtained 
with the laser method where a video- 
replay of laser spots (projected on 
fish bodies) carrying real-time size 
information was used. The auto-focus 
system performed poorly because of a 
delay in obtaining focus and because 
of some technical problems. 



Precision and accuracy of fish length measurements 
obtained with two visual underwater methods 



Marie-Joelle Rochet 

Ldboratoire MAERHA, IFREMER 
Rue de nie d'Yeu, B.P 21105 
44311 Nantes, Cedex 03, France 
E-mail address mirochetifiifremerfr 

Jean-Francois Cadiou 

IFREMER (DNIS/SM/IM)-Centre Mediterranee 

B P 330 

83507 La Seyne, France 

Verena M. Trenkel 

Laboratoire MAERHA, IFREMER 
Ruede llle d Yeu, BP 21105 
44311 Nantes, Cedex 03, France 



Manuscript submitted 22 July 2003 
to the Scientific Editor's Office. 

Manuscript approved for publication 
20 April 2005 by the Scientific Editor. 

Fish. Bull. 104:1-9(20061. 



Visual sampling of marine systems by 
SCUBA divers and underwater vehi- 
cles is increasingly used to estimate 
animal abundances, to observe natu- 
ral behavior and response behavior 
to fishing gear in situ, and to assess 
community interactions (e.g., Bublitz, 
1996; Auster et al., 1997; Davis et al., 
1997; Uiblein et al., 2002; Trenkel et 
al., 2004). Visual methods also allow 
estimates of population-size struc- 
tures without the bias caused by the 
size selectivity of fishing gear. Visual 
techniques have been used in the wild 
for measuring the length of animals by 
SCUBA divers (e.g., Yoshihara, 1997; 
Pfister and Goulet, 1999; Harvey et 
al., 2002a) or by submersibles (Love 
et al., 2000; Yoklavich et al., 2000). 
They have also been employed for esti- 
mating the length frequency of the 
catch of live tuna to be fattened after 
capture (Harvey et al., 2003), and in 
aquaculture to estimate the size range 
offish (Petrell et al., 1997). Until now 
these techniques were mainly used 
in shallow waters or tanks. Because 
of the optical characteristics of sea 
water — its turbidity, the variations in 
light intensity with depth and water 
movements and fish movements, these 
methods are subject to measurement 
errors. Estimating the order of mag- 
nitude of this measurement error has 
been the focus of many studies (van 



Rooij and Videler, 1996; Yoshihara, 
1997; Harvey et al., 2001, 2002a, 
2002b, 2003). 

Efficient methods for measuring 
fish length in situ can also be used 
in deeper waters not accessible to 
divers. Parallel laser projected from 
a video camera onto the seafioor or 
fish bodies permit accurate measure- 
ments (Love et al., 2000; Yoklavich et 
al., 2000). Albert et al. (2003) mea- 
sured fish lengths on a video screen 
and then transformed these mea- 
surements into real length knowing 
the distance of the camera from the 
ground, its tilt angle, and the hori- 
zontal opening angle of the camera. If 
fish are not on or close to the bottom, 
it is necessary to know their distance 
off the bottom to apply this method. 
Auster et al. (1997) and Norcross and 
Mueter (1999) measured fish size on a 
video screen when the fish appeared 
between the skids of their ROV. The 
screen measurement is then related 
to the known distance of the skids. 
This method relies on the fish and 
skids being in the same horizontal 
plane and on the fish being perpen- 
dicular to the axis of the camera. 
Krieger (1992) used a submersible to 
estimate the size of rockfish. 

Two methods were tested during 
the VITAL cruise in the Bay of Bis- 
cay, in late August and early Septem- 



Fishery Bulletin 104(1) 




232 mm 



232 mm 



Figure 1 

The four laser pointers mounted on a crown around 
the main camera in front of the ROV Victor 6000. 
The inner circle is the camera lens. 




Figure 2 

Laser spots (indicated by arrows) visible on and under a fish. 
These laser spots documented on videotape provided size infor- 
mation both in real time and during video replay. 



ber 2002.^ Victor 6000, a remotely operated vehicle 
(ROV) equipped with several video cameras and re- 
corders, was operated at depths ranging from 1100 to 
1500 m. Fish size was measured both by using a pair 
of parallel laser beams, and an auto-focus video camera 
linked to software for estimating object size based on 
the focal distance of the object in focus. In this study 
three sources of measurement variability were investi- 
gated: 1) systematic errors inherent to each method; 2) 
variability due to observer differences; 3) variability due 
to continuous fish movements and horizontal body orien- 
tation. To estimate these error components separately, 
rigid and articulated artificial objects ("artificial fish") 
of known size were measured repeatedly by several 
independent observers. Individuals belonging to several 
deep-sea fish species were also repeatedly measured. 
Mixed-effects models and heteroscedastic error models 
were fitted to the resulting measurements to compare 
the magnitude of errors due to different sources. 



Materials and methods 

Measurement devices 

Both measurement devices were installed close to each 
other on the ROV Victor. The laser-beam crown was 
mounted on the main camera, which was itself attached 



1 Trenkel, V. M., N. Bailly. O. Berthele, O. Brosseau, R. Causse, 
F. de Corbiere, O. Dugornay, A. Ferrant, J. D. M. ordon, D. 
Latrouite, D. Le Piver, B. Kergoat, P. Lorance, S. Mahevas, 
B. Mesnil, J.-C. Poulard, M.-J. Rochet, D. Tracey, J.-P. Vach- 
erot, G, Veron, and H. Zibrowius. 2002. First results of a 
quantitative study of deep-sea fish on the continental slope of 
the Bay of Biscav: visual observations and trawling. ICES 
CM 20b2/L:18, 2002, 15 p. 



to the pan and tilt unit. The METRAU© (SONY, model 
FCB-IX 47P) autofocus camera was mounted on the 
same pan and tilt unit. 

Laser-beam pointers Four red laser pointers (10 mW, 
635 nanometers [nm]) were mounted around the main 
camera housing (Fig. 1). The distance between each 
two opposite lasers was 232 mm. Red light is strongly 
attenuated by water but because of the relatively high 
power of the laser light-emitting diode (LED), a range 
of up to 7 m is reachable in clear waters. 

To measure the fish and objects, the laser beams were 
projected on the target (Fig. 2). The laser spots, visible 
on the video, give size information both in real time and 
during video replay. The principle is simple, but several 
limitations exist. First, the measurement is correct only 
for an object located in a plane perpendicular to the 
laser axis. Second, the target should be large enough to 
be reached by at least two laser beams; the more laser 
impacts that are seen on the object to be measured, the 
easier the measurement. 

For the measurement to be accurate, there must be 
a strict parallelism between the laser beams. This is 
complicated by the fact that the laser component itself 
(the diode with its optic lens) does not necessarily have 
a beam parallel to the axis of the component package. 
Further, designing an accurate alignment mechanism 
that is compatible with offshore and deep underwater 
operating conditions is difficult. The residual error af- 
ter alignment is about 0.15°, which entails an error of 
10 mm for the distance between two opposite spots at 
a distance of 4 meters (i.e., 4% of size). 

METRAU camera The METRAU system is based on 
the autofocus video camera. The imaging device is an 
original equipment manufacturer (OEM) camera module 
similar to those used in off-the-shelf camcorders. The 



Rochet et al : Precision and accuracy of fish length measurements obtained with two visual underwater methods 



" .'*^ 



ms' 



..Jg^l 3:02:26 



-*•- 



■. y 




^ 



Figure 3 

A METRAU video camera image with overlaid grid. 



camera has a built-in automatic focus unit which adjusts 
lens settings to provide a sharp image. The camera is 
remotely controlled by a RS323 digital link and sends 
data back over this link, including zoom position and 
focal value (see details in Cadiou et al., 2004). A previ- 
ous calibration in air and in a test tank provided cor- 
relation rules between raw data and field angle or focal 
distance. 

When the system is operated, the data received by the 
computer are processed in real time. The object distance 
and the field angle are computed and a scale is overlaid 
on the video image (Fig. 3). 

According to optical laws, the depth of focus decreases 
as the focal length increases. This means that in order 
to obtain an accurate measure of focal distance, a nar- 
row field angle is required. In addition, the depth of 
focus increases when the focal distance moves towards 
infinity. Consequently, a domain of validity of the mea- 
surement can be defined. With the METRAU camera, 
there must be a target distance under three meters and 
a field angle of less than 6°. These constraints have to 
be combined with the following conditions: a steady 
image that would allow the automatic focal servo to 
stabilize; and avoidance of scenes with several image 
planes. In turbid waters, particles can create disturbing 
focal planes and affect the measurement process. 

Measurement experiments 

Artificial objects and live fish Objects of known size 
were used to estimate the potential bias in the length 
measurements obtained with the two devices. Three 
rigid objects — a can, a bottle, and a plastic tube mea- 
suring respectively 13, 30, and 66 cm — were repeatedly 
measured to evaluate device performance and observer- 
induced variability in the absence of errors induced by 
fish movement and variations in horizontal observation 
angles. 



Fish movement makes the horizontal observation 
angle vary continuously. As a result, it is difficult to 
judge if and when an individual fish is perpendicular 
to the measurement axis. Further, fish seldom lie in 
a straight plane. Some species continuously flex their 
tail, others bend their whole body. To mimic the mobil- 
ity of a real fish, a mobile object was built consisting 
of several pieces of Ertalyte (Quadrant Engineering 
Plastic Products, Bridgeport. CT) plates linked together 
with rope rings. This artificial fish was designed to be 
neutrally buoyant so that it could be moved by water 
currents and undulate like a real swimming fish. The 
"artificial fish" had three distinctively colored parts. 
Thus depending on how many parts were measured, a 
small (13 cm), medium (17 cm), or large (41 cm) "artifi- 
cial fish" was the result. The real size of rigid objects 
and of the artificial fish was unknown to the observers 
throughout the measurement experiment. 

In addition to measuring each of the rigid objects 
and the artificial fish, 351 individuals belonging to 21 
deep-sea fish species were measured with both methods. 
The body sizes of these species ranged from 5 to 110 cm. 
Each individual fish was measured up to nine times. 
Altogether 2373 measurements were carried out. 

Real time and postoperation measurements While the 
ROV was in operation, four to five observers were able 
to watch the video images. Real time measurements 
were performed by estimating sizes directly from the 
screen, without using any measuring instrument. Each 
observer was asked to write down his or her length 
estimate without announcing it, so that independent 
measurements were obtained. All artificial objects were 
measured by both trained and novice observers; real fish 
were measured only by trained observers, namely scien- 
tists and ROV pilots. All objects and fish were measured 
at distances of 2 to 5 meters. 

Postoperation measurements were also performed 
on registered videos and digital images. For the laser 
method, the video tape was replayed. The tape was 
stopped when the image with an object or fish seemed 
to be in the best possible position. The fish or object was 
then measured with a ruler on the still video image. 
Postoperation measurements made with a ruler were 
also performed on digital snapshots taken from the 
videos in real time for both the laser and the METRAU 
method. A ruler was used rather than computer image 
analysis because it was easy and cost-efficient and it 
was felt appropriate for this trial appraisal of measure- 
ment methods. The bias introduced by this method was 
assumed to be negligible compared to observer-induced 
and fish-movement-induced errors. 

Operational constraints prevented a full factorial 
design where all observers could use all methods and 
measure all objects. 

Data analysis 

Variance components for observers and fish movements 

The measurement variability due to observer differences 



Fishery Bulletin 104(1) 



and fish movements was estimated from the measure- 
ments of the artificial objects, for which the true size 
was known. Because the measurement variance was 
expected to be larger for the mobile artificial fish than 
for the rigid objects, an extended linear mixed-effects 
model (Pinheiro and Bates, 2000) was used to account 
for this expected heteroscedasticity. The model included 
true length as a fixed effect and observer as a random 
effect. Fixed objects and mobile objects (artificial fish) 
were allowed separate variances. The resulting model 
was 



L,f= 1.1 + pL* + o, + f^ -t-f,,. 



(1) 



where L, ^ = length measured by observer / of an object 
of class J=|fixed, mobile! and of true size 

L*; 
o, ~ iV(0,a(o)); and 
e,j ~ N(0,a), whereas /", ~ N(0, a,(/;)). 

The model was fitted to the data from the eight observers 
who had measured all objects with the laser method. 

Accuracy and precision for artificial objects An extended 
linear model including heteroscedastic variance terms 
was used to compare the precision and accuracy of 
the two methods. The model included fixed effects for 
true length and the measurement method. The esti- 
mated fixed effects allow assessment of the potential 
measurement bias of each method. The measurement 
errors for fixed and mobile artificial objects were mod- 
eled separately for each method. This allowed us to 
compare the precision of the methods. Thus, the fitted 
model was 



measurements. The model included a fixed individual 
fish effect (each fish had a different, unknown size) 
and a random observer effect; and fish species were al- 
lowed heteroscedastic variances to account for species 
behavior differences (Lepidion versus Bathypterois). 
The stationary species, B. dubius, is easier to measure 
compared to the more lively L. eqiies. The model was 



^,,.,.1 = ^'n +0, +S, + £„,, 



(3) 



where Z>,, , ^ = the length measurement obtained by ob- 
server / for individual fish n belonging 
to species /; 
o, ~ MO, a(o)); and 
f,„, ~ MO, a), whereas s, ~ N(o, a, (s,)). 

Unfortunately, it was not possible to carry out a direct 
comparison between the precision of size estimates 
of fish and artificial objects because the latter were 
measured seven to 11 times, whereas the former were 
measured only two to seven times. Random subsamples 
could be carried out to obtain comparable sample size; 
unfortunately, subsamples from large samples would 
still have a larger variance than small samples. 

All models were fitted by using Splus 6.0 for Unix 
(MathSoft, Seattle, WA). For heteroscedastic models, be- 
cause of identifiability constraints, the fitting algorithm 
provided estimates of the ratio between the standard 
deviations of each class in relation to the standard 
deviation of a specified class instead of the full set of 
standard deviations. 



Results 



H + PL* + f,k + f,*. 



(2) 



where k = the measurement method and j the object 
class as before. As in model 1, f^,, ~ N(0, o). In contrast, 
f^i. ~ N(0, o^f^if^i.)) allowed for separate variances for each 
object-type and method pair. Only two trained observ- 
ers used both measurement methods for all objects. 
Because there was no significant difference between 
their measurements, and in order to reduce the number 
of parameters to be estimated, no observer effect was 
included in this model. 



Precision of fish measurements The precision of fish- 
length estimates was compared for two species, Bathyp- 
terois dubius and Lepidion eques. These species were 
selected for this analysis because they are abundant and 
relatively easy to measure, compared to other species 
that move faster or flex their body more often. Twenty- 
four individuals belonging to these two species were 
measured repeatedly by up to five observers using the 
real-time laser measurement method. 

Because true fish size was unknown, measurement 
accuracy could not be estimated. For estimating the pre- 
cision of fish measurements, an extended linear model 
with heteroscedastic errors was fitted to the fish length 



Precision and accuracy of measurements varied among 
objects and methods (Table 1). The best precision was 
obtained with the video-replays of laser measurements, 
whereas METRAU generally did not perform very well, 
especially on snapshots. The precision was generally 
much lower for mobile objects than for rigid objects. Mea- 
surement bias was generally low for the laser method, 
whereas the METRAU method systematically under- 
estimated the size of objects. A variety of fish species 
with various sizes were measured. CVs for individual 
fish measurements varied from 3% to 23% (Table 2). 
Species were grouped according to their motion behavior 
(l=sitting on bottom motionless, 2 = station holding or 
drifting, 3 = slow swimming, 4=fast swimming [Lorance 
and Trenkel-]). CVs were found to differ between groups, 
increasing with mobility (mean CV in group 1: 8.9%; 
group 2: 9.7%; group 3: 12.9%; group 4 was excluded 
because there was only one individual, P<10"''). 



- Lorance, P., and V. Trenkel. In preparation. Natural 
behaviour and reaction to an approaching ROV of large 
mid-slope species. IFREMER, Centre do Brest, B.P. 70, 
29280 Plouzane. France. 



Rochet et al.: Precision and accuracy of fish length measurements obtained with two visual underwater methods 









Table 1 






Summary of length measurements for artificial objects by five visual methods. Lengths (in cml are given along with their 
coefficient of variation i'i) and number of observations in parentheses, a.f = artificial fish. The laser method entailed viewing 
laser points (that permit accurate length data) projected on fish. The METRAU method is based on an autofocus video camera. 


Object 


True size 


Laser 


Laser + video 


Laser + snapshot 


METRAU 


METRAU + snapshot 


Can 


13 


14.31 (13'7f, 26) 


13.76 (3%, 5) 


13.55 (4%, 5) 


11.25(7%, 10) 


10.96 (7%, 5) 


Bottle 


30 


30.87(7%, 26) 


31.90(2%, 5) 


31.47 (2%, 5) 


26.90(18%, 10) 


32.05(41^5,5) 


Tube 


66 


65.62 (9%, 26) 


66.40(2'/,, 5) 


66.66(1%, 5) 


44.25 (14%, 8) 


40.75(25'?, 4) 


Short a.f 


13 


13.30 (13%. 211 


13.32 (6%, 3) 


22.10(23%, 2) 


11.50(6%, 2) 


7.55 (33%, 6) 


Medium a.f 


17 


16.49 (14%, 21) 


17.11 (5%, 3) 


24.86(25%, 2) 


11.50(18%, 2) 


9.25(32%, 6) 


Large a.f 


41 


40.80 (17%, 21) 


44.87 (14%, 3) 


63.23(25%, 2) 


30.00 (14%, 4) 


23.21(31%, 6) 











Table 2 














Summary offish length measurements, n = 


number of individual fis 


h measui 


•ed for each species, in 


= total number of 


measure- 


ments. m In = mean number 


of ob 


servatior 


s per indiv 


idual. Mean 


length = 


average individual fish length (cm) 


per species. 


Mean CV = average individual coefficient of 


variation per species. Group = behavioral group 


of the 


species (l=sitting 


-m bottom 


motionless, 2=station holding 


or drifting, 3 = 


slow swimming, 4=fast 


swimming). 










Species 




n 


in 


in In 




Mean length 




Mean CV 




Group 


Alepocephalus bairdi 




2 


4 


2.0 




48.4 




8.2% 




2 


Balhypterois 




87 


586 


6.7 




19.2 




8.9% 




1 


Breviraja caerulea 




3 


13 


4.3 




29.6 




4.8% 




2 


Caelorinchus labiatus 




15 


73 


4.9 




28.7 




10.2% 




2 


Cataetyx latyceps 




1 


5 


5.0 




24.1 




12.4% 




2 


C him a era monstrosa 




6 


23 


3.8 




91.4 




9.2% 




3 


Coelorhyncus labiatus 




2 


8 


4.0 




25.4 




14.9% 




2 


Coryphaenoides rupestris 




21 


77 


3.7 




45.4 




12.1% 




2 


Cottunculus thomsoni 




2 


17 


8.5 




29.1 




16.1% 




1 


Galeus melastomus 




1 


4 


4.0 




30.7 




2.8% 




4 


Hoplostethus atlanticus 




6 


27 


4.5 




32.6 




9.4% 




2 


Hydrolagus affinis 




1 


6 


6.0 




74.9 




23.3% 




3 


Hydrolagus mirabilis 




3 


6 


2,0 




78.4 




11.9% 




3 


Lepidion eques 




161 


814 


5.1 




26.6 




9.6% 




2 


Mora rnoro 




3 


8 


2.7 




60.0 




7.9% 




2 


Nezumia aequalis 




10 


22 


2.2 




32.8 




8.9% 




2 


Notacanthus 




2 


5 


2.5 




44.6 




10.6% 




2 


Phycis blenoides 




1 


4 


4.0 




44.8 




15.3% 




2 


Syphobranchus kaupii 




7 


23 


3.3 




27.2 




16.2%' 




3 


Trachyrincus in u rrayi 




4 


7 


1.8 




38.9 




3.5% 




2 


Ti-achyscorpia crintulata echinata 


14 


67 


4.8 




40.7 




8.1% 




1 



Measurement variance for observers and fish movements 

The standard deviation of the observer random effect, 
aio), amounted to approximately 20% of the residual 
standard deviation. The standard deviation of the 
random effect for mobile objects, On,obih'^fmobih^^ was 
40'7c higher than that for rigid objects, indicating that 
the measurement variability was lower for fixed than for 
mobile objects, as expected. The slope and intercept of 



the fixed effects did not significantly differ from 1 and 
0, respectively. Thus the laser method is unbiased and 
performs equally well for all sizes in the range tested 
(Table 3). 

Accuracy and precision of measurements for artificial objects 

Size measurements carried out using the laser beams 
in real time or on registered videos were found to be 



Fishery Bulletin 104(1) 



Table 3 

Estimated fixed-effect coefficients and standard devia- 
tions for model 1 for the measurements of rigid and mobile 
objects by eight independent observers using the laser 
method. SE = standard error; CL =confidence limit. 



Coefficient 


Estimate 


SE 


P-value 


ii 

ft 


1.53 
0.976 


0.97 
0.023 


0.12 
<0.0001 


Standard deviation 


Lower 
CL 


Estimated 


LIpper 
CL 



oio) 0.296 1.126 4.435 

a 4.369 5.535 7.014 

On,jL,,,^lo,„oH.iM.nMc'' 0-523 0.710 0.964 

^ Standard deviations were estimated in relation to the standard 
deviation for mobile objects. 



unbiased, whereas the METRAU-based measurements 
underestimated the true length of objects by as much 
as 7 cm for real-time measurements and 10 cm for time- 
delayed measurements (Table 4, Fig. 4). In addition, the 
variance of METRAU measurements was systemati- 
cally larger than the corresponding laser measurements 
(Table 5: ratios larger than 1). For rigid objects, laser- 
based video-replays had a lower variance than real time 
measurements. This lower variance for postoperational 
measurements was due to the allowance of videos to be 
replayed as many times as necessary in order to select 
the best image where an object was perpendicular to 
the optical axis. Use of a ruler also improves the mea- 
surement. The high estimation variance obtained for 
mobile objects measured with the laser method on digi- 
tal snapshots was partially due to one outlier (Fig. 4B). 
The object was measured at a relatively great distance 
in somewhat turbid water. The outlier was not removed 
because these kinds of errors are to be expected under 
common measurement conditions in the field. Generally 
the most precise results were obtained for video-replays 
with the laser beam method. 







Table 4 








Estimated coefficients for model 


2 for 


measurements | 


of rigid and 


mobile 


objects by five varia 


nts 


of the two 


methods. 












Coefficient 




Estimate 


SE 




P-value 


f'/oser 




0.106 


0.934 




0.91 


I^METRAU 




-7.198 


1.621 




<0.0001 


f-^hser+video 




1.466 


0.937 




0.15 


' laser+snapsliol 




1.348 


0.932 




0.19 


>'METRAU*snap 


-hot 


-9.516 


1.806 




<0.0001 


/' 




0.976 


0.023 




<0.0001 


a 




4.592 









Precision of fish measurements 

The variance of the random effect for B. dtibius was 
about 66% of the variance estimated for L. eques. For 
live fish, the standard deviation of the observer random 
effect was approximately 16% of the residual standard 
deviation (Table 6). This standard deviation is lower 
than that obtained for objects of known size because 
the residual variance was larger owing to the small 
number of repeated measurements obtained for each fish. 
It was not easy to repeatedly measure fish because of 
escapement behavior. In addition, only trained observers 
took part in this experiment, which reduced observer 
variability. 



Discussion 



Table 5 

Estimates of the standard deviations of length mea- 
surements for rigid and mobile objects obtained by five 
variants of the two methods, from model 2. Number of 
measurements are given in parentheses. Estimates sig- 
nificantly different from 1 are in bold font. 



Method 


Rigid objects 


Mobile objects 


Laser 


1.12(21) 


1.00' (91 


Laser + video 


0.21(15) 


1.03(91 


Laser + snapshot 


0.19(151 


3.43(6) 


METRAU 


2.11(201 


1.06(81 


METRAU + snapshot 


3.16(14) 


1.58(181 



All standard deviations are relative to the standard deviation 
for laser measurements of mobile objects. 



The potential sources of errors and variability in visual 
fish length measurements are 1) the design and calibra- 



Table 6 

Estimates and 9b''i confidence limits for the standard 
deviations of the components of model 3 for fish size mea- 
surements obtained for two species by five independent 
observers using the laser method. 



Standard deviation 



Lower 



Esti- 
mate 



Upper 



aio) 
o 



'^B.dubius^^Bdub,us''l "L.eques'^L.eques^ 



0.068 
1.096 

0.41 



0.278 1.130 
1.702 2.642 
0.66 1.06 



Standard deviations provided in relation to the standard devia- 
tion (or Lepuiion measurements. 



Rochet et al Precision and accuracy of fish length measurements obtained with two visual underwater methods 



16 


—r- Can 25 
^ m 20 


D 2^°^'  




14 
12 
10 


1 = 2 15 


r^ 




^ ^ g 10 


^^ (=: 


S 
M+S 




L M L+V L+S M+S 


L M L+V L+S 


50 


— 30 

„ Bottle 

B 25 


E Medium  




40 
30 
20 


20 


 » ■.... 




B 


M+S 


= B " f - 




L M L+V L+S M+S 


L M L+V L+S 


70 
60 
50 
40 
30 


^ Tube 

C.H ^ s 70 


Large 

s 


s 

M+S 


 60 
~ — 50 

~ i i '° 

M H 30 

H 20 


= i 




L M L+V L+S M+S 


L M L+V L+S 


Figure 4 




Boxplots for the distributions of length measurements (cm) obtained by five 


measure- 


ment methods for each of three fixed (left) and three mobile (right) objects 


Boxes = 


interquartile range, white line = median, whiskers = extremes (excluding 


outliers). 


Dotted line is the true size of the object. L = laser method; M = METRAU 


method; 


L+V = laser video-replay; L + S = laser snapshot measurement; M + S = METRAU j 


snapshot measurement. 





tion of the measurement devices, 2) differences among 
observers, 3) orientation and position offish in relation 
to the camera, and 4) swimming motion. We investigated 
each of these featuress. 

Among the two methods tested during this study, 
the METRAU method performed poorly. It has several 
disadvantages. First, METRAU system needs the fish 
brought into focus when it is perpendicular to the field 
of vision. This may take time during which the fish can 
escape. Second, the registered video images do not in- 
clude the superimposed calculated scales. Thus, unlike 
the laser method, it is not possible to replay the video to 
identify the best image. Postoperational measurements 
can be performed only by using the digital snapshots 
registered in real time. Third, this method had a higher 
variance than the laser method, probably because of the 
technical constraints just mentioned. Fourth, during the 
VITAL cruise the estimates were systematically biased 
downwards. Measurements carried out after the cruise 
in a laboratory pool confirmed this systematic underes- 
timation to be -20% of the real sizes. This result may 
be due to errors in the software which processes the 
output of the camera. Although the errors could prob- 
ably be fixed, and the hardware improved to generate 
a video signal with the overlaid scale for recording, the 
other disadvantages are more difficult to eliminate, be- 



cause of the intrinsic limitations of the system. Hence 
the method is not promising for estimating the size of 
fish in the wild. 

By contrast, the laser beam method performed rath- 
er well, at least for rigid objects. We obtained CVs of 
7-13% for rigid objects in real time and 1-4% with 
image postprocessing (Table 1), both of which compare 
well with CVs for silhouette measurements obtained 
with a single camera placed in a laboratory pool (with 
scale bars placed on the bottom of the pool) and with 
computer image processing (1%, Harvey et al., 2002b), 
or even with stereo-video measurements of silhouettes 
(0.6-7.5%, Harvey and Shortis, 1996). Length mea- 
surements were always unbiased and postoperational 
measurements on video images reached a high precision 
for rigid objects and for small- to medium-size mobile 
objects. Thus the method seems suited for measuring 
the size of animals of low mobility, like invertebrates, 
along visual observation transects. 

The variance due to differences between observers 
was about 20% of the residual variance. This variance 
was reduced to 16% when measurements were per- 
formed by trained observers. This is true for real-time 
measurements. For video-replays of the laser-beam data, 
the variance due to observers was very small because of 
the use of a ruler instead of subjective extrapolation of 



Fishery Bulletin 104(1) 



the known laser distance. The use of automated analy- 
sis of video images may further reduce the observer 
error source. 

The major difficulty in measuring fish length /;; 
situ is caused by fish mobility, which causes them to 
be in variable orientations and positions in relation to 
the camera, and also to be flexed. We addressed both 
variance components together by comparing measure- 
ments of mobile objects ("artificial fish") and rigid 
objects. With the laser beam method, the measurement 
standard deviation of rigid objects was estimated to 
be 20% of the standard deviation of mobile objects 
(95% confidence limits: 6-75%). These components 
may be of the same order of magnitude as those for 
fish measurements, although fish measurements could 
not be estimated in our study because the true size 
of the fish was unknown. An attempt to disentangle 
both components is provided by estimating the differ- 
ence between the precision obtained for species with 
contrasting behaviors. Bathypterois dubius individu- 
als lie motionless on the bottom and seldom move, 
but because they stand on their fins they are never 
exactly perpendicular to the camera. By contrast, L. 
eques swims close to the bottom and tends to escape 
when the ROV is approaching too closely. This species 
continuously moves its tail; therefore it is very diffi- 
cult to obtain an image with the whole body properly 
orientated and straight. The standard deviation of B. 
dubius length measurements was estimated to be 66% 
of that of L. eques. This difference is smaller than the 
difference between rigid and mobile objects above; 
therefore we conclude that the major part of variance 
is due to the orientation of the fish in relation to the 
camera. Similarly, the estimated CVs of 21 species 
grouped by motion behavior differed only slightly. This 
is consistent with previous studies which have shown 
that relative errors of single-camera or stereo-video 
measurements of silhouettes or frozen fish could reach 
10% to 30%, depending on the distance to the camera, 
when the angle to the camera was increased from 0° to 
60°, whereas the measurement CVs increased fourfold 
(Harvey and Shortis, 1996; Petrell et al., 1997; Harvey 
et al., 2002b). By contrast, error due to tail flexion 
and muscle contractions during swimming motions 
was estimated at -5% in a comparison of "linear" to 
"sinusoidal" length of dorsally photographed sharks 
(Klimley and Brown, 1983) and at 0.5% for repeated 
stereo-video measurements of swimming tunas (Har- 
vey et al., 2003). 

In conclusion, the major source of measurement error 
for live fish may be their orientation and position in 
relation to the camera. For animals that are sessile or 
lying immobile on the ocean floor, this would be much 
reduced if the camera and laser beams were mounted 
vertically instead of obliquely. Thus the laser-beam 
method may be potentially useful for measuring ben- 
thic animals. For mobile animals, however, stereo-video 
methods (Harvey et al., 2001; Harvey et al., 2002a; van 
Rooij and Videler, 1996) may be more promising, and 
are continuously improving (Harvey et al., 2003). 



Acknowledgments 

We thank all observers for taking part in the experi- 
ment, and the ROV pilots for their willingness and skill 
at pursuing fish and in obtaining positions suitable for 
being measured. An anonymous referee gave very helpful 
comments on a previous version of this manuscript. 

Literature cited 

Albert, O. T., A. Harbitz, and A. S. Koines. 

2003. Greenland halibut observed by video in front 
of survey trawl: behaviour, escapement, and spatial 
patterns. J. Sea Res. 50:117-127. 
Auster, P.. R. Malatest, and C. Donaldson. 

1997. Distributional responses to small-scale habitat 
variability by early juvenile silver hake, Merluccius 
bilniearis. Environ. Biol. Fish. 56:195-200. 
Bublitz, C. 

1996. Quantitative evaluation of flatfish behaviour during 
capture by trawl gear. Fish. Res. 25:293-304. 

Cadiou, J.-F., V. Trenkel, and M.-J. Rochet 

2004 Comparison of several methods for in situ size 
measurements of moving animals. In Proceedings of 
the fourteenth (2004) international offshore and polar 
engineering conference; Toulon, France, May 23-2S, 
2004, p. 438-444. Int. Soc. Offshore Polar Engineers, 
Danvers, MA. 

Davis, C. L., L. Carl, and D. Evans. 

1997. Use of a remotely operated vehicle to study habitat 
and population density of juvenile lake trout. Trans. 
Am. Fish. Soc. 126:871-875. 

Harvey, E., M. Cappo, M. Shortis, S. Robson, J. Buchanan, and 
P. Speare. 

2003. The accuracy and precision of underwater measure- 
ments of lentgh and maximum body depth of southern 
bluefin tuna iTIiunnus maccoyii) with a stereo-video 
camera system. Fish. Res. 63:315-326. 
Harvey, E., D. Fletcher, and M. Shortis. 

2001. A comparison of the precision and accuracy of esti- 
mates of reef-fish lengths determined visually by divers 
with estimates produced by a stereo-video system. Fish. 
Bull. 99:63-71. 
2002a. Estimation of reef fish length by divers and by 
stereo-video. A first comparison of the accuracy and 
precision in the field on living fish under operational 
conditions. Fish. Res. 57:255-265. 
Harvey, E.. and M. Shortis. 

1996. A system for stereo-video measurement of sub-tidal 
organisms. Mar. Technol. Soc. J. 29:10-22. 
Harvey, E.. M. Shortis, M. Stadler, and M. Cappo. 

2002b. A comparison of the accuracy and precision of mea- 
surements from single and stereo-video systems. Mar. 
Technol. Soc. J. 36:38-49. 
Klimley, A. P., and S. T. Brown. 

1983. Stereophotography for the field biologist: mea- 
surement of lengths and three-dimensional positions of 
free-swimming sharks. Mar. Biol. 74:175-185. 
Krieger, K. J. 

1992. Distribution and abundance of rockfish determined 
from a submersible and by bottom trawling. Fish. Bull. 
91:87-96. 
Love, M. S., J. E. Caselle, and L. Snook. 

2000. Fish assemblages around seven oil platforms in the 
Santa Barbara Channel area. Fish. Bull. 98:96-117. 



Rochet et a\. Precision and accuracy of fish length measurements obtained with two visual underwater methods 



Norcross, B., and F.-J. Mueter. 

1999. The use of an ROV in the study of juvenile flat- 
fish. Fish. Res. 39:241-251. 
PetreU, R. J., X. Shi, R. K. Ward, A. Naiberg, and C. R. Savage. 
1997. Determining fish size and swimming speed in cages 
and tanks using simple video techniques. Aquacult. 
Engineer. 16:63-84. 
Pfister. R. D., and D. Goulet. 

1999. Nonintrusive video technique for in situ sizing of 
coral reef fishes. Copeia 1999:789-793. 

Pinheiro, J. C, and D. M. Bates. 

2000. Mixed-effects models in S and S-PLUS, 528 p. 
Springer, New York, NY. 

Trenkel, V. M., P. Lorance, and S. Mahevas. 

2004. Do visual transects provide true population den- 
sity estimates for deep-water fish? ICES J. Mar. Sci. 
61:1050-1056. 



Uiblein, F., P. Lorance, and D. Latrouite. 

2002. Variation in locomotion behaviour in northern cut- 
throat eel iSynaphobranchus kaupi) on the Bay of Biscay 
continental slope. Deep-Sea Res. 49l 11:1689-1703. 

van Rooij, J. M., and J. J. Videler. 

1996. A simple field method for stereo-photographic 
length measurement of free-swimming fish: merits and 
constraints. J. E.xp. Mar. Biol. Ecol. 195:237-249. 

Yoklavich, M. M., H. G. Greene, G. M. Cailliet, D. E. Sullivan, 
R. N. Lea, and M. S. Love. 

2000. Habitat associations of deep-water rockfishes in a 
submarine canyon: an example of a natural refuge. Fish. 
Bull. 98:625-641. 
Yoshihara, K. 

1997. A fish body length measuring method using an 
underwater video camera in combination with laser 
discharge equipment. Fish. Sci. 63:676-680. 



10 



Abstract — Humpback whales iMegap- 
tera novaeangliae) are significant 
marina consumers. To examine the 
potential effect of predation by hump- 
back whales, consumption (kg of prey 
daily) and prey removal (kg of prey 
annually) were modeled for a current 
and historic feeding aggregation of 
humpback whales off northeastern 
Kodiak Island, Alaska. A current prey 
biomass removal rate was modeled by 
using an estimate of the 2002 hump- 
back whale abundance. A historic 
rate of removal was modeled from a 
prewhaling abundance estimate (pop- 
ulation size prior to 1926). Two pro- 
visional humpback whale diets were 
simulated in order to model consump- 
tion rate. One diet was based on the 
stomach contents of whales that were 
commercially harvested from Port 
Hobron whaling station in Kodiak, 
Alaska, between 1926 and 1937. and 
the second diet, based on local prey 
availability as determined by fish 
surveys conducted within the study 
area, was used to model consumption 
rate by the historic population. The 
latter diet was also used to model 
consumption by the current popula- 
tion and to project a consumption 
rate if the current population were to 
grow to reach the historic population 
size. Models of these simulated diets 
showed that the current population 
likely removes nearly 8.83x10*' kg 
of prey during a 5-month humpback 
whale feeding season, which could 
include around 3.26 x lO*" kg of juve- 
nile pollock (Theragra chalcogramma). 
2.55 X 10'' kg of capelin iMalloti/s vil- 
losiis). if these species are consumed 
in proportion to their availability. The 
historic humpback whale population 
may have removed over 1.76 x 10'' kg 
of prey annually. 



The effect of predation (current and historical) 
by humpback whales (Megaptera novaeangliae) 
on fish abundance near Kodiak Island^ Alaska 



Briana H. Witteveen 

Robert J. Foy 

Kate M. Wynne 

School of Fisheries and Ocean Sciences 

University of Alaska Fairbanl<s 

118 Trident Way 

KQdial<, Alaska 99615 

E-mail address (for B H Witteveen) bwideveenia'sfosuafedu 



Manuscript submitted 22 April 2004 
to the Scientific Editor's Office. 

Manscript approved for publication 
25 April 2005 by the Scientific Editor. 

Fish. Bull. 104:10-20 (2006). 



Numerous studies have revealed that 
an increased awareness of trophic-level 
interactions is essential in assessing 
the status of complex marine ecosys- 
tems (Overholtz et al., 1991; Hair- 
ston and Hairston, 1993; Pascual et 
al., 1993; Estes, 1994; Kenney et al., 
1997; Trites et al., 1997). Such stud- 
ies have shown that predator-prey 
relationships in marine systems can 
have direct and indirect effects on all 
ecosystem members, but predictions of 
their effects cannot be made without 
multispecies models. 

Cetaceans are top predators in 
marine ecosystems and consume sig- 
nificant amounts of prey. Knowledge 
of the distribution, abundance, and 
foraging habits of cetaceans is, there- 
fore, an essential element of any pe- 
lagic ecosystem study (van Franeker, 
1992). Many species preyed upon by 
cetacean populations are targeted by 
other marine predators and commer- 
cial fisheries or are linked to fisheries 
through complex food webs. Previous 
studies have reported that prey re- 
moval due to cetacean consumption 
approaches or exceeds removals due to 
commercial fishing (Laws, 1977; Lae- 
vastu and Larkins, 1981; Bax, 1991, 
Markussen et al., 1992; Nordoy et 
al., 1995; Kenney et al., 1997). Such 
high levels of consumption can have 
significant effects on the distribution 
and abundance of prey species and 
the structure of marine communities 
(Perez and McAlister, 1993; Kenney 
et al., 1997; Croll et al., 1998). There- 
fore, examining consumption by ceta- 
ceans contributes information about 



complex ecosystem relationships and 
the long-term sustainability of ma- 
rine resources (Perez and McAlister, 
1993; Kenney et al., 1997; Tamura 
and OhsumiM. 

Humpback whales [Megaptera no- 
vaeangliae) feed in the waters off 
Kodiak Island and, because they are 
considered apex predators, may in- 
fluence the structure of the Kodiak 
Island marine ecosystem (Fig. 1) 
(Trites et. al., 1997; Croll et. al., 
1998). Modeling the amount of prey 
consumed (kg of prey annually) by 
feeding humpback whales is, there- 
fore, a useful tool for evaluating their 
role as marine predators. 

Cetaceans, in general, are described 
as opportunistic in their food selec- 
tion, although species tend to select 
broad categories of prey such as 
cephalopods, fish, or zooplankton (To- 
milin, 1954; Nemoto, 1959; Klumov, 
1966; Sigurjonsson and Vikingsson, 
1998). Humpback whales are clas- 
sified as generalists and target a 
wide variety of prey species (Nemoto, 
1970; Perry et al., 1999). They have 
been shown to be seasonal feeders on 
euphausiids [Thysanoessa spp.) and 
schooling fish species up to 30 cm in 
length, including capelin (Mallotus 
villosus). Pacific herring iClupea pal- 



1 Tamura, T., and S. Ohsumi. 2000. Re- 
gional assessments of prey consumption 
by marine cetaceans in the world. In- 
ternational Whaling Commission docu- 
ment SC/52/E6, 45 p. Website: www. 
icrwhale.org/eng/SC52E6.pdf [Accessed 
on 30 November 20021. 



Witteveen et al : Effect of prey removal by Megaptera novaeangliae on fisti abundance 



1-" 1 nil w 

I 



I 



A 





Muniiot Bity 
Cbiniak Bav 



2 

Woody Island 



Long Island 



iKilomelers 



Figure 1 

Map of Kodiak Island study area. Study area is shown as shaded area and subareas (1-4) are outlined and 
numbered. In detail is the nearshore subarea between Woody Island and Long Island. 



last), walleye pollock {Theragra chalcogramma). Atka 
mackerel iPleurogrammus monopterygius), cod (Gadus 
spp.). sardines {Sardinops spp. ), and sandlance (Ammo- 
dytes spp.) (Nemoto, 1957, 1959; Mitchell, 1973; Payne 
et al., 1990). The variety, as well as the amount, of 
prey removed from Kodiak waters may therefore be 
significant. Resource removal from Kodiak waters is of 
particular importance when considering the high value 
of Kodiak Island commercial fisheries, which totaled 
63.3 million dollars in exvessel (wholesale) value in 
2002 (NMFS2). 

Modeling consumption by humpback whales as they 
recover from severe population declines could shed light 
on patterns of change seen in prey and sympatric con- 
sumer populations, such as marine birds and pinnipeds 
(Merrick, 1997; Anderson and Piatt, 1999). Commercial 
whaling in the 1900s significantly reduced the number 
of humpback whales, both within coastal Kodiak waters 
and throughout the North Pacific (Rice, 1978). Following 



the protection of humpback whales in 1965, however, 
their numbers in the central North Pacific increased, 
possibly by as much as 10%, between the early 1980s 
and early 1990s for some North Pacific stocks (Baker 
and Herman, 1987; Calambokidis et al.'^). Removal and 
subsequent recovery of a marine predator of this magni- 
tude may cause large variations in the biomass removal 
of prey in the ecosystem, as has been hypothesized in 
other studies (Laws, 1985; Springer et al., 2003). How- 
ever, no empirical evidence exists to demonstrate such 
trophic interactions in the Gulf of Alaska. In this article, 



2 NMFS (National Marine Fisheries Service). 2002. Unpubl. 
data. Website: http://www.st.nmfs.gov/pls/webpls/MF_ 
LPORT.YEARD. RESULTS (Accessed on 31 May 2003.1 



3 Calambokidis, J., G. H. Steiger, J. M. Straley, T. Quinn, L. 
M. Herman, S. Cerchio, D. R. Salden, M. Yamaguchi, F. Sato, 
J. R. Urban, J. Jacobson, O. von Ziegesar, K. C. Balcomb, 
C. M. Gabriele, M. E. Dalheim, N. Higashi, S. Uchida, J. 
K. B. Ford, Y. Miyamura, P. Ladron de Guevara, S. A. Miz- 
roch, L. Schlender, and K. Rasmussen. 1997. Abundance 
and population structure of humpback whales in the North 
Pacific basin. Cascadia Research Cooperative Final Con- 
tract Report 50ABNF500113 to Southwest Fisheries Science 
Center, La Jolla, CA 92038, 72 p. Website: http://www. 
cascadiaresearch.org/reports/rep-NPAC.pdf. [Accessed on 
19 April 1999.] 



12 



Fishery Bulletin 104(1) 



we model the historic and current consumption rate by 
humpback whales within waters of northeastern Kodiak 
Island in order to assess the impact these whales have 
as predators on local prey populations. 



Materials and methods 

Study area 

The study area encompassed waters of northeastern 
Kodiak Island, including Chiniak and Marmot Bays 
(Fig. 1). The study area was divided into four subareas 
of approximately equal size in order to equalize sampling 
effort and maximize coverage of the study area. Subar- 
eas were also used to separate sightings of humpback 
whales for the purpose of weighting diet composition 
in relation to prey availability. An additional subarea, 
including the waters near Woody and Long Islands, was 
not considered a survey subarea but was designated in 
the poststudy period for calculating diet composition 
("nearshore," Fig. 1). 

Sightings and abundance of humpback whales 

Data on humpback whale sightings were collected during 
vessel surveys conducted between June and September 
in 2001 and 2002. Individual whales were identified from 
photographs of the black and white pigment patterns 
(and other natural markings) on the ventral surface 
of their tail flukes (Katona et al., 1979). A humpback 
whale sighting was defined as a sighting of an individual 
whale on a single day. Therefore, no whale was counted 
twice on one day, but may have been counted multiple 
times during the study period. Humpback whale sight- 
ings were summed by month and then by subarea for 
calculation of whale diet (see "Materials and methods" 
section: "Composition of simulated diets"). 

These sightings and fluke photographs were used 
in an associated study to estimate current humpback 
whale abundance within the study area (Witteveen, 
2003). The estimate determined from this associated 
study was used in conjunction with historic catch data 
from the Port Hobron whaling station to estimate his- 
toric humpback whale abundance. The whaling grounds 
of Port Hobron encompassed most of eastern Kodiak 
waters — an area approximately four times that of the 
study area. To account for the size difference between 
whaling grounds and the study area, catch values were 
divided by four under the assumption of a random har- 
vest throughout the grounds. The prewhaling and cur- 
rent estimates of humpback whale population size in 
the study area are 343 individuals (95% CI: 331, 376) 
and 157 individuals (95% CI: 114, 241), respectively 
(Witteveen, 2003). 



humpback whales. The diets were simulated because 
direct observation of humpback whale feeding behavior 
is rare and, even when observed, cannot produce a pre- 
cise account of the prey species being eaten. 

Diet A simulated historic target species and was 
based on the stomach contents of 39 humpback whales 
harvested at the Port Hobron whaling station from 
southeast Kodiak waters between 30 May and 9 August 
1937 as analyzed by Thompson (1940). 

Diet B simulated current target species and assumed 
no prey selectivity. It was based on the assumption 
that humpback whales will eat prey of a suitable size 
(<30 cm) in proportion to the relative occurrence of the 
prey in areas used by humpback whales. Euphausiid 
proportions in the diet were based on historic stomach 
contents and assumed to be constant over time (no cur- 
rent euphausiid abundance estimate is available). 

Information on seasonal prey availability was collect- 
ed from mid-water trawl surveys that were conducted 
within eastern Kodiak waters in July 2001 and from 
June through September 2002. Multiple passes with a 
commercial mid-water trawl net with a 22-mm mesh 
codend liner were made through acoustic scattering 
layers, ensuring an accurate representation of mid-wa- 
ter fish composition and occurrence. Species composi- 
tion, species counts, and fish size were determined for 
each tow and grouped within the study subareas. Only 
data from tows conducted during the study period in 
2001 and 2002 in areas utilized by humpback whales 
were included in our analysis. Therefore, prey surveys 
overlapped humpback whale sightings both temporally 
and spatially. A separate series of acoustic and purse- 
seine (center panel with a 3.2-mm mesh net) surveys 
was used to determine prey availability within the 
nearshore subarea from June through September 2002 
(Foy*). Prey composition determined by these surveys 
was assumed to be homogeneous throughout the near- 
shore habitat within the study area. 

To calculate diet B, the occurrence of fish smaller 
than 30 cm was determined from the mid-water trawl 
surveys within each subarea and month for both 2001 
and 2002. Tow data were first separated by subarea 
and month. Percent composition of prey species in each 
tow was calculated by dividing the total number of 
fish of each species caught by the total number of all 
fish caught in each tow, excluding species larger than 
30 cm (Nemoto, 1959) and species that were not previ- 
ously documented as prey, such as flatfish and other 
nonschooling fishes (Nemoto, 1957, 1959; Klumov, 1963; 
Krieger and Wing, 1984, 1986; Perry et al., 1999). 

To calculate diet B for the entire study area, prey 
proportions were weighted by the number of whales 
in each subarea. The weighted proportions were then 
summed across all months and subareas and multiplied 
by one minus the percentage of assumed euphausiid 



Composition of simulated diets 

Two diets were simulated: one that reflected the historic 
diet and the other that reflected the current diet for 



■• Foy, R. 2002. Unpubl. data. Fishery Industrial Technol- 
ogy Center, University of Alaska Fairbanks, Kodiak, AK 
99615. 



Witteveen et al,: Effect of prey removal by Megaptera novaeangliae on fish abundance 



13 



l>i :4'(r\\ 


i53"3'o"w is: 4:ii"\\ 
1 1 




i522r(rw 


1^2 


ri)-\\ 








' ' \ 


N 

A 




Afognak Island 




• 
• • 






', ^■--^x' 














•\ 


Mannot Bay 










+ + 






^u^_ 


 r'u'l 
+ 


y 
>^-'' 


• + ', 

* • • •• ' 


4- 

• 




-' - '" t •■ "^ / 








.•. - > 1 

Chiniak Ba\ 






Kodiak Island 




• • 










• 


1 ' 


5 10 20 
' 1 


30 40 
1 1 


1 





Figure 2 

A close-up of the study area showing locations of humpback whale (Megaptera novaeangliae) sightings 
and prey tows ( + ) for 2001 and 2002. Only mid-water trawl locations are shown. 



occurrence within the diet. Thus, diet B simulated a 
weighted availability of prey species based on temporal 
and spatial overlap between prey surveys and humpback 
whale sightings within the study period (Fig. 2). 

Consumption rate 

A seasonal consumption rate was estimated for both 
the current humpback whale population and the pre- 
whaling humpback whale population. The prewhaling 
consumption rate was estimated by using diet A only. 
Diet B was used to estimate the consumption rate by 
the current humpback whale population and to project 
the consumption rate by a humpback whale population 
at the prewhaling abundance level. 

The active metabolic rate (kcal/day) of feeding 
humpback whales was estimated in this study as 
£ = 192A/o '5 where Kleiber's (1961) model for basal 
metabolic rate (BMR; E=70M"'^) was modified by us- 
ing average oxygen consumption estimates for feeding 
baleen whales, where M is average body weight (kg) 
(Wahrenbrock et al., 1974: Sumich, 1983; Perez and 
McAlister, 1993). 

Daily prey consumption was then estimated as 



1 



K 1,000 



where / = total prey consumption (kg/day); 

E = estimated daily energy requirements (kcal/ 

day); and 
K = the estimated energy density (kcal/gram wet 

weight) of presumed prey. 

The average body mass for humpback whales (Mi was 
set equal to 30,408 kg (Trites and Pauly. 1998). The 
total energy density iK) of each diet was calculated by 
multiplying the average seasonal energy density of each 
prey species sampled in the study area by the percentage 
of that species within each diet and summing across all 
species. Values of A' for individual prey species came from 
proximate compositions that were determined from prey 
collected during 2002 trawl surveys for all months within 
the study period (Foy*). For each month, energy density 
was calculated by multiplying percent lipid by 9.4 kcal/g 
and percent protein by 4.3 kcal/g, which are conversion 
factors based on heat produced during metabolism of food 
(Schmidt-Nielson, 1997). Carbohydrates were considered 
to be bound and not available for nutrition (Gaskin, 



Fishery Bulletin 104(1) 



1982). The average seasonal energy density of each prey 
species was calculated by summing all values of energy 
dens.ty and dividing by the number of months in the 
study period. Previously published proximate composition 
values for surf smelt {Hypomesus pretious) and energy 
density data for euphausiids (Thysanoessa spp.) were 
used (Davis et al., 1997; Payne et al., 1999). 

Seasonal prey consumption for the population was es- 
timated by multiplying 1 by estimates of abundance (A^ ) 
and the total number of days in the humpback whale 
feeding season. Consumption estimates were calculated 
for both the upper and lower 95% confidence limits on 
the abundance estimates to show a possible range of 
consumption. The length of the feeding season was pre- 
sumed to be 152 days (Perez and McAllister, 1993). 



Results 







Table 1 






Number of sightings of humpback whales iMegaptera 
novaeangliae) for 2001 and 2002 by subarea and month in 
the Kodiak Island study area. 


Area 






2001 and 2002 




June 


July 


August 


September 


Total 


1 


10 


7 


10 


3 


30 


2 


29 


89 


22 


3 


143 


3 


20 


8 





12 


40 


4 





3 





7 


10 


Nearshore 


5 


14 








19 


Total 


64 


121 


32 


25 


242 



Analysis of sightings showed that humpback whales 
were not uniformly distributed within the study area 
(Table 1). Occurrence of humpback whales within sub- 
areas was variable, indicating within-season shifts of 
habitat use. Peak humpback whale sightings occurred 
in subarea 2 in July of both years. No humpback whales 
were sighted in the nearshore area after the month of 
July in either year. 

Only two prey items were identified in the 27 stom- 
achs that contained appreciable quantities of prey of 
39 stomachs analyzed by Thompson (1940). Surf smelt 
were found in 21 of 27 (78%) stomachs and euphausiids 
were found in 6 of 27 (22%) stomachs (Table 2). These 
percentages represent diet A. Energy densities of these 
two species combined to give a total energy density of 
1.31 kcal/gram (Table 3). 

The fish species in areas used by humpback whales 
in 2001 and 2002, as shown by mid-water trawl sur- 
veys, were pollock (36.96%), capelin (28.89%), eula- 



chon (7.60%), Pacific sandlance (4.44%), Pacific sandfish 
(0.08%), and Pacific herring (0.03%) (Table 2). These 
percentages represent diet B. Calculated energy densi- 
ties of prey species ranged from a high (eulachon) of 
2.52 kcal/gram to a low of 1.12 kcal/gram (juvenile 
pollock). The total energy density for diet B was 1.19 
kcal/gram (Table 3). 

Based on energetic content of the above diets, the 
model indicated that each humpback whale in the study 
area would consume 338 kg/day on diet A and 370 kg/ 
day on diet B. Using a prewhaling estimate of 343 (95% 
CI=331-376) animals in the study area, we determined 
that humpback whales feeding on diet A prior to 1927 
would have removed an estimated 1.76x10" kg of prey 
annually (95% CI= 1.70x10" to 1.93x10"), including 
nearly 3.87 xlO^ (3.74x10^ to 4.24x106) kg of euphausi- 
ids and approximately 1.37x10" (1.32x10" to 1.50x10") 
kg of surf smelt (Table 4). If diet B accurately reflects 
prey selection by the estimated 157 (95% 01 = 114-241) 









Table 2 








Composition 


and 


relative occurrence of prey species 


represented in simulated humpback whale (Megaptera novaeangliae) diet A 


(historic) and diet B (current). 










Diet 




Prey species 




Common name 


Percent of total diet 


A 




Hypomesus pretious 
Thysanoessa spp. 




surf smelt 
euphausiid spp. 


Total 


78.00% 
22.00% 
100% 


B 




Theragra chalcogramma 
Mallotus villosus 
Thysanoessa spp. 
Thaleichthys pacificus 
Ammodytes hexapterus 
Trichodon trichodon 
Clupea harengus pallasi 




walleye pollock 
capelin 

euphausiid spp. 
eulachon 
Pacific sandlance 
Pacific sandfish 
Pacific herring 


Total 


36.96% 

28.88% 

22.00% 

7.60% 

4.44% 

0.08% 

0.03% 

100% 



WItteveen et al Effect of prey removal by Megaptera novoeong/iae on fisfi abundance 













Table 3 






Monthly and average energy 


densities 


kcal/g 


"am 1 of prey 


species represented 


in simulated humpbat 


k whale {Megaptera novuc- 


angliae) diets A and B based 


on lipid and protein composition. Energy densities 


were used to estimate 


consumption by humpback 


whales. Average values 


in parentheses 


have been adjusted to reflect standard deviations of lipid and protein composition. N/A = | 


not available. 
















Species 










Energy densities (kcal/gram) 




June 




July 


August 


September 


Average 


Capelin 




1.1285 




1.2632 


1.1956 


1.4298 


1.2542(1.1665, 1.3755) 


Pacific sandlance 




1.4179 




1.4179 


1.4179 


1.4179 


1.4179(1.3211, 1.5590) 


Pacific sandfish 




0.8661 




1.2126 


1.1165 


1.1165 


1.0779(1.0449, 1.1300) 


Eulachon 




2.1582 




2.5218 


2.6758 


2.7424 


2.5245(2.3761,2.6860) 


Herring 




2.0999 




2.0999 


1.9454 


2.1205 


2.0664(1.9432,2.2942) 


Juvenile pollock 




1.0144 




1.0657 


1.1380 


1.2461 


1.1160(0.9730, 1.29941 


Euphausiids 




N/A 




N/A 


N/A 


N/A 


0.7430 


Surf smelt 




N/A 




N/A 


N/A 


N/A 


1.4698 







Table 4 










Daily and annual (over a 152-day feeding season) consumption of prey from two different diets off northeastern Kodiak Island 
by humpback whales ^Megaptera novaeangUae) at two levels of population abundance: the current population of 157 and the 
historic population of 343 (also presumed to be the carrying capacity to which the current population will recover). Diet A is the 
simulated diet of the historic population through analysis of stomach contents of 39 whales in 1937; Diet B is the simulated diet 
of the historic and current population based on currently available prey of suitable size for consumption. 


Prey species 


Daily prey 
removal (kg) 




Annual prey removal (kg) 






Mean 




Lower limit 




Upper limit 


Historic population 














Diet A 














Surf smelt 


90,301 


13,725,715 




13,245,515 




15,046,264 


Euphausiids 


25,469 


3.871,355 




3,735,914 




4,243,818 


Total 


115,770 


17.597,070 




16.981.429 




19,290,083 


Diet B 














Euphausiids 


27,934 


4,246,006 




4.097.458 




4,654,514 


Walleye pollock 


46,924 


7,132,406 




6.882.876 




7,818,614 


Capelin 


36,671 


5,573,943 




5,378.936 




6,110,211 


Eulachon 


9,652 


1,467,057 




1,415,731 




1,608,202 


Pacific sandlance 


5,635 


856,475 




826,511 




938,876 


Pacific sandfish 


98 


14,907 




14,386 




16,342 


Pacific herring 


33 


5,038 




4,862 




5,523 


Total 


126,974 


19,300,028 




18,624,808 




21,156,882 


Current population 














Diet B 














Euphausiids 


12,786 


1.943,507 




1,411,209 




2,983,345 


Walleye pollock 


21,478 


3,264,687 




2,370,537 




5,011,399 


Capelin 


16,785 


2,551,.338 




1,852,564 




3,916,385 


Eulachon 


4,418 


671,510 




487,593 




1.030.789 


Pacific sandlance 


2,579 


392,031 




284,659 




601,780 


Pacific sandfish 


45 


6,824 




4,9.55 




10,474 


Pacific herring 


15 


2,306 




1,675 




3,540 


Total 


58.119 


8,834,124 




6,414,587 




13,560,661 



16 



Fishery Bulletin 104(1) 



humpback whales currently feeding in the study ar- 
ea, these whales would be removing nearly 8.83x10^ 
(6.41x106 to 1.36x10") kg annually, including 3.26xl0« 
{2.37x10*5 to 5.01x10*5) kg of pollock, nearly 2.55 xlO*^ 
(1.85x10*5 to 3.92x10*5) kg of capelin, and 6.71x105 
(4.88x105 to 1.03x10*5) kg of eulachon. If the same diet 
were consumed by a population of humpback whales al- 
lowed to return to prewhaling abundance, the projected 
population would remove 1.9x10" (1.86x10" to 2.12x10") 
kg of prey annually, including approximately 7.13x10*5 
(6.88x10*5 to 7.82x10*5) kg of pollock, 5.57x10*5(5.38x10*5 
to 6.11x10*5) kg of capelin, and 4.25x10*5 (4.10x10*5 to 
4.65x10*5) kg of euphausiids (Table 4). 



Discussion 

Consumption rate 

Estimating the energy requirements of large cetaceans is 
inherently difficult and values presented in the present 
study may be subject to substantial uncertainty. Previ- 
ous studies in which consumption rates for cetaceans 
were estimated have used a range of values to adjust 
BMR (£=70M0 '5) for active metabolism. These values 
generally range from approximately 1.5 to 3 times BMR 
(Hinga, 1979; Lockyer, 1981; Sigurjonsson and Vikings- 
son, 1998). Our value of 192 is 2.7 times larger than 70 
and is, therefore, a reasonable estimate because it fits 
within this range and is based on the observed oxygen 
consumption rates of baleen whales. However, the con- 
sumption estimates are highly sensitive to perturbations 
of model input; a 5% error in this value would cause 
deviation of the same percentage (5%) in final consump- 
tion values. Further, all values in our consumption model 
are assumed to be constant when body mass, physiologi- 
cal status, and assimilation efficiency are likely subject 
to large seasonal fluctuations (Innes et al., 1987; Perez 
and McAlister, 1993, Kenney et al.. 1997; Trites et al., 
1997; Sigurjonsson and Vikingsson, 1998). Our model, 
however, did account for seasonal changes in the energy 
density of local prey sources; previous models, on the 
other hand, did not account for these changes (Perez 
and McAlister, 1993). Further research is necessary to 
obtain reliable field estimates of metabolic rates if model 
uncertainty is to be reduced. 

The historic prevalence of surf smelt in diet A could 
imply a dramatic change in surf smelt availability, 
misidentification, or an overestimation of smelt found 
in stomachs. Thompsons (1940) analysis resulted from 
"samples of stomach contents" obtained from catcher 
vessels; therefore, these samples may have completely 
missed less prevalent species. Further, stomach samples 
may have only reflected the most recent meal of the 
whale and therefore be biased toward a single species. 
This potential bias, however, could have been minimized 
by sampling stomachs throughout the season (May 30- 
August 09) (Thompson, 1940). Diet B was dominated by 
walleye pollock, a species not present in historic diet A. 
The increased importance of juvenile pollock in contem- 



porary humpback whale diet B could reflect changes in 
prey species availability and use, foraging selectivity, 
or reflect our diet reconstruction method. 

Diet B is considered provisional for two reasons. First, 
it is assumed that humpback whales eat prey species 
in proportion to their availability within foraging ar- 
eas. Humpback whales select preferred prey species 
and consumption, therefore, may be disproportional 
to availability. That is, they may be selectively forag- 
ing from all available prey sources. Previous foraging 
studies have described humpback whale distribution 
as being correlated with areas of capelin (Whitehead 
and Carscadden 1985; Piatt et al. 1989) and sandlance 
abundance (Payne et al. 1986; Kenney et al. 1996) and 
this correlation may indicate a possible preference for 
small forage fish species. Given that in the decades 
since whaling, the Gulf of Alaska has shifted from a 
system dominated by forage fish to one dominated by 
pollock and other groundfish (Merrick 1997; Anderson 
and Piatt 1999; Benson and Trites 2002), a shift in 
prevalence from surf smelt in the historic diet to pol- 
lock in the current diet is not unexpected. Pollock have 
been shown to be a dominant prey source of humpback 
whales harvested in Russia (Klumov, 1963). Addition- 
ally, humpback whales in southeastern Alaska have 
been observed near schools of juvenile pollock and are 
believed to eat pollock to an unknown, but potentially 
large, extent in some years (Gabriele^). 

The second source of uncertainty in diet B stems 
from the assumption that our mid-water trawl surveys 
provide unbiased samples of all available prey. Because 
these surveys were not designed to sample zooplankton, 
they may have produced a biased estimate of euphausiid 
availability. This bias may not be significant, however, 
because the 22% value we used in diet B was based on 
historic usage and falls within the range of euphausiid 
consumption (5-30% of the total diet) estimated in 
other humpback whale studies (Perez and McAlister, 
1993; Kenney et al., 1997). 

Further, diet B was constructed from the results of 
mid-water trawl surveys that may underestimate the 
availability of some forage fishes, particularly Pacific 
sandlance. Pacific sandlance are often small enough 
to swim through the meshes in the net or are found in 
benthic habitats and cannot be captured by mid-water 
trawl methods. To minimize this potential sampling 
bias, we supplemented our trawl surveys with purse 
seine sampling in the nearshore subarea. Despite this 
effort we may have underestimated the prevalence of 
Pacific sandlance in the area because it was found to 
dominate the diets of other coastal piscivores; stomach 
contents of 34 coho salmon (Oncorhynchus kisutch) and 
Pacific halibut iHippoglossus stenolepis) in 2002 (Wit- 
teveen^) and regurgitants from blacklegged kittiwakes 



= Gabriele, C. 2001-2002. Personal commun. Glacier Bay 
National Park. P.O. Box 140. Gustavus, AK 99826-0140. 

'5 Witteveen, B. H. 2002. Unpubl. data. Fishery Industrial 
Technology Center, University of Alaska Fairbanks, Kodiak, 
AK 99615. 



Witteveen et al : Effect of prey removal by Megaplera novaeonglioe on fisfi abundance 



17 



in 2001 (n=96) and 2002 (72=147) were dominated by 
Pacific sandlance (Murra et al., 2003). 

Ecological effects from humpback whale prey 
consumption 

Although estimates of consumption are highly dependent 
on estimates of population abundance and metabolic 
rates, these values indicate that humpback whales were, 
and still are. significant predators within the Kodiak 
Island ecosystem. 

Historic commercial whaling reduced the population 
in our study area to an estimated low of 27 animals 
by 1938 (Witteveen, 2003). The removal of so many 
large consumers likely had significant impacts on the 
surrounding ecosystem. As modeled, reducing historic 
consumption to that of current levels would release 
nearly 10,000 tons of prey within the study area in a 
single feeding season. Such a release could have caused 
a trophic cascade effect. 

Cetacean removals in the Southern Ocean have dem- 
onstrated how trophic cascades can affect marine eco- 
systems through removal of large marine predators, 
including whales (Laws, 1985). It has been hypothesized 
that a similar reorganization of the marine community 
may have occurred in the Bering Sea and Gulf of Alas- 
ka, although the mechanisms of such a cascade are not 
well understood (Merrick, 1997; Trites, 1997; Springer 
et al. 2003). Removal of whales during commercial 
harvest reduced predation on certain fish, cephalopod, 
and zooplankton species, which were then available to 
other consumers. This large number of unconsumed 
prey, when combined with environmental factors such 
as the 1977 regime shift, may have contributed to the 
growth of sympatric marine predator populations from 
the late 1940s to late 1970s. It is hypothesized that 
whale stock resurgence, coupled with the 1977 regime 
shift that favored the proliferation of groundfish species, 
may have reduced prey availability to other piscivores 
in the system and may have led to declines seen in har- 
bor seal (Phoca vitulina), Steller sea lion, northern fur 
seal iCallorhinits iirsinus), common murre (Ur-ia aalge), 
thick-billed murre iU. lomvia). and red-legged kittiwake 
{Rissa brevirostris) populations (Merrick, 1995, 1997; 
NRC, 1996; Trites, 1997). The Gulf of Alaska and Ber- 
ing Sea ecosystems may still be affected by changes 
caused by baleen whale removals and their recovery 
(NRC, 1996). 

Assuming that the Kodiak Island study area was 
similarly affected by this trophic reorganization, an es- 
timate of the current consumption by humpback whales 
would help elucidate the role that a humpback whale 
recovery is playing in ecosystem dynamics. If our diet 
composition and subsequent consumption estimates are 
accurate, our results indicate that the diet of hump- 
back whales in Kodiak waters directly overlaps those 
of sympatric piscivores and the biomass that is removed 
may be substantial. The top species modeled in the 
humpback whale diet represent important sources of 
energy for multiple higher-trophic-level species and are 



known to be significant dietary species for Steller sea 
lions (Wynne^), harbor seals (Jemison**). tufted puffins 
(Fratercula cirrhata) (Piatt et al., 1997), blacklegged 
kittiwakes (Murra et al., 2003), adult pollock. Pacific 
halibut, and arrowtooth flounder (Livingston, 1993; 
Yang, 1995; Merrick, 1997; Best and St. Pierre^). 

Our model indicates that humpback whales within 
the study area may currently be consuming a signifi- 
cant amount of fish, including over 3.26 x 10'^ kg of juve- 
nile pollock and nearly 3.62 x 10'' kg of small forage fish, 
such as capelin, eulachon and Pacific sandlance, during 
a 152-day feeding season. In comparison, tufted puffins 
consume less juvenile pollock (6.40x10'' kg) between 
mid-July and mid-September, but this amount still 
accounts for one-tenth of the age-0 pollock stock in the 
Gulf of Alaska during early July (Hatch and Sanger, 
1992). In addition, gadid removal by Steller sea lions 
in 1998 was estimated to be 1.79 xlO» kg, or 12% of 
the total gadid biomass that is removed by commercial 
fisheries for that year (Winship and Trites, 2003). This 
amount, although nearly 55 times the amount of pol- 
lock removal due to consumption by humpback whales, 
includes all gadid (not only pollock) species removals in 
all Alaskan waters. More importantly, these fish are 
likely larger (>60 cm vs. ^30 cm) than fish targeted by 
humpback whales. 

Although humpback whales generally feed on smaller 
age classes than are targeted by commercial fisheries or 
Steller sea lions (Perez and McAlister'; Kenney et al., 
1997), consumption of younger age classes may affect 
future recruitment into the fishery. Barrett et al. (1990) 
stated that consumption of young cod (Gadus morhua) 
and saithe (PoUachius virens) by shags (Phalacrocorax 
aristotelis) and cormorants (P. carbo) in the Northeast 
Atlantic could be a limiting factor in recruitment in 
years of low stock size, even if consumption of these 
species was overestimated by an order of magnitude. 
Thus, it is noteworthy that the removal by humpback 
whales of an estimated 3.26x10'' kg of pollock (age 0-2) 
equals 30% of the 2002 commercial pollock harvest of 
1.09x10'' kg (ages 3 to 8) for the entire Kodiak Island 
management area and 2.1% of the 2002 spawning bio- 
mass of pollock for the entire Gulf of Alaska, which was 
estimated at 1.58x108 kg (NMFSi"; NPFMCi'^^), 

These comparisons are based on mean estimates of 
prey removal and do not take into account model uncer- 



' Wynne, K. M. 2002. Unpubl. data. Fishery Industrial 
Technology Center, Univ. Alaska Fairbanks, Kodiak, AK 
99615. 

* Jemison, L. A. 2001. Summary of harbor seal diet data 
collected in Alaska from 1990-1999. In Harbor seal inves- 
tigations in Alaska iR. J. Small, ed.), p. 314-22. Ann. Rep. 
NOAA Grant NA 87Fx0300. Alaska Departmart of Fish 
and Game, P.O. Box 240020, Douglas, AK 99824. 

^ Best, E. A., and G. St. Pierre. 1986. Pacific halibut as 
predator and prey. International Pacific Halibut Commis- 
sion Technical Report 21, 27 p. Website: http://www.iphc. 
washington.edu/halcom/pubs/techrep/tech0021.pdf [Accessed 
on 31 May 2003.1 

10. n. 12 ggg next page for footnote text. 



18 



Fishery Bulletin 104(1) 



tainty. When uncertainty is considered, comparison to 
even the lower end of estimates of prey removal are still 
of note. For example, assuming that removal of juvenile 
pollock is equal to the lower estimate, or 2.37x10'' kg, 
the removal of pollock by humpback whales could still 
equal 21,7% of the 2002 commercial pollock catch and 
1.5% of 2002 spawning biomass. Thus, it follows that 
if true consumption is actually closer to the upper esti- 
mates, the impact of prey removal by humpback whales 
would likely increase. 

The humpback whale represents only one of a myriad 
of marine consumers within the Kodiak Island ecosys- 
tem whose ecological role cannot be determined without 
sophisticated multispecies models and an analysis of 
ecosystem interactions. This study was designed to 
provide essential baseline data and a model for estimat- 
ing prey removal by foraging humpback whales. Our 
results show that the potential for biomass removal due 
to consumption by humpback whales is significant and 
that the foraging strategies of these whales warrant 
further investigation. Continued research efforts can 
improve estimates of biomass removal by identifying 
target prey, determining the degree of prey selectivity, 
and assessing variable foraging efficiency. 



Acknowledgments 

The authors are grateful for the insight and guidance 
provided by Janice M. Straley, Terry J. Quinn II, Bren- 
dan P. Kelly, and Chris Gabriele. Field and labora- 
tory assistance was provided by Lisa Baraff and Katie 
Brenner. Financial support for this project was provided 
by the Rasmuson Fisheries Research Center and NOAA 
Grant no. NA16FX1270 to the University of Alaska Gulf 
Apex Predator-Prey Project. All research was conducted 
until provisions of NMFS Scientific Research Permit 
473-1433. 



Literature cited 

Anderson, P. J., and J. F. Piatt, 

1999. Community reorganization in the Gulf of Alaska 
following ocean climate regime shift. Mar. Ecol. Prog. 
Ser. 189:117-123. 



10 NMFS 2002. 2002. Gulf of Alaska groundfish quotas 
and preliminary catch in round metric tons. NMFS/AKR 
Fish Management, Juneau, AK 99802. 

" North Pacific Fisheries Management Council (NPFMC). 
2003. Stock assessment and fishery evaluation (SAFE) 
report for the groundfish resources of the Gulf of Alaska, 
846 p. Website: http://www.afsc.noaa.gov/refm/stocks/His- 
toric_Assess.htm (Accessed on 24 February 2004.] 

12 NPFMC. 2002. SAFE report for the groundfish resources 
of the Gulf of Alaska, Assessment of Walleye Pollock Stock 
in the Gulf of Alaska, 90 p. Website: http://www.afsc. 
noaa.gov/refm/stocks/Historic_Assess.htm [Accessed on 
24 February 2004.] 



Baker, C. S., and L. M. Herman. 

1987. Alternative population estimates of humpback 

whales (Megaptera novaeangliae) in Hawaiian waters. 

Can. J. Zool. 65:2818-2821. 
Barrett. R. T., N. Rov, J. Loen, and W.A. Montevecchi. 

1990. Diets of shags Phalacrocorax aristotelis and cor- 
morants P. carbo in Norway and possible implications 
in gadoid stock recruitment. Mar. Ecol. Prog. Ser. 
66:205-218. 

Bax, N. J. 

1991. A comparison of the fish biomass flow to fish, fish- 
eries, and mammals in six marine ecosystems. ICES 
Mar. Sci. Symp. 193:217-224. 

Benson, A. J., and A. W. Trites. 

2002. Ecological effects of regime shifts in the Bering 

Sea and eastern North Pacific Ocean. Fish Fish. 

3:95-113. 

Croll, D. A., B. R. Tershy, R. P. Hewitt, D. A. Demer, P. C. Fiedler, 

S. E. Smith, W. Armstrong. J. M. Popp, T. Keikhefer, V. R. 

Lopez, J. LTrban, and D. Gendron. 

1998. An integrated approach to the foraging ecology 
of marine birds and mammals. Deep-Sea Res. 45: 
1353-1371. 
Davis, N. D., K. W. Meyers, and Y. Ishida. 

1997. Caloric value of high-seas salmon prey organ- 
isms and simulated salmon ocean growth and prey 
consumption. North Pacific Anadromous Fish Com- 
mission iNPAFCi 1:146-162. 
Estes, J. A. 

1994. Top-level carnivores and ecosystem effects: ques- 
tions and approaches. In Linking species and ecosystems 
(C. G. Jones and J. H. Lawton, eds.) p. 151-158. Chap- 
man Hall, New York, NY. 
Gaskin, D. E. 

1982. The ecology of whales and dolphins, 459 p. Heine- 
mann, London. 
Hairston, N. G.. and N. G. Hairston. 

1993. Cause-effect relationships in energy flow, trophic 
structure, and interspecific interactions. Am. Nat. 
142:379-411. 
Hatch, S. A., and G. A. Sanger. 

1992. Puffins as samplers of juvenile pollock and other 
forage fish in the Gulf of Alaska. Mar. Ecol. Prog. 
Ser. 80:1-14. 

Hinga, K. R. 

1979. The food requirements of whales in the Southern 
Hemisphere. Deep-Sea Res. 26:569-577. 
Innes, S., D. M. Lavigne, W. M. Eagle, and K. M. Kovacs. 

1987. Feeding rates of seals and whales. J. Anim. Ecol. 
56:115-130. 
Katona, S., P. Baxter, O. Brazier, S. Kraus, J. Perkins, and H. 
Whitehead. 

1979. Identification of humpback whales by fluke 
photographs. In Vol. 3: behavior of marine animals 
(H. E. Winn and B. L. 011a, eds.), p. 33-44. Pages 
Plenum Press, New York, NY. 
Kenney, R. D., P. M. Payne, D. W. Heinmann, and H. E. Winn. 

1996. Shifts in northeast shelf cetacean distributions 
relative to trends in the Gulf of Maine/Georges Bank 
finfish abundance. In The northeast shelf ecosys- 
tem: assessment, sustainability, and management (K. 
Sherman, N. A. Jaworski, and T. J. Smayda, eds.). 
p. 169-197. Blackwell Science, Cambridge, MA. 

Kenney, R. D., G. P. Scott, T. J. Thompson, and H. E. Winn. 

1997. Estimates of prey consumption and trophic impacts 



Witteveen et al.: Effect of prey removal by Megaptera novaeangliae on fisfi abundance 



19 



of cetaceans in the USA Northeast continental shelf 
ecosystem. J. Northwest Atl. Fish. Sci. 22: 55-171. 
Kleiber, M. 

1961. The fire of life; an introduction to animal energet- 
ics, 454 p. John Wiley and Sons, New York, NY. 
Klumov, S. K. 

1963. Food and helminth fauna of whalebone whales in 
the main whaling regions of the world ocean. Trudy 
Instituta Okeanologii 71:94-194. 
1966. Plankton and the feeding of the whalebone whales 
(Mystacoceti). Trans. Inst. Oceanol. 51:142-156. 
Krieger, K. J., and B. L. Wing. 

1984. Hydroacoustic surveys and identification of hump- 
back whale forage in Glacier Bay, Stephens Passage, and 
Frederick Sound, Southeastern Alaska, Summer 1983. 
NOAA Tech. Memo. NMFS F/NWC-66, 60 p. Auke Bay 
Laboratory National Marine Fisheries Service P.O. Box 
210155 Auke Bay, Alaska 99821. 

1986. Hydroacoustic monitoring of prey to determine 
humpback whale movements. NOAA Tech. Memo. 
NMFS F/NWC-98, 62 p. Auke Bay Laboratory National 
Marine Fisheries Service P.O. Box 210155 Auke Bay, 
Alaska 99821. 
Laevastu, T., and H. A. Larkins. 

1981. Marine fisheries ecosystems, its quantitative evalu- 
ation in management, 167 p. Fishing News Books, 
Farnham, Surrey. LTK. 
Laws, R. M. 

1977. Seals and whales in the Southern Ocean. Phil. 
Trans. R. Soc. of London Series B Biol. Sci. 279:81-96. 

1985. The ecology of the Southern Ocean. Am. Sci. 
73:26-40. 

Livingston, P. A., A. Ward, G. M. Lang, and M-S. Yang. 

1993. Groundfish food habits and predation on com- 
mercially important prey species in the eastern Bering 
Sea from 1987 to 1989. NOAA Tech. Memo. NMFS 
F/NWC-54, 192 p. 
Lockyer, C. 

1981. Estimation of the energy costs of growth, main- 
tenance and reproduction in the female minke whale 
(Balaenoptera acutorostrata), from the Southern Hemis- 
phere. Rep. Int. Whal. Comm. 31:337-343. 
Markussen, N. H., M. Ryg, and C. Lydersen. 

1992. Food consumption of the NE Atlantic minke whale 
(Balaenoptera acutorostrata) population estimated with 
a simulation model. ICES J. Mar. Sci. 49:317-323. 
Merrick, R. L. 

1995. The relationship of the foraging ecology of Steller 
sea lions (Eumetopias jubatusi to their population decline 
in Alaska. Ph.D. diss., 172 p. Univ. Washington, 
Seattle, WA. 
1997. Current and historical roles of apex predators in 
the Bering Sea ecosystem. J. Northwest Atl. Fish. 
Sci. 22:343-355. 
Mitchell, E. 

1973. Draft report on humpback whales taken under spe- 
cial scientific permit by eastern Canadian land stations, 
1969-1971. Rep. Int. Whal. Comm. 23:138-154. 
Murra, K. A., C. L. Buck. S. D. Kildaw, J. B. Gamble, C. T. 
Williams. 

2003. Forage location, diet, and productivity of black- 
legged kittiwakes in 2001 and 2002 in Chiniak Bay, 
Kodiak Alaska (Abstract). In Marine science in the 
Northeast Pacific: science for resource dependent com- 
munities joint science symposium (book of abstracts); 
January 13-17, 2003, Anchorage, Alaska. 



NRC (National Research Council). 

1996. The Bering Sea ecosystem, 307 p. National Acad- 
emy Press, Washington D.C. 

Nemoto, T. 

1957. Foods of baleen whales in the northern Pacific. Sci. 

Rep. Whales Res. Inst. Tokyo 12:33-89. 
1959. Foods of baleen whales with reference to whale 
movements. Sci. Rep. Whales Res. Inst. Tokyo 14:244- 
290. 
1970. Feeding patterns of baleen whales in the ocean. In 
Marine food chains (J. H. Steele, ed.) p. 241-252. Oliver 
and Boyd, Edinburgh, Scotland. 
Nord0y, E. S., L. P. Falkov, P.-E. Matensson, and A. S. Blix. 

1995. Food requirements of northeast Atlantic minke 
whales. In Whales, seals, fish, and man (A. S. Blix, 
L. Wall0e, and 0. Ulltang, eds.), p. 307-317. Elsevier 
Science, Amsterdam, The Netherlands. 
Overholtz, W. J., S.A. Murawski and K.L. Foster. 

1991. Impacts of predatory fish, marine mammals, and 
seabirds on pelagic fish ecosystems of the northeastern 
USA. In Multispecies models relevant to management 
of living resources (N. Daan and M.P. Sissenwine, eds.) 
p. 198-208. ICES Mar. Sci. Symp. 193. International 
Council for the Exploration of the Sea (ICES), Copen- 
hagen V, Denmark. 
Pascual, M.. R. Hilborn, L. Fritz, H. Xi, and J. Moss. 

1993. Modeling the trophic relationships between fish and 
marine mammal populations in Alaskan waters. In Is it 
food? Addressing marine mammal and seabird declines, 
p. 30-44. Alaska Sea Grant Rep. 93-01. Alaska Sea 
Grant Program, Univ. Alaska, Fairbanks, AK. 
Payne, S. A., B. A. Johnson, and R. S. Otto. 

1999. Proximate composition of some northeastern Pacific 
forage fish species. Fish. Oceanogr. 8:159-177. 
Payne, P. M., J. R. Nicolas, L. O'Brien, and K. D. Powers. 

1986. The distibution of the humpback whales Megap- 
tera novaeangliae on Georges Bank and in the Gulf of 
Maine in relation to densities of the sand eel Amniodytes 
americanus. Fish. Bull. 84:271-277. 
Payne, P. M., D. N. Wiley, S. B. Young, S. Pittman, P. J. Clapham. 
and J. W. Jossi. 

1990. Recent fluctuations in the abundance of baleen 
whales in the southern Gulf of Maine in relation to 
changes in selected prey. Fish. Bull. 88:687-696. 
Perez, M. A., and W .B. McAlister. 

1993. Estimates of food consumption by marine mam- 
mals in the eastern Bering Sea. NOAA Tech. Memo. 
NMFS-AFSC-14, 36 p. National Marine Fisheries Ser- 
vice Alaska Fisheries Science Center 7600 Sand Point 
Way N.E., Building 4, Seattle, WA 98115. 
Perry, S. L., D. P. DeMaster, and G. K. Silber. 

1999. The great whales: History and status of six species 
listed as endangered under the Endangered Species Act 
of 1973. Mar. Fish. Rev. 61(l):l-74. 
Piatt, J. F., D. A. Methven, A. E. Burger, R. L. McLagan. 
V. Mercer, and E. Creelman. 

1989. Baleen whales and their prey in a coastal 
environment. Can. J. Zool. 67:1523-1530. 
Piatt, J. F., D. D. Roby, L. Henkel, and K. Neuman. 

1997. Habitat use, diet, and breeding biology of tufted 
puffins in Prince William Sound, Alaska. Northwest. 
Nat. 78:102-109. 

Rice, D. W. 

1978. The humpback whale in the North Pacific: distribu- 
tion, exploitation and numbers. Appendix 4 in Report 
on a workshop on problems related to humpback whales 



20 



Fishery Bulletin 104(1) 



(Megaptera novaeangliae) in Hawaii (K. S. Norris and 
R. R. Reeves, eds.), p. 29-44. U.S. Dep. Commer., Nat. 
Tech. Info. Serv. PB-280 794, Springfield, VA. 
Schmidt-Nielson, K. 

1997. Animal physiology: adaptation and environment, 
S"' ed., 607 p. Cambridge Press, Cambridge, UK. 

Sigurjonsson, J., and G. A. Vikingsson. 

1998. Seasonal abundance of and estimated food consump- 
tion by cetaceans in Icelandic and adjacent waters. J. 
Northwest Atl. Fish. Sci.. 22:271-287 

Springer, A. M., J. A. Estes, G. B. van Vilet, T. M. Williams, D. F. 
Doak, E. M. Banner, K. A. Forney, and B. Pfister. 

2003. Sequential megafaunal collapse in the North Pacific 
Ocean: An ongoing legacy of industrial whaling? Proc. 
Acad. Nat. Sci. 100:1222:3-12228. 
Sumich, J. L. 

1983. Swimming velocities, breathing patterns, and esti- 
mated costs of locomotion in migrating gray whales, 
Eschrichtius robustus. Can. J. Zool. 61:647-652. 
Thompson, R.J. 

1940. Analysis of stomach contents taken during the 
years 1937 and 1938 from the North Pacific. M.Sc. 
thesis, 82 p. Univ. Washington, Seattle. WA. 
Tomilin, A. G. 

1954. Adaptive types in the order Cetacea (the prob- 
lem of ecological classification of Cetacea). Zool. Zh. 
33:677-692. 
Trites, A. W. 

1997. The role of pinnipeds in the ecosystem. In Pin- 
niped populations, eastern north Pacific: status, trends, 
and issues. A symposium of the 127"^ annual meeting of 
the American Fisheries Society (G. Stong, J. Goebel, and 
S. Webster, eds.), p. 31-38. New England Aquarium, 
Conservation Department, Boston, MA. 

Trites, A. W. and D. Pauly 

1998. Estimating mean body masses of marine mam- 



mals from maximum body lengths. Can. J. Zool. 
78:886-896. 
Trites, A. W., V. Christensen, and D. Pauly. 

1997. Competition between fisheries and marine mam- 
mals for prey and primary production in the Pacific 
Ocean. J. Northwest Atl. Fish. Sci. 22:173-187. 
van Franeker, J. A. 

1992. Top predators as indicators for ecosystem events in 

the confluence zone and marginal ice zone of the Weddell 

and Scotia seas, Antarctica, November 1988 to January 

1989 (EPOS Leg 2). Polar Biol. 12:93-102. 

Wahrenbrock, E. A., G. F. Maruschak, R. Eisner, and D. W. 

Kenney. 

1974. Respiration and metabolism in two baleen whale 
calves. Mar. Fish. Rev. 36:1-9. 
Whitehead, H., and J. E. Carscadden. 

1985. Predicting inshore whale abundance — whales 
and capelin off the Newfoundland coast. Can. J. Fish. 
Aquat. Sci. 42:976-981. 
Winship, A. J., and A. W. Trites. 

2003. Prey consumption of Steller sea lions iEunietopias 
jubatus) off Alaska: how much prey do they require? 
Fish. Bull. 101:147-163. 
Witteveen, B. H. 

2003. Abundance and feeding ecology of humpback whales 
(Megaptera novaeangliae! in Kodiak. Alaska. M.Sc. 
thesis, 109 p. Univ. Alaska Fairbanks. Fairbanks, 
AK. 
Yang, M. S. 

1995. Food habits and diet overlap of arrowtooth flounder 
(Atheresthes stomias) and Pacific halibut iHippoglossus 
stenolepis) in the Gulf of Alaska. In Proceedings of the 
international symposium on North Pacific flatfish, p. 
205-223. Alaska Sea Grant Program Report 95-04, 
Univ. of Alaska, Fairbanks, AK. 



21 



Abstract — Survey standardization 
procedures can reduce the variabil- 
ity in trawl catch efficiency thus 
producing more precise estimates 
of biomass. One such procedure, 
towing with equal amounts of trawl 
warp on both sides of the net, was 
experimentally investigated for its 
importance in determining optimal 
trawl geometry and for evaluating the 
effectiveness of the recent National 
Oceanic and Atmospheric Adminis- 
tration (NOAA) national protocol on 
accurate measurement of trawl warps. 
This recent standard for measuring 
warp length requires that the differ- 
ence between warp lengths can be no 
more than 49f of the distance between 
the otter doors measured along the 
bridles and footrope. Trawl perfor- 
mance data from repetitive towing 
with warp differentials of 0. 3, 5, 7, 
9, 11, and 20 m were analyzed for 
their effect on three determinants 
of flatfish catch efficiency: footrope 
distance off-bottom, bridle length in 
contact with the bottom, and area 
swept by the net. Our results indi- 
cated that the distortion of the trawl 
caused by asymmetry in trawl warp 
length could have a negative influ- 
ence on flatfish catch efficiency. At a 
difference of 7 m in warp length, the 
NOAA 4'7f threshold value for the 83- 
112 Eastern survey trawl used in our 
study, we found no effect on the acous- 
tic-based measures of door spread, 
wing spread, and headrope height off- 
bottom. However, the sensitivity of the 
trawl to 7 m of warp offset could be 
seen as footrope distances off-bottom 
increased slightly (particularly in the 
center region of the net where flatfish 
escapement is highest), and as the 
width of the bridle path responsible 
for flatfish herding, together with the 
effective net width, was reduced. For 
this survey trawl, a NOAA threshold 
value of 4% should be considered a 
maximum. A more conservative value 
(less than 4%) would likely reduce 
potential bias in estimates of relative 
abundance caused by large differences 
in warp length approaching 7 m. 



Variation in trawl geometry due to 
unequal warp length 

Kenneth L. Weinberg 

David A. Somerton 

National Marine Fisheries Service 

Alaska Fisheries Science Center 

7600 Sand Point Way N E 

Seattle, Washington 98115 0070 

E-mail address (for K L, Weinberg): l<en-weinberg(gJnoaa gov 



Manuscript submitted 20 October 2004 
to the Scientific Editor's Office. 

Manuscript approved for publication 
6 June 2005 by the Scientific Editor. 

Fish. Bull. 104:21-34(20061. 



Standardization of trawl survey pro- 
cedures can reduce the variability in 
abundance indices between samples, 
survey vessels, and over time by reduc- 
ing the variability in trawl catch effi- 
ciency. Such standardization was the 
focus of the recently developed U.S. 
National Oceanic and Atmospheric 
Administration (NOAA) protocols for 
the operation of its groundfish bottom 
trawl surveys (Stauffer, 2004). The 
first of these protocols concerns the 
measurement of towing cables or 
warps. For vessels towing with two 
warps, the NOAA protocols specify 
that the difference in length between 
port and starboard warps may not 
exceed 4% of the wire length between 
otter doors measured along the bridles 
and footrope. The need for adopting 
such a critical value was considered 
essential because of the belief that 
unequal warp lengths — from inac- 
curate measurement or subsequent 
stretching — would lead to distortion of 
trawl geometry and a change in catch 
efficiency, particularly for operations 
that use trawl winches with the brakes 
set or locked. The adopted value, how- 
ever, was chosen somewhat arbitrarily 
because experimental data showing the 
dependency of trawl geometry or fish- 
ing performance on warp symmetry 
was lacking for any of the bottom trawl 
surveys subject to the protocols. 

In this study, we examine the ef- 
fect of unequal warp lengths on the 
geometry of the 83-112 Eastern trawl 
which is used by the Alaska Fisher- 
ies Science Center (AFSC) to conduct 
the annual eastern Bering Sea shelf 
survey. Although we monitor a full 
suite of trawl dimensions, such as 



door spread, wing spread, and hea- 
drope height that are typically mea- 
sured on trawl surveys, our attention 
was primarily focused on the distance 
between the footrope and lower bri- 
dles with the sea floor. Prior studies 
with this trawl have demonstrated 
that escapement under the footrope 
(Somerton and Otto, 1999; Munro 
and Somerton, 2002; Weinberg et 
al., 2004) and herding by the bridles 
(Somerton and Munro, 2001) are the 
most important determinants of catch 
efficiency for flatfishes and other 
benthic species. Although we under- 
stand that catch efficiency depends 
on animal behavior as well as trawl 
geometry, a goal of our study was to 
assess whether the 4% critical value 
is appropriate to prevent an appre- 
ciable degradation of catch efficiency 
due to warp asymmetry, the result of 
unequal trawl warp lengths. 



Materials and methods 

Experimental design 

The experiment was conducted during 
14-17 September 2003 along the Alaska 
Peninsula in Bristol Bay approximately 
85 km NE of Amak Island (55°58'N, 
162°55'W) on smooth, relatively level 
bottom at a depth of 82 m. Trawling 
was performed with the chartered 
38-m stern trawler FV Vesteraalen. 
The Vesteraalen is powered by a single 
1725-hp engine and is equipped with 
split Rapp Hydema (Rapp Hydema AS, 
Bodo, Norway) trawl winches carrying 
2.5 cm (1") diameter, compacted, solid- 
core trawl warp. 



22 



Fishery Bulletin 104(1) 



door legs 
K extension 

■l52m ~T 



bridle 
* 55 rn* 






61 


)m 








U 1 m V«^ 1 




^X^j^.^-^'^*^oo'fope center 












^ tootrope corner 












,^^^^ tootrope wing 






50 m 


40 m 
bridle 


25 m 
bridle 




^ -^ 


bridle 








otter 


toor 











Figure 1 

Schematic diagram of the 83-112 Eastern survey trawl and rigging shown 
from the side (upper panel) and from above (lower panel). The experimentally 
determined mean door spread and wing spread dimensions shown apply when 
towing at a depth of 82 m with 274 m of trawl warp on each side. Bottom 
contact sensor units, shown as oversized triangles along the bridles and 
footrope, are labeled by position as discussed in the text. 



The 83-112 Eastern is a low-rise, 2-seam, flatfish 
trawl designed for use on smooth, soft bottom. The 
nylon net is constructed of 10.1-cm stretch mesh in 
the wing and body, 8.9 cm in the intermediate, 
and double 8.9-cm mesh lined with a 3.1-cm mesh 
in the codend. It is towed behind a pair of 1.8x2.7 
m steel "V" doors, weighing approximately 816 kg 
apiece, which are attached to the net by two 3-m- 
long door legs (consisting of 1.6-cm long-link chain); 
a 12.2-m-long door leg extension (consisting of 1.9-cm 
diameter stranded wire); and a pair of 55-m-long, 
bridles (consisting of 1.6-cm diameter bare stranded 
wire) on each side of the net (Fig. 1). The 25.5-m-long 
(83 ft) headrope has 41 evenly spaced, 20.3-cm diameter 
floats that provide 116.4 kg of total lift. The 34.1-m-long 
(112 ft), 5.2-cm diameter footrope is constructed of 1.6-cm 
diameter stranded-wire rope protected by a single wrap 
of both 1.3-cm diameter polypropylene line and split 
rubber hose. The footrope is weighted with 51.8 m of 
chain (0.8-cm proof-coil) attached at every tenth link, 
forming 168 loops to which the netting is hung. An ad- 
ditional 0.6-m-long, 1.3-cm long-link chain extension 
connects each lower bridle to the trawl wing tips to 
help keep the footrope close to the bottom. Because the 
wire length between otter doors measured along the 
bridles and footrope is 175.6 m, the critical value for 
differential warp length for this trawl established by 
the NOAA 4% rule is 7 m. 

Prior to the experiment, the warp length to be used 
was measured to the nearest 0.1 m with a calibrated, 
in-line wire counter (Olympic 750-N, Vashon, WA). A 



zero reference point was painted on each trawl warp, 
as determined by first setting the wire counter to zero 
with the trawl door just beneath the water surface and 
then measuring out 274 m (150 fml, the survey stan- 
dard warp length for a fishing depth of 82 m. To verify 
this benchmark measurement, the process was repeated 
and all replicate measurements were found to be within 
0.3 m, which is less than the 1 m specified by the NO- 
AA trawl survey protocols for replicate measurements. 
Subsequent reference marks, measured by tape, were 
then placed along each warp at 3, 5, 7, 9, 11, and 20 m 
from the zero mark. 

The experiment consisted of examining the effect of 
seven differences in warp length, henceforth referred to 
as "offsets," where one warp was positioned at values 
of 0, 3, 5, 7, 9, 11, or 20 m longer. An experimental set 
consisted of all seven offsets, chosen in random order. A 
treatment consisted of towing the trawl, with the winch 
brakes locked, at the specified offset for 5 minutes at 
3 knots while maintaining a constant vessel heading. 
Treatments were preceded by a 2-min equalization pe- 
riod at the specified offset. A haul consisted of two 
treatment sets: one with a port offset and the other 
with a starboard offset — the order having been chosen 
randomly for each haul. To minimize the effect of bot- 
tom currents on trawl symmetry, hauls were made in 
pairs along the axis of the dominant current direction, 
either with or against the current — again, the order 
having been chosen randomly. This direction was deter- 
mined by deploying a current meter (Nobska MAVS-3, 
Woods Hole, MA) 3 m above the bottom for one day 



Weinberg and Somerton: Variation in Irawl geometry due to unequal warp length 



23 



prior to the start of the experiment. Hauls were 
made with the trawl codend open to eliminate any 
catch effects on trawl geometry. 

Several measures of trawl geometry and per- 
formance were taken during each treatment. The 
distance between the doors (door spread), the wing 
tips (wing spread), and the center of the headrope 
to the sea floor were measured acoustically with 
Scanmar sensors (Scanmar, Asgardstrand, Nor- 
way) to 0.1 m at 4-s intervals. Water flow, both 
perpendicular and tangential to the headrope, 
was measured to 0.1 knot at 24-s intervals with a 
Scanmar trawl speed sensor placed at the center 
of the headrope. Vessel position was measured 
with satellite navigation at 2-s intervals. Bridle 
tension was measured in kilograms at 2-s inter- 
vals using in-line tension recorders (Billings Ind. 
TR-999, N. Falmouth, MA) attached behind the 
door legs. Bottom current velocity in cm/s and 
direction data were recorded at 10-s intervals. 

Footrope off-bottom distance was measured at 
five positions simultaneously by placing bottom 
contact sensors (BCS) at the center of the foo- 
trope, at the corners located 3 m to either side 
of the center, and on each wing 1 m behind the 
wing tip (Fig. 1). These sensors are self-contained 
units consisting of a tilt meter, which measured 
angle to the nearest half degree at 0.5 s elapsed 
time intervals, and a data logger housed in a wa- 
tertight stainless steel container that fits inside 
a steel sled (Somerton and Weinberg, 2001). One 
side of this sled clips into a clamp on the footrope, 
which allows that end of the sled to pivot freely 
about the footrope while the other end drags along 
the bottom (Fig. 2). In this way, changes in the 
distance of the footrope from the bottom produced 
changes in the recorded tilt angle. Conversion 
from tilt angle to distance off-bottom was accom- 
plished by applying a calibration function derived 
for each BCS unit by fitting a quadratic function 
to data from an experiment in which angles asso- 
ciated with known distances from a hard surface 
were measured. The BCS unit extended 44 cm 
behind the footrope and weighed (BCS, sled, and 
footrope clamp) 8.9 kg in seawater. The clamp ex- 
tended beneath the footrope by 2 cm and, depend- 
ing on the extent of penetration into the sediment, 
could raise the footrope off the bottom (Fig. 2). 
Because the degree of penetration is unknown, 
no adjustments to our calibration functions were 
made. 

Bridle off-bottom distance was measured at six posi- 
tions simultaneously by placing BCS units on the lower 
bridle at distances of 25, 40, and 50 m forward of the 
wing tip on both sides of the trawl (Fig. 1). However, the 
BCS units used on the bridles differed from those used 
on the footrope. These units were mounted on a trian- 
gular frame designed to hold the BCS perpendicular 
to the bridle (Fig. 2; Somerton, 2003). The triangular 
frame measured 49 cm in its longest dimension and was 




Figure 2 

Bottom contact sensors shown mounted to the footrope (upper) 
and bridle (middle). The footrope shown in the "on-bottom" 
position without lateral tension and on a hard surface is 
elevated 2 cm by the footrope clamp (lower i. 



held in place by a cable stop that also extended beneath 
the bridle by about 2 cm. The weight of a bridle BCS 
unit and frame was 8.7 kg in seawater. 

Data analyses 

Three tilt angle measurements from each BCS unit were 
averaged for each 1.5-s interval, converted to distance 
off-bottom by applying the calibration function deter- 
mined for that unit, and then the off-bottom distances 



24 



Fishery Bulletin 104(1) 



were averaged for each experimental treatment on each 
haul. Mean off-bottom distances for each bridle and foot- 
rope position, except the footrope center, were grouped by 
offset distance (i.e., long side vs. short side of an experi- 
mental treatment) where, for example, the distance 
measurements from the 25-m position on the starboard 
bridle collected on a starboard treatment (long side) was 
grouped with the measurements from the 25-m position 
on the port bridle during a port treatment. Under these 
groupings, we assumed equality and subsequently refer 
to the offsets by whether they were on the long or short 
side (e.g., long 5 m) rather than whether they were on 
the port or starboard side (e.g., port 5 m). 

To allow interpolation between the experimental off- 
sets, cubic spline models (Venables and Ripley, 1994) 
were fitted to door spread, wing spread, headrope 
height, and the off-bottom distances for each bridle and 
footrope position as a function of offset. Bootstrapped 
empirical 95% confidence intervals (CI; Efron and Tib- 
shirani, 1993) were estimated for all measured catego- 
ries as described in the following example for off-bottom 
distance as follows: 1) assuming that mean off-bottom 
distances within a set were correlated because of local 
environmental conditions, we chose 1000 bootstrap rep- 
licates by sampling entire sets of measurements with 
replacement; 2) we fitted a cubic spline, weighted by the 
inverse of the variance, to each bootstrap sample, and 
then predicted the mean off-bottom distance for each 
1 m of offset; and 3) we ranked the predicted values, 
then chose the 25"^ highest and lowest values as the CI 
bounds. Similarly, we estimated the empirical 95% CIs 
about the mean off-bottom distance at zero offset in the 
same manner except that only zero offset treatments 
were considered. 



angle-of-attack between the bridle and the direction 
of travel (a). For flatfish, the vertical distance of the 
bridle from the sea floor at which a herding response is 
initiated (reaction height) will vary with species, size, 
viewing conditions, arousal state, and other variables, 
but for illustrative purposes, we considered a reaction 
height of 1 cm where video observations indicated that 
this value is appropriate for a small flatfish, initially 
at rest and unaware of the approaching bridle (Somer- 
ton, unpubl. data). Thus the value of the bridle contact 
length was determined as the distance between the 
wing tip and the point along the bridle at which the 
interpolated value of off-bottom distance reached the 
reaction height. The angle-of-attack, a, was not mea- 
sured during the experiment; however, when the trawl 
is symmetric, a can be modeled as 

a = Sin"'(0.5(D-W)/B), 

where D = the distance between doors; 

W = the distance between wing tips; and 
B = the distance between the wing tip and the 
door. 

It is not clear, however, how a will differ between the 
long and short side of the trawl when warp offsets occur. 
For illustrative purposes, we assumed that o is sym- 
metrical and remains constant during all experimental 
values of warp offset. Thus, for each side of the trawl 
and for each offset, a value of bridle contact length and 
o were first calculated as above, then the width of the 
herding area on each side of the trawl was computed as 
the bridle contact length times sin (o) and the two areas 
were summed together. 



Modeling trawl shape 

To help visualize the distortion of the trawl that occurs 
in response to offset, we created three views of the trawl; 
1) the shape of the lower bridle when viewed laterally, 2) 
the shape of the headrope when viewed from above, and 
3) the shape of the footrope when viewed from in front 
of the trawl. In addition, we calculated the area swept 
by the bridles (herding area) and the effective net width 
(i.e., the greatest lateral dimension of the net). 

Bridle shape and herding area 

The shape of the lower bridle, when viewed laterally 
(i.e., off-bottom distance as a function of position along 
the bridles), was approximated by linear interpola- 
tion between the mean off-bottom distances measured 
at the wing, and the 25-, 40-, and 50-m bridle posi- 
tions. Such shape functions were calculated for both 
the long side and the short side of the trawl at each 
offset increment. 

The herding area can be considered as a function 
of the bridle contact length, that is, the length of the 
bridle that is sufficiently close to the bottom to elicit a 
herding response (Somerton and Munro, 2001) and the 



Headrope shape and effective net width 

The curved shape of the headrope when viewed from 
above can be approximated as a quadratic (parabolic) 
function when the warps are equal in length and there 
are no external forces to distort the symmetry of the net 
(Fridman, 1969). If the shape remains parabolic as the 
warp offset is increased, then the headrope shape can 
be uniquely determined with three geometric measure- 
ments: 1) the length of the headrope; 2) the distance 
between wing tips; and 3) the tangent to the headrope at 
its center. The first quantity was measured at the start 
of the experiment. The second quantity was measured 
acoustically on all hauls, and then averaged by treatment. 
The third quantity was calculated as the quotient of the 
tangential water velocity (U) divided by the perpendicular 
velocity (V) measured by the headrope speed sensor (i.e., 
U/V). With these quantities, the headrope shape was 
determined as described in Appendix A. 

Although the shape of the headrope determined by 
this method is assumed to be parabolic, the headrope 
becomes increasingly asymmetric about the direction 
of travel as the degree of warp offset increases because 
the wing tip on the short side of the net precedes that 
on the long side. When this distortion occurs, the mea- 



Weinberg and Somerton: Variation in trawl geometry due to unequal warp length 



25 




-10 10 

U (crn/sec) 

Figure 3 

Current velocity vectors during each of the 
13 experimental tows are shown with arrows. 
Vessel direction during these tows is shown 
with the dark lines. The axes, labeled U and 
V, indicate the latitudinal and meriodonial 
components of the velocity vectors in cm/s. 




2500 



2000 - 



1500 - 



1000 



Warp offset (m) 

Figure 4 

Mean tension lin kilograms) measured on both port and starboard 
door legs is shown plotted as a function of warp offset in meters. 
Note that on the side with the shorter warp (negative offset), 
tension is higher than when the warps are equal and that on the 
side with the longer warp, the tension is lower. 



sured distance between wing tips becomes increasingly 
greater than the effective net width (i.e., the distance 
from wing tip to wing tip projected on a plane perpen- 
dicular to the direction of travel). The method used to 
estimate effective net width was based on headrope 
geometry and is described in Appendix A. 

Footrope shape viewed from the net mouth 

Because of footrope geometry, the importance of footrope 
bottom contact to overall net efficiency varies along 
the length of the footrope. This feature is true not only 
because escapement probability likely changes with the 
angle of the footrope in relation to the direction of travel 
but also because the proportion of the net width spanned 
by a unit length of footrope varies. To help visualize the 
latter effect better, we projected the off-bottom distances 
from their positions along the footrope onto a plane that 
was perpendicular to the direction of travel and spanned 
by the effective net width, using the footrope shape 
model determined for each offset increment described 
in Appendix A. 



Results 

Twelve successful tows consisting of 24 sets of treat- 
ments were completed during the experiment. Bottom 
current velocities ranged from about 5 cm/s to about 



30 cm/s and current direction was approximately paral- 
lel to the trawl towing direction (Fig. 3); consequently 
the mean current velocity perpendicular to the towing 
direction was quite small (1.8 cm/s). 

Bridle tension and geometry 

For the single haul in which both tension meters worked 
successfully, the bridle tension, combined over both 
sides, did not change significantly when regressed on the 
warp offset (df=ll, P=0.69). This result indicates that 
any distortion of the trawl due to the offset treatments 
was not sufficient to appreciably affect the combined 
tension and, therefore, the hydrodynamic and frictional 
drag of the net and the bridles. However, changing the 
relative length of the warps resulted in a progressive 
transfer of the tension to the shorter warp (Fig. 4). For 
example, based on the average combined bridle tension 
(3248 kg), when the difference in warp lengths was 
11 m, the shorter warp carried 72% of the total net and 
bridle drag. 

Headrope speed through the water 

The velocity component perpendicular to the headrope 
(U) decreased with increasing warp offset, whereas the 
absolute value of the velocity tangential (velocities from 
the port side of the sensor are opposite in sign to veloci- 
ties from the starboard side) to the headrope (V) at its 



26 



Fishery Bulletin 104(1) 



center increased (Fig. 5). We interpret this as 
indication that the speed sensor at the headrope 
was rotated in relation to the direction of travel 
as the warp offset was increased. Calculating the 
angle between the perpendicular at the center 
of the headrope and the direction of travel as 
tan"'(U/V), the angle increased from 11° at 3 m 
offset to 61° at 20 m offset. Although the absolute 
value of the tangential velocity was measured at 
values >0 at zero offset, the tangential velocity 
was not significantly different than zero it-tesi. 
P=0.71). This result indicates that the alignment 
of the experimental tows in relation to the pre- 
vailing current was sufficient to reduce the cross 
current to negligible levels. 

Net and door measurements 

At zero offset, the mean door spread obtained 
was 61.9 m. The mean wing spread was 17.1 m, 
and the mean headrope height was 2.0 m. Differ- 
ences in the means of all three quantities were 
not apparent at lower offsets, however headrope 
height and wing spread were more sensitive to 
changes in large offsets because both were signifi- 
cantly (P<0.05l greater than the zero offset means 
when offset was increased to 11 m, whereas door 
spread did not differ significantly until the offset 
was approximately 14 m (Fig. 6). 

Bridle and footrope distance off-bottom 
by position 



Bridle and footrope off-bottom distance varied consider- 
ably with position, not only with respect to the mean 
value but also with respect to the sensitivity of the mean 
to changes in offset. At zero offset, the mean off-bottom 
distance of the bridle declined from 12.3 cm at 50 m from 
the wing tips, to 3.2 cm at 40 m and 2.0 cm at 25 m (Fig. 
7). Mean off-bottom distance remained small along the 
footrope, varying from 1.7 cm at 1 m behind the wing tip 
to 2.5 cm at the corner and 1.9 cm at the center. 

The mean response to changes in offset varied greatly 
by position. Along the bridles, the most sensitive location 
was at 50 m, where off-bottom distance increased on the 
short side and decreased on the long side with increasing 
offset (Fig. 7). At 40 m. a similar pattern was repeated, 
but for most offsets on the long side, the off-bottom dis- 
tance was near the minimum recorded, indicating that 
the bridle was resting on the bottom. At 25 m, the bridle 
was nearly always in contact with the bottom and off- 
bottom distance was insensitive to variations in warp 
offset. Along the footrope, the most sensitive position was 
the corner where off-bottom distance increased greatly 
with offset, particularly with positive offsets due to the 
relaxation in warp tension. At the center of the footrope, 
off-bottom distance was also sensitive to warp offset, re- 
sponding almost identically on the long and short sides. 
At 1 m behind the wing tip, sensitivity to warp offset 
was quite low and the off-bottom distance indicated that 




Figure 5 

Current velocity (in m/sec) measured at the center of the hea- 
drope for each tow is shown separated into the component per- 
pendicular to the headrope (O) and the component tangential to 
the headrope ( + ). The solid and dashed lines connect the means 
at each offset increment. Note, for clarity, that the offsets are 
incremented by plus or minus 0.1 m for the tangential and 
perpendicular components. 



the footrope was in contact with the bottom except for 
large offsets on the long side. 

An alternate method of assessing the sensitivity of 
geometry of the 83-112 Eastern trawl to changes in 
offset is to determine if the mean off-bottom distance 
at 7 m, the maximum offset allowed under the NOAA 
protocols for the 83-112 Eastern trawl, differs statisti- 
cally from the mean off-bottom distance at zero offset. 
Based on the bootstrapped confidence intervals (Fig. 7), 
off-bottom distance is significantly different from what 
it is at zero offset at the 50-m and 40-m bridle positions 
and at the center and corner footrope positions but is 
not significantly different at the wing and the 25-m 
bridle position. 

Bridle shape and herding area 

To understand better how the change in tension that 
accompanies offsets in warp leads to changes in bridle 
shape, we show the mean off-bottom distances plotted 
against the BCS positions on the wing and bridles for 
both the short side and long side of the trawl. From this 
perspective it is clear that as the tension is increased, 
off-bottom distance increases on the forward part of the 
bridle. Likewise, as the tension is reduced, the off-bottom 
distance decreases (Fig. 8). For flatfish, the effect of 
these changes in off-bottom distance is a change in the 
area subjected to herding stimuli. For the case where 
the reaction height is 1 cm, the bridle contact length 
is determined by the intersection of the line depicting 



Weinberg and Somerton Variation in trawl geometry due to unequal warp length 



27 



the reaction height with the lines depicting the bridle 
shape at each offset (Fig. 8). The change in these lengths 
on the short and long sides of the trawl is asymmetric 
with changes in warp offset (Fig 9). For the long side, 
bridle contact length increases linearly with positive 
offset. However, for the short side, bridle contact length 
decreases nonlinearly with warp offset — the greatest 
changes occurring with small offsets. This difference 
likely leads to a change in the total width of the herding 
area with changes in warp offset. If, for example, it is 
assumed that the angle-of- attack (a) is the same for the 
long and short sides of the trawl, then the width of the 
herded area declines to a minimum at about 8 m offset, 
at which the herded area is reduced by W.SVr compared 
to that at zero offset. 

Headrope shape and effective net width 

With increasing difference in warp length, the model we 
used to describe headrope shape predicts three distinct 
changes in shape. First, the headrope is distorted so 
that the wing tip on the short side of the trawl precedes 
that on the long side in the direction of travel (Fig. 10). 
The difference in the forward position of the wing tips, 
however, is much less than the warp offset. For example, 
an 11-m difference in warp length resulted in an offset 
in the position of the wing tips of only 2-3 m. This dif- 
ference occurs because the increased tension on the short 
warp changes the catenary in both the bridles and the 
warps (i.e., both become effectively longer as the sag is 
reduced). Second, the headrope is distorted so that its 
center is increasingly displaced away from the midpoint 
between wings and toward the short side of the trawl. 
When this displacement occurs, the perpendicular at 
the center of the headrope is no longer aligned with the 
direction of travel. Third, the headrope is distorted so 
that the effective width of the net (i.e., the wing spread 
projected to the line perpendicular to the towing direc- 
tion) becomes increasingly shorter than the distance 
measured by the acoustic net sensors. The difference 
between the effective and the measured net width is 
negligible for offsets up to 7 m but rapidly increases at 
greater offsets (Fig. 11). 

Footrope shape viewed from In front of the net 

The distance of the footrope off-bottom, when viewed from 
a position in front of the net, increases with increasing 
offset; however, the location of the maximum off-bottom 
distance in relation to the midpoint between wings, 
shifts slightly with increasing offset (Fig. 12). With off- 
sets of 9 m or less, the position of maximum off-bottom 
distance is at the corner of the footrope on the long side 
of the trawl. However, with increasing offset, the shift 
in the position of the footrope corner changes because 
of the rotation of the trawl in relation to the direction 
of travel; and at a 20-m offset the footrope corner on 
the long side of the trawl is positioned, when viewed on 
a plane perpendicular to the direction of travel, almost 
exactly midway between the wing tips. 




14 16 18 20 




10 12 Id 16 18 20 



i ^ + + ^ 



pqqrt 



10 12 14 16 18 



Offset (m) 

Figure 6 

Mean door spread, wing spread, and headrope height 
are shown ( + ) plotted against offset increment. The 
means for all values of offset increment were fitted with 
a cubic spline function (solid curve). Bootstrapped 959^ 
confidence bounds are shown with shading. Also shown 
are the mean door spread, wing spread, and headrope 
height for treatments with zero offset (solid horizontal 
line) and the corresponding bootstrapped 95% confidence 
bounds Idashed horizontal lines). 



28 



Fishery Bulletin 104(1) 



Bridle 50 m forward of wing tip 



"H 



:l4 i 



* * 



:'i^/- 



: i~ 



-20 -15 -10 -5 5 10 15 20 



18 
16 

14-^ 

10 



Bridle 40 m forward of wing tip 



"--t - I 



t 






£ ^ T . m 



-i 



-20 -15 -10 



10 15 20 



Bridle 25 m forward of wing tip 



3.5- 




10 
9 
8 
7 
6 
5 
4 
3 
2 
1 


30 
25 
20 
15 



Footrope 1 m aft of wing tip 




Footrope corner 




20 
18 
16 

14- 

12 

10 

8 

6 

4 

2 



Footrope center 



/ 



-20 



-15 



-10 



10 



15 



Offset (m) 



Figure 7 

Mean bridle and footrope distance at each bottom contact sensor position is shown ( + i plotted against offset increment. 
The means for all values of offset increment were fitted with a cubic spline function (solid curve). Bootstrapped 95':r 
confidence bounds are shown with shading. Also shown is the mean distance off-bottom for only treatments with zero 
offset (solid horizontal line) and the bootstrapped 95% confidence bounds (dashed horizontal lines). 



Weinberg and Somerton Variation in trawl geometry due to unequal warp length 



29 







Long 




25- 








20- 








15 - 






/° 


10- 






/.. 


5 




^^^^^ 




0- 









10 



20 



30 40 



10 20 30 40 

Distance from wing tip (m) 



50 





Shon 


,20 


25 




Ir 


20- 




r 


15 - 




i: 


10- 


J 


1/ 


5- 


^^. 


/ 








- 







50 



Figure 8 

Mean off-bottom distance is shown plotted against the 
distance measured from the wing tip to the positions of 
the wing and the three bridle bottom contact sensors. 
This approximation to the shape of the bridle when 
viewed laterally is shown for each of the offset incre- 
ments for both the side with the longer warp and. the side 
with the shorter warp. The dashed line represents the 
hypothetical reaction height of a fish. The intersection 
of the dashed line with the solid line for each configura- 
tion defines the bridle length that is sufficiently close 
to the bottom to elicit a herding response. 



50 - 






, - - -° ' 


_ 45 - 

B) 

1 40- 


o--o--°-"" 

.0- ' 
.- O' 


03 


\ 


§ 35- 


\ 


03 


\ 


^ 30 - 


o\ 1^ 


'-' — D 


25 - 






5 10 15 20 


24.5 - 


\ 1 


- 24.0 - 

Q. 


\ 


P 23.5 - 


\ 1 


JD 


\ 1 


0) 


\ 1 


% 23.0 - 


\ 1-10.3% y 


1 22.5- 


\ ^y^ 


22.0 - 


^ — "^^^ 




5 10 15 20 


Offset (m) 


Figure 9 


The length of the bridle with the off-bottom distance 


<1 cm is shown plotted against the offset increment 


in meters for both the short warp (solid line) and long 


warp (dashed line) sides of the trawl (upper panel). In 


both cases the lines are represented using cubic spline 


smoothing functions. The width of the swept bridle path 


as a function of warp offset is represented in the lower 


panel. 



Discussion 

There are two distinct approaches forjudging whether a 
difference in warp length between the sides of a survey 
trawl will lead to a significant bias in estimates of rela- 
tive abundance. In both approaches we focused on the 
adequacy of the maximum 7-m offset allowed for the 
83-112 Eastern trawl under NOAA trawl survey pro- 
tocols. In the first approach, we simply asked whether, 
given the sampling effort used in the experiment, any 
of the measured dimensions at 7-m offset were statis- 
tically different from zero offset. In our experiment, 
none of the three standard measures of trawl geometry 
(i.e., door spread, wing spread, and headrope height) 



differed from mean values at zero offset. This finding 
indicates that either these dimensions are fairly robust 
to changes in warp offset or that the acoustic measure- 
ment of these dimensions was insufficiently precise to 
detect a difference. Off-bottom distance, however, was 
significantly different at the two forward positions on 
the bridles and along the footrope at the center and 
corner positions. 

From the perspective of trawl survey standardiza- 
tion, however, the detectability of changes in geometry 
is not of primary importance; these changes, however, 
may produce a significant effect on estimates of relative 
abundance. Bias in these estimates could result either 
because the change in trawl geometry leads to an inac- 



30 



Fishery Bulletin 104(1) 




30 
25 
20 
15 
10 
5 


30 
25 

20 

15 

10 

5 



30 
25 
20 
15 
10 
5 




3 m offset 




5 m offset 




-15 -10 -5 5 10 15 -15 -10 -5 5 10 15 



7 m offset 




9 m offset 




-15 -10 -5 5 10 15 -15 -10 -5 5 10 15 



1 1 m offset 




30  




20 m offset 




25- 








20 - 








15 




,.'] 




10  




K j 




5 




\ I 




0- 




\ — r 





-15 -10 -5 5 10 15 -15 -10 -5 5 10 15 

Meters 

Figure 10 

Estimated net shape (curve i and position of headrope center (O) 
at varying levels of warp offset. The mean distance between wing 
tips that was acoustically measured during the experimental tows 
is indicated with a dashed line. The ma.ximum lateral dimensions 
of the net are indicated with solid vertical lines. The distance 
between these lines is the effective net width needed for estimat- 
ing the area swept by the net. 




5 7 9 

Offset (m) 

Figure 11 

The mean distance between wing tips mea- 
sured acoustically lOl during the experimental 
tows (net width) and the calculated effective 
net width (-t-) are shown plotted against the 
offset increment in meters. Note that there 
is little difference between the two measures 
of net width until the offset increment is 
increased to 9 m. 



curate measurement of swept area or because 
it leads to a change in catch efficiency. 

Relative abundance indices produced for the 
eastern Bering Sea shelf survey are based on 
catch per area swept between the trawl wings. 
As the net distorts on account of differential 
warp length, the effective net width will become 
increasing less than the width that is acousti- 
cally measured during the survey. Thus, net 
width will become increasingly overestimated 
and relative abundance of fish species therefore 
will be underestimated. At 7-m offset, however, 
the measured net width differed from the ef- 
fective net width by only 0.5%, and therefore 
this source of error is unlikely to contribute to 
bias in the swept area estimates. However, the 
difference between measured and effective net 
width increases rapidly at greater offsets and 
could present a problem if a less restrictive 
threshold value of offset were used. 

Catch efficiency of the 83-112 Eastern trawl 
depends primarily on 1) herding by the bridles, 
doors, and the mud clouds they create; 2) the 
escapement under the footrope; and 3) the es- 
capement through the mesh in the body of the 
net. The relative importance of these three 
processes, however, will vary for the major spe- 
cies groups that are targeted in the surveys. 
Gadoids (primarily walleye pollock [Theragra 
chalcogramma] and Pacific cod [Gadus macro- 
cephalus] appear to have little or no herding 
response to the 83-112 Eastern trawl (Somer- 
ton, 2004) and rarely pass under the footrope 
(Somerton, unpubl. data). However, both Pa- 



Weinberg and Somerton; Variation in trawl geometry due to unequal warp length 



31 




20 












15  


10 


ig 




sh 


)rt 


10  












5 












U \ 




" ~ r 





5 10 15 

7 m offset 



long 










5 m offset 






20 












15 


long 




sh 


)r1 


10  












5 












U 1 






, 


^ 





5 10 

9 m offset 



loiig 




15 

Distance from wing tip (m) 

Figure 12 

Mean off-bottom distances (in cm) at the five bottom contact sensor positions along the 
footrope (circles) are shown for each warp offset. The positions are projected onto the wing 
tip to wing tip plane to depict the footrope as it would appear if one were looking into the 
net from the direction of travel. The vertical solid lines indicate the positions of the wing 
tips on the long warp and short warp sides of the trawl. The dashed line indicates the 
midpoint between wing tips. Note that the projection considers the reduction in effective 
net width with increasing offset. 



cific cod and walleye pollock are found gilled in the 
body of the net; therefore some mesh escapement may 
occur, especially if any distortion of the net results in 
altered water flow through the meshes. Video observa- 



tions of crabs (snow and Tanner crabs [Chionoecetes 
sp.] and king crabs [Paralithodes sp.] show that they 
also exhibit little or no herding response to the 83-112 
Eastern trawl (Weinberg, unpubl. data). However, both 



32 



Fishery Bulletin 104(1) 



Chionoecetes species (Somerton and Otto, 19991 and red 
king crab (Weinberg et al., 20041 do escape under the 
footrope. Flatfishes (including yellowfin sole [Limanda 
aspera], flathead sole [Hippoglossoides elassodon], and 
rock sole [Lepidopsetta biliueata] display a strong herd- 
ing response to the 83-112 Eastern trawl (Somerton 
and Munro, 2001); as much as 49% of the catch con- 
sisted of fish that were herded by the bridles into the 
net path. Likewise, flatfishes are readily capable of 
escaping under the footrope and, for species such as 
yellowfin sole, at least 25% of the largest individuals 
escape in this manner (Munro and Somerton, 2002). 
Thus, the capture efficiency of the trawl is species- 
specific. The change in catch efficiency due to warp 
offset is likely minimal for species not captured as a 
function of herding and footrope escapement behaviors; 
however, because flatfish are susceptible to herding and 
are adept at footrope escapement, their catch rate could 
potentially be affected most by warp offsets. 

Bridle efficiency (i.e., the fraction of fish in the area 
between the wing tips and doors that are herded into 
the path of the net) for flatfish catch is strongly influ- 
enced by the size of the herding area or area swept by 
the bridles because flatfish are stimulated to herd by 
the close approach or direct contact of the lower bridle. 
Although we were able to measure the off-bottom dis- 
tance along the bridle and thereby predict the shape of 
the bridle, there is still considerable uncertainty as to 
the exact size of the herding area because the reaction 
height of a fish will vary with species, size, physiological 
state, state of arousal to the approaching bridle, viewing 
conditions for the fish, and, perhaps, other variables. 
Additionally, there is uncertainty in the estimate of the 
size of the herding area because it is based on the as- 
sumption of symmetry in the bridle angle-of-attack — a 
symmetry that is increasingly untenable with increasing 
offset. Despite this uncertainty, it is likely that the loss 
of herding area on one side of the trawl is not countered 
by an increase on the other side; thus some overall loss 
of herding efficiency is to be expected. In the hypotheti- 
cal case chosen in our study, the reduction in herded 
area was 10.3%, which when applied to a strong herding 
flatfish such as rock sole (the herded component of the 
catch has been estimated to be about 49%, Somerton 
and Munro, 2001), the expected reduction in catch with 
an 8-m offset would be roughly 5%. 

Flatfish escapement under the footrope will be influ- 
enced not only by the increase in off-bottom distance 
but also by the location along the footrope where the 
increase occurs. At a 7-m offset, the footrope off-bottom 
distance is about 2 cm higher in the footrope corner 
on the long side of the trawl and approximately 1 cm 
higher at the center and opposing corner than at zero 
offset (Fig. 12). Footrope off-bottom distances increased 
appreciably with greater offsets. Weinberg et al. (2002) 
demonstrated that flatfish escapement can increase 
with similar increases in footrope off-bottom distance; 
however their study focused on a different trawl and 
considered escapement for the entire footrope rather 
than by position along the footrope. Although we are 



unaware of any studies that quantify escapement rate 
by position along the footrope, our video observations 
indicate that fiatfish are less likely to escape under the 
footrope near the wings than in the center (Somerton, 
unpubl. data). Because it is the center portion of the foo- 
trope where most of the increases in off-bottom distance 
occur in the 83-112 trawl, flatfish escapement is likely 
increased and potentially could represent a significant, 
but presently unquantifiable, loss in catch and a source 
of bias in estimates of relative abundance. 

In conclusion, most aspects of the 83-112 Eastern 
trawl geometry were significantly degraded by warp 
offset differences equal to or greater than 7 m compared 
to zero offset. More importantly, the locations where the 
detectable differences occurred could affect catch effi- 
ciency; therefore a NOAA threshold value of 4% should 
be considered a maximum value for the 83-112 Eastern 
trawl and perhaps a more conservative value (less than 
4%) would be prudent. However, given today's standard- 
ized survey procedures for measuring warp and for real- 
time monitoring of warp offset, the probability of warp 
offsets even approaching 7 m is highly unlikely on our 
surveys when locked-winches are used. Likewise, we 
argue that any appreciable differences in warp lengths 
between sides due to stretching are unrealistic because 
AFSC charter vessels use large diameter, compressed, 
solid-core wire. In fact, a review of the 413 hauls made 
during the 2004 EBS survey revealed that only three 
tows had a recorded maximum 1-m length difference 
between sides (Weinberg, unpubl. data). 



Acknowledgments 

We would like to thank Captain Brad Lougheed and the 
FV Vesteraalen crew, Niles Griffen, Waldemar Janezak, 
Van Ngo, and Todd Becker for their outstanding profes- 
sionalism, positive attitudes, and constant attention to 
detail; AFSC scientists Stan Kotwicki and Dennis Benja- 
min for their assistance at sea; reviewers Guy Fleischer 
and Henry Milliken for their helpful comments; and the 
AFSC editorial staff, including Gary Duker, Jim Lee, 
and Kama McKinney for their help during the in-house 
review and manuscript preparation process. 



Literature cited 

Efron, B., and R. Tibshirani. 

1993. An introduction to the bootstrap, 436 p. Chapman 
and Hall, New York, NY. 
Fridman, A. L. 

1969. Theory and design of commercial fishing gear. 
(Transl. from Russian by Israel Program Sci. Transl., 
Jerusalem) 1973, 489 p. [Available as TT 71-50129 
from Natl. Tech. Inf. Serv., Springfield, VA.] 
Munro, P. T, and D. A. Somerton. 

2002. Estimating net efficiency of a survey trawl for 
flatfishes. Fish. Res. 55:267-279. 
Somerton, D. A. 

2004. Do Pacific cod (Gadus macrocephalus) and wall- 



Weinberg and Somerton: Variation in trawl geometry due to unequal warp length 



33 



eye pollock (Theragra chalcogramrna) lack a herding 

response to the doors, bridles and mudclouds of survey 

trawls? ICES J. Mar. Sci. 61:1186-1189. 
2003. Bridle efficiency of a survey trawl for flatfish: 

measuring the length of the bridles in contact with the 

bottom. Fish. Res. 60:273-279. 
Somerton, D. A., and P. T. Munro. 

2001. Bridle efficiency of a survey trawl for flatfish. Fish. 

Bull. 99:641-652. 
Somerton, D. A., and R. S. Otto. 

1999. Net efficiency of a survey trawl for snow crab, 

Chionoecetes opilio, and Tanner crab, C. hairdi . Fish. 

Bull. 97:617-625. 
Somerton, D. A., and K. L. Weinberg. 

2001. The affect of speed through the water on footrope 

contact of a survey trawl. Fish. Res. 53:17-24. 



Stauffer. G. 

2004. NOAA protocols for groundfish bottom trawl sur- 
veys of the nation's fishery resources. NOAA. Tech. 
Memo. NMFS-F/SPO-65, 205 p. Alaska Fisheries 
Science Center, 7600 Sand Point Way N.E., Seattle, 
WA, 98115. 
Venables, W. N., and B. D. Ripley. 

1994. Modern applied statistics with S-plus. 462 p. 
Springer-Verlag, New York, NY. 
Weinberg, K. L., D. A. Somerton, and P. T. Munro. 

2002. The effect of trawl speed on the footrope capture 
efficiency of a survey trawl. Fish. Res. 58:303-313. 
Weinberg, K. L., R. S. Otto, and D. A. Somerton. 

2004. Capture probability of a survey trawl for red 
king crab iParalithodes camtschaticus). Fish. Bull. 
102:740-749. 



Appendix 1: Estimating headrope and footrope 
shape when the warps differ in length 

If a trawl headrope has the same shape as a flexible 
twine under a uniformly distributed load, then the shape 
of the headrope can be approximated as a quadratic 
(parabolic) function (Fridman, 1969; p. 84) as 



y = cx 



(1) 



where c is a constant controlling the shape (Fig. Al). 
As the headrope is distorted by a differential in warp 
length, not only does the value of c change, but the 
headrope is displaced along the path of the parabola, so 
that its center is no longer aligned with the vertex of the 
parabola. A unique solution to the shape of the headrope 
when it is distorted in this manner can be determined 
from three types of data: the total headrope length (L), 
the measured distance between the wing tips (W), and 
the measured slope (tangent) of the parabola at the 
center of the headrope (tan). The third quantity can be 
obtained from the V (perpendicular to the footrope) and 
U (tangential to the footrope) velocities measured by the 
headrope speed sensor as the quotient U/V. With these 
quantities, the solution can be obtained as follows. 

A small length interval measured along the headrope 
can be expressed as 




X(tn) 

Figure Al 

Shape of a trawl headrope as described 
by a parabola. The total length of the 
headrope (shown with a solid line) is 
equal to L. The measured width of the 
trawl (shown with a dashed line) is 
equal to W. The circle indicates the 
center of the headrope where a speed 
sensor is located. The speed sensor 
measures water speed both perpendicu- 
lar and parallel to the headrope. 



ds = (dy +dx 



(2) 



Which, after substitution of the derivative of Equation 
1, is 



ds = [l + (2cx)^fdx. 



(3) 



The length of any segment of the headrope, measured 
from the port end (->;/„„ ^r'' i^ then 



S= \[l + (2cx)^y ds. 



(4) 



A segment equal to the total length of the headrope 
is obtained by integrating up to the starboard end 



The solution is approximated numerically in two stag- 
es. First, for a trial value of c. Equation 4 is integrated 
from trial values of Ji:,^,,^,^ up to the value of x at which 
S=L/2 (i.e., -Vjjjj^^/p). The tangent at this position is then 
evaluated as 2cx,. 



nldlr 



(based on the derivative of Eq. 
1). This process is then repeated iteratively to find the 
value of .T:,„„.j,r for the specified value of c at which the 
calculated tangent equals the tangent value determined 
from the headrope speed sensor. The value of -v, is 



34 



Fishery Bulletin 104(1) 



then determined by integrating Equation 4 from Xi„,^.^,^ 
to the value of .v at which S=L. The wing tip to wing 
tip distance IJ,,,,,,^,,^ is then calculated as 

1 



D. 



•ingtip 



-yio 



,)^+(a:„ 



(5) 



where .v,,^^^. and yi„,^,^,, are obtained from Equation 1. In 
the second stage, c is varied and the above process is 
repeated iteratively until the value of O,, ,„„„^j is found 
that is closest to the measured net spread. At this point 
the calculated values of Z),,,„„„^, and the tangent at the 
headrope center will equal the measured values. 

The headrope-shape model for each offset was used 
to project the off-bottom distances measured at the five 
positions along the footrope onto a plane orientated 
perpendicular to the direction of travel to depict the 



shape of the footrope as it would appear from a position 
in front of the trawl. To do this, a shape function was 
developed for the footrope. Assuming that the coordi- 
nates of the endpoints of the footrope (-v,,,,, ^,^, -v^^,^^,,^, >'/o„,pr' 
y ^) were the same as the headrope. Equation 5 was 
iteratively integrated with varying values of c until the 
estimated value of the footrope length (S) equaled the 
true length (34.1 m): 



S= \ [l + (2cx)-)' dx. 



(6) 



Once c is determined, Equation 4 is integrated to find 
the value of .r associated with the value of S at each BCS 
(bottom current sensor) position on the footrope. 



35 



Abstract — Three aspects of a survey 
bdttdin trawl performance — 1) trawl 
geometry (i.e., net spread, door spread, 
and headrope height); 2) footrope dis- 
tance off-bottom; and 3) bridle dis- 
tance off-bottom — were compared 
among hauls by using either of two 
autotrawl systems (equal tension and 
net symmetry) and hauls conducted 
with towing cables of equal length 
and locked winches. The effects of 
environmental conditions, vessel 
heave, crabbing (i.e., the difference 
between vessel heading and actual 
vessel course over ground), and 
bottom current on trawl performance 
with three trawling modes were 
investigated. Means and standard 
deviations of trawl geometry mea- 
sures were not significantly different 
between autotrawl and locked-winch 
systems. Bottom trawls performed 
better with either autotrawl system 
as compared to trawling with locked 
winches by reducing the variance 
and increasing the symmetry of the 
footrope contact with the bottom. 
The equal tension autotrawl system 
was most effective in counteracting 
effects of environmental conditions 
on footrope bottom contact. Footrope 
bottom contact was most influenced 
by environmental conditions during 
tows with locked winches. Both of 
the autotrawl systems also reduced 
the variance and increased the 
symmetry of bridle bottom contact. 
Autotrawl systems proved to be 
effective in decreasing the effects of 
environmental factors on some aspects 
of trawl performance and, as a result. 
have the potential to reduce among- 
haul variance in catchability of survey 
trawls. Therefore, by incorporating 
an autotrawl system into standard 
survey procedures, precision of survey 
estimates of relative abundance may 
be improved. 



The effect of autotrawl systems on 
the performance of a survey trawl 

Stan Kotwicki* 

Kenneth L. Weinberg* 

David A. Somerton 

National Marine Fisheries Service 

Alaska Fisheries Science Center, 

7600 Sand Point Way N.E. 

Seattle, WA 98115 

E-mail address (for K L Weinberg, contact author) ken weinbergig'noaa gov 

'Equal authorship 



Manuscript submitted 27 October 2004 
to the Scientific Editor's Office. 

Manuscript approved for publication 
6 June 2005 by the Scientific Editor. 

Fish. Bull. 104:35-45 12006). 



Bottom trawl survey operating pro- 
cedures are standardized in order to 
reduce the variability of catch per unit 
of effort (CPUE) estimates. Many of 
the current standardization proce- 
dures address the efficiency of the 
trawl gear and the maintenance of 
constant catchability among samples 
and over time. Despite these efforts, 
variability in trawl catchability can 
be exacerbated by uncontrollable envi- 
ronmental conditions. Variables such 
as surface and bottom currents, sea 
state, wind direction, varying sub- 
strate types and inclinations, and 
depth of tow may all contribute to dif- 
ferences in gear efficiency by influenc- 
ing the area swept by the net (Rose 
and Nunnallee, 1998), the herding 
efficiency of the bridles (Somerton 
and Munro, 2001; Somerton, 2003), 
and escapement beneath the footrope 
(Weinberg et al., 2002). 

Many bottom trawl surveys con- 
ducted by the National Marine Fish- 
eries Service, such as the Alaska 
Fisheries Science Center's (AFSC) 
eastern Bering Sea (EBS) shelf sur- 
vey, operate with trawl winch brakes, 
set or locked, and tows are made with 
equal amounts of towing cable (warp) 
on both sides of the vessel. Other 
than by controlling towing speed and 
direction, these surveys are unable 
to compensate for changing environ- 
mental conditions. In contrast, auto- 
trawl systems are widely used by the 
commercial fleet and are purported 
to improve fishing performance by 
stabilizing trawl geometry over vary- 
ing environmental conditions, such 



as rough weather when vessel heave 
produces an upward lift on the trawl 
door resulting in loss of ground shear 
and wing spread, or over rough bot- 
tom when doors and nets have a 
greater probability of snagging. If 
autotrawl systems are able to reduce 
some of the variability in gear effi- 
ciency that is due to environmental 
variability, such as sea state and cur- 
rents, then including the use of auto- 
trawl systems as a standard survey 
bottom trawl procedure may improve 
the precision of survey results. 

In simple terms, autotrawls are dy- 
namic systems that operate on the 
principle of ensuring that the trawl 
is being towed in a direction perpen- 
dicular to the center of the footrope 
and headrope in order to optimize its 
performance. We are aware of two 
styles of autotrawl systems current- 
ly marketed. The first is a tension- 
controlled system that reacts to the 
difference in warp tension between 
winches by equalizing hydraulic pres- 
sure (equal tension). When the ten- 
sion on either side exceeds that of the 
other side (a user-defined threshold) 
due to factors such as increased drag, 
currents, sediments, or steep slopes, 
the system lengthens that warp to 
equalize the pressure between the 
two winches. Conversely, when the 
tension decreases on one warp, the 
system compensates by shortening 
that warp to equalize pressure be- 
tween the two winches. The second 
autotrawl style is a symmetry-con- 
trolled system that actively adjusts 
warp length in response to cross flow 



36 



Fishery Bulletin 104(1) 



signals from a sensor mounted on the trawl headrope. 
This system operates on the principle that net skewing 
can be caused by a crosscurrent. If the net is pulled 
square to the direction of flow then, its geometry will be 
symmetrical and trawl performance will be optimized. 
In the late summer of 2003, the AFSC conducted an 
experiment to examine the effect of these two types of 
autotrawl systems on the geometry of a survey trawl, 
comparing them to towing with equal amounts of warp 
on each side with the winches locked. The study consid- 
ers three aspects of trawl performance: 1) the factors of 
trawl geometry influencing the area and volume swept 
by the trawl (door spread, wing spread, and headrope 
height); 2) the bottom-tending performance of the foo- 
trope; and 3) the bottom-tending performance of the 
lower bridles. 



Materials and methods 

Operations and instrumentation 

The experiment was conducted during 19-25 September 
2003 aboard the chartered 38-m-long commercial stern 
trawler FV Vesteraalen on smooth, relatively level bottom 
in 115 m of water at a site approximately 70 km north 
of Unimak Pass in the Bering Sea (55°10'N, 166°15'W). 
The Vesteraalen is powered by a single 1725-hp engine 
and is equipped with split Rapp Hydema (Rapp Hydema 
AS, Bod0, Norway) trawl winches carrying 2.5 cm (1") 
diameter, compacted, solid-core trawl warp. The winches 
are controlled by a Scantrol 2000 (Scantrol, Bergen, 
Norway) winch control system capable of quickly switch- 
ing to different towing modes as requested by the vessel 
operator. For this experiment towing was performed 
with the codend open and with three different winch 
control modes; locked winches with equal amounts of 
warp on the port and starboard side (locked); a tension- 
controlled autotrawl, which maintains equal tension on 
both warps by adjusting warp length based on the drag 
forces on each side (tension); and a symmetry-controlled 
autotrawl, which adjusts warp length according to side 
current forces in order to "optimize" water flow through 
the net (symmetry). The symmetry-controlled system 
requires a real-time speed sensor capable of detecting 
the direction of water flow across the headrope. We used 
an acoustically linked Scanmar (Scanmar, Asgardstrand, 
Norway) trawlspeed sensor that transmits flow data at 
24-sec intervals both perpendicular and tangential to 
the headrope at its center. For this experiment and 
when in symmetry mode, the Scantrol system adjusted 
warp length at 30-sec intervals in response to changes 
in tangential velocity. 

The experiment was conducted with the AFSC stan- 
dardized trawl for the BBS shelf survey, the 83-112 
Eastern bottom trawl. The 83-112 Eastern is a low- 
rise, 2-seam flatfish trawl designed to fish on smooth, 
soft bottom. The nylon net is constructed of 10.1-cm 
stretch mesh in the wing and body, 8.9-cm mesh in 
the intermediate, and double 8.9-cm mesh lined with 




Figure 1 

Schematic diagram of the 83-112 Eastern bottom 
trawl and rigging shown from above. Mean door 
spread (68.0 m) and mean wing spread (17.8 ml 
were calculated from e.xperimental tows made at 
a depth of 115 m with the winches locked and 366 
m of trawl warp out on each side. Bottom contact 
sensor units, shown as oversized triangles along 
the bridle and footrope, are labeled by position, 
as discussed in the text. 



3.1-cm mesh in the codend. It is towed behind a pair 
of 1.8x2.7 m steel "V" doors, weighing approximately 
816 kg apiece, which are attached to the net by two 
3-m-long, 1.6-cm long-link chain door legs, a 12.2-m- 
long, 1.9-cm diameter stranded-wire door leg exten- 
sion, and a pair of 55-m-long, 1.6-cm diameter bare 
stranded-wire bridles on each side (Fig. 1). The 25.5- 
m-long (83 feet) headrope has forty-one evenly spaced, 
20.3-cm diameter floats providing 116.4 kg of total lift. 
The 34.1-m-long (112 feet), 5.2-cm diameter footrope is 
constructed of 1.6-cm diameter stranded-wire rope that 
is protected with a single wrap of both 1.3-cm diameter 
polypropylene line and split rubber hose. The footrope 
is weighted with 51.8 m of chain (0.8-cm proof-coil) at- 
tached at every tenth link, forming 168 loops to which 
the netting is hung. An additional 0.6-m-long, 1.3-cm 
long-link chain extension connects each lower bridle to 
the trawl wing tips to help keep the footrope close to 
the bottom. 



Kotwicki et al.: Effect of autotrawl systems on tfie performance of a survey trawl 



37 



Prior to experimental towing, trawl warps were 
measured and marked at 366 m, the amount of 
warp used on the EBS survey when stations are 
fished at a depth of 115 m, the depth of our study 
site. Warps were measured and marked in ac- 
cordance with AFSC protocol (Stauffer, 2004) 
by using in-line wire counters (Olympic 750-N, 
Vashon, WA) while at the same time, calibration 
of the geometric winch counters associated with 
the autotrawl system was performed. 

A haul consisted of towing at a vessel speed 
of 3 knots while steering a steady course over 
ground. Tow direction was selected by attempt- 
ing to expose the trawl to the maximum amount 
of side current. Vessel speed and position were 
measured at 2-sec intervals with satellite navi- 
gation. Each haul consisted of two treatment 
sets in which three 15-min towing treatments 
(locked winches [locked], tension-controlled au- 
totrawl [tension], symmetry-controlled autotrawl 
[symmetry]) were conducted, allowing at least 
two minutes between treatments for the net to 
equilibrate. Randomizing the order of treatments 
within each treatment set reduced the influence 
of treatment order on trawl performance intro- 
duced by sea state, wind, and tidal currents. 
Likewise, towing at the same site for the dura- 
tion of the experiment eliminated bias that could 
be attributed to varying substrate. 

Wing spread, door spread, and headrope height 
were measured acoustically with Scanmar sensors 
at 4-sec intervals to the nearest 0.1 m. Footrope 
distance from the sea floor (cm) was measured 
at 0.5-sec intervals and averaged over 1.5-sec 
periods at five positions along the footrope si- 
multaneously by placing bottom contact sensors 
(BCS) at the center, at both trawl corners (located 
3 m to either side of the center), and on each wing 
1 m aft of the wing tips (Fig. 1). These sensors 
are self-contained units consisting of a tilt meter 
capable of measuring angle to the nearest 0.5° 
and a data logger housed in a watertight stainless steel 
container that fits inside a steel sled (Somerton and 
Weinberg, 2001). One side of the sled clips into a clamp 
that pivots freely on the trawl footrope and the other end 
drags along the bottom (Fig. 2). Changes in the distance 
of the footrope from the bottom produce changes in the 
recorded tilt angle. Conversion from tilt angle to distance 
off-bottom was accomplished with a calibration function 
determined for each BCS unit by fitting a quadratic 
function to data derived from a separate calibration 
experiment in which tilt angles were recorded when the 
footrope clamp was elevated set distances from a hard 
surface. The BCS unit extended out from the footrope 44 
cm and its combined weight (consisting of BCS, sled, and 
footrope clamp) was 8.9 kg in seawater. The thickness 
of the clamp beneath the footrope was 2 cm. Because 
underwater video equipment was unavailable for this 
experiment, the extent to which this clamp penetrates 
into variable substrates was not estimated. 











Figure 2 

Bottom contact sensors mounted to the footrope (top) and the 
bridle (bottom). 



Bridle distance from the bottom was measured at 
three positions simultaneously on both port and star- 
board sides by placing BCS units 25, 40, and 50 m 
forward of each wing tip. The BCS units and sled were 
mounted on triangular frames designed to hold them 
perpendicular to the bridle (Fig. 2, Somerton, 2003). 
The triangular frame measured 49 cm in its longest 
dimension. The combined weight of a BCS unit and 
frame was 8.7 kg in seawater. 

In addition to trawl mensuration, data were also 
collected on certain environmental variables during the 
different towing modes. Three variables were studied; 1) 
vessel heave measured at the trawl block; 2 ) the relative 
degree of offset of the warps from the heading of the ves- 
sel (crabbing); and 3) bottom current velocity both paral- 
lel and perpendicular to the direction of the vessel. 

The effect of sea state on the vertical displacement 
and attitude of the vessel is transmitted to the footrope 
from the trawl blocks through the trawl warps and 



38 



Fishery Bulletin 104(1) 



likely causes variable bridle and footrope contact with 
the bottom. Vessel heave at the starboard trawl block 
was used as a proxy for sea state. Heave, pitch, and 
roll data were collected at 1-sec intervals with a heave 
sensor (VT TSS, DMS-25, Watford, UK) mounted in the 
bridge along the centerline of the vessel. Heave data 
at the starboard block were then predicted, given the 
X, y, z coordinates of the block from the heave sensor, 
as distance (cm) from its equilibrium position. In the 
analyses, the standard deviation of the heave was used 
as the index of sea state. 

Net crabbing was subjectively assessed on a four-point 
scale by a single observer, then numerically coded as 
follows; 1) none — the net trailed straight behind the 
vessel; 2) slight — the warp could be seen entering the 
water between the side rail and the aft gantry; 3) mod- 
erate — the point of entry of the warp into the water was 
blocked from view by the aft gantry; and 4) severe — the 
warp was observed entering the water behind the stern 
ramp. Conditions usually remained constant and one 
observation was made per towing-mode treatment. How- 
ever, in some instances conditions changed rapidly and 
warranted more than one code. In such cases the aver- 
age of the observation codes was used in the analyses. 

Bottom current direction and velocity (cm/sec) were 
measured at 10-sec intervals with an oceanographic 
current meter (Nobska, MAVS-3, Woods Hole, MA) 
moored in the vicinity of our trawling activity three 
meters from the bottom. The current data were parti- 
tioned into two directional components, one parallel to 
the course of the vessel and the other perpendicular, 
and averaged for each sample treatment. 

Data analyses 

Comparison of the means and standard deviations among 
treatments Our null hypothesis, that the three treat- 
ments had the same effect, was tested with a Krus- 
kal-Wallis one-way ANOVA for all measures of trawl 
performance. We selected this test because of the skewed 
nature of the data. To describe the effect of the three 
treatments on trawl geometry features the mean and 
standard deviation (SD) were calculated for the wing 
spread, door spread, and headrope height off-bottom 
from each treatment in each haul. 

To describe the bottom-tending performance of the 
footrope we calculated the following statistics for each 
treatment: 

1 mean footrope distance off-bottom — the sum of mean 
distances off-bottom along the footrope, from five 
footrope BCS units; 

2 standard deviation of the footrope distance off- 
bottom — the sum of standard deviations along the 
footrope, from five footrope BCS units; and 

3 symmetry of the footrope distance off-bottom — the 
sum of the absolute difference between the means 
of the two wing positions and the absolute differ- 
ence between the means of the two corner positions 
on the footrope. 



To describe the bottom-tending performance of the 
lower bridle, we considered only the BCS unit positioned 
40 m from the wing tip. The variability in bottom-tend- 
ing performance of the bridles 40 m from the wing tip 
may have the highest impact on the catchability of the 
trawl because the BCS at the 25 m position was always 
on the bottom and the BCS at the 50 m position was 
always off the bottom. Performance of the bridles at the 
40 m position was characterized by 

1 mean bridle distance off-bottom — the sum of the 
mean distances off-bottom from the BCS units lo- 
cated 40 m forward of the wing tip on both lower 
bridles; 

2 standard deviation of the bridle distance off-bot- 
tom — the sum of the standard deviations from the 
BCS units located 40 m forward of the wing tip on 
both lower bridles; and 

3 symmetry of the bridle distance off-bottom — the 
absolute difference between means from the BCS 
units located 40 m forward of the wing tip on both 
lower bridles. 

Assessment of the effect of environmental factors If 

differences among towing mode treatments were found 
by the ANOVA (P<0.05), then the effects of heave, 
crabbing, and bottom current on gear performance 
within each treatment were explored. Multiple regres- 
sion analyses were performed for each of the treatment 
statistics with environmental variables as dependents. 
At the start of the analyses, the models included all 
four dependent variables (heave, crabbing, current par- 
allel, and current perpendicular to the tow direction). 
Variance inflation factors (VIFs) were calculated for 
all variables to test for multicollinearity (Neter et al., 
1996). Dependent variables in all models had VIFs 
ranging from 1 to 1.4, indicating that no serious multi- 
collinearity existed among dependent variables. Models 
were simplified (backward deletion) until all P-values 
of the individual slopes were lower than 0.05. This 
procedure enabled us to establish which variables had 
a statistically significant impact on performance of the 
trawl within each treatment. 



Results 

A total of 21 successful hauls were performed during 
the experiment, yielding 42 possible treatment sets 
for analyses. The number of successful treatment sets 
included in the analyses varied for each of our perfor- 
mance statistics as a result of either malfunctions or 
poor survey trawl performance caused by conflicts with 
abandoned fishing gear. 

Comparison of the means and standard deviations 
among treatments 

Trawl geometry A total of 41 treatment sets were used 
to analyze the six trawl geometry statistics (means and 



Kotwicki et al.: Effect of autotrawl systems on the performance of a survey trawl 



39 













Table 1 








Means of the trawl geometry statistics (ml and P-values of the Kruskal-Wallis 
treatments. Locked=lockeci winches; symmetry= symmetry-controlled autotrawl 


test for differences between three towing-mode 
and tension=tension-controlled autotrawl. 


Statistic 








Locked 


Symmetry 




Tension 


P-value 


Wing 


mean 






17.76 


17.77 




17.85 


0.6952 




SD 






0.69 


0.70 




0.73 


0.9157 


Door 


mean 






68.03 


67.67 




67.71 


0.8380 




SD 






1.87 


1.73 




2.14 


0.0977 


Height 


mean 






1.82 


1.81 




1.81 


0.7618 




sn 






0.21 


0.20 




0,2(1 


0.8618 













Table 2 








Means of the footrope 


and bridle 


statistics (cm) and P-va 


ues of the Kruskal-Wallit 


test for differences between three 


tow 


ing-mode 


treatments. Locked=locked winches 


symmetry: 


=symmetry-controlled autotrawl; 


and tension=tension-controlled autotrawl. 


Statistic 






Locked 




Symmetry 


Tension 




P-value 


Footrope mean 






13.98 




13.09 


12.08 




0.0441 


SD 






6.39 




5.67 


5.24 




0.0391 


symmetry 






1.95 




1.44 


1.45 




0.0554 


Bridle mean 






7.50 




6.69 


6.28 




0.0286 


SD 






4.91 




4.03 


3.37 




0.0046 


symmetry 






1.59 




0.93 


0.88 




0.0030 



standard deviations of the wing spread, door spread, 
and net height). No significant differences were detected 
among the three towing modes (Table 1); consequently, 
no environmental variables were tested for their influ- 
ence on trawl geometry. 

Footrope distance off-bottom Analyses based on 33 suc- 
cessful treatment sets produced varying results among 
the three measures describing the bottom tending perfor- 
mance of the footrope. Mean footrope distance off-bottom 
differed significantly among the three towing modes 
(Table 2, Fig. 3). Footrope distance was lowest for the 
tension treatment followed by the symmetry treatment 
and locked winches. The greatest observed difference 
occurred between the locked (13.98 cm) and the tension 
treatment (12.08 cm). Standard deviation in footrope 
distance off-bottom also differed significantly among 
treatments. Differences were similar to those observed 
for the mean, in that the lowest SD was observed for 
the tension treatment (5.24 cm) and the highest SD was 
observed for the locked treatment (6.39 cm). Because the 
variance equals the square of the standard deviation, 
this seemingly small reduction in standard deviation 
corresponds to a fairly large reduction in the variance 
(-30%). Symmetry in the footrope off-bottom distance 
did not differ significantly (P=0.0554) among the three 
towing modes. 



Bridle distance off-bottom A total of 34 successful 
treatment sets were used to analyze the bridle distance 
off-bottom data. Mean bridle distance off-bottom differed 
significantly among treatments (Table 2, Fig. 4); it was 
lowest for the tension treatment (6.28 cm) and highest 
with the winches locked (7.50 cm). Standard deviation 
of bridle distance off-bottom also differed significantly 
among the three towing modes. SD values were signifi- 
cantly lower in the autotrawl towing modes, being lowest 
in the tension treatment (3.37 cm) and highest in the 
locked treatment (4.91 cm). Bridle distance off-bottom 
was most symmetrical in the symmetry and tension 
towing modes. Symmetry and tension means were simi- 
lar (0.93 cm and 0.88 cm, respectively) and significantly 
lower than that observed for the locked-winches treat- 
ment (1.59 cm). 

Assessment of the effect of environmental factors 

Footrope distance off-bottom The effect of environmen- 
tal factors on our three measures describing the bottom 
tending performance of the footrope produced varying 
results. Mean distance off-bottom was significantly 
affected by heave only, during all treatments (Table 3, 
Fig. 5). The standard deviation of the footrope distance 
off-bottom was also significantly affected by heave, crab- 
bing, and bottom current parallel to the direction of the 



40 



Fishery Bulletin 104(1) 



15 


h 


Mean 




1.T 


^ I 


-p 




12 


- 






11 


: 








2,2 


- 


Symmetry 


2 


- 


" 


18 


'. 








i- 


1 6 


- 


J T 


1 4 


- 






1 ? 


_ 


-*- 



Figure 3 

Means and standard errors of the footrope distance off-bottom, standard deviation of the 
footrope distance off-bottom, and footrope symmetry for the locked winches (L), symmetry 
iS). and tension iT) treatments. 



8.7 


r Mean 






58 






SD 


8.3 


- 






5.4 


r 




~r 


7.9 


-p 






5 






>^ 


E 75 
u 

7.1 
6.7 
6.3 
5 9 


n 


I 




E 4.6 
u 
4.2 

3.8 

3.4 

3 








L 

i 
i 


L S 


T 








L S T 




19 
17 
15 


- 




Symmetry 








_ 




)( 






E 
o 


1.3 
1.1 
0.9 

n 7 


- 




-r 


iS 










L S 


T 










Figure 4 






Means and standard errors of the 


bridle distance off-bott 


3m, standard deviation of the 


bridle distance off-bottom, 


and bridle 


symmetry for the 


locked winches (L), symmetry (S), | 


and tension (T) treatments 













tow during some treatments (Table 3, Fig. 6). Heave 
had a greater effect on footrope SD during the locked- 
winches treatment (slope=0. 06101 than during the sym- 
metry treatment (slope = 0.0381). Similar effects were 



also detected for crabbing (slopes = 1.2355 and 0.8493, 
respectively). Current speed in the direction of the tow 
affected footrope SD only during the symmetry mode. 
Heave, crabbing, and current velocity had no effect on 



Kotwicki et al : Effect of autotrawl systems on the performance of a survey trawl 



41 



Table 3 

Multiple regression model slope coefficients with P-value <0.05. for locked winches, symmetry-controlled, and tension-controlled 
towing modes (y=bottom current parallel to the direction of the tow, AbsX=absolute value of the crosscurrent). 



Statistic 





Treatment 


Heave 


Crabbing 


y 


mean 


locked 


0.0.572 


— 







symmetry 


0.0713 


— 


— 




tension 


0.0.544 


— 


— 


SD 


locked 


0.0610 


1.2355 


— 




symmetry 


0.0381 


0.8493 


-0.0446 




tension 


— 


— 


— 


symmetry 


locked 


— 


— 


0.0255 




symmetry 


— 


— 


— 




tension 


— 


— 


— 


mean 


locked 


0.0680 


— 


— 




symmetry 


— 


— 


— 




tension 


0.0491 


— 


— 


SD 


locked 


0.0796 


0.9536 


— 




symmetry 


— 


— 


— 




tension 


0.0411 


0.3023 


— 


symmetry 


locked 


— 


— 


— 




symmetry 


— 


— 


— 




tension 


— 









AbsX 



ff2 



P-value 



Footrope 



Bndle 



— 


9.04 


0.0495 


— 


15.64 


0.0227 


— 


16.14 


0.0205 


— 


58.93 


<0.0001 


— 


33.96 


0.0066 


— 


— 


>0.05 


— 


15.42 


.0238 


— 


— 


>0.05 


— 


— 


>0.05 


— 


19.66 


0.0086 


— 


— 


>0.05 


— 


14.88 


0.0242 


— 


60.30 


<0.0001 


— 


— 


>0.05 


— 


33.83 


0.0017 


0.0556 


26.58 


0.0018 


— 


— 


>0.05 


0.0361 


13.25 


0.0343 



footrope SD during the tension mode. The sym- 
metry in footrope distance off-bottom was sig- 
nificantly affected by the current parallel to the 
direction of the tow during the locked-winches 
treatment only (Table 3, Fig. 7). 

Bridle distance off-bottom Mean bridle distance 
off-bottom was affected by heave only during 
some of the treatments (Table 3, Fig. 8). Heave 
had a greater impact on the mean bridle distance 
off-bottom during the locked-winches treatment 
(slope = 0.0680) than during the tension treat- 
ment (slope = 0.0491). The standard deviation of 
the bridle distance off-bottom was affected by 
heave and crabbing during two of the treatments 
(Table 3, Fig. 9). Heave had an effect on the SD of 
the bridle distance off-bottom during both locked 
(slope = 0.0796) and tension (slope = 0.0411) treat- 
ments. A significant effect due to crabbing was 
also detected during these same two treatments 
(slopes 0.9536 and 0.3023, respectively). During 
the symmetry mode no effect due to any of the envi- 
ronmental conditions was detected, even though 
SD was almost always higher than that observed in 
the tension mode, with the exception of extreme heave 
conditions. Symmetry in the bridle distance off-bottom 
was affected only by crosscurrent (Table 3, Fig. 10). 
Crosscurrent had an effect on the symmetry during 
the locked {slope = 0.0556) and tension (slope=0.0361) 
treatments. 



00 


- 








* AL □ S oT 








^ Q •- 


1H 


- 














14 


- 




° /^ !3-^*^^^^r ....•■■•• ■" 






,-.-iA- 




10 


r-."-"- 


"-'< 













6 


-, 


, 


. 







20 



40 



60 80 



100 



Heave (cm) 
Figure 5 

Comparison of regression lines illustrating the relationship 
between the mean footrope distance off-bottom and heave for 
locked winches (L), symmetry (S), and tension (T) treatments. 
Treatments in which heave had a statistically significant effect 
on the mean are identified with an asterisk (*). 



Discussion 

The objective of this study was to identify towing modes 
that could potentially reduce variance in the catchability 
of the 83-112 Eastern bottom trawl among hauls. Three 
separate aspects of trawl performance affecting catch- 
ability were investigated: trawl geometry, bottom tend- 



42 



Fishery Bulletin 104(1) 




40 60 

Heave (cm) 

15 
12 



Q 
CO 



oa 



Aft) 



-8-9° 



0° 



ifipt 



0% 



I il Oo 



On o 



ti_^ 



-30 



-20 



-10 







10 



20 



30 



Current parallel to tow direction (cm/s) 

Figure 6 

Comparison of regression lines illustrating the relationship between the standard deviation (SD) of the footrope distance 
off-bottom and environmental variables for locked winches (L), symmetry (S), and tension (T) treatments. Treatments in 
which the environmental variable had a statistically significant effect on the SD are identified with an asterisk (*). 





5 

4 























AL 


DS 


oT 1 








a 


A 









































f 














D 


*A 








A 










C) 


































>. 


3 


- 








D 








o 




A 


a 




A 




(1) 






















A 


D 




r+J 




. L 


(- 
























__ — 








F 


■^ 


- 


A 










9 n 









O 
















■f^ 








_A_£__-- 






AO 












rn 




; A 


Q- 


3 


^ 


A 


""^^"a^ 


-^i^,e. 


_^ 


A 








□ 




- T 










1 


- 




A 





8 CO 
o A* 


D 







^AAQ 

o 


o 
in 
o 


-- 4- 

D 

4j 










a 







- 












, . , , 








, 




D 







-30 



-20 



-10 







10 



20 



30 



Current parallel to tow direction (cm/s) 

Figure 7 

Comparison of regression lines illustrating the relationship 
between the symmetry of the footrope distance off-bottom and 
current parallel to the direction of the tow for locked winches 
(L), symmetry (S), and tension IT) treatments. Treatments 
in which the current had a statistically significant effect on 
the symmetry are identified with an asterisk (*). 



ing performance of the footrope, and bottom-tending 
performance of the lower bridles. 

Trawl geometry 

Because wing and door spread and net height vari- 
ability all influence catchability of a trawl (Rose and 
Nunnallee, 1998), surveys would stand to benefit 
from autotrawl systems if variances in the trawl 
geometry features were reduced. Our study showed 
that neither autotrawl system improved trawl geom- 
etry over the conventional locked-winches towing 
mode. 

Footrope 

Previous studies have demonstrated that escape- 
ment under the footrope is an important factor 
determining the catchability of the 83-112 Eastern 
survey trawl (Somerton and Otto 1999; Munro and 
Somerton, 2002; Weinberg et al., 2004). Periodic 
separation between the footrope and the bottom 
contribute to variability in catchability and have 



Kotwicki et al : Effect of autotrawl systems on ttie performance of a survey trawl 



43 



been shown to be influenced by a variety of factors, 
such as net speed through the water (Somerton 
and Weinberg, 2001) that can increase footrope 
distances off-bottom. Weinberg et al. (2002), using 
a different survey bottom trawl, found capture 
probability of some benthic species to decrease as 
a result of footrope separations with the bottom. 
These findings likely apply to the 83-112 Eastern 
trawl as well. Because capture probability is gener- 
ally size dependent (Munro and Somerton, 2002), 
an unstable footrope may not only increase the 
variance of overall biomass estimates, but may bias 
the size distribution data generated by a survey 
as well. If autotrawl systems can help maintain 
footrope contact with the bottom, then the use of 
an autotrawl during surveys may be warranted. 

In our experiment, the tension-controlled auto- 
trawl system provided the best footrope contact 
overall and the lowest standard deviation (Fig 3). 
High heave and moderate to strong crabbing were 
largely responsible for the differences observed in 
the bottom-tending performance of the footrope 
among treatments (Figs. 5-7). The tension treat- 
ment was most effective in counteracting the ef- 
fects of environmental conditions, whereas the 
locked-winches treatment was the least stable of 
the three treatments given the changing environ- 
mental conditions. 

We found both autotrawl systems have a po- 
tential to improve bottom trawl survey biomass 
estimates by increasing the dynamic stability of 
the trawl, thereby reducing the variance in its 
catchability. The equal-tension system proved bet- 
ter than the symmetry system in improving the 
overall stability of the footrope bottom-tending 
performance given the low bottom current ve- 
locities observed. Symmetry autotrawl systems, 
designed to react to crosscurrent conditions at 
towing depth, could prove to be more effective 
under stronger current conditions, but during our 
experiment, current velocities may have been too 
low (<35 cm/s) for us to be able to detect a differ- 
ence between the symmetry and tension towing 
modes. 

Bridles 

Flatfish are stimulated to herd by close proximity 
to or actual contact with the lower bridle (Main 
and Sangster, 1981a), whereas semipelagic species 
such as Atlantic cod (Gadus morhua) are probably 
stimulated to herd by the sight of the otter doors 
and mud clouds created by the doors and lower bri- 
dles (Main and Sangster, 1981b). For this reason, 
the length of the lower bridle in contact with the 
bottom and the frequency of this contact affect the 
herding efficiency of the trawl (Somerton, 2003). If 
the among- and within-tow variability of the bridle 
bottom-tending distance could be reduced by using 
an autotrawl system, then standardizing survey 



16 



12 



AL dS ot 




A. S 







10 



40 



60 



80 



100 



Heave (cm/s) 

Figure 8 

Comparison of regression lines illustrating the relationship 
between mean bridle distance off-bottom and heave for locked 
winches (L), symmetry (Si, and tension (T) treatments. Treat- 
ments in which the heave had a statistically significant effect 
on the mean are identified with an asterisk (*). 



2 






b. 


1 '^'- 


DS 


OT 



















A 


A 


.^ 


8 
6 




D 


a A '^ D 


o 
o 


s- 






- 


□ 




a 
□ 


T 




4 


A 

O 
O 




2 


- 






















E 

u 
Q 



20 



40 60 

Heave (cm) 



80 



100 




1,5 2 

Crabbing (cm) 

Figure 9 

Comparison of regression lines illustrating the relationship 
between standard deviation (SD) in bridle distance off-bottom 
and environmental variables for locked winches (L), symme- 
try (S), and tension (T) treatments. Treatments in which the 
environmental variable had a statistically significant effect 
on the SD are identified with an asterisk (*). 



44 



Fishery Bulletin 104(1) 



5 










o 
a 

A 


AL as 


OT 1 


4 


a 




3 


- 


6 






A 
O 

A 

A D 


i 


^^— L' 


2 




O 
D □ 




o 


O 
D 

D 

- 


Q,,----^ 

n M o O 


a 


T 

S 


1 



o ° ° 

D 





10 20 30 

Crosscurrent (cm/s) 



40 



Figure 10 

Comparison of regression lines illustrating the relation- 
ship between bridle symmetry and crosscurrent for locked 
winches (L). symmetry (S), and tension (T) treatments. 
Treatments in which crosscurrent had a statistically sig- 
nificant effect on bridle symmetry are identified with an 
asterisk (*). 



procedures to include towing with an autotrawl system 
would likely be beneficial. 

Our results showed that both autotrawl systems re- 
duced the mean distance of the bridles off-bottom and 
the standard deviation of this distance and increased 
the symmetry between bridles (Fig. 4) over the locked- 
winches treatment. Our experiment also demonstrated 
that both autotrawl systems increased the stability of 
lower bridle bottom contact in changing environmental 
conditions, but we were unable to discern which of the 
systems was better. The data indicate that bridle bot- 
tom-tending performance was the least affected by en- 
vironmental conditions during the symmetry treatment; 
however the standard deviation in the bridle distance 
off-bottom was almost always lower during the tension 
treatment, the exception being under extreme heave 
conditions (Fig. 9). The locked-winches treatment had 
the highest standard deviation and was affected the 
most by heave, crabbing, and crosscurrent velocity. 

Autotrawl systems counteract the effects of warp 
length differential created by trawl crabbing. During 
our locked-winches treatment crabbing likely caused 
unequal tension in the warps similar to that seen while 
towing straight behind the boat with unequal warp 
lengths. Weinberg and Somerton (2006) reported that 
warp tension changes significantly with offset in warp 
lengths. To compensate for the crabbing angle and to 
assure that the trawl is pulled square to the direction 
of the tow during crabbing, one warp should be shorter 
than the other. Tension-controlled autotrawl systems ad- 
just the length of the warps to equalize the tension on 
them. Symmetry-controlled autotrawl systems change 
the length of the warps to minimize crosscurrent and 
also increase the symmetry of the trawl in relation to 
the tow direction. 



In summary, autotrawl systems proved to be effective 
in decreasing some of the adverse effects of environ- 
mental factors on some aspects of the 83-112 bottom 
trawl performance and, as a result, have the potential 
to reduce variance in among-haul catchability of the 
survey trawl. For this trawl, footrope and bridle dis- 
tances off-bottom were significantly different among the 
three towing modes, albeit the differences in actual dis- 
tances off-bottom were small. Trawls deploying heavier 
groundgear, thus enabling more constant contact with 
the bottom, may not be as affected. Further investiga- 
tions are needed to assess the effects of the two types of 
autotrawl systems on other types of survey trawl gear, 
such as high-opening bottom trawls, trawls using differ- 
ent footropes, shrimp trawls, and midwater trawls. The 
effect of using autotrawls in areas with stronger current 
and different depths also needs to be investigated for 
the different types of trawl gear. 

For bottom trawl surveys, we are concerned with 
two potential shortcomings in the symmetry-controlled 
autotrawl system tested. First, videos of trawling in 
rough terrain have revealed footrope distortion occur- 
ring when the footrope or doors snag on the bottom 
(Weinberg, unpublished data). Typically, the side op- 
posite the snag is pulled forward while the snagged 
side remains stationary or is pulled ahead at a slower 
pace. This uneven pull on the net causes asymmetry in 
the headrope shape, by skewing it from the general tow 
direction (Weinberg and Somerton, 2006). Distortion 
of the headrope introduces error in the current direc- 
tion and velocity values obtained by the current sensor 
mounted on the headrope from which warp length is 
determined, and thus impacts trawl performance. Our 
second concern involves the overall warp adjustment 
period required for the trawl to equilibrate to current 



Kotwicki et al,: Effect of autotrawl systems on the performance of a survey trawl 



45 



direction and velocity based on the current sensor in- 
put. Because many surveys standardize to a 15-min 
tow duration, a proportionally large percentage of tow 
time may be involved with adjusting warp lengths in 
pursuit of optimal symmetry and therefore vary trawl 
catchability during the tow. Under commercial trawling 
conditions, the manufacturer of the symmetry winch 
control system recommends using a 2-min minimum 
signal collection and analysis period between warp ad- 
justments. After each adjustment, new sensor signals 
would be received and evaluated, followed by another 
warp adjustment, if necessary. We are uncertain as 
to how many warp adjustments may be necessary to 
orient the trawl in relation to crosscurrent flow, but 
based on our experiences and the frequency with which 
warp lengths changed during our experiment, it is con- 
ceivable that several adjustment periods spanning the 
majority of the survey tow may be necessary, leaving 
minimal time for the net to actually fish symmetrically. 
Furthermore, each 2-min warp adjustment would be 
based on a maximum of only five current flow readings 
because the current sensor refreshes data at 24-s inter- 
vals. Should signal loss occur, then fewer data points 
would be available. 

We recommend taking a cautious approach before 
switching a trawl survey from locked winches to auto- 
trawl. A change in survey method may require an ex- 
tensive calibration experiment between the two trawl- 
ing methods in order to maintain the continuity of a 
survey time series. Furthermore, autotrawl systems, 
like other mechanical devices, require service, appro- 
priate inspections and periodic testing to ensure that 
they are functioning correctly within manufacturer 
specifications. Autotrawl calibration parameters are 
dependent upon accurate measurements of the diameter 
and length of each winch drum, warp diameter, and 
construction (e.g., compacted vs. traditional or wire 
core vs. fiber core), layers of warp, and the number of 
windings per layer on each drum. Hydraulic pumps 
and lines, electric motors, valves, solenoids, computer- 
ized control panels, and geometric counters must all be 
inspected to assure proper operation. Of course, should 
surveys operate with a symmetry-style autotrawl sys- 
tem, then all of the above procedures would hold true 
in addition to the need for accurate calibration of the 
current sensor and the proper mounting of the sensor 
to the headrope. 



Acknowledgments 

The success of this project is due to the efforts of many. 
We extend our gratitude to Tim Cosgrove, Brad Lougheed, 
and the crew of the FV Vesteraalen. Our knowledge of 



winch control systems was greatly enhanced by the guid- 
ance of Doug Dixon and Ed Ramberg, as was our ability 
to communicate at-sea with a wide range of instru- 
ments thanks to Scott Furnish, David Roetcisoender, 
Jon French, and Dennis Benjamin. We are also grateful 
to our reviewers including Captain John Gruver, Mark 
Wilkins, Gary Walters, and Peter Munro. 



Literature cited 

Main, J., and G. I. Sangster. 

1981a. A study of the sand clouds produced by trawl 
boards and their possible effect on fish capture. Scott. 
Fish. Res. Rep. 20:1-20. 
1981b. A study of fish capture process in a bottom trawl 
by the direct observations from a towed underwater 
vehicle. Scott. Fish. Res. Rep. 23:1-23. 
Munro, P. T., and D. A. Somerton. 

2002. Estimating net efficiency of a survey trawl for 
flatfishes. Fish. Res. 55:267-279. 

Neter, J., M. H. Kutner, C. J. Nachtsheim, and W. Wasserman. 
1996. Applied linear regression models, third ed., 720 
p. Irwin, Chicago, IL. 
Rose, C. S., and E. P. Nunnallee. 

1998. A study of changes in groundfish trawl catch- 
ing efficiency due to differences in operating width, 
and measures to reduce width variation. Fish. Res. 
36:139-147. 

Somerton. D. A. 

2003. Bridle efficiency of a survey trawl for flatfish: 
measuring the length of the bridles in contact with the 
bottom. Fish. Res. 60:273-279. 

Somerton D. A., and P. T. Munro. 

2001. Bridle efficiency of a survey trawl for flatfish. Fish. 
Bull. 99:641-652. 
Somerton, D.A.. and R. S. Otto. 

1999. Net efficiency of a survey trawl for snow crab, 
Chionoecetes opilio, and Tanner crab, C. bairdi. Fish. 
Bull. 97:617-625. 

Somerton, D.A., and K. L. Weinberg. 

2001. The affect of speed through the water on footrope 
contact of a survey trawl. Fish. Res. 53:1724. 

Stauffer, G. 

2004. NOAA protocols for groundfish bottom trawl sur- 
veys of the nation's fishery resources. NOAA Tech. 
Memo. NMFS-F/SPO-65, 205 p. Alaska Fish. Sci. 
Center, 7600 Sand Point Way NE, Seattle, WA 98115. 

Weinberg, K. L., R. S. Otto, and D. A. Somerton. 

2004. Capture probability of a survey trawl for red 
king crab iParalithodes camtschaticus). Fish. Bull. 
102:740-749. 
Weinberg, K. L., and D. A. Somerton. 

2006. Variation in trawl geometry due to unequal warp 
length. Fish. Bull. 104:21-34. 
Weinberg, K. L., D. A. Somerton. and P. T Munro. 

2002. The effect of trawl speed on the footrope capture 
efficiency of a survey trawl. Fish. Res. 58:303-313. 



46 



Abstract — The California market 
squid iLoligo opalescens) has been 
harvested since the 1860s and it 
has become the largest fishery in 
California in terms of tonnage and 
dollars since 1993. The fishery began 
in Monterey Bay and then shifted to 
southern California, where effort has 
increased steadily since 1983. The 
California Department of Fish and 
Game (CDFGl collects information 
on landings of squid, including ton- 
nage, location, and date of capture. 
We compared landings data gathered 
by CDFG with sea surface tempera- 
ture (SST), upwelling index (UI). the 
southern oscillation index (SOI), and 
their respective anomalies. We found 
that the squid fishery in Monterey 
Bay expends twice the effort of that 
in southern California. Squid land- 
ings decreased substantially follow- 
ing large El Nino events in 1982-83 
and 1997-98, but not following the 
smaller El Niiio events of 1987 and 
1992. Spectral analysis revealed 
autocorrelation at annual and 4.5- 
year intervals (similar to the time 
period between El Nino cycles). But 
this analysis did not reveal any 
fortnightly or monthly spawning 
peaks, thus squid spawning did not 
correlate with tides. A paralarvae 
density index (PDI) for February 
correlated well with catch per unit 
of effort (CPUE) for the following 
November recruitment of adults to 
the spawning grounds. This stock- 
recruitment analysis was significant 
for 2000-03 iCPUE = 8.42 + 0.4lPDI. 
adjusted coefficient of determina- 
tion, r2 = 0.978, P=0.0074). Surveys 
of squid paralarvae explained 97. 89^ 
of the variance for catches of adult 
squid nine months later. The regres- 
sion of CPUE on PDI could be used to 
manage the fishery. Catch limits for 
the fishery could be set on the basis 
of paralarvae abundance surveyed 
nine months earlier. 



The fishery for California market squid 
{Loiigo opalescens) (Cephalopoda: Myopsida), 
from 1981 through 2003 



Louis D. Zeidberg 

Monterey Bay Aquarium Research Institute 
7700 Sandholdt Rd. 
Moss Landing, California 95039-9644 
E-mail address: zelo g mbari org 

William M. Hamner 

Dept. ol Ecology and Evolutionary Biology 

Univ, California, Los Angeles 

621 Charles E. Young Drive South 

Box 951606 

Los Angeles, California 90095-1606 

Nikolay P. Nezlin 

Southern California Coastal Water Research Proiect 

7171 Fenwick Lane 

Westminster, California 92683-5218 



Annette Henry 

California Department of Fish and Game 
Southwest Fisheries Science Center 
Marine Region, La Jolla Field Office 
8604 La Jolla Shores Drive 
La Jolla, California 92037 



Manuscript submitted 7 May 2004 
to the Scientific Editor's Office. 

Manuscript approved for publication 
20 June 2005 by the Scientific Editor 

Fish. Bull. 104:46-.59 12006). 



The recent discovery of falsification in 
Chinese fisheries reporting has led to 
the realization that the majority of the 
world's fisheries surpassed sustain- 
ability in 1988 (Watson and Pauly, 
2001). The food chain has been fished 
down by removal of apex predators 
like swordfish and snapper beyond 
sustainability, and fisheries have 
subsequently shifted to prey items 
like sardine and mackerel (Pauly et 
al., 1998). We have reached the point 
where cephalopods are regularly the 
largest biomass of all commercial spe- 
cies harvested. Since 1970, groundfish 
landings of flounders, cods, and had- 
docks have either decreased or main- 
tained their levels while landings in 
cephalopod fisheries have increased 
(Caddy and Rodhouse, 1998). Some of 
the larger cephalopod landings may be 
due to increased demand, but lower 
levels of predation and competition 
from finfish, and the shorter lifespan 



of squid may have allowed cephalopods 
to increase in abundance worldwide. 
Loiigo opalescens is a small squid 
(130 mm mantle length) that occupies 
the middle trophic level in Califor- 
nia waters, and it may be the state's 
most important forage species. Mar- 
ket squid are principal forage items 
for a minimum of 19 species of fishes, 
13 species of birds, and six species of 
mammals (Morejohn et al., 1978). The 
effective management of this fishery 
is of paramount importance not only 
to the fishermen involved but also to 
the millions of fishes, birds, and mam- 
mals that compete for this resource. 
Because cephalopods eat mostly zoo- 
plankton (Loukashkin, 1976), if we 
also deplete the squid population, it 
is not clear how oceanic food chains 
will respond. If the subannual popu- 
lation of L. opalescens fails to recruit 
a large biomass in a given year, the 
long-lived predators of this species in 



Zeidberg et al , The fishery for Loligo opalescens from 1981 through 2003 



47 



the California Current may encounter severe metabolic 
stress. 

Since the decline of the anchovy fishery, market squid 
probably constitutes the largest biomass of any single 
marketable species in the coastal environment of Cali- 
fornia (Rogers-Bennett. 2000). In the 1999-2000 season, 
fishermen landed 105,005 metric tons of California mar- 
ket squid [Loligo opalescens) with an exvessel (whole- 
sale) revenue of $.36 million (California Department of 
Fish and Game [CDFGJ landing receipts). These squid 
deposit egg capsules on sandy substrates at depths of 
15-50 m in Monterey Bay (Zeidberg et al., 2004) and 
20-90 m in the Southern California Bight. The majority 
of squid landings occur around the California Channel 
Islands, from Pt. Dume to Santa Monica Bay, and in 
southern Monterey Bay. The fishery comprises chiefly 
light-boats with high wattage illumination to attract 
and aggregate spawning squid to the surface, and seine 
vessels that net the squid (Vojkovich, 1998). 

Management to date has followed methods that are 
not dependent upon an estimate of population abun- 
dance because no estimate of squid biomass exists. In 
addition to limiting the catch and the number of ves- 
sels, management of the fishery has included weekend 
closures north of Point Conception since 1983, and these 
closures have recently extended to all of California 
coastal waters. This regulation is designed to allow 
a 48-hour period each week for undisturbed spawn- 
ing. For Monterey Bay, the weekend closure resulted 
in highest landings on Mondays and decreasing daily 
landings through Friday (Leos, 1998). Since 2000, light 
boat and seine vessel operators have been required to 
complete logbooks for CDFG, such that CPUE can be 
estimated from data on the cumulative effort required 
to land squid. 

Because of their short lifespan, many squid popula- 
tions have been more effectively correlated with local 
oceanographic conditions than have pelagic fish spe- 
cies with life spans of 4-8 years. Squid landings from 
all regions of the world fluctuate in conjunction with 
the temperatures of the previous season. Mclnnis and 
Broenkow (1978) found positive temperature anomalies 
preceded good Loligo opalescens landings by 18 months, 
and poor squid catches followed periods of anomalous 
low temperatures in Monterey Bay. Robin and Denis 
(1999) found similar results. Warmer waters (mild win- 
ters) were followed by increased cohort success for Lo- 
ligo forbesi in the English Channel, but this effect was 
not constant throughout the year. Conversely, Roberts 
and Sauer (1994) found Loligo vulgaris reynaudii land- 
ings in South Africa to increase with upwelling that 
coincided with La Nina (cold water) conditions in the 
equatorial Pacific. Rocha et al. (1999) also found an 
increase in squid paralarvae of many species during 
upwelling conditions on the Galacian-coast. 

Modern instruments for monitoring coastal ocean con- 
ditions, including weather buoys and satellites, provide 
a vast amount of information on the physical environ- 
ment of fish and squid populations. The correlation 
between cold, upwelled nutrient-rich water at the sea 



surface resulting from Eckman transport and phyto- 
plankton blooms a few days later is well established 
(Nezlin and Li, 2003). Mesoscale eddies generated by 
coastal processes and islands also serve to concentrate 
phytoplankton (Falkowski et al., 1991; Aristegui et al., 
1997; DiGiacomo and Holt, 2001). The subsequent effect 
upon zooplankton grazers rapidly follows the cycles of 
upwelling and relaxation (Wing et al., 1995; Graham 
and Largier, 1997; Hernandez-Trujillo, 1999). 

Waluda et al. (1999) found that the CPUE for the II- 
lex argentinus fishery was not related to monthly local 
sea surface temperature (SST), but CPUE was inversely 
related to SST on the hatching grounds for the previous 
July, when hatchlings were in their exponential growth 
phase (Yang et al., 1986; Grist and des Clers, 1998). 
The largest catches followed cold water. Waluda et al. 
(2001) found a large CPUE when the Brazilian Current 
dominated and frontal waters diminished in the location 
where squid hatching occurs. Agnew et al. (2000, 2002) 
found that CPUE for Loligo gahi was inversely corre- 
lated with SST for hatching areas six months earlier. 
Sakurai et al. (2000) found that Todarodes pacificiis 
CPUE was highest following periods when there were 
large regions of hatchling-favorable habitat (17-23°C 
waters). They found a positive correlation between the 
density of paralarvae and the catch per unit of effort 
(CPUE) of adults in the same year (r~ = 0.91) and also 
in the CPUE of the following year (r2 = 0.77). 

The CDFG has an extensive database of landings 
data from 1981 to the present for market squid. Be- 
cause there is no record of effort prior to 2000 and be- 
cause the market is driven by demand, it is difficult to 
use landings and vessel-day data to calculate a CPUE 
and therefore estimate biomass. Fishermen report that 
even if squid are available, they may not be harvested 
when processors are not accepting squid (Brockman^). 
However, there is no other database as large and wide- 
spread temporally and spatially as fishery data. Even 
though there are no data recorded when boats attempt 
to catch squid and fail, we can still use landings and 
vessel-days to create a CPUE. This CPUE therefore is 
not a methodically collected estimate of biomass, but 
is still a robust enough estimate of abundance to draw 
preliminary conclusions as we wait for logbook data to 
accumulate. 

It is important to determine the effects of the envi- 
ronment on the California market squid fishery so that 
we can predict future landings from present conditions. 
Our investigation uses California market squid landings 
for 1981-2003 to examine correlations of landings and 
CPUE with physical oceanography. We compare land- 
ings data (time, location, vessel-days, and landings [in 
pounds]) to sea surface temperature (SST), upwelling 
index (UI), the Southern Oscillation index (SOI), the in- 
dex of sea surface temperature in the eastern equatorial 
tropcial Pacific NIN03, and their respective anomalies. 
We also compare CPUE to a paralarvae density index 



Brockman, D. 2002. Personal commun. Davie.s Locker 
Sportfishing, 400 Main St. Newport Beach, CA 92611. 



48 



Fishery Bulletin 104(1) 



(PDI) based upon distributions determined 
in the Southern California Bight (Zeidberg 
and Hamner, 2002). 



Materials and methods 



35° - 







-125° 



The CDFG database for commercial Cali- 
fornia market squid landings from 1981 to 
present includes weight, date, location (based 
on CDFG 10x10 nm blocks), and gear type. 
Accounting for general physical oceanographic 
properties (Harms and Winant, 1998; Bray 
et al., 1999: Brink et al., 2000; Hickey et 
al., 2003) and following our previous stud- 
ies (Nezlin et al., 2005), we organized the 
landings data into six areas to look at subtle 
differences between them: MB = northern 
coastal (because the majority of the land- 
ings in this area occur in southern Monterey 
Bay), CC=central coastal, SB = Santa Barbara 
Channel, SCB = Southern California Bight, 
SM= Santa Monica, and SD = San Diego. Also 
we grouped the fishery into two larger regions 
April (APR, equal to MB above) and October 
(OCT, a combination of the other five areas) 
based upon the month of greatest recruit- 
ment (Fig. 1). For the purpose of our study, 
recruitment is the aggregation of reproductive 
adults on the spawning grounds. When CDFG 
reports squid data, they make a distinction 
at Point Conception, thus our MB and CC 
areas are grouped as the "north" and our SB, 
SCB, SM, and SD are named "south." For this 
fishery we defined CPUE as the recorded tons 
landed in a day, divided by the number of 
seine vessels that landed these squid. Those 
days in which there were no landings were 
assigned a value of zero. This CPUE is impor- 
tant because, although not truly a quantifying effort, it 
does provide a means for estimating the abundance of 
squid by providing some basis for the amount of time 
taken to make a landing. Lampara, brail, and light boat 
data were not included because of increased variability in 
landings and effort and the fact that these vessels have 
dwindled from ten to zero percent since 1981. 

The landings and boat data for each area were summed 
for each block by day. For example, assume that on a 
particular day fishermen caught 10,000 metric tons with 
four boats in the area of SM, 18,000 tons from three 
boats in SCB, and 12,000 tons from three boats in SB. 
We would calculate a CPUE of 2500 tons/vessel-day in 
SM, 6000 tons/VD in SCB, and 4000 tons/VD in SB, 
respectively. Thus for every date for which there was a 
landing we were able to calculate CPUE value for each 
area. Until 2002, there had never been a landing in Mon- 
terey in January, when one vessel captured 75 tons. Data 
such as this produce misleadingly high CPUEs; therefore 
all months with less than seven vessel-days for the en- 
tire 22-year period were removed from the analysis. 



April 

recruitment 

(APR) 



October 

recruitment 

(OCT) 



km 

--F= 




-120 = 



Figure 1 

The California coast with the fishery areas for the California market 
squid (Loligo opalescens) identified. Areas were classified according 
to physical oceanographic features: Northern Coast (MB), Central 
Coast (CC), Santa Barbara Channel (SB), Southern California Bight 
(SCB), Santa Monica Bay (SM), and San Diego (SD). Regions were 
also classified by fishery recruitment month: April recruiting (APR: 
same as MB area) and October recruiting (OCT: CC, SB, SCB, SM, 
and SD areas combined). Block 526 (indicated with a slender arrow) 
is where the majority of the MB-APR landings occur. Shaded area 
indicates the location of the paired-net surveys used to generate the 
paralarvae density index (PDI). 



Physical oceanography data were gathered from the In- 
ternet for sea surface temperature (SST),'- upwelling in- 
dex (UI),'^ southern oscillation index (SOI),^ and NIN03.5 
Upwelling index (UI) is an Ekman offshore water trans- 
port (mVs per 100 m of the coastline) estimated from 
fields of atmospheric pressure (Bakun, 1973). Southern 
Oscillation index (SOI) is the difference between the 
standardized measurements of the sea level atmospheric 



- NOAA (National Oceanic and Atmospheric Administration). 
National Data Buoy Center. 2000. Website: http://facs. 
scripps.edu/surf/buoys.html [Accessed on 24 April 2003.1 

^ NOAA Pacific Fisheries Environmental Laboratory. 2003. 
Website: http://www.pfeg.noaa.gov/products/las.html 
[Accessed on 20 March 2003.1 

* Australian Government Bureau of Meteorology. 2005. 
Website: http://www.bom.gov.au/climate/current/soihtml. 
shtml [Accessed on 29 March 2003.1 

^ IRI (International Research Institute for Climate Pre- 
diction). 2005. Website: http://ingrid.ldgo.columbia.edu/ 
SOURCES/.Indices/.nino/.EXTENDED/.NIN03/ [Accessed 
on 20 April 2004.1 



Zeidberg et a\ The fishery for Loligo opalescens from 1981 through 2003 



49 











Table 1 














Totals of vessel-days, landi 


ngs (metric tons), and CPUE (tons/vessel-day) for two ti 


me per 


iods 


of the Califor 


nia market squid 


{Loligo opalescens) fishery (1981- 


2003). The majority of the fishery occurred in Monterey Bay, 


1981-89, and 


in southern Cali- 


fornia. 1990 


-2003. The six 


smal 


areas represent 


physical oceanographic features and the 


larger regions 


(APR and OCT) are | 


grouped by the month of greatest 


recruitment of t 


pawning adults to the fishery (see 


bolder border in Fig. 


1). 


MB and APR are 


synonymous 


terms. MB = northern coastal area, pi 


edominantly southern 


Monterey B 


ay; CC 


=central coast 


SB 


= Santa Barbara; 


SCB = Southern California Bight; 


SM = Santa Monica; and SD = San Diego 


VD=ve.ssel 


days. 












Region 




Vessel-days 




Landing 




Percent landings 


CPUE 


Years 


and area 




(VD) 


Percent VD 


(tons) 






(tons) 




(tons/VD) 


1981-89 


APR 






















MB 




4918 


87.4 


27242 






66.5 




5.5 




OCT 






















CC 




38 


0.7 


1040 






2.5 




27.4 




SB 




186 


3.3 


3811 






9.3 




20.5 




SCB 




169 


3.0 


3475 






8.5 




20.6 




SM 




122 


2.2 


1811 






4.4 




14.8 




SD 




194 


3.4 


3597 






8.8 




18.5 


Subtotal 






709 


12.6 


13735 






33.5 




19.4 


Total 






5627 




40977 










7.3 


1990-2003 


APR 






















MB 




6508 


22.5 


92323 






13.6 




14.2 




OCT 






















CC 




1283 


4.4 


33964 






5.0 




26.5 




SB 




3822 


13.2 


104172 






15.3 




27.3 




SCB 




8037 


27.7 


212986 






31.3 




26.5 




SM 




4408 


15.2 


113569 






16.7 




25.8 




SD 




4918 


17.0 


124147 






18.2 




25.2 


Subtotal 






22468 


77.5 


588838 






86.4 




26.2 


Total 






28976 




681161 










23.5 



pressure in Tahiti and Darwin. NIN03 is determined by 
averaging the SST anomalies over the eastern tropical 
Pacific (5°S-5°N; 150°W-90°W). The buoys used were 
the Monterey buoy (46042, 36°N 122°W) for the MB 
region, the east Santa Barbara buoy (46053, 34.24°N 
119.85°W) for the SB region, and the Santa Monica buoy 
(46025, 33°N 119°W) for the remaining regions. 

We performed a spectral analysis of the entire time 
series to look for significant periodicities in the daily 
data for the entire 22-year data set. CPUE values were 
natural log-transformed and smoothed with a Parzen 
window (Ravier and Fromentin, 2001). We used a time 
series analysis method of cross correlation to deter- 
mine the lag period in months between CPUE and the 
physical features of SST, SOI, NIN03, and UI and 
their anomalies from averaged seasonal cycles. Using 
this lag period we calculated linear regression of the 
CPUE from SST. 

Sea surface temperature (SST) time series was ob- 
tained from infrared satellite measurements with ad- 
vanced very high resolution radiometers (AVHRRs) 
on National Oceanic and Atmospheric Administration 
(NOAA) meteorological satellites. The data were pro- 
cessed at the University of Miami's Rosenstiel School 



of Marine and Atmospheric Science (RSMAS) and the 
NOAA National Oceanographic Data Center (NODC) 
within the scope of Pathfinder Project (version 4.1, 
available from the Jet Propulsion Laboratory Physical 
Oceanography Distributed Active Archive Center [JPL 
PO DAAC]).!' 

We performed a stock-recruitment analysis from a 
paralarvae density index (PDI). Paralarvae were col- 
lected with paired nets (505-f<m mesh) without bridles, 
deployed like bongo-nets, and towed in a double oblique 
mode to 100 m depth. Samples were taken in February 
from 1999 to 2003, every 7.5 km on transects in regions 
SCB and SM (Zeidberg and Hamner, 2002). Flow meters 
were used to standardize the number of paralarvae per 
1000 m'^ of water. The PDI is the average number of 
paralarvae/1000 m^ from all tows. We used linear re- 
gression to compare the February PDI with the CPUE 
for the large November adult recruitment event in the 
SCB and SM regions of the same year. 

Statistics were performed with Statview 3.0 (Abacus 
Concepts, Berkeley, CA) or Statistica 6.0 (Statsoft, Tul- 
sa, OK). Interpretations of ^test, regression, ANOVA, 



http;//podaac.jpl. nasa.gov/ [Accessed on 15 March 2003.] 



50 



Fishery Bulletin 104(1) 



spectral analysis, and cross correlation time series were 
made in accordance with Zar (1984). 



Results 

Decadal-regional analysis 

The 22-year fishery data for Loligo opalescens were 
divided into two periods: 1981-89 and 1990-2003 
(Table 1) due to a southward shift in the fishery after 
1989. For the first period (1981-89), 87% of the effort 
and 66% of the landings were predominantly focused 
in the MB or APR region, specifically in the southern 
portion of Monterey Bay. The amount of squid captured 
in 1981 and 1982 was not matched again in Monterey 
Bay until 2002. The MB region was the most focused 
region; 62% of the total catch and 83% of the CPUE 
came from a very small area (block 526, Fig. 1) — ^just 
off Monterey harbor. CPUE in this region and time 
period was low, 5.54 tons/vessel-day (Table 1). SM had 
4.43% of the landings and 2.2% of the vessels, yielding 



S 120 

100- 
i 80- 

1 60 
§ 40 
f 20 
-> 



.0 



J: Q ii - i - . l .., - :^^ ~<t- 



1981 1983 1985 1987 1989 1991 



1993 1995 1997 1999 2001 2003 




1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 
Strong Week Week Strong Strong 



El Nino 
Year 



El Nino La Nina 



Figure 2 

Fi.shery data for Loligo opalescens, summed by year for 1981-2003. (A) 
landings, (B) vessel days, and (Cl catch per unit of effort (CPUEi by 
year for Monterey Bay (April [APRI— black circles) and southern Cali- 
fornia (October [OCT!— unfilled circles) regions. Scale ofy-axis changes 
between A, B, and C. Note the lack of landings in OCT in 1984 and APR 
in 1998. following strong El Nino events. 



a CPUE of 14.85 tons/vessel-day. The SB, SCB, and SD 
areas were similar in the 1980s with landings around 
9%, vessels at 3%, and CPUEs varying from 18.5 to 
20.5 tons/vessel-days. The CC area had the smallest 
percentage of landings (2.5%) and vessels (0.7%), but the 
highest CPUE (27.4%) — most likely due to squid being 
hauled as a secondary target species in this region. 
Few fishermen choose to harvest squid in the central 
California area because of rough seas and rocky, gear- 
fouling ocean bottoms. 

The focus of the market squid fishery shifted to 
southern California in the 1980s and landings sur- 
passed those of Monterey Bay in 1990 (Fig. 2, Table 1). 
Whereas vessel-days per year decreased by 17.6% in 
the MB area, vessel-days increased 20-fold in the other 
areas. For the period 1990-2003. SM and SB ranked 
third and fourth for landings and vessel-days because 
of hauls made on the northern coasts of the Chan- 
nel Islands and off the Malibu and Redondo Canyons, 
respectively. CC is the area least targeted, with only 
5% of landings and vessels. CPUE for this period was 
26 tons/vessel for all areas except MB, where it was 
little more than half that at 14 tons/ 
vessel. CPUE in APR/MB has nearly 
tripled since 1981-82. CPUE in the 
OCT region has increased more mod- 
estly, except in SM. 

Since 1999, annual landings have de- 
creased in OCT (91,229 tons to 22,180 
tons) and increased in APR (289 to 
14,521 tons— with 22,770 tons in 2002, 
Fig. 2A). Effort has increased as well in 
the last 23 years (Fig. 2B). With the ex- 
ception of MB 1981-82, all areas saw the 
number of vessel-days/month increase 
until the 1997-98 season. The number of 
vessel-days has decreased since 1999 — in 
OCT (4011 to 1573)— and has increased 
in APR (20 to 978). The average number 
of days between landings for individual 
boats in APR (2.3) and OCT (2.1) was 
not significantly different (^-testg 05,2,, 
df=977, f-value 0.87, P=0.39). 

There have been increases in CPUE 
concomitant with gains in experience, 
and advances in technology have en- 
hanced the ability of fishermen to lo- 
cate squid. There has been a "ratcheting 
up," both in terms of size of vessel and 
in the amount of sonar (from single to 
dual frequency [50 to 200 kHz]) used in 
the competition among fishermen. How- 
ever, CPUE decreased substantially in 
all regions in 1984 and 1998, the second 
years of the two biggest El Niiio events 
recorded. Milder El Niiio events in 1987 
and 1992 preceded dips in CPUE val- 
ues in 1988 and 1993 (Fig. 2C). Aver- 
age CPUE was calculated for the APR 
and OCT regions by splitting the data 



Zeidberg et al The fishery for Loligo opa/escens from 1981 through 2003 



t/l o 

O) o 

c o 






6000 


S 5000 


^ 4000 
1 3000 
■2 2000 


1000 





35 


Q 30 

1 25 
§ 20 
w 15 

2 10 
o 5 






by the frequencies determined from the 
spectral analysis. These splits resulted in 
three separate means for CPUE in APR 
(7.5-year frequency) and five means for 
the OCT region (4.5-year frequencies). 
Anomalies of CPUE from these means 
were compared to the climatic indices, 
and had significant linear regressions to 
NIN03, SOI, and UI anomalies, but ex- 
plained less than 5% of the variance (data 
not shown). 

1981-2003 squid fishery data 



In comparisons of landings (Fig. 3A) by 
month, the six areas fell into two catego- 
ries: APR and OCT. Effort in vessel-days 
and CPUE showed similar trends. The 
Loligo opalesce/TS squid fishery generally 
occurs from April through November in 
APR. Although landings peak in May, by 
then there are so many vessels in oper- 
ation that CPUE has dwindled to half 
that of April (Fig. 3, B and C). There is a 
second landings pulse in August. 

In the five areas of the OCT grouping, 
landings typically begin in October, build 
to a peak in January, and diminish to 
lows in August (Fig. 3A). A large uni- 
modal pulse of squid landings occurred in 
November for all areas except SCB. The 
SCB had a bimodal recruitment pulse, the 
two largest recruitment events in all of 
California: one in November and a larger 
one in January. In SD, like SCB, land- 
ings peaked in January, but there was no 
strong November signal in this region. 

CPUE by month for APR was typically half that of 
OCT. The APR CPUE varied between 8 and 20 tons/ 
vessel-day, for months with more than seven vessel- 
days, whereas CPUE for the southern California (OCT) 
areas ranged from 17-36 tons/vessel-day (Fig. 3C). 

Time-series analysis 

The periods of the ten highest peaks of the variance 
spectrum were determined for all six areas (Table 2). 
The largest peaks from the spectral analysis occurred 
at periods of 372 and 356 days, or roughly 1 year for 
all areas. There was a 7.5-year peak in the MB and CC 
areas. There was a 4.5-year peak (for all areas except 
MB) that was similar to the period of the four El Nifio 
events that occurred in this area during the 22-year 
period. There was a 3.7-year peak for all areas except CC 
and SM. The seven-day cycle is most likely a stochastic 
factor of fishermen working within weekend closures 
because data before 1998 did not have this periodicity. 
There was no 28- or 14-day cycle in any of the areas; 
this finding likely indicates that spawning squid do not 
respond to tidal currents or lunar light. 



120- 


A 

o 










0,. 




80 - 




Q. 






p 




40- 






o. 








0- 




— • — 


r''^ , — 


-<^~<>-^^ 


! •— 


--*- 



1 

B 

--f— 


2 

O 


3 4 


5 

O  
1 


6 

O 

1 


7 
O 


8 


9 


10 

o 

r 


11 


12 


-•-APR 
o OCT 


 O 






-9- 


O 





10 



iC 

- a o.. o- ° 

o O • 

•o •» o ° 

— *l- — I 1 1 1 1 1 1 1 1 , ^ 



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

Month 

Figure 3 

Fishery data for Loligo opalescent:, summed by month for 1981-2003. 
(A) Landings (metric tons). (B) Effort in vessel-days (VD). (C) Catch 
per unit of effort (tons/VD). Monterey Bay (April [APR] — black circles) 
and southern California (October [OCT] — unfilled circles). Scale of 
y-axis changes between A, B, and C. Largest landings occurred one 
month later in May in the APR region and in November in the OCT 
region when SST was 11.7"C and 16.1°C, respectively. 



The most significant cross correlations of time lag 
analysis for CPUE to SST are listed in Table 3. In all 
cases of biological significance, CPUE lagged SST by 
4, 5, or 10 months. MB CPUEs were highest (in May) 
when SST was low four months earlier (Jan), and hence 
gave a negative correlation. In all other regions, the 
four or five month correlation was positive, with CPUE 
high (Nov) when SSTs were high four months earlier 
(Jul). For the southern California areas there was a 
negative correlation with SSTs 10 months earlier (Janl. 
Therefore, cold winters and warm summers correlate 
with larger landings. Recruitment of spawning adults 
to the fishery occurs during the productive seasons in 
both APR, MB, and OCT. Productivity in APR and MB 
co-occurs with the spring-summer upwelling season, 
and in OCT productivity correlates with winter storms 
that lead to a deeper mixed layer. There were signifi- 
cant cross correlations with SOI, NIN03, and UI, but 
not with the anomalies of SOI and UI. Interestingly 
correlations for NIN03 were greater than those for SOI 
(Fig. 4), indicating that the CPUE of Loligo opalescens 
is more closely related to the "oceanic teleconnection" 
than to "atmospheric teleconnection," 



52 



Fishery Bulletin 104(1) 



Table 2 

Periods of greatest spectral variance in the daily CPUE 
data of the market squid tLoligo opnlescens) fishery for 
1981-2003. Significance: P<0.01. Numbers in bold are 
repeated in more than one area. Harmonics of factors of 2: 
2, 4, ... 4096 (blank spaces I are omitted because they are 
inherent in spectral analysis and not relevant to this spe- 
cies. MB=northern coastal area, predominantly southern 
Monterey Bay; CC = central coast; SB = Santa Barbara; 
SCB = Southern California Bight; SM = Santa Monica; and 
SD = San Diego. 





Top ten periods of spectral 


variance 








MB 






CC 




Rank 


Days 


Years 


Month 


Days 


Years 


Month 


1 


372.4 


1 


12.2 








2 


356.2 


1 


11.7 


372.4 


1 


12.2 


3 








1638.4 


4.5 


53.7 


4 


7 





0.2 


2730.7 


7.5 


89.5 


5 


315.1 


0.9 


10.3 


455.1 


1.2 


14.9 


6 


2730.7 


7.5 


89.5 


481.9 


1.3 


15.8 


7 


341.3 


0.9 


11.2 


356.2 


1 


11.7 


8 


455.1 


1.2 


14.9 


390.1 


1.1 


12.8 


9 


3.5 





0.1 








10 


1365.3 


3.7 


44.8 












SB 






SCB 




Rank 


Days 


Years 


Month 


Days 


Years 


Month 


1 


372.4 


1 


12.2 


372.4 


1 


12.2 


2 


356.2 


1 


11.7 


1638.4 


4.5 


53.7 


3 


390.1 


1.1 


12.8 


356.2 


1 


11.7 


4 


1365.3 


3.7 


44.8 


1365.3 


3.7 


44.8 


5 


1638.4 


4.5 


53.7 


390.1 


1.1 


12.8 


6 

7 
8 


910.2 


2.5 


29.8 








264.3 


0.7 


8.7 








9 








819.2 


2.2 


26.9 


10 


182 


0.5 


6 


682.7 


1.9 


22.4 


Rank 




SM 






SD 




Days 


Years 


Month 


Days 


Years 


Month 


1 


372.4 


1 


12.2 


356.2 


1 


U.7 


2 


1638.4 


4.5 


53.7 








3 


182 


0.5 


6 


372.4 


1 


12.2 


4 








182 


0.5 


6 


5 








1638.4 


4.5 


53.7 


6 


356.2 


1 


11.7 


682.7 


1.9 


22.4 


7 


390.1 


1.1 


12.8 


341.3 


0.9 


11.2 


8 


315.1 


0.9 


10.3 


1365.3 


3.7 


44.8 


9 


204.8 


0.6 


6.7 


390.1 


1.1 


12.8 


10 


7 





0.2 


178.1 


0.5 


5.8 







Table 3 




Results of the time 


series analysis. Significant correla- 


tion coefficients occurred when CPUE lagged behind sea | 


surface temperature 


(SST) from buoys and advanced very 


high resolution radiometers (AVHRRs) by 4 


-10 months for 


all areas, exc 


ept CC. 


Negative correlation coefficients deni- 


onstrate that high CPUE corresponds with low water tem- | 


peratures in 


the lags 


zed month from column two; positive | 


values may 


indicate 


a direct relationship. 


MB=northern 


coastal area 


predominantly southern Monterey Bay; CC = 


central coast 


; SB = Santa Barbara; SCB = Southern Califor- 


nia Bight; SM=Santa Monica; and SD = San Diego. 


CPUE 


Lagg( 


?d SST 


Correlation 


region 


(months) source 


coefficient 


MB 


4 


MB buoy 


-0.481 




5 


AVHRR 


-0.358 


SCB 


4 


SM buoy 


0.206 




10 


SM buoy 


-0.344 




9 


AVHRR 


-0.340 


SD 


4 


SM buoy 


0.220 




10 


SM buoy 


-0.398 




10 


AVHRR 


-0.387 


SM 


5 


SM buoy 


0.176 




10 


SM buoy 


-0.340 




10 


AVHRR 


-0.334 


SB 


4 


ESB buoy 


0.387 




4 


SM buoy 


0.415 




9 


AVHRR 


-0.372 



Assuming a 6-9 month lifespan for L. opalescens, 
we used linear regression to compare SST from buoy 
data for the previous 6-10 months. We performed com- 
parisons up to 10 months because squid eggs take 30 
days to hatch at 12°C, which is a typical time period 
for eggs to hatch in winter in Southern California and 
spring-summer in Monterey Bay. The only significant re- 
gression occurred in the SM region with SSTs 10-months 
earlier (r- = 0.46, P=0. 00.33, Fig. 5). We compared satel- 
lite-derived (AVHRR) estimates of SST for 1985-2002 
from areas with high densities of paralarvae and juve- 
niles (within 8 km of shore) with CPUE by using cross 
correlation time series analysis. Although there were 
significant correlations, linear regression yielded no sig- 
nificant predictions for landings or CPUE from SST. 

Stock recruitment analysis 

We compared a paralarvae density index (PDI) with 
CPUE for the SCB and SM regions (Fig. 6, shaded area 
of Fig. 1). Collections of paralarvae were made in Febru- 
ary (Fig. 6). After the initial surveys of 1999, methods 
developed in Zeidberg and Hamner (2002) resulted in 
34-50 stations for oblique bongo tows to collect paralar- 
vae in SCB and SM regions. Paralarvae/1000 m-' from 



Zeidberg et al : The fishery for Loligo opalescens from 1981 through 2003 



53 



APR CPUE/SOI 




10 12 14 16 -16 -14 -i: -10 

Time lag (months) 



10 12 14 16 



Figure 4 

Time-series analysis: cross-correlation between CPUE and the global climatic indices: Southern Oscilla- 
tion index (SOI for atmospheric pressure differences between Tahiti and Darwin, A and B) and NIN03 
(SST anomaly in eastern equatorial Pacific, C and D) for regions OCT (A, C) and APR/MB (B, D). The 
correlation between CPUE and N1N03 is greater than the correlation between CPUE and SOI in both 
regions. CPUE lags NIN03 by 9-11 months in APR and 4 months in OCT. Thus, the effects of an El 
Nifio event cause declines in CPUE for Loligo opalescens in Southern California 4 months later and 
in Monterey Bay 9-11 months later (long arrows). High correlation coefficients at a 10-month lag in 
Southern California (OCT) may be due to a second generation of market squid responding to changes 
in SST (shorter arrow). 



all stations were averaged to create the February PDI 
(Fig. 6 lower right), and this PDI was then compared to 
the November recruitment of spawning adults (CPUE) to 
the fishery for the same year. Linear regression was not 
significant for 1999-2003 {r'- = 0.522. P=0.1683). How- 
ever, if 1999 was treated as an outlier because it lacked 
nearshore sampling sites where 76% of the paralarvae 
were captured subsequently, the regression explains 
97.8% of the variance, and the F-value of the ANOVA 
ratio test for this regression is significant, P=0.007 
(Fig. 7J. From 1992 through 2002, SCB (36.2%) and SM 
(16.2%) represented nearly half of the landings for the 
state, and therefore this technique (regression of CPUE 
on PDI) could apply throughout the state. 



Discussion 

We report landings, effort, and catch per unit of effort 
for Loligo opalescens in California for 1981-2003. It is 
important to reiterate that CPUE is an approximation of 
abundance in the fishery and fails to estimate biomass of 
squid in California waters. Vessels that attempt to cap- 



40  


"-« CPUE= 131.19 -7.24 (SST); r2 = 0.46 


f 35 - 


• 


o 


W^ "-^ 


r 30  


^*^^ --., 1990 


o 


CD 

Z. 25  
o 


. ^^^ ~ - 


.t: 


^^ ^^^w_ " ^ — '"^^ 


§ 20 - 


""^ ^"V,^ 


CD 

Q- 15 - 


2002 •^"v ^V^ 


x: 


• ^ ^^V^ 


5 10 - 


1986 X ^\,^ 




• 


13 5 14 14,5 15 15.5 16 16 5 17 


Sea surface temperature (SST) 


10 months prior 


Figure 5 


Linear regression and 95'* confidence intervals for CPUE 


in November, the highest recruitment month, from SST of 


the previous January in the SM region. CPUE=131.19 - 


7.24 xSST; r2 = 0.46, P=0.0033. 



54 



Fishery Bulletin 104(1) 




2000 



c^ •■ 




.•o;Ooo« ^^ 



'f*\ 






V- 



2001 



\ 



2002 



"'°o 



■•»-\ 



>.*' 



oas® ~. 



\ 



"^■r^.w 





0- 1 
1 -10 
10-100 
i 100- 1000 



 •••••«> 




C,e* ', 



f*\ 



•^• 



i-o 



"^^ ^1000-10000 

^ Paralarvae/1 000 m3 



V 



2003 






^•" 



■j^-S 



^°° Paralarvae density index (PDI) 

1 80 

o 

o 

° 60- 



V 



g Strong 

> 40 La Nina 
i5 year 



Weak 

El Nirio 

year 



2000 2001 
rear 



Figure 6 

Density profiles (exponential bubble ploti for Loligo opalesceus paralarvae in the 
Southern California Bight, February 1999-2003 (shaded area of Fig. 1). Size of 
circles corresponds to number of paralarvae/1000 m-' seawater sampled. Data 
for 1999-2001 reprinted with permission from Springer-Verlag, originally in 
Zeidberg and Hamner (2002) Mar. Biol. 141(1):111-122. Data from all tows were 
averaged to obtain a paralarvae density index (PDI) for each year, lower right. 
1999 La Niiia (cold) and 2002 El Nifio (warm) events are labeled above bars in 
the index. Note the lack of any sample sites within 8 km of shore or with >100 
paralarvae/1000 m-= in 1999. 



ture squid and fail cannot be tracked with this method, 
and squid that are not harvested commercially are not 
accounted for in this report. Loligo opalescens reproduces 
by aggregating from small, foraging groups of hundreds 
of individuals to groups of millions of individuals. It is 
possible, therefore, that a large decrease in biomass 
can be masked by a larger percentage of the population 
aggregating in seemingly similar-size spawning masses. 
Such species are vulnerable to highly mobile fisheries 
(Oostenbrugge et al., 2002). 



Trends in the fishery 

The fishery for market squid (Loligo opalescens) has 
increased in all parts of the study area since 1983 
because of an increase in demand for calamari inter- 
nationally and because of the collapse of other fisheries 
both within and outside California waters. The major- 
ity of fishing activity has shifted from Monterey Bay to 
the Southern California Bight since the 1980s. Fishing 
activity in the bight experienced a second increase in 



Zeidberg et al. The fishery for Loligo opalescem from 1981 through 2003 



55 



40- 




i- 35- 
Q_ ro 

o y 
rS 30- 

CD -S 25 • 
■5 g 
If 20- 


X^ 


if '^- 


•/^ CPUE = 8.423 + 0.407 (PDI); 
^^ r- = 0.978, P = 0.007 


i^ 10. 
o 


"• 


5 . 




10 20 30 40 50 60 70 80 


Paralarvae density index (PDI) 


February 


Figure 7 


Stock-recruitment model: linear regression of catch per 


unit of effort I CPUE) for spawning adults in November 


on the February paralarvae density index I PDI) in the 


SM and SCB areas for 2000-2003. 



the 1990s, reflecting fishery participants from Alaska, 
Washington, and Oregon. The most economically harm- 
ful trend has been the substantial decrease in landings 
during the second year of strong El Niiio events, and the 
slight decrease in landings after weak ones. 

The initial impetus of performing the spectral analy- 
sis was to determine if the squid were migrating to the 
spawning grounds in relation to a lunar or tidal signal. 
It is important to note that the spectral analysis with 
CPUE and landings data (not shown) did not show that 
squid recruit to spawning sites in a fortnightly cycle. 
There was no 14-day period in any area. Spectral analy- 
sis demonstrated periodicities for CPUE of Loligo opal- 
escens on scales ranging from days to years. The most 
common periods for all areas were annual. Varying from 
315 to 390 days, annual cycles made up more than half 
of the top ten signals in the analysis. The 4.5-year cycle 
corresponds well with the El Nirio events of 1982-83, 
1987. 1992, and 1997-98 (Hayward et al., 1999). In 
each of these cases the CPUE anomalies were negative 
(Zeidberg, 2003). The longest period was 7.5 years in 
the MB and CC areas. There were evident leaps in the 
mean CPUE based on mean CPUE ±5 months in MB 
at mid-1988 and the end of 1995, when out-of-state fish- 
ermen began to harvest squid in California (Zeidberg, 
2003). Although these leaps may correspond to changes 
in the biomass of the squid, they are more likely due to 
enhancements in the capacity of the fishery to capture 
squid as acoustic and communication technology has 
improved. The 3.7-year period is probably a statistical 
harmonic of the 7.5-year period. 

Paralarvae density index (PDI) can predict CPUE 

Zeidberg and Hamner (2002) have sampled the SCB 
and SM areas for Loligo opalescens paralarvae since 



1999 and we used that data to create a paralarvae 
density index (PDI). CPUE appears to be a better 
indicator of stock abundance than landings data for 
squid (Sakurai et al., 2000). Adults recruiting to the 
fishery in November, measured in CPUE, can be pre- 
dicted by linear regression from the PDI of February. 
A regression of the CPUE data from the PDI data for 
1999-2003 is not significant, but if 1999 is treated as 
an outlier the remaining four points (2000-03) create 
a regression that explains 97.8% of the variance. Our 
1999 sampling of paralarvae may not be representative 
of the fishery because it was the first sampling year 
and the sampling sites were located farther offshore 
than those sampled in 2000-03. In 1999 there were no 
sites within 7.4 km of shore, where 76% of the paralar- 
vae were captured in the following four years of sam- 
pling. Despite these caveats, this method could provide 
the first opportunity to manage California's market 
squid fishery according to scientifically gathered bio- 
logical indicators and with very few of the inherent 
assumptions needed for many other types of forecast- 
ing (Mangel et al., 2002). As the years of logbook data 
accumulate, estimates of CPUE will be more closely 
related to the actual biomass of the species. By the end 
of February, we can have a prediction for the CPUE 
for the following year's adult recruitment. Paralarvae 
may be the best stage of the life cycle for a fishery 
prediction because juveniles can escape trawls, fewer 
assumptions need to be made than with estimates from 
spawning females (Macewicz et al., 2004), and there 
is sufficient time (6-9 months) to develop predictions. 
These predictions could help managers set catch limits 
and aid fishermen in deciding how to invest in gear for 
the following season. 

In addition to our paralarvae sampling, CalCOFI 
has sampled the waters of California for zooplankton 
in a manner similar to ours since 1949. Paralarval 
distributions for Loligo opalescens have been described 
from these data (Okutani and McGowan, 1969). The 
greatest difference between the two sampling efforts 
is the number of stations that are in close proximity 
to land. The majority of the paralarvae (76%) captured 
by Zeidberg and Hamner (2002) were at stations less 
than 8 km from shore, but there is only one CalCOFI 
station at this proximity to land. After reviewing their 
surveys and models of larval dispersal (Franks, 1992; 
Botsford et al., 2001; Siegel, 2003), we predict that a 
PDI calculated from CalCOFI samples will be substan- 
tially lower than ours, but given the long time period 
of the CalCOFI sampling program, any significant cor- 
relations could be more powerful statistically than ours. 
Furthermore, fishermen could be employed to perform 
bongo tows for paralarvae in proximity to shore to com- 
plement CalCOFI data. If the CalCOFI bongo net data 
were sorted for Loligo opalescens paralarvae, and fisher- 
men collected paralarvae nearshore, Monterey Bay and 
southern California CPUE could be predicted months 
in advance. Separate management of the two regions 
would be necessary given the time lag of recruitment 
(APR and OCT). 



56 



Fishery Bulletin 104(1) 



Comparison of fishery data with physical data 

We found a correlation between CPUE of the largest 
recruitment month with SST buoy data from 10 months 
earlier in the SM area only. There may be physical 
features specific to this area that increase the cor- 
relation between spawning recruitment and SST. For 
example, SM is a small area, it is close to the buoy, 
most of the area is sandy bottom, and it contains the 
Redondo Canyon. Thus if further attempts to match 
physical oceanography to the biology of a pelagic species 
were to occur, the Santa Monica Bay could be the most 
ideal location. But this correlation between CPUE of the 
largest recruitment month with SST in the SM area may 
be a seasonal effect because the regression is significant 
for SST only and not an SST anomaly. Furthermore, we 
caution that the significance of the correlation between 
CPUE and SST in SM may be a type-I error because it 
was the only significant test of the 30 tests run with 
an alpha level of 0.05. The size of the recruitment event 
was not strongly related to small deviations from aver- 
age monthly SST; thus the timing of squid recruitment 
to spawning grounds in APR and OCT may be tied to 
annual fluctuations of prey availability and correlations 
with temperature may be coincidental. The 10-month 
lag corresponds to the egg-laying date of 9-month-old 
squid. The lack of a greater number of correlations may 
be due to the small spatial resolution of the buoy data 
and the enormous variability of SST data due to meso- 
scale oceanographic features in the large fishery areas. 
In some areas the nearest buoy was quite distant from 
the fishery zone. 

To address the spatial distance of spawning grounds 
from buoys, we compared SSTs derived from satel- 
lite AVHRR images with CPUE. AVHRR data were 
collected from just the shelves and slopes of the six 
fishery areas because these are the most important 
areas for the growth of hatchlings and juveniles. Cross- 
correlation time series analyses were significant at 
5-10 month lags (Table 3), but this significance did not 
translate into any predictive capabilities with linear 
regression. 

Similarly, cross-correlations of CPUE with SOI and 
NIN03 were significant at a 10-month lag in Monterey 
Bay and a 4-month lag in the Southern California 
Bight. Thus the Monterey fishery (10% of landings) is 
offset by six months (roughly one short cohort) from 
the SCB fishery. The correlation coefficients for NIN03 
were greater than those of SOI, corroborating the idea 
that the direct influence of the coastal waves ("oceanic 
teleconnection") is the main source of the changes in 
the hydrographic and ecological features of the Califor- 
nia Current system (Huyer and Smith, 1985; Rienecker 
and Mooers, 1986; Lynn et al., 1995; Chavez, 1996; 
Ramp et al., 1997) rather than the ENSO (El Niiio 
Southern Oscillation)-related changes of atmospheric 
circulation ("atmospheric teleconnection") (Simpson, 
1983; Simpson, 1984a, 1984b; Mysak, 1986; Breaker 
and Lewis, 1988; Breaker et al., 2001; Schwing et al., 
2002). 



Loliginid life cycles and future management of squid fisheries 

The correlation between SST and CPUE in the following 
season may have resulted from the unique development 
pattern of teuthids. The use of CPUE as an index of 
abundance of the population (Sakurai et al., 2000), in 
combination with studies of squid growth in relation to 
SST (Jackson and Domeier, 2003), could explain large 
fluctuations in landings data from year to year. In terms 
of bottom-up forcing, individual squid health and the 
resulting population size result from a combination 
of prey availability and metabolic rates. Squids grow 
exponentially in the first two months of life and then 
logarithmically until senescence. In rearing tanks and 
given a constant food supply, loliginids also grow faster 
in warmer temperatures (Yang et al., 1986; Forsythe 
et al., 2001) as their metabolic rates increase (O'Dor, 
1982). Grist and des Clers (1998) predicted that annual 
fluctuations in SST that cause differential growth in 
squids can lead to younger cohorts hatched in warm 
temperatures and surpassing in size older cohorts born 
in colder seasons. Thus in October, a large 6-month-old 
squid that hatched in April and developed in warm 
water may spawn with a smaller 9-month-old squid that 
hatched in the cold waters of January. 

However, in the California system and possibly in 
other upwelling systems the situation is more complex 
than in rearing tanks. For example, Jackson and Do- 
meier (2003) demonstrated that due to the influences of 
El Nino and La Nifia cycles and upwelling. the mean 
mantle length of Loligo opalescens is shortest when 
larvae are hatched in the warmest temperatures and 
longest when hatched in cold waters. Mantle length is 
also positively correlated with the upwelling index. In 
the ocean, squid do not have a constant food supply. 
The high productivity and cold temperatures caused by 
upwelling and La Nina combined to create a period of 
rich food resources and lower metabolic rates for squid, 
probably enhancing the recovery of the fishery in 1999. 
During the El Nifio event the squids were small and less 
abundant because they had a high metabolic rate due 
to increased temperatures and were exposed to lower 
levels of available prey due to decreased ocean produc- 
tivity. Seasonal maxima of phytoplankton in Monterey 
Bay occur in summer; but in the southern part of the 
Southern California Bight productivity peaks in win- 
ter (Nezlin et al., 2002). These differences may be an 
indicator of why the fishery operates in Monterey Bay 
from April to November, coinciding with the upwelling 
season, and in the Southern California Bight from No- 
vember to May, coinciding with less stratification of the 
water column and more mixing due to winter storms 
and colder air temperatures. 

Lowry and Carretta (1999) corroborated the tempera- 
ture-induced plasticity of mantle length (ML) from beaks 
of squid in California sea lion (Zalophus californianus) 
scats and spewings. MLs of squid prey were half the 
size during El Nino years on San Clemente and Santa 
Barbara islands. However, at San Nicholas Island dur- 
ing El Nino events, there were both small- and regular- 



Zeidberg et al : The fishery for Loligo opolescens from 1981 through 2003 



57 



size squid prey; this finding may indicate that the squid 
stock moved offshore to find productive waters. Alter- 
natively, San Nicholas sea lions may have been feeding 
on squids from Baja California. Zeidberg and Hamner 
(2002) suggested the possibility of a northern shift of 
the squid population in El Niiio years, as has been 
found for most zooplankton (Colebrook, 1977). 

However, the growth plasticity and fluctuating re- 
productive success for Loligo opalescens should not be 
underestimated. The possibility remains that the huge 
fluctuations in squid landings during strong ENSO 
events is due to the entire biomass of market squid 
waning and waxing rather than to population migra- 
tions away from traditionally fished spawning grounds. 
Triennial groundfish surveys demonstrate that market 
squid experienced a coast-wide population decrease, not 
a poleward migration during the 1997-98 El Nino.*^ 

With the exception of El Nino years, the fishery in- 
creased its landings each year until 2000. However, it 
remains unknown if the capacity of the fishery is close 
to reaching the total biomass of squid in California. 
The California sardine {Sardinops sagax) fishery col- 
lapsed in the 1960s, and a twenty-year moratorium was 
required before there was recovery to a fraction of prior 
spawning biomass (Wolf, 1992). Whether over-fishing or 
large scale, multidecadal climatic regime shifts caused 
this collapse is matter of debate (Chavez et al., 2003), 
but without an effective management plan, squid will 
continue to be fished because of market demand. Mar- 
kets are driven by economic forces and traditionally 
do not control themselves in a biologically sustainable 
manner. A full recovery of the squid fishery occurred 
from 1998 to 2000 and thus spanned four generations of 
squid given a 6-9 month lifecyle; for the California sar- 
dine (with a 6-8 year lifecycle), a proportionally similar 
recovery period would be 24-32 years (Parrish^). 

In 1998-99 the fishery for Loligo opalescens decreased 
to low levels during the El Nino event, then recovered 
to record levels in the following years. This was most 
likely due to the plasticity of squid development in rela- 
tion to water temperature and upwelling and the short 
(4-6 month) life span of squid. One should not assume 
that the ability of this species to recover from environ- 
mental stress like El Nino applies also to the recent 
anthropogenic stresses associated with increasing fish- 
ery capacity. It remains to be seen if the large decline 
in southern California landings in the last five years 
(119,780-24,449 tons/year) is due to the small El Nino 
of 2002-03, the climate-regime shift in 1998, overfish- 
ing, or some other factor such as increased water strati- 
fication due to global warming. Although the short-lived 
squid may not be able to recover from overexploitation 



" CDFG (California Department of Fish and Game). 2005. Fi- 
nal market squid fishery management plan. Website: http:// 
www.dfg.ca.gov/mrd/msfmp/index.html [Accessed on 6 June 
2005.] 

' Parrish, R. 2005. Personal commun. Fisheries and 
Marine Ecosystems Program, Pacific Fisheries Environ- 
mental Laboratory, 1352 Lighthouse Ave. Pacific Grove, CA 
93950-2097. 



in short order, the huge number of long-lived birds, fish, 
and marine mammals (Morejohn et al., 1978; Lowry and 
Carretta. 1999) that depend on squid as a key forage 
species may not be able to recover rapidly from lack of 
management foresight. The recent establishment of the 
marine reserve system in the Channel Islands elimi- 
nates 139f of key squid fishing grounds. This ecosystem- 
based management approach may assist in protecting 
not only the squid but also their predators. 



Acknowledgments 

This research was funded by a California Fish and Game 
award (no. FG7334MR) and a Coastal Environmental 
Quality Initiative Program grant (no. 783828-KH-19900) 
and was supported in part by the David and Lucile Pack- 
ard Foundation through the Monteray Bay Aquarium 
Research Institute. Bruce Robison, Rich Ambrose, Peggy 
Fong, and Dick Zimmer provided thoughtful reviews. 
Andrea Steinberger assisted with data management. 



Literature cited 

Agnew, D. J., J. R. Beddington, and S. L. Hill. 

2002. The potential use of environmental information 

to manage squid stocks. Can J. Fish. Aquat. Sci. 

59:1851-1857. 
Agnew, D. J., S. Hill, and J. R. Beddington. 

2000. Predicting the recruitment strength of an 
annual squid stock: Loligo gahi around the Falkland 
Islands. Can J. Fish. Aquat. Sci. 57:2479-2487. 

Aristegui, J., P. Tett, A. Hernandez-Guerra, G. Basterretxea, 
M. F. Montero, K. Wild. P. Sangra, S. Hernandez-Leon, M. 
Canton, J. A. Garcia-Braun, M. Pacheco, and E. D. Barton. 
1997. The influence of island-generated eddies on chloro- 
phyll distribution: a stufy of mesoscale variation around 
Gran Canaria. Deep Sea Res. 44:7-96. 
Bakun, A. 

1973. Coastal upwelling indices, west coast of North 
America, 1946-1971, 103 p. NOAA Tech. Rep. NMFS 
SSRF-671. 
Botsford L. W., A. Hastings, and S. D. Gaines. 

2001. Dependence of sustainability on the configuration 
of marine reserves and larval dispersal distance. Ecol. 
Lett. 4(2):144-150. 

Bray N. A., A. Keyes, and W. M. L. Morawitz. 

1999. The California Current system in the Southern 
California Bight and Santa Barbara Channel. J. Geo- 
phys Res 104 C4:7695-7714. 

Breaker, L. C, and P. A. W. Lewis. 

1988. A 40-50 day oscillation in sea-surface temperature 
along the central California coast. Estuar. Coast. Shelf 
Sci. 26:395-408. 
Breaker, L. C, P. C. Liu, and C. Torrence. 

2001. Intraseasonal oscillations in sea surface tempera- 
ture, wind stress, and sea level off the central California 
coast. Cont. Shelf Res. 21(6-7):727-750. 
Brink, K. H., R. C. Beardsley, J. Paduan, R. Limeburner, M. 
Caruso, and J. G. Sires. 

2000. A view of the 1993-1994 California Current based 
on surface drifters, floats, and remotely sensed data. J. 
Geophys. Res. Oceans. 105(C4):8575-8604. 



58 



Fishery Bulletin 104(1) 



Caddy, J. F., and P. G. Rodhouse 

1998. Cephalopod and groundfish landings: evidence for 

ecological change in global fisheries? Rev. Fish Biol. 

Fish. 8:431-444. 
Chavez. F. P. 

1996. Forcing and biological impact of onset of the 1992 
El Nino in central California. Geophys. Res. Let. 
23(3):265-268. 

Chavez, F. P., J. Ryan, S. E. Lluch-Cota, and M. Niquen-C. 

2003. From anchovies to sardines and back: multidecadal 
change in the pacific ocean. Science 299:217-221. 
Colebrook, J. M. 

1977. Annual fluctuations in biomass of taxonomic 
groups of zooplankton in the California Current 1955- 
59. Fish. Bull. 75:357-368. 
DiGiacomo, P. M., and B. Holt. 

2001. Satellite observations of small coastal ocean eddies 
in the Southern California Bight. J. Geophys. Res. 106 
(C10):22, 521-22, 543. 
Falkowski, P. G.. D. Ziemann, Z. Kolber, P. K. Bienfang. 

1991. Role of eddy pumping in enhanced primary pro- 
duction on the ocean. Nature 352:55-58. 

Forsythe, J. W., L. S. Walsh, P. E. Turk, and P. G. Lee. 

2001. Impact of temperature on juvenile growth and 
age at first egg-laying of the Pacific reef squid Sepio- 
teuthis lessoniana reared in captivity. Mar. Biol. 
138:103-112. 

Franks, P. J. S. 

1992. Sink or swim: accumulation of biomass at fronts. 
Mar. Ecol. Prog. Ser. 82:1-12. 

Graham, W. M., and J. L. Largier. 

1997. Upwelling shadows as nearshore retention sites: 
the example of northern Monterey Bay. Cont. Shelf 
Res. 17:509-532. 

Grist, E. P. M., and S. des Clers. 

1998. How seasonal temperature variations may influ- 
ence the structure of annual squid populations. IMA 
J. of Math. Appl. Med. Biol. 15:187-209. 

Harms, S., and C. D. Winant. 

1998. Characteristic patterns of the circulation in the 
Santa Barbara Channel. J. Geophys. Res. V103, 
C2:3041-3065. 

Hayward T. L., T. R. Baumgartner, D. M. Checkley, R. Durazo, 
G. Gaxiola-Castro, K. D. Hyrenbach, A. W. Mantyla, M. M. 
Mulhn, T. Murphree, F. B. Schwing, P. E. Smith, M. J. 
Tegner. 

1999. The state of the California Current in 1998-1999: 
Transition to cool-water conditions. Calif Coop Oceanic 
Fish Invest Rep 40:29-62. 

Hernandez-Trujillo, S. 

1999. Variability of community structure of Copepod 
related to El Ninol982-83 and 1987-88 along the west 
coast of Baja California Peninsula, Mexico. Fish. 
Oceanog. 8:284-295. 
Hickey B. M., E. L. Dobbins, and S. E. Allen. 

2003. Local and remote forcing of currents and tem- 
perature in the central Southern California Bight. J. 
Geophys. Res. Oceans 108(C3):3081-3084. 
Huyer, A., and R. L. Smith. 

1985. The signature of El Nifio off Oregon in 1982-83. 
J. Geophys. Res. 90:7133-7142. 
Jackson, G. D., and M. L. Domeier. 

2003. The effects of an extraordinary El Nino/La Niiia 
event on the size and growth of the squid Loligo opalescens 
off Southern California. Mar. Biol. 142:925-935. 



Leos, R. R. 

1998. The biological characteristics of the Monterey Bay 
squid catch and the effect of a two-day-per-week fishing 
closure. CalCOFI Reports 39:204-211. 

Loukashkin, A. S. 

1976. On biology of market squid, Loligo opalescens, a 
contribution toward the knowledge of its food habits and 
feeding behavior. CalCOFI Reports 18:109-111. 

Lowry, M. S., and J. V. Carretta. 

1999. Market squid {Loligo opalescens) in the diet of Cali- 
fornia sea lions {Zalophus californianus) in southern Cal- 
ifornia (1981-1995). CalCOFI Reports 40:196-207. 

Lynn, R. J., F. B. Schwing, T. L. Hayward. 

1995. The effect of the 1991-1993 ENSO on the California 
Current system. CalCOFI Reports 36:57-71. 
Macewicz, B. J., J. R. Hunter, N. C. H. Lo, E. L LaCasella. 

2004. Fecundity, egg deposition, and mortality of market 
squid [Loligo opalescens). Fish. Bull. 102:306-327. 

Mangel, M., B. Marinovic, C. Pomeroy, and D. CroU. 

2002. Reqiuem for Ricker: unpacking MSY. Bull. Mar. 
Sci. 70:763-781. 

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

1978. The importance of Loligo opalescens in the food 
web of Marine vertebrates in Monterey Bay, Calif Fish 
Game Fish Bull. 169:67-98. 
Mclnnis, R. R., and W. M. Broenkow. 

1978. Correlations between squid catches and oceanog- 
raphy conditions in Monterey Bay, Calif. Fish Game 
FishBull. 169:161-170. 
Mysak, L. A. 

1986. El Nino, interannual variability and fisheries in 
the northeast Pacific Ocean. Can. J. Fish. Aquat. Sci. 
43:464-497. 
Nezlin, N. P., W. M. Hamner, and L. D. Zeidberg. 

2005. Influence of El Nino 1997-1998 on the Pelagic 
Ecosystem off California: remote sensed analysis. In 
California and the World Ocean '02 (O. T. Magoon, H. 
Converse, B. Baird, B. Jines, and M. Miller-Henson, 
eds.), p. 1233-1252. 

Nezlin, N. P., and B. L. Li. 

2003. Time series analysis of remote-sensed chlorophyll 
and environmental factors in the Santa Monica-San 
Pedro Basin off Southern California. J. Mar. Syst. 
39:185-202. 

O'Dor, R. K. 

1982. Respiratory metabolism and swimming perfor- 
mance of the squid Loligo opalescens. Can J. Fish. 
Aquat. Sci. 39:580-587. 
Okutani T., and J. A. McGowan. 

1969. Systematics, distribution and abundance of the 

epiplanktonic squid (Cephalopoda, Decapoda) larvae of 

the California Current April, 1954-March, 1957. Bull. 

Scripps Inst. Oceanog. 14:1-90. 

Oostenbrugge, J. A. E., J. J. Poos, W. L. T. van Densen, and 

M. A. M. Machiels. 

2002. In search of a better unit of effort in the coastal 
liftnet fishery with lights for small pelagics in Indonesia. 
Fish. Res. 59:43-56. 
Pauly, D., V. Christensen, J. Dalsgaard, R. Froese, and F. Torres Jr. 
1998. Fishing down marine food webs. Science 279: 
860-863. 
Ramp, S. R., J. L. McClean, C. A. Collins, A. J. Semtner, and 
K. A. S. Hays. 

1997. Observations and modeling of the 1991-1992 EI 
Nifio signal off central California. J. Geophys. Res. 
102(C3):5553-5582. 



Zeldberg et a\ The fishery for Loligo opolescens from 198f through 2003 



59 



Ravier, C, and J. M. Fromentin. 

2001. Long-term fluctuations in the eastern Atlantic 
and Mediterranean bluefin tuna population. ICES J. 
Mar. Sci. 58:1299-1317. 

Rienecker, M. M., and C. N. K. Mooers. 

1986. The 1982-1983 El Niiio signal off" northern 
California. J Geophys. Res. 91:6597-6608. 
Roberts, M. J., and W. H. H. Sauer. 

1994. Environment: the key to understanding the South 
African chokka squid iLoligo vulgaris reynaudii) life 
cycle and fishery? Antarctic Science 6:249-258. 
Robin. J. P., and V. Denis. 

1999. Squid stock fluctuations and water temperature: 
temporal analysis of English Channel Loliginidae. J. 
Appl. Ecol. 36:101-110. 

Rogers-Bennett, L. 

2000. Review of some California fisheries for 1999: 
market squid, dungeness crab, sea urchin, prawn, 
abalone, groundfish, swordfish and shark, ocean 
salmon, nearshore finfish. Pacific sardine. Pacific her- 
ring. Pacific mackerel, reduction, white seabass, and 
recreational. CalCOFI Rep. 41:8-26. 

Rocha, F., A. Guerra, R. Prego, U. Piatkowski. 

1999. Cephalopod paralarvae and upwelling condi- 
tions off Galician waters ( NW Spainl. J. Plank. Res. 
21:21-33. 

Sakurai, Y., H. Kiyofuji, S. Saitoh, T. Goto, and Y. Hiyama. 

2000. Changes in inferred spawning areas of Toda- 
rodes pacificus (Cephalopoda: Ommastrephidae) due 
to changing environmental conditions. ICES J. Mar. 
Sci. 57:24-30. 

Schwing, F. B., T. Murphree, L. deWitt, and P. M. Green. 

2002. The evolution of oceanic and atmospheric anom- 
alies in the northeast Pacific during the El Nifio 
and La Nina events of 1995-2001. Prog. Oceanog. 
54(l-4):459-491. 

Siegel, D. A., B. P. Kinlan, B. Gaylord, and S. D. Gaines. 

2003. Lagrangian descriptions of marine larval dis- 
persion. Mar. Ecol. Prog. Ser. 260:83-96. 

Simpson, J. J. 

1983. Large-scale thermal anomalies in the California 
Current during the 1982-1983 El Nino. Geophys. Res. 
Let. 10:937-940. 

1984a. El Nino-induced onshore transport in the Cali- 
fornia current during 1982-1983. Geophys. Res. Let. 
ll(3):241-242. 

1984b. A simple model of the 1982-1983 Californian "El 
Nino". Geophys. Res. Let. ll(3):243-246. 



Vojkovich, M. 

1998. The California fishery for market squid (Loligo 
opalescens). CalCOFI Rep. 39:55-61. 

Waluda, C. M., P. N. Trathan, and P. G. Rodhouse. 

1999. Influence of the oceanographic vriability on recruit- 
ment in the Illex argentinus i Cephalopoda: Ommas- 
trephidae) fishery in the South Atlantic. Mar. Ecol. 
Prog. Ser. 183:159-167. 

Waluda, C. M., P.G. Rodhouse, G. P. Podesta, P. N. Trathan, and 
G. J. Pierce. 

2001. Surface oceanography of the inferred hatching 
grounds of Illex argentinus (Cephalopoda: Ommastrephi- 
dae) and influences on recruitment variability. Mar. 
Biol. 139:671-679. 
Watson, R., and D. Pauly. 

2001. Systematic distortions in world fisheries catch 
trends. Nature 414:534-536. 

Wing, S. R., J. L. Largier, and L. W. Botsford. 

1995. Settlement and transport of benthic invertebrates 
in an intermittent upwelling region. Limnol. Oceanog. 
40(2):316-329. 
Wolf, P 

1992. Recovery of the Pacific sardine and the California 
sardine fishery. CalCOFI Reports 33:76-86. 
Yang. W. T, R. F. Hixon, P. E. Turk, M. E. Krejci, W. H. Hulet, 
and R. T. Hanlon. 

1986. Growth, behavior, and sexual maturation of the 
market squid, Loligo opalescens, cultured through the 
life cycle. Fish. Bull. 84:771-798. 
Zar, J. H. 

1984. Biostatistical analysis, 2"'^ ed., 718 p. Prentice 
Hall, Englewood Cliffs, NJ. 
Zeidberg, L. D., and W. M. Hamner. 

2002. Distribution of squid paralarvae, Loligo opal- 
escens (Cephalopoda: Myopsida), in the Southern Cali- 
fornia Bight in the three years following the 1997 El 
Niiio. Mar. Biol. 141:111-122. 

Zeidberg, L. D. 

2003. The early life history and fishery of the Califor- 
nia market squid, Loligo opalescens. Ph.D. diss., 136 
p. Univ. California Los Angeles, Los Angeles, CA. 

Zeidberg, L. D., W. Hamner, K. Moorehead, and E. Kristof 

2004. Egg masses of Loligo opalescens (Cephalop- 
oda: Myopsida) in Monterey Bay, California follow- 
ing the El Nino event of 1997-1998. Bull. Mar. Sci. 
74:129-141. 



60 



Abstract — The variability in the 
supply of pink shrimp (Farfante- 
penaeus duorarum) postlarvae and 
the transport mechanisms of plank- 
tonic stages were investigated with 
field data and simulations of trans- 
port. Postlarvae entering the nursery 
grounds of Florida Bay were collected 
for three consecutive years at chan- 
nels that connect the Bay with the 
Gulf of Mexico, and in channels of 
the Middle Florida Keys that connect 
the southeastern margin of the Bay 
with the Atlantic Ocean. The influ.x of 
postlarvae in the Middle Florida Keys 
was low in magnitude and varied sea- 
sonally and among years. In contrast, 
the greater postlarval influx occurred 
at the northwestern border of the Bay, 
where there was a strong seasonal 
pattern with peaks in influx from 
July through September each year. 
Planktonic stages need to travel up to 
150 km eastward between spawning 
grounds (northeast of Dry Tortugast 
and nursery grounds (western Florida 
Bay) in about 30 days, the estimated 
time of planktonic development for 
this species. A Lagrangian trajectory 
model was developed to estimate the 
drift of planktonic stages across the 
SW Florida shelf. The model simu- 
lated the maximal distance traveled 
by planktonic stages under various 
assumptions of behavior. Simulation 
results indicated that larvae traveling 
with the instantaneous current and 
exhibiting a diel behavior travel up 
to 65 km and 75% of the larvae travel 
only 30 km. However, the eastward 
distance traveled increased substan- 
tially when a larval response to tides 
was added to the behavioral variable 
(distance increased to 200 km and 
85% of larvae traveled 150 km). The 
question is, when during larval devel- 
opment, and where on the shallow SW 
Florida shelf, does the tidal response 
become incorporated into the behavior 
of pink shrimp. 



Variability in supply and cross-shelf transport 
of pink shrimp iFarfantepenaeus duorarum) 
postlarvae into western Florida Bay 

Maria M. Criales' 

John D. Wang^ 

Joan A. Browder^ 

Michael B. Robblee'* 

Thomas L. Jackson^ 

Clinton Hittle" 

' Rosenstiel School of Marine and Atmospheric Science, MBF 
University ol Miami 
4600 Rickenbacker Causeway 
Miami, Florida 33149 
E-mail address (for M M Cnales) mcrialesa'Tsmas miami edu 

^ Rosenstiel School of Marine and Atmospheric Science, AMP 
University of Miami 
4600 Rickenbacker Causeway 
Miami, Florida 33149 

3 NOAA Fisheries, Southeast Fisheries Science Center 
75 Virginia Beach Drive 
Miami, Florida 33149 

■* United States Geological Survey 
Center for Water and Restoration Studies 
3110 SW9'^ Avenue 
Ft Lauderdale, Florida 33315 



Manuscript submitted 16 September 2003 
to the Scientific Editor's Office. 

Manuscript approved for publication 
30 June 2005 by the Scientific Editor. 

Fish. Bull. 104:60-74 (2006). 



Patterns of recruitment of coastal spe- 
cies are highly variable, mainly because 
of the complex interaction of biotic anii 
abiotic factors across the different life 
history stages. These factors include 
but are not limited to reproductive 
dynamics, larval dispersal and behav- 
ior, physiological tolerances, and the 
hydrometeorological regime in which 
their life stages develop (e.g., Shanks, 
1995; Cowen, 2002). The commercially 
valuable tropical penaeid shrimps that 
use different habitats during their life 
cycle (offshore spawning grounds and 
estuarine nursery habitats) have to 
cope with a large suite of physical 
processes and stimuli (Rothlisberg 
et al., 1995, 1996). The pink shrimp 
iFarfantepenaeus duorarum) of Dry 
Tortugas is one of the most economi- 
cally and ecologically important spe- 
cies in southwest Florida. The pink 
shrimp supports an important year- 
round fishery of about 4000 metric 
tons in an area of 10,000 km- between 
Dry Tortugas and Key West (Iversen 
et al., 1960; Klima et al., 1986). The 



Tortugas fishery is directly depen- 
dent on young shrimp that migrate 
from inshore nursery areas onto the 
offshore fishing grounds (Sheridan, 
1996; Browder et al., 2002). Recruit- 
ment shows no relationship to spawn- 
ing size; therefore harvest fluctuations 
are apparently due to environmental 
conditions rather than fishing opera- 
tions (Nance and Patella, 1989). To 
effectively manage this species, it is 
necessary to have accurate informa- 
tion on the processes linking nursery 
and spawning ground populations. 

Pink shrimp population dynamics 
are affected by physical processes and 
environmental conditions occurring 
in the southwestern (SW) Florida 
Shelf of the Gulf of Mexico, the At- 
lantic coastal zone, and the Florida 
Bay. Early research on gonad develop- 
ment (Cummings, 1961), distribution 
of larval stages (Jones et al., 1970), 
and analysis of length frequency 
distributions of fishery stock data 
(Iversen et al., 1960; Roberts, 1986) 
indicated that the center of spawn- 



Criales et al Variability in supply and cross shelf transport of Farfanlepenaeus duorarum postlarvae into Florida Bay 



61 



ing is northeast of the Dry Tortugas. If gravid females 
spawn northeast of the Dry Tortugas, larvae need to 
travel up to 150 km to reach the main nursery ground 
in western Florida Bay. Females spawn on the continen- 
tal shelf at about 30 m of depth, where larvae develop, 
passing through several changes in feeding habitats, 
behavior, and physical stages (nauplii, zoeae, myses) 
(Dobkin, 1961; Ewald, 1965; Jones et al, 1970). Postlar- 
vae undergo between three and eight additional plank- 
tonic stages before settlement. Larvae develop rapidly, 
needing only about 30 days to become postlarvae ready 
to settle to the bottom (Ewald, 1965; Dobkin, 1961). 
Planktonic stages (larvae and postlarvae) approach 
the coast and postlarvae enter the nursery grounds of 
Florida Bay at about 9-10 mm total length (Tabb et al., 
1962; Allen et al., 1980; Criales et al., 2000). Larval 
development and ocean hydrodynamics must be tightly 
linked to successfully bring these planktonic stages to 
their coastal nursery grounds. 

Mechanisms of transport used by planktonic stages 
of penaeid shrimps are highly variable, depending on 
the species, different environmental conditions, oceanic 
physical processes, and complexity of larval behaviors 
(e.g., Dall, 1990; Rothlisberg et al., 1995, 1996; Wenner 
et al., 2005). Physical oceanographic processes signifi- 
cantly affect the transport of planktonic stages from 
spawning to nursery grounds (Yeung and Lee, 2002; 
Criales et al., 2003). Two main immigration routes have 
been hypothesized for pink shrimp postlarvae entering 
Florida Bay: 1) postlarvae may drift south-southeast 
downstream with the Florida Current and enter Florida 
Bay through the tidal channels of the Lower and Middle 
Florida Keys (Rehrer et al., 1967; Munro et al., 1968), 
and 2) postlarvae may move northeast across the SW 
Florida shelf and enter the Bay at its northwestern 
boundary (Jones et al., 1970; Criales and Lee, 1995). 
The most widely recognized pathway for postlarvae 
to reach Florida Bay up-to-now has been by transport 
up the Atlantic side through the tidal channels of the 
Middle Florida Keys (Munro et al., 1968; Criales and 
McGowan, 1994; Criales et al., 2003). The favorable 
Ekman transport generated by the southeastern winds 
along the west-east oriented coast, and coastal counter- 
current flow generated by cyclonic eddies provide favor- 
able onshore transport mechanisms along the Florida 
Keys coast (Criales and Lee, 1995; Lee and Williams, 
1999). In contrast, larval transport across the broad, 
shallow SW Florida Shelf has not been well studied and 
questions exist about the feasibility of this pathway. 
Subtidal frequency flows are weak in the SW Florida 
shelf and mainly in the alongshore (north-south) direc- 
tion as a direct response to wind events (Koczy et al., 
1960; Weisberg et al., 1996; Lee et al., 2001). Tidal 
currents are strong mainly in the cross-shelf direction 
(Wang, 1998; Smith, 2000). Freshwater discharges from 
the Everglades affect a broad area of the SW Florida 
shelf (Lee et al., 2001; Jurado, 2003). Isopleths less 
than 32 are typically confined to the region between 
Cape Sable and Cape Romano, and from 32 to 36 extend 
from near Cape Romano to the vicinity of Dry Tortugas 



in a highly variable annual pattern (Lee et al., 2001; 
Johns and Szymanski'). 

For tropical penaeid shrimps that undergo larval 
development offshore, but whose nursery grounds are 
inshore, migratory behavior is a key factor for their 
advection to nursery grounds (Dall et al., 1990; Shanks, 
1995). The simplest migratory behavior is vertical 
movement, and three types of vertical migrations are 
known to mediate horizontal transport of larvae: on- 
togenic, diel, and tidal (for reviews see Sponaugle et 
al., 2002). For some Australian penaeid species (ba- 
nana prawn [Fenneropenaeus merguiensis], grooved 
tiger prawn [Penaeus seinisulcatus], and eastern king 
prawn [Melicertus plebejus]) it has been shown that 
early planktonic stages (protozoeae and myses) perform 
diel vertical migration cued by light and that later in 
development (as postlarvae) the migration is cued by 
tides (Rothlisberg, 1982; Rothlisberg et al., 1983, 1995) 
and there is no cross-shelf displacement of larvae dur- 
ing the 15 days of diel behavior. Previous studies of 
pink shrimp in South Florida have clearly indicated 
ontogenic behavior for pink shrimp; postlarvae have a 
higher degree of mobility than earlier protozoeae and 
myses (Temple and Fischer, 1965; Eldred et al., 1965; 
Jones et al., 1970; Criales and Lee, 1995). On the other 
hand, diel behavior is not so well determined. Although 
protozoeae, myses and postlarvae were more abundant 
at the surface during the night than during the day, 
day and night differences have not been statistically 
significant in any of these previous studies. The effect 
of diel, tidal, or ontogenic combinations of behavior 
on cross-shelf transport from South Florida spawning 
grounds to western Florida Bay nursery grounds has 
not previously been explored. The postlarvae of many 
penaeid shrimps, including pink shrimp, are known to 
synchronize vertical migration with the tides at the 
entrance to estuarine nursery grounds (for reviews see 
Garcia and Le Reste 1981, Dall et al., 1990). This pro- 
cess is known as selective tidal stream transport (STST) 
(Forward and Tankersley, 2001). Penaeid postlarvae as- 
cend in the water column during the flood and sit on the 
bottom during the ebb to maximize up-estuary move- 
ment (e.g., Rothlisberg et al., 1995). This behavior has 
been shown for pink shrimp postlarvae inside Florida 
Bay (Tabb et al., 1962; Roessler and Rehrer, 1971), but 
not along the border of the bay with the Gulf of Mexico. 
When during the life cycle and where on the shelf this 
tidal behavior begins and what the environmental cues 
are — these questions remain unanswered. 

The purpose of our research was 1) to determine pat- 
terns of supply of pink shrimp postlarvae into Florida 
Bay through two distinct regions, 2) to define the most 
important transport route for planktonic stages from 
the Dry Tortugas into Florida Bay, 3) to examine al- 
ternative behavioral responses of larvae and postlar- 
vae, and 4) to propose a recruitment mechanism for 



' Johns, E., and D. Szymanski. 2003. Mixing it up in Florida 
Bay. Florida Bay News, summer 2003:1-3. 



62 



Fishery Bulletin 104(1) 



26°N (s 



25°N 



24-N 




83 W 



82'W 



81 "W 



80°W 



Figure 1 

Map of the study area (with bathymetryl showing the channel net sampling stations 
at the northwestern and southeastern borders of the nursery grounds of Florida Bay, 
ADCP moorings, and CMAN and COMP stations. The Tortugas fishing grounds (area 
enclosed by dashed line) includes the center of spawning for Farfantepenaeus duorarum 
(filled ellipse). Northwestern stations: MG = Middle Ground and SK=Sandy Key; Florida 
Key stations: WH=Whale Harbor and PH = Panhandle; A and B = onshore and offshore 
ADCP moorings respectively; l = Long Key CMAN station; 2 = NW Florida Bay COMP 
station. Small map at the left corner indicates major currents in the Gulf of Mexico 
and off the coast of Florida. CC = Caribbean Current; LC = Loop Current; FC = Florida 
Current; GS = Gulf Stream. 



pink shrimp across the SW Florida shelf that combines 
the effect of hydrodynamics with larval behavior. Four 
modes of behavior-related transport were simulated 
under hydrodynamic conditions occurring on the SW 
Florida shelf in order that the resulting distances trav- 
eled under each condition might be contrasted. 



Material and methods 

Pink shrimp postlarvae were collected in two regions of 
Florida Bay to evaluate postulated hypotheses of eastern 
and western gateways and pathways of larval transport 
into the bay. The two study sites consisted of two large 
channels connecting northwestern Florida Bay with the 
SW Florida shelf of the Gulf of Mexico (Sandy Key Chan- 
nel [SK], and Middle Ground [MG]); and of more confined 
tidal channels in the Middle Florida Keys that connect 
the Bay with the Atlantic Ocean (Whale Harbor [WH], 



and Panhandle Key [PH]) (Fig. 1). Adjacent mud banks 
that are occasionally exposed at low tide and over-topped 
at higher tide stages define these channels. The averaged 
depth, tidal flow, and cross sectional area of the four 
channels are summarized in Table 1 to show the differ- 
ent levels of water flow through these pathways. Chan- 
nels depths are similar, but western channels (Middle 
Ground and Sandy Key) have higher tidal flows and 
larger cross sectional areas than the Florida Keys chan- 
nels (Table 1). Tidal fluctuations are primarily semidiur- 
nal at all stations and have weaker diurnal constituents 
(Smith, 1998, 2000). Acoustic Doppler velocity meters 
(ADVMs) that measure continuous velocity and depth 
(tide and stage) and associated CTD instruments that 
measure conductivity and temperature were installed at 
each channel in January 2002. A boat-mounted acoustic 
Doppler current profiler (ADCP) was used to calculate 
total discharge across the cross-section of the channel 
during monthly sampling (Hittle et al., 2001). 



Cnales et al : Variability in supply and cross shelf transport of Farfanlepenaeus duororum postlarvae into Florida Bay 



63 



Postlarvae were collected monthly in each channel 
during two nights around the new moon from January 
2000 to December 2002. At the PH station sampling 
began in June 2000 after the original site {Captain Key 
channel) was abandoned because it had insufficient wa- 
ter flow for effective sampling. Two moored subsurface 
channel nets (net 1 and net 2) of 0.75-m- opening, 1-mm 
mesh size, and SOO-nm mesh in the codend were sus- 
pended with floats at 0.5 m depth. Nets were deployed 
each night before dusk and removed shortly after dawn 
each day. General Oceanic flowmeters (2030R16, Low 
Speed Rotor, Miami, FL) were mounted in the mouth 
of the nets and the volume of water filtered through 
the nets was calculated for each net. Farfanlepenaeus 
duorarum postlarvae were sorted, identified, and pre- 
served in 90% ethanol. The raw catch in each sample 
was standardized to numbers of postlarvae per 1000 m'^ 
of water filtered. The average number of postlarvae over 
the two sampling nights was used as the mean month- 
ly concentration for each net. The average of monthly 
postlarval concentration for each region (northwestern 
Florida Bay vs. Florida Keys) was compared by using a 
nonparametric two-way analysis of variance (ANOVA) 
(Anderson, 2001). 

Three 12-hour experiments were conducted in sum- 
mer 2002 in the SK channel to document the behavioral 
response of pink shrimp postlarvae to ebb and flood 
tides. Consecutive pairs of night (i.e. dark) flood and 
dark-ebb plankton samples were taken hourly from 
19:00 to 07:00 h from 9 to 10 July (new moon), 23 to 
24 July (full moon), and 8 to 9 August (new moon). In 
addition, plankton samples were taken for 10 consecu- 
tive hours daily on 8 August to verify the response of 
postlarvae to light. The nonparametric Kruskal-Wallis 
test and analysis of variance (ANOVA) were used to 
determine differences in concentration of postlarvae 
between dark-ebb and dark-flood periods. 

To evaluate the possible effect of environmental vari- 
ables on larval supply to Florida Bay and the pattern 
of seasonality in postlarval concentrations, available 
time series of winds and sea surface temperature (SST) 
on the coastal shelf were examined in relation to our 
time series of monthly measured larval concentrations. 
In particular we were looking for a pattern that might 
help to determine the reason for the marked summer 
peak in the concentration of postlarvae at the western 
border of Florida Bay. Time series of hourly winds and 
sea surface temperature (SST) for the 3-year sampling 
period were obtained from the Coastal Marine Auto- 
mated Network (CMAN) station at Long Key, and from 
the Coastal Ocean Monitoring and Prediction System 
(COMP) station in NW Florida Bay (Fig. 1). Tempera- 
ture and wind time series from the Long Key CMAN 
station and the NW Florida Bay COMP station were 
highly correlated with each other (7-2 = 0.9; P<0.01). The 
longer Long Key time series were used for coastal SST 
and wind analysis. Wind speed and direction over the 
Keys, as measured at CMAN sites, are highly coherent 
(Peng et al., 1999) and useful for explaining currents 
on the SW shelf (Lee and Williams 1999). Monthly av- 



Table 1 

Mean cross section depth, peak tidal flow, and cross sec- 
tion area of the four channels sampled for pink shrimp 
iFarfantepenaeus duorarum) postlarvae at the north- 
western border of Florida Bay, Middle Ground (MG) and 
Sandy Key (SKl, and at the southeastern edge in the 
Middle Florida Keys, Whale Harbor (WH), and Panhan- 
dle Key (PH). Area was calculated at zero (m) of mean 
sea level. 



Station 

MG 
SK 
WH 
PH 



epth 


Peak tidal flow 


C 


ross area 


(m) 


(m^/sec) 




(m-) 


3.0 


1420 




2723 


3.1 


570 




1345 


3.3 


280 




407 


2.0 


30 




74 



erages of wind vectors and SST were calculated from 
hourly CMAN time series data (years 2000 to 2002) to 
examine the effect of winds and SST on the monthly 
postlarval collections. 

Time series of current data from two established sta- 
tions with moored ADCPs and temperature and salinity 
sensors were used to drive our transport model. Ini- 
tially, these data were examined to determine whether 
prevailing currents alone could explain the transport of 
larvae from the Tortugas spawning grounds to Florida 
Bay nursery grounds. These stations also were a source 
of salinity data used in one set of simulations. These 
stations were located on the inner SW shelf of the Gulf 
of Mexico, about 30 km from our MG station (Lee et al., 
2001) (Fig. 1). This array monitored coastal currents as 
part of the Florida Bay Circulation and Exchange Study 
(Lee et al., 2001). The ADCP moorings were located at 
depths of 6.4 m (mooring A=onshore) and 11.6 m (moor- 
ing B=offshore) and recorded data every 30 minutes for 
a 3-year period (A=21 September 1997 to 15 October 
2000; B=22 September 1997 to 17 October 2000). The 
two ADCP moorirr^s were about 30 km apart. Lee et 
al. (2001) reported insignificant differences between 
currents in the vertical for cross-shelf transport in the 
shallow SW Florida shelf. Wind and current vectors 
were resolved into cross-shelf (u=east [+] and west [-]) 
and alongshore ((;=north [-I-] and south [-] constituents). 
The east-west and north-south displacement of current 
and wind constituents (half-hour current data and the 
hourly wind data) was the product of each current and 
wind constituent by the respective time interval. Cor- 
relation analysis was conducted on currents and wind 
time series. 

A harmonic analysis was conducted on the three-year 
ADCP raw data to define tidal constituents and current 
magnitude. Period (Pi) and tidal excursion (Ti) were 
calculated for each constituent from amplitudes (Ai) and 
frequencies of the constituents as follows: 

Ti = AiPi/n. 



64 



Fishery Bulletin 104(1) 



A simple Lagrangian trajectory model was devel- 
oped to estimate the drift of planktonic stages. The 
model used the observed currents at ADCP moorings 
A and B to calculate trajectories. Because of the lack 
of information on spatial current variations and the 
difficulty of extrapolating vertical current profiles 
from one depth and bottom relief to other conditions, a 
simple two-dimensional (horizontal) simulation model 
was used. For the computations we selected the high- 
est bin from the ADCP that had good data throughout 
the tidal cycle at each station. This bin typically was 1 
m below the mean water level at that location and was 
the highest bin not affected by instrument sidelobe 
interference. Because instantaneous bottom currents 
were closely aligned in direction with surface currents 
and because magnitudes were only 25% lower, the 
tidal currents were barotropic and we used, therefore, 
the surface currents to estimate the largest possible 
distance traveled. 

The larval transport was calculated with the equation 

dxjdt = w,, 

where .r, = position vector of the larvae; 
t = time; and 
«, = the local current velocity. 

An Euler (forward in time) integration rule was used 
to numerically solve this equation. Because only two 
ADCPs (A and B) provided the current data for this 
large region, no attempt was made to extrapolate a 
current field from them. Simulations were run by using 
current meters A and B independently and by assum- 
ing that the current field was spatially homogeneous. 
The trajectories therefore were two-dimensional in 
the horizontal plane and the result was identical to a 
progressive vector diagram. Comparison between the 
two sets of trajectories (A and B) provided an estimate 
of the possible variability. The model was used to ex- 
plore the potential transport of planktonic stages under 
various assumptions of behavior controlled by environ- 
mental cues: 1) a behavioral response to salinity and 
light, 2) a diel behavior, 3) a diel and tidal behavior 
throughout the planktonic phase, and 4) an ontogenic 
change that began with diel behavior and added a 
tidal behavior at the 15''^ day. All four hypotheses of 
larval behavior in relation to transport were simulated 
in order that their effects on distance traveled could 
be compared and contrasted. In all simulations, we 
assumed that larvae traveled only at night. We also 
assumed that the source of the pink shrimp larvae 
was located immediately northeast of the Dry Tortugas 
about 150 km from western Florida Bay (Cummings, 
1961; Jones at al., 1970; Roberts, 1986). The program 
simulated distances traveled by particles for a period 
of 30 days (e.g., days 1-30, 31-60, 61-90, etc.), a pe- 
riod that corresponds to the estimated developmental 
period for pink shrimp from the time of hatching to the 
postlarval stage when larvae are ready for settlement 
(Dobkin, 1961; Ewald, 1965). 



Results 

Patterns of postlarval supply, SST, and winds 

The monthly influx of postlarvae through the Middle 
Florida Keys channels (WH and PH) exhibited a highly 
variable temporal pattern from year to year. Postlarvae 
were observed every month through the three years 
(Fig. 2A). Peaks of postlarvae through the Middle Keys 
channels occurred in May, July, August, and October 
2000; in January, July, and October 2001; and in Janu- 
ary, March, June, and September 2002. In contrast, the 
monthly influx of postlarvae through the northwestern 
stations (SK and MG) showed a strong seasonal pattern 
with one distinct high peak centered in summer from 
July through September for each year of the 3-year 
period (Fig. 2B). The number of postlarvae entering 
through northwestern Florida Bay was much higher 
than through the Keys stations. The mean concentration 
of postlarvae per station over the 3-year period indicated 
that concentrations of postlarvae entering northwestern 
Florida Bay through SK and MG channels were about 
eight times greater than through the Florida Keys chan- 
nels of WH and PH (Fig. 3). Results from a two-way 
ANOVA indicated that there was a significant effect 
of site (northwestern stations vs. Florida Key stations) 
and month on the supply of postlarvae entering Florida 
Bay (Table 2). 

Winds showed a seasonal pattern; the spring and 
summer were dominated by weak southeasterly winds 
and the fall and winter, by strong northerly winds 
(Fig. 2, C-D). The monthly average alongshore showed 
a weak northward constituent in spring-summer of 
each year (Fig. 2C). The monthly average cross-shelf 
winds were consistently negative (toward the west) 
and showed no seasonality (Fig. 2D). The average tem- 
perature over the three-year time series was 26.1°C, 
and winter temperatures in 2000-01 were lower than 
in 2001-02 (Fig. 2E). In summer, during the period of 
peak postlarval immigration through the northwestern 
stations (MG and SK), the alongshore wind constituent 
was mainly northward, the cross-shelf wind was west- 
ward, and the SST was above average during the three- 
year period (Fig. 2, A-E). Postlarval concentrations at 
the northwestern stations were correlated with SST and 
alongshore winds (Fig. 2, B-E; Table 3). Postlarval con- 
centrations at the Florida Keys stations (WH and PH) 
were not correlated with either winds or SST. 

Subtidal and tidal currents at the SW Florida shelf 

Advective displacement derived from two ADCP velocity 
records indicated that the net current is primarily in the 
alongshore direction (Fig. 4, A and B). The alongshore 
flow from onshore mooring A was northward, had a total 
mean velocity of 0.0062 m/sec, and a total water dis- 
placement of 589.3 x 10' m over the three years (Fig. 4A). 
The cross-shelf flow was westward, had a mean veloc- 
ity of -0.0005 m/sec, and a total water displacement of 
-289.7 X 10^ m. The alongshore flow from offshore moor- 



Cnales et al Variability in supply and cross-shelf transport of Farfantepenaeus duorarum postlarvae into Florida Bay 



65 



Florida Key stations 




J FMAI^ JJ/JJ AS0NDJFI\/1ANi1J JASONDJ Fti^AIVlJ JASOND 
2000 2001 2002 

Figure 2 

Monthly time series of pink shrimp iFarfantepenaeus duorarum ) postlarval 
concentrations, winds, and sea-surface temperature: (A) mean concentration 
of pink shrimp postlarvae at the Florida Key stations of Whale Harbor (WHi 
and Panhandle (PH); (B) mean concentration of pink shrimp postlarvae at 
the northwestern Florida Bay stations of Middle Ground (MG) and Sandy 
Key (SK); concentrations of two nets are depicted as 1 and 2; (C) alongshore 
winds d'); (D) cross-shelf winds iu); (E) sea-surface temperature (SST). 
Winds and SST data are from Long Key CMAN station. 



ing B was southward, had a mean velocity of -0.0038 
m/sec and a total water displacement of -364.7x10' m 
(Fig. 4B). The cross-shelf flow was eastward, had a mean 
speed of 0.0015 m/sec, and a total water displacement of 



145.5 X 10'^ m and a resultant southeastward current. The 
alongshore winds measured at Long Key were southward 
and had a total water displacement of 16.1 x 10^ m, and 
the cross-shelf winds were westward and had a total 



66 



Fishery Bulletin 104(1) 




SK MG 

Northwestern stations 



WH PH 

Florida Key stations 



Figure 3 

Mean concentration of pink shrimp (Farfantepenaeus duorarum ) 
postlarvae over a three-year period at each sampling station: 
northwestern Florida Bay stations — Middle Ground (MG) and 
Sandy Key (SK); Florida Key stations— Whale Harbor (WH) 
and Panhandle (PH). 



Table 2 

Results of a nonparametric two-way ANOVA of the effect 
of site and month on pink shrimp {Farfantepenaeus duo- 
rarum) postlarvae entering Florida Bay. P<0.05 is indi- 
cated with an asterisk. SS = sum of squares; MS=mean 
of squares. 



Factor 



df SS 



MS 



Month 

Site 

Month X site 

Error 



80 

1 

1 

461 



756.7 

65.9 

1191.5 

772.2 



9.5 

65.9 

1191.5 

1.7 



5.6 0.007* 

39.3 0.002* 

711.3 0.007* 



Table 3 

Correlation coefficients of pink shrimp {Farfantepenaeus 
duorarum) postlarval concentrations with sea surface 
temperature (SST), cross-shelf wind (t/), and alongshore 
wind (V) at the four sampling stations of MG= Middle 
Ground, SK= Sandy Key, WH= Whale Harbor and PH = 
Panhandle. Environmental data are from Long Key 
CMAN station. Significant correlations (P<0.05 ) are indi- 
cated with an asterisk. 





SST 


U 


V 


MG 


0.82* 


-0.025 


0.67* 


SK 


0.73* 


0.11 


0.57* 


WH 


0.15 


0.16 


0.36 


PH 


-0.17 


0.14 


-0.09 



water displacement of -86.3 x 10^ m (Fig. 4C ). The 
subtidal currents in this region are very weak as 
also observed by other investigators (Koczy et al., 
1960; Rehrer et al., 1967). Correlation coefficients 
between winds and currents in the alongshore 
direction iv) were significant for both onshore and 
offshore currents time series (Table 4). Correlation 
coefficients between winds and currents in the 
cross-shelf direction were higher in the onshore 
than in the offshore data series. This analysis 
indicated that prevailing currents, overall, were 
not favorable to passive transport of larvae from 
the Tortugas to western Florida Bay. 

A harmonic analysis of the 3-year ADCP data 
showed that semidiurnal tidal constituents (M.2, 
S.,, K.,, and N„, see Table 5 for explanation of 
abbreviations) are dominant on the SW Florida 
shelf. M., was the strongest tidal constituent and 
its east-west constituent explained 95% of the 
total current variance. The east-west amplitude 
of the M., constituent was 0.32 m/s for both moor- 
ings, and the north-south was 0.07 m/s for moor- 
ing A and 0.04 m/s for mooring B. The east-west 
amplitude (0.32 m/s) was much stronger than the 
long-term averaged subtidal cross-shelf constitu- 
ent at both stations (0.001 m/s). The east-west 
tidal excursion of the M, constituent was similar for 
both moorings, but a few meters larger on the offshore 
station (Table 5). This result indicates that it is reason- 
able to consider that there are similar tidal excursions 
on the SW shelf up to 50 km. 

Ebb versus flood catches 

Concentration of postlarvae collected hourly over a com- 
plete dark portion of a tidal cycle showed clearly that 
dark-ebb catches were negligible (<10%) by comparison 
with dark-flood catches (>90'7f ) (Fig. 5, A-C). Hourly con- 
centrations of postlarvae collected on the dark flood were 
92.7% from 9 to 10 July, 90.2% from 23 to 24 July, and 
86.9% from 8 to 9 August 2002. A nonparametric Krus- 
kal-Wallis test showed significant differences in catches 
of postlarvae between dark-ebb and dark-flood peri- 
ods (Kruskal-Wallis; H=15.5, n = 36, P<0.001; (ANOVA: 
7^=18.6; P<0.001). Samples taken during the day on 8 
August confirmed the hypothesis that postlarvae are 
present mostly at night in the water column (Table 6). 
A total of 18 postlarvae (31.5 postlarvae/1000 m^) were 
captured during 10 consecutive daylight hours, against 
2657 (3702.5 postlarvae/1000 m^) captured during 10 
consecutive hours of darkness. Although concentrations 
of postlarvae tend to increase with the tidal current, 
these differences were not significant (Kruskal-Wallis: 
//=10.3, /! = 36, P=0.5). From 9 to 10 July, no postlar- 
vae were captured during the hours of highest current 
speed (2:00 to 3:00 h), which coincided with high rain 
and winds (Fig. 5A). It is not clear whether these fac- 
tors caused a change of behavior in the postlarvae or a 
malfunction of the net. From 23 to 24 July, concentra- 
tion of postlarvae on the dark flood followed the cur- 



Cnales et al Variability in supply and cross-shelf transport of Farfontepenaeus duoroium postlarvae into Florida Bay 



67 



rent speed; highest catches occurred 
at the maximum speed and lowest 
catches at the minimum speed (Fig. 
5B). From 8 to 9 August, the highest 
peak of postlarvae occurred at the 
end of the dark-flood period when 
the current had decreased in speed 
(Fig. SCI. 

Transport simulations 

Results of channel net sampling 
showed that the greatest postlarval 
influx occurred at the northwestern 
border of the bay, where there was a 
strong seasonal pattern with maxima 
from July through September each 
year. Postlarval concentrations at 
the northwestern stations were cor- 
related with the alongshore winds 
and with surface temperature but 
did not correlate with the cross-shelf 
winds. However, postlarvae need to 
travel up to 150 km across the shelf 
to reach their nursery grounds. 

Cross-shelf transport mechanisms 
were explored by using a Lagrang- 
ian trajectory model that simulated 
the maximum distance traveled by 
planktonic stages moving at night 
for a 30-day period. Four scenarios 
of transport were modeled under dif- 
ferent assumptions of behavior: 



600x10-" 



400x10' 



200x103 . 



A 


/ 
N 


y 




Mooring A •''' 


__^ 


<V*'>*-i~'--^' " 


W ^ 




-100x10' 



-200x10-' 



-300x10' 



-400x10' • 

-500x10' 

20x10' 



Mooring B 



"-V 



r -20x10'  



-40x10'  



-60x10'  



-80x10^ 



-100x10' 




Long Key CMAN 



Scenario 1 With the assumption 
that planktonic stages (larvae and 
postlarvae) of pink shrimp respond 
to light and salinity changes (Hughes 
1969a, 1969b). the first simulation 
postulates that larvae and postlar- 
vae move horizontally with a vertical 
migratory behavior cued by changes 
in salinity. Larvae and postlarvae 
swim to the surface at night when 
an increase in salinity is detected 
and remain near the bottom when 
the salinity decreases. 

Onshore mooring The maximum distance traveled in 
the cross-shelf direction was 98 km eastward and 70% 
of larvae traveled up to 30 km (Fig. 6A). The average 
eastward distance in all simulations was 22 km. The 
maximum displacement occurred in fall-winter. Along- 
shore distances were as much as 25 km northward 
(70% of larvae) and 15 km southward (30%) (Fig. 6, A 
and B). 

Offshore mooring The maximum distances traveled 
in the cross-shelf direction was 100 km (84%^ of larvae 
traveled 40 km). Alongshore distances were as much as 
15 km northward (70%) and 20 km southward (30%) 
(Fig. 6, C and D). 



1011121 23456789 101 1121 23456789 101 1121 2345678910 
1997 1998 1999 2000 

Figure 4 

Time series of current and wind displacement. October 1997-October 2000. 
Current data are from two ADCPs located at the SW Florida shelf, and wind 
data are from Long Key CMAN station; lA) onshore ADCP mooring A, (B) 
offshore ADCP mooring B; see ADCP locations in Figure 1. (C) Wind displace- 
ment. Alongshore constituents are represented by interrupted lines (N=north 
l-^], S=south [-]) and cross-shelf by continuous lines (E = east [+], W=west [-]). 



Scenario 2 The second simulation assumes that plank- 
tonic stages travel at night using the instantaneous 
current. 

Onshore mooring Distances traveled in the cross- 
shelf direction reached a maximum of 68 km and 75% 
of larvae traveled only 30 km eastward (Fig 6A). The 
average eastward distance in all simulations was 15 km. 
The maximum displacement occurred in spring-summer 
(May to September 1998) during the first year and in 
fall-winter (October 1999 to February 2000) during the 
second and third year. Distances recorded in the along- 
shore direction were as far as 40 km northward (48% of 
larvae) and 30 km southward (52%) (Fig. 6, A and B). 



68 



Fishery Bulletin 104(1) 



Offshore mooring Maximum distance traveled in 
the cross-shelf direction reached 60 km eastward and 
80% of larvae traveled less than 30 km. Distances re- 
corded in the alongshore direction were as far as 40 
km northward (51%) and 40 km southward (49%) (Fig. 
6, C and D). 

Scenario 3 Under the assumption that pink shrimp 
larvae and postlarvae migrate vertically in a tidal cycle, 
the third simulation postulates that larvae and post- 
larvae swim in the water column near the surface at 
night during the flood tide and remain near the bottom 
during the ebb tide. In this simulation it is assumed 
that planktonic stages move by using the eastward cur- 



Table 4 

Correlation coefficients of wind and current components. 
U and V are the east-west and north-south components 
respectively. Nearsurface currents are from onshore (A) 
and offshore (B) ADCP moorings over the SW Florida 
shelf Wind data are from Long Key CMAN station. Signif- 
icant correlations (P<0.05l are indicated with an asterisk. 



Mooring A 



Mooring B 



onshore 



offshore 



[/-current V-current f7-current V-current 



{/-wind 
V-wind 



-0.25* 
0.26* 



-0.22* 
0.55* 



-0.10 
0.12 



-0.06 
0.60* 



rent (flood tide) during the postulated 30 days of larval 
development. 

Onshore mooring The maximum distance traveled in 
the cross-shelf direction was 200 km eastward and 86% 
of the larvae exceeded 150 km. The average eastward 
distance in all simulations was 132 km. The maxi- 
mum larval displacement occurred from December 1999 
through March 2000. Distance traveled in the along- 
shore direction was 45 km northward (70%) and 5 km 
southward (30%) (Fig. 6, A and B). 

Offshore mooring The maximum larval displace- 
ment in the cross-shelf direction was 200 km eastward 
and 85% of the larvae reached 150 km. The maximum 
eastward displacement occurred in fall and in winter. 
Distance traveled in the alongshore direction was as 
much as 40 km northward (82%) and 5 km southward 
(18%). The maximum distance traveled was recorded in 
March-April and June 2000 (Fig. 6, C and D). 

Scenario 4 In this simulation, it is assumed that there 
is a change in behavior for pink shrimp larvae — an 
assumption similar to the one taken in simulations for 
some Australian penaeid species (Rothlisberg, 1982; 
Rothlisberg et al., 1995, 1996). Early larval stages (pro- 
tozoea and myses) migrate vertically in a diel cycle and 
there is no cross-shelf displacement during the first 15 
days of development. Later in development, postlarvae 
migrate by using tidally induced behavior superim- 
posed on the diel behavior for the remaining 15 days 
of planktonic development, and the eastward current 
(flood tide). 

Onshore mooring The maximum displacement of 
larvae in the cross-shelf direction was 100 km eastward 











Table 5 












Harmonic analysis results conducted on 


three 


years of ADCP data 


September 1997 


-October 2000). Moorings 


were located 


at the SW Florida shelf: A (onshore) and B (offshore) moorings. 


U and V are the east-west and north-south tidal consistuents. 


respectively. Explained 


variance of constituent U 


was 95<:{ and of V 


was 50% (for A), 


and ge-Zr and 29% 


(for 


B), 


respectively. 


M2 = semidiurnal lunar: 


S2= semidiurnal solar; 


N2 


=semidiurna 


larger lunar elliptic; Kl-diurnal lunisolar 


Ml 


= diurnal smaller | 


lunar elliptic; 01 = diurnal principal lunar 


,K2= 


semidiurnal lunisolar. 












Tidal 


Period 






Tidal constituent U 


Tidal constituent V 


Amplitude 


Tidal excursion 


Amplitude 




Tidal excursion 


constituent 


(h) 






(m/sec) 




(m) 


(m/sec) 






(ml 


A M2 


12.42 






0.321 




4561.7 


0.070 






993.5 


N2 


12.66 






0.056 




810.8 


0.013 






191.5 


S2 


12.00 






0.100 




1368.2 


0.017 






239.3 


Kl 


23.93 






0.049 




1332.9 


0.015 






408.7 


Ml 


24.84 






0.005 




139.5 


0.001 






31.3 


01 


25.82 






0.039 




1139.1 


0.010 






289.9 


K2 


11.97 






0.025 




345.6 


0.010 






135.8 


B M2 


12.42 






0.323 




4593.0 


0.038 






539.4 


N2 


12.66 






0.057 




828.3 


0.007 






105.9 


S2 


12.00 






0.105 




1439.7 


0.009 






129.3 


Kl 


23.93 






0.052 




1434.4 


0.008 






213.9 


Ml 


24.84 






0.003 




93.9 


0.002 






54.1 


01 


25.82 






0.043 




1278.1 


0.002 






68.0 


K2 


11.97 






0.023 




319.5 


0.005 






63.1 



Cnales et al.: Variability m supply and cross-shelf transport of Foi fantepenaeus duorarum postlarvae into Florida Bay 



69 



-40 



and 86% of larvae reached 80 km. The aver- 
age eastward distance traveled in Simula- 60 
tions was 66 km. The maximum distances 
traveled in the alongshore direction was 25 40  
km northward (70% of larvae) and 10 km 
southward (30%). The maximum distance 20 
traveled was recorded in March-April and 
June 2000 (Fig. 6, A and B). 

Offshore mooring The maximum displace- 
ment of larvae in the cross-shelf direction -20 
was 90 km eastward and 85% of larvae 
reached 80 km. Distance traveled in the 
alongshore direction reached 20 km north- 
ward (86% of larvae) and 5 km southward 
(14%) (Fig. 6, C and D). 



Discussion 

The monthly influx of pink shrimp postlar- 
vae entering Florida Bay through its north- 
western border showed a strong seasonal 
pattern of annual high peaks in summer 
over the 3-year period. Postlarval concentra- 
tions were correlated with alongshore winds 
and sea surface temperature. Alongshore 
winds were seasonal, with a weak northward 
constituent in spring-summer of each year 
changing to a strong southward in fall and 
winter. This seasonal pattern agrees with 
the general circulation described for the SW 
Florida shelf, highly dependent on synoptic- 
scale winds, coupled with a strong seasonal- 
ity and strong tidal currents (e.g., Weisberg 
et al.. 1996; Wang, 1998; Smith, 2000; Lee 
et al., 2001). Although tidal currents seem 
to be the main vehicle of eastward transport 
for planktonic stages, alongshore winds may 
be fundamental for moving larvae northward 
along the SW Florida shelf by avoiding a 
drift with the Florida Current or with the 
cyclonic circulation of the gyres that form 
southwest of the Dry Tortugas (Lee et al., 
1994; Fratantoni et al., 1998). Under these 
circumstances larvae may reach Florida Bay 
by the shortest route in summer using tidal 
currents and winds across the shallow SW 
Florida shelf and entering Florida Bay by 
its northwestern border. The summer sea- 
sonality of postlarval immigration may also 
be amplified by the seasonality of spawning 
because higher temperatures induce higher 
spawning activity (Cripe, 1994) and conse- 
quently more recruits to enter the Bay during 
favorable onshore conditions. 

Another alternative explanation for the summer lar- 
val immigration is that larvae may take advantage of 
a distinct annual tidal cycle produced in summer every 
year as a result of the interaction of the periods of the 
diel vertical migration and the tidal constituent (S.,, 



A 

Flood 


i 


° 
Storm 


^ 
ll 

^ ^ 


• 

new 


Ebb ^^3/ 


^ 







July 9 -10, 2002 



- 1 



- -1 




19h 20h 21ti 22ti 23h 24ti 



n 



Figure 5 

Hourly concentration of pink shrimp (Farfantepenaeus duorarum) 
postlarvae at Sandy Key station (SK) during a complete dark (night) 
tidal cycle conducted during (A) new moon, 9-10 July 2002, (B) full 
moon, 23-24 July 2002, and (C) new moon, 8-9 August 2002. Right 
y-axis indicates concentration of postlarvae/m''; lefty-axis is tidal 
current speed (cm/sec) measured by acoustic Doppler velocity meter 
(filled circle and solid line) and by flowmeter (empty circles and 
interrupted line). Positive values in they-axes indicate transport into 
Florida Bay (flood), and negative values transport out of the Bay (ebb). 
Horizontal bars on the bottom indicate hours of darkness versus light. 



K,) with the annual cycle of the length of the night 
(Criales et al., 2005). Larvae moving vertically in the 
water column with a diel behavior can be transported 
up to 70 km onshore in summer because the eastward 
current of the tidal constituents matches the diel cycle 
over extended intervals in the shorter summer nights 



70 



Fishery Bulletin 104(1) 









Table 6 








Data from ebb-flood experiment conducted at 
D = dark. VWF = vohime of water filtered. 


Sandy Key (SK) station during 20 


consecutive hours. 


E = ebb, F=flood, L = light. 


Date 


Collection 
time 


Tide 
stage 


Light vs. 
dark 


VWF 

(m3) 


Number of 
postlarvae 


Concentrations 
(postlarval m'l 


08/08/2002 


11:00 


E 


L 


395.8 


7 


17.7 


08/08/2002 


12:00 


E 


L 


535.3 


2 


3.7 


08/08/2002 


13:00 


F 


L 


1360.1 





0.0 


08/08/2002 


14:00 


F 


L 


1262.3 





0.0 


08/08/2002 


15:00 


F 


L 


971.3 


8 


8.2 


08/08/2002 


16:00 


F 


L 


531.2 


1 


1.9 


08/08/2002 


17:00 


F 


L 


0.0 





0.0 


08/08/2002 


18:00 


F 


L 


0.0 





0.0 


08/08/2002 


19:00 


E 


L 


450.6 





0.0 


08/08/2002 


20:00 


E 


L 


575.6 





0.0 


08/08/2002 


21:00 


E 


D 


530.9 


5 


9.4 


08/08/2002 


22:00 


E 


D 


477.6 


14 


29.3 


08/08/2002 


23:00 


E 


D 


60.7 


7 


115.4 


08/09/2002 


0:00 


E 


D 


263.3 


87 


330.4 


08/09/2002 


1:00 


F 


D 


648.6 


141 


217.4 


08/09/2002 


2:00 


F 


D 


944.3 


247 


261.6 


08/09/2002 


3:00 


F 


D 


1164.1 


491 


421.8 


08/09/2002 


4:00 


F 


D 


977.5 


529 


541.2 


08/09/2002 


5:00 


F 


D 


737.8 


1006 


1363.6 


08/09/2002 


6:00 


F 


D 


315.1 


130 


412.5 



(Criales et al., 2005). Therefore, pink shrimp and other 
fish and invertebrate species that use tidal currents 
for transport with a daily vertical behavior may take 
advantage of this annual tidal cycle to improve their 
chances of reaching coastal nursery habitats. 

The monthly influx of postlarvae through the Mid- 
dle Florida Keys channels exhibited a highly variable 
seasonal pattern from year to year and from station 
to station. This section of the Keys coastal waters is 
frequented by coastal cyclonic eddies that originate 
at the Dry Tortugas and move downstream at 5-17 
km/day along the edge of the shelf at intervals of 1 to 
3 months (Fratantoni et al., 1998; Yeung et al., 2001). 
For planktonic stages, these eddies may serve as a 
delivery mechanism from offshore spawning grounds 
to the southeastern border of Florida Bay, allowing an 
onshore transport by the coastal countercurrent flow 
generated by the cyclonic circulation (Lee et al., 2001; 
Yeung et al., 2001; Criales et al., 2003). Episodic, me- 
soscale events associated with boundary current fronts 
and eddies may cause high variability in transport (Lee 
et al., 1994; Limouzy-Paris et al., 1997; Yeung and Lee, 
2002). This variability was reflected in the influx of 
pink shrimp postlarvae through Middle Keys channels 
(Criales et al., 2003). The influx of postlarvae at WH 
and PH channels in the present study was also highly 
variable and there was no correlation with winds or sea 
surface temperature. Therefore, we hypothesized that 



the high variability in postlarval influx detected at 
the Florida Keys channels reflects the temporal and 
spatial variability associated with the passage of coastal 
eddies. 

Simulations of transport based on current observa- 
tions indicated that passive larvae could not consistent- 
ly be advected the estimated 150 km eastward across 
the shelf between spawning and nursery grounds in 
30 days. This is true primarily because of the weak 
eastern current and the reversing nature of tides. In 
contrast, planktonic stages (larvae and postlarvae) mov- 
ing at night with the eastward current (flood tide) can 
consistently travel 100 to 200 km in 30 days. Hypotheti- 
cally, 85% of the larvae can be transported far enough 
from the known spawning grounds of Dry Tortugas to 
the nursery grounds in western Florida Bay in 30 days. 
Previous works conducted inside Florida Bay (Tabb et 
al., 1962; Roessler and Rehrer, 1971) and our own data 
obtained along the western border of Florida Bay dem- 
onstrated the ability of pink shrimp postlarvae to re- 
spond to the dark flood tide and to distinguish between 
day and night. Over 90% of postlarvae were caught in 
the dark flood period and only a few postlarvae were 
caught during daylight hours. This behavior needs to be 
investigated for early larval stages to define the exact 
age at which larvae begin reacting to change in tides 
and to environmental cues that trigger the vertical 
movement in relation to tidal stage. For other penaeid 



Cnales et al Variability in supply and cross shelf transport of Farfontepenaeus duorarum postlarvae into Florida Bay 



A Onshore ADCP A, Cross-shelf (+East. -West) 






 ■♦QCioooS 






-•♦••OOOOOOO- 



-r-T — I — ^ — r— 1 — t I 1 I I I I I I I I — I I I I I — n — rn — i i i — i — i i i i i — i i i 

9 10 11 i; 1 : 3 J 5 6 7 8 9 10 11 12 I 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 



C Offshore ADCP B, Cross-shelf (+East, -West) 






▼ T * 



»▼▼▼- 



^^0^0 TTT ▼ AA 



▼ T 



^aaa^SOa ••• « gd^^'^ "^^aaA 



.-88ooooflfil 2ooooa* 



'.^^^^m^ 



oo- 



I I I I I I I I I I I I I I I I I I I 1 I I — n — n — I I I I I I I I I I I I 

9 10 11 12 1 234 56789 10 11 12 t 234 56789 10 11 12 1 234 56789 10 



1997 



1998 



1999 



2000 



1997 



1998 



1999 



2000 



B Onshore ADCP A, Alongshore (-i-North, -South) 



▼oQ ' 






SjjJi 



'^ 



«o%^ 8 ^5 'a 



o 
ooo 

• •• • 



8 •^ 8 



.• • 






I I I I I I I I I I I I 1 I I I I I I I I I I I I I I I I I I I I 

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



D Offshore ADCP B, Alongshore (+North, -South) 



T^ T^ O 



?▼» _ -TAA^A, 






 8. 



o o 
o 



1997 



1998 



1999 



2000 



9 10 11 12 1 234 56789 10 11 12 1 234 56789 10 11 12 1 234 56789 10 

1997 1998 1999 2000 



• 


scenario 1 


o 


scenario 2 


▼ scenario 3 | 


A 


scenario 4 | 



Figure 6 

Outputs of transport simulations of pink shrimp iFarfantepenaeus duorarum) planktonic stages (larvae and postlarvae) 
developed from in situ current observations and under different scenarios of larval behavior. Each point represents 30 
days of travel from September 1997 to October 2000; (A and Bi outputs from onshore ADCP A currents, cross-shelf and 
alongshore constituents, respectively; (C and D) outputs from offshore ADCP B currents, cross-shelf and alongshore 
constituents, respectively. Scenario 1 (filled circles) = planktonic stages migrating with the instantaneous current and 
a diel behavior (at night) cued by changes in salinity; scenario 2 (empty circles) = planktonic stages migrating with the 
instantaneous current and a diel behavior (at night); scenario 3 (filled triangles) = planktonic stages migrating with a 
diel (at night) and a tidal (eastward current) behavior; scenario 4 (empty triangles) = planktonic stages migrating by an 
induced ontogenic change of behavior (a diel behavior during the first 15 days and a tidally induced behavior (eastward 
current) superimposed on the diel behavior (at night) for the remaining 15 days). 



species (Fenneropenaeus merguiensis, Penaeus plebejus. 
and Penaeus semisulcatus) in the Gulf of Carpentaria, 
Australia, an ontogenic change in behavior has been 
(iocumented for planktonic stages (Rothlisberg, 1982; 
Rothlisberg et al., 1995. 1996; Condie et al., 1999). 
Results of these studies showed that during the first 
two weeks of planktonic development, larvae migrate 
vertically in a diel cycle; later in development, the verti- 
cal migration is tidally induced. Under those conditions 
there was little or no systematic cross-shelf displace- 
ment of larvae during the first two weeks; afterward, 
cross-shelf displacement occurred rapidly (Rothlisberg 
et al., 1995; 1996). Simulation of transport for pink 
shrimp larvae using this ontogenic change of behavior 



indicated that planktonic stages can be transported up 
to 100 km eastward and 85% of larvae reach 80 km. If 
this behavior applies to pink shrimp, the location of the 
spawning grounds needs to be reconsidered in favor of 
areas closer to Florida Bay. Alternatively, the behavior 
of early planktonic stages may have been underesti- 
mated. Results of this research indicated once again 
the extreme importance of defining larval behavior and 
including behavior in dispersal models. 

The cue(s) that penaeid postlarvae use to migrate in 
the water column during the flood tide and to return 
to the bottom on the ebb tide are not completely un- 
derstood and could be species specific. Hughes (1969a, 
1972) demonstrated in laboratory experiments that 



72 



Fishery Bulletin 104(1) 



pink shrimp postlarvae react to changes in salinity by 
changing swimming direction. Postlarvae were more 
active in the water column with increases in salinity, 
and this finding implies a shoreward displacement with 
the flood tide. Similar responses to salinity changes 
have been found for postlarvae oi Farfantepenaeus cali- 
forniensis, Farfantepenaeus brevirostris, Litopenaeus 
stylirostris, and Litopenaeus vannamei from the Mexi- 
can Pacific (Mair, 1980). The importance of a salinity 
cue for transport has been questioned for other penaeid 
species that inhabit hypersaline estuaries (southeast 
African, western Australia) in which penaeid postlar- 
vae would need to move against a salinity gradient 
(Penn, 1975; Forbes and Benfield, 1986; Rothlisberg 
et al., 1995). Our simulations of transport guided by 
salinity changes indicated that planktonic stages could 
travel distances in the range of only 30 km in 30 days. 
This result may indicate that salinity is not the only 
environmental factor controlling long cross-shelf migra- 
tions of pink shrimp. However, Hughes (1969a, 1969b) 
in early experiments suggested that a salinity cue 
could apply to postlarvae near the nursery grounds. 
Changes in water pressure have been proposed as the 
only environmental factor that triggers the vertical 
migration of postlarval shrimps in the Gulf of Car- 
pentaria, Australia (Penn, 1975; Forbes and Benfield, 
1986; Rothlisberg et al., 1995). Laboratory experiments 
and numerical models have shown that tiger shrimps 
iPenaeus semisulcatus and Penaeus esculentus) larvae 
switch behavior when the change in water pressure 
with tides becomes a significant fraction of the total 
pressure (Rothlisberg et al., 1996; Condie et al., 1999; 
Vance and Pendrey''^). This behavior only occurred in 
larvae above a certain size. However, it still remains to 
be determined whether, in a natural ecological context, 
the rates of relative changes of pressure are consistent 
with the tidal cycle periodicity and are detectable at 
absolute amounts in order to permit a behavioral re- 
sponse. 

By means of simulations of transport, we have identi- 
fied a potential STST mechanism for planktonic pink 
shrimp to migrate the estimated 150 km in 30 days 
from spawning to nursery grounds over the Florida 
shelf Organisms inhabiting coastal ecosystems domi- 
nated by tides have the potential to control their cross- 
shelf movement through STST (Shanks, 1995). The 
extent of the transport depends on the speed of the 
tidal current and the time that organisms spend in the 
water column. Success in reaching the nursery grounds 
depends upon the stage in larval or postlarval develop- 



^ Vance D. J., and R. C. Pendrey. 2001. Vertical migration 
behaviour of postlarval penaeid prawns: a laboratory study 
of the effect of tide, water depth and day/night. In The 
definition of effective spawning stocks of commercial tiger 
prawns in the Northern prawn fishery and king prawns in the 
eastern king prawn fishery-behaviour of postlarval prawns, 
p. 28-52. Fisheries Research and Development Corporation 
(FRDC ) Final Report ( Project 97/108 ). 68 p. CSIRO Marine 
Research Laboratories, P.O. Box 120, Cleveland, Qld. 4163, 
Australia. 



ment when the tidal behavior is added. A dependence 
on tidal currents for the entire larval transport period 
was postulated for Melicertus latisultacus in Western 
Australia (Penn, 1975). The shrimp Lucifer faxoni is 
the only species for which a STST mechanism across 
the shelf has been shown (Woodmansee, 1966). Strong 
tidal currents and several coastal sources of fresh wa- 
ter define the spawning grounds of the wide and shal- 
low SW Florida shelf (Lee et al., 2001; Jurado, 2003). 
Under these conditions, parts of the SW Florida shelf 
may behave as an estuary in which planktonic pink 
shrimp may easily recognize tides by means of endog- 
enous behavior or environmental variables. 

From this study we determined that the greatest 
influx of postlarval pink shrimp occurred at the north- 
western border of Florida Bay in summer. Postlarvae 
entering Florida Bay through the channels of the Mid- 
dle Florida Keys occurred at a much lower magnitude 
and there were only sporadic peaks and no appar- 
ent seasonality in influx. The transport mechanism 
of planktonic stages of pink shrimp across the SW 
Florida shelf seems to depend heavily on semidiurnal 
tides and larval behavior, and much less on seasonal 
winds. The response of postlarvae to the tidal currents 
was clearly observed at the western margin of Florida 
Bay. Such behavior needs to be explored in early stages 
to define the age at which larvae begin to respond to 
tides, the location on the Florida shelf at which this 
response occurs, and the specific environmental cues 
linked to such behavior. With this information, more 
realistic simulations of transport can be made with a 
complete hydrodynamic model that would incorporate 
spatial and vertical variations in currents. Depend- 
ing on the resulting transport, the location of spawn- 
ing grounds may be better defined, leading to better 
protection of this valuable fishery resource. Informa- 
tion on recruitment variability and key environmen- 
tal factors affecting larval transport are essential to 
accurately interpret stock assessments, maintain the 
ecological integrity of both spawning grounds and 
nursery grounds, and to effectively manage the pink 
shrimp fishery. 



Acknowledgments 

We are especially grateful to Thomas Lee and Elizabeth 
Williams (RSMAS, University of Miami), Elizabeth 
Johns, Ryan Smith, Shailer Cummings, and Nelson Melo 
(NOAA/AOML, Miami), and Ned Smith (Harbor Branch 
Oceanographic Institute) for providing ADCP data and 
valuable comments; to William Richards (NMFSC/ 
NOAA Miami) and Robert Cowen (RSMAS, University 
of Miami) for their constructive comments and support; 
to Hernando Cardenas (NMFSC) for his assistance 
sorting plankton samples; to Andre Daniels (USGS, 
Miami), for his valuable support of fieldwork; and to 
the personnel involved in collecting and managing data 
from C-MAN and COMP stations. This research was 
funded by the NOAA South Florida Ecosystem Resto- 



Criales et al Variability in supply and cross shelf transport of Forfantepenaeus duoraium postlarvae into Florida Bay 



73 



ration Prediction and Modelling (SFERPM) program 
through a cooperative agreement between Southeast 
Fisheries Science Center, USGS and CIMAS, RSMAS. 
University of Miami, and the DOI Critical Ecosystems 
Studies Initiative of the Everglades Restoration Pro- 
gram. The sampling that supported this research was 
conducted pursuant to National Park Service Permit 
no. EVER-2003-SCI-0109 and Florida Fish and Wildlife 
Conservation Commission Special Activity License no. 
03SR-502. 



Literature cited 

Allen, D. M., J. H. Hudson, and T. J. Costello. 

1980. Postlarval shrimp iPenaeus) in the Florida Keys: 

species, size, and seasonal abundance. Bull. Mar. Sci. 

30:21-33. 
Anderson, M. J. 

2001. A new method for non-parametric multivariate 
analysis of variance. Aust. Ecol. 26: 32-46. 

Browder, J. A., Z. Zein-Eldin, M. M. Criales, M. B. Robblee, 
S. Wong, T. L. Jackson, and D. Johnson. 

2002. Dynamics of pink shrimp (Forfantepenaeus duor- 
ariim) recruitment potential in relation to salinity and 
temperature in Florida Bay. Estuaries 25:1355-71. 

Condie S. A., N. R. Loneragan, and D. J. Die. 

1999. Modelling the recruitment of tiger prawns Penaeus 
esculentus and P. semisulcatus to nursery grounds in the 
Gulf of Carpentaria, northern Australia: implications 
for assessing stock-recruitment relationships. Mar. 
Ecol. Prog. Ser. 178:55-68. 

Cowen, R. B. 

2002. Larval dispersal and retention and consequences for 
population connectivity. In Coral reef fishes: dynamics 
and diversity in a complex ecosystem (P. Sale, ed.), p. 
149-170. Academic Press, London. 

Criales, M. M., M. J. Bello, and C. Yeung. 

2000. Diversity and recruitment of penaeoid shrimps 
at Bear Cut, Biscayne Bay, Florida. Bull. Mar. Sci. 
67:773-788. 

Criales, M. M., and T. N. Lee. 

1995. Larval distribution and transport of penaeoid 
shrimps during the presence of the Tortugas Gyre in 
May-Junel991. Fish. Bull. 93:471-482. 
Criales, M. M., and M. F. McGowan. 

1994. Horizontal and vertical distribution of penaeidean 
and caridean larvae and micronektonic shrimps in the 
Florida Keys. Bull. Mar. Sci. 54:843-856. 
Criales, M. M., J. Wang, J. A. Browder, and M. B. Robblee. 

2005. Tidal and seasonal effect of transport of pink shrimp 
postlarvae. Mar. Ecol. Prog. Ser. 286:231-238. 
Criales, M. M., C. Yeung. D. Jones. T. L. Jackson, and W. J. 
Richards. 

2003. Variation of oceanographic processes affecting 
the size of pink shrimp (Farfantepenaeus duorarum i 
postlarvae and their supply to Florida Bay. Estuar. 
Coast. Shelf Sci. 57 (3): 457-468. 

Cripe, G. M. 

1994. Induction of maturation and spawning of pink 
shrimp, Penaeus duorarum, by changing water tempera- 
ture, and survival and growth of young. Aquaculture 
128:255-260. 
Cummings, D. C. 

1961. Maturation and spawning of pink shrimp, Penaeus 



duorarum Burkenroad. Trans. Am. Fish. Soc. 90: 
462-468. 
Dall, W., B. J. Hill, P. C. Rothiisberg, and D. J. Staples. 

1990. The biology of Penaeidae. Adv. Mar. Biol. 27, 
489 p. 
Dobkin, S. 

1961. Early development stages of pink shrimp, Penaeus 
duorarum. from Florida waters. Fish. Bull. 61: 
321-349. 
Eldred, J. W., G. T. Martin, and E. A. Joyce Jr. 

1965. Seasonal distribution of penaeid larvae and postlar- 
vae of the Tampa Bay area, Florida. Fl. Board Conserv. 
Mar. Res. Lab., Tech. Ser. 44, 47 p. 
Ewald, J. J. 

1965. The laboratory rearing of pink shrimp, Penaeus 
duorarum Burkenroad. Bull. Mar. Sci. 15(2):436- 
449. 
Forbes, A. T.. and M. C. Benfield. 

1986. Tidal behaviour and post-larval penaeid prawns 
(Crustacea: Decapoda: Penaeidae) in a southeast African 
estuary. J. Exp. Mar. Biol. Ecol. 102:23-34. 
Forward, R. B., and R. A. Tankersley. 

2001. Selective tidal-stream transport of marine ani- 
mals. Oceanog. Mar. Biol. 39:305-353. 
Fratantoni, P S., T. N., Lee, G. P. Podesta, and F Muller-Karger. 
1998. The influence of Loop Current perturbations 
on the formation and evolution of Tortugas eddies 
in the southern Straits of Florida. J. Geophys. Res. 
103:24759-79. 
Garcia, S., and L. Le Reste. 

1981. Life cycles, dynamics, exploitation and management 
of coastal penaeid shrimp stocks. FAO Tech. Paper 
203, 215 p. FAO, Rome. 
Hittle, C, E. Patino, and M. Zucker. 

2001. Freshwater flow from estuarine creeks into north- 
eastern Florida Bay. USGS Water-Resources Invest. 
Rep. 01-4164, 32 p. 
Hughes, D. A. 

1969a. Responses to salinity change as a tidal trans- 
port mechanism of pink shrimp, Penaeus duorarum 
Burkenroad. Biol. Bull. 136:45-53. 
1969b. Evidence for the endogenous control of swim- 
ming in pink shrimp, Penaeus duorarum. Biol. Bull. 
136:398-404. 
1972. On the endogenous control of tide-associated 
displacements of pink shrimp, Penaeus duorarum 
Burkenroad. Biol. Bull. 142:271-280. 
Iversen, F. S., A. E. Jones, and C. P. Idyll. 

1960. Size distribution of pink shrimp, Penaeus duor- 
arum, and fleet concentrations on the Tortugas fishery 
grounds. U.S. Fish. Wild. Ser. Spec. Sci. Rep. 356: 
1-62. 
Jones, A. C., D. E. Dimitriou, J. J. Ewald, and J. H. Tweedy. 

1970. Distribution of early developmental stages of pink 
shrimp, Penaeus duorarum. in Florida water. Bull Mar. 
Sci. 20:634-661. 
Jurado, J. L. 

2003. The dynamics of diatom blooms and silicon cycling 
in coastal waters of the southwestern Florida shelf and 
northwestern Florida Bay. Ph.D. diss, 268 p. Univ. 
Miami, Miami, FL. 
Klima E. F., G. A. Matthews, and F. J. Patella. 

1986. Synopsis of the Tortugas pink shrimp fishery, 1960- 
1983, and the impact of the Tortugas Sanctuary. North 
Am. J. Fish. Manag. 6:301-310. 



74 



Fishery Bulletin 104(1) 



Koczy, F. F., M. O.Rinkel, and S. J. Niskin. 

1960. The current patterns on the Tortugas shrimp 
grounds. Proc. Gulf Carib. Fish. Inst. 12"' Annual 
Sess., Nov. 1959:112-125. 
Lee, T. N. M. E. Clarke. E. Williams, A. F. Szmant, and 
T. Berger. 

1994. Evolution of the Tortugas Gyre and its influence 
on recruitment in the Florida Keys. Bull. Mar. Sci. 
54:621-646. 

Lee, T. N, E. Johns, D. Wilson, E. Williams, and N. P. Smith. 
2001. Transport processes linking south Florida coastal 
ecosystems. In Linkages between ecosystems in the 
South Florida hydroscape (J. Porter and K. Porter, eds.), 
p 309-342. CRC Press, Boca Raton, FL. 
Lee, T. N., and E. Williams. 

1999. Mean distribution and seasonal variability of 

coastal currents and temperature in the Florida Keys 

with implications for larval recruitment. Bull Mar. 

Sci. 64:35-56. 

Limouzy-Paris, C, H. C. Graber, D. L. Jones, A. W. Ropke. and 

W. J. Richards. 

1997. Translocation of larval coral reef fishes via sub- 
mesoscale spin-off eddies from the Florida Current. Bull. 
Mar. Sci. 60: 966-83. 
Mair, J. McD. 

1980. Salinity and water-type preferences of four species 
of postlarval shrimps iPenaeus) from west Mexico. J. 
Exp. Mar. Biol. Ecol. 45:69-82. 
Munro, J. L., A. C. Jones, and D. Dimitriou. 

1968. Abundance and distribution of the larvae of the 
pink shrimp iPenaeus duorarum] on the Tortugas Shelf 
of Florida. August 1962-October 1964. Fish. Bull. 
67:165-181. 
Nance, E. F, and F. J. Patella. 

1989. Review of the Tortugas shrimp fishery from May 
1987 to January 1989. NOAA Tech. Memo., NMFS- 
SEFC-238, 11 p. 
Peng, G., C. N. K. Mooers, and H. Graber. 

1999. Coastal winds in South Florida. J. Appl. Meteor. 
38:1740-1757. 
Penn, J. W. 

1975. The influence of tidal cycles on the distribu- 
tion pathway of Penaeus latisulcatus Kishinouye in 
Shark Bay, Western Australia. J. Mar. Freshw. Res. 
26:93-102. 
Rehrer, R., A. C. Jones, and M. A. Roessler. 

1967. Bottom water drift on the Tortugas Grounds. Bull. 
Mar. Sci. 17:562-575. 
Roberts, T. W. 

1986. Abundance and distribution of pink shrimp in and 
around the Tortugas Sanctuary, 1981-1983. North 
Am. J. Fish. Manag. 6:311-327. 
Roessler M. A., and G. Rehrer. 

1971. Relation of catches of postlarval pink shrimp in 
Everglades National Park, Florida, to the commercial 
catches on the Tortugas Grounds. Bull. Mar. Sci. 21 
(4):790-805. 
Rothlisberg, P. C. 

1982. Vertical migration and its effect on dispersal 
of penaeid shrimp larvae in the Gulf of Carpentaria, 
Australia. Fish. Bull. 80:541-554. 
Rothlisberg, P. C, J. A. Church, and C. Fandry. 

1995. A mechanism for near-shore density and estua- 
rine recruitment of post-larval Penaeus plebejus Hess 



(Decapoda, Penaeidae). Estuar. Coast. Shelf Sci. 40: 
115-138. 

Rothlisberg, P. C. J. A. Church, and A. M. G. Forbes. 

1983. Modelling advection of vertically migrating shrimp 
larvae. J. Mar. Res. 41:511-538. 

Rothlisberg, P. C, P. D. Craig, and J. R. Andrewartha. 

1996. Modelling penaeid prawn larval advection in Alba- 
tross Bay, Australia: defining the effective spawning 
population. Mar. Freshw. Res. 47:157-168. 

Shanks, A. L. 

1995. Mechanisms of cross-shelf dispersal of larval inver- 
tebrates and fish. In Ecology of marine invertebrate 
larvae (L. R. McEdward, ed.), p. 323-367. CRC Press, 
Inc., Boca Raton, FL. 

Sheridan. P. 

1996. Forecasting the fishery for pink shrimp Penaeus 
duorarum, on the Tortugas Grounds, Florida. Fish. 
Bull. 94:743-755. 

Smith, N. P 

1998. Tidal and long-term exchanges through channels 
in the Middle and Upper Florida Keys. Bull. Mar. Sci. 
62:199-211. 

2000. Transport across the western boundary of Florida 
Bay. Bull. Mar. Sci. 66:291-304. 

Sponaugle, S., R. K. Cowen, A Shanks, S. G. Morgan, J. M. Leis, 
J. Pineda, G. W. Boehlert. M. J. Kingsford, K.C. Linderman, 
C. Grimes, and J. L. Munro. 

2002. Predicting self-recruitment in marine populations 
biophysical correlations and mechanisms. Bull. Mar. 
Sci. 70(1) suppl.:341-375. 
Tabb, D. C, D. L Dubrow, and A. E. Jones. 

1962. Studies on the biology of pink shrimp Penaeus 
duorarum Burkenroad, in Everglades National Park, 
Florida. Fla. Board Conser. Tech. Ser. 37:1-32. 
Temple, R., and C. C. Fischer. 

1965. Vertical distribution of the planktonic stages of 
penaeid shrimp. Inst. Mar. Sci. Publ. 10:59-67. 

Wang, J. D. 

1998. Subtidal flow patterns in western Florida Bay. Es- 
tuarine Coast. Shelf Sci. 46:901-915. 
Weisberg, R. H., B. D. Black, and J. Yang. 

1996. Seasonal modulation of the West Florida continental 
shelf circulation. J. Geophys. Res. 23:2247-2250. 
Wenner, E. L., D. M. Knott, C. A. Barans, S. Wilde, J. O. Blanton, 
and J. Amft. 

2005. Key factors influencing transport of white shrimp 
{Litopenaeus setiferus) postlarvae into the Ossabaw Sound 
system, Georgia, USA. Fish. Ocean. 14(3)175-194. 
Woodmansee, R. A. 

1966. Daily vertical migration of Lucifer. Planktonic 
numbers in relation to solar and tidal cycles. Ecology 
47:847-850. 

Yeung, C. D. L. Jones, M. M. Criales, T. L. Jackson, and W. J. 
Richards. 

2001. Influence of coastal eddies and counter-currents on 
the influx of spiny lobster, Panulirus argus, postlarvae 
into Florida Bay: influence of eddy transport. Mar. 
Freshw. Res. 52:1217-32. 

Yeung. C, and T. L. Lee. 

2002 Larval transport and retention of the spiny lobster, 
Panulirus argus, in the coastal zone of the Florida Keys, 
USA. Fish. Oceanog. 11:86-309. 



75 



Abstract — The population biology 
and status of the painted sweeplips 
iDiagramma pictum) and spangled 
emperor (Lethrinus ncbiilosus) in the 
southern Arabian Gulf were estab- 
lished by using a combination of 
size-frequency, biological, and size- 
at-age data. Transverse sections of 
sagittal otoliths were characterized 
by alternating translucent and opaque 
bands that were validated as annuli. 
Comparisons of growth characteris- 
tics showed that there were no sig- 
nificant differences (P>0.05) between 
sexes. There were well defined peaks 
in the reproductive cycle, spawning 
occurred from April to May for both 
species, and the mean size at which 
females attained sexual maturity was 
31.8 cm fork length (Lp) for D. pictum 
and 27.6 cm (Lp) for L. nebulosus. The 
mean sizes at first capture (21.1 cm 
Lp for D. pictum and 26.4 cm Lp for 
L. nebulosus) were smaller than the 
sizes for both at first sexual maturity 
and those at which yield per recruit 
would be maximized. The range of 
fishing-induced mortality rates for 
D. pictum (0.37-0.62/yr) was sub- 
stantially greater than the target 
(F„p, = 0.07/yr) and limit lF,,^„ = 0.09/ 
yr) estimates. The range of fishing- 
induced mortality rates for L. nebu- 
losus (0.15/yr to 0.57/yr) was also in 
excess of biological reference points 
(F„p^ = 0.10/yr and F,,^„ = 0.13/yr). In 
addition to growth overfishing, the 
stocks were considered to be recruit- 
ment overfished because the biomass 
per recruit was less than 20'^* of the 
unexploited levels for both species. 
The results of the study are important 
to fisheries management authorities in 
the region because they indicate that 
both a reduction in fishing effort and 
mesh-size regulations are required 
for the demersal trap fishery. 



Biology and assessment of the painted sweetlips 
iDiagramma pictum (Thunberg, 1792)) and the 
spangled emperor (Lethrinus nebulosus 
(Forsskal, 1775)) in the southern Arabian Gulf 

Edwin M. Grandcourt 

Thabit Z. Al Abdessalaam 

Ahmed T. Al Shamsi 

Franklin Francis 

Marine Environment Research Centre 

Environmental Research and Wildlife Development Agency 

Corniche Road 

P O Box 45553 

Khalidlya 

Abu Dhabi, United Arab Emirates 

E-mail address (for E M Grandcourt) egrandcourtS'erwdagovae 



Manuscript submitted 29 December 
2003 to the Scientific Editors Office. 

Manuscript approved for publication 
11 July 2005 by the Scientific Editor. 

Fish. Bull. 104:75-88 (2006). 



The painted sweetlips iDiagramma 
pictum (Thunberg, 1792)). is a member 
of the family Haemuli(iae and is 
widely distributed throughout the 
Indo-West Pacific, from the Red Sea 
and East Africa to Japan and New 
Caledonia (Randall et al., 1997). 
Adults are found in shallow coastal 
waters and coral reefs down to a depth 
of 80 m, and juveniles are often found 
in weedy areas (Smith and McKay, 
1986). The diet of this species consists 
of benthic invertebrates and fishes 
(Sommer et al., 1996). It is a rela- 
tively large tropical species attaining 
100 cm fork length and 6 kg in total 
weight (Torres, 1991); consequently it 
is exploited throughout its range with 
a variety of gears, including hand- 
lines, traps, and nets (Fischer and 
Bianchi, 1984). Diagramma. pictum 
has a gonochoristic reproductive mode 
and spawning occurs annually with 
one clear seasonal peak (Breder and 
Rosen, 1966). 

Fishes of the family Lethrinidae 
are abundant in the coastal tropical 
and subtropical Indo-Pacific (Young 
and Martin, 1982). The spangled em- 
peror (Lethrinus nebulosus (Forsskal, 
1775)), is distributed throughout the 
Indo-West Pacific from the Red Sea 
and East Africa to southern Japan 
and Samoa. It is found in a variety 
of habitats including coral reefs, sea 
grass beds and mangroves from near 
shore to a depth of 75 m (Randall, 



1995). Adults are either solitary or 
are found in small schools, and ju- 
veniles form large schools in shal- 
low, sheltered sandy areas. The diet 
of this species is mainly composed 
of moUusks, crustaceans, polychaete 
worms, and echinoderms (Fischer and 
Bianchi, 1984). 

As with other representatives of 
the family Lethrinidae, L. nebulosus 
is a protogynous hermaphrodite, and 
sexual transformation from female to 
male occurs over a wide range of sizes 
(Young and Martin, 1982; Ebisawa, 
1990). Lethrinids are considered to 
have long spawning seasons, running 
from spring to at least early fall, with 
spring and fall peaks. Spawning oc- 
curs after dark for most species in 
aggregations along the inner or outer 
edge of the fringing reef (Johannes, 
1981). Lethrinus nebulosus is a large 
tropical species reaching 80.0 cm total 
length and 8.4 kg total weight (Ran- 
dall, 1995) and is exploited through- 
out its range with a variety of gears 
(Fischer and Bianchi, 1984). 

Both species form an important part 
of fisheries landings in the southern 
Arabian Gulf, where they are mainly 
caught with dome-shaped wire traps 
that have a hexagonal mesh of ap- 
proximately 3.5 cm in diameter. 
Traps are either set individually or 
in strings from traditional wooden 
dhows, sets are made to a maximum 
depth of 40 m, and vessels fish an 



76 



Fishery Bulletin 104(1) 







i *> 


\\ 


W':- 


-^^: 


r 


'__■; "■ 





V ^ - ,^ 





Figure 1 

Study site (stippled area) showing the location of the Emirate of Abu Dhabi, off the coast of 
which data were collected for the painted sweetlips iDiagramtna pictum) and spangled emperor 
iLethrinus nehulosus) from commercial catches. 



average of 210 traps each. Collection of catch-and-effort 
data for the fisheries of the Emirate of Abu Dhabi in the 
United Arab Emirates was initiated in 2001. However, 
many species, including D. pictum and L. nebulosus, 
are recorded to the family level, and therefore the use 
of statistical catch-at-age methods can not be used for 
conducting assessments at the species level. Landings 
of haemulids (predominantly D. pictum) and lethrinids 
(predominantly L. nebulosus), totaled 719 and 2911 
metric tons, respectively, in the Emirate of Abu Dhabi 
during 2003 (Grandcourt et al.M. Despite the limited 
time scale for which catch and effort data are available, 
there has been an overall increase in fishing effort and 
catches since 2001. 

Many of the fish populations in the Arabian Gulf have 
been heavily exploited (Samuel et al., 1987), and fishing 
effort may have already been above optimum levels for 
most demersal species (Siddeek et al., 1999). The expan- 
sion of the fishing fleet of the United Arab Emirates and 
the lack of appropriate data on most stocks underscore 
the need to assess the fisheries resources of the region. 
The goal of this study was to evaluate the status of £1. 
pictum and L. nebulosus and to provide biological refer- 



Grandcourt, E. M., F. Francis, A. Al Shamsi, K. Al Ali, and 
S. Al Ali. 2004. Annual fisheries statistics for Abu Dhabi 
Emirate 2003, 87 p. Environmental Research and Wildlife 
Development Agency, P.O. Box 45553, Government of Abu 
Dhabi, United Arab Emirates. 



ence points and other pertinent information required for 
management. Specific objectives included establishing 
key demographic parameters by using validated age 
estimates, identifying reproductive characteristics and 
conducting yield-per-recruit analyses for the selected 
study species. 



Materials and methods 

Study site and sampling protocol 

Size-frequency data were collected from commercial 
catches made off the coast of the Emirate of Abu Dhabi 
in the United Arab Emirates (Fig. 1) between September 
2000 and March 2003. Fish were selected at random 
from landings and fork length (Lp) was recorded to the 
nearest cm by using a measuring board. Monthly target 
sample sizes were 500 individuals per species. 

Biological data was collected from specimens pur- 
chased from commercial catches between June 2002 and 
May 2003. Samples were obtained from 30 individuals 
of each species from a representative size range during 
the last week of each month. Standard length (Lg), fork 
length (Lp), and total length (L.j.) were recorded to the 
nearest mm by using a measuring board. Whole wet 
weight was measured with an electronic balance and 
recorded to the nearest g. The sex of a fish was deter- 
mined by macroscopic examination of the gonad, which 



Grandcourt et al : Biology and assessment of Diagramma pictum and Lethrinus nebulosus in the southern Arabian Gulf 



77 



was removed and subsequently weighed to 0.1 g with 
an electronic balance. 

Sagittal otoliths were extracted, cleaned in water, 
dried, and stored in seed envelopes. One of each pair 
of sagittae was weighed to 0.1 mg, burnt on a hotplate 
until it changed to a dark brown color, and embedded 
in epoxy resin. Transverse sections through the nucleus 
(of approximately 200-300 jim thickness) were obtained 
by using a twin blade saw. Sections were mounted on 
glass slides and examined with a low-power microscope 
and transmitted light. 

Age and growth 

The number of opaque bands in transverse sections was 
recorded in addition to the optical characteristics of the 
outer margin (opaque or translucent). The proportion 
of samples with opaque or translucent margins was 
calculated for each month and used to infer the timing 
and periodicity of increment formation. The age at which 
the first opaque band formed was calculated as the time 
between the mean birth date and the time of formation 
of opaque bands. Subsequently, the absolute age was 
calculated as the age at formation of the first band plus 
the number of opaque bands outside the first band and 
the time between the formation of the last band and cap- 
ture. In order to establish the relationship of the timing 
of opaque zone formation with trends in sea water tem- 
perature, time series data were converted by using the 
scaling process given in Newman and Dunk (2003). 

Growth was investigated by fitting the von Berta- 
lanffy growth function (von Bertalanffy, 1938) to size- 
at-age data using standard nonlinear optimization 
methods. The model was fitted to pooled data and each 
sex separately. The von Bertalanffy growth function is 
defined as follows: 



L, = L., il -e 



-k U-l„) 



«'), 



where L, = length at time t\ 

L^= the asymptotic length; 
k = the instantaneous growth coefficient; and 
t^ = the hypothetical time at which length is 
equal to 0. 

Growth curves were compared between sexes for each 
species by using the analysis of residual sums of squares 
method of Chen et al. (1992). 

The growth performance index (P (Gayanilo and Pau- 
ly, 1997) was calculated in order to provide a basis for 
the comparison of growth characteristics in terms of 
length: 

*' = <J>- 2/3 logio (a), 

where <J> = logjg ik) + 0.67 logm iWj and W, = aLj. 

The constant, a, was derived from length-weight rela- 
tionships and k and L^ were obtained from the von 
Bertalanffy growth function. 



Parameters of the length-weight relationship were ob- 
tained by fitting the power function W = oLp* to length 
and weight data where W is the total wet weight, a is a 
constant determined empirically, Lp is the fork length, 
and b is close to 3.0 for species with isometric growth. 
Ratios of total length (Lj) to fork length (Lp) were also 
calculated for each species. 

Reproduction 

The mean size at first sexual maturity was estimated for 
females by fitting the logistic function to the proportion 
of mature fish in 2-cm (Lp) size categories. The mean 
size at first maturity was taken as that at which 50% 
of individuals were mature. Gonadosomatic indices, 
calculated by expressing gonad weight as a proportion 
of total body weight, were plotted against the sample 
period by month to establish the timing and seasonality 
of spawning. The mean birth date was estimated from 
patterns in reproductive indices. 

Population sex ratios were examined by using x" good- 
ness-of-fit tests. Independent tests were conducted to 
determine whether sex ratios differed significantly from 
unity for whole samples and for size and age categories 
within samples. The probability level was set at 0.05 
(a=0.05) and Yates's correction factor was used on ac- 
count of there being only one degree of freedom for each 
comparison. Juvenile retention was calculated as the 
proportion of fish in aggregated size-frequency samples 
below the mean size at first sexual maturity. 

Mortality and selectivity 

Size-at-age data were used to construct age-length keys 
following the method of Ricker (1975) and these were 
used to convert aggregated length-frequency data into 
age-frequency distributions. The number of fish above 
the age at which fish were fully recruited to the fishery 
was calculated as a proportion of the total number of 
fish. The annual instantaneous rate of total mortality 
(Z) was subsequently determined with the age-based 
catch curve method (Beverton and Holt, 1957). The 
natural logarithm of the number of fish in each age 
class was plotted against the corresponding age, and 
Z <±95% CI) was estimated from the descending slope 
of the best fitting line by using least-squares linear 
regression. Initial ascending points representing fish 
that were not fully recruited to the fishery were excluded 
from the analyses. 

The annual instantaneous rate of total mortality was 
also estimated with the length-converted catch method 
of Pauly (1983). Pooled length-frequency samples were 
converted into relative age-frequency distributions by 
using parameters of the von Bertalanffy growth func- 
tion. The natural logarithm of the number of fish in 
each relative age group divided by the change in rela- 
tive age was plotted against the relative age, and Z 
(±95% CI) was estimated from the descending slope of 
the best fitting line with least-squares linear regres- 
sion. The estimates of Z from age-based and length- 



78 



Fishery Bulletin 104(1) 



converted catch curves were compared by 
using a modified Ntest (Sokal and Rohlf, 
1995 . 

Backwards extrapolation of the length-con- 
verted catch curves was used to estimate 
the probability of capture data. Selectivity 
curves were generated by fitting a logistic 
function to the plot of the probability of cap- 
ture against size, from which values of the 
parameters L^g, L^g, and the size at which 
fish were fully recruited to the fishery (Ljqq) 
were obtained. 

Estimates of the annual instantaneous 
rate of natural mortality (M) were obtained 
for each species with the empirical equation 
derived by Hoenig (1983). Maximum age 
estimates of 31 years for D. pictiim and 21 
years for L. nebulosus from the literature 
(Loubens, 1980; Edwards and Shaher. 1991) 
were used because the maximum ages and 
sizes obtained in our study were considerably 
lower than other reported values. 

The annual instantaneous rate of fish- 
ing-induced mortality (F) was calculated by 
subtracting the natural mortality rate (M) 
from the total mortality rate (Z) derived from 
age-based catch curves (F=Z-M). The calcu- 
lation was also made for the upper and lower 
95% confidence intervals for estimates of Z 
in order to derive a range of fishing mortal- 
ity rate estimates. The exploitation rate (E) 
was calculated as the proportion of the fish- 
ing mortality in relation to total mortality 
{E=FIZ). 

Assessment of the fishery 



Relative yield and biomass-per-recruit analyses were 
used to assess the fishery. Growth (k and L^), mortal- 
ity (M), and selectivity (Z-f^) parameters were used as 
model inputs, and knife-edge selection was assumed. 
The Beverton and Holt (1966) yield-per-recruit (YPR) 
model modified by Pauly and Soriano (1986) was used to 
estimate the sizes at maximum yield per recruit ^L^^^) 
and to predict the effects of increasing the mean size 
at first capture (Z-5q) to the mean size at first sexual 
maturity (L^^^) and that at which yield per recruit 
would be maximized (^,„ax*- Estimates of exploitation 
rates representing 1) a marginal increase of relative 
yield per recruit which is 0.1 of its value at the origin 
(£gi) and 2) maximum yield iE^^^^) were also derived 
from the model. The exploitation rates corresponding to 
i^^pt and ^iin,,, (^upt and£|,|^|,) were calculated and used 
to estimate the relative biomass per recruit for each 



,^J^ V 



^^^E> 



■■^/>---l 





Figure 2 

Photomicrographs of transverse sections through the sagittal oto- 
liths of (A) Diagramma pictum. 56.0 cm Lj, and (B) Lethrinus nebu- 
losus, 45.7 cm Lp. Dots show opaque zones (scale bar=l niml. 



and i'niax frorn relative biomass- 



species for Ljq, L^ 
per-recruit curves. Precautionary target iF^^) and limit 
^^limit' biological reference points were calculated as 
0,5 and 2/3 M, respectively, and used to infer resource 
status by direct comparison with the fishing mortality 
rates established for the study species. 



Results 

Age and growth 

Alternating translucent and opaque growth increments 
were observed in transverse sections of the sagittal 
otoliths of D. pictum and L. nebulosus when viewed 
with transmitted light under low-power magnification 
(Fig. 2). For both species, one growth increment consist- 
ing of an opaque and translucent zone was formed on 
an annual basis. Opaque bands formed in the summer 
months between May and September in association with 
increasing sea water temperatures (Fig. 3); conversely 
translucent zones were deposited in the autumn and 
winter (October to February) in association with decreas- 
ing sea water temperatures. 

The maximum age estimates determined from counts 
of opaque bands were 13 and 14 years for D. pictum 
and L. tiebulosus, respectively. Size-at-age relationships 
were asymptotic in form and there was considerable 
individual variability in growth (Fig. 4) (parameters of 
the von Bertalanffy growth function are given in Table 
1). A comparison of the growth characteristics between 
sexes revealed that there were no significant differ- 
ences in parameter estimates for both species (P=0.125, 
df=319 for D. pictum and P=0.878, df =324 for L. nebu- 



Grandcourt et a\ Biology and assessment of Diogiamma pictum and Lethnnus nebulosus in the southern Arabian Gulf 



79 



losus). Values of the growth performance index 
<P for growth in length were 2.81 for D. pictum 
and 2.80 for L. nebulosus. 

The length-weight relationship provided a 
good fit to length and weight data for D. pictum 
(W=lxlO-5xL2 -'9) (r-'=0.994) and L. nebulosus 
(W=3xlO-^xL-»») (r2 = 0.992). Ratios of total 
length (Lj) to fork length (Lp) were 1.11:1.0 for 
D. pictum and 1.07:1 for L. nebulosus. 



Reproduction 

The mean size at first sexual maturity (i,„a, ) for 
D. pictum was 30.7 cm Lp for males (24.3-37.2 cm 
95% CI) and 31.8 cm Lp for females (24.3-38.9 
cm 95% CI). Those for L. nebulosus were 28.6 
cm Lp (23.7-33.8 cm 95% CI) and 27.6 cm Lp 
(19.6-35.6 cm 95% CI) for males and females, 
respectively. 

There was a peak in the gonadosomatic index 
for both D. pictum and L. nebulosus females in 
April, the main spawning period lasted until the 
end of May (Fig. 5), and the mean birth date 
was estimated as 1 May. There was a significant 
(P<0.05) female bias in the overall sex ratios 
(male to female) of 1:2.8 for D. pictum and 1:2.6 
for L. nebulosus. The results of chi-square good- 
ness-of-fit tests conducted for the sex ratios in 
both age and size categories revealed that the 
female bias was consistent across all categories 
for L. nebulosus although the bias was removed 
in the oldest age and largest size classes of D. 
pictum (Tables 2 and 3). 

The proportion of fish in aggregated size-fre- 
quency samples that were below the mean size 
at first sexual maturity for females (juvenile re- 
tention rate) was 35.1% for D. pictum and 10.9% 
for L. nebulosus. 



Mortality and selectivity 



1 1 T 


A 


1  


y^^^^'-^-r^^ 


09 - 


— •—Temp / ^^ 


08  


x^ .** \ 


7  


  o -■  % opaque f • \ 
/ • * \ 


6  


/ ^ \ 


5  


/ !» . \ 


0.4  


/ ^ 


0,3  


/ ** 


0.2  


/ 


01  


/ ■' 

o / o 


- 


8' • — -^ « 


,. -0 1  




' 


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


Scaling i 


B 


9 • 


— • — Temp / 11 


8 - 


X * \ 


0,7 - 


- . -o- • - % opaque ¥ _.■' \ 


0,6  


/ \ 


0,5 • 


/ «\ 


0,4  


/ * 


3  


J 


02 - 


o / 6 


0,1 - 


/ -.d 


- 


•— -^t>— >^ 


-0 1 - 




< 1 1 1 1 1 1 I 1 1 1 1 


01 23456789 10 11 12 


Month 


Figure 3 


The proportion of otoliths with opaque outer margins for (A) 


Diagramma pictum (n = 348l and (B) Lethnnus. nebulosus (n = 343) 


and monthly sea temperatures off the Emirate of Abu Dhabi. 


Note that the values have been converted to a standardized 


scale to enable comparison of the trends. 



Modal age groups in age-frequency distributions 
derived from age-length keys and size-frequency data 
were 3 years for D. pictum and 5 years for L. nebulosus 
(Fig. 6). The proportion offish above the age at which 
fish were fully recruited was 13.8% and 45.7% for D. 
pictum and L. nebulosus, respectively. There were no 
significant differences between the total mortality rate 
estimates derived from age-based and length-converted 
catch curves for D. pictum {t=0.81, P=0.43, 15 df) and 
L. nebulosus (^=0.03, P=0.98, 11 df) (Fig. 7). Fishing 
mortality rates were in excess of the natural mortality 
rates, accounting for 79% and 64% of the total mortality 
for D. pictum and L. nebulosus, respectively (Table 4). 

The selectivity range derived from plots of the prob- 
ability of capture at size was 25.0 cm for D. pictum 
(18.0-43.0 cm) and 34.0 cm for L. nebulosus (13.0-47.0 
cm) (Fig. 8). Values of the sizes where the probability of 
capture was 50% (Ljq), 75% (L-g), and 100% (Lj,,,,) are 
given in Table 5. For both species, fish were recruited 





Table 1 






Parameters of the von Bertalanffy growth function, coef- 
ficients of determination (/■-), and sample sizes in) for Dia- 
gramma pictum and Lethrinus nebulosus in the southern 
Arabian Gulf 


D 


pictum 




L. 


nebulosus 


Males 


Fe- 
males 


All 


Males 


Fe- 
males All 


;; 0.29 


0.23 


0.24 


0.10 


0.11 0.11 


L^cm(Lp) 60.6 


63.8 


63.0 


69.9 


65.2 66.2 


^olyr) -1.2 


-1.5 


-1.4 


-3.3 


-2.9 -3.0 


r2 0.86 


0.83 


0.84 


0.91 


0.88 0.79 


n 81 


244 


325 


86 


244 330 



80 



Fishery Bulletin 104(1) 




Figure 4 

The von Bertalanffy growth function fitted to size-at- 
age relationships for (A) Diagramma pictuni and (B) 
Lethriiius nebulosus. 



4.0 1 


A 

1 




3.0  












1 


2,0  


' 


• Females 










— »— Males 


1.0  


' 




, 


r 


T 


£ 




L J 


L.  I v: 




[ 


1 0,0 1 

c 

u 

ro 

E 

o 

S 7.0-1 

CO 

S 6  


^^~ a ^~' "*■♦ •^ J 


r-'-'-N.. m"-""'^ 




2 

B 


3 


1 5 6 7 8 9 10 11 12 


5,0 - 








4,0  
3  




I 


^ 


\ 


2,0  


- / 






\ .,,,... Females 
\ — •— Males 


1.0 ' 






 \ . 




M f f f f M 


1 


2 3 4 5 6 7 8 9 10 11 12 


Month 


Figure 5 


Mean monthly gonadosomatic indices (±SE) for (A) 


Diagramm pictum (n = 359) and (B) Lethrinus nebulosus 


(/?=360l. 

















Table 2 














Results of chi-square 


goodness 


-of-fit tests on 


sex 


ratios 


within age categories 


forZ) 


agra 


mma 


pictum 


and Lethrinus 


nebulosus 


(* significant at it- 


0.05). 
























Age category (yr) 






No 


of males 




No. of females 






Chi 


square 


total) 


P 


D. pictum 




























0-2 








29 






130 








64.2 




<0.01* 


3-5 








38 






91 








21.8 




<0.01* 


6-13 








16 






29 








3.8 




>0.05 


L, nebulosus 




























0-2 








34 






93 








27.4 




<0.01* 


3-6 








29 






88 








29.8 




<0.01* 


7-14 








27 






69 








18.4 




<0.01* 



Grandcourt et al Biology and assessment of Diagiamma pictum and Lethnnus nebulosus in the southern Arabian Gulf 



81 



2000 



1500 



1000 



500 



n = 5006 




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



2000  



1500 



1000 



500 



n = 13.129 




1 2 3 4 5 6 7 8 9 10 11 12 13 14 
Age group (yr) 

Figure 6 

Age-frequency distributions for I A) Diagramma pictum 
and (Bi Lethrinus nebulosus derived from age-length 
keys and size-frequency data. 



to the fishery at a mean size iLc^g) which was smaller 
than the mean size at which first sexual maturity was 
attained (L,„j,(). 

Assessment of the fishery 

The size at which yield per recruit would be maximized 
*-^max' w^s ■^4.4 cm (Lp) for D. pictum and 36.9 cm (Lp) 
for L. nebulosus. For both species, these values were 
considerably greater than the mean size at first capture 
and the mean size at first sexual maturity (Fig. 9). 

The exploitation rate for D. pictum (0.79/yrl was 
greater than that which would maximize yield per re- 
cruit (0.57/yr) at the existing mean size at first capture 
(Table 6). Furthermore, the same yield per recruit could 
be achieved at a much lower exploitation rate and at 
an increased relative biomass per recruit (Fig. 10). 
The yield-per-recruit function also indicated that an 
increase in the size at first capture to that which would 







Table 3 






Results of chi 


-square 


goodness-of-fit 


tests on 


sex ratios 


within size categories 


for Diagramma 


pictum and Lethn- 1 


nus nebulosus 


(*=sign 


ficantat a = 0.05). 




Size 






Chi- 




category 


No. 


f No. of 


square 




(cm Lp) 


males females 


(total) 


P 


D. pictum 










20-34 


26 


128 


67.6 


<0.01* 


35-49 


41 


93 


20.2 


<0.01* 


50-64 


27 


43 


3.7 


>0.05 


L. nebulosus 










15-29 


40 


121 


40.8 


<0.01* 


30-44 


41 


89 


17.7 


<0.01* 


45-54 


26 


73 


22.3 


<0.01* 





Table 4 


Mortality and exploitation rates (±95 CI) for Diagramma 
pictum and Lethrinus nebulosus. 


D. pictum L. nebulosus 


Total mortality 

rate/yr(Z) 




(Age-based 
catch curve) 


0.63 (0.50-0.75) 0.56 (0.35-0.77) 


(Length- 
converted 
catch curve) 


0.69(0.58-0.79) 0.56(0.53-0.60) 


Natural mortality 
rate/yr (Mt 


0.13 0.20 


Fishing mortality 
rate/yr (F) 


0.50(0.37-0.62) 0.36(0.15-0.57) 


Exploitation 

rate/yr (£) 


0.79 0.64 



Table 5 

Probability of capture ( selectivity) for Diagramma pictum 
and Lethrinus nebulosus. 



Probability of capture 



50% (L50) 



D. pictum 


L. nebulosus 


cm (Lp) 


cm (Lp) 


21.1 


26.4 


30.7 


35.1 


40.3 


43.8 



82 



Fishery Bulletin 104(1) 



• Length converted 

* Age based 




7 
6 

5 - 

t 

4  

3 

2 

1 



B 



»oo' 




10 12 14 16 



• Length converted 
A Age based 



5 10 15 20 

Age (yr)- Relative age (yr) 



25 



Figure 7 

Age-based catch curves (InF against A^e) and length- 
converted catch curves (In F/dt against Relative age) 
for Diagramma pictum (A) and Lethrinus nebulosus (B). 
Dashed and solid lines show the regression equation 
{y = b.x+a) fitted to data for age-based and length-con- 
verted catch curves, respectively. Only solid points are 
included in the regressions. 



Table 6 

Exploitation rates iEg^ and E^^^^) at the existing size 
at first capture (L^p), the size at first capture at sexual 
maturity iL^^,) and the size at first capture at maximum 
yield per recruit (L^^^) for Diagramma pictum and Leth- 
rinus nebulosus. 


D. pictum 


L. nebulosus 


'•ate L50 ^mat i^ax 


^50 ^mat -^max 


£u,(/yr) 0.50 0.56 0.75 
E„3, (/yr) 0.57 0.65 0.81 


0.51 0.56 0.75 
0.63 0.65 0.86 




10 20 30 



50 



60 



70 




B 



30 40 50 

Fork length (cm) 



60 



70 



Figure 8 

Selectivity curves for (A) Diagramma pictum and (B) 
Lethrinus nebulosus, showing the mean size at first 
capture at probabilities of 0.5 (^5,,), 0.75 (L-5). and the 
size at which fish are fully recruited to the fishery <Lj(„,). 



maximize yield per recruit would be associated with 
a substantial increase in yield at the current level of 
exploitation. An increase in the size at first capture to 
that at which sexual maturity occurs (L^^^) was also 
predicted to be associated with an increase in yield, 
although to a lesser degree (Fig. 10). 

The relative biomass per recruit for D. pictum at 
the current exploitation rate was less than 10% of 
that at the theoretical unexploited level. An increase 
in the mean size at first capture to that which would 
maximize yield per recruit was predicted to be asso- 
ciated with a small increase in biomass per recruit. 
Changes in the exploitation rate were predicted to 
have a greater impact on biomass per recruit, which 
was estimated to be above 50% of the unexploited 
level at the optimum exploitation rate (£„„,) (Table 7). 



Grandcourt et a\ : Biology and assessment of Diagramma pictum and Lethnnus nebulosus in the southern Arabian Gulf 



83 




3000 - 
2500 - 


B 






i-m 


n=13.180 


2000 - 




L 


™, 




A 


1500 - 




Lso 




/ 


' \ 


1000  






r 


y 


\ 


500  
1 




/ 






V 



10 



20 



30 40 50 

Size (cm Lp) 



60 



70 



Figure 9 

Aggregated length-frequency distributions for (A) 
Diagramma pictum and (B) Lethrinus nebulosus 
showing the mean sizes at first capture (Lj,,). sexual 
maturity iL^^^), and the size at first capture which 
would maximize yield per recruit (i^ax'- 



-N 






■— ^r 


0.15 - 


^ 


V / 




. • 




^y=- 




0.1 - 








0.05  


/ 


f L50 


E \sj^^ 


e ^ 


n - 


/ 


 - Lmal 




<:;:^ 





B 







0.2 



0.4 



0.6 



0.8 



0.6 



0.4 



33 
- 0.2 !L 



(1) 
> 




u.us  

0.04  
0.03  




^^ 









^ 




^"^^^ 




0.02  




^ 


v 






0.01  


/ L50 

/ - Lmat 




X 


^^ 








toi 




^^^^^:ssw 




• 




, 




^^^ 



0.6 



0.4 



0.2 







0.2 0,4 0,6 0.8 1 

Exploitation rate (/yr) 

Figure 10 

Relative yield and respective biomass-per-recruit curves 
(descending lines) for (A) Diagramma pictum and (B) 
Lethrinus nebulosus showing the existing exploitation 
rate (£) and the exploitation rate at which the slope of 
the yield-per-recruit curve is 0.1 of its value at the origin 
iEqj). Curves show the effect of increasing the existing 
mean size at first capture (LjqI to the mean size at first 
sexual maturity (imat* ^^^ ^^^ ^'^^ which would maximize 
yield per recruit ih^^^). 



Estimates of precautionary target and limit exploita- 
tion rates (E ^ and £|,j„,t) for D. pictum were 0.34 and 
0.40, respectively. 

The exploitation rate for L. nebulosus (0.64/yr) was 
marginally greater than that which would maximize 
yield per recruit (0.63/yr) at the existing mean size at 
first capture (Table 6). The yield-per-recruit function 
indicated that an increase in the size at first capture 
to that which would maximize yield per recruit would 
be associated with an increase in yield at the current 
level of exploitation. An increase in the size at first 
capture to that at which sexual maturity occurs iL^^^) 
was also predicted to be associated with an increase in 
yield, although to a lesser degree (Fig. 10). 

The relative biomass per recruit for L. nebulosus at 
the current exploitation rate was less than 20% of that 



Table 7 

Relative biomass per recruit at precautionary exploita- 
tion rates (E ^ and E^^^^^) at the existing size at first cap- 
ture (L5Q ), the size at first capture at sexual maturity (Lj^^^^j), 
and the size at first capture at maximum yield per recruit 
'•^max' '*"" Diagramma pictum and Lethrinus nebulosus. 

Relative biomass per recruit 



Exploitation 
rate 



D. pictum 



L. nebulosus 



^50 



mat max 



^50 



mat max 



£i,m,t</yt 



0.52 0.55 0.60 
0.44 0.48 0.53 



0.50 0.51 0.53 
0.39 0.40 0.45 



84 



Fishery Bulletin 104(1) 



at the theoretical unexploited level. An increase in the 
mean size at first capture to that which would maxi- 
mize yield per recruit was predicted to be associated 
with a small increase in biomass per recruit. Changes 
in the exploitation rate were predicted to have a greater 
impact on biomass per recruit, which was estimated to 
be above SO^i: of the unexploited levels at the optimum 
exploitation rate iE^^ ^) (Table 7). Estimates of precau- 
tionary target and limit exploitation rates (£,,„, and 
^iimit' f°'" ^- nebulosus were 0.33 and 0.43, respectively. 
For both D. Pictum and L. nebulosus, the exploitation 
rates predicted from the yield per recruit function (Eg j 
and Efnax' were equal to or in excess of values where 
F=M (0.5) for all sizes at first capture. 

The range of fishing mortality rates estimated for D. 
pictum (0.37-0.62/yr) was substantially greater than 
both the target (F„p, = 0.07/yr) and limit (F|,„„j=0.09/yr) 
biological reference points. The range of fishing mortal- 
ity rates for L. nebulosus (0.15/yr to 0.57/yr) were also 
in excess of the biological reference points for this spe- 
cies (F„pj=0.10/yr and F,,„„,=0.13/yr). 



Discussion 

Age and growth 

The formation of alternating translucent and opaque 
growth zones in fish otoliths has been associated with a 
variety of factors, including seasonal variations in water 
temperature, photoperiod, feeding, and reproduction 
(Moe, 1969; Reay, 1972; Panella, 1980; Manickchand- 
Heileman and Phillip, 2000). Although the mechanisms 
of growth-increment formation are poorly understood, 
the deposition of the opaque zone in tropical species 
generally occurs in the spring and summer months 
during periods of accelerated growth, whereas the trans- 
lucent zone is formed when there is reduced metabolic 
activity (Beckman and Wilson, 1995). The formation of 
opaque and translucent zones in the sagittal otoliths 
of D. pictum and L. nebulosus determined in our study 
follows this generalized pattern. 

The southern Arabian Gulf exhibits marked sea- 
sonal variability in oceanographic characteristics; sea 
water temperatures can exceed 34°C in summer and 
fall to 2rC in the winter (Sheppard et al,, 1992). The 
close association of the formation of opaque zones with 
increasing seawater temperature indicates that tem- 
perature could be the principal environmental signal 
stimulating the deposition of these growth zones. Other 
allied variables, such as productivity and subsequent 
food availability, may also be associated with seasonal 
growth-rate oscillations and the formation of growth 
increments. The validation of the annual periodicity 
of increment formation adds to a growing body of evi- 
dence (Fowler, 1995) and dismisses the misconception 
that annuli do not form in the otoliths of reef fish due 
to a lack of seasonality in the tropics (e.g., Sparre and 
Venema, 1992). Nevertheless, the edge analysis method 
used should ideally be conducted over a 2-year cycle and 



could have been more rigorous with the use of larger 
sample sizes and by conducting the analyses for indi- 
vidual ages. Furthermore, it is important to distinguish 
between the validation of increment periodicity and 
absolute age (Campana, 2001). Although our study has 
provided empirical evidence for an annual pattern of 
increment formation, the absolute age of the study spe- 
cies remains to be validated. Validation of the absolute 
age could be achieved through independent means such 
as a mark-recapture study and the chemical marking of 
juvenile fish of known age. 

The maximum number of opaque bands counted for 
D. pictum in the present study (13) was considerably 
less than the maximum age of 31 years estimated by 
Loubens (1980) for this species in New Caledonia. Like- 
wise, our longevity estimate of 14 years for L. nebulosus 
was less than that of Mathews and Samuel (1991) (20 
years) and of Edwards and Shaher (1991) (21 years) for 
this species in the northern Arabian Gulf and Gulf of 
Aden, respectively. Our longevity estimates therefore 
were most likely to have been underestimated owing 
to the absence of fish close to the maximum reported 
sizes for these species. 

A method of validating growth parameters involves 
the comparison of growth performance indices (<f>') in 
terms of growth in length with other estimates obtained 
for the same or a similar species (Gayanilo and Pauly, 
1997). Values of <P for D. pictum available from other 
studies have ranged from 2.88 (Loubens, 1980) to 3.24 
(Baillon and Kulbicki, 1988), and an estimate of 3.07 
has been obtained for this species in the Gulf of Aden 
(Edwards et al., 1985). The estimate obtained in our 
study (2.81) compares with the lower end of this range. 
Values of <& for L. nebulosus have ranged from 2.55 
(Kuo and Lee, 1986) to 3.41 (Baillon, 1991), and an es- 
timate of 2.87 obtained for this species in the northern 
Arabian Gulf in the waters off Kuwait (Baddar, 1987) 
compares well with our estimate of 2.80. Although the 
growth parameters in our study would appear to be of 
the right order (by comparison), improvements in our es- 
timates could have been made by the addition of larger 
specimens close to the maximum size for both species. 

Despite its widespread use, the von Bertalanffy 
growth function may not be suitable for hermaphro- 
ditic populations (Appeldoorn, 1996). Growth analyses 
have shown distinct differences in the sizes of equal-age 
males and females of protogynous species (Moe, 1969; 
Nagelkerken, 1979; Garratt et al., 1993), and experi- 
ments have shown what are considered to be growth 
accelerations leading to sex change (Ross et al., 1983). 
Failure to account for growth spurts in yield models 
can result in significant over estimation of both maxi- 
mum yield and optimal effort (Bannerot et al., 1987). 
Although L. nebulosus is a protogynous hermaphro- 
dite (Young and Martin, 1982; Ebisawa, 1990), the 
results of the analysis of residual sums of squares in 
our study indicated that there are no differences in the 
growth characteristics between sexes and that the use 
of growth parameters from pooled data would therefore 
be justified in yield-per-recruit analyses. 



Grandcourt et al : Biology and assessment of Diagiamma pictum and Lethnnus nebulosus in the southern Arabian Gulf 



85 



Reproduction 

Simulation models and evidence of the effects of fish- 
ing have shown that protogynous species are far more 
vulnerable to fishing pressure than comparable gono- 
choristic stocks (Huntsman and Schaaf, 1994). For 
protogynous species, in which males tend to be larger 
than females on average, there are indications that 
size-selective fishing mortality may result in the dif- 
ferential loss of larger males (Sadovy, 1996) and the 
possibility that insufficient males remain in the repro- 
ductive population to fertilize eggs from all females 
(Koenig et al., 1996). In this context, L. nebulosus may 
be particularly vulnerable to such effects because the 
female-biased sex ratios were consistent throughout all 
the age categories and size classes. The overall female 
bias and removal of the bias in the oldest and largest 
age category for D. pictum is generally representative 
of the sexual structure of a protogynous population. 
Given these characteristics, histological confirmation 
of the reproductive mode of this species should be 
considered. 

The well-defined spawning period of D. pictum and 
L. nebulosus between April and May supports the view 
that seasonal reproductive cycles are common among 
tropical fishes (Robertson, 1990; Montgomery and Gal- 
zin, 1993; Sadovy, 1996). There were high levels of ju- 
venile retention for D. pictum (35.1%) because fish were 
recruited to the fishery before the mean size at which 
sexual maturity occurred, indicating a need to increase 
the mesh size of traps. 

Mortality and selectivity 

Estimates of natural mortality derived from other stud- 
ies of Z). pictum range from 0.43/yr (Edwards et al., 
1985) to 0.67/yr (Baillon and Kulbicki, 1988). The com- 
paratively low value of M obtained in our study (0.13/yr) 
can be attributed to the difference in methods used, but 
errors in other estimates from the empirically derived 
formula of Pauly (1980) may have occurred because the 
relationship has tended to overestimate M, especially 
for slow growing species (e.g., Ralston, 1987; Russ et 
al., 1998). Similarly, our value of the instantaneous 
natural mortality rate for L. nebulosus (0.20/yr) was 
lower than other estimates that range from 0.279/ 
yr (Edwards et al., 1985) to 1.18/yr (Baillon, 1991). 
Although estimates of M derived from the Hoenig ( 1983) 
relationship have been shown to provide a reasonable 
approximation of M in tropical demersal fishes (Hart 
and Russ, 1996; Newman et al., 1996), errors in this 
parameter were potentially the greatest source of error 
in our assessment. 

Upward bias in estimates of the total mortality rate 
(Z) may have occurred if larger fish were less vulnera- 
ble to the fishing gear or if adult fish underwent migra- 
tions, for example. A survey of the biomass of demersal 
species in the Arabian Gulf waters of the United Arab 
Emirates showed that there were no seasonal changes 
in the abundance of L. nebulosus and the haemulid 



Plectorhinchus sordidus (Shallard-). This finding indi- 
cates that ontogenetic or spawning-associated migra- 
tions would unlikely be altering the size and age com- 
position and subsequent estimates of Z for the species 
investigated. Although size-specific selectivity cannot 
be discarded as a possible explanation for the small 
proportion of larger and older fish in size-frequency and 
biological samples, the impact of fishing on the size and 
age structure of the respective populations is considered 
a more likely reason for these observations and is likely 
to be the probable cause given the historic absence of 
regulation in the trap fishery. 

Because the size at first capture (21.1 cm Lp) was 
considerably smaller than the size at which yield per 
recruit would be maximized (44.4 cm L-p) for D. pictum, 
an increase in the mesh size for the trap fishery should 
be considered by management authorities especially 
given the high rate of juvenile retention for this species. 
The same is applicable for L. nebulosus with a mean 
size at first capture of 26.4 cm Lp and size at maximum 
yield per recruit of 36.9 cm Lp. 

Assessment of the fishery 

The use of yield-per-recruit models may be particularly 
restrictive for fast growing tropical species with high 
rates of natural mortality because the curves may not 
reach a maximum within a reasonable range of fishing 
mortality values (Gayanilo and Pauly, 1997). Although 
the species examined in our study were relatively slow 
growing and had low rates of natural mortality, failure 
of the yield-per-recruit model may still have occurred at 
the upper end of the fishing mortality range. 

Gulland (1970) suggested that in an optimally ex- 
ploited stock, fishing mortality should be about equal 
to natural mortality, resulting in an exploitation rate 
of 0.5/yr. However, exploitation rates should be very 
conservative for relatively long-lived reef fish (Newman 
and Dunk, 2003), especially given that potential yields 
may be over estimated by a factor of 3-4 where F=M 
(Beddington and Cooke, 1983). With a range from 0.5 
to 0.86/yr, the exploitation rates derived from yield-per- 
recruit analyses (£„ j and S,,,.,^) are considered to have 
overestimated the associated fishing mortality rates. 

The relative yield-per-recruit analyses indicated that 
an increase in the size at first capture of D. pictum and 
L. nebulosus to that which would maximize yield per re- 
cruit would be associated with increases in yields at the 
existing exploitation rates. However, the existing exploi- 
tation rate for D. pictum (0.79/yr) was greater than that 
which would maximize yield per recruit {E^^^^=0.57/yr). 
Although the exploitation rate for L. nebulosus (0.64/yr) 
was comparable to that which would maximize yield per 



- Shallard, B. 2003. Distribution and abundance of demersal 
fish stocks in the UAE, 211 p. Technical Report 1. Fish 
Resources Assessment Survey Project of Abu Dhabi and UAE 
Waters. Bruce Shallard and Associates and Environmental 
Research and Wildlife Development Agency, P.O. Box 45553, 
Government of Abu Dhabi, United Arab Emirates. 



86 



Fishery Bulletin 104(1) 



recruit (0.63/yr), given that E^^^ was probably overes- 
timated, the results indicate that growth overfishing is 
also occurring for this species. 

The specified precautionary target (F„p, = 0.5.M) and 
limit (F,|^jj,=2/3.M) values are considered to be more 
appropriate biological reference points in light of the 
constraints of the yield-per-recruit model. The range 
of fishing mortality rates estimated for D. pictum 
(0.37-0.62/yr) was substantially greater than both the 
target (F„pj=0.07/yr) and limit (F,„„„ = 0.09/yr) biologi- 
cal reference points, and the existing exploitation rate 
(0.79/yr) was more than double the optimum level (0.34/ 
yr). The range of fishing mortality rates for L. nebulo- 
sus (0.15/yr to 0.57/yr) were also in excess of the biologi- 
cal reference points for this species {Fppj=0.10/yr and 
F,,^,t=0.13/yr) and the exploitation rate (0.64/yr) was 
approximately double the optimum target level (0.33/ 
yr). This result clearly indicates growth overfishing for 
both species and, in combination with the results of the 
yield-per-recruit analyses, demonstrates that effort re- 
ductions are also required in the fishery because target 
reference points cannot be achieved by modification of 
the gear-selectivity characteristics alone. 

A critical limitation of the yield-per-recruit model is 
the assumption that there is no relationship between 
the size of the spawning stock biomass and recruit- 
ment (Buxton, 1992). Even if the size at first capture 
is less than the size at first sexual maturity, the stock 
size may approach zero at high levels of fishing mor- 
tality in spite of predictions of high levels of yield per 
recruit. It is therefore important to consider the size of 
the spawning stock biomass across the range of fishing 
mortality rates when interpreting results. The relative 
biomass per recruit of D. pictum and L. nebulosus at 
the estimated fishing mortality rates was particularly 
low at less than 10% and 20%, respectively, of unex- 
ploited levels. If the critical spawning stock biomass 
is between 20% and 50% of the unexploited levels, as 
suggested by King (1995), recruitment overfishing is 
likely to be occurring for both species. This is most 
clearly seen in the age structure for D. pictum, for 
which only 13.8% of the total number offish were above 
the age at which fish were fully exploited by the gear 
(ages 4-13 years). For this species, the majority of the 
yield was derived from the newly recruited age class 
representing fish that had just become fully vulnerable 
to the gear. 

The relative biomass per recruit at the exploitation 
rates that were equivalent to F^^^ corresponded to 52% 
and 51% of the theoretical unexploited biomass levels 
for D. pictum and L. nebulosus, respectively. The associ- 
ated levels of fishing mortality are therefore considered 
appropriate target reference points, given the present 
fisheries policy which is aimed at resource conserva- 
tion and stock rebuilding. Accordingly, the results of 
our study are important to fisheries management au- 
thorities in the region because they indicate that both 
a substantial reduction in fishing effort and an increase 
in mesh-size of traps are currently necessary for the 
previously unregulated demersal trap fishery. 



Acknowledgments 

We are very grateful to Sultan Al Ali, Khaled Al AH, 
John Hoolihan, and Kunath Gopalan for their assis- 
tance with data collection and Abdulla Aboubaker for 
helping with data entry. Howard Choat and Will Rob- 
bins assisted in the training of technicians in otolith 
processing. The comments of the anonymous reviewer 
helped considerably in improving the manuscript. This 
study was funded by the Government of the United 
Arab Emirates through the Environmental Research 
and Wildlife Development Agency's Marine Environment 
Research Centre. 



Literature cited 

Appeldoorn, R. S. 

1996. Model and method in reef fishery assessment. In 
Reef fisheries (N. V. C. Polunin and C. M. Roberts, eds.), 
p. 219-248. Chapman and Hall, London. 

Baddar. M. K. 

1987. A preliminary study of the population dynamics 
of sheiry, the starry pigface bream, Lethrinus nebu- 
losus. Kuwait Bull. Mar. Sci. 9:215-220. 

Baillon, N. 

1991. Otolithometrie en milieu tropical: application a trois 
especes du lagon de Nouvelle-Caledonie. ORSTOM, Inst. 
Fr. Rech. Sci. Dev. Coop., Trav. Doc. 113, 296 p. 
Baillon, N., and M. Kulbicki. 

1988. Aging of adult tropical reef fish by otoliths: a 
comparison of three methods on Diagramma pictum. In 
Proceedings of the sixth international coral reef sym- 
posium, Townsville, Australia, 8"'-12"' August 1988 
(J. H. Choat, ed.), p. 341-346. 6th International Coral 
Reef Symposium Executive Committee, Townsville, 
Queensland, Australia. 

Bannerot, S. P., W. W. Fox, and J. E. Powers. 

1987. Reproductive strategies and the management of 
snappers and groupers in the Gulf of Mexico and the 
Caribbean. In Tropical snappers and groupers: biology 
and fisheries management (J. J. Polovina and S. Ralston, 
eds.), p. 561-603. Westview Press, Boulder, CO. 
Beckman, D. W., and C. A. Wilson. 

1995. Seasonal timing of opaque zone formation in fish 
otoliths. In Recent developments in fish otolith research 
(D. H. Secor, J. M. Dean, and S. E. Campana eds.), p. 
27-43. Univ. South Carolina Press, Colombia, SC. 
Beddington, J. R., and J. G. Cooke. 

1983. The potential yield offish stocks. FAO Fish. Tech. 
Pap. 242, 47 p. FAO, Rome. 
Beverton, R. J. H., and S. J. Holt 

1957. On the dynamics of exploited fish populations, 533 p. 

Chapman and Hall. London. 
1966. Manual of methods for fish stock assessment, part 
II. Tables of yield function. Fisheries Biology Technical 
Paper 38, 67 p. FAO, Rome. 
Breder, C. M., and D. E. Rosen. 

1966. Modes of reproduction in fishes, 941 p. T.F.H. 
Publications, Neptune City, NJ. 
Buxton, C. D. 

1992. The application of yield-per-recruit models to two 
South African sparid reef species, with special consid- 
eration to sex change. Fish. Res. 15:1-16. 



Grandcouil et a\ Biology and assessment of Diagiamma pictum and Lethnnus nebulosus in the southern Arabian Gull 



87 



Chen, Y. D.. D. A. Jackson, and H. H. Harvey. 

1992. A comparison of von Bertalanffy and polynomial 
functions in modelling fish growth data. Can. J. Fish. 
Aquat. Sci. 49:1228-123.5. 

Campana, S. E. 

2001. Accuracy, precision and quality control in age deter- 
mination, including a review of the use and abuse of age 
validation methods. J. Fish. Biol. 59:197-242. 

Ebisawa, A. 

1990. Reproductive biology oC Lethrinus nehulosua (Pices: 
Lethrinidae) around the Okinawan waters. Nippon 
Suisan Gakkaishi 56 (121:1941-1954. 

Edwards, R. R. C, A. Bakhader, and S. Shaher. 

1985. Growth, mortality, age composition and fishery 
yields of fish from the Gulf of Aden. J. Fish. Biol. 
27:13-21. 
Edwards, R. R. C, and S. Shaher. 

1991. The biometrics of marine fishes from the Gulf of 
Aden. Fishbyte 9(2):27-29. 

Fischer, W., and G. Bianchi. 

1984. FAO species identification sheets for fishery pur- 
poses. Vol. II: Western Indian Ocean (Fishing Area 
51), 257 p. Prepared and printed with the support 
of the Danish International Development Agency 
(DANIDA). FAO. Rome. 

Fowler, A. J. 

1995. Annulus formation in the otoliths of coral reef fish — 
a review. /;; Recent developments in fish otolith research 
(D. H. Secor, J. M. Dean, and S. E. Campana, eds.l, p. 
45-63. Univ. South Carolina Press, Colombia, SC. 

Garratt, P. A., A. Govender, and A. E. Punt. 

1993. Growth acceleration at sex change in the protogy- 
nous hermaphrodite Chrysoblephus puniceus (Pisces: 
Sparidae). S. Afr. J. Mar. Sci. 13:187-193. 

Gayanilo, F. C. Jr., and D. Pauly. 

1997. FAO-ICLARM stock assessment tools. Reference 
manual, 262 p. FAO. Rome. 
Gulland, J. A. 

1970. The fish resources of the ocean. FAO Fisheries 
Technical Paper 97, 425 p. FAO, Rome. 
Hart, A. M., and G. R. Russ. 

1996. Response of herbivorous fish to crown of thorns 
starfish Acanthaster planci outbreaks. Ill: Age, 
growth, mortality and maturity indices oi Acanthurus 
nigrofuscus. Mar. Ecol. Prog. Ser. 136:25-35. 

Hoenig, J. M. 

1983. Empirical use of longevity data to estimate 
mortality rates. Fish. Bull. 82:898-902. 
Huntsman, G. R., and W. E. Schaaf. 

1994. Simulation on the impact of fishing on reproduc- 
tion of a protogynous grouper, the graysby. N. Am. J. 
Fish. Manag. 14:41-52. 

Johannes, R. E. 

1981. Words of the lagoon, 320 p. Univ. Calif. Press, 
Los Angeles, CA. 
King, M. 

1995. Fisheries biology, assessment and management, 
341 p. Blackwell Science Ltd., London. 

Koenig, C. C, F. C. Coleman, L. A. Collins, Y. Sadovy, and P. L. 
Collin. 

1996. Reproduction in gag iMycteroperca microlepis) 
(Pices: Serranidae) in the eastern Gulf of Mexico and 
the consequences of fishing spawning aggregations. In 
Biology, fisheries and culture of tropical groupers and 
snappers (F. Arreguin-Sanchez, J. L. Munro, M. C. 



Balgos, and D. Pauly, eds.), p. 307-323. ICLARM 
Conf. Proc. 48. 
Kuo, C. L., and S. S. Lee. 

1986. Age and growth of common porgy, Lethrinus nebula- 
sus (Forsskal) in shelf waters off northwest Australia. J. 
Fish. Soc. Taiwan. 3:39-57. 
Loubens, G. 

1980. Biologic de quelques especes de poissons du lagon 
Neo-Caledonien. III. Croissance. Cah. Indo-Pac. 
2:101-153. 
Manickchand-Heileman, S. C, and D. A. Phillip. 

2000. Age and growth of the yellow edge grouper, Epineph- 
elus flavolimbatus, and the yellowmouth grouper, Myc- 
teroperca interstitialis, off Trinidad and Tobago. Fish. 
Bull. 98:290-298. 
Mathews, C. P., and M. Samuel. 

1991. Growth, mortality and length-weight parameters for 
some Kuwaiti fish and shrimp. Fishbyte 9(2):30-33. 
Moe, M. A. 

1969. Biology of the red grouper, Epinephelus Mario, 
(Valenciennes) from the eastern Gulf of Mexico. Flor- 
ida Dept. Nat. Resour. Mar. Res. Lab., Prof. Pap. Ser. 
10:1-95. 
Montgomery, W. L., and R. Galzin. 

1993. Seasonality in gonads, fat deposits and condition of 
tropical surgeonfishes (Teleostei: Acanthuridae). Mar. 
Biol. 115:529-536. 
Nagelkerken, W. P. 

1979. Biology of the graysby, Epinephelus cruenfatus, 
of the coral reef of Curasao. Stud. Fauna Cura9ao other 
Carib. Islands 186, 118 p. 

Newman, S. J., and I. J. Dunk. 

2003. Age validation, growth, mortality, and additional 
population parameters of the goldband snapper iPristi- 
pomoides multidens) off the Kimberley coast of north- 
western Australia. Fish. Bull. 101:116-128. 

Newman, S. J., D. McB. Williams, and G. R. Russ. 

1996. Age validation, growth and mortality rates of the 
tropical snappers (Pices: Lutjanidae), Lutjanus adetii 
(Castelnau, 1873) and L. qumquelineatus (Bloch, 1790) 
from the Central Great Barrier Reef, Australia. Mar. 
Freshw. Res. 47(4):575-584. 

Panella, G. 

1980. Growth patterns in fish sagittae. In Skeletal 
growth of aquatic organisms (D. C. Rhoads and R. A. 
Lutz, eds.), p. 519-560. Plenam Press, New York, NY. 

Pauly, D. 

1980. On the interrelationships between natural mor- 
tality, growth parameters and mean environmental 
temperature in 175 fish stocks. J. Cons. Int. Explor. 
Mer 39:175-199. 
1983. Length-converted catch curves: a powerful tool 
for fisheries research in the tropics (part 1). Fishbyte 
1(2):9-13. 
Pauly, D., and M. L. Soriano. 

1986. Some practical extensions to Beverton and Holt's 
relative yield-per-recruit model. In The first Asian 
fisheries forum (J. L. Maclean, L. B. Dizon, and L. V. 
Hosillo, eds.), p. 491-496. Asian Fisheries Society, 
Manila, Philippines. 
Randall, J. E. 

1995. Coastal fishes of Oman, 439 p. Univ. Hawaii 
Press, Honolulu, HI. 
Randall, J. E., G. R. Allen, and R. C. Steene. 

1997. Fishes of the Great Barrier Reef and Coral Sea, 
507 p. Univ. Hawaii Press, Honolulu, HI. 



Fishery Bulletin 104(1) 



Ralston, S. 

1987. Mortality rates of snappers and groupers. In 
Tropical snappers and groupers: biology and fisheries 
management (J. J. Polovina and S. Ralston eds.), p. 
375-404. Westview Press, Boulder, CO. 
Reay, P. J. 

1972. The seasonal pattern of otolith growth and its 
application to back-calculation studies in Ammodytes 
tobianus. J. Cons. Int. Explor. Mer 34:485-504. 
Ricker, W. E. 

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

1990. Differences in the seasonalities of spawning and 
recruitment of some small neotropical reef fishes. J. 
Exp. Mar. Biol. Ecol. 144:49-62. 
Ross, R. M., G. S. Losey, and M. Diamond. 

1983. Sex change in coral reef fish:dependence of 
stimulation and inhibition on relative size. Science 
221:574-575. 
Russ, G. R., D. C. Lou, J. B. Higgs, and B. P. Ferreira. 

1998. Mortality rate of a cohort of the coral trout, Plec- 
tropomus leopardus, in zones of the Great Barrier Reef 
Marine Park closed to fishing. J. Mar. Freshw. Res. 
49:505-11. 
Sadovy, Y. J. 

1996. Reproduction of reef fishery species. In Reef 
fisheries (N. V. C. Polunin and C. Roberts, eds.), p. 
15-59. Chapman and Hall, London. 
Samuel, M., C. P. Mathews, and A. S. Bawazeer. 

1987. Age and validation of age from otoliths for warm 
water fishes from the Arabian Gulf. In Age and growth 
of fish (R. C. Summerfelt and G. E. Hall, eds.), p. 
253-265. IOWA State Univ. Press, Ames. lA. 



Sheppard, C, A. Price, and C. Roberts. 

1992. Marine ecology of the Arabian region. Patterns 
and processes in extreme tropical environments, 347 
p. Academic Press. London. 
Siddeek, M. S. M., M. M. Fouda, and G. V. Hermosa Jr. 

1999. Demersal fisheries of the Arabian Sea, the Gulf 
of Oman and the Arabian Gulf. Estuar. Coastal Shelf 
Science 49 (suppl. A):87-97. 
Smith, M. M., and R. J. McKay. 

1986. Haemulidae. In Smiths' sea fishes (M. M. Smith 
and P. C. Heemstra, eds.), p. 564-571. Springer-Verlag, 
Berlin. 
Sokal, R. R., and F. J. Rohlf 

1995. Biometry. The principles and practice of statistics 
in biological research. 3'''' ed., 887 p. W. H. Freeman 
and Co., New York, NY. 

Sommer, C, W. Schneider, and J. M. Poutiers. 

1996. FAO species identification field guide for fishery 
purposes. The living marine resources of Somalia, 
376 p. FAO, Rome. 

Sparre, P., and S. C. Venema. 

1992. Introduction to tropical fish stock assessment. 
Part 1. FAO Fisheries Technical Paper No. 306.1, rev. 
1, 376 p. FAO, Rome. 
Torres, F. S. B., Jr. 

1991. Tabular data on marine fishes from Southern 
Africa. Part I. Length-weight relationships. Fishbyte 
9(l):50-53. 
von Bertalanffy, L. 

1938. A quantitative theory of organic growth. Hum. 
Biol. 10:181-213. 
Young, P. C, and R. B. Martin. 

1982. Evidence for protogynous hermaphroditism in some 
lethrinid fishes. J. Fish. Biol. 21:475-84. 



89 



Abstract — Many modern stock assess- 
ment methods provide the machinery 
for determining the status of a stock 
in relation to certain reference points 
and for estimating how quickly a 
stock can be rebuilt. However, these 
methods typically require catch data, 
which are not always available. We 
introduce a model-based framework 
for estimating reference points, stock 
status, and recovery times in situ- 
ations where catch data and other 
measures of absolute abundance are 
unavailable. The specific estima- 
tor developed is essentially an age- 
structured production model recast 
in terms relative to pre-exploitation 
levels. A Bayesian estimation scheme 
is adopted to allow the incorpora- 
tion of pertinent auxiliary informa- 
tion such as might be obtained from 
meta-analyses of similar stocks or 
anecdotal observations. The approach 
is applied to the population of goli- 
ath grouper \Epinephelus itajara) off 
southern Florida, for which there 
are three indices of relative abun- 
dance but no reliable catch data. The 
results confirm anecdotal accounts of 
a marked decline in abundance during 
the 1980s followed by a substantial 
increase after the harvest of goliath 
grouper was banned in 1990. The ban 
appears to have reduced fishing pres- 
sure to between 10% and 50% of the 
levels observed during the 1980s. 
Nevertheless, the predicted fishing 
mortality rate under the ban appears 
to remain substantial, perhaps owing 
to illegal harvest and depth-related 
release mortality. As a result, the 
base model predicts that there is less 
than a 40% chance that the spawn- 
ing biomass will recover to a level 
that would produce a 50% spawning 
potential ratio. 



A catch-free stock assessment model 
with application to goliath grouper 
iEpinephe/us itajara) off southern Florida 



Clay E. Porch 

Anne-Marie Ekiund 

Gerald P. Scott 

Southeast Fisheries Science Center 
National Marine Fisheries Service, NOAA 
75 Virginia Beach Drive 
Miami, Florida 33149-1099 
E-mail address clay porchiSnoaa. gov 



Manuscript submitted 30 April 2004 
to the Scientific Editor's Office. 

Manuscript approved for publication 
14 July 2005 by the Scientific Editor. 

Fish. Bull. 104:89-101 (2006). 



The last decade has witnessed con- 
siderable interest in the so-called 
precautionary approach to resource 
management, where human activities 
are curtailed to prevent further envi- 
ronmental degradation without the 
burden of proving that these activities 
are to blame. Fisheries applications 
of the precautionary approach typi- 
cally hinge on the notion that fishing 
pressure should be reduced in a pre- 
determined fashion as certain "limit" 
reference points are approached (FAO, 
1995; Caddy, 1998; Restrepo et al., 
1998). In the United States, the Mag- 
nuson-Stevens Fishery Conservation 
and Management Act (Public Law 
94-265) mandates the development 
of fishery management plans (FMPs) 
that specify criteria for determin- 
ing when a stock is overfished and 
the remedial measures necessary to 
ensure a timely recovery. The National 
Standard Guidelines developed by the 
National Marine Fisheries Service to 
implement the Act require each FMP 
to include an "MSY control rule" that 
comprises two reference points, known 
as the maximum fishing mortality 
threshold (MFMT) and the minimum 
stock size threshold (MSST). When 
the abundance of the stock dips below 
the MSST, special provisions must be 
made to rebuild the stock to the level 
that would support the maximum sus- 
tainable yield within a time frame 
that is as short as possible and that is 
commensurate with the intrinsic pro- 
ductivity of the stock and the needs of 
the fishing community. 



Many modern stock assessment 
methods provide the machinery for 
determining limit reference points 
as well as for appraising where the 
stock is in relation to them and how 
quickly it can be rebuilt. However, 
these methods typically require data 
on total catch or absolute abundance, 
which are not always available. In the 
case of goliath grouper (Epinephelus 
itajara), for example, a recent review 
panel concluded that the catch sta- 
tistics were unreliable and not use- 
ful for assessment purposes (Anon.'). 
Several ad hoc control rules have 
been developed to accommodate such 
"data-poor" situations. One of the 
more common is simply to define the 
MSST in terms of historical indices of 
abundance that supposedly represent 
a desirable stock condition (Annala, 
1993; Cadrin et al., 2004). An advan- 
tage of this type of approach is that it 
is model-free, and nothing is assumed 
concerning the recovery rate of the 
stock. Being model-free, however, is a 
disadvantage with respect to the re- 
quirements of the Magnuson-Stevens 
Act, inasmuch as the recovery time 
cannot be estimated. Moreover, there 
may be other types of information 
that could influence the perception of 
the status of the stock, and it would 
be useful to integrate that informa- 
tion formally into the assessment. 



Anon. 2003. Goliath grouper data 
workshop report. SEDAR3-DW-1, 11 p. 
South Atlantic fishery Management 
Council, 1 Southpark Circle, Charles- 
ton SC 29406. 



90 



Fishery Bulletin 104(1) 



The purpose of this article is to introduce a model- 
based framework for estimating reference points, stock 
status, and recovery times in situations where catch 
data and other measures of absolute abundance are 
unreliable. The specific estimator developed in this 
study is essentially an age-structured production model 
recast in terms relative to pre-exploitation levels. A 
Bayesian estimation scheme is adopted to allow the 
incorporation of pertinent auxiliary information such as 
might be obtained from meta-analyses of similar stocks 
or anecdotal observations. The approach is applied to 
the population of goliath grouper off southern Florida, 
which is believed to have been severely depleted during 
the 1980s and has been protected from all harvest since 
1990 (GMFMC-). 



Materials and methods 



v^ = the relative vulnerability of the remaining 
age classes (which implicitly includes factors 
such as gear selectivity, size limit regula- 
tions, and the fraction of the stock exposed 
to the fishery). 

Relative recruitment (r) is modeled as a first-order log- 
normal autoregressive process. 



 P,-f v-l + Ir.y 



(3) 



where f/^ 

Pr 

n 



the median expectation 
the correlation coefficient; and 
normal-distributed random variates having 
mean and standard deviation a, (ostensi- 
bly representing the effect on recruitment 
of fluctuations in the environment). 



Population dynamics model 

The study period begins when the stock is believed to 
be near virgin levels, such that the relative abundance 
N of each age class a at the beginning of the first year 
is given by 



Na.l 



1 

N„ 



Na-u^ ■" '/(l-e"'"'). 



-Af J . 



a = a^ 

a^<a<A (1) 



where a^ = the youngest age class in the analysis; 

A = a "plus-group" representing age classes A 

and older; and 
M = the natural mortality rate. 

The relative abundance at the beginning of subsequent 
years (y) is modeled by the recursion 



N„ 



A^„-i,v-ie 



-n-l'a-l-A^a-l 



•'',i-l,v-l^ ^ 



N 



A,y-1"= 



-fy-ifA-'^A 



c,. <a<A 
a = A 



(2) 



where r^ = the annual recruitment to age class a,, rela- 
tive to virgin levels; 
F = the fishing mortality rate on the most vul- 
nerable age class; and 



2 GMFMC (Gulf of Mexico Fishery Management Council). 
1990. Amendment number 2 to the fishery manage- 
ment Plan for the reef fish fishery of the Gulf of Mexico, 
31 p. Gulf of Mexico Fishery Management Council. 3018 
North US Highway 301, Suite 1000, Tampa, FL 33619. 



The median is modeled by the Ricker or Beverton-Holt 
spawner-recruit functions recast in terms of the maxi- 
mum lifetime reproductive rate a and relative spawning 
biomass s: 



^'r 



«.v-a,« '"" 


Ricker 


««.v-a. 


Beverton and Holt 


l + (a-l)s„ „ 



(4) 



'.v=ISa-""^'""^''""'^«,.v/I^.-"^"'^A^«-.- 



where E = an index of the per-capita number of eggs 
produced by each age class {E)\ and 
t^ = the fraction of the year elapsed at the time 
of spawning. 

The shapes of these two curves are essentially the same 
as the conventional relationships (Fig. 1); however their 
domain is implicitly limited to the interval Osssl (see 
Appendix 1 for a derivation). 

The fishing mortality rate on the most vulnerable 
age class F is also modeled as a first-order lognormal 
autoregressive process, 






(5) 



where f)p = the median level; 

pp. = the correlation coefficient; and 
11 = normal-distributed random variates having 
mean and standard deviation Op.. 

The median annual is generally assumed to be propor- 
tional to an index of fishing effort /: 



/'f = <Pf, 



(6) 



Porch et al A catch-free assessment model with application to Epinephelus itajora 



91 



where 4> can vary among three eras of exploitation; 
a "prehistoric" period, during which little data are 
available; a "modern" period, when presumably there 
are some data on abundance or mortality rates; and a 
"future" period, when fishing mortality rates are con- 
trolled (input). The absence of data during the "prehis- 
toric era" generally precludes the estimation of annual 
deviations in recruitment (f) or fishing mortality rate 
(b) during that period. 

The average weight or fecundity of the plus group is 
expressed as a function of the average age of the plus- 
group. Initially, it is assumed that the age composition 
of the plus-group is in equilibrium consistent with Equa- 
tion 1, in which case the average age of the plus-group 
at the beginning of the first year is approximately 



-M,, 



a^i = A + 



1-e 



-M^ 



(7) 



Subsequently, the age of the plus-group is updated as 



AN 



A-l.y 



-'^v''.4-l-'W.4-l 



/ — .1 V AT - i" J' X~ ^1 A 



'.4.y+l 



N 



(8) 



A,y+\ 



Reference points 

Equations 1-4 describe the relative dynamics of a popu- 
lation apart from its absolute abundance. As such they 
are suitable for developing management plans where 
the fishing mortality rate is controlled directly (e.g., by 
reducing effort) and the biomass reference points are 
expressed on a relative scale. When the virgin spawn- 
ing biomass itself is used as the reference point, the 
estimated value of s^ is a direct measure of the status 
of the stock. For example, if the management goal is 
to maintain spawning biomass at or above 50% of the 
virgin level, then estimates of s below 0.5 may trigger 
some action to reduce fishing pressure. 

A related reference point is the equilibrium spawning 
potential ratio (Goodyear, 1993), defined as the expected 
lifetime fecundity per recruit at a given F (i/>^) divided 
by the expected lifetime fecundity in the absence of 
fishing (ly'fl): 



Wo 



a=0 



(9) 



-J^Fi^+M, 



As shown in Appendix 2, the corresponding equilibrium 
level of relative spawning biomass (denoted by a tilde) 
may be computed as 



1-1- 



log. 


P 


log, 


a 


ap 


-1 



a-1 



Ricker 

Beverton and Holt 



(10) 



Beverton and Holt 


0-8 


/T^"^ 


^^^ 


0.6 


// ."""^ 




0,4 


\fx 


-♦-2 
-*-10 


0,2 


i/ 


-•-40 


oK 




0.2 0-4 0,6 


0.8 1 


Ricker 




4 -1 


y^^*"^^ 


-•-2 


3 - 


/ X. 


-*-10 
-•-40 


2 - 
1 - 


/x^^^^"^^:!::: 


^ 


.] 


k::-— - 




0.2 0.4 0.6 


0.8 1 


S 




Figure 1 




Examples of scaled Beverton-Holt 


and Ricker 


spawner-recruit relationships for various values of | 


a (maximum lifetime fecundity). The 


labels s and 


r refer to relative spawning biomass 


and relative 


recruitment, respectively. 





Note that s is independent of the vulnerability vector 
I'. Accordingly, MSST definitions based on s will have 
the desirable property of being insensitive to changes 
in fishery behavior. 

Other management plans employ reference points 
such as ^,„„,. or 
recruit statistic 



Fqj, which are based on the yield per 



^V-ly- 



1-e 



-(R' +M„) 



-I /;•,+*', 



(11) 



Fv+M„ 



where w^ is some measure related to the average weight 
of the catch. Inasmuch as there are no terms involving 
the absolute abundance of the stock, the calculation of 
such statistics poses no special problems for the relative 
framework presented in the present study. Prescriptions 
based on the maximum sustainable yield (MSY) are 
slightly more complicated because equilibrium yield is 
the product of equilibrium recruitment R and equilib- 
rium yield per recruit: 



A 1 .p-'/^'n+^a' -Y.F",*I^, 

Y = RpJ^w,,Fv„ ^ .^ e -0 . (12) 



Fv+M„ 



92 



Fishery Bulletin 104(1) 



However, the fishing mortality rate that maximizes 
Equation 12 also maximizes Equation 12 divided by the 
virgir. recruitment /fg (a constant). Thus, i^^v/sv ™3y be 
obtained from 



max 

F 






e ■-" 



(13) 



where s Jp has been substituted for RplR^y 

The values of p and s corresponding to F„,^,^ , Fq ,, 
or F,„,,. may be calculated by means of Equations 9 and 
10, respectively. Note however, that p is no longer the 
target value specified by management, but a derivative 
of the targeted values of F. This means that MSST defi- 
nitions based on s, ,, , s,, ,, and s,., will vary somewhat 

III (l\ '.I 1 I'lsv -^ 

with the behavior of the fishery. In some cases this 
could lead to risk prone situations where the percep- 
tion of stock status changes simply because the fishery 
targets different age groups (i.e., the definition of MSST 
changes rather than the abundance of the resource). In 
the case of MSY. a more stable alternative is to define 
the MSST in terms of a "spawn at least once" policy 
where mature animals are regarded as fully vulnerable 
to the fishery and immature animals area regarded as 
invulnerable. 

Parameter estimation 

The equations above include numerous "unknowns" rep- 
resenting the processes of reproduction, mortality, and 
growth. In the case of "data-poor" stocks like goliath 
grouper, there are insufficient data to estimate all of 
these unknown parameters with an acceptable level of 
precision. However, it is often possible to increase the 
precision of the estimates through the use of Bayesian 
prior probability densities constructed to reflect expert 
opinion (e.g., Wolfson et al., 1996; Punt and Walker, 
1998) or based on meta-analyses involving similar spe- 
cies (e.g., Liermann and Hilborn, 1997; Maunder and 
Deriso, 2003). 

The Bayesian approach to estimation seeks to develop 
a "posterior" probability density for the parameters 
that is conditioned on the data D, P{0 I D). By applica- 
tion of Bayes rule it is easy to show that 



P{0\D)c<P{D\0}P{0). 



(14) 



where P(D\0) is the sampling density (likelihood func- 
tion); and P(0) is the prior density (in this case the 
analyst's best guess of the probability density for 0). 

Estimates for may be obtained by integrating the 
posterior (the classical Bayes moment estimator; cf. 
Gelman et al., 1995) 



Oj = \e^P{D\0)P(0)dej, 0, 6 



(15) 



or by minimizing its negative logarithm (the highest 
posterior density estimator; Bard, 1974) 



min{-log.P(I>l0)-log,P(0)|. 



(16) 



In the present model, a prior needs to be specified 
for the parameters reflecting recruitment (o and £ ), 
mortality (M, (p. d^,, i',, ), fecundity (£„', and growth in 
weight ((f„). It is assumed in the present study that the 
parameters are statistically independent with respect 
to prior knowledge, such that the joint prior is merely 
the product of the marginal priors for each parameter. 
The exceptions are the process error functions for the 
annual recruitment and fishing mortality rate devia- 
tions, f^ and (3^. These are assumed to be autocorrelated 
lognormal variates with negative- log density functions 
of the form 



-logP(f) = 



2a 



fr + 



I*' 



-pr£.y 



- log cr^. 



(17) 



where p, = the correlation coefficient; and 
o'\ = the variance of log^.j;^. 

For stability reasons, it is assumed that £g = 0. 

It is possible, at least in principle, to conduct an as- 
sessment based on prior specifications alone. However, it 
may be difficult to develop sufficiently informative pri- 
ors for some of the parameters, particularly for the fish- 
ing mortality rates. The preferred approach, of course, 
is to condition the estimates on data. With the present 
model it is assumed that catch data are either unavail- 
able or unreliable, otherwise a standard age-structured 
production model (cf. Restrepo and Legault, 1998) would 
be more appropriate. However, time series of catch-per- 
unit-of-effort data or fishery-independent surveys are 
often available even when total catches are not. To the 
extent that changes in these data (r) are proportional to 
changes in the abundance of the population as a whole 
(N), they may be modeled as 



C,..=Q,Y.^,,,Na.ye 



-(f,r„+M„l(, y,^^ 



(18) 



y, ^, - NonnahO.a^., ), 



where / 



t. = 



and 



ndexes the particular survey time series; 

= the proportionality coefficient scaling the 
time series to the relative abundance of the 
population; 

the fraction of the year elapsed at the time 
of the survey; 

the standard deviation of the fluctuations in 
logp c, owing to observation errors or changes 
in the distribution of the stock; and 
the relative vulnerability of each age class to 
the fishery and the /"' survey, respectively. 



The corresponding negative logarithm of the sampling 
density is 



Porch et al : A catch-free assessment model with application to Epinephelus ita/ara 



93 



log P(c\0} = 

'Iog,,(c,,,)-log,, 



II 



2a- 



+ logCT,., 



(19) 



Anecdotal observations may be treated in similar 
fashion. For example the perceptions of constituents on 
the abundance of the resource relative to virgin levels 
{/}) can be modeled as 



<F,';r,*M„)A 



V 1 AT -''■»'•. <.+'"a 



n,- 



r„,y 



a 

7„ ^. ~ Nor?7ial{0,(j„). 



-IM 



(20) 



Table 1 








Anecdotal reports from nine individuals concerning the 
abundance of adult-size goliath grouper i Epinephelus ita- 
jara) in 1990 in relation to their abundance in the 1950s 
(expressed as percent reduction). 


Interviewee Method 






■/f reduction 


1 diver 






60 


2 angler 






70 


3 diver 






75 


4 angler 






90 


5 diver 






90 


6 diver 






95 


7 diver 






95 


8 diver 






97 


9 diver 






98 



Here A^ is the relative contribution of each age class in 
forming the perception of total abundance (e.g., fisher- 
men may never encounter very young fish), A is the time 
of the year most reflective of the period upon which the 
perceptions were based (e.g., the peak of the fishing 
season), and a„ is the standard deviation of the fluctua- 
tions in log. n^,. 

It is not generally possible to obtain consistent esti- 
mates for all of the elements of the covariance matrix 
V (i.e., pp, cfip, p,., cfi^, cfl. , and o'~^). In the case of survey 
data, the variances associated with sampling variability 
are often estimated extraneous to the population model 
(e.g., during the standardization procedure). However, 
there may be additional variance owing to fluctuations 
in the distribution of the stock in relation to the survey 
area (IWC, 1994). To accommodate such possibilities, 
the survey variance parameters are modeled as 



The model outlined above was implemented by using 
the nonlinear optimization package AD Model Builder 
(version 4.5, Otter Research Ltd., Sidney, Canada), which 
provides facilities for estimating the mode and shape of 
the posterior distribution. Confidence intervals for the 
probability of recovery were generated directly from the 
posteriors approximated by the likelihood profile method 
(the accuracy of which was checked by replicating the 
prior distributions without data and by comparing the 
modes of the posterior with the HPD estimates). For 
some quantities confidence limits were computed by us- 
ing normal approximations centered at the HPD esti- 
mate with variances obtained by inverting the Hessian 
matrix. This approach reduced computing time consid- 
erably, but the approximations were poor for confidence 
intervals broader than 80 percent owing to the thick 
tails and skewed nature of the posterior distributions. 



9 



■■x„.,+P„<^ • 



(21) 



where the x^c.,.y and /^ , , 



the annual observation vari- 
ances (estimated outside 
the model); 

a- reflects some overall process 
variance (estimated within 
the model); and 

/j = constant multipliers (usu- 
ally fixed by the analyst 
based on a careful con- 
sideration of the inherent 
variability of the underly- 
ing processes). 

The recruitment variance and correlation coefficient are 
generally inestimable without a good index of recruit- 
ment and may have to be fixed to some moderate values 
(say 0^=0.4 and p=0.5). 



Application to goliath grouper 

Goliath grouper are large, long-lived predators found 
predominantly in the tropical western Atlantic and 
Caribbean Sea. They are among the least wary of reef 
fishes, easily approached by spearfishers and readily 
caught in traps or by hook and line gear. Not surpris- 
ingly, they have declined considerably throughout much 
of their range (Sadovy and Eklund, 1999). Although 
there are few data on the historic abundance of these 
animals in southern Florida, anecdotal reports suggest 
that they were much more abundant during the 1950s 
and 60s than they are now (Table 1). Concerns of over- 
fishing prompted regulators in the U.S. to impose a 
moratorium on the harvest of goliath grouper that has 
remained in effect since 1990. To date, the duration of 
the moratorium has not been specified owing to the pau- 
city of information on their potential recovery rates. 

Spawner-recruit relationship There does not appear 
to be any reliable information on the nature of the 



94 



Fishery Bulletin 104(1) 



spawner-recruit relationship for any grouper species. 
A Beverton and Holt model was assumed in this study 
because it is difficult to envision a mechanism for the 
strong density dependence in mortality rates required 
by the Ricker model. A prior for the value of o (Fig. 2) 
was constructed from a subset of the values collected 
by Myers et al. (1999) that corresponds to larger, highly 
fecund fishes with long life spans (the 'periodic' strate- 
gists of Rose et al., 2001). 



• observed 
— fitted 




Figure 2 

Lognormal prior for the maximum lifetime fecundity 
parameter (o) derived from the values in Myers et al. 
(1999) that correspond to species categorized as periodic 
strategists by Rose et al. (2001). The lognormal density 
was fitted to the values of «-l (with median 9.8 and 
log-scale variance 1.31) and then was shifted 1 unit to 
provide a prior for a. 



Fecundity and growth To date there are insufficient 
data for estimating a fecundity-at-age relationship for 
goliath grouper. We followed Legault and Eklund^ and 
substituted the weight at age relationship: 



E,. 



(<' =1.31xl0"'^/3°'^** 



a<6 

a>6 



(22) 



Z = 200.6(l-e-''^26,a.0.49lj 



1 - 


A /\ M 


B ^^-^ ^^^ 


0.8 - 


/ \ °^ " 


^^ ^ 


0.6 - 


/ \ °^ " 


^ 


4 - 


/ \ ° ''  




>• 0.2  


/ >^ °^ " 




■S 0- 


J ^ ^ 






1 1 -1 U 1 . , . 




o 0.1 0.2 0.3 1 0.2 0.3 4 5 


? M 0, 


Relativ 


C I 
^-"-■"""^^ 1 - 


) 


0,8 - 


\. 0.8 - 






0.6 - 


\^^ 0.6 - 






0.4 - 


N,^ 4 






0.2 - 


^\^ 2  








^ 


_— — 1 


2 4 6 8 10 20 40 60 80 100 


02 1-03 (%) 


Figure 3 


Priors for the mortality rate parameters: (A) lognormal prior for natural 


mortality rate, (B) truncated normal prior for proportionality factor 0j, (C) 


truncated normal prior for multiplier $2 (D) gamma prior for percent reduction 


in F associated with the 1990 harvest ban. The upper and lower boundaries 


for each parameter are as given on the horizontal axes. 



where w = weight in kg; and 

/ = length in cm expressed as a von Bertalanffy 
function of age (see Bullock et al., 1992). 

Natural mortality The maximum observed age of 37 
years (Sadovy and Eklund, 1999) suggests a value for 
M of about 0.11/yr according to the method of Hoenig 
(1983). Legault and Eklund"* suggested a plausible range 
of 0.037 yr to 0.19/yr (midpoint 0.11) based on an analy- 
sis of the fraction surviving to various maximum ages. 
To reflect this uncertainty, a lognormal prior with a 
median of 0.11 and CV of 0.4 was used (Fig. 3A). 

Fishing mortality rate and relative vulnerability A large 
fraction of the recreational landings of goliath grouper 
appear to come from the Ten Thousand Islands area in 
Southwest Florida, where most of the animals caught 
have been between the ages of one and five years. How- 
ever, large animals were often targeted by commercial 
and recreational fishermen in other areas. Accordingly, 
we assumed the vulnerability of goliath grouper gener- 
ally increased with age according to 
the sigmoid-shaped logistic curve 



(23) 



1-i-e 



-(n-a„, l/rf 



Estimates for the parameters ar,o and 
d were obtained by fitting the curve 
(weighted by cumulative mortality at 
age ) to the relative frequency of ages 
in two different data sets. The first 
data set included mostly juveniles 
animals between the ages of and 
5, obtained during creel censuses 
of recreational catches in the Ten 
Thousand Islands area of the Ever- 
glades National Park (see Porch et 



3 Legault, C. M., and A.-M. Eklund. 
1998. Generation times for Nassau 
grouper and jewfish with comments 
on M/K ratios. Sustainable Fisher- 
ies Division Contribution SFD-97/98- 
lOA, 5 p. Southeast Fisheries Science 
Center, 75 Virginia Beach Drive, Mi- 
ami, Florida 33149. 



Porch et al : A catch-free assessment model with application to Epinephelus ita/ara 



95 



al.^). The second data set included mostly adult animals 
obtained opportunistically from recreational and com- 
mercial catches in the eastern Gulf of Mexico (Bullock 
et al., 1992). The SEDAR stock assessment review panel 
based their advice on models that used the former selec- 
tion curve (Kingsley'); however the effect of using the 
latter curve was examined as a sensitivity analysis. The 
two curves are contrasted in Fig. 4A. 

The fishing mortality rate on the most vulnerable age 
class was modeled as follows: 



0/,, 1900 <y< 1979 

'5v'i*2^i979' 1980<y<1990 
^3^1980-89 1990 <y 



(24) 



where /, = a time series of historical effort; and 
(/)j, (t>2, 03, 6y are parameters to be estimated. 

In the present study, effort was assumed to track the 
U.S. Census'' for the number of people living in South 
Florida coastal counties between 1900 and 1980. From 
1980 to 1989 this assumption was no longer required 
owing to the availability of several time series of relative 
abundance (see below). Instead, interannual variations 
in fishing mortality were modeled according to Equation 
5 with median (^.,Fjc,-c|, log-scale variance a7,.=0.15 and 
correlation coefficient p^, = 0.5, which essentially amounts 
to a mild constraint on year-to-year changes in F. The 
nonzero correlation coefficient is intended to reflect the 
momentum in effective fishing effort from one year to the 
next that arises from a combination of market demands 
and the tendency of many fishermen to target only the 
species they are most adept at catching. Even so, the 
relatively large variance term admits substantial inter- 
annual variations if the data warrant them. Moreover, 
runs with p^,= 0.0 (no year-to-year momentum) did not 
produce substantially different results. 

The effect of the harvest moratorium was modeled 
as a percentage ^g of the average fishing mortality rate 
in the 1980-89 period. Relatively uninformative priors 
were used for (p^ and (p^ (Fig. 3, B and C). A somewhat 



* Porch, C. E., A-M. Eklund and G. P. Scott. 2003. An assess- 
ment of rebuilding times for goliath grouper. Sustainable 
Fisheries Division Contribution SFD-2003-0018. Southeast 
Fisheries Science Center, 75 Virginia Beach Drive, Miami, 
Florida 33149. 26 p. 

5 Kingsley, M. C. S., ed. 2004. The Goliath Grouper in 
southern Florida: assessment review and advisory report. 
Report prepared for the South Atlantic Fishery Management 
Council, the Gulf of Mexico Fishery Management Council, 
and the National Marine Fisheries Service, 17 p. South 
Atlantic fishery Management Council, 1 Southpark Circle, 
Charleston SC"29406. 

^ Population of Florida Counties by Decennial Census: 1900 
to 1990, 4 p. 1995. Compiled and edited by Richard L. 
Forstall. Population Division, U.S. Bureau of the Census. 
Washington, DC 20233 



more informative prior with bounds between 0.01 and 
0.5 was used for ip., based on the opinions of members 
of the SEDAR panel (Fig. 3D). 

Survey information Porch and Eklund (2004) have 
developed relative indices of abundance from two visual 
surveys: the personal observations of a professional 
spearfisher (DeMaria") and a volunteer fish-monitor- 
ing program administered by the Reef Education and 
Environmental Foundation (REEF 2000). In addition. 
Cass-Calay and Schmidt*^ have standardized catch rate 
data collected in the Ten Thousand Islands area by the 
Everglades National Park (ENP). The two visual surveys 
are assumed to reflect the abundance of mature fish 
ages 6 and older (based on diver reports of size). The 
ENP catch rate index, on the other hand, is assumed to 
reflect the relative abundance of juveniles with relative 
vulnerabilities given by the dome-shaped gamma func- 
tion (normalized to a maximum of 1): 



■'ENP. a 



° ^,l-°/Oin.y. 
^100% 



(25) 



where Ojoo', - ^he most vulnerable age; and 
CV = the coefficient of variation. 

Estimates for ajQ,,,; (3.47) and CV (0.34) were obtained 
by fitting the mortality-weighted gamma curve to the 
frequency of ages -7 in the Ten Thousand Islands data 
mentioned earlier (for more detail see Porch et al.^). The 
resulting curve is shown in Figure 4B. 

Anecdotal Impressions of stock status Johannes et al. 
(2000) pointed out that local fishermen often disagree 
with the conclusions drawn by scientists in data-poor 
situations and suggest that many times additional data 
will prove the fishermen correct. As mentioned ear- 
lier, expert judgements about the relative abundance 
of a stock can be treated as data or represented by a 
"prior." We collected information on the value of s at 
the time moratoriums began (1990) by interviewing 
fishermen and divers who had been active in southern 
Florida since the early 1960s or before. Specifically, 
interviewees were asked to state their perception of 
the percent reduction in goliath grouper populations 
from the time they began diving to the time the mora- 
torium on catch was imposed (1990). The average per- 
cent reduction reported for large goliath (approximately 
age 6 and older) was 86% (standard deviation of about 
13%, Table 1). This information was modeled as data in 
accordance with Equation 20. 



 DeMaria, D. 2004. Personal, obsery. P.O. Box 420975, 
Summerland Key, FL 33042. 

** Cass-Calay, S. L., and T. W. Schmidt. In review. Stan- 
dardized catch rates of juvenile goliath grouper, Epinephelus 
itajara, from the Everglades National Park Creel Survey, 
1973-1999. 



96 



Fishery Bulletin 104(1) 



Results 

The model was able to fit the ENP index of juvenile goli- 
ath grouper very well but could not reconcile the conflict- 
ing trends indicated by the DeMaria and REEF indices 




Figure 4 

Selection curves used to represent the vulnerability of 
goliath grouper [Epinephelus itajarat to (A) the overall 
fishery and (B) Everglades National Park anglers. The 
logistic curves shown in (Al were fitted to either age- 
composition data derived from the Everglades National 
Park (ENPl creel census or opportunistic samples from 
offshore fishing trips (Bullock et al., 1992). 



DeMaria 




1970 



Year 
Figure 5 

Base model fitted to the four indices of abundance for goliath grouper iEpinepheliis 
itajara) in southern Florida. 



for adult goliath grouper (Fig. 5). The estimated trends 
in spawning biomass were rather uncertain (Fig. 6A), 
but nevertheless indicated a rapid decline to about 5% of 
virgin levels by the time the harvest ban was imposed in 
1990, followed by a significant increase. The estimates of 
fishing mortality were also somewhat uncertain, but gen- 
erally indicated a gradual increase in fishing mortality 
to moderate levels during the 1970s followed by a rapid 
increase during the 1980s (Fig. 6B). The harvest mora- 
torium was estimated to have been about 83% effective 
in reducing fishing mortality, nevertheless losses owing 
to human activities (e.g., illegal harvest and release mor- 
tality of animals caught at depth) were still estimated 
to be substantial (F =0.05/yr). If, in accordance with 
the Gulf of Mexico Management Council's generic Sus- 
tainable Fisheries Act amendment, the limit reference 
point is taken to be the equilibrium spawning biomass 
corresponding to a spawning potential ratio of 50% , then 
the model indicates that current fishing mortality rates 
are near F^^c, and that there is less than a 50% chance 
the stock will recover within 15 years (Fig. 7). 

Sensitivity runs were conducted to examine the im- 
plications of 1) dropping one or more of the indices, 
2) increasing the assumed minimum age represented 
in the REEF and DeMaria indices from 6 to 10, 3) 
assuming that the historical period began in 1950 
rather than 1900 and using the anecdotal informa- 
tion as a tuning index and (4) using the alternate 
fishery selection curve that was fitted to the data from 
Bullock et al. (1992), where adult animals were much 
more vulnerable to the fishery than were juveniles. Of 
these, the results were most sensitive to removal of the 
DeMaria index — the projected trends being much more 
optimistic (Fig. 8). This is because the DeMaria index 
indicates that the adult population increased rapidly 

during the first few years of 
the harvest ban, but then 
suffered a set back in 1999 
and has since leveled off. In 
contrast, the REEF index in- 
dicates that the population 
continued to increase during 
that time. Thus, when the 
DeMaria index is removed, 
the model allows for a faster 
postmoratorium increase in 
the adult population by esti- 
mating a low fishing mortal- 
ity rate of about 0.01/yr (i.e., 
a harvest ban that is 97% 
effective). The fishing mor- 
tality rate estimates for the 
1980s are also lower without 
the DeMaria index inasmuch 
as the DeMaria index indi- 
cates a more precipitous de- 
cline during that time than 
the ENP index (the REEF 
index does not begin until 
1994). 




2000 



Anecdotal 



1950 1960 1970 1980 1990 2000 



Porch et al.: A catch-free assessment model with application to Epinephelus ita/aro 



97 




1950 1960 1970 1980 1990 2000 2010 2020 

Year 




1950 1960 1970 



1980 1990 

Year 



2000 2010 2020 



Figure 6 

Base model predictions of (A) spawning biomass of goli- 
ath grouper (£. itajara) in southern Florida in relation 
to the equilibrium level associated with a spawning 
potential ratio of 50%, s/Sjq,.,, and (B) the apical fishing 
mortality rate on goliath grouper (£. itajara), F^ .^i. 
The lines with small dashes represent approximate 
80% confidence limits and the dashed horizontal lines 
represent the levels associated with an SPR of 50% . 



The sensitivity run with the alternate selection curve 
also produced more optimistic results (Fig. 8). Inasmuch 
as the model now attributes most of the fishing mortal- 
ity to age classes well beyond the age at first maturity 
(see Fig. 41, the spawning stock biomass is estimated 
to have been reduced to a lesser extent (to about 10% of 
virgin levels by 1990 as compared to 5%). Thus, other 
things being equal, recovery requires less time. The lev- 



el of F, 



increased with the alternate selection curve 



because fewer age classes are affected by fishing. 



Discussion 

All of the model formulations examined depicted the 
same qualitative patterns: escalating fishing mortality 
rates and rapidly declining spawning biomass, particu- 
larly during the 1980s, followed by a sharp decrease 
in fishing mortality and strong recovery in spawning 
biomass after the 1990 harvest ban. These trends are 
remarkably consistent with the anecdotal observations 
shown in Table 1 and Figure 5 as well as with the expert 
testimony given during the SEDAR stock assessment 



1 00 1 




~ 80  

o 




CO 

« 60  




>. 




■i 40 • 




J3 
o 




D- 20  


^-^^^^ 




y^ 


1995 2000 2005 2010 2015 2020 


Year 


Figure 7 


Predicted probability that the stock of goliath grouper 


(E. itajara) in southern Florida will have recovered 


to levels exceeding the equilibrium spawning biomass 


associated with a spawning potential ratio of 50%. 



review. The estimated rapid increase in fishing mortality 
during the 1980s appears to reflect a real increase in 
effort that occurred due to elevated demand and selling 
prices (Sadovy and Eklund, 1999), as well as the wide- 
spread use of the LORAN-C navigational system (which 
made it easier for fishermen to relocate productive off- 
shore shipwrecks). Thus, it seems safe to conclude that 
the population was overfished at the time the harvest 
ban was imposed and is currently undergoing a sub- 
stantial recovery. Less clear is the extent to which the 
population has recovered since the harvest ban. 

Using the base model, we estimated that the harvest 
ban has reduced fishing pressure by more than 50%, 
but probably less than 90% (Fig. 9). Thus, there is a 
strong chance that the current fishing mortality rate, 
although greatly reduced as compared to the 1980s, 
remains greater than i^jo'; 'i-^-' above 0.05/yr). This in 
turn translates into less than a 40% chance that the 
population will recover to levels above SgQ,- within the 
next 15 years. Several fishermen have testified that the 
harvest ban is probably less than 90% effective because 
goliath grouper are still taken illegally in places and 
because animals caught and released in deeper water 
often do not survive^; therefore this result does not ap- 
pear unrealistic. 

More optimistic results, implying a 70% to 80% 
chance of recovery within 15 years, were obtained when 
the DeMaria index was excluded or when selection was 
oriented more towards older animals. There does not 
appear to be a strong a priori case for excluding the 
DeMaria index in favor of the REEF and ENP indices. 
Although the coverage is rather limited, the trends of 
the DeMaria index are consistent with those of the ENP 
index (with a suitable time lag) and with anecdotal 
accounts of the trends in other areas." The issue of 
selection is more vexing. It can be argued that the age- 
composition data from the ENP creel census adequately 
reflects the composition of the juvenile catch inasmuch 



98 



Fishery Bulletin 104(1) 



(D 



20 
1 5 - 
1 
5 
00 



I -, 





0) 

> 

o 
o 

0) 6 
^ 4 
io 0.2 



Exclude DeMarIa index 




Selection favors older fish 




Year 



20 -J 










1 5  






^l^ 




1  




? 


}H 




5 • 


J  




1950 


1970 


1990 


2010 



Year 



Year 

Figure 8 

Trends in relative spawning biomass (s/Sjqsj), apical fishing mortality rate (F). 
and the probability of recovery is>s^Q,.^) for two sensitivity runs — one exclud- 
ing the DeMaria index (left) and the other with the selection curve favoring 
older fish (right). 



as it comes from the center of juvenile abundance; how- 
ever most aciults were caught outside this area of abun- 
dance. Thus, the relative contribution of juveniles and 
adults to the overall catch is unclear and the directional 
bias in the fitted logistic selection curve is uncertain. 



1 - 

^ 08 - 






» / \ 






. 


' / \ } 




probabili 

o 




posterior 


1 / > 










> 04 - 
ra 

0) 

'^ 02 - 


_^ '^ 




- 


1 1 1 1 1 1 


20 40 60 80 100 


Percent reduction in F(1 - ^3) 


Figure 9 


Posterior and prior distributions for the effectiveness of 


the 1990 harvest ban in reducing the fishing mortality 


rate F on southern Florida goliath grouper (£. itajara) 


populations (in relation to the fishing mortality rate 


levels observed during the 1980s). 



The only other age composition information that has 
come to light comes from the study by Bullock et al 
(1992). which was not designed to provide a random 
sample of the catch and is probably biased towards 
larger animals caught on offshore wrecks. In principle, 
one could reflect this uncertainty more formally either 
by developing a prior for the selectivity parameters or 
else by weighting the results from the two selection 
models. The scientists on the SEDAR stock assessment 
review panel based their advice on the selection curve 
derived from the ENP data," which is equivalent to 
placing negligible weight on the curve derived from the 
Bullock et al. (1992) data; however they recognized the 
selection curve as an important source of uncertainty 
that is difficult to address without adequate data. 

It is important to emphasize that the Bayesian ap- 
proach adopted in the present study allows one to ex- 
plicitly model the uncertainty about parameters such as 
M, for which no data may exist, but a prior distribution 
covering the plausible range of values may be specified. 
There is, of course, the potential for introducing bias 
when one or more of the priors are based on expert 
opinion or otherwise subjective information. However, 
the same sorts of bias can be introduced by conducting 
sensitivity analyses where the unknown parameters are 
fixed to various values selected by the analysts. Fur- 
thermore, if unbiased data are in short supply, analyses 



Porch et al.: A catch-free assessment model with application to Epinephelus ita/ora 



99 



based on completely uninformative priors will be useless 
for generating advice because the range of plausible 
outcomes is too large. Accordingly, we view the use 
of subjective priors primarily as a vehicle for provid- 
ing more realistic limits on uncertainty and prefer to 
express the model outcomes in terms of probability 
statements. For example, the point estimate from the 
base model indicated that the population would never 
recover to Sgg,, because the fishing mortality rate under 
the harvest ban was still slightly above F^g.,. However, 
consideration of the uncertainty led to the conclusion 
that the chance of recovering to Sggr; within 15 years 
was nearly 40%. 

Some sources of uncertainty have not been adequately 
accounted for in the above assessment. For example, the 
relationship between fecundity and age is unknown. 
We used weight-at-age as a proxy for the relative fe- 
cundity-at-age in our analysis, but it is often the case 
that fecundity increases with age faster than weight. 
If this is true for goliath grouper, then our projections 
would be too optimistic. It should also be remembered 
that the results apply strictly to the goliath grouper 
population in southern Florida. It is believed that the 
center of abundance for the population in U.S. waters 
is off southern Florida, particularly in the Ten Thou- 
sand Islands area, but goliath grouper are known to 
have occurred throughout the coastal waters of Gulf of 
Mexico and along the east coast of Florida, and on up 
through the Carolinas. Inasmuch as goliath grouper 
are not highly migratory, it is possible it may take 
some additional time for the species to fully occupy its 
historical range, thus delaying the overall recovery of 
the U.S. population. 

The primary advantage of the catch-free assessment 
model proposed in the present study is that it does not 
require knowledge of the total number of removals. In 
this light it is worth noting that 623 of the 905 stocks 
included in the 2000 annual report to Congress on the 
Status of Fisheries were listed as having unknown sta- 
tus, often because catch data were either unavailable or 
deemed unreliable. Thus we expect the proposed method 
will become increasingly useful as fishery scientists are 
asked more and more to develop FMPs for poorly moni- 
tored fisheries. The fact that the model estimates the 
population's relative abundance, rather than its absolute 
abundance, is of little consequence when, as is often 
the case, adjustments to the target fishing mortality 
rate or catch quota are made in relation to the levels 
in previous years (Caddy, 2004). Moreover, certain bi- 
ases tend to cancel out when dimensionless quantities 
like relative abundance are used. If, for example, only 
a consistent fraction of the population were sampled, 
then the absolute estimates of abundance would be 
biased but the relative estimates would not (Prager et 
al., 2003). 

The greatest drawback of the catch-free method is 
probably its inability to provide direct estimates of 
the equilibrium catch levels associated with particular 
reference points (e.g., MSY). This situation could per- 
haps be ameliorated by obtaining estimates of absolute 



abundance from a comprehensive short-term survey 
covering the entire range of the animal, in which case 
the relative outputs from the model (including relative 
catch) could be appropriately scaled. Alternatively, a 
long-term monitoring program at select sites located 
throughout the known range of the animal could be es- 
tablished to detect changes in relative abundance under 
various closely monitored trial levels of catch. 



Acknowledgments 

The present paper benefitted greatly from the scrutiny 
given to a related document'' by members of the SEDAR 
stock assessment review panel (R. Allen, L. Barbieri, 
J. Brodziak, M. Cufone, D. DeMaria, M. Kingsley, D. 
Murie, M. Murphy, J. Neer, J. Rooker, R. Taylor, E. 
Toomer, and J. Wheeler). S. Turner, L. Brooks, and two 
anonymous reviewers also gave helpful comments on 
the manuscript. S. Cass-Calay and T. Schmidt secured 
the goliath grouper length-composition data from the 
Everglades National Park creel survey; J. Brusher and J. 
Schull provided the age-length data from their sampling 
program in the Ten Thousand Islands area. 



Literature cited 

Annala, J. 

1993. Fishery assessment processes in New Zealand's ITQ 

system. In Proceedings of the international symposium 

on management strategies for exploited fish populations. 

21-24 October 1992, Anchorage Alaska (G. Kruse, D. M. 

Eggers, R. J. Marasco, C. Pautzke, and T. J. Quinn II, 

eds. ), p. 791-806. Alaska Sea Grant College Program, 

Univ. Alaska, Fairbanks. AL. 
Bard, Y. 

1974. Nonlinear parameter estimation. Academic Press, 

San Diego, CA, 341 p. 
Bullock, L. H., M. D. Murphy, M. F. Godcharles, and M. E. Mitchell. 

1992. Age, growth, and reproduction of jewfish Epineph- 
elus itajara in the eastern Gulf of Mexico. Fish. Bull. 
90:243-249. 

Caddy, J. F 

1998. A short review of precautionary reference points and 
some proposals for their use in data poor situations. FAQ 
Fish. Tech. Pap. 379, 30 p. FAO, Rome. 
2004. Current usage of fisheries indicators and reference 
points, and their potential application to management 
of fisheries for marine invertebrates. Can. J. Fish. 
Aquat. Sci. 61:1307-1324. 
Cadrin, S. X., J. A. Boutillier, and J. S. Idoine. 

2004. A hierarchical approach to determining reference 
points for Pandalid shrimp. Can. J. Fish. Aquat. Sci. 
61:1373-1391. 
FAO (Food and Agricultural Organization). 

1995. Precautionary approach to fisheries. FAO Fish. 
Tech. Pap. 350, 210 p. FAO, Rome. 
Gelman, A., J. B. Carlin, H. 8. Stern, and D. B. Rubin. 

1995. Bayesian data analysis, 526 p. Chapman and 
Hall, London. 
Goodyear, C. P. 

1993. Spawning stock biomass per recruit in fisheries 



100 



Fishery Bulletin 104(1) 



management: foundation and current use. In Risk 
evaluation and biological reference points for fish- 
eries management (Smith, S. J., J. J. Hunt, and D. 
Rivard, ed.), p. 67-81. Can Spec. Pub). Fish. Aquat. 
Sci 120. 

Hoenig, J. M. 

1983. Empirical use of longevity data to estimate mor- 
tality rates. Fish. Bull. 82:898-903. 

IWC (International Whaling Commission). 

1994. Report of the Scientific Committee, Annex D. Report 
of the sub-committee on management procedures. Rep. 
Int. Whal. Comm. 44:74-92. 

Johannes, R. E., M. M. R. Freeman, and R. J. Hamilton. 

2000. Ignore fishers" knowledge and miss the boat. Fish 
Fish. 1:257-271. 

Liermann, M., and R. Hilborn. 

1997. Depensation in fish stocks: a hierarchic Bayesian 
meta-analysis. Can. J. Fish. Aquat. Sci. 54: 1976- 
1984. 

Maunder, M. N., and R. B. Deriso. 

2003. Estimation of recruitment in catch-at-age 
models. Can. J. Fish. Aquat. Sci. 60:1204-1216. 

Myers, R. A., K. G. Bowen, and N. J. Barrowman. 

1999. Maximum reproductive rate offish at low popula- 
tion sizes. Can. J. Fish. Aquat. Sci. 56: 2404-2419. 
Porch, C. E., and A.-M. Eklund. 

2004. Standardized visual counts of goliath grouper 
off South Florida and their possible use as indices of 
abundance. Gulf Mex. Sci. 22:155-163. 

Prager, M. H., C. E. Porch, K. W. Shertzer, and J. F. Caddy. 

2003. Targets and limits for management of fisheries: 
A simple probability-based approach. N. Am. J. Fish. 
Manag. 23:349-361. 



Punt, A. E.. and T. I. Walker. 

1998. Stock assessment and risk analysis for the school 
shark (Galeorhinus galeus) off southern Australia. Mar. 
Freshw. Res, 49:719-731. 
Restrepo, V. R., and C. M. Legault. 

1998. A stochastic implementation of an age-structured 
production model. In Fishery stock assessment models 
(F. Funk, T. J. Quinn II, J. Heifetz, J. N. lanelli. J. 
E. Powers. J. F. Schweigert, P. J. Sullivan, and C.-I. 
Zhang, eds.), p. 435-450. Alaska Sea Grant College 
Program Report No. AK-SG-98-01, Univ. Alaska. Fair- 
banks, AL. 
Restrepo, V. R., G. G. Thompson, P. M. Mace, W. L. Gabriel, L. L. 
Low. A. D. MacCall, R. D. Methot, J. E. Powers, B. L. Taylor, P. 
R. Wade, and J. F. Witzig. 

1998. Technical guidance on the use of precautionary 
approaches to implementing National Standard 1 of the 
Magnuson-Stevens Fishery Conservation and Manage- 
ment Act. NOAA Technical Memo. NMFS-F/SPO-31, 54 
p. National Technical information Center, 5825 Port 
Royal Road, Springfield, VA 22161. 

Rose, K. A., J. H. Cowan, K. O. Winemiller, R. A. Myers, and 
R. Hilborn. 

2001. Compensatory density dependence in fish popu- 
lations: importance, controversy, understanding and 
prognosis. Fish Fish. 2:293-327. 
Sadovy, Y., and A-M. Eklund 

1999. Synopsis of biological data on the Nassau grouper, 
Epinephelus striotiis iBloch, 1792), and the jewfish, 
E. itajara (Lichtenstein, 1822). NOAA Tech. Report 
NMFS 146, 65 p. 

Wolfson, L. J., J. B. Kadane, and M. J. Small. 

1996. Bayesian environmental policy decisions: two case 
studies. Ecol. Appl. 6:1056-1066. 



Appendix 1: reparameterized spawner-recruit 
relationships 



The number of young fish recruiting to a population (R) 
is often relateci to the aggregate fecundity of the spawn- 
ing stock (S) by using one of two functional forms: 



fl = 



abS 



Ricker 

Beverton and Holt 



(A.l) 



b + S 



The parameter a is the slope of the curve at the origin 
and the parameter b controls the degree of density 
dependence. Notice that the domain of both functions 
extends from zero to infinity, whereas in practice there 
must be some limitation on S and R even in the absence 
of fishing owing to environmental constraints (call them 
jSq and i?0' respectively). This being so, we obtain 



12.- 



,*s„ 



l-i-So/6. 



Ricker 

Beverton and Holt 



(A.2) 



The ratio S„/Ry represents the maximum expected life- 
time fecundity of each recruit and a represents the sur- 
vival of recruits in the absence of density dependence. 
Accordingly, the product o = qSq/Rq may be interpreted 
as the maximum possible number of recruits produced 
by each spawner over its lifetime (Myers et al., 1999). 

The dimensionless character of a makes it useful for 
interspecies comparisons, or for borrowing values from 
species with similar life history strategies. Solving for 
b in terms of a one obtains 



log,, a I Sf, 



Ricker 



Sy/(l-a). 



Beverton and Holt 
Substituting Equation A. 3 into Equation A.l gives 



(A.3) 



R- 



aSa 



-s/s„ 



aS, 







Ricker 

Beverton and Holt 



(A.4) 



l-Ka-l)S/S„ 



Porch et al. A catch free assessment model with application to Epinephelus ita/aia 



101 



and, since a = a/J^/Sy, 



R- 



^^^A«l-S«o 



Rn 



aS/Sn 



'l + (a-l)S/S„ 



Ricker 



Beverton and Holt 



(A. 5) 



Dividing through by i?y and defining s as S/Sq gives 
Equation 4. 



Appendix 2: formula for equilibrium 
spawning biomass 

The spawning potential ratio (p) is defined as the number 
of spawners produced by each recruit at equilibrium with 
a given fishing mortality rate F divided by the number 



of spawners per recruit under virgin conditions (F-O). 
This may be written 



Vo Sq / /?o ^f ^ ^0 



(A.6) 



where the tilde signifies equilibrium values. 
At equilibrium we also obtain from Equation 4 



sa^-' 



as 



(l + s(a-l)) 



Ricker 

Beverton and Holt 



(A.7) 



Dividing both sides of Equation A.7 by r, substituting p 
for Equation A.6, and solving for s gives Equation 10. 



102 



Abstract — Fish assemblages were 
investigated in tidal-creek and sea- 
grass habitats in the Suwannee River 
estuary, Florida. A total of 91.571 
fish representing 43 families were 
collected in monthly seine samples 
from January 1997 to December 
1999. Tidal creeks supported greater 
densities of fish (3.89 fish/m-; 83% 
of total) than did seagrass habitats 
(0.93 fish/m- 1. We identified three 
distinct fish assemblages in each 
habitat: winter-spring, summer, and 
fall. Pinfish iLagodon rhomboides), 
pigfish iOrthopristis chrysoptera I, and 
syngnathids characterized seagrass 
assemblages, whereas spot fLeiosto- 
rnus xanthurus), bay anchovy (Anchoa 
mitchiUi), silversides (Menidia spp.), 
mojarras iEucinostornus spp.), and 
fundulids characterized tidal-creek 
habitats. Important recreational and 
commercial species such as striped 
mullet (Mugil cephahis) and red drum 
(Sciaeiiops ocellatus) were found 
primarily in tidal creeks and were 
among the top 13 taxa in the fish 
assemblages found in the tidal-creek 
habitats. Tidal-creek and seagrass 
habitats in the Suwannee River estu- 
ary were found to support diverse fish 
assemblages. Seasonal patterns in 
occurrence, which were found to be 
associated with recruitment of early- 
life-history stages, were observed for 
many of the fish species. 



Fish assemblages found in tidal-creek and 
seagrass habitats in the Suwannee River estuary 

Troy D. Tuckey 

Florida Fish and Wildlife Conservation Commission 

Flonda Wildlife Research Institute 

Apalachicola Field Laboratory 

East Point, Florida 32328 

Present address: School of Marine Science 

Virginia Institute of Marine Science 

College of William and Mary 

PO. Box 1346, Route 1208 Create Road 

Gloucester Point, Virginia, 23062 
E-mail address, tuckeytg'vims edu 

Mark Dehaven 

Department of Agnculture and Consumer Services 
Division of Aquaculture 
11350 SW 153^'^ Ct 
Cedar Key Florida 32625 



Manuscript submitted 20 July 2004 
to the Scientific Editor's Office. 

Manuscript approved for publication 
19 July 2005 by the Scientific Editor. 

Fish. Bull. 104:102-117 12006). 



The Suwannee River estuary, located 
on the gulf coast of Florida, is rela- 
tively pristine and supports commer- 
cial and recreational fisheries. It is an 
unusual estuary, with an orientation 
along the open coastal shoreline, and 
its habitats include oyster bars, mud- 
flats, seagrasses, tidal creeks, and 
an extensive salt marsh (Comp and 
Seaman, 1985). In other estuaries of 
the United States, fish assemblages, 
species abundance, and habitat asso- 
ciations within estuaries have been 
studied extensively. Particular atten- 
tion has focused on estuaries as nurs- 
ery habitats for young-of-the-year 
(YOY) fishes that use seagrasses, 
tidal-creeks, and marshes during their 
early-life stages (Shenker and Dean, 
1979; Bozeman and Dean, 1980; Liv- 
ingston, 1984; Cowan and Birdsong, 
1985; Gilmore, 1988; McGovern and 
Wenner, 1990; Baltz et al., 1993; 
Peterson and Turner, 1994; Rooker 
et al., 1998). In addition, comparisons 
between different habitats within estu- 
arine systems have been conducted to 
evaluate the value of each habitat as a 
nursery (Weinstein and Brooks, 1983; 
Sogard and Able, 1991; Rozas and 
Minello, 1998; Paperno et al., 2001). 
Aside from basic species-composition 
studies of marsh fishes (Kilby, 1955; 
Nordlie, 2000) and fishes that inhabit 



shallow waters near Cedar Key (Reid, 
1954), only one recent study (Tsou 
and Matheson, 2002) has investigated 
the distribution patterns of fishes in 
the Suwannee River estuary. Tsou 
and Matheson (2002) found that the 
nekton community structure for the 
Suwannee River estuary had a strong 
seasonal pattern that was consistent 
among years and followed patterns 
for water temperature and river dis- 
charge. They found assemblages that 
were associated with warm and cold 
seasons, and wet and dry seasons, 
but they did not examine habitat spe- 
cific assemblages (Tsou and Matheson, 
2002). For proper management of fish- 
ery resources, it is beneficial to have 
detailed, current information concern- 
ing the status of all life-history stages 
and associated habitats of species that 
reside in the area, as well as informa- 
tion concerning species interactions 
and associated food webs. Although 
human development in the Suwannee 
River estuary is not a current threat, 
the potential withdrawal of freshwater 
from the Suwannee River for human 
consumption is a possibility and could 
impact fish assemblages found in the 
estuary (Tsou and Matheson, 2002). 

This article describes habitat-spe- 
cific assemblages by examining the 
fish fauna collected in seagrass habi- 



Tuckey and Dehaven: Fish assemblages found in tidal creek and seagrass habitats in the Suwannee River estuary 



103 



tats with those collected in tidal-creek habitats in 
the Suwannee River estuary. We performed com- 
parisons of monthly collections offish found along 
tidal-creek shorelines and those found in seagrass 
habitats to define fish assemblages and incor- 
porated abiotic parameters as potential factors 
influencing the assemblages. We also compared 
length distributions of species in each habitat to 
examine the success of YOY recruitment and the 
subsequent influence of YOY recruitment on fish 
assemblages in order to understand the nursery 
function of each habitat. 



Methods 

Study location 

This study took place in the Suwannee River estu- 
ary, which lies along the gulf coast of Florida, 
extending from just north of the Suwannee River 
to Cedar Key (Fig. 1). The Suwannee River emp- 
ties directly into the Gulf of Mexico forming an 
unusual open estuary that stretches 13 kilometers 
north of the river mouth, southeastward to the 
islands of Cedar Key, and extends approximately 
8 kilometers offshore (Leadon'). The Suwannee 
River estuary is shallow (water depth <2.2 m below 
mean sea level), and has semidiurnal tides with 
a tidal range of 0.7 m. The shoreline is relatively 
undeveloped; the city of Cedar Key (pop. 898) along 
the southeastern edge of the estuary and the small 
town of Suwannee approximately 4.8 kilometers 
inland along the Suwannee River are the only 
populated areas. The remainder of the coastline, 
consisting of the Lower Suwannee and Cedar Keys 
National Wildlife Refuges and the Cedar Key State 
Preserve, is owned by the public. 

Study design 



Randomly selected sites were sampled monthly within 
tidal-creek and seagrass habitats from January 1997 
through December 1999. Juvenile and small adult fish 
from each site were collected using a 21.3-mxl.8-m 
nylon seine with 3.2-mm mesh and a center bag measur- 
ing 1.8-mx 1.8-mx 1.8-m. Sampling methods depended 
on the habitat sampled, and all seines were deployed 
during daylight hours. 

Collections in tidal creeks consisted of six hauls per 
month in 1997 and increased to nine hauls per month 
in 1998 and 1999. Tidal creeks consisted of soft mud, 
deep channels, oyster bars, and steep banks. Shoreline 
vegetation included saltmarsh cord grass {Spartina al- 



ri 




tt.jt Suwannee River 


Suwannee , ■- 
River estuary 


-29 20'N ^i> jfe .^ 


y. ' 






'•••. • ^4 


'( 


"k 


?:t ^ 


Gulf •? ' >^ c^f; 

of • * * ^C ^ 


Mexico • " '"**m.j. / 
■29 10'N • • .'^!fb^^ ^^^^ 

North • • •■ 44i^. 


Key J?_^, ,• i? 

.•  • •• *^ 

Land * >*••• •♦ 


A Tidal-creek seine hauls \f 


• Seagrass seme hauls 


15 3 6 9 12 83-W 



Leadon, C. J. 1979. Unpubl. manuscr. Environmental 
effects of river flows and levels in the Suwannee River sub- 
basin below Wilcox and the Suwannee River estuary, Florida, 
59 p. Suwannee River Water Management District Interim 
Report, 922.5 County Road 49, Live Oak, FL 32060. 



Figure 1 

udy area showing location of tidal-creek and seagrass seine hauls in 
e Suwannee River estuary, located on the gulf coast of Florida. 

terniflora) and needle rush (Juncus roemerianus) near 
the creek mouths and changed to a variety of freshwa- 
ter marsh grasses and terrestrial vegetation upstream. 
The seine was set from a boat in a semicircular pattern 
along the shoreline, retrieved onshore, and sampled an 
average area of 68 m- per haul. Shoreline areas inun- 
dated with vegetation were not sampled if the water 
depth was greater than 0.5 m in order to reduce the 
interference of vegetation during sample collections. 
Sampling in tidal-creeks was limited to the shoreline 
because the water was too deep in the channels to 
deploy the seine. In addition, despite the importance 
of oyster bar habitat, oyster bars located inside tidal 
creeks were not sampled because they interfered with 
the proper deployment of the seine. 

Seagrass habitats generally surrounded the major is- 
lands near Cedar Key, including North Key and nearby 
islands (Fig. 1). In addition, vegetated patches extended 
from North Key northwestward to the mouth of the Su- 
wannee River and were present in shallow areas ap- 
proximately three kilometers west of the Suwannee River 
(Fig. 1). Dominant seagrass species in the Suwannee 
River estuary included turtle grass (Thalassia testudi- 



104 



Fishery Bulletin 104(1) 



num.), manatee grass {Syringodium filliforme), and shoal 
gprass (Halodule ivrightii). Percent coverage was estimat- 
ed visually, or if water clarity was insufficient to visually 
inspect the bottom, bottom samples were collected at 
3-m intervals during deployment of the seine. For ana- 
lytical purposes, areas sampled had to contain at least 
ten percent seagrass to he considered seagrass habitat. 
Samples were classified as "vegetated" or "unvegetated" 
after sampling; and monthly collections varied from one 
to seven hauls per month depending on seagrass cov- 
erage. Seagrass habitats were sampled by pulling the 
seine into the current or wind, whichever was strongest. 
We kept the opening of the net at a constant width by 
maintaining tension on a 15.5-m line that was attached 
between each end pole of the seine while the seine was 
hauled for a distance of 9.1 m. The distance covered by 
the seine was measured by a weighted line from the 
starting point. The net was retrieved by bringing the end 
poles together and pulling the net at an angle around a 
vertical pole that closed the wings of the net and forced 
the catch into the bag. A typical seine haul over seagrass 
habitat covered approximately 140 m-. 

In the field, all fish were identified to the lowest pos- 
sible taxon. counted, and released. Up to 40 individuals 
of species of special interest (important to the commer- 
cial or recreational fishery) and 10 individuals of all 
other species were measured to the nearest millimeter 
standard length (SL). For quality-control purposes, 
three specimens of each species collected were returned 
to the laboratory so that species identification could be 
confirmed. At each site, Secchi depth and water depth 
were measured, and water temperature (°C), salinity, 
dissolved oxygen level (mg/L), and pH were measured 
by using a Hydrolab Surveyors'- water-quality instru- 
ment (Hach Environmental, Loveland, CO). 

Data analysis 

Multivariate analyses were used to compare fish com- 
munity structures along tidal-creek shorelines to those 
found in seagrass habitats (Field et al., 1982). Average 
monthly abundance estimates (number of fish divided 
by the number of hauls) were calculated separately for 
each fish species in each habitat type. Average monthly 
abundance estimates were then converted to percent 
composition to correct for bias introduced by the two 
different net-deployment methods and for the different 
levels of effort in each habitat. Fishes that were not 
identified to species were eliminated (<0.1% of total 
fish collected) except where species complexes, such as 
silversides (Menidia spp.), mojarras (Eucinostomus spp. 
<50 mm SL), menhaden {Brevoortia spp.), and minnows 
{Notropis spp.) were substituted. Species complexes 
were used when meristic characters for juveniles were 
insufficient to distinguish between two or more pos- 
sible species (Eucinostomus spp. and Notropis spp.) or 
where there was possible hybridization (Menidia spp. 
and Brevoortia spp.). 

Fish-community comparisons based on percent spe- 
cies composition by habitat and month were conducted 



by using algorithms in PRIMER (Plymouth Routines 
in Multivariate Ecological Research, vers. 5, Plymouth 
Marine Laboratory, UK) for the study of community 
structure (Clarke and Warwick, 1994). To identify 
fish assemblages, hierarchical agglomerative cluster 
analysis was performed with the Bray-Curtis similar- 
ity matrix calculated on fourth-root transformed per- 
centage data. The fourth-root transformation reduced 
the dominance of abundant species and increased the 
influence of less abundant species in the community 
analysis. The cluster analysis was run on one matrix 
consisting of all transformed fish abundance esti- 
mates collected in tidal-creek and seagrass habitats 
combined. 

To identify species that were responsible for the pat- 
terns observed in the cluster analysis, similarity and 
dissimilarity percentage breakdowns were conducted 
by using the SIMPER procedure in PRIMER (Clarke, 
1993). Average similarities between assemblages were 
analyzed to determine the contribution of each species 
to the overall similarity. This procedure reduced the 
number of species required to explain the patterns 
observed in the cluster diagram and allowed for a sim- 
plified interpretation of the species assemblages. The 
higher the similarity value, the more alike samples 
were within assemblages. Alternatively, dissimilarity 
values were examined to identify species that were 
characteristic of a particular assemblage. Species that 
have high average dissimilarity values and low stan- 
dard deviations are those that contribute consistently 
to samples within their group, with the result that they 
can be used to distinguish between groups. 

The relationship between environmental variables 
and fish community structure was examined by us- 
ing the BIO-ENV procedure. A Spearman rank cor- 
relation test was used to compare ranked values 
from the aforementioned biota similarity matrix to 
ranked values from an environmental similarity ma- 
trix, which was created from environmental variables 
measured in this study. Comparisons were based on 
normalized Euclidean distance. The environmental 
variables used to create the environmental similar- 
ity matrix included pH, water temperature, salinity, 
water depth, and Secchi depth. Dissolved oxygen was 
strongly correlated with water temperature and was 
therefore not included in the analysis because it would 
produce results similar to those produced by water 
temperature. Abiotic variables were standardized by 
subtracting each mean and dividing by the standard 
error to remove any bias associated with the different 
measurement scales. 

The influence of recruitment of YOY fishes in defining 
fish assemblages was examined. By relating increases 
in abundance of species that were identified to be im- 
portant contributors through the SIMPER procedure to 
decreases in their average length in both habitats, we 
were able to identify the timing of juvenile recruitment. 
Length-frequency distributions showed that the seine 
continued to catch larger individuals and therefore the 
decrease in average length was not due to a decrease 



Tuckey and Dehaven: Fish assemblages found in tidal-creek and seagrass habitats in the Suwannee River estuary 



105 




-1 — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — 1— 
Jan Mar May Jul Sep Nov Jan Mar May Jul Sep Nov Jan Mar May Jul Sep Nov 

1997 1998 1999 

Figure 2 

Monthly mean values (± standard error) for water temperature, dissolved oxygen, and 
salinity from January 1997 to December 1999. Circles represent values measured in 
seagrass habitats; triangles represent values measured in tidal-creek habitats. 



in the number of larger fish. The mean length of each 
species in each sample was calculated by habitat, and 
differences in mean length were tested by using fish 
assemblages as the main factor in a one-way ANOVA. 
Multiple comparisons tests (Tukey's test) were then 
conducted to examine significant ANOVA results and 
these tests determined which assemblages contained 
fishes with significantly different lengths. 



Results 

Environmental conditions 

The combination of water temperature, salinity, and 
water depth had the highest correlation (p,^=0.659) 
with the fish assemblages of any possible combination 
of measured abiotic variables. Seasonal patterns were 
observed for water temperature and dissolved oxygen, 



whereas salinity fluctuated in seagrass habitats and 
generally increased during 1999 in tidal-creek habitats 
(Fig. 2). Water temperatures ranged from 7.5° to 32.5°C 
(mean=23.0°C, SE = 0.99) in the seagrass habitats and 
ranged from 10.3° to 33.3°C (mean=23.1°C, SE = 0.91) in 
the tidal-creek habitats. Minimum values of dissolved 
oxygen coincided with the highest water temperatures 
in each habitat and ranged from 2.9 to 12.7 mg/L in the 
seagrass habitats and from 2.9 to 13.6 mg/L in the tidal- 
creek habitats. Salinity, however, was lower during all 
seasons in the tidal-creek habitats than in the seagrass 
habitats (Fig. 2). Mean salinity in the seagrass habitats 
was 27.1%f (SE = 0.91) and ranged from nearly fresh 
(1.3^?^) to marine (34.8%f ) depending on river discharge, 
whereas in the tidal creeks, mean salinity was 9.5%c 
(SE = 0.81) and ranged from 0.0%t. to 29. 0%^. An unusu- 
ally high period of rain during February and March of 
1998 decreased salinity values in the tidal-creek and 
seagrass habitats. 



106 



Fishery Bulletin 104(1) 




Jan Mar May Jul Sep Nov Jan Mar May Jul Sep Nov Jan Mar May Jul Sep Nov 



1997 



1998 



1999 



Figure 3 

Cumulative number of species collected in both tidal-creek and seagrass 
habitats (open circles) and number of species collected in tidal-creeks 
(trianglesi and seagrass habitats (filled circles) by year and month 
from January 1997 to December 1999. 



Fish fauna 

At the conclusion of the three-year study, 111 fish species 
and 4 additional species complexes had been collected. 
During the first year of the study, 61 species were col- 
lected in samples taken in tidal-creek habitats, and 
48 species were collected in seagrass habitats (Fig. 3). 
Thirteen new species were added to the species list from 
tidal creek samples and 20 new species were added to 
the list from samples taken in seagrass habitats during 
1998. During the final year of sampling only six addi- 
tional species were collected in tidal creeks and 12 new 
species were collected in seagrass habitats. Overall, tidal 
creeks contained greater relative (uncorrected for gear 
efficiency) densities offish (3.89 fish/m'-) compared with 
seagrass habitats (0.93 fish per m-). In seagrass habi- 
tats, 15,395 individuals were collected in 118 samples 
that covered approximately 16,520 m- of seagrass habi- 
tat (Table 1). In tidal-creek habitats, a total of 76,176 
individuals were collected from 288 samples that covered 
approximately 19,584 m- of tidal-creek shoreline habitat 
(Table 2). Thirty five species were restricted to seagrass 
habitats, and another 35 species were collected only in 
tidal creeks. The remaining 45 species were collected 
in both habitats at least once during the study. Overall, 
twelve families were restricted to seagrass habitats: 
phycid hakes (Phycidae), toadfishes (Batrachiodidae), 
batfishes (Ogcocephalidae), flyingfishes (Exocoetidae), 
cardinalfishes (Apogonidae), barracudas (Sphyraenidae), 
wrasses (Labridae), combtooth blennies (Blenniidae), 
mackerels (Scombridae), triggerfishes (Balistidae), box- 
fishes (Ostraciidae), and porcupinefishes (Diodontidae; 
Table 1). Seven families were restricted to tidal-creek 
habitats: minnows (Cyprinidae), sunfishes (Centrarchi- 
dae), killifishes (Cyprinodontidae), gars (Lepisosteidae), 



eagle rays (Myliobatidae), pikes (Esocidae), and livebear- 
ers (Poeciliidae; Table 2). 

Fish assemblages 

A clear separation of fish assemblages was identified 
and indicated by two main branches that corresponded 
to fishes found in seagrass habitats and those found in 
tidal-creek habitats (Fig. 4). There were two months iden- 
tified from seagrass samples (January 1997 and March 
1998) that had a species composition that was more 
closely linked to samples taken from tidal creeks. 

Seagrass habitats 

Seasonal fish assemblages in seagrass habitats were evi- 
dent during all three years of the study, which included 
winter-spring, summer, and fall assemblages (Fig. 4). 
The winter-spring assemblage consisted principally 
of pinfish {Lagodon rhomboides). pigfish (Orthopris- 
tis chrysoptera). dusky pipefish (Syngnathus floridae), 
southern puffer (Sphoeroides nepheliis), and gulf pipe- 
fish [Syngnathus scovelU). which together accounted for 
more than 95% of the cumulative percent similarity. The 
summer assemblage had a higher average similarity 
value (43.67) than did the winter-spring assemblage 
(42.56) and consisted of more than 21 species. Eleven 
of these species — silver perch (Bairdiella chrysoura), 
S. floridae, bay anchovy {Anchoa mitchilli), L. rhom- 
boides, Eucinostomus spp., spotted seatrout (Cynoscion 
nebulosus), S. scovelli, planehead filefish (Monacanthiis 
hispidus), striped anchovy {Anchoa hepsetus). inshore liz- 
ardfish {Synodus foetens), and O. chrysoptera — accounted 
for more than 75% of the cumulative similarity of the 
summer assemblage. The fall assemblage had the high- 



Tuckey and Dehaven Fish assemblages found in tidal creek and seagiass habitats in the Suwannee Rivei estuary 



107 













Table 1 




















Taxonomic list of individuals collected in seagrass habitats 


for each species by month and total number collected, a 


11 years 


combined. 






























Family 


Species 


Jan 


Feb 


Mar 


Apr 


May 


Jun 


Jul 


Aug 


Sep 


Oct 


Nov 


Dec 


Total 


Dasyatidae 


Dasyatis sabina 




















1 





1 


3 


1 


1 


7 




Dasyatis say 

















2 




















2 


Ophichthidae 


Myrophis punctatus 











1 


























1 


Clupeidae 


Harengula jaguana 




















6 


230 





162 








398 




Bi-evoortia spp. 








1 





























1 




Opisthonema ogliniini 





C) 











4 


1 


4 


35 


9 








53 




Sardinella aurita 





u 

















16 














16 


Engraulidae 


Anchoa hepsetus 














1 


540 


2 


1 


235 


13 


3 





795 




Anchoa mitchilli 








4 


3 





38 


1567 


3838 


632 


3504 


53 





9639 


Ariidae 


Arius felis 

















1 


1 


14 


45 


1 








62 




Bagre marinus 


























1 











1 


Synodontidae 


Synod us foetens 














1 


2 


3 


8 


3 


1 


1 


1 


20 


Gadidae 


Urophycis floridana 



































6 


6 


Batrachiodidae 


Opsanus beta 


























1 











1 


Ogcocephalidae 


Ogcocephalus radiatus 











1 


























1 


Exocoetidae 


Hyporhamphus meeki 














1 


7 


2 

















10 


Belonidae 


Strongylura marina 

















1 


3 


1 














5 


Atherinidae 


Membras niartinica 





10 








1 


10 


49 


21 


13 


15 


3 





122 




Menidia spp. 
































1 


2 


3 


Syngnathidae 


Hippocampus erectus 





























1 








1 




Hippocampus zosterac 





1 














1 


1 


2 





1 


2 


8 




Syngnathus floridae 


2 


7 


1 


9 


2 


34 


33 


53 


48 


26 


52 


38 


305 




Syngnathus louisianae 











1 





3 


4 


2 


1 


4 


3 





18 




Syngnathus scoveHi 





3 


6 


4 





11 


32 


4 


9 


10 


13 


13 


105 


Serranidae 


Mycteroperca microlepis 





























1 








1 




Centropristis striata 

















1 


2 


7 


12 


3 


74 


12 


111 




Serraniculus pumilio 
































10 





10 




Serranus subligarius 


























2 











2 


Apogonidae 


Astrapogon alutus 





























1 








1 


Carangidae 


Caranx hippos 





























2 








2 




Chloroscombrus chrysurus 




















2 


12 


5 











19 




Oligoplites saurus 

















7 


2 


2 


4 


2 








17 




Selene vomer 





























2 








2 


Lutjanidae 


Lutjanus griseus 























1 





1 








2 




Lutjanus synagris 




















3 








1 


2 





6 


Gerreidae 


Eucinostomus gula 


























6 


6 


4 


1 


17 




Eucinostomus harengulus 

















1 





2 


1 











4 




Eucinostomus spp. 

















34 


172 


162 


73 


13 


54 


36 


544 


Haemulidae 


Haemulon plumieri 




















1 


6 


11 


4 


3 


7 


32 




Orthopristis chrysoptera 


1 





6 


26 


738 


30 


38 





8 


4 





1 


852 


Sparidae 


Diplodus holbrooki 














7 


17 





4 


1 


7 








36 




Lagodon rhomboides 


17 


65 


63 


87 


21 


54 


24 


27 


40 


36 


32 


87 


553 


Sciaenidae 


Bairdiella chrysoura 











3 


24 


159 


273 


59 


184 


44 


2 


1 


749 




Cynoscion arenarius 




















1 


24 














25 




Cynoscion nebulosus 

















4 


70 


37 


10 


6 


1 





128 




Leiostom u s xa n th u ru s 


2 


34 


1 

















2 











39 




Menticirrhus a m erica nus 

















2 


13 


119 


2 





1 





137 




























continued 



108 



Fishery Bulletin 104(1) 



Table 1 (continued) 


Family 


Species 


Jan 


Feb 


Mar 


Apr 


Maj 


Jun 


Jul 


Aug 


Sep 


Oct 


Nov 


Dec 


Total 


Sciaenidae 


Menticirrhus saxatilis 








1 








7 





5 











2 


15 


ifont.i 


Sciaenops ocellatus 
































9 





9 


Ephipipidae 


Chaetodipterus faber 




















2 


12 


2 











16 


Mugilidae 


Mugil cephalus 





5 


1 





























6 




Mugil curema 











1 


























1 


Sphyraenidae 


Sphyraena borealif: 














4 























4 


Labridae 


Halichoeres bivittatus 























3 














3 




Lachnolaimus maximus 


























2 


1 








3 


Blennhdae 


Cliasmodes saburrae 




















9 





5 





2 





16 




Parablennius mannoreus 























1 








1 





2 




Hypsoblennius hentzi 


























1 








2 


3 




Hypleurochilus geminatus 




















2 

















2 


Gobiidae 


Gobionellus boleo.'ioma 




















9 

















9 




Gobiosoma bosc 




















29 











1 





30 




Gobiosoma longipala 





























1 








1 




Gobiosoma i-obiiatuni 











2 








7 


2 





1 





1 


13 




Microgobius gulnsiiti 




















3 


2 








2 





7 




Microgohiu s thala usin us 























77 














77 


Scombridae 


Scomberomorus maculatu 


■; 























1 











1 


Triglidae 


Prionotus scitulus 





2 











2 


5 


4 


2 


1 





1 


17 




Prionotus tribulus 























2 








1 





3 


Bothidae 


Paralichthys albigutta 








2 


2 








2 





2 











8 




Etropus crossotus 





3 


1 











2 


1 


1 








1 


9 




Etropus microstomus 








1 





























1 


Cynoglossidae 


Symphiiriis plagiusa 








1 








1 


1 


9 











1 


13 


Soleidae 


Achirus lineatus 




















1 


7 














8 




Trinectes maculatus 


























1 











1 


Balistidae 


Aluterus schoepfi 

















1 


1 


2 





1 








5 




Monacanthus citiatus 























2 


10 


8 


2 


19 


41 




Monacanthus hispidus 














2 


10 


1 


36 


3 


9 


7 


25 


93 


Ostraciidae 


Lactophrys quadricornis 








2 








2 





6 


6 


3 


3 


1 


23 


Tetraodontidae 


Sphoeroides nephelus 








2 


16 


4 


2 


2 


2 


1 


4 


2 


2 


37 


Diodontidae 


Chilomycterus schoepfi 





1 


2 





4 


2 





11 


5 


10 


5 


9 


49 


Column total 




22 


131 


95 


156 


810 


989 


2382 


4839 


1429 


3921 


349 


272 


15,395 


Number of hauls 




6 


10 


9 


4 


5 


15 


10 


11 


11 


11 


10 


16 


118 



est average similarity level at 48.90, and nine species — 
S. floridae, M. hispidus, black seabass (Centropristis 
striata), L. rhomhoides, Eucinostomus spp., fringed file- 
fish {Monacanthus ciUatus), S. scovelli, striped burrfish 
(Chilomycterus schoepfi), and S. nephelus — accounted for 
more than 91% of the cumulative similarity. Lagodon 
rhomhoides and S. floridae were characteristic of all 
three assemblages, and Eucinostomus spp., S. scovelli, 
and M. hispidus were important contributors to the 
summer and fall assemblages. Other abundant species 
from seagrass assemblages included southern kingfish 
{Menticirrhus americanus), rough silverside {Membras 
niartinica), and gobiids, particularly green goby {Micro- 
gobius thalassinus) and naked goby (Gobiosoma bosc). 



A significant drop in salinity during March 1998 
corresponded to an alteration in the fish community 
collected in seagrass habitats. The species present in 
the seagrass habitats during March 1998 were more 
consistent with species collected in tidal creeks and 
included A. mitchilli, Brevoortia spp., and spot (Leiosto- 
mus xanthurus), none of which were collected in March 
of 1997 or 1999 in seagrass habitats. Samples collected 
during March 1997 and 1999 contained individuals 
more characteristic of seagrass habitats, such as C. 
schoepfi, L. rhomhoides, O. chrysoptera, and iS. nephelus 
in 1997 and scrawled cowfish (Acanthostracion quadri- 
cornis), O. chrysoptera, L. rhomhoides, S. floridae, and 
S. scovelli in 1999. 



Tuckey and Dehaven Fish assemblages found In tidal-creek and seagrass habitats in the Suwannee River estuary 



109 











Table 2 




















Taxonomic list of individuals collected in tidal-creeks for each 


species by month and total number collected 


all years combined. 


Family 


Species 


Jan 


Feb 


Mar 


Apr 


Ma> 


Jun 


Jul 


Aug 


Sep 


Oct 


Nov 


Dec 


Total 


Dasyatidae 


Dasyatis sabina 








6 


1 


1 











1 


1 


3 





13 


Myliobatidae 


Rhinoptera bonasus 














3 























3 


Lepisosteidae 


Lepisosteus osseus 

















2 


2 


1 














5 




Lepisosteus platyrhincus 








1 


1 





1 


1 





2 


2 








8 


Ophichthidae 


Myrophis punctatus 














1 























1 


Clupeidae 


Harengula jaguana 














2 


1 


2 


9 


10 


2 








26 




Brevoortia spp. 


15 


139 


93 


400 


394 


7 


5 














1 


1054 




Sardinella aurita 




















22 

















22 


Engraulidae 


Anchoa hepsetus 


3 








1 


175 


482 


71 


33 


7 





3 





775 




Anchoa mitchilli 


86 


497 


825 


108 


612 


329810,329 


6970 


5776 


7436 


1637 


2107 39,681 


Cyprinidae 


Notemigoniis crysoleucas 





2 


1 





























3 




Notropis spp. 














4 


9 




















13 


Ariidae 


Arius felis 























2 


3 











5 


Esocidae 


Esox niger 











1 


1 























2 


Synodontidae 


Synod us foe tens 











3 


7 











1 





3 


1 


15 


Belonidae 


Strongylura marina 











2 


4 


1 


8 


1 


1 


1 








18 




Strongylura notata 


2 

















3 


2 


8 





1 





16 




Strongylura timucu 














11 


1 


8 


3 


11 











34 


Cyprinodontidae 


Adinia xenica 


7 


7 





4 


1 





12 


23 





153 


14 


25 


246 




Cyprinodon variegatus 


4 








1 




















2 


3 


10 




Lucania goodei 








11 





























11 




Lucania parva 





2 


2 








1 


2 


15 


4 








6 


32 


Fundulidae 


Fundulus confluentus 


2 


























1 





3 


6 




Fundulus grandis 


33 


39 


6 


28 


9 


30 


21 


127 


9 


39 


170 


218 


729 




Fundulus majalis 


29 


12 


24 


7 





18 





42 


3 


10 


13 


65 


223 




Fundulus seminolis 


7 


2 





1 





14 





14 


5 


6 





1 


50 


Poeciliidae 


Gambusia holbrooki 


6 





59 


2 


4 


1 





1 


12 


4 


1 


7 


97 




Heterandria formosa 





1 























1 








2 




Poecilia latipinna 


1 











2 


2 





17 





4 


4 


12 


42 


Atherinidae 


Membras martinica 








4 





9 


566 


2286 


28 


114 


7 


1 





3015 




Menidia spp. 


383 


429 


201 


265 


145 


695 


982 


571 


814 


602 


749 


358 


6194 


Syngnathidae 


Syngnathus floridae 





1 




















1 











2 




Syngnathus louisianae 


























1 





1 





2 




Syngnathus scovelli 


2 


1 


2 





1 





1 


7 


5 


1 


2 


11 


33 


Serranidae 


Diplectrum hivittatum 
































1 





1 


Centrarchidae 


Enneacanthus gloriosus 








2 





























2 




Elassoma zonatum 














4 























4 




Lepomis gulosus 





1 


1 





























2 




Lepomis niacrochirus 


























1 











1 




Lepomis marginatus 











1 


























1 




Lepomis microlophus 














1 























1 




Lepomis punctatus 


1 








1 


1 


1 


2 


2 


3 


2 








13 




Micropterus salmoides 








3 


1 


14 


4 





5 


8 


2 








37 


Carangidae 


Chloroscombrus chrysurus 




















1 


3 


15 


2 








21 




Otigoplites saurus 

















8 


53 


25 


47 


17 


6 





156 




Trachinotus falcatus 





























3 








3 




























continued 



110 



Fishery Bulletin 104(1) 



Table 2 (continued) 


Family 


Species 


Jan 


Feb 


Mar 


Apr 


May 


Jun 


Jul 


Aug 


Sep 


Oct 


Nov 


Dec 


Total 


Lutjanidae 


Lutjanus griseus 

















2 


2 


9 


8 


6 


9 


1 


30 


Gerreidae 


Eucinostoinus giila 


1 




















4 


7 


25 


49 


3 


89 




Eucmostomus harengulu^ 





25 


5 


2 


6 


3 


41 


150 


88 


100 


65 


45 


530 




Eucinostoinus spp. 


17 


38 


6 


1 


1 


87 


1035 


1082 


503 


862 


691 


358 


4681 


Haemulidae 


Orth opristis ch rysoptera 























2 














2 


Sparidae 


Archosargus 
probatocephatus 


7 


1 


1 





1 


1 





1 








1 





13 




Lagodon rhomboides 


87 


231 


146 


142 


67 


114 


111 


85 


38 


29 


53 


10 


1113 


Sciaenidae 


Bairdiella chrysoura 








7 


7 


200 


154 


58 


246 


117 


4 


3 





796 




Cynoscion arenarius 











5 


236 


102 


90 


91 


46 


76 


5 





651 




Cynoscion nebulosus 


3 


1 


1 





5 


11 


21 


53 


62 


23 


36 


2 


218 




Leiostomus xanthurus 9526 


2044 


834 


771 


274 


42 


125 


24 


6 


5 


2 


31 


13,684 




Menticirrhus americanus 














3 


12 


50 


36 


12 


10 


5 





128 




Micropogonias undulatiis 


4 


4 





5 


2 











1 





6 


2 


24 




Pogonias cromis 














2 


2 





3 





1 





1 


9 




Sciaenops ocellatus 


26 


28 


20 


15 


3 


7 


4 


1 


1 


48 


105 


65 


323 


Ephipipidae 


Chaetodipterus faber 

















1 


3 


1 


3 











8 


Mugilidae 


Mugil cephalus 


175 


159 


59 


25 


2 


39 


22 


7 


18 


1 


6 


6 


519 




Mugil curema 


1 














25 


2 














1 


29 




Mugil gyrans 


























2 








4 


6 


Gobiidae 


Bathygobius soporator 











4 








2 


1 


18 


5 


6 


3 


39 




Gobionellus boleosoma 








3 


6 


9 





1 


6 


15 





4 





44 




Gobiosonia bosc 


8 


3 


14 


7 


1 


40 


21 


17 


52 


20 


18 


67 


268 




Gobiosonia robustum 





6 


3 


1 





8 


7 


10 


6 





4 


15 


60 




Microgobius gulosus 


5 


23 


7 


5 


4 


7 





25 


5 


8 


3 


14 


106 




Microgobius thalassinus 




















1 


1 


1 





1 





4 


Triglidae 


Prionotus scitulus 

















1 




















1 




Prionotus tribulus 











1 














1 


11 


4 





17 


Bothidae 


Paralichthys albigutta 


5 


1 











2 








1 











9 




Paralichthys lethostigma 








1 








1 




















2 




Etropus crossotus 

















1 




















1 


Cynoglossidae 


Symphurus plagiusa 


6 














2 


2 


1 


13 


5 


1 


5 


35 


Soleidae 


Achirus lineatus 

















2 


4 


9 


10 


8 


2 





35 




Tnnectes maculatus 


1 


8 


10 





5 


13 


1 


13 


2 


4 


2 


1 


60 


Tetraodontidae 


Sphoeroides nephelus 











1 


6 














4 





1 


12 


Column total 


10,453 


3705 


2358 


1826 


2233 


5821 1 


5.414 


9779 


7898 


9551 


3685 


3453 


76,176 


Number of hauls 




24 


24 


24 


24 


24 


24 


24 


24 


24 


24 


24 


24 


288 



Recruitment of YOY fishes had an influence on 
defining fish assemblages in seagrass habitats. The 
winter-spring assemblage was dominated by YOY L. 
rhomboides and O. chrysoptera (Table 3), which had 
significantly shorter standard lengths than did the 
other assemblages (Table 4). The summer assemblage 
showed an increase in the number of species and an in- 
crease in their abundance, but there were no significant 
differences in length for YOY for any species between 
assemblages. 



Tidal-creek habitats 

Three fish assemblages (winter-spring, summer, and 
fall) were identified from samples taken in tidal-creek 
habitats and reflected similar seasonal patterns com- 
pared with fish assemblages identified from seagrass 
habitats (Fig. 4). The winter-spring assemblage had 
an average similarity level of 51.76 and consisted of L. 
xanthurus, Menidla spp., A. initchilli, L. rhomboides, M. 
cephalus, and red drum (Sciaenops ocellatus). These six 



Tuckey and Dehaven Fish assemblages found in tidal creek and seagrass habitats in the Suwannee River estuary 



111 



Bray-Curtis similarity 







r- 


sg 12 


Seagrass (fall) 




!g_i-' 

so 11 






y 


ig.l : 
ig.11 

sg.?. 

sg_6 
sg_9 

?g 8 


Seagrass 






!g_' 


(summer) 






sg 10 

!g_9 

59.6. 

sg_6. 






y 


S0_9. 
sg 10 

sg-5. 
sg_5. 

SSJ. 


Seagrass 






sg_.'. 

so 3 


(winter and 


spring) 




sg_4 

sg_'. 
sg.i 
sg_i. 
sa.5. 






^ 


=9.3. 
S9_1. 
so 3 






r 


tc 6 
tc 7 

tc 6 
tc 9 
tc 6 
tc To 


Tidal-creek 






'c_11. 


(summer) 






tc 8 
tc 9 
tc 7 
tc 9 

tc e 

tc_10. 






> 


tc 12 

tc a 


Tidal-creek 






Ic 12 


(tall) 






Is 11 
tc 11 






r 


tc 1 
tc 2 
1c 1 
lc.2. 
tc 3 


Tidal-creek 






1c_l_ 


(winter and 


spnng) 




tc 5 
tc 4 
tc 5 
tc S 
Ic 3 
Ic 4 






V- 


tc 4 
tc.S. 




Figure 4 

Results of cluster analysis for fish assemblages found in the Suwannee 
River estuary. SG = seagrass habitats, TC = tidal-creek habitats, l = January, 
2 = February, etc., and the year is denoted by the last two digits of the year. 
For example, SG_1_97 corresponds to samples collected during January of 
1997 in seagrass habitats. Note: Clusters are free to rotate at the point at 
which the lines branch. 



species accounted for more than 68% of the cumulative 
similarity within the winter-spring assemblage. The 
summer assemblage had an average similarity level 
of 62.29 and was characterized by ten species that 
accounted for 70.65% of the cumulative similarity; A. 
mitchilli, Menidia spp., Eucinostomus spp., sand seatrout 
(Cynoscion arenarius), B. chrysoura, C. nebulosus, L. 
rhomboides, E. harengulus, M. martinica, and leather- 
jacket (Oligoplites saurus). The fall assemblage had an 
average similarity level of 60.36 and was characterized 
hy Eucinostomus spp., Menidia spp., gulf killifish (Fiin- 
dulus grandis), A. mitchilli, E. harengulus, diamond kil- 
lifish (Adinia xenica), clown goby (Microgobius gulosus). 



and M. cephalus, which accounted for more than 67% of 
the cumulative similarity. Species tolerant of low salin- 
ity, such as A. xenica, marsh killifish {Fundulus con- 
fluentus), mosquitofish (Gambusia holbrooki), and least 
killifish {Heterandria formosa) were commonly collected 
in tidal creeks. There were 35 species collected, includ- 
ing groups such as cyprinids, cyprinodontids, poeciliids, 
lepisosteids, and centrarchids, which were restricted 
entirely to tidal creeks (Table 2). 

Seasonal recruitment of juvenile fishes to tidal-creek 
habitats was evident and remained consistent through- 
out the study resulting in three clearly defined assem- 
blages. Young-of-the-year L. xanthurus recruited to 



112 



Fishery Bulletin 104(1) 









Table 3 










Results of SIMPER procedure showing the average percent dissimilarity id'^i) for important species between seagrass assem- 
blages, di is the average contribution of the ith species to the disssimilarity between groups; bi/SD(di) is the ratio between the 
average contribution of the /'*' species and the standard deviation of 6i; Cum di'7c is the cumulative contribution to the total dis- 
similarity. Species are listed in decreasing contribution to average dissimilarity. 


Species 


Average 


abundance 


di 


(V/SD(cV) 


di^-i 


Cumulative di% 


Eucinosto??ius spp. 


winter-spring 


summer 


4.8 


1.38 


6.31 


6.31 


0.24 




9.9 


Lagodon rho/nboides 


10.75 




1.83 


4.5 


1.69 


5.91 


12.22 


Bairdiella chrysoura 


1.01 




12.76 


4.38 


1.25 


5.75 


17.97 


Orth opristis ch rysoptera 


26.58 




0.94 


3.66 


1.17 


4.81 


22.78 


Monacanthus hispidus 


0.09 




1.89 


3.25 


1.22 


4.27 


27.05 


Syngnathus floridae 


1.2 




3.88 


3.24 


1.1 


4.25 


31.3 


Centropristis striata 


0.01 




1.82 


2.68 


0.99 


3.52 


34.82 


Anchoa hepsetus 


0.03 




11.95 


2.67 


0.8 


3.51 


38.33 


Syngnathus scovelli 


0.8 




1.13 


2.48 


1.36 


3.25 


41.58 


Sphoeroides nephelus 


1.24 




0.19 


2.27 


1.2 


2.99 


44.57 


Anchoa mitchilli 


0.33 




4.46 


2.27 


0.95 


2.98 


47.55 


Chilomycterus schoepfi 


0.18 




0.46 


2.12 


1.07 


2.78 


50.33 


Anchoa niilchilli 


winter-spring 


fall 


10.87 


2.32 


13.31 


13.31 


0.33 




478.1 


Lagodon rhomboides 


10.75 




4.57 


6.43 


1.36 


7.87 


21.18 


Orth opristis ch rysoptera 


26.58 




1.33 


4.73 


1.05 


5.79 


26.97 


Bairdiella chrysoura 


1.01 




8.26 


3.39 


1.58 


4.15 


31.12 


Leiostomus xanthurus 


0.67 




0.14 


3.26 


0.61 


3.99 


35.11 


Syngnathus scovelli 


0.8 




0.71 


2.83 


1.19 


3.46 


38.57 


Syngnathus floridae 


1.2 




3.26 


2.81 


1.27 


3.44 


42.01 


Eucinostomus spp. 


0.24 




5.38 


2.68 


1.24 


3.28 


45.29 


Harengula jaguana 







9.74 


2.57 


1.24 


3.15 


48.44 


Sphoeroides nephelus 


1.24 




0.12 


2.55 


0.91 


3.12 


51.56 


Anchoa mitchilli 


summer 




fall 


6.13 


1.98 


9.30 


9.30 


4.46 




478.1 


Eucinostomus spp. 


9.9 




5.38 


3.11 


1.34 


4.72 


14.02 


Bairdiella chrysoura 


12.76 




8.26 


2.70 


1.21 


4.10 


18.12 


Anchoa hepsetus 


11.95 




0.48 


2.36 


0.98 


3.59 


21.71 


Syngnathus floridae 


3.88 




3.26 


2.35 


1.05 


3.56 


25.27 


Monacanthus hispidus 


1.89 




0.38 


2.34 


1.14 


3.55 


28.82 


Harengula jaguana 


2.79 




9.74 


2.15 


1.25 


3.27 


32.09 


Centropristis striata 


1.82 




0.64 


2.06 


1.01 


3.13 


35.22 


Syngnathus scovelli 


1.13 




0.71 


1.82 


1.35 


2.77 


37.99 


Lagodon rhomboides 


1.83 




4.57 


1.75 


1.08 


2.65 


40.64 


Chilomycterus schoepfi 


0.46 




0.52 


1.72 


1.13 


2.61 


43.25 


Cynoscion nebulosus 


1.54 




1.95 


1.66 


1.31 


2.53 


45.78 


Monacanthus ciliatus 


0.45 




0.45 


1.62 


0.83 


2.47 


48.25 


Orthopristis chrysoptera 


0.94 




1.33 


1.62 


1.04 


2.46 


50.71 



Tuckey and Dehaven: Fish assemblages found in tidal creek and seagrass habitats in the Suwannee River estuary 



113 









Table 4 








Results of ANOVA comparing 
PRIMER. * <0.05, **<0.01, **■ 


standard length 
<0.001. 


of each species 


between habitats 


and seasonal 


assemblages identified through 


Species 


Factor 


df 






F 


P 


Tukey HSD 


Bairdiella chrysoura 


habitat 
season 


1 
2 






0.02 
0.38 


0.8750 
0.6819 




Cynoscion nehulosus 


habitat 


1 






2.74 


0.1008 






season 


2 






8.03 


*** 


winter > summer 


Lagodon rliomboides 


habitat 
season 


1 
2 






6.61 
80.12 


* 
*** 


tidal-creek > seagrass 
summer > fall > winter 


Leiostomus xanthurus 


habitat 


1 






6.84 


* 






season 


2 






29.87 


*** 


summer > fall, winter 


Mugil cephalus 


habitat 
season 


1 
2 






1.23 

2.54 


0.2713 
0.0862 




Orthopristis chrysoptera 


habitat 


1 






2.96 


0.0996 






season 


2 






4.6 


* 


summer > winter 


Sciaenops ocellatus 


habitat 


1 






1.98 


0.1633 






season 


2 






6.28 


** 


summer, winter > fall 



tidal creeks and dominated samples collected during 
January and February (Tables 2 and 5). Recruitment 
of YOYL. rhomboides and C. arenarius also contributed 
to the winter-spring species assemblage (Tables 4 and 
5). The summer assemblage was influenced by recruit- 
ment of YOY F. grandis and C. nebulosus, which had 
significantly shorter standard lengths than they had 
in the winter-spring assemblage. The recruitment of 
S. ocellatus helped to characterize the fall assemblage. 
Emigration of larger individuals could also account for 
a decrease in mean length; however length-frequency 
plots showed that larger individuals remained vulner- 
able to the gear and that the reduction in mean length 
was due to recruitment of YOY fishes. 



Discussion 

Fishes collected in seagrass habitats in this study were 
similar to those found in other studies of seagrass hab- 
itats; resident species were present year-round and 
there were seasonal pulses of juveniles that used the 
seagrass habitats as a nursery (Reid, 1954; Livingston, 
1982; Weinstein and Brooks, 1983). The assemblages we 
identified were the result of the staggered influx of YOY 
fishes of different species to seagrass habitats through- 
out the year. For example, YOY L. rhomboides and O. 
chrysoptera recruited during winter and spring, whereas 
other abundant species such as YOY B. chrysoura and 
Eucinostomus spp. entered the nursery during summer 
and fall. We found an increase in species abundance 
and species richness during summer and fall similar 
to that found by Reid (1954), who conducted his study 
near Cedar Key. The same pattern was evident in other 
estuarine systems (Cowan and Birdsong, 1985; Rooker 



et al., 1998), demonstrating that recruitment of many 
juvenile fish species to seagrass habitats during summer 
and fall allows the juveniles to use the protection pro- 
vided by the growing seagrasses (Stoner, 1983) and to 
use the food resources found within them (Carr and 
Adams, 1973). 

Early-life-history stages of species with commercial 
or recreational importance were found in each habi- 
tat, but seagrass habitats contained a greater variety 
of juveniles from offshore reef species than did tidal 
creeks. Along the southeastern United States, juveniles 
of many economically important species use a variety 
of habitats in estuaries as nurseries, including man- 
groves, oyster reefs, marshes, tidal creeks, and seagrass 
habitats (Coleman et al., 1999, Coleman, et al., 2000). 
In our study. YOY reef fish taxa, such as serranids, 
lutjanids, and haemulids, were more abundant in sea- 
grass habitats than they were in tidal-creek habitats, 
except for gray snapper [Lutjanus griseus). Juveniles of 
several reef species (C. striata, Mycteroperca microlepis, 
Serraniculus pumilio, Serranus subligaris, and Lachno- 
laimus maximus) were found only in seagrass habitats. 
However, a complicating factor in our study was the 
elimination of oyster habitats from our sampling design. 
Oyster reefs are known to harbor juvenile C. striata 
and M. microlepis (Coleman et al., 2000) and they may 
have been under-estimated in our study because we 
did not sample these habitats. Other economically im- 
portant species, such as C. nebulosus, also recruited to 
the seagrass habitats and are known to reside in them 
much of their life (Reid, 1954; McMichael and Peters, 
1989; Mason and Zengel, 1996). These economically im- 
portant species use seagrass habitats in the Suwannee 
River estuary as a nursery and eventually enter local 
fisheries. Consequently, the maintenance of healthy 



114 



Fishery Bulletin 104(1) 







Table 5 












Results of SIMPER procedure showing the average percent dissimilarity {d9r) for important species between tidal-creek assem- 
blages, di is the average contribution of the ;■'' species to the disssimilarity between groups; di/SDidi) is the ratio between the 
average contribution of the ("' species and the standard deviation of di; Cum di9c is the cumulative contribution to the total dis- 
similarity. Species are listed in decreasing contribution to average dissimilarity. 


Species 


Average abundance 


6i 


di/SD(di) 


di% 


Cumulative bi'/c 


Leiostomus xanthurus 


winter-spi-ing 


summer 


4.17 


2.07 


7.03 




7.03 


177.1 


8.43 


Anchoa mitchilli 


4.93 


243.61 


3.83 


2.4 


6.46 




13.49 


Brevoortia spp. 


10.87 


0.74 


2.65 


1.42 


4.46 




17.95 


Eucinostomus spp. 


0.24 


22.28 


2.62 


1.74 


4.42 




22.37 


Cynoscion arenarius 


1.62 


3.02 


1.94 


1.77 


3.26 




25.63 


Mugil cephalus 


4.52 


0.88 


1.91 


1.57 


3.21 




28.84 


Membra^ martinica 


0.04 


21.61 


1.9 


1.10 


3.19 




32.03 


Lagodon rhombnides 


6.46 


2.92 


1.83 


1.46 


3.08 




35.11 


Leiostomus xanthurus 


winter-Spring 


fall 


4.33 


2.32 


7.85 




7.85 


177.1 


0.96 


Eucinostomus spp. 


0.24 


24.63 


4.03 


2.63 


7.30 




15.15 


Brevoortia spp. 


10.87 


0.03 


2.58 


1.41 


4.68 




19.83 


Eucinsotomus herengulus 


0.07 


2.86 


2.51 


2.73 


4.56 




24.39 


Fundulus grandis 


1.26 


12.02 


2.19 


1.72 


3.97 




28.36 


Fundulus majalis 


0.77 


3.47 


1.66 


1.40 


3.02 




31.38 


Adinia xenica 


0.23 


1.82 


1.62 


1.55 


2.94 




34.32 


Cynoscion nebulosus 


0.07 


1.16 


1.60 


1.68 


2.90 




37.22 


Fundulus g/-andis 


summer 


fall 


2.59 


2.34 


4.84 




4.84 


1.16 


12.02 


Anchoa mitchilli 


243.61 


14.92 


2.52 


1.69 


4.70 




9.54 


Sciaenops oceUatus 


0.53 


5.36 


2.09 


1.39 


3.91 




13.45 


Eucinostomus spp. 


22.28 


24.63 


1.89 


1.27 


3.53 




16.98 


Fundulus majalis 


0.26 


3.47 


1.88 


1.45 


3.51 




20.49 


Bairdiella chrysoura 


4.06 


1.53 


1.81 


1.71 


3.38 




23.87 


Adinia xenica 


0.97 


1.82 


1.76 


1.9 


3.28 




27.15 


Cynoscion arenarius 


3.02 


0.87 


1.7 


1.78 


3.18 




30.33 



seagrasses in this area is important to the preservation 
of these resources. 

Tidal-creek habitats in the Suwannee River estuary 
provided resources for many species that had restricted 
distributions related to salinity tolerances and includ- 
ed taxa that were also found in the nearby seagrass 
habitats. We found recreationally important freshwater 
taxa, such as Micropterus salmoides and Lepomis punc- 
tatus, in tidal-creek habitats. Other groups restricted to 
tidal-creeks were those tolerant of low salinity and in- 
cluded the fundulids, poeciliids, and cyprinodontids. In 
addition, some economically valuable species were more 
abundant in tidal-creeks than in seagrass habitats, in- 



cluding M. cephalus. C. arenarius, S. ocellatus, and L. 
griseus. Tidal creeks also supported a greater density 
of fishes than did seagrass habitats — a density that 
could have resulted from habitat preferences, differen- 
tial mortality between habitat types, or gear avoidance. 
Because the seine was set along the shoreline in tidal 
creeks, fishes were trapped between the seine and the 
shoreline, perhaps making them more vulnerable to the 
gear, whereas in seagrass habitats, the seine was pulled 
along the bottom with the end open prior to retrieval 
and fishes could have used the opening to escape. 

The results of our study showed that there was a more 
consistent assemblage of fishes in tidal creeks, whereas 



Tuckey and Dehaven Fish assemblages found in tidal-creek and seagrass habitats in the Suwannee River estuary 



115 



fish assemblages found in seagrass habitats had greater 
variability in the species present and in the abundance 
of those species. The more consistent assemblage of 
fishes found in tidal creeks could be explained by the 
persistence of vegetation throughout the year in tidal 
creeks, which may have contributed to reduced preda- 
tion and may have provided direct or indirect sources 
of food. Vegetation coverage in seagrass habitats was 
seasonal and Strawn (1961) found that above-ground 
seagrass biomass declined during winter and increased 
during summer and fall. The increased complexity re- 
sulting from blade density and seagrass species hetero- 
geneity offered by the growing seagrasses is known to 
affect fish abundance and composition (Stoner, 1983). 
The fish community structure in our study reflected 
this seasonal change; fewer fish species were present 
during winter and spring than in summer and fall. As 
seagrass biomass increased, fish species composition 
and total numbers also increased, resulting in greater 
variability within seagrass fish assemblages. 

We found that the combination of water tempera- 
ture, salinity, and water depth, more than any other 
combination of abiotic variables, helped to explain the 
fish community structure found in the Suwannee River 
estuary. Although water temperatures between the two 
habitats were similar, tidal creeks typically had soft 
mud sediments instead of sand and mud, marsh-grass 
species instead of submerged aquatic vegetation, deeper 
average depths, and lower salinity values. Water tem- 
perature has been shown to correlate with timing of 
recruitment for YOY fishes, which is ultimately re- 
lated to adult spawning patterns (Subrahmanyam and 
Coultas, 1980; Nelson, 1998; Paperno, 2002). Because 
water temperatures were similar in each habitat, dif- 
ferences in fish-community structures were more likely 
related to salinity tolerances, factors that correlate 
with salinity and water depth. Water depth in the Su- 
wannee River estuary varies seasonally; lowest water 
levels occur during winter (Strawn, 1961). The result 
is a confounding effect of water temperature and water 
depth that probably act in concert to limit distribution 
of fishes. A strong indicator that salinity may be the 
major abiotic factor that determines fish distributions, 
and ultimately species assemblages, was the low-salin- 
ity event during March 1998 that changed the seagrass 
fish assemblage to one more closely resembling a tidal- 
creek assemblage. If vegetation type were the primary 
factor controlling species assemblages in these habitats, 
tidal-creek species would remain in tidal-creeks and not 
invade seagrass habitats when salinity values changed 
to more favorable conditions. Therefore, varying salini- 
ties allowed different groups of fishes to use habitats 
according to their salinity tolerance (Wagner, 1999). 

Nordlie (2003) examined 20 studies of estuarine salt 
marsh fish communities in eastern North America and 
characterized communities based on the life history 
patterns exhibited by the species. General life history 
categories were originally established by McHugh (1967) 
and included permanent residents, marine nursery, ma- 
rine transients, diadromous, and freshwater transients. 



The 45 species that had overlapping distributions among 
habitats in our study were consistent with the classifi- 
cations for marine nursery or marine transient species. 
Marine transient species do not require estuarine habi- 
tats for development, but venture into estuaries during 
periods of low rainfall, whereas marine nursery species 
require estuarine conditions for development. The two 
exceptions in our study (Gobionellus bolesoma and M. 
gulosus) were considered primary residents of saltmarsh 
communities, but were frequently found in estuaries. 

We collected 80 fish species in tidal creeks in the 
Suwannee River estuary — more species than have been 
found in most other studies of tidal creeks — and this 
number could be related to the long-term duration of 
sampling. For example, Peterson and Turner (1994) 
observed 29 fish species inhabiting Louisiana marshes 
in a one-year study, whereas we found 51 additional 
species in our tidal-creek habitats. Similarly, Hettler 
(1989) found 35 species in a one-year study of saltmarsh 
fishes in North Carolina, and Weinstein (1979) recorded 
61 species from his one-year study of the Cape Fear 
River, North Carolina. Furthermore, Cain and Dean 
(1976) found 51 species in a one-year examination of 
fishes in an intertidal creek in South Carolina. The 
first year of our study resulted in the collection of 61 
species from tidal-creek habitats. It is likely that three 
years of sampling in our study increased our chances 
of collecting rare species, which resulted in the higher 
level of species richness. 

Another reason for the high species diversity and 
abundance of fishes that we found in tidal creeks could 
be attributed to our sampling along the tidal-creek edge, 
which is known for its structural complexity (Montague 
and Wiegert, 1990) and importance as a foraging and 
refuge area (Baltz et al., 1993; Kneib and Wagner, 1994; 
Peterson and Turner, 1994). For example, Baltz et al. 
(1993) collected fishes in Louisiana marsh edges to look 
at the importance of the marsh-edge microhabitat and 
found that the 15 most abundant fishes were concentrat- 
ed near the marsh edge and consisted mostly of early- 
life-history stages. They hypothesized that the fishes 
aggregated near the marsh edge to take advantage of 
the protection provided by the vegetation and the avail- 
able food resources. Our sampling targeted the tidal- 
creek edge, and the gear we used selected for juveniles 
and small-adult species, which could explain the higher 
diversity than that seen in other studies. Another pos- 
sibility is that our randomly chosen sampling sites cov- 
ered a greater variety of microhabitats along tidal-creek 
shorelines than did the sampling of Weinstein (1979), 
Hettler (1989), and Peterson and Turner (1994), which 
could also explain the higher species richness. Despite 
differences in sampling methods, the collection of 80 fish 
species in tidal creeks appears to be unusual. 

The withdrawal of fresh water from the Suwannee 
River would likely change the salinity regime in the 
Suwannee River estuary, which may in turn reduce spe- 
cies diversity in the region by reducing habitat availabil- 
ity to groups tolerant of low salinity. Furthermore, the 
high abundance of juvenile fishes that use low-salinity 



116 



Fishery Bulletin 104(1) 



tidal creeks as a nursery would be altered and the re- 
sponses could vary on a species-specific basis (Tsou and 
Matheson, 2002). A decline in the amount of freshwater 
inflow into the tidal-creeks could lead to an overall shift 
towards a more saline environment and result in the 
expansion of seagrass habitats. However, Strawn (1961) 
showed that the distribution of seagrasses at Cedar Key 
was affected by water depth, water clarity, and the inter- 
action of temperature and tides during winter months, 
making the prospect of seagrass expansion unlikely. 
Although a decrease in the amount of fresh water may 
result in an increase in water clarity through a reduction 
in dissolved nutrient input and reduced primary produc- 
tivity, as has been seen in Apalachicola Bay, Florida (Liv- 
ingston, 2003), the extreme low tides, cold temperatures, 
wave action, and sediment geochemistry in the Suwannee 
River estuary may negate the effects of increased light 
penetration (Koch, 2001). Therefore, a decrease in fresh 
water may result in an increase in high-salinity bare 
substrate that has been shown to be less suitable as a 
fish nursery than either seagrass or tidal-creek habitats 
(Sogard and Able, 1991; Rozas and Minello, 1998). 

Tidal-creek and seagrass habitats in the Suwannee 
River estuary contained diverse fish communities that 
reflected seasonal changes associated with recruitment 
of YOY fishes. Many of these species are the targets of 
commercial and recreational activities, which support 
local economies. Although much of the land surround- 
ing the Suwannee River estuary has been preserved, 
measures must be taken to ensure that the supply of 
fresh water from the Suwannee River is also preserved 
to maintain the integrity of the aquatic environment 
and the associated estuarine fish community. 



Acknowledgments 

We appreciate the effort of our coworkers at the Florida 
Marine Research Institute's Cedar Key Field Laboratory 
for assisting with the collection of data for this study and 
to Fred Vose and Cynthia Cooksey for the initial concept 
of this study. This paper benefitted from reviews by D. 
Nemeth, D. Adams, B. Winner, F. Vose, J. Whittington, 
K. Tisdel, R. McMichael, J. Leiby, J. Quinn. T. Tsou, and 
two anonymous reviewers. This project was supported 
in part by funding from the Department of the Interior, 
U. S. Fish and Wildlife Service, Federal Aid for Sport 
Fish Restoration Project number F-43, and Florida rec- 
reational fishing license revenues. 



Literature cited 

Baltz, D. M., C. Rakocinski, and J. W. Fleeger. 

1993. Microhabitat use by marsh-edge fishes in a Loui- 
siana estuary. Environ. Biol. Fish. 36:109-126. 
Bozeman, E. L. Jr., and J. M. Dean. 

1980. The abundance of estuarine larval and juvenile 
fish in a South Carolina intertidal creek. Estuaries 
3:89-97. 



Cain, R. L., and J. M. Dean. 

1976. Annual occurrence, abundance and diversity of 
fish in a South Carolina intertidal creek. Mar. Biol. 
36:369-379. 
Carr, W. E. S., and C. A. Adams. 

1973. Food habits of juvenile marine fishes occupying 
seagrass beds in the estuarine zone near Crystal River, 
Florida. Trans. Am. Fish. Soc. 102:511-540. 
Clarke, K. R. 

1993. Non-parametric multivariate analyses of changes 
in community structure. Austr. J. Ecol. 18:117-143. 

Clarke, K. R., and R. M. Warwick. 

1994. Change in marine communities: an approach to 
statistic analysis and interpretation, 144 p. National 
Environmental Research Council, UK. 

Coleman, F. C, C. C. Koenig, A. Eklund, and C. B. Grimes. 

1999. Management and conservation of temperate reef 
fishes in the grouper-snapper complex of the southeastern 
United States. Am. Fish. Soc. Symp. 23:233-242. 

Coleman, F. C, C. C. Koenig, G. R. Huntsman, J. A. Musick, 
A. M. Eklund, J. C. McGovern, R. W. Chapman, G. R. Sedberry, 
and C. B. Grimes. 

2000. Long-lived reef fishes: the grouper-snapper com- 
plex. Fisheries 25(3):14-21. 

Comp, G. S., and W. Seaman Jr. 

1985. Estuarine habitat and fishery resources of 
Florida. /;; Florida aquatic habitat and fishery resources, 
IW. Seaman jr., ed.), p. 337-436. Florida Chapter of 
the American Fisheries Society, Eustis, FL. 
Cowan, J. H. Jr., and R. S. Birdsong. 

1985. Seasonal occurrence of larval and juvenile fishes 
in a Virginia Atlantic coast estuary with emphasis on 
drums (Family Sciaenidae). Estuaries 8:48-59. 
Field, J. G., K. R. Clarke, and R. M. Warwick, 

1982. A practical strategy for analyzing multispecies 
distribution patterns. Mar. Ecol. Prog. Ser. 8:37-52. 
Gilmore, R. G. Jr. 

1988. Subtropical seagrass fish communities: population 
dynamics, species guilds and microhabitat associations 
in the Indian River Lagoon, Florida. Ph.D. diss., 199 p. 
Florida Inst. Technology, Melbourne, FL. 

Hettler, W. F, Jr. 

1989. Nekton use of regularly-flooded saltmarsh cord- 
grass habitat in North Carolina, USA. Mar. Ecol. Prog. 
Ser. 56:111-118. 

Kilby. J. D. 

1955. The fishes of two gulf coastal marsh areas of 
Florida. Tulane Stud. Zool. 2:175-247. 
Koch, E. W. 

2001. Beyond light: Physical, geological, and geochemical 
parameters as possible submersed aquatic vegetation 
habitat requirements. Estuaries 24:1-17. 

Kneib, R. T, and S. L, Wagner. 

1994. Nekton use of vegetated marsh habitats at differ- 
ent stages of tidal inundation. Mar. Ecol. Prog. Ser. 
106:227-238. 
Livingston, R. J. 

1982. Trophic organization of fishes in a coastal sea- 
grass system. Mar. Ecol. Prog. Ser. 7:1-12. 

1984. The relationship of physical factors and biologi- 
cal response in coastal seagrass meadows. Estuaries 
7:377-390. 



Tuckey and Dehaven: Fish assemblages found in tidal creek and seagrass habitats in the Suwannee River estuary 



117 



2003. Trophic organization in coastal systems, 388 p. 
CRC Press, Boca Raton, FL. 
Mason, W. T., Jr., and S. A. Zengel. 

1996. Foods of juvenile spotted seatrout in seagrasses at 
Seahorse Key, Florida. Gulf Mex. Sci. 2:89-104. 
McGovern. J. C, and C. A. Wenner. 

1990. Seasonal recruitment of larval and juvenile fishes 
into impounded and non-impounded marshes. Wetlands 
10:203-221. 
McHugh, J. L. 

1967. Estuarine nekton. In Estuaries (Lauff, G. H., ed.), 
p. 581-620. AAAS, Publ 83, Washington, D.C. 
McMichael. R. H.. Jr., and K. M. Peters. 

1989. Early life history of spotted seatrout, Cynoficion neb- 
ulosus (Pisces: Sciaenidae), in Tampa Bay, Florida. Es- 
tuaries 12:98-110. 

Montague, C. L., and R. G. Wiegert. 

1990. Salt marshes. In Ecosystems of Florida (R. L. 
Myers and J. E. Ewel, eds.), p. 481-516. Univ. Central 
Florida Press, Orlando, FL. 

Nelson, G. A. 

1998. Abundance, growth, and mortality of young-of- 
the-year pinfish, Lagodon rhomboides, in three estu- 
aries along the gulf coast of Florida. Fish. Bull. 
96:315-328. 

Nordlie, F. G. 

2000. Patterns of reproduction and development of 
selected resident teleosts of Florida salt marshes. Hy- 
drobiologia 434:165-82. 

2003. Fish communities of estuarine salt marshes of 
eastern North American, and comparisons with temper- 
ate estuaries of other continents. Reviews Fish. Biol. 
Fisheries 13:281-325. 

Paperno, R. 

2002. Age-0 spot {Leiostomus xanthuriis) from two 
estuaries along central Florida's east coast: compari- 
sons of the timing of recruitment, seasonal changes 
in abundance, and rates of growth and mortality. Fl. 
Sci. 65:85-99. 

Paperno, R., K. J. Mille, and E. Kadison. 

2001. Patterns in species composition offish and selected 
invertebrate assemblages in estuarine subregions near 
Ponce De Leon Inlet, Florida. Estuar. Coast. Shelf 
Sci. 52:117-130. 

Peterson, G. W., and R. E. Turner. 

1994. The value of salt marsh edge vs. interior as a 
habitat for fish and decapod crustaceans in a Louisiana 
tidal marsh. Estuaries 17:235-262. 



Reid, G. K. Jr. 

1954. An ecological study of the Gulf of Mexico fishes, 
in the vicinity of Cedar Key, Florida. Bull. Mar. Sci. 
Gulf Car. 4:3-94. 
Rooker, J. R., S. A. Holt, M. A. Soto, and G. J. Holt. 

1998. Postsettlement patterns of habitat use by sciae- 
nid fishes in subtropical seagrass meadows. Estuaries 
21:318-327. 
Rozas L. P., and T. J. Minello. 

1998. Nekton use of salt marsh, seagrass, and nonveg- 
etated habitats in a South Texas (USA) estuary. Bull. 
Mar. Sci. 63:481-501. 

Shenker, J. M., and J. M. Dean. 

1979. The utilization of an intertidal salt marsh creek 
by larval and juvenile fishes: abundance diversity and 
temporal variation. Estuaries 2:154-163. 

Sogard, S. M., and K. W. Able. 

1991. A comparison of eelgrass, sea lettuce macroalgae, 
and marsh creeks as habitats for epibenthic fishes and 
decapods. Estuar. Coast. Shelf Sci. 33:501-519. 
Stoner, A. W. 

1983. Distribution of fishes in seagrass meadows: role 
of macrophyte biomass and species composition. Fish. 
Bull. 81:837-846. 
Strawn, K. 

1961. Factors influencing the zonation of submerged 
monocotyledons at Cedar Key, Florida. J. Wild. Manag. 
25:178-189. 
Subrahmanyam, C. B., and C. L. Coultas. 

1980. Studies on the animal communities in two north 
Florida salt marshes Part III. Seasonal fluctuations 
of fish and macroinvertebrates. Bull. Mar. Sci. 
30:790-818. 

Tsou, T., and R. E. Matheson Jr. 

2002. Seasonal changes in the nekton community of the 
Suwannee River estuary and the potential impacts of 
freshwater withdrawal. Estuaries 25:1372-1381. 

Wagner, C. M. 

1999. Expression of the estuarine species minimum in 
littoral fish assemblages of the lower Chesapeake Bay 
tributaries. Estuaries 22:304-312. 

Weinstein, M. P. 

1979. Shallow marsh habitats as primary nurseries 
for fishes and shellfish. Cape Fear River, North Caro- 
lina. Fish. Bull. 77:339-357. 
Weinstein, M. P., and H. A. Brooks. 

1983. Comparative ecology of nekton residing in a tidal 
creek and adjacent seagrass meadow: community compo- 
sition and structure. Mar. Ecol. Prog. Ser. 12:15-27. 



118 



Abstract — Examination of 203 adult 
bluefish iPomatomus saltatrix) from 
Long Island, New York, in 2002 and 
2003 and 66 from the Outer Banks, 
North Carolina, in 2003 revealed 
the presence of dracunculoid nem- 
atodes (PhiloinetJ-a saltatrix) in 
the ovaries of female fish. Percent 
prevalence reached 88% in July and 
then decreased after the peak of the 
spawning season. Bluefish contained 
up to 100 parasites per fish. Infec- 
tion was associated with a range of 
disorders, including hemorrhage, 
inflammation, edema, prenecrotic 
and necrotic changes, and follicular 
atresia, that may prevent proper de- 
velopment of oocytes and probably 
affect bluefish fecundity. Historical 
occurrences, life cycle, and geographi- 
cal distribution of this nematode 
remain largely unknown, but may 
play important roles in recruitment 
processes of bluefish. 



Prevalence, intensity, and effect of 
a nematode iPhilometra saltatrix) 
in the ovaries of bluefish iPomatomus saltatrix) 

Lora M. Clarke 

Marine Sciences Research Center 
Stony Brook University 
Stony Brook, New York 11794-5000 
E-mail address Lora ClarkefQ*msrc sunysb edu 

Alistair D. M. Dove 

Department of Microbiology and Immunology 

Cornell College of Veterinary Medicine 

c/o Marine Sciences Research Center 

Stony Brook University, Stony Brook, New York 11794-5000 

David O. Conover 

Marine Sciences Research Center 

Stony Brook University 

Stony Brook, New York 11794-5000 



Manuscript submitted 14 July 2004 
to the Scientific Editor's Office. 

Manuscript approved for publication 
25 July 2005 by the Scientific Editor. 

Fish. Bull. 104:118-124 (2006). 



Factors influencing recruitment vari- 
ability in marine fishes are often com- 
plex and poorly understood. Slight 
variations in mortality rates, growth 
rates, and stage durations in the 
early life stages can result in tenfold 
or greater fluctuations in abundance 
(Houde, 1987). Recruitment variation 
appears to be driven by a combina- 
tion of factors, such as environmental 
and oceanographic processes (Munch 
and Conover, 2000), diet (Friedland 
at al., 1988; Marks and Conover, 
1993; Juanes and Conover, 1995), 
growth and development (McBride 
and Conover, 1991; Hare and Cowen, 
1997) and habitat use (Able et al., 
2003). The importance of parasitism 
and disease has, however, seldom been 
considered. 

In the Northwest Atlantic, the 
bluefish [Pomatomus saltatrix) is dis- 
tributed from Florida to the Gulf of 
Maine and is both commercially and 
recreationally important. This highly 
migratory species has at least two 
distinct spawning seasons. The first 
occurs in the spring, from March to 
May, south of Cape Hatteras, North 
Carolina (NO (Kendall and Walford, 
1979; Collins and Stender, 1987) and 



the second occurs off the coast of New 
York (NY) from late June to August 
(Norcross et al., 1974; Sherman et 
al., 1984), peaking in July (Chiarella 
and Conover, 1990). Ichthyoplankton 
surveys have indicated that a third 
spawning event occurs south of Cape 
Hatteras, NC, in the autumn, but 
juveniles spawned during this time 
frame have rarely been captured (Col- 
lins and Stender, 1987). 

During the collection of bluefish 
ovaries for another study, the nema- 
tode Philometra saltatrix Ramach- 
andran, 1973 was detected in the 
ovaries of adult bluefish. Previous 
studies of Philometra spp. in other 
host species have indicated that their 
presence can have a negative effect 
on fecundity (Oliva et al., 1992; Hesp 
et al., 2002), implying that a more 
complete understanding of parasites 
may be important to understanding 
reproductive success. Although fac- 
tors such as female size and condi- 
tion are often considered in deter- 
mining reproductive success, the role 
of parasitism is rarely investigated 
(Marshall et al., 1998; Marteinsdottir 
and Begg, 2002). The potential effect 
of this nematode on the reproductive 



Clarke et al Prevalence of the nematode Philometra saltatrix in the ovaries of Pomatomus saltan i 



119 



Long Island, NY 



41N 




'05 -^t, iO -"6 ts' -"5 50 "5" 25" 



Outer Banks, NC 




'^^'T^^/ Cape 



Hatteras 



-Tros -Tew -Te' 15' -75' sc -sas' 



Figure 1 

Map of the sampling area, Long Island, New York, and Outer Banks, North Carolina, where adult 
bluefish iPomatomus xaltatrix) were collected from commercial gill-netters, trawlers, and seafood 
markets in 2002-03 for examination of infestation bv the nematode Philometra saltatrix. 



potential and early life history success of bluefish is 
unknown. No information other than location of oc- 
currence (Ramachandran, 1973) and a brief abstract 
describing the presence of philometrids in the heart 
of juvenile bluefish is available (Cheung et al.M. The 
purpose of this study is to investigate the prevalence, 
intensity, and effect of Philometra saltatrix in the ova- 
ries of bluefish. 



Materials and methods 

Adult bluefish were collected from commercial gill-net- 
ters, trawlers, and seafood markets on Long Island, 
NY, and the Outer Banks, NC (Fig. 1). In NY, fish were 
caught off the southern coast of Long Island from Shin- 
necock Inlet to Montauk Point, approximately 1-15 km 
offshore. In NC, fish were caught approximately 1-40 
km off the coast, and the majority of fish were caught 
30-40 km east of Oregon Inlet. 



Cheung, P. J., R. F. Nigrelli, and G. D. Ruggieri. 1984. Philo- 
metra saltatrix infecting the heart of the 0-class bluefish. 
Pomatomus saltatrix (L.), from the New York coast. In S. F. 
Snieszko commemoration fish disease workshop, p. 27. Joint 
Workshop of Fish Health Section, AFS, and Midwest Disease 
Group, Little Rock, AR. 



Sampling dates were determined by the availability 
of fish through the local fishermen. In 2002, bluefish 
were sampled from mid-July through early October 
off the southern coast of Long Island, NY (80 females, 
108 males). In 2003, fish were collected in NC in April 
(43 females, 21 males) and in NY from the end of June 
through September (123 females, 42 males). 

Fork length (FL), fish weight, gonad weight, preva- 
lence of both live and dead worms, total worm weight, 
and gonadosomatic indices (GSI) were recorded for each 
fish. Worms were often intertwined making it difficult 
to count the number of worms in each ovary; therefore, 
total worm weight per ovary was used as a proxy for in- 
tensity. Representative samples were fixed in a solution 
of 95% glacial acid and 5% formalin for identification. 
Initially, examinations of both male and female fish 
were conducted, but after preliminary evidence showed 
that nematodes were not present in the gonads of male 
fish, future examinations were restricted to female fish. 

Haphazardly selected ovaries were preserved in 10% 
formalin and processed according to standard histo- 
logical methods (Luna, 1968) to investigate pathologies 
associated with the parasite. Transverse sections were 
cut from the same region in the center of each ovary. 
These were examined under a light microscope and 
images were captured with a Spot Insight digital CCD 
and processed with ImagePro Plus software (Media 
Cybernetics, Silver Spring, MD). 



120 



Fishery Bulletin 104(1) 




Jul 02 Aug 02 Sep 02 Ocl 02 Apr 03 Jun 03 Jul 03 Aug 03 Sep 03 

Month 

Figure 2 

Monthly prevalence of live Philometru t^altatrix in the 
ovaries of bluefish (Pomatomus saltatrix). Sample sizes 
are noted on the top of the bars. 




[ZZ12002 
IB 2003 



T " ' " I — ' — \ 

JullH Jul 24 Jul 30 Aug 5 Aug n Aug 1 7 Aug 23 Aug 29 Sep 4 

Date 

Figure 3 

Daily prevalence of live Philnmetra saltatrix in the 
ovaries of bluefish tPomatoniuti saltatrix) in 2002 and 
2003, 



Results 

Description and location of worm 

Philoinetra saltatrix was identified in the gonads of 
female fish ranging in size from 363 to 815 mm (FL) in 
both NC and NY samples. The majority of worms found 
in the ovaries were gravid females. Gravid female worms 
were visible macroscopically and most often visible even 
before the initiation of ovary dissection. Female worms 
reached a maximum of 150 mm in length and approxi- 
mately 300 iim in width. Dead worms were present in 
all months sampled and were encapsulated by several 
melanised layers of fibrotic tissue. Approximately 20 
juvenile male and female worms were identified in the 
ovary of a female fish in July 2003. These immature 
worms reached a maximum length of 2 mm. It should 
be noted, however, that not all fish were examined 
microscopically; therefore, it is possible that other male 
and juvenile worms were present in other hosts but not 
detected. 

Prevalence and intensity 

Bluefish in the western North Atlantic represent a 
single genetic stock (Graves et al., 1992), allowing us 
to combine data from New York and North Carolina to 
examine seasonal trends. Pooled by month of capture, 
the prevalence of infection varied seasonally during the 
two sampling years. In NC, prevalence of live nematodes 
was 7% in April, the spring spawning season (Fig. 2). 
Prevalence of live worms in NY in June was 4.8% and 
reached a maximum of 79% (2002) and 42% (2003) in 
July during the summer spawning season (Fig. 2). It is 
important to note that sampling in 2002 did not begin 



until mid-July but was conducted for the entire month 
in 2003. If only the second half of July is examined for 
both years, the prevalence is 79% and 83% for 2002 and 
2003, respectively. This peak in prevalence was followed 
by a slow decrease in live worms until late August and 
early September after which time live worms were no 
longer detected. 

Examination of prevalence by sampling day showed 
greater evidence of interannual variation. In the middle 
of July 2002, 100% of bluefish examined were infected 
with live Philometra saltatrix, whereas in 2003 the 
highest percentage of fish sampled that were infected 
was 91% (Fig. 3). Furthermore, in 2002 prevalence 
peaked in the middle of July, whereas in 2003 preva- 
lence was highest at the end of the month. Additionally, 
live worms were detected until the end of August in 
2002 but were detected two weeks later in 2003. 

Intensity of parasite infection, as indicated by total 
live and dead worm weight per fish, reached a maxi- 
mum 0.145 g (mean of 0.081 g ±0.1367) and was high- 
est during the month of July in both years of the study 
(Fig. 4). Patterns of intensity were similar to those of 
prevalence; intensity was highest during the summer 
spawning season and then later decreased. One bluefish 
caught just south of Cape Hatteras, NC, in the begin- 
ning of May 2003 had the highest intensity observed 
with over 100 nematodes present (Fig. 5, A and B). This 
fish was not included in the data analysis because it 
was the only fish examined during that month, but this 
extreme instance of infection sets the upper bound on 
the level of observed intensity. 

Binary logistic regression was used to test whether 
the presence or absence of the nematode was related 
to fork length (P=0.096), fish weight (P=0.292), or GSI 
(P=0.783), but no significant relationships were found. 



Clarke et al Prevalence of the nematode Philometra saltatnx in the ovaries of Pomatomus saltatnx 




Jul 02 Aug 02 Sep 02 Oct 02 Apr 03 Jun 03 Jul 03 Aug 03 Sep 03 

H/lonth 

Figure 4 

Monthly intensity of live and dead Philometra saltatrix 
in the ovaries of bluefish (Pomatomus saltatrix). 



However, a significant positive correlation was found 
between intensity and fish weight (R= 0.169, P=0.000), 
intensity and fork length (i?=0.221, P=0.003), and in- 
tensity and GSI (i? = 0.262, P=0.000). The Bonferroni 
method was applied to account for the multiple com- 
parisons. Adjusted P-values are given. 

Observations in adult males and YOY 

Although not the target of this study, we also detected 
the presence of Philometra saltatrix in the pericardial 
cavity of one adult male bluefish collected in NC and 
three adult males in NY, Fish size ranged from 411 to 
429 mm FL. Worms were not detected in the gonads of 
male fish. 

Additionally, we detected worms in the pericardial 
cavity of three young of the year (YOY) bluefish caught 
in the Long Island Sound. Fish ranged in size from 140 
to 185 FL. 

Histopathology 

Worms found in the ovary were surrounded by oocytes 
at various stages of development. The guts of these 
worms were most often filled with host erythrocytes. 
A diversity of pathological responses was noted, and 
significant variability was found among host individuals 
(Fig. 5). In many cases, infection was associated with 
interstitial hemorrhage in the connective tissues of the 
ovary. Occasionally, hemorrhage into the lumen of the 
ovary was also observed, but it was not clear whether 
this was caused by feeding activities of the worm or from 
tissue damage by other means. In many cases, moderate 
lymphocytic infiltration was observed in ovarian connec- 
tive tissues. Some specimens showed marked edema in 



the perifollicular spaces. Prenecrotic changes, including 
pyknosis and cellular swelling, were observed in connec- 
tive tissue cells, and necrosis of both connective tissue 
and oocytes (atresia) was observed in several instances. 
Granulomatous inflammation and fibrosis occurred in 
association with dead worms and, occasionally, atretic 
follicles. Fibrotic capsules surrounding dead worms were 
polylaminate and melanised and appeared to represent 
female worms that had expelled larvae and had subse- 
quently died. Encapsulated worms that still contained 
larvae were also observed occasionally. 



Discussion 

Bluefish along the east coast of the United States appear 
to be heavily infected with the ovarian nematode Philo- 
metra saltatrix. Although many studies provide descrip- 
tions of various philometrid species, very few provide 
information about the prevalence, intensity, or effect 
of these nematodes. Ramachandran (1973) described 
the presence of several female and male worms in a 
single ovary but provided no information on pathological 
changes associated with infection. 

Intensity and prevalence of Philometra saltatrix cycle 
seasonally and appear to be synchronous with the blue- 
fish spawning cycle. Although live worms were detected 
from April to October, the levels of infection were higher 
in July than in any other month sampled and this tim- 
ing is coincident with the peak of the summer spawning 
season off NY (Chiarella and Conover, 1990). The life 
cycle of Philometra saltatrix is unknown and, therefore, 
it is not clear whether initial infection coincides with 
the spawning season or if nematodes are acquired at an 
earlier time and reside in some other host tissue site 
before migrating to the ovaries during the spawning 
season, perhaps stimulated by hormonal cues from the 
host. Prevalence and intensity were much lower during 
the spring spawning season south of Cape Hatteras, 
NC. It is difficult to determine whether this was due 
to geographical differences or the limited sampling that 
month; it is possible that we missed the peak spawning 
period in 2003. The rapid decline in both prevalence 
and intensity after the summer spawning period indi- 
cates that Philometra saltatrix are released or migrate 
out of the host fish synchronously with the release of 
fish eggs. We speculate that the reproduction cycle of 
the nematode is closely matched to that of the host 
bluefish. 

Other studies have reported that infection occurs af- 
ter first host maturity and that female nematodes have 
only been observed in the gonads of females (Oliva et 
al., 1992; Hesp et al., 2002). Although not the focus of 
this study, the pericardial cavity was also found to be 
infected by Philometra saltatrix in adult males in NC 
and NY and in YOY bluefish in NY waters, in contradic- 
tion to these other findings. It is not clear if these YOY 
infections represent separate infections or if they are 
the same infections observed in adult females. Review- 
ing the reported lifecycle of other spirurid nematodes 



122 



Fishery Bulletin 104(1) 














N 



cl^^l*^^. 



*5> 

100 |jm E 





'- v>^l^li\- 



Figure 5 

Observations associated with Philometra saltatrix infection in bluefish (Pomatomus saltatrix) ovaries. 
(A) Gross, bisected ovary from infected bluefish, showing heavy infection with Philometra (B) Low 
power view of transverse section (approximately 1.5 cm in diameter) of heavily-infected bluefish ovary. 
M=muscular capsule, F=ovarian follicles. W=nematodes in lumen of ovary, displacing oocytes. (C) 
Histological section of healthy ovarian tissue showing dense packing of healthy oocytes in various 
developmental stages, little connective tissue, and no cellular infiltrates. (D) Histological section of 
infected bluefish ovary, showing relatively fewer oocytes (Ol and the presence oi Philometra (P). In 
addition, there is severe interstitial hemorrhage (H) and necrosis (N) of ovarian tissue, with heavy 
cellular infiltration and atresia (A) of ovarian follicles. 



and Philometra spp. (Williams and Jones, 1994), we 
speculate that nematode larvae are released at spawn- 
ing time, pass through the copepod intermediate stage, 
and a second perhaps paratenic intermediate host (or 
perhaps pass through only one intermediate host), and 
infect YOY bluefish. It is important to note that most 
YOY bluefish larger than approximately 40 mm TL 
are piscivorous (Marks and Conover 1993); thus, it is 
possible that the infected YOY bluefish observed in 
this study were infected through a second intermedi- 
ate host. The existence of a second intermediate fish 
host is supported by the observed positive relationship 
between host body size and parasite intensity. After ini- 
tial infection, these worms reside in non-ovarian sites 
such as the pericardial cavity (as we observed), where 
they may remain quiescent until maturation of the host. 
At that time, worms may migrate through the tissues 



to the ovary for spawning and the completion of their 
life-cycle. This proposed life history would explain the 
ontogenetic and seasonal patterns of worm distribution 
that we observed and explain the rapid appearance of 
well-developed worms in the ovary at the onset of the 
spawning season. 

The prevalence and intensity of infections we ob- 
served in bluefish were significantly higher than those 
reported in most other fish species. For example, inten- 
sity of infection in Glaucosoma hebraicum by Philometra 
lateolabracis ranged from 1 to 7 nematodes (mean of 
2 nematodes) (Hesp et al., 2002), whereas prevalence 
of Philometra margolisi in the gonads of red grouper, 
Epinephelus morio, ranged from 14 to 28% depending on 
locality (Moravec et al., 1995) and prevalence of Philo- 
metra lateolabracis in Parupeneus indicus was 5.3% 
(intensity of 1-2 nematodes) (Moravec et al., 1988). 



Clarke et a\. Prevalence of the nematode Philometra saltan ix in the ovaries of Pomatomus saltalnx 



123 



Although most studies indicate the occurrence of these 
nematodes to be lower than in bluefish, in a later study 
Moravec et al. (1997) reported prevalence of Philometra 
margolisi in the gonads of Epinephelus morio to be as 
high as 88% with an intensity of up to 84 nematodes. 
Additionally, although most studies report the absence 
of male worms (Hesp et al., 2002), we found up to 20 
males in the ovary of one female fish. The high per- 
centage of adult female bluefish that are infected may 
indicate that Philometra saltatrix are significantly af- 
fecting the reproductive success of bluefish. 

Despite some studies showing deleterious effects of 
other parasites on fish ovaries (see for example, Adler- 
stein and Dorn, 1998), studies on the effects of philome- 
trids on fish ovaries remain scarce. Indeed, in a recent 
review of histopathological assessments of gonadal tis- 
sue in wild fishes, Blazer (2002) made no mention of 
Philometra spp., despite the fact that it is probably the 
most common adult helminth observed in fish gonads. 
Annigeri (1962) described damage to ovaries of Oto- 
lithiis argenteiis by an unidentified philometrid, and 
Ramachandran (1975) mentioned necrosis in the ovaries 
of Mugil cephalus caused by Philometra cephalus. The 
infection of testes of pink snapper Pagriis auratus by 
Philometra lateolabracis resulted in partial or extensive 
atrophy of those gonads (Hine and Anderson 1981). 
Oliva et al. (1992) alluded briefly to lower fecundity 
in Paralabrax humeralis as a result of infection with 
Philometra but provided no data in support of this as- 
sertion. Hesp et al. (2002) studied the dynamics and 
effects of P. lateolabracis in the gonads of the Austra- 
lian predatory marine fish Glaucosoma hebraicum, but 
did not observe significant pathological abnormalities 
associated with infection. 

Histopathological changes associated with philome- 
trid infection in bluefish were significant. The hem- 
orrhaging, edema, inflammation, atresia, and necro- 
sis observed most likely reduce oocyte number and 
quality, leading to lower fecundity. The presence of 
erythrocytes in the guts of nematodes indicates that 
the parasites were feeding on the blood of the host; 
this diversion of nutrients to the parasite may exacer- 
bate the impacts of the worm on ovarian tissue. Fur- 
thermore, intense infections can reduce the effective 
volume of the ovary and thus lead to lower fecundity 
(Fig. 5, A and B). 

The high prevalence, intensity, and pathological dam- 
age observed could be an important factor in recruit- 
ment variation in bluefish. Although modest interannual 
differences in prevalence and timing of infection were 
observed in this study (Fig. 4), Philometra saltatrix was 
not observed by researchers conducting studies of the 
GSI of bluefish from NY waters in the 1980s (Chiarella 
and Conover, 1990; Conover, personal observ. ). Hence, 
it is possible that the abundance of Philometra may 
fluctuate greatly over long time scales. The reason for 
the current prevalence of the worm and its effect on 
bluefish recruitment is unknown. Future studies of 
this host-parasite system should seek to account for 
temporal variations of this kind. 



Acknowledgments 

We thank R. Knotoff and numerous other commercial 
fishermen from Shinnecock Inlet, NY, for helping us 
to obtain samples. B. Burns and J. Pearce of the NC 
Division of Marine Fisheries helped collect North Caro- 
lina bluefish. We thank R. Clarke, C. Knakal, and S. 
Abrams for laboratory and field assistance. We also 
thank two anonymous reviewers for helpful comments 
on the manuscript. Funding was provided by the Rut- 
gers/NOAA Bluefish-Striped Bass Research Program 
and the New York State Department of Environmental 
Conservation. 



Literature cited 

Able K. W., P. Rowe, M. Burlas, and D. Byrne. 

2003. Use of ocean and estuarine habitats by young-of- 
year bluefish iPomatonuis saltatrix) in the New York 
Bight. Fish. Bull. 101:201-214. 
Adlerstein, S. A., and M. W. Dorn. 

1998. The effect of Kudoa paniformis infection on the 
reproductive effort of female Pacific hake. Can. J. Zool. 
76:2285-2289. 
Annigeri, G. G. 

1962. A viviparous nematode, Philometra sp., in the 
ovaries of Otolithua argenteus (Cuvier). J. Mar. Biol. 
Assoc. (India) 3:263-265. 
Blazer, V. S. 

2002. Histopathological assessment of gonadal tissue in 
wild fishes. Fish Physiol. Biochem. 26: 85-101. 
Chiarella, L., and D. O. Conover. 

1990. Spawning season and first year growth of adult 
bluefish (Pomatomus saltatrix) from the New York 
Bight. Trans. Am. Fish. Soc. 119:455-462. 
Collins, M. R., and B. W. Slender. 

1987. Larval king mackerel (Scomberomorus caralla), 
Spanish mackerel, (S. maculatus) and bluefish (Poma- 
tomus saltatrix) off the southeast coast of the United 
States, 1973-1980. Bull. Mar. Sci. 41:822-834. 

Friedland, K. D., G. C. Garman, A. J. Bejda, A. L. Studholme, 
and B. 011a. 

1988. Interannual variation in diet and condition in 
juvenile bluefish during estuarine residency. Trans. 
Am. Fish. Soc. 117:474-479. 

Graves, J. E., J. R. McDowell, A. M. Beardsley, and D. R. Scoles 
1992. Stock structure of the bluefish Pomatomus sal- 
tatrix along the mid-Atlantic coast. Fish Bull. 90 
(4):703-710. 
Hare, J., and R. K. Cowen. 

1997. Size, growth, development, and survival of the 
planktonic larvae of Pomatomus saltatrix. Ecology 
78:2415-2431. 
Hesp, S. A., R. P. Hobbs, and I. C. Potter. 

2002. Infection of the gonads of Glaucosoma hebraicum 
by the nematode Philometra lateolabracis: occurrence 
and host response. J. Fish Biol. 60:663-673. 
Hine, P. M., and C. D. Anderson. 

1981. Diseases of the gonads and kidneys of New Zealand 
snapper, Chrysophrys auratus Forster (F. Sparidae). In 
Wildlife diseases of the Pacific Basin and other coun- 
tries (M. E. Fowler, ed.), p. 166-170. Academic Press, 
London. 



124 



Fishery Bulletin 104(1) 



Houde, E. D. 

1987. Early life dynamics and recruitment variability. 
Am. Fish. Soc. Symp. 2:17-29. 

Juanes, F., and D. O. Conover. 

1995. Size-structured piscivory; advection and linkage 
between predator and prey recruitment in young-of-the- 
year bluefish. Mar. Ecoi. Prog. Ser. 128:287-304. 
Kendall. A. W. Jr., and L. A. Walford. 

1979. Sources and distribution of bluefish. Poinatoinus 
saltatrix, larvae and juveniles off the east coast of the 
United States. Fish. Bull. 77:213-227. 
Luna, L. G. 

1968. Manual of histological staining of the Armed Forces 
Institute of Pathology, 258 p. McGraw-Hill Book Com- 
pany, New York, NY. 
Marks, R., and D. O. Conover. 

1993. Ontogenetic shift in the diet of young-of-the-year 
bluefish iPomatomus saltatrix) during the oceanic phase 
of the early life history. Fish. Bull. 91:97-106. 
Marshall, C. T., O. S. Kjesbu, N. A. Yaragina, P. Solemdal, 
O. Ulltang. 

1998. Is spawner biomass a sensitive measure of the 
reproductive and recruitment potential of Northeast 
Arctic cod? Can. J. Fish. Aquat. Sci. 55:1766-1783. 
Marteinsdottir, G., and G. A. Begg. 

2002. Essential relationships incorporating the influ- 
ence of age, size and condition on variables required 
for estimation of reproductive potential in Atlantic cod 
Gadus morhua. Mar. Ecol. Prog. Ser. 235:235-256 
McBride, R. S., and D. O. Conover. 

1991. Recruitment of young-of-the-year bluefish {Poina- 
toinus saltatrix) to the New York Bight: variation in 
abundance and growth of spring and summer-spawned 
cohorts. Mar. Ecol. Prog. Ser. 78:205-216. 
Moravec, F., P. Orecchia, and L. Paggi. 

1988. Three interesting nematodes from the fish Paru- 
peneus indices (Mullidae, PerciformesI of the Indian 
Ocean, including a new species, Ascarophis pariipenei 
sp. N. (Habronematoidea). Folia Parasitol. 35:47-57. 

Moravec, F., V. M. Vidal-Martinez, and L. Aquirremacedo. 

1995. Philometra margolisi n.sp. (Nematoda, Philome- 



tridae) from the gonads of the red grouper, Epinephelus 
morio (Pisces, Serranidae), in Mexico. Can. J. Fish. 
Aquat. Sci. 52(suppl. 1) 161-165. 
Moravec, F., V. M. Vidal-Martinez, J. Vargas-Vazquez, C. Vivas- 
Rodriguez, D. Gonzalez-Solis, E. Mendoza-Franco, R. Sima- 
Alvarez, and J. Guemez-Ricalde. 

1997. Helminth parasites o{ Epinephelus morio (Pisces: 
Serranidae) of the Yucatan Peninsula, southeastern 
Mexico. Folia Parasitol. 44:255-266. 
Munch, S. B., and D. O. Conover. 

2000. Recruitment dynamics of bluefish, Pomatomus sal- 
tatrix, on the continental shelf from Cape Fear to Cape 
Cod, 1973-1995. ICES J. Mar. Sci. 57:393-402. 
Norcross, J. J., S. L. Richards, W. H. Massraann, and E. B. 
Joseph. 

1974. Development of young bluefish iPomatomus sal- 
tatrix) and distribution of eggs and young in Virginian 
coastal waters. Trans. Am. Fish. Soc. 103:477-497. 

Oliva, M. E., A. S. Borquez, and A. N. Olivares. 

1992. Sexual status of Paralahrax humeralis (Serra- 
nidae) and infection by Philometra sp. (Nematoda: 
Dracunculoidea). J. Fish Biol. 40:979-980. 

Ramachandran, P. 

1973. Philometra saltatrix sp. n., infecting the gonads 
of the common bluefish Pomatomus saltatrix (L.) off 
the New England coast of the United States. Zool. 
Anz. 191:325-328. 

Ramachandran, P. 

1975. Philometra cephalus sp. n. infecting the gonads of 
the striped mullet, Mugil cephalus L. from the Arabian 
Coast of Kerala, India, with a note on its pathology. Zool. 
Anz. 194:140-144. 

Sherman, K., W. Smith, W. Morse, M. Berman. J. Green, and 
L. Ejsymont. 

1984. Spawning strategies of fishes in relation to cir- 
culation, phytoplankton production, and pulses in zoo- 
plankton off the northeastern United States. Mar. 
Ecol. Prog. Ser. 18:1-19. 
Williams, H., and A. Jones. 

1994. Parasitic worms offish, p. 139-146. Taylor and 
Francis Ltd., Bristol. PA. 



125 



Abstract — Size-related differences 
in power production and swim speed 
duration may contribute to the 
observed deficit of nursing calves in 
relation to lactating females killed 
in sets by tuna purse-seiners in the 
eastern tropical Pacific Ocean (ETP). 
Power production and swim-speed 
duration were estimated for north- 
eastern spotted dolphins iStenella 
attenuata), the species (neonate 
through adult) most often captured 
by the fishery. Power required by 
neonates to swim unassisted was 
3.6 times that required of an adult 
to swim the same speed. Estimated 
unassisted burst speed for neonates 
is only about 3 m/s compared to about 
6 m/s for adults. Estimated long-term 
sustainable speed is about 1 m/s for 
neonates compared to about 2.5 m/s 
for adults. Weight-specific power 
requirements decrease as dolphin 
calves increase in size, but power 
estimates for 2-year-old spotted dol- 
phin calves are still about 40% higher 
than power estimates for adults, to 
maintain the same speed. These esti- 
mated differences between calves and 
adults are conservative because the 
calculations do not include accommo- 
dation for reduced aerobic capacity in 
dolphin calves compared to adults. 
Discrepancies in power production are 
probably ameliorated under normal 
circumstances by calves drafting next 
to their mothers, and by employing 
burst-coast or leap-burst-coast swim- 
ming, but the relatively high speeds 
associated with evasion behaviors 
during and after tuna sets likely 
diminish use of these energy-saving 
strategies by calves. 



Duration of unassisted swimming activity 
for spotted dolphin iStenella attenuata) calves: 
implications for mother-calf separation 
during tuna purse-seine sets 



Elizabeth F. Edwards 

Southwest Fisheries Science Center 

National Marine Fisheries Service 

8604 La Jolla Shores Drive 

La Jolla, California 92037 

E mail address Elizabeth Edwardsigi noaa.gov 



Manuscript submitted 15 February 2005 
to the Scientific Editor's Office. 

Manuscript approved for publication 
25 July 2005 by the Scientific Editor. 

Fish. Bull. 104:125-135 (2006). 



Dolphin calves draft in echelon posi- 
tion next to their mothers for the first 
few weeks after birth and continue 
to return to drafting in echelon posi- 
tion frequently throughout at least 
their first year (EdwardsM. "Echelon 
position' is the physical positioning 
of the calf within a few centimeters 
of the mother, near her mid-section, 
with fin motions reduced or absent 
(e.g., Norris and Prescott, 1961). This 
position takes advantage of the moth- 
er's flow field, reducing or effectively 
eliminating the energy cost to the 
calf of moving forward through the 
water (Weihs, 2004). Because draft- 
ing appears to be ubiquitous among 
dolphin calves, particularly during 
the neonate stage, it appears likely 
that drafting is an essential factor 
in maintaining physical association 
between calves and their mothers, 
especially when the calves are small. 
In the eastern tropical Pacific 
Ocean (ETP) a situation occurs in 
which it becomes important to con- 
sider the consequences for calves of 
losing their drafting association with 
their mother. In this area, a tuna 
purse-seine fishery targets schools of 
large yellowfin tuna that associate 
closely with schools of dolphins, pri- 
marily the spotted dolphin (SteneUa 
attenuata) (NRC, 1992). The associ- 
ated schools of tunas and dolphins 
are located by a helicopter sent out 
from the purse-seine vessel and are 
subsequently captured through the 
actions of several high-powered speed- 
boats, which are released from the 
purse-seiner to overtake and herd the 



associated animals into the closing 
arc of the purse-seine (NRC, 1992). 
Examination of the dolphins found 
dead in the net has revealed that 
75% to 95% of the lactating females 
killed in the sets are not killed with 
an accompanying calf (Archer et al., 
2004). This observed calf deficit is a 
potentially important factor in the 
lack of recovery of ETP dolphin popu- 
lations despite over a decade of very 
low fishery mortality (Wade et al.^), 
but little is known about why or how 
the separation of mothers and calves 
occurs (Archer et al., 2004). 

The chase, encirclement, and re- 
lease procedure of purse-seine fish- 
ing operations tends to be relatively 
prolonged — chases averaging about 
30 minutes, capture and confine- 
ment about 90 minutes, and postre- 
lease swimming at least 90 minutes 
(Myrick and Perkins, 1995; Chivers 



1 Edwards, E. 2002. Behavioral contri- 
butions to separation and subsequent 
mortality of dolphin calves chased by 
tuna purse-seiners in the eastern tropi- 
cal Pacific Ocean. National Oceano- 
graphic and Atmospheric Administra- 
tion Administrative Report LJ-02-28, 
33 p. Southwest Fisheries Science 
Center, 8604 La Jolla Shores Drive, La 
Jolla, CA 92037. 

- Wade, P., S. Reilly, and T. Gerrodette. 
2002. Assessment of the population 
dynamics of the northeastern offshore 
spotted and eastern spinner dolphin 
populations through 2002. National 
Oceanographic and Atmospheric Ad- 
ministration Administrative Report 
LJ-02-13, 58 p. Southwest Fisheries 
Science Center, 8604 La Jolla Shores 
Drive, La Jolla, CA 92037. 



126 



Fishery Bulletin 104(1) 



and Scott^). The set procedure also tends to induce rela- 
tively high, sustained swimming speeds in the dolphins. 
Swimming speeds of two spotted dolphins carrying ve- 
locity tags during and after tuna purse-seine sets in 
the ETP averaged 1.7-2.8 m/s during chase and 2.6-3.1 
m/s after release, compared to undisturbed speeds of 
1.2-1.9 m/s (Chivers and Scott^). Compared to adults, 
dolphin calves have smaller, less-coordinated muscles 
and lower aerobic capacity, especially at birth but per- 
sisting through the first year or more (Dolar et al., 1999; 
Dearoff et al., 2000; Noren et al., 2001; Noren et al., 
2002; Eitner et al, 2003; Noren et al., 2004). Therefore, 
it appears possible that the observed calf deficit in the 
kill may result at least in part from energetics-related 
separation of calves from mothers during the chase or 
after release, as well as the inability of calves to main- 
tain speed with adults while swimming alone. 

The present study examines the possibility of ener- 
getics-related separation, by estimating the length of 
time (duration) during which spotted dolphin calves of 
different sizes ranging from neonate through two years 
of age can swim unassisted at various velocities. These 
velocity-durations are then compared to adult swim- 
ming capacities, in order to determine whether calves 
swimming without assistance may experience energy- 
based difficulty keeping up with adult dolphins during 
evasion from tuna purse-seine sets. 



Materials and methods 

Duration limits for unassisted swimming by spotted 
dolphin calves were calculated by combining estimates 
of mass-specific energy cost of steady, submerged, unas- 
sisted swimming by neonate through adult spotted 
dolphins, with reported swimming speed durations of 
adult dolphins of various species (Table 1). Swimming 
duration for adults was used because this information 
was not available for dolphin calves. The data were com- 
bined based on the assumption that the duration a given 
unit of dolphin muscle can sustain a given mass-spe- 
cific energy cost is the same for both adults and calves 
within any species of dolphin. It is much more likely that 
swimming capacity of calves within any dolphin species 
is significantly lower than that of adults and that dif- 
ferences are particularly pronounced at birth {Dolar et 
al., 1999; Dearoff et al., 2000; Eitner et al., 2003; Noren 
et al., 2001, 2002, 2004), but the exact differences are 
unknown. The swimming duration estimates presented 
in this study therefore represent maximum durations. 
It is also possible that species differ in power produc- 
tion capacities, based on observations of blood chemistry 
differences between species (Ridgway and Johnston, 
1965), but it is not unreasonable to assume that relative 



differences between species are maintained throughout 
the size ranges of individuals within a species, so that 
younger dolphins are probably less adept than adults 
in any particular species. Thus, although it is not yet 
possible to quantify differences between calf and adult 
swimming-duration capacities in Stenella attenuata in 
nature, they are likely greater than estimated in the 
present study, so that problems apparent from this study 
are likely greater during actual tuna purse-seine sets. 
The energy cost of a steady, submerged, unassisted 
swimming rate was estimated in this study for veloci- 
ties ranging from 1.5 to 6.0 m/s in order to encompass 
speeds from slow undisturbed swimming to the maxi- 
mum likely to occur during attempted evasion from 
tuna purse-seine sets. Average velocities observed dur- 
ing tracking of dolphins before and after experimental 
sets varied from about 1.5 m/s m to about 3 m/s (Chiv- 
ers and Scott') but short term speeds attained during 
evasion maneuvers were likely to have exceeded these 
averages (Table 1). Total body energy costs were esti- 
mated first and then converted to mass-specific costs by 
dividing total estimated energy costs by the estimated 
total muscle weight of each size of modeled dolphin. 
Muscle-mass-specific measures are more appropriate 
than simply dividing by total body weight for compari- 
sons between sizes because muscle fraction of body 
weight increases with body weight in Stenella attenuata 
in the ETP, from 35-40% in neonates to 55-60% in 
adults (Edwards, 1993). 

Data sources 

Total body energy costs were estimated for eight mod- 
eled Stenella attenuata ranging in size (age) from new- 
born through adult. Sizes were selected to emphasize 
changes during the early months (Table 2). Size-at-age 
to two years old was estimated from Hohn and Ham- 
mond (1985). Size of an adult reproductive female was 
estimated from Perrin and Reilly (1984). Total body wet 
weight, wetted surface area of body, fins, and flukes, 
maximum body diameter, and fraction of total body 
weight composed of muscle were estimated for each size 
of modeled dolphin by using regression equations devel- 
oped from morphological measurements of ETP dolphins 
(Edwards 1993, Edwards"*). 

The morphological measurements were taken from 
35 spotted dolphins ranging in size from 0.71 to 2.1 m 
total length (tip of rostrum to fluke notch). This size 
range encompasses all life-history stages (near-term 
fetus through mature adult) of the spotted dolphins 
found in the eastern tropical Pacific Ocean. Spotted 
dolphins are about 0.8 m total length at birth (Hohn 
and Hammond, 1985). Specimens included 22 females 
and 13 males (Edwards, 1993). Male specimens included 



■' Chivers, S., and M. Scott. 2002. Tagging and tracking of 
Stenella specie.s during the 2001 Chase and Encirclement 
Stress Studies cruise. National Oceanographic and Atmo- 
spheric Administration Administrative Report LJ-02-33. 
23 p. Southwest Fisheries Science Center, 8604 La JoUa 
Shores Drive, La Jolla, CA. 92037. 



^ Edwards, E. 2002. Energetics consequences of chase by 
tuna purse-seiners for spotted dolphins (Stenella attenuata) 
in the eastern tropical Pacific Ocean. National Oceano- 
graphic and Atmospheric Administration Administrative 
Report LJ-02-29. 32 p. Southwest Fisheries Science Center, 
8604 La Jolla Shores Drive, La Jolla, CA 92037. 



Edwards: Duration of unassisted swimming by Stenella attenualo 



127 



three fetuses, six immature, and four mature individu- 
als. Female specimens included three fetuses, six im- 
mature, one mature resting, one mature lactating, and 
11 mature pregnant individuals. All specimens were 
killed during tuna fishing operations in the eastern 
tropical Pacific Ocean. Two of the specimens were col- 
lected in February 1980, one in July 1983, nine in July 
1985, two in August 1985, seventeen in December 1985, 
and four were collected without a date noted (Edwards, 
1993). All specimens were processed according to the 
same procedures prior to dissection. Immediately after 
the sets, the dolphin specimens were brought on board 
the vessel and frozen whole in the brine wells of the 
vessel. The specimens were transported frozen to port 
and then transported frozen to the Southwest Fisheries 
Science Center. Specimens were kept frozen until thawed 
in fresh water (about 27°C) just prior to dissection. Not 
all measurements were made on all specimens; therefore 
sample sizes differ between the regression equations 
presented below. 

Energetics model The energetics model used to esti- 
mate total body cost of swimming was taken from 
Edwards (1992, based on Magnuson, 1978), except that 
1) new data were used to estimate dolphin body param- 
eters and 2) the estimate of fin plus induced drag was 
replaced by the multiplier 3 (see below). The model 
used standard hydrodynamic equations and methods 
(Hoerner, 1965; Hertel, 1969; Webb, 1975) to estimate 
hydrodynamic drag on a fully submerged streamlined 
body of revolution moving steadily in turbulent flow. 
Body surface area was increased to specifically include 
the surface area of fins and flukes (Fish''), and drag 
estimates were increased to account for body and fin 
movements. Because energy to move forward (thrust 
energy) must exactly balance the drag experienced by a 
steadily swimming animal, estimating total drag energy 
is equivalent to estimating thrust energy, i.e., the energy 
cost to swim (Fish and Rohr, 1999). 



studies by Fish (1998), Webb (1975), and Yates (1983). 
P was estimated as a function of total hydrodynamic 
drag (Di, in dynes) and velocity (V, in m/s) as 

P,„=D,V/lOl 

where the factor 10" converts (D,V) to watts. 

Total drag was estimated as a function of drag due to 
body, fins, and movements of body parts as 

D,=0.5pV'S,.C,, = 1.5pV%,C„ 

where p = density of seawater (1.025 g/cm'^); 
iS,,, = wetted surface area; 
Cj = coefficient of total drag; and 
3 = drag augmentation factor. 

S,^ includes surface area of body plus fins and flukes, 
where estimated planar area of fins was increased by 
69c to account for the curvature of the fins, based on 
measurements of individual slices from fins and flukes 
from one small and one large dolphin, 1.32 m and 1.93 m 
in length, respectively (Edwards, unpubl. data). Drag 
augmentation factor generally varies between 3 and 5 
(e.g., Lighthill, 1971; Fish, 1993) and was assumed equal 
to the value of 3 in the present study, based on studies of 
gliding vs. actively swimming dolphins (Skrovan et al., 
1999). The factor 3 accounts for the increase over gliding 
drag caused by body movements during active swim- 
ming. Use of a squared relationship between V and JD, 
is supported by the observed relationship between total 
drag and velocity in free swimming Tursiops truncatus 
(Skrovan et al. 1999, Eq. 6). Use of a cubic relation- 
ship between V and P, is supported by observations of 
swimming kinematics of Tursiops truncatus swimming 
between 1 and 6 m/s (Fish, 1993). 
Sj,, (in cm^) was estimated as 

S, =0.299L2°5, 



Model formulation Total power (P,, in watts) required 
to overcome drag during steady, submerged swimming 
(Hertel, 1969) by a modeled dolphin of a given total 
length (L, rostrum to fluke notch) was estimated as 

where P^ = mechanical power (in watts) required to 
overcome hydrodynamic drag; 
£„, = muscle efficiency; and 

E = "propeller efficiency" (efficiency of propul- 
sion by flukes). 



based on measurements from 19 Stenella attenuata 
ranging in size from 0.71 to 2.01 m, where L is total 
length (in cm). 

Cj was estimated from the formula for drag of sub- 
merged streamlined bodies of revolution moving at 
constant velocity (Hoerner, 1965; Hertel, 1969; Webb, 
1975) as 

Cj=Cf\\ + [l.b(d I Lf'^) + l[(d I L\^]^, 

Cf = the coefficient of friction drag; and 
d = 0.12 



£„, was assumed to be 0.2 from studies of muscle efficien- 
cies in terrestrial animals (e.g., Goldspink 1988), man 
(Alexander, 1983; quoting Dickinson, 1929) and dolphins 
(Fish, 1993, 1996). E^ was assumed to be 0.85 based on 



5 Fish, F. 2002. Personal comniun. Liquid Life Laboratory, 
West Chester Univ. Pennsvlvania, West Chester, PA. 



where d = maximum body diameter (in cm) based on 
measurements from 24 Stenella attenuata 
ranging in size from 0.71 to 2.01 m. 

C, was estimated from the formula for submerged stream- 
lined bodies of revolution moving at constant velocity in 
turbulent flow (e.g., Webb, 1975) as 



128 



Fishery Bulletin 104(1) 







Table 1 






Reported duration of various 


swimming speeds for several species of dolphins. 




Velocity 








(m/s) 


Duration 


Publication 


Species 




Burst (seconds) 






11.2 


leap 


Rohretal. (2002) 


Tursiops 


11.1 


2 


Lang (1975) 


Stenella 


10.6 


2 


Lang (1975) 


S ten el! a 


10.3 


2 


Lang (1975) 


Stenella 


9.7 


leap 


Rohretal. (2002) 


Tursiops 


8.8 


<2 


Rohret al. (2002) 


Delphinus 


8.8 


leap 


Rohretal. (2002) 


Tursiops 


8.3 


7.5 


Lang (1975) 


Tursiops 


8.2 


<3 


Rohretal. (2002) 


Tursiops 


8.2 


<120 


Auetal. (1988) 


Stenella 


8.0 


7.6 


Lang and Norris 1 1966) 


Tursiops 


8.0 


<3 


Rohretal. (2002) 


Delphinus 


7.8 


2 


Lang (1975) 


Lagenor 


7.7 


2-3 


Lang and Prior (1966) 


Stenella 


7.5 


leap 


Lang (1975) 


Tursiops 


7.4 


leap 


Lang (1975) 


Tursiops 


7.3 


leap 


Au and Weihs (1980) 


Stenella 


6.8 


10 


Lang and Norris ( 1966) 


Tursiops 


6.7 


1.4 


Rohretal. (2002) 


Delphinus 


6.7 


<3 


Rohretal. (2002) 


Delphinus 


6.3 


10 


Lang and Norris ( 1966 ) 


Tursiops 


6.4 


240 


Au and Weihs (19801 


Stenella 


6.2 


<2 


Rohretal. (2002) 


Tursiops 


5.9 


50 


Lang and Norris (1966) 


Tursiops 


6.0 


2 


Rohretal. (2002) 


Delphinus 


5.6 


<10 


Rohretal. (2002) 


Tursiops 


5.3 


6.6 

Maximum (minutes) 


Lang and Norris (1966) 


Tursiops 


4.7 


11 


Au and Perry man (1982) 


Stenella 


4.4 


21 


Au and Perryman (1982) 


Stenella 


4.2 


<8.5 

Prolonged (hours-days) 


Rohret al.( 20021 


Delphinus 


3.5 


1 hour 


Au and Perryman ( 1982 ) 


Stenella 


3.2 


9.3 hour 


Leatherwood and L. (1979) 


Stenella 


3.1 


extended period 


Lang and Norris ( 1966 ) 


Tursiops 


3.1 


90min 


Chivers and Scott'' 


Stenella 


3.0 


1.5 hours 


Au and Perryman ( 1982 ) 


Stenella 


2.8 


15 min 


Chivers and Scott-' 


Stenella 


2.6 


106 mm 


Chivers and Scott"* 


Stenella 


2.1 


days 


Chivers and Scott^ 


Stenella 


1.9 


hours 


Chivers and Scott'' 


Stenella 


1.8 


hours 


Chivers and Scott-' 


Stenella 


1.7 


32 min 


Chivers and Scott^ 


Stenella 


1.6 


4 days 


Hui(1987) 


Delphinus 


1.6 


hours-days 


Chivers and Scott-' 


Stenella 


1.5 


hours 


Chivers and Scott-' 


Stenella 


1.2 


hours 


Chivers and Scott-' 


Stenella 


1.2 


<50 hours 


Perrinetal. (1979) 


Stenella 



Edwards: Duration of unassisted swimming by Stenella attenuafa 



129 





No. of 


Length 


Weight 


Swimming 




animals 


(m) 


ikgl 


location 


Swimming condition 


3 






tank 


maximum observed leap velocity 


1 


1.86 


52.7 


lagoon 


accelerating, 25 m course 


1 


1.86 


52.7 


lagoon 


accelerating, 25 m course 


1 


1.86 


52.7 


lagoon 


accelerating, 25 m course 


3 






tank 


average maximum leap velocity 


1 






wild 


maximum observed speed 


3 






tank 


average leap velocity 


1 


1.91 


89 


lagoon 


61 m course in 300 m lagoon 


6 


2.6 


197 


small tank 


maximum speed, 8 m course 


300 






wild 


evading helicopter 


1 


1.91 


89 


lagoon 


61 m course; maximum speed observed 


1 


1.83 


105 


tank 


5 or 8 m course 


1 


2.09 


91 


tank 


accelerating, long narrow tank 


1 






captive 


70 m circular pool; maximum speed 


1 


1.91 


89 


ocean 


swimming in speedboat waves 


1 


1.91 


89 


ocean 


swimming in speedboat waves 


not specified 






wild 


during chase by speedboats 


1 


1.91 


89 


ocean 


swimming in speedboat waves 


"A school" 






wild 


average maximum observed speed 


1 


1.83 


105 


captive 


mean high swim speed 


1 


1.91 


89 


lagoon 


61 m course; maximum observed speed 


not specified 






wild 


after escaping purse seine 


6 


2.6 


197 


tank 


mean high swim speed 


1 


1.91 


89 


ocean 


swimming in speedboat waves 


41 






wild 


average maximum speed evading aircraft 


4 






wild 


maximum speed after release 


1 


1.91 


89 


ocean 


swimming in speedboat waves 


school no. 5 






wild 


maximum observed speed and duration 


school no. 5 






wild 


swimming with ocean swells 


"A school" 






wild 


average maximum velocity evading aircraft 


school 






wild 


average speed evading vessel 


1 


2.05 




wild 


average speed estimated from radiotag 


1 


1.91 


89 


ocean 


swimming in speedboat waves 


D230 






wild 


postrelease velocity 


school no. 2 






wild 


average speed evading vessel 


D230 






wild 


velocity during chase by seiner 


D19 






wild 


postrelease velocity 


12 






wild 


average speed, 1992, 1993 


D230 






wild 


average nonchase velocity, night 


D230 






wild 


average nonchase velocity, day 


D19 






wild 


velocity during chase by senior 


2 


1.76 


95 


tank 


small, shallow round tank 


2 






wild 


average speed, 2001 


D19 






wild 


average nonchase velocity, day 


D19 






wild 


average nonchase velocity, night 


26 






wild 


minimum distance traveled, radiotag 



130 



Fishery Bulletin 104(1) 











Table 2 












Dolphin mode 


parameters 


See 


text for formulas 


and rationale. 


















Wetted 


Maximum 




Dolphin 




Total length 


Body weight 




Muscle weight 


surface area 


diameter 


Fineness 


no. 


Age 


(cml 


(kg) 




(kg) 


(cm2) 


(cm) 


ratio 


1 


new born 


85 


6.40 




2.62 


2700 


13.5 


6.30 


2 


1 week 


87 


6.85 




2.81 


2832 


13.8 


6.29 


3 


1 month 


90 


7.58 




3.18 


3036 


14.3 


6.28 


4 


3 months 


98 


9.76 




4.29 


3615 


15.7 


6.24 


5 


6 months 


110 


13.76 




6.33 


4581 


17.7 


6.20 


6 


1 year 


129 


22.08 




11.04 


6351 


21.0 


6.14 


7 


2 years 


154 


37.37 




20.18 


9131 


25.4 


6.07 


8 


adult 


190 


69.73 




41.84 


14045 


31.7 


.=1.99 



C/ =o.o72i^-"^ 

where R is Reynold's number, estimated here as 
R = LV/v, 

where v = kinematic viscosity ( = 0.01 Stokes). 

The calculations above generated estimates of the 
power required for a whole dolphin of a given size to 
swim at a given velocity (P,, in watts). Power required 
per kilogram of wet weight muscle (P„, ), for a given 
velocity was estimated as 

where M„, = total muscle mass (wet weight in kg), esti- 
mated from total body mass (M,, wet weight in kg) as 



M„ 



-2.97M' 



based on In-ln regression of measurements from a sample 
of 26 Stenella attenuata from the ETP ranging in size 
from 0.71 to 2.06 m. Total muscle was used rather than 
some portion of measured musculature because the com- 
plex and interconnected muscle and connective tissue of 
dolphins makes it difficult to isolate any particular por- 
tion as uniquely responsible for locomotion (Pabst, 1990). 
M-i was estimated from total length (L, in cm) as 

M, =0.0000119L-^' 

based on In-ln regression of measurements from a sample 
of 23 Stenella attenuata from the ETP ranging in size 
from 0.71 to 2.01 m. 



Results 

Model corroboration 

Model estimates of the cost of swimming compared 
reasonably well with the cost of swimming in published 



reports for other species of dolphins swimming 1-6 m/s, 
in cases where the published reports can be appropri- 
ately compared with the present model. Although a 
number of studies present a variety of estimates of drag, 
thrust power, and metabolic power at various swimming 
speeds for a variety of dolphins (reviewed by Fish and 
Rohr, 1999), comparisons of present results with many 
of these earlier studies would be inappropriate because 
either the estimates were derived from completely dif- 
ferent models from the one used in the present study 
or from generally similar models but where different 
assumptions were made about model parameters, such 
as propulsive efficiency, metabolic efficiency, drag for- 
mulation, and body structure (Edwards'). Only appro- 
priately comparable published results are discussed in 
the present study. 

All weight-specific measurements in the following 
paragraph refer to total body mass. At estimated opti- 
mum velocities ranging from about 1.2 m/s in neonate 
spotted dolphins to about 1.7 m/s in adults (Edwards^), 
model estimates are approximately 3 W/kg for all sizes 
of spotted dolphin, compared to measurements (de- 
rived under various methods) of about 2.5-5.5 W/kg 
for adults of various species of dolphins, either resting 
or swimming about 2 m/s (Hui, 1987; Worthy et al., 
1987; Williams et al., 1992; Fish, 1993; Yadzi et al., 
1999). Observed average total metabolic rates calculat- 
ed from oxygen consumption by two Tursiops (average 
weight 162 kg), swimming 2 and 3 m/s, were approxi- 
mately 2.5 and 3.7 W/kg (Yadzi et al., 1999), compared 
to model estimates of approximately 2.9 and 5.9 W/kg 
for an adult spotted dolphin (about 70 kg) swimming 
at the same speeds. Model estimates of thrust power 
output for an adult spotted dolphin (about 70 kg) also 
compared well with thrust power estimated as a func- 
tion of velocity from videos of five Tursiops swimming 
between 1 and 6 m/s (Fish, 1993). Given an average 
adult Tursiops weight of 230 kg, average estimated 
thrust power for Tursiops swimming 1, 3, 5, and 6 m/s 
was 1, 3, 14, and 23 W/kg, compared to spotted dolphin 
model estimates of 1, 3, 13, and 21 W/kg, respectively. 
These comparisons refer to mechanical power output 



Edwards: Duration of unassisted swimming by Stenello oltenuata 



131 



only, because Fish's (1993) analysis does 
not include metabolic efficiency. Assum- 
ing the same metabolic efficiency of 0.2 
for the Tursiops in Fish's study as used 
in the present model, an estimate of total 
power by his Tursiops swimming 6 m/s is 
115 W/kg (i.e., 23x(l/0.2)), compared to 
the model estimate of about 125 W/kg for 
the adult spotted dolphin. 

Mass-specific cost of swimming 



300 



250 - 



200 



E 

5' 150 



100 



50 



Model estimates of mass-specific power 
requirements (watts per kilogram muscle, 
W/kg,„) for unassisted swimming by spot- 
ted dolphins in the ETP increase quickly 
with velocity, regardless of dolphin size, 
but increase much more quickly for smaller 
dolphins (Fig.l). Model results indicated 
that a neonate (85 cm) spotted dolphin 
must produce 3.6 times more power per 
kilogram of muscle than an adult swim- 
ming at the same speed. The factors are 
3.5, 3.3, 2.8, 2.4, 1.8, and 1.4 times adult 
power for 87-, 90-, 98-, 110-, 129-. and 
154-cm spotted dolphins. The factors do 
not vary with velocity because the relative differences 
between body dimensions in dolphins of different sizes 
remain constant regardless of swimming speed. 

For example, at speeds typical of ETP dolphins at- 
tempting to evade speedboats during an impending 
tuna-purse-seine set or during escape from the net 
after release (about 3 m/s) (Chivers and Scott^), the 
model predicts that an 85-cm neonate swimming un- 
assisted by its mother would require a muscle power 
of about 105 W/kg„, versus 30 W/kg„, for a 190-cm 
adult. The difference is still relatively marked even for 
two-year-old calves (154 cm), which would require an 
estimated 42 W/kg,,^ to maintain a steady submerged 
unassisted swim speed of 3 m/s, i.e., 1.4 times adult 
power requirements. 

Velocity duration limits 



Velocity duration limits for adults, drawn from litera- 
ture sources, appear to range from a few seconds for 
burst speeds above about 5 m/s, to several minutes for 
maximum speeds of about 4 m/s, to hours at prolonged 
speeds below about 3.5 m/s (Table 1). 

Estimated mass-specific muscle power required of 
adult spotted dolphins to achieve these speeds ranges 
from 208 W/kg,,, for burst speeds of 6 m/s greater, to 67 
W/kg,,, for maximum sustainable speeds around 4 m/s, 
and 4-10 W/kg„, for prolonged duration speeds between 
1 and 2 m/s (Fig.l). 

If one assumes that these power-duration limits apply 
also to dolphin calf muscle, neonate dolphins (85 cm) 
for example, could generate a burst duration power of 
about 200 W/kg,„ for a few seconds at unassisted swim- 
ming speeds of about 3.0 m/s, maximum duration power 



Estimated duration limits 



-Seconds 



-Minutes 



-Hours 



00 




m/s 



Figure 1 

Estimated swimming duration limits for eastern tropical Pacific Ocean 
spotted dolphins iStenella attenuata) of various sizes (total length, tip 
of rostrum to fluke notch), swimming at various speeds. 



of about 70 W/kg,,, for some minutes at about 2.2 m/s, 
and prolonged duration power of about 5-10 W/kg„, for 
hours at about 1.0 m/s (Fig.l). For two-year-old spotted 
dolphins (154 cm), estimated swimming duration limits 
were about 5 m/s for burst effort at 200 W/kg„,, about 
3.5 m/s for maximum effort at 70 W/kg„,, and about 2 
m/s for prolonged effort at 5-10 W/kg,„. Intermediate 
ages produced intermediate results. 



Discussion 

Model results demonstrated clearly that swimming capa- 
bility of dolphin calves is much lower than that of adults. 
The especially pronounced difference between neonates 
and adults appears a likely basis for development of the 
ubiquitous drafting behavior observed in both captive 
and wild neonates. Given the large difference in swim- 
ming capacities, it is difficult to imagine how small 
dolphin calves could maintain any long-term association 
with the mother without the hydrodynamic benefits of 
drafting, even under normal circumstances. 

During the relatively fast-paced evasion swimming 
associated with tuna purse-seine sets in the ETP, it 
appears that spotted dolphin calves swimming unas- 
sisted are likely to have serious difficulty maintaining 
the same speed as adults — difficulties being especially 
pronounced for younger dolphins. It is also likely that 
the differences in swimming capacity presented in this 
study are underestimates because they do not include 
effects such as positive bouyancy (Cockroft and Ross, 
1990; Mann and Smuts, 1999) and soft, flexible fins 
and flukes that are present for several hours after birth 
(McBride and Kritzler, 1951; Tavolga and Essapian, 



132 



Fishery Bulletin 104(1) 



1957; Wells, 1991). Other differences persist for weeks 
or months, including undeveloped musculature (Ei- 
tner et al., 2003) and uncoordinated swimming and 
respiratory movements (Tavolga and Essapian, 1957; 
Taylor and Saayman, 1972; Reiss, 1984; Cockroft and 
Ross, 1990; Peddemors, 1990; Herzing, 1997; Mann 
and Smuts. 1999). Aerobic capacity is also likely to 
be reduced throughout the first few months, and not 
likely to reach adult levels until two to three years of 
age (Dolar et al., 1999; Dearoff et al., 2000; Noren et 
al., 2001, 2002, 2004). The results presented in this 
study also do not include the added costs of increased 
drag due to swimming near the surface, and repeatedly 
piercing it. which will occur much more often during 
evasion of tuna sets than during normal swimming. 
These other effects increase the likelihood that dolphin 
calves, particularly the younger individuals, will have 
problems coping with the swimming conditions induced 
during tuna purse-seine sets. 

Difficulties with unassisted swimming for dolphin 
calves may be ameliorated to some extent by employing 
drafting (Weihs, 2004), burst and coast (Au and Weihs, 
1980), or leap-burst-coast swimming behaviors (or a 
combination of these behaviors) (Weihs, 2002). Theo- 
retically, these strategies could significantly reduce the 
cost to calves of moving through the water. However, it 
is not clear that any of these energy-saving strategies 
will be consistently attainable by spotted dolphin calves 
during herd movements associated with evading or es- 
caping sets by tuna purse-seiners in the ETP. 

Drafting can only be sustained through respiratory 
leaps if mother and calf leave and reenter the water 
with equal speed and efficiency (Weihs, 2004). Because 
evasion of tuna purse-seine sets involves sustained 
high-speed swimming characterized by repeated full- 
body respiratory leaps from the water (Au and Wiehs. 
1980), calves are likely to tire and lose coordination 
more quickly than adults, and the smallest calves will 
be the first to experience problems. These factors may 
have contributed to the observed failure of a neonate 
dolphin calf in the ETP to successfully maintain a 
drafting relationship with its assumed mother during 
a respiratory leap while attempting to evade a vessel 
(Weihs, 2004). 

Once the drafting relationship is disrupted, the calf 
appears likely to fall behind because of its physical 
limitations, unless its mother alters her speed so that 
the calf can reestablish the drafting relationship. How- 
ever, review of dolphin mother-calf behavior indicates 
that the mother is not likely to voluntarily leave other 
adults during attempted evasion of tuna purse-seine 
sets in the ETP (EdwardsM. Thus, the faster or longer 
the chase or postrelease escape period (or combination 
of all three factors), and the younger the calf, the more 
likely it appears that the calf will become separated 
from its mother and be left behind during periods of 
fast swimming by the adults. Although fast swimming 
for a few minutes may not pose a great problem for 
many calves, longer periods appear increasingly likely 
to lead to significant calf loss in purse-seine operations. 



The duration of the high-speed period is at least as 
important as the speed maintained during the period, 
given the power-duration relationship. If fast swim- 
ming is concluded quickly, it is more likely that calves 
could achieve the required power during the short time 
required. As fast-swimming persists, power capacity 
decreases rapidly (Fig. 1), so that more and more calves 
are likely to be lost because they have exceeded their 
ability to keep up with the adults in their school. 

Even under normal circumstances, burst and coast, 
or burst-leap-coast, swimming patterns are not likely to 
provide sustained benefits to free-swimming calves until 
they approach adult size, coordination, and swimming 
capacity, because the calves will still tire more quickly 
than their associated adults. In the case of tuna purse- 
seine set evasion, sustained use of these swimming 
patterns may not be employed even by adults because 
these behaviors become less efficient as swim speed 
increases (Weihs, 2002). Video studies of swimming 
behaviors of adult Tursiops have shown that burst and 
coast periods decrease with increasing drag and cease 
altogether if the drag is large (Skrovan et al., 1999). In 
a study of dolphin gaits, an adult Tursiops experiencing 
increased drag due to an instrument pack did not use 
burst and coast propulsion during horizontal swimming 
speeds and depths at which four other adult Tursiops 
(without instrument packs) did employ short periods 
of burst and coast swimming. Even animals swim- 
ming without instruments at 1.5-3.7 m/s incorporated 
only short periods of burst and coast, and glide periods 
rarely exceeded 2 seconds. At the speeds characteristic 
of tuna purse-seine sets, dolphins tend to swim steadily 
or employ burst-leap-coast swimming behaviors (Weihs, 
2002, Au and Weihs, 1980). With their smaller power 
production capacities, calves will need to shift into 
leap-burst-coast mode at slower speeds than will adults 
(Weihs, 2002), and again will tire more quickly. 

A potential shortcoming of the results presented in 
the present study is that the model produces absolute 
values from single calculations, without any associated 
statistics or sensitivity analyses. It is more likely that 
a range of values for morphological, physiological and 
behavioral characteristics occurs within real spotted 
dolphin populations in the ETP, so that a range of re- 
sponses is much more likely than a single response for 
a given size, swimming speed, and duration of chase. 
Lack of statistical and sensitivity analyses can be a 
problem where responses tend to be subtle and there- 
fore difficult to discern. However, the model results 
presented in this study with respect to estimated size- 
related differences in power production capacities of 
spotted dolphins in the ETP are far from subtle, and 
model results are reasonably similar to energy mea- 
sures derived from observations on real dolphins indi- 
cating that these model results are not unrealistic. In 
general, results presented here imply that older calves 
may be able to swim unassisted as fast and for as long 
as their associated adults during or after most sets (or 
during and after most sets) by tuna purse-seine vessels 
in the ETP. The markedly smaller energy production 



Edwards Duration of unassisted swimming by Stenella atlenuota 



133 



capacities of younger calves, particularly neonates and 
infants, compared to adults appears to be a reasonable 
factor contributing to mother-calf separation during or 
after sets by tuna vessels in the ETP. 

Management and research implications 

From these observations, it appears that energetic limi- 
tations may contribute significantly to the observed 
calf deficit in the retrieved purse-seine by facilitating 
mother-calf separation during purse-seine set activ- 
ity. Separation risk appears to increase strongly with 
decreasing calf size and with increased speed and dura- 
tion of evasion-related swimming. If separation is pro- 
longed, subsequent mortality of milk-dependent calves 
appears likely due to predation or starvation because 
adoption by a surrogate mother is unlikely in the ETP 
(EdwardsM. Examination of existing aerial photographs 
of spotted dolphins attempting to evade helicopters and 
research vessels, as well as directed future experiments 
in which aerial observations are conducted specifically to 
identify and monitor mother-calf pairs over time during 
various chase scenarios, would be very helpful in evalu- 
ating the extent to which calf size effects predicted here 
occur during relatively high-speed evasion behaviors by 
ETP spotted dolphin schools. Estimation of the rate at 
which dolphins of various ages are likely to encounter 
tuna purse-seine sets will be another important factor 
in relating the results of this analysis to potential calf 
loss during purse-seine sets. 

Assuming that power limitations are a significant 
factor in calf loss during evasion of tuna purse-seine 
sets, management strategies that could be implemented 
to minimize calf loss include 1) avoiding setting on 
schools that contain calves, particularly young calves, 
2) minimizing the speed or duration of chase, and 3) 
minimizing the length of time dolphins are retained in 
the net so that mothers and calves separated during 
the set may reunite more quickly, if possible. Minimiz- 
ing the length of time in the net is already a desired 
fishery goal, related to minimizing cost and the poten- 
tial for dolphin mortality in the net. Minimizing chase 
duration is also already a desired fishery goal, because 
shorter chases are both less expensive and tend to be 
more successful. Minimizing chase speed is not a re- 
alistic goal because it would likely lead to the escape 
of the dolphins and the targeted tuna. Thus, the only 
potential improvement that might be made under cur- 
rent conditions (i.e., while setting on dolphin calves is 
permitted) would be to concentrate effort on identifying 
and avoiding dolphin schools with calves, presumably by 
scrutinizing the school from the vessel's helicopter prior 
to initiating a set. This may or may not be possible, 
depending on the ability of observers in helicopters to 
spot calves from the air. 

Given the model results presented here, it appears 
that further research and perhaps management actions 
should be implemented to better understand and reduce 
the risk of separation of mothers and calves during sets 
by tuna purse-seiners in the ETP. 



Acknowledgments 

This research was conducted under Section 304 (a)(1): 
Stress studies, of the 1997 International Dolphin Con- 
servation Program Act amendments to the U.S. Marine 
Mammal Protection Act, and subsequent Congressional 
directives. 



Literature cited 

Alexander. R. 

1983. Animal mechanics. Blackwell Scientific Publica- 
tions, London, England. 
Archer, F., T. Gerrodette, S. Chivers, and A. Jackson. 

2004. Annual estimates of the unobserved incidental 
kill of pantropical spotted dolphin (Stenella attenu- 
ata attenuata) calves in the eastern tropical Pacific 
Ocean. Fish. Bull. 102:233-244. 
Au, D., and W. Ferryman. 

1982. Movement and speed of dolphin schools responding 
to an approaching ship. Fish. Bull. 80:371-379. 
Au, D., M. Scott, and W. Ferryman. 

1988. Leap-swim behavior of "porpoising" dolphins. 
Cetus 8:7-10. 
Au, D., and D. Weihs. 

1980. At high speeds dolphins save energy by leaping. Na- 
ture 284:548-550. 
Cockroft, v., and G. Ross. 

1990. Observations in the early development of a captive 
bottlenose dolphin calf In The bottlenose dolphin (S. 
Leatherwood and R. Reeves, eds.), p. 461-478. Aca- 
demic Press, San Diego, CA. 
Dearoff, J., W. McLellan, R. Dillaman, D. Frierson, and D. 
Fabst. 

2000. Precocial development of axial locomotor muscle 
in bottlenose dolphins (Tursiops truncatusl J. Morph. 
244:203-215. 
Dickinson, S. 

1929. The efficiency of bicycle pedaling, as affected by 
speed and load. J. Phys. Lond. 76:242-255. 
Dolar, M., F. Suarez, P. Fonganis, and G. Kooyman. 

1999. Myoglobin in pelagic small cetaceans. J. E.xp. 
Biol. 202:227-236. 
Edwards, E. 

1992. Energetics of associated tunas and dolphins in 
the eastern tropical Pacific Ocean: a basis for the 
bond. Fish. Bull. 90:678-690. 

1993. Allometry of energetics parameters in spotted 
dolphin iStenella attenuata) from the eastern tropical 
Pacific Ocean. Fish. Bull. 91:428-439. 

Eitner, S., J. Dearoff, W. McLellan, and D. Fabst. 

2003. Postural role of lateral axial muscles in develop- 
ing bottlenose dolphins {Tursiops truneatus). Proc. R. 
Soc. Lond. B 271:909-918. 
Fish, F 

1993. Power output and propulsive efficiency of swim- 
ming bottlenose dolphins (Tursiops truneatus). J. Exp. 
Biol. 185:179-183. 

1996. Transitions from drag-based to lift-based propulsion 
in mammalian swimming. Am. Zool. 36:628-641. 

1998. Comparative kinematics and hydrodynamics of 
odontocete cetaceans: morphological and ecological 
correlates with swimming performance. J. Exp. Biol. 
201:2867-2877. 



134 



Fishery Bulletin 104(1) 



Fish, F., and J. Rohr. 

1999. Review of dolphin hydrodynamics and swimming 
performance. Tech. Rapt. 1801, 136 p. SPAWAR Sys- 
tems Center, SSC San Diego, San Diego, CA. 
Goldspinlt, G. 

1988. Muscle energetics. In Mechanics and energetics 
of animal locomotion (R. Ale.xander and G. Goldspink, 
eds.), p. 57-81. Chapman and Hall, London, UK. 
Hertel, H. 

1969. Hydrodynamics of swimming and wave-riding 
dolphins. In The biology of marine mammals (H. Ander- 
son, ed.), p. 31-63. Academic Press, New York, NY. 
Herzing, D. 

1997. The life history of free-ranging Atlantic spot- 
ted dolphins iStenella frontalis): age classes, color 
phases, and female reproduction. Mar. Mamm. Sci. 
13(4)1.576-595. 
Hoerner, S. 

1965. Fluid-dynamic drag, 460 p. Hoerner, 148 Busteed 
Drive, Midland Park, NJ. 

Hohn, A., and P. Hammond. 

1985. Early postnatal growth of the spotted dolphin, 
Stenella attenuata, in the offshore eastern tropical 
Pacific. Fish. Bull. 83:553-566. 
Hui, C. 

1987. Power and speed of swimming dolphins. J. Mamm. 
68:126-132. 
Lang, T. 

1975. Speed, power, and drag measurements of dolphins 
and porpoises. /;; Swimming and flying in nature (T. 
Wu and C Brokaw, eds.), p. 553-572. Plenum Press, 
New York, NY. 
Lang, T., and K. Norris. 

1966. Swimming speed of a Pacific bottlenose por- 
poi.se. Science 151(3710):588-590. 

Lang, T., and K. Pryor. 

1966. Hydrodynamic performance of porpoises (Stenella 
attenuatta). Science 152:531-533. 
Leatherwood, S., and D. Ljungblad. 

1979. Nightime swimming and diving behavior of a radio- 
tagged spotted dolphin, Stenella attenuata. Cetology 
34:1-6. 
Lighthill, Sir J. 

1971. Large-amplitude elongated-body theory of fish 
locomotion. Proc. R. Soc. Lond. B. 179:125-138. 
Magnuson, J. 

1978. Locomotion by scombroid fishes; hydromechanics, 
morphology and behavior. /;; Fish physiology, vol. 7 
(W. Hoar and J. Randall, eds.), p. 240-313. Academic 
Press, New York, NY. 
Mann, J., and B. Smuts. 

1999. Behavioral development in wild bottlenose dolphin 
newborns (Tursiops sp.}. Behavior 136:529-566. 
McBride, A., and H. Kritzler. 

1951. Observations on pregnancy, parturition, and post- 
natal behavior in the bottlenose dolphin. J. Mamm. 
32.251-266. 
Myrick, A., and P. Perkins. 

1995. Adrenocortical color darkness as indicators of con- 
tinuous premortem stress in chased and purse-seined 
male dolphins. Pathophys. 2:191-204. 
NRC (National Research Council). 

1992. Dolphins and the tuna industry, 176 p. Committee 
on Reducing Porpoise Mortality from Tuna Fishing, 
National Research Council. National Academy Press, 
Washington, D.C. 



Noren, S., V. Cuccurulo, and T. Williams. 

2004. The development of diving bradycardia in bottle- 
nose dolphins (Turs/op.s- truncatua). J. Comp. Physiol. 
B 174:139-147. 
Noren, S., G. Lacave, R. Wells, and T. Williams. 

2002. The development of blood oxygen stores in bot- 
tlenose dolphins (Tursiops truncatus): implications for 
diving capacity. J. Zool. Lond. 258:105-113. 
Noren, S., T. Williams, D. Pabst, W. McLellan, and J. Dearoff 

2001. Development of diving in marine endotherms: 
preparing the skeletal muscles of dolphins, penguins, 
and seals for activity during submergence. J. Comp. 
Physiol. 171:127-134. 

Norris, K., and J. Prescott. 

1961. Observations on Pacific cetaceans of Califor- 
nian and Mexican waters. Univ. Calif. Publ. in Zool. 
63:291-402. 
Pabst, A. 

1990. Axial Muscles and Connective Tissues of the Bottle- 
nose Dolphin. In The bottlenose dolphin (S. Leather- 
wood and R. Reeves, eds. I. p. 51-67. Academic Press, 
San Diego, CA. 
Peddemors, V. 

1990. Respiratory development in a captive-born bottle- 
nose dolphin. S. Afr. J. Zool. 5:178-184. 
Perrin, W., W. Evans, and D. Holts. 

1979. Movements of pelagic dolphins iStenella spp.) in 
the eastern tropical Pacific as indicated by results of 
tagging operations, 1969-1976. NOAA Tech. Rept. 
NMFS SSRF-737, 14 p. 
Perrin, W., and S. Reilly. 

1984. Reproductive parameters of dolphins and small 
whales of the family Delphinidae. Rep. Int. Whal. 
Comm. (special issue 6):97-133. 
Reiss, D. 

1984. Observations on development of echolocation in 
young bottlenose dolphins. In Animal sonar ( P. Natchti- 
gall and P. Moore, eds.), p. 121-127. Plenum Press, 
New York, NY. 
Ridgway, S., and D. Johnston. 

1965. Blood oxygen and ecology of porpoises of three 
genera. Science 151:456-458. 
Rohr, J., F. Fish, and J. Gilpatrick. 

2002. Maximum swim speeds of captive and free- 
ranging delphinids: critical analysis of extraordinary 
performance. Mar. Mamm. Sci. 18(1):1-19 

Skrovan, R., T. Williams. P. Berry, P. Moore, and R. Davis. 

1999. The diving physiology of bottlenose dolphins (Tur- 
siops truncatus] II. Biomechanics and changes in bouy- 
ancy at depth. J. Exp. Biol. 202:2749-2761. 
Taylor, C, and G. Saayman. 

1972. Social organization and behavior of dolphins 
and baboons. Ann. Cape Prov. Mus. Nat. Hist. 9(2): 
11-49. 
Tavolga, M., and F. Essapian. 

1957. The behavior of the bottle-nosed dolphin (Tursiops 
truncatus): mating, pregnancy, parturition, and mother- 
infant behavior. Zoologica 42:11-31. 
Webb, R 

1975. Hydrodynamics and energetics of fish pro- 
pulsion. Bull. Fish. Res. Board Can. 190:1-158. 
Weihs, D. 

2002. Dynamics of dolphin porpoising revisited. Integr. 

Comp. Biol. 42:1071-1078. 
2004. The hydrodynamics of dolphin drafting. J.Biol. 
3(8):l-23. 



Edwards Duration of unassisted swimming by Stenella atlenuota 



135 



Wells, R. 

1991. Bringing up baby. Nat. Hist. August:56-62. 
Williams, T., W. Friedl, M. Fong, R. Yamada, P. Sedivy, and J. 

Haun. 

1992. Travel at low energetic cost by swimming and 
wave-riding bottle-nosed dolphins. Nature 355 (6363): 
821-823. 

Worthy, G., S. Innes, B. Braune, and R. Stewart. 

1987. Rapid acclimation of cetaceans to an open-system 
respirometer. //; Approaches in marine mammal ener- 



getics (A. Huntley, D. Costa, G. Worthy and M. Castellini, 
eds.l, p. 115-126. Soc. Mar. Mamm. Spec. Publ. 1. Soc. 
Mar. Mamm., P.O. Box 368. Lawrence, KS 66044. 

Yadzi, P., A. Kilian, and B. CuHk. 

1999. Energy expenditure of swimming bottlenose 
dolphins. Mar. Biol. 134:601-607. 

Yates, G. 

1983. Hydrodynamics of body and caudal fin pro- 
pulsion. In Fish biomechanics (P. Webb and D. Weihs, 
eds.), p.177-213. Praeger Publishers, New York, NY. 



136 



Abstract — We assayed allelic varia- 
tion at 19 nuclear-encoded microsat- 
ellites among 1622 Gulf red snapper 
iLutjanus campechanus) sampled 
from the 1995 and 1997 cohorts at 
each of three offshore localities in 
the northern Gulf of Mexico (Gulf). 
Localities represented western, cen- 
tral, and eastern subregions within 
the northern Gulf. Number of alleles 
per microsatellite per sample ranged 
from four to 23, and gene diversity 
ranged from 0.170 to 0.917. Tests of 
conformity to Hardy-Weinberg equi- 
librium expectations and of genotypic 
equilibrium between pairs of micro- 
satellites were generally nonsignifi- 
cant following Bonferroni correction. 
Significant genie or genotypic het- 
erogeneity (or both) among samples 
was detected at four microsatellites 
and over all microsatellites. Levels 
of divergence among samples were 
low (Fj,.;. sO.OOl). Pairwise exact tests 
revealed that six of seven "significant" 
comparisons involved temporal rather 
than spatial heterogeneity. Contem- 
poraneous or variance effective size 
(N^y) was estimated from the tempo- 
ral variance in allele frequencies by 
using a maximum-likelihood method. 
Estimates of N^,y ranged between 1098 
and >75,000 and differed significantly 
among localities; the N^.y estimate for 
the sample from the northcentral Gulf 
was >60 times as large as the esti- 
mates for the other two localities. The 
differences in variance effective size 
could reflect differences in number 
of individuals successfully repro- 
ducing, differences in patterns and 
intensity of immigration, or both, and 
are consistent with the hypothesis, 
supported by life-history data, that 
different "demographic stocks" of red 
snapper are found in the northern 
Gulf. Estimates of N^,y for red snap- 
per in the northern Gulf were at least 
three orders of magnitude lower than 
current estimates of census size (A^). 
The ratio of effective to census size 
{N^,/N) is far below that expected in an 
ideal population and may reflect high 
variance in individual reproductive 
success, high temporal and spatial 
variance in productivity among subre- 
gions or a combination of the two. 



Manuscript submitted 17 July 2004 
to the Scientific Editor's Office. 

Manuscript approved for publication 
25 July 2005 by the Scientific Editor. 

Fish. Bull. 104:136-148 (2006). 



Population structure and variance effective size 
of red snapper (Lutjanus campechanus) 
in the northern Gulf of Mexico* 



Eric Saillant 

John R. Gold 

Center for B)osystematics and Biodiversity 
Texas A&M University, TAMU-2258 
College Station, Texas 77843-2258 
E-mail address (for E Saillant) esaillantiStamu edu 



Red snapper (Lutjanus campechanus) 
is a highly exploited marine fish found 
primarily on the continental shelf of 
the Gulf of Mexico (Hoese and Moore, 
1977). Red snapper abundance in the 
northern Gulf of Mexico (hereafter. 
Gulf) has decreased by almost 90% 
in the past two decades (Goodyear 
and Phares') owing to overexploita- 
tion by commercial and recreational 
fishermen, high juvenile mortality due 
to the shrimp-trawl fishery, and habi- 
tat change (Christman-; Gallaway et 
al., 1999). An important question for 
management and conservation of red 
snapper resources regards delineation 
of geographic stock structure. Should 
separate stocks exist, management of 
the fishery, including assessment and 
allocation, could be subdivided to avoid 
subregional over-exploitation and to 
maintain potentially adaptive genetic 
variation (Carvalho and Hauser, 1995; 
Hauser and Ward, 1998). A second 
important question for management 
is whether sufficient genetic resources 
exist to ensure long-term integrity of 
red snapper stocks. Preliminary esti- 
mates of the number of red snapper 
adults in the northern Gulf range from 
7.8 to 11.7 million (Cowan-'; Porch""), 
which may indicate a priori that suf- 
ficient genetic resources are available. 
However, recent studies in other, com- 
mercially exploited marine fishes have 
shown that genetic effective size iN^.) 
can be three-five orders of magnitude 
smaller than estimates of census size 
or N (Hauser et al., 2002; Turner et 
al., 2002). Briefly, A^,, is defined as the 
number of individuals in an "ideal" 



population that would experience the 
same magnitude of genetic drift as the 
actual population (Hartl and Clark, 
1989). N^. is an important biological 
parameter, in part because it reflects 
the relative effects of genetic drift and 
selection on nonneutral loci, and in 
part because it can indicate long-term 
risk of extinction from genetic factors 
(Turner et al., 2002). As long-term 
sustainability requires maintenance of 
sufficient genetic resources (Allendorf 
and Waples, 1996), populations (or 
stocks) with small N, potentially may 



* Contribution 133 from the Center for 
Biosystematics and Biodiversity, Texas 
A&M University, College Station, Texas 
77843-2258. 

1 Goodyear, C. P., and P. Phares. 1990. 
Status of red snapper stocks of the Gulf 
of Mexico: report for 1990, 72 p. Na- 
tional Marine Fisheries Service, South- 
east Fisheries Centre, Miami Laboratory, 
CRD 89/90-05, 75 Virginia Beach Drive, 
Miami, FL 33149-1099. 

- Christman, M. C. 1997. Peer review 
of red snapper (Lutjanus caii^pechanus) 
research and management in the Gulf 
of Mexico: statistics review, 51 p. Of- 
fice of Science and Technology, National 
Oceanic and Atmospheric Administra- 
tion, National Marine Fisheries Service, 
1315 East-West Highway 9th Floor F/CS, 
Silver Spring, MD 20910. 

■^ Cowan, J. H. 2004. Personal commun. 
Department of Oceanography and 
Coastal Sciences, Coastal Fisheries 
Institute, Louisiana State University, 
Baton Rouge, LA 70803. 

■• Porch, C. E. 2004. Personal commun. 
National Marine Fisheries Center, 
Southeast Fisheries Science Center, 
75 Virginia Beach Drive, Miami, FL 
33149. 



Saillant and Gold: Population structure and variance effective size of Lut/anus campechanus in tfie nortfiern Gulf of Mexico 



137 



Sampl 
in the 
centra 



suffer reduced capacity to respond to changing 
or novel environmental pressures (Frankham, 
1995; Higgins and Lynch, 2001). 

At present, red snapper resources in the 
northern Gulf are managed under a single- 
stock hypothesis (GMFMC-^ «). This hypoth- 
esis is supported by a number of prior genetic 
studies that employed allozymes (Johnson"), 
mitochondrial (mt)DNA (Gold et al., 1997; Gar- 
ber et al., 2004), and microsatellites (Gold et 
al., 2001b). In each study, genetic homogene- 
ity was observed across sampling localities, 
leading to the inference that gene flow was 
sufficient to maintain statistically identical 
allele distributions across the sampling area. 
All of these studies, however, either involved 
individuals of mixed cohorts or were based on 
relatively small sample sizes. Alternatively, 
tag-and-release and ultrasonic tracking (Fable, 
1980; Szedlmayer and Shipp, 1994; Szedlmay- 
er, 1997) have indicated that adult red snapper are 
sedentary and exhibit high site fidelity (but see Patter- 
son et al., 2001). In addition, Pruett et al. (2005) used 
nested-clade analysis of red snapper mtDNA sequences 
and found evidence of different temporal episodes of 
both range expansion and restricted gene flow due to 
isolation by distance. They suggested that the spatial 
distribution of red snapper in the northern Gulf had 
a complex history that likely reflected glacial advance 
and retreat, habitat availability and suitability, and 
that the latter (i.e., physical conditions, and habitat 
availability and suitability) could partially restrict 
gene flow among present-day red snapper. 

Our objectives were to more rigorously assess genetic 
stock structure in the northern Gulf by employing a 
large sample size of individuals from discrete cohorts. 
We report allelic variation at 19 nuclear-encoded mic- 
rosatellites sampled from each of two cohorts at three 
different localities in the northern Gulf. Genetic homo- 
geneity among localities was tested and contemporane- 
ous or variance effective size (A/',,) at each locality was 
estimated from the temporal variance in allele frequen- 
cies (Waples, 1989) by using a maximum likelihood 
method (Wang, 2001). 



N 

t 






^ 


"T^ 










^ 




AL \ ^'^ 


") 


MS 


\ TX 




\ U 


r— -^— — 


A ' 


V/~A 






J^ 


^^Jjpv^ 


\ 


\ 




^ 


% Alabama 


FL 




V. 


r 

1 


Louisiana 
Texas 

t 


'^ 



33°N 



27°N 



96 W 



84°W 



= GMFMC I Gulf of Mexico Fishery Management Coun- 
cil). 1989. Amendment number 1 to the Reef Fish Fishery 
Management Plan. 356 p. Gulf of Mexico Fishery Manage- 
ment Council, 2203 N. Lois Avenue, Suite 1100, Tampa, FL 
33609. 

•^ GMFMC (Gulf of Mexico Fishery Management Coun- 
cil). 1991. Amendment number 3 to the Reef Fish Fishery 
Management Plan, 17 p. Gulf of Mexico Fishery Manage- 
ment Council, 2203 N. Lois Avenue, Suite 1100, Tampa, FL 
33609. 

' Johnson, A. G. 1987. An investigation of the biochemical 
and morphometric characteristics of red snapper tLutjanus 
canipecltanus) from the southern United States. Unpubl. 
manuscr., 26 p. National Marine Fisheries Service, Panama 
City Laboratory. 3.500 Delwood Beach Road, Panama Citv, 
FL 32407-7499. 



Figure 1 

ing localities for adult red snapper iLiitJaniis campechanus) 
northern Gulf of Mexico: northwestern Gulf (Texas), north- 
1 Gulf (Louisiana), and northeastern Gulf (Alabama). 

Materials and methods 

Adult red snapper were sampled between 1999 and 
2001 by angling 40-50 km offshore of Port Aransas 
(Texas), Port Fourchon (Louisiana), and Dauphin Island 
(Alabama). These localities represent western, cen- 
tral, and eastern subregions, respectively, within the 
northern Gulf (Fig. 1) but hereafter for convenience 
are referred to as Texas, Louisiana, and Alabama. 
Individual fish were aged by otolith-increment analysis 
(following Wilson and Nieland, 2001) and individuals 
belonging to the 1995 and 1997 cohorts were selected 
for genetic analysis. Sample sizes for the 1995 and 1997 
cohorts at each locality were 203 and 211 (Texas), 286 
and 272 (Louisiana), and 376 and 274 (Alabama). Tissue 
samples (heart and muscle) were removed from each 
fish and stored as described in Gold et al. (2001b). The 
genotype of all fish was determined at 19 microsatel- 
lites by using PCR primers and methods described in 
Gold et al. (2001b). 

Summary statistics, including number of alleles, allel- 
ic richness (a measure of number of alleles independent 
of sample size), and unbiased gene diversity (expected 
heterozygosity) were computed for each microsatellite in 
each sample, with F-stat, version 2.9.3 (Goudet, 1995). 
Homogeneity of allelic richness and gene diversity 
among samples was tested with Friedman rank tests. 
Departure of genotypic proportions from Hardy-Wein- 
berg equilibrium expectations was measured within 
samples as Weir and Cockerham's (1984) f; probability 
of significance (Pjjw' was assessed with a Markov-chain 
method (Guo and Thompson, 1992), as implemented in 
Genepop (Raymond and Rousset, 1995) and by using 
5000 dememorizations, 500 batches, and 5000 itera- 
tions per batch. Genotypic disequilibrium between pairs 
of microsatellites within samples was tested by exact 
tests, as implemented in Genepop and by employing the 
same Markov-chain parameters as above. Sequential 
Bonferroni correction (Rice, 1989) was applied for all 
multiple tests performed simultaneously. 



138 



Fishery Bulletin 104(1) 



Homogeneity of allele and genotype distributions 
among samples was examined with exact tests; signifi- 
cance of probability values was assessed by a Markov- 
chain method, as implemented in Genepop and using the 
same Markov-chain parameters as above. The degree of 
differentiation between pairs of samples was estimated 
as Weir and Cockerham's (1984) 9. as implemented in 
F-Stat. Sequential Bonferroni correction (Rice, 1989) 
was applied for all multiple tests performed simultane- 
ously. Spatial (geographic) differences among samples 
was assessed from multilocus data by estimating the 
likelihood that any given individual could be assigned 
to the sample locality from which it was drawn. The 
Bayesian method of Rannala and Mountain (1997), as 
implemented in Geneclass vers. 2.0 (Piry et al., 2005), 
was used to "assign" sampled individuals to a locality; 
the probability that an individual belonged to a given 
locality was calculated by using the resampling algo- 
rithm in Paetkau et al. (2004) and was based on 1000 
simulated individuals. A locality was excluded as a 
potential origin of a given individual if the probability 
of the individual belonging to that locality fell below a 
threshold level of 0.05. 

Temporal changes in allele frequencies between the 
two cohorts were used to estimate variance effective 
size (Afj.y.) at each locality. This "temporal" method (Wa- 
ples, 1989) estimates effective size from the temporal 
variance in allele frequencies over the time interval 
between sampling, thus providing a contemporaneous 
estimate of A^,.. The pseudo-maximum-likelihood method 
described in Wang (2001) was used to obtain estimates 
and 95% confidence intervals of N^y by using the pro- 
gram MLNE available at http://www.zoo.cam.ac.uk/ 
ioz/software.htm#MLNE. The 95% confidence intervals 
were obtained as the range of support associated with 
a drop of two logarithm units of the likelihood func- 
tion, as inferred from the likelihood distribution (Wang, 
2001). We used the analytical method developed by 
Jorde and Ryman (1995, 1996) to account for effects of 
overlapping generations on temporal-method estimates 
of N,. In a population with overlapping generations, the 
magnitude of temporal allele-frequency change is depen- 
dent in part on age-specific survivorship (/,) and birth 
rate (&,). Survivorship was calculated by assuming an 
equal probability (S) of surviving from one year class to 
the next and equal probability of survival of males and 
females. The value of -S (0.56 for Texas and 0.604 for 
Louisiana and Alabama) was estimated by using age- 
structure data of red snapper to calculate age-specific 
survivorship </^=S' ') for each age class ;. Birth rate was 
estimated by calculating mean individual (wet) weight 
at each age class, as an indicator of relative gamete con- 
tribution. Individual weights averaged across males and 
females within each age class were determined by using 
von Bertalanffy equations (Fischer et al., 2004) for red 
snappers at each locality; this mean value was then 
multiplied by /, to obtain the proportional contribution 
of each age class to offspring (p, ); p, values were then 
summed over k age classes. Mean individual weights at 
each age class were divided by 



k 

i=l 

to produce a standardized birth rate (6,), corrected to 
reflect a nongrowing population with stable age struc- 
ture, i.e.. 

Both age-structure and individual (wet) weight data 
were from the commercial and recreational catch of 
red snapper in the northern Gulf were provided by D. 
Nieland of Louisiana State University. Resulting life- 
history tables were used to calculate a correction factor 
(C) for overlapping generations by using 100 iterations 
of Equation 5 in Jorde and Ryman (1996). The value C 
can be defined as a correction term that is determined by 
the particular values of/, and 6, of the population under 
study. G, the mean generation length in years, was calcu- 
lated by using Equation 10 in Jorde and Ryman (1996). 
Values of C and G obtained for each locality were sub- 
sequently used to correct estimates of TV, by N^,^. = N^, x 
[C/G], where N^, is the pseudo-maximum-likelihood 
estimate of variance effective size obtained by follow- 
ing Wang (2001). C and G values, respectively, for the 
three localities were 10.1 and 6 (Texas), 12.1 and 6.1 
(Louisiana), and 10.5 and 6.8 (Alabama). 



Results 

Summary statistics (number of alleles, allelic richness, 
gene diversity; and results of tests of HW equilibrium) 
for each sample are given in Appendix Tables 1 and 
2. Number of alleles among all samples ranged from 
4 to 7 at Prs260 to 20-23 at P;-s248, and averaged 
(±SD) 11.67 ±5.15 (1995 cohort) and 11.30 ±5.02 (1997 
cohort). Allelic richness generally paralleled the number 
of alleles. Gene diversity among all samples ranged 
between 0.178-0.238 (Lca20) and 0.898-0.915 (Prs257), 
and averaged (±SD) 0.597 ±0.224 (1995 cohort) and 
0.602 ±0.217 (1997 cohort). No significant difference 
in allelic richness (P=0.35) or gene diversity (P=0.07) 
was detected. 

Four of 114 tests of conformity to Hardy- Weinberg 
equilibrium expectations were significant following 
Bonferroni correction. These included two tests in the 
1995 cohort {Prs275 in the Texas sample and Prsl37 
in the Alabama sample) and two tests in the 1997 co- 
hort (Lcn22 in the Texas sample and Prs229 in the 
Louisiana sample). F^^, values over all loci for all four 
samples ranged between 0.008 and 0.029 (Appendix 
Tables 1 and 2). A total of 21 of 1026 (pairwise) tests 
of genotypic disequilibrium were significant (P<0.05) 
after Bonferroni correction. All 21 involved different 
pairs of loci (i.e., only one out of six possible tests for a 
given pair combination was significant) except for Lca64 
and Prs328 in the 1995 cohort from Alabama and the 
1997 cohort from Texas, and Lca64 and P7-s248 in both 
cohorts sampled from Texas. 



Saillant and Gold: Population structure and variance effective size of Lul/anus campechanus in the northern Gulf of Mexico 



139 







Table r 




Probability of genie and genotypic homogeneity at 19 
microsatellites among spatial and temporal sample.s of 
red snapper ^Liitjaiuis campechanus} sampled from the 
northern Gulf of Mexico. Probability values are based on 
exact tests; significance was assessed via a Markov-chain 
method (cf text). Boldface indicates significance following 
sequential Bonferroni correction. 


Microsatellite 




Genie 
homogeneity 


Genotypic 
homogeneity 


Lfa20 




0.018 


0.031 


Lca22 




0.001 


0.005 


Lca43 




0.044 


0.065 


Lca64 




0.109 


0.127 


Lca91 




0.000 


0.001 


Lea 107 




0.416 


0.292 


Prs55 




0.179 


0.212 


Prsl37 




0.706 


0.788 


Prs221 




0.931 


0.930 


Prs229 




0.024 


0.047 


Prs240 




0.000 


0.000 


Prs248 




0.101 


0.154 


Prs257 




0.053 


0.085 


Prs260 




0.098 


0.111 


Prs275 




0.819 


0.893 


Prs282 




0.050 


0.063 


Prs 303 




0.014 


0.002 


Prs328 




0.184 


0.289 


Prs333 




0.137 


0.113 


Overall 




0.000 


0.000 



Significant heterogeneity (exact tests) among samples 
in either allele or genotype distributions (or both) was 
found overall and, after Bonferroni correction, at four 
individual microsatellites (Table 1). Pairwise compari- 
sons (exact tests) of allele and genotype distributions 
between samples paralleled one another and revealed 
that almost all of the genetic heterogeneity was due 
to the 1995 cohort from Texas and the 1997 cohort 
from Alabama (Table 2). This result indicated that the 
observed genetic heterogeneity is more temporal (be- 
tween cohorts) than spatial (among localities). Temporal 
rather than spatial heterogeneity also was indicated by 
the nonsignificant exact tests among localities sampled 
in 1997 and by the average Fgj values among localities 
(both cohorts) of less than 0.001. 

Results of assignment tests are given in Table 3. On 
average, 53% of the individuals were "assigned" (i.e., 
had the highest probability of belonging) to their origi- 
nal locality. This proportion was significantly higher 
(P<0.001) than that expected if multilocus genotypes 
were distributed randomly with respect to geographic 
location. However, the estimated probabilities of be- 
longing to all three localities were higher than 0.05 for 
96.4-99.8% of the individuals, indicating that none of 
the three localities could be rejected as a potential ori- 



Table 2 

Pairwise F^-j. values (upper diagonal) and probability 
that F^j.=0 (lower diagonal) for twelve samples (four 
cohortsxthree localities) of red snapper iLutjanus cam- 
pechanus) from the northern Gulf of Mexico. Boldface 
indicates significance following sequential Bonferroni 
correction. TX=Texas; LA=Louisiana; AL-Alabama. 

TX95 LA 95 AL 95 TX 97 LA 97 AL 97 



TX95 
LA 95 
AL95 
TX97 
LA 97 
AL97 



0.002 
0.001 
0.000 

0.036 
0.000 



0.0012 

0.031 
0.013 
0.756 
0.000 



0.0010 
0.0006 

0.001 

0.737 
0.000 



0.0013 
0.0007 
0.0008 

0.079 
0.045 



0.0010 0.0020 
-0.0002 0.0009 
-0.0001 0.0015 

0.0005 0.0002 
— 0.0006 

0.073 — 





Table 3 


Results of assignment tests (percentage of fish assigned 
to a given locality) based on red snapper (Lutjaiius 
campechanus) sampled from three geographic localities in 
the northern Gulf of Mexico. TX=Texas; LA=Louisiana; 
AL=Alabama. 


Origin of sample 
(sample size) 


Highest likelihood of belonging to 


TX LA AL 


TX(414) 
LA (558) 
AL(651) 


53.0 24.0 22.0 
21.0 53.0 26.0 
21.0 26.0 53.0 



Table 4 

Estimates of variance effective size (N^,y) and 95*^ confi- 
dence intervals for red snapper iLutjaniis campechanus) 
sampled at three geographic localities in the northern 
Gulf of Mexico. Estimates were generated using the 
pseudo-maximum-likelihood method of Wang (20011. Val- 
ues are corrected for overlapping generations, following 
Jorde and Ryman (1995). 



Locality 



Af., 



95% lov 



95<>r high 



Texas 


1098 


Louisiana 


>75.000 


Alabama 


1235 



6.52 
3275 

777 



2706 

>75,000 

2515 



gin. In addition, for four individuals from the Alabama 
sample (0.6%), all three localities were excluded as the 
potential origin. 

The pseudo-maximum-likelihood (temporal-method) 
estimates of variance effective size (N^y), corrected for 
overlapping generations, and their 95% confidence in- 
tervals for all three localities are shown in Table 4. 



140 



Fishery Bulletin 104(1) 



Estimates for the samples from Texas (A/py=1098) and 
Alabama (N^,y=1235) were essentially the same, falling 
well v/ithin the 95% confidence intervals of one another. 
An exact, maximum-likelihood estimate could not be 
generated for the sample from Louisiana because the 
value of N^.y with highest likelihood was >75,000 and 
the likelihood of higher values of N^,y could not be com- 
puted. This estimate is more than an order of magni- 
tude greater than the N ^• estimates for the other two 
localities and is significantly higher than those based 
on 95% confidence intervals. 



Discussion 

Genetic population structure 

Results obtained from pairwise exact tests indicated 
that the majority of genetic differentiation detected 
among the twelve spatial-temporal samples of red snap- 
per was due to allele and genotype distributions in 
the 1995 cohort sampled from the northwestern Gulf 
(Texas) and in the 1997 cohort sampled from the north- 
eastern Gulf (Alabama). In addition, exact tests among 
cohorts sampled in 1997 were nonsignificant and F^y. 
values among localities (both cohorts) averaged less than 
0.001. These results indicate that the genetic differences 
observed in the present study are temporal (between 
cohorts within localities I and not spatial (among locali- 
ties). A "hint" of spatial differentiation was suggested by 
assignment tests. A total of 53% offish were reclassified 
(assigned) to their original locality, a proportion that 
differed significantly from that expected if genotypes 
were distributed randomly among localities. However, 
for 98% of the fish, none of the three localities could be 
unequivocally excluded as the locality of origin. 

The above results are in general agreement with oth- 
er, genetics-based studies of red snapper in the northern 
Gulf in that little to no significant geographic hetero- 
geneity in genetic markers, ranging from allozymes to 
mtDNA to microsatellites, has been detected (John- 
son^; Gold et al., 1997; Garber et al., 2004; Gold et 
al., 2001b). The one exception was a study by Bortone 
and Chapman* where significant heterogeneity in both 
temporal and spatial restriction-fragment patterns of 
the mitochondrially-encoded 16S ribosomal (r)RNA 
gene was reported. Bortone and Chapman'* suggested 
that the observed genetic heterogeneity likely stemmed 
from nonrandom sampling where individuals related 
by descent had remained in close spatial proximity to 
one another. In general, the "consensus" inference has 
been that gene flow among present-day red snapper 



in the northern Gulf is sufficient to offset divergence 
by genetic drift of the (presumed) selectively neutral 
genetic markers assayed. Such gene flow could involve 
movement of adults (Patterson et al., 2001), hydrody- 
namic transport of pelagic eggs and larvae (Goodyear-'), 
or both. 

The foregoing notwithstanding, there are a number of 
caveats (discussed in Pruett et al., 2005) to the infer- 
ence that significant gene flow occurs among present- 
day red snapper in the northern Gulf Briefly, tag-and- 
recapture and ultrasonic-tracking studies (Fable, 1980; 
Szedlmayer and Shipp. 1994; Szedlmayer, 1997) have 
indicated that adult red snapper are largely sedentary 
and nonmigratory. Significant movement of adults in 
the northeastern Gulf was reported by Patterson et 
al. (2001), but movement per se was mostly unidirec- 
tional (west to east) and the average distance covered 
in roughly a year was only -30 kilometers. Movement 
of (pelagic) red snapper eggs and larvae likely occurs, 
but neither egg nor larval type nor length of larval life 
are effective predictors of gene flow in marine fishes 
(Shulman and Bermingham, 1995) and larval exchange 
rates of marine species generally appear overestimated 
(Cowen et al., 2000). In addition, regardless of the life- 
history stage at which gene flow might occur in red 
snapper, movement across the continental shelf should 
be more-or-less linear and would be expected to fol- 
low a pattern of isolation by distance where fish from 
proximal localities are more similar genetically than 
fish from more distal ones. However, the correlations 
between genetic and geographic distance expected from 
isolation by distance have not been found (Gold et al., 
1997; 2001b; this article). Finally, salient differences 
in geologic structure, habitat structure, and ecological 
conditions (Rezak et al., 1985; Gallaway et al., 1998), 
significant differences in salinity due to freshwater 
outflow from river systems in the northcentral Gulf 
(Morey et al., 2003), and the present-day occurrence 
during the summer months of a major hypoxic zone 
that extends out along the continental shelf from the 
Mississippi Delta westward (Rabalais et al.'°; Ferber, 
2001) potentially could serve as barriers to movement 
and gene flow. 

Despite these caveats, the bulk of the genetics data 
has indicated essentially no difference among present- 
day red snapper sampled across the northern Gulf This 
is consistent with the unit stock hypothesis and with 
the inference that observed genetic homogeneity is due 
to substantial gene flow. However, it is important to 



' Bortone, S. A., and R. W. Chapman. 1995. Identification 
of stock structure and recruitment patterns for the red snap- 
per, Lutjanus campechanus. in the Gulf of Mexico. Final 
report for Marfin Program Grant Number NA17FF0379-03, 
39 p. Southeast Regional Office, National Marine Fisher- 
ies Service, 263 13"^ Avenue South, Saint Petersburg, FL 
33701. 



•' Goodyear, C. P. 1995. Red snapper stocks in U.S. waters 
of the Gulf of Me.xico, 171 p. National Marine Fisheries 
Service. SE Fisheries Center, Miami Laboratory, CRD 95/96- 
05, 75 Virginia Beach Drive, Miami, FL 33149-1099. 

i"Rabalais, N. N., R. E. Turner, D. Justic, Q. Dortch, and W. 
J. Wiseman. 1999. Characterization of hypoxia: topic I 
report for the integrated assessment on hypoxia in the Gulf 
of Mexico. NOAA Coastal Ocean Program Decision Analy- 
sis Series No. 15, 203 p. NOAA Coastal Ocean Program, 
1305 East-West Hwy,. N/SC12 SSMC4, Silver Spring, MD 
20910. 



Saillant and Gold Population structure and variance effective size of Lutianus campechonus in ttie norttiern Gulf of Mexico 



141 



note the following. First, it is possible that gene flow 
among present-day red snapper in the northern Gulf is 
limited but there has been insufficient time for semi- 
isolated lineages to completely sort into monophyletic 
assemblages. Pruett et al. (2005), on the basis of re- 
sults of nested-clade analysis of mtDNA haplotypes ob- 
tained from representative samples of the same cohorts 
(and localities) studied in the present study, hypoth- 
esized that semi-isolated assemblages of red snapper 
in the northern Gulf may exist over the short term, 
yet over the long term comprise a larger metapopu- 
lation tied together by periodic gene flow. Similarity 
in allele frequencies of genetic markers (such as used 
here and in previous studies of red snapper) presumed 
to be neutral to natural selection in theory could be 
maintained in such a metapopulation during periods 
when gene flow was limited or even absent. Second, all 
the genetic markers studied to date are presumed to 
be selectively neutral and to be affected primarily by 
the interaction(s) between gene flow and genetic drift. 
Genes affecting life-history and other traits that are in- 
fluenced by natural selection need not necessarily follow 
the same pattern(s), and geographic differences in adap- 
tively useful alleles at such genes can be maintained 
even in the face of substantial gene flow (Conover et 
al., 2005). It is thus not implausible that red snapper 
across the northern Gulf could differ in allele frequency 
at adaptively useful genes yet be homogeneous at selec- 
tively neutral ones. 

Contemporaneous effective size (N^^y) and present-day 
demographic dynamics 

Estimates of contemporaneous or variance effective 
size {N^y) for the Texas and Alabama localities (-1100) 
were essentially the same, but were at least an order of 
magnitude less than the N^,y estimate (>75,000) for the 
Louisiana locality. These estimates reflect differences 
in effective population size under the assumption that 
no immigration into a locality has occurred during the 
study interval, an assumption at odds with the general 
absence of allele-frequency heterogeneity among locali- 
ties as well as the low estimates of F>,.j. between pairs 
of samples. Short-term immigration (within the time 
interval of the study) could increase the variance in 
allele frequencies, thus resulting in an overestimate of 
A^^;. (Wang and Whitlock, 2003); whereas longer-term 
immigration at a (more-or-less) constant rate from a 
source population would have the opposite effect. The 
observed differences among localities could thus reflect 
differences in effective sizes, differences in patterns and 
intensity of immigration, or both. Temporal variation in 
allele frequencies also could occur if only a fraction of 
potential spawners at a locality actually contributed to 
recruitment and if such "temporal" subpopulations dif- 
fered in allele frequencies between years. Regardless, 
the differences in 7V,^• may indicate that different demo- 
graphic dynamics currently exist among localities. 

Wang and Whitlock (2003) recently extended previous 
maximum-likelihood methods to allow simultaneous es- 



timation of N^.y and m (rate of migration), provided data 
from multiple loci were available and all sources of im- 
migrants into a focal population were known. Because 
of the latter, we were able to generate estimates of N^.y 
and m only for the sample from the Louisiana locality 
(focal population), using the samples from the Texas 
and Alabama localities as source populations. Surpris- 
ingly, the estimate of N^,y for the Louisiana sample 
(4887, 95% confidence intervals of 1543 and 31,254) was 
at least -15 times smaller than the estimate based on 
no migration; m was estimated to be 0.0097 (95% confi- 
dence intervals of <0.001 and 0.0355). Clearly, more ex- 
tensive sampling across the northern Gulf is warranted 
to obtain estimates of N^.y and ;?; at other localities and 
to place this finding into perspective. 

Effective size (/Vp)/census size(AV) ratios 

Estimates of N^,y for all three sample localities were two 
or more orders of magnitude less than the current, pre- 
liminary estimates of adult census size (7.8-11.7 million) 
across the northern Gulf (Cowan-^; Porch^). Given that 
empirically derived N^JN ratios from a variety of verte- 
brates are 0.10-0.11 on average (Frankham, 1995), this 
result is somewhat surprising in that red snapper have a 
long reproductive life-span and overlapping generations 
(Wilson and Nieland, 2001), life-history features that 
are expected to increase N^./N by limiting variance in 
lifetime reproductive success among individuals (Jorde 
and Ryman, 1995; Waite and Parker, 1996). The issue 
is of importance in that census sizes of many commer- 
cially exploited marine fish populations are generally 
orders of magnitude larger than sizes where genetic 
resources might be lost (Franklin, 1980; Schultz and 
Lynch, 1997). However, species or populations with 
exceedingly small N^JN ratios potentially could be in 
danger of losing genetic resources, resulting in reduced 
adaptation and population productivity (Hauser et al., 
2002). In addition, low N^JN ratios may explain in part 
why there often is a poor relationship between spawn- 
ing stock size and recruitment (Hauser et al., 2002). 
To date, N^,/N ratios smaller than 10" ' have been found 
for four other exploited marine fish species (Hauser et 
al., 2002; Turner et al., 2002; Hutchinson et al., 2003; 
Gomez-Uchida and Banks^M. 

Factors that theoretically can lower genetic effec- 
tive size with respect to census size include fluctuating 
adult number and year-class strength (Hedgecock, 1994; 
Vucetich et al., 1997), and variance in reproductive suc- 
cess. The latter can arise from biased sex ratio, high 
variance in male or female reproductive success, vari- 
ance in productivity among habitats, or any combina- 
tion of these factors (Nunney, 1996, 1999; Whitlock and 
Barton, 1997). Virtually any of these factors could lower 
N./N ratios in red snapper. Biased sex-ratio, however. 



" Gomez-Uchida. D., and M. A. Bank.s. 2003. Oregon State 
Univ., Hatfield Marine Science Center, Department of Fish- 
eries and Wildlife, 20.30 SE Marine Sci. Dr. Newport, OR 
97365. 



142 



Fishery Bulletin 104(1) 



seems unlikely, because the ratio of males and females 
across three years of red snapper catch data was 0.97 
and did not differ significantly from unity (Nieland^-). 
Variation in population number and in year-class 
strength, alternatively, seems likely, given the annual 
differences in commercial and recreational landings and 
the annual differences in abundance of age-0 and age- 
1 red snapper, respectively (Schirripa and Legault^'*). 
Variance in reproductive success is far more difficult to 
assess but can include mating systems (Nunney, 1993) 
that lead to differences in reproductive success between 
males and females, and a "sweepstakes" process (Hedge- 
cock, 1994) where size-dependent fecundity, combined 
with random but family-specific early mortality (Hauser 
et al., 2002), leads to a large variance in the number of 
(surviving) offspring per parent. The latter could be ef- 
fected in red snapper by nonrandom removal of related 
subadults or juveniles either by localized overfishing 
or by shrimp trawling. Finally, variance in productiv- 
ity among habitats across the northern Gulf can be 
inferred from subregional differences in red snapper 
growth rates (Fischer et al., 2004) and from subregional 
ecological differences (Gallaway et al., 1998) that dis- 
tinguish the northeastern Gulf from the northwestern 
Gulf. Future ecological and behavioral studies to gen- 
erate estimates of variance in individual reproductive 
success or variation in productivity among localities are 
clearly warranted. 

Demographic stocks 

The differences in variance effective size (A''^,i) among the 
geographic samples of red snapper indicate present-day 
differences in demographic dynamics that may include 
the number of individuals that produce surviving off- 
spring, and hence by inference, census size. The factors, 
ecological or otherwise, promoting these demographic 
differences are difficult to assess but likely relate in 
some way to variation in food availability, habitat qual- 
ity, or mortality (or a combination of all three factors). 
Accordingly, one might expect one or more of these 
factors to differ among the sample localities, given the 
differences in variance effective size among localities. 
In addition, one might expect other demographic param- 
eters to differ as well. 

Our study was part of a larger, multidisciplinary proj- 
ect that involved studies of age-and-growth and repro- 
duction of red snapper at the three localities. The age- 
and-growth studies of Fischer et al. (2004) documented 
that fork length, total weight, and age-frequency dis- 
tributions differed significantly among localities. Red 
snapper sampled at the Texas locality were significantly 



'- Nieland, D. 2005. Personal commun. Coastal Fisheries 
Institute, Louisiana State University, Baton Rouge, LA 
70803-7503. 

" Schirripa, M. J., and C. M. Legault. 1999. Status of the 
red snapper in U.S. waters of the Gulf of Mexico. Report 
SFD-99/00-75, 86 p. Southeast Fisheries Science Center, 
75 Virginia Beach Drive, Miami, FL 33149-1099. 



smaller at age and reached smaller maximum size than 
did red snapper sampled at the Louisiana and Alabama 
localities; fish sampled at the latter two localities did 
not differ in size-at-age or maximum size. There also 
was a significantly higher proportion of smaller, young- 
er fish at the Texas locality than at the other two. The 
studies of reproductive capacity (Woods et al., 2003) 
involved only fish sampled from the Louisiana and 
Alabama localities but revealed that females sampled 
from the Alabama locality reached sexual maturity at a 
smaller size and younger age than did females sampled 
from the Louisiana locality. The differences in growth 
rate are likely a function in part of the more produc- 
tive, nutrient-rich waters found at the Louisiana and 
Alabama localities and caused by the plume produced 
from the Mississippi River (Fischer et al., 2004), a hy- 
pothesis reinforced by the observation (Grimes, 2001) 
that between 70-80'7f of fishery landings in the north- 
ern Gulf of Mexico come from waters surrounding the 
Mississippi River delta. The differences in female age 
and size at maturity, alternatively, are thought to indi- 
cate a stressed population and to reflect a compensatory 
response to growth overfishing or declining population 
size, or a response to both (Trippel, 1995; Woods et al., 
2003). Collectively, the life-history differences and the 
differences in genetic-based estimates of effective size 
strongly suggest that red snapper at the three localities 
represent three different, demographic stocks. 

A critical issue is whether the demographic differ- 
ences in life history observed among red snapper in the 
northern Gulf are genetic or phenotypic (environmen- 
tally induced) in origin. Most discussions of stock struc- 
ture in commercially exploited marine fishes involve 
an explicit genetics component (Gold et al., 2001a), and 
typically, the absence of genetic heterogeneity within a 
fishery leads to management planning for a single unit 
stock. However, life-history traits can change rapidly in 
response to environmental pressures (e.g., size-selective 
fishing), and it has been hypothesized that the pool of 
genotypes that code for life-history traits is a highly 
dynamic property of populations, and moreover, that lo- 
cal adaptation(s) differentiating populations can evolve 
even in the presence of extensive gene flow (Conover et 
al., 2005). Thus, demographically different stocks could 
differ genetically, but not necessarily in selectively neu- 
tral markers that respond primarily to the interaction 
between gene flow and genetic drift. The issue also is of 
importance to management planning because phenotypi- 
cally plastic responses due to environmental differences 
generally can be reversed fairly quickly, whereas genetic 
responses are typically much slower (Hutchings, 2004; 
Conover et al., 2005). 

The geographic differences in red snapper in growth 
rates and shifts in timing of female maturity in all 
likelihood are due to a mix of genetic and environmen- 
tal factors, as are most life-history traits in a variety 
of animal species, including fishes (Mousseau and Roff, 
1987; Conover and Munch, 2002). A significant genetic 
component to growth rate is well documented in a va- 
riety of fishes under aquaculture (Dunham et al., 2001) 



Saillant and Gold: Population structure and variance effective size of Lut/anus compechanus in tfie nortfiern Gulf of Mexico 



143 



and genotypes for smaller size and younger age at ma- 
turity clearly exist (Gjerde, 1984; Tipping, 1991; Trip- 
pel, 1995). These considerations indicate that a genetic 
component to growth rate and age at maturity may 
exist in red snapper, and if so, stock-structure consid- 
erations solely on the basis of homogeneity in selectively 
neutral genetic markers may not be warranted. 



Acknowledgments 

We thank A. Fisher for providing individual age data, 
D. Nieland for providing life-history data, T. Turner 
for extensive assistance with statistical analysis, the 
Texas A&M supercomputing facility staff for help with 
parallel processing of MLNE, and S.C. Bradfield, C. 
Burridge, and L. Richardson for technical assistance. We 
also thank W. Patterson and T. Turner for helpful com- 
ments on early drafts of the manuscript, and J. Cowan 
and C. Porch for providing estimates of red snapper 
census size. Work was supported by the Gulf and South 
Atlantic Fisheries Development Foundation (Grant 70- 
04-20000/11824), by the Marfin Program of the U.S. 
Department of Commerce (Grant NA87-FF-0426I, and 
by the Texas Agricultural Experiment Station (Project 
H-6703). This article is number 46 in the series "Genet- 
ics Studies in Marine Fishes." 



Literature cited 

Allendorf, F. W., and R. S. Waples. 

1996. Conservation and genetics of salmonid fishes. In 
Conservation genetics: case histories from nature (J. C. 
Avise, and J. L. Hamrick. eds.l p. 238-280. Chapman 
and Hall. New York, NY. 
Carvalho, G. R.. and L. Hauser. 

1995. Molecular genetics and the stock concept in 
fisheries. In Molecular genetics in fisheries (G. R. 
Carvalho, and T. J. Pitcher eds.l p. 55-80. Chapman 
and Hall, London. 
Conover, D. O., S. A. Arnott, M. R. Walsh, and S. B. Munch. 

2005. Darwinian fishery science: lessons from the Atlan- 
tic silverside. Can. J. Fish. Aquat. Sci. 62:730-737. 
Conover, D. O., and S. B. Munch. 

2002. Sustaining fisheries yields over evolutionary time 
scales. Science 297:94-96. 
Cowen, R. K., K. M. M. Lwiza, S. Sponaugle, C. B. Paris, and 
D. B. Olson. 

2000. Connectivity of marine populations: open or closed? 
Science 287:857-859. 

Dunham, R. A., K. Majumdar, E. Hallerman, D. Bartley, 
G. Mair, G. Hulata, Z. Liu, N. Pongthana. J. Bakos, D. Penman, 
M. Gupta, P. Rothlisberg, and G. Hoerstgen-Schwark. 

2001. Review of the status of aquaculture genetics. In 
Aquaculture in the third millennium (R. P. Subasinghe, 
P. Bueno, M. J. Phillips, C. Hough, S. E. McGladdery, 
and J. R. Arthur, eds.) p. 137-166. FAG, Rome. 

Fable, W. A. 

1980. Tagging studies of red snapper iLiitjanus campecha- 
nus) and vermilion snapper [Rhomboplites auroru- 
bens) off the south Texas coast. Contrib. Mar. Sci. 
23:115-121. 



Ferber, D. 

2001. Keeping the Stygian waters at bay. Science 291: 
968-973. 
Fischer, A. J., M. S. Baker Jr., and C. A. Wilson. 

2004. Red snapper, Lutjanus campechanus, demographic 
structure in the northern Gulf of Mexico based on spatial 
patterns in growth rates and morphometries. Fish. 
Bull. 102:593-603. 
Frankham, R. 

1995. Effective population-size adult-population size 
ratios in wildlife: a review. Genet. Res. 66:95-107. 
Franklin, I. R. 

1980. Evolutionary change in small populations. Chap- 
ter 8 m Conservation biology: an evolutionary-eco- 
logical perspective (M. Soule, and B. Wilcox, eds.) p. 
135-149. Sinauer, Sunderland, MA. 
Gallaway, B. J., J. G. Cole, R. Meyer, and P. Roscigno. 

1999. Delineation of essential habitat for juvenile red 
snapper in the northwestern Gulf of Mexico. Trans. 
Am. Fish. Soc. 128:713-726. 
Gallaway, B. J., M. Longnecker, J. G. Cole, and R. M. Meyer. 

1998. Estimates of shrimp trawl bycatch of red snapper, 
Lutjanus campechanus, in the Gulf of Mexico. In Fish- 
ery stock assessment models (F. Funk F, T. J. Quinn II, 
J. Heifetz, J. N. lanelli, J. E. Powers, J. F. Schweigert, P. 
J. Sullivan, and C.-I. Zhang, eds.) p. 817-839. Alaska 
Sea Grant College Program, No. AK-SG-98-01. Fair- 
banks, AK. 
Garber, A. F, M. D. Tringali. and K. C. Stuck. 

2004. Population structure and variation in red snapper 
{Lutjanus campechanus) from the Gulf of Mexico and 
Atlantic Coast of Florida as determined from mitochon- 
drial DNA control region sequence. Mar. Biotechnol. 
6:175-185. 
Gjerde, B. 

1984. Response to individual selection for age at sexual 
maturity in Atlantic salmon. Aquaculture 38:229- 
240. 
Gold, J. R., C. P Burridge. and T. F. Turner. 

2001a. A modified stepping-stone model of population 
structure in red drum, Sciaenops ocellatus (Sciae- 
nidae) from the northern Gulf of Mexico. Genetica 
111:305-317. 
Gold, J. R., E. Pak, and L. R. Richardson. 

2001b. Microsatellite variation among red snapper [Lut- 
janus campechanus) from the Gulf of Mexico. Mar. 
Biotechnol. 3:293-304. 
Gold, J. R., F. Sun, and L. R. Richardson. 

1997. Population structure of red snapper from the Gulf 
of Mexico as inferred from the analysis of mitochondrial 
DNA. Trans. Am. Fish. Soc. 126:386-396. 

Goudet, J. 

1995. F-STAT (Version 1.2): A computer program to cal- 
culate F-Statistics. J. Hered. 86:485-486. 
Grimes, C. B. 

2001. Fishery production and the Mississippi River 
discharge. Fisheries 26:17-26. 
Guo, S., and E. Thompson. 

1992. Performing the exact test of Hardy-Weinberg propor- 
tion from multiple alleles. Biometrics 48: 361-372 
Hartl, D. L., and A. G. Clark. 

1989. Principles of population genetics, 2"'' edition, 669 
p. Sinauer Assoc. Inc.. Sunderland, MA. 
Hauser, L., and R. D. Ward. 

1998. Population identification in pelagic fish: the limits 
of molecular markers. In Advances in molecular ecol- 



144 



Fishery Bulletin 104(1) 



ogy (G. R. Carvalho, ed.), p. 2191-2242. lOS Press, 
Amsterdam. 
Hause-, L., G. J. Adcock, P. J. Smith. J. H. B. Ramirez, G. R. 
Carvalho. 

2002. Loss of microsatellite diversity and low effective 
population size in an overexploited population of New 
Zealand snapper (Pagrus auratus). Proc. Natl. Acad. 
Sci. USA. 99:11742-11747. 

Hedgecoek, D. 

1994. Does variance in reproductive success limit effective 
population sizes of marine organisms? In Genetics and 
evolution of aquatic organisms (A. R. Beaumont, ed.) p. 
122-134. Chapman and Hall, New York, NY. 
Higgins, K., and M. Lynch. 

2001. Metapopulation extinction caused by mutation 
accumulation. Proc. Natl. Acad. Sci. USA 98: 2928- 
293.3. 
Hoese, H. D., and R. H. Moore. 

1977. Fishes of the Gulf of Mexico, 327 p. Texas A&M 
Univ. Press, College Station, TX. 
Hutchings. J. A. 

2004. Life history consequences of over-exploitation to 
population recovery in northwest Atlantic cod, Gadus 
morhua. Can. J. Fish. Aquat. Sci. 62:824-832. 
Hutchinson, W. F., C. van Gosterhout, S. L Rogers, and G. R. 
Carvalho. 

2003. Temporal analysis of archived samples indi- 
cates marked genetic changes in declining North Sea 
cod [Gadus morhua). Proc. R. Soc. London, B 270: 
2125-2132. 

Jorde, P. E., and N. Ryman. 

199.5. Temporal allele frequency change and estima- 
tion of effective size in populations with overlapping 
generations. Genetics 139:1077-1090. 
1996. Demographic genetics of brown trout iSalmo 
trutta) and estimates of effective population size 
from temporal change of allele frequencies. Genetics 
143:1369-1381. 
Morey, S. L., W. L. Schroeder, J. J. O'Brien, and J. Zavala- 
Hidalgo. 

2003. The annual cycle of riverine influence in the east- 
ern Gulf of Mexico. Geophys. Res. Letters 30:OCE 
5.1-5.4. 

Mousseau, T. A., and D. A. Roff 

1987. Natural selection and the heritability of fitness 
components. Heredity 59:181-198. 
Nunney, L. 

1993. The influence of mating system and overlapping 
generations on effective population size. Evolution 
47:1329-1341. 
1996. Influence of variation in female fecundity on effec- 
tive population size. Biol. J. Linn. Soc. 59: 411-425. 
1999. The effective size of a hierarchically structured 
population. Evolution 53:1-10. 
Paetkau, D., R. Slade, M. Burden, and A. Estoup. 

2004. Direct, real-time estimation of migration rate using 
assignment methods: a simulation-based exploration of 
accuracy and power. Mol. Ecol. 13:55-65. 

Patterson, W. F., J. C. Watterson, R. L. Shipp, and J. H. Cowan Jr. 

2001. Movement of tagged red snapper in the northern 

Gulf of Mexico. Trans. Am. Fish. Soc. 130:533-545. 

Piry, S., A. Alapetite, J. M. Cornuet, D. Paetkau, L. Baudouin, 

and A. Estoup. 

2005. GeneClass2: a software for genetic assignment 
and first generation migrants detection. J. Hered. 
95:536-539. 



Pruett, C. L., E. Saillant, and J. R. Gold. 

2005. Historical population demography of red snap- 
per iLutjanus campechanus} from the northern Gulf of 
Mexico based on analysis of sequences of mitochondrial 
DNA Mar. Biol. 147:593-602. 
Rannala, B., and J. L. Mountain. 

1997. Detecting immigration by using multilocus 
genotypes. Proc. Nat. Acad. Sci. USA 94:9197-9201. 
Raymond, M., and F. Rousset. 

1995. Genep(ip (version 1.2): a population genetics soft- 
ware for exact test and Ecumenicism. J. Hered. 86: 
248-249. 
Rezak, R., T. J. Bright, and D. W. McGrail. 

1985. Reefs and banks of the northwestern Gulf of Mexico, 
259 p. John Wiley and Sons, New York, NY. 
Rice, W. R. 

1989. Analyzing tables of statistical tests. Evolution 
43:223-225. 
Schultz, S. T., and M. Lynch. 

1997. Mutation and extinction: the role of variable muta- 
tional effects, synergistic epistasis, beneficial mutations, 
and degree of outcrossing. Evolution 51:1363-1371. 
Shulman, M. J., and E. Bermingham. 

1995. Early life histories, ocean currents, and the pop- 
ulation genetics of Caribbean reef fishes. Evolution 
49:897-910. 
Szedlmayer, S. T. 

1997. LTltrasonic telemetry of red snapper, Lutjanus 
campechanus, at artificial reef sites in the northeast 
Gulf of Mexico. Copeia 1997:846-850. 
Szedlmayer, S. T., and R. L. Shipp. 

1994. Movement and growth of red snapper, Lutjanus 
campechanus, from an artificial reef area in the north- 
eastern Gulf of Mexico. Bull. Mar. Sci. 55:887-896. 

Tipping, J. M. 

1991. Heritability of age at maturity. North Am. J. 
Fish. Manag. 11:105-108. 
Trippel, E. A. 

1995. Age at maturity as a stress indicator of fish- 
eries. BioScience 45:759-771. 

Turner, T. F., J. P. Wares, and J. R. Gold. 

2002. Genetic effective size is three orders of magnitude 
smaller than adult census size in an abundant, estua- 
rine-dependent marine fish tSciaenops ocellatus). Ge- 
netics 162:1329-1339. 

Vucetich, J. A.. T. A. Waite. and L. Nunney. 

1997. Fluctuating population size and the ratio of effective 
to census population size. Evolution 51: 2017-2021. 
Waite, T. A., and P. G. Parker. 

1996. Dimensionless life histories and effective popula- 
tion size. Conserv. Biol. 10:1456-1462. 

Wang, J. 

2001. A pseudo-likelihood method for estimating effective 

population size from temporally spaced samples. Genet. 

Res. 78:243-257. 
Wang, J., and M. C. Whitlock. 

2003. Estimating effective population size and migration 
rates from genetic samples over space and time. Ge- 
netics 163:429-446. 

Waples, R. S. 

1989. A generalized approach for estimating effec- 
tive population size from temporal changes in allele 
frequency. Genetics 121:379-391. 
Weir, B. S., and C. C. Cockerham. 

1984. Estimating F-statistics for the analysis of popula- 
tion structure. Evolution 38:1358-1370. 



Saillant and Gold: Population structure and variance effective size of Lutianus campechanus in the northern Gulf of Mexico 



145 



Whitlock, M. C, and N. H. Barton. 

1997. The effective size of a subdivided population. Ge- 
netics 146:427-441. 
Wilson. C. A., and D. L. Nieland. 

2001. Age and growth of red snapper, Lutjanus 
campechanus, from the northern Gulf of Mexico off 
Louisiana. Fish. Bull. 99:65.3-664. 



Woods, M. K., A. J. Fischer, J. H. Cowan, Jr., and D.L. Nieland. 
2003. Size and age of female red snapper Lutjanux 
campechanus in the northern Gulf of Mexico. Proc. 
Gulf Caribb. Fish. Inst. 54:526-537. 



Appendix Table 1 

Summary statistics at 19 nuclear-encoded microsatellite loci for the 1995 cohort of red snapper (Lutjanus campechanus} sampled 
at three localities in the northern Gulf of Mexico, n is sample size, no. of A is the number of alleles, A^ is allelic richness, H^ is 
gene diversity (expected heterozygosity), Pfj^y is probability of conforming to expected Hardy- Weinberg genotypic proportions, 
and Fjg is an inbreeding coefficient measured as Weir and Cockerham's ( 1984 ) /; Boldface indicates significant departures from 
HW equilibrium following (sequential) Bonferroni correction. 


Locus 


Texas 


Louisiana 


Alabama 


Locus 


Texas 




Louisiana 


Alabama 


Lca20 








Pr.s240 










n 


199 


286 


373 


n 


140 




276 


372 


No. of A 


5 


5 


6 


No. of A 


20 




21 


23 


Ar 


3.39 


3.37 


3.00 


An 


15.84 




14.94 


14.96 


He 


0.170 


0.215 


0.172 


He 


0.917 




0.901 


0.885 


^HW 


0.088 


0.816 


0.007 


^HW 


0.010 




0.129 


0.096 


F,s 


0.053 


-0.009 


0.097 


F,s 


0.065 




0.079 


-0.012 


Lca22 








Prs248 










n 


198 


281 


376 


n 


195 




285 


372 


No. of A 


14 


17 


14 


No. of A 


21 




21 


23 


^H 


8.45 


9.62 


8.92 


An 


13.40 




12.61 


12.88 


He 


0.686 


0.741 


. 0.712 


He 


0.889 




0.851 


0.874 


P 


0.176 


0.317 


0.013 


^HW 


0.468 




0.393 


0.291 


F,s 


0.013 


0.002 


0.055 


P,s 


-0.010 




0.006 


0.047 


Lca43 








Prs257 










n 


202 


275 


340 


n 


165 




273 


269 


No. of A 


10 


11 


9 


No. of A 


16 




16 


16 


An 


6.41 


5.98 


6.19 


An 


12.95 




12.57 


12.50 


H, 


0.535 


0.553 


0.530 


He 


0.903 




0.909 


0.904 


^HW 


0.585 


0.763 


0.669 


^HW 


0.3.50 




0.140 


0.392 


F,, 


-0.028 


-0.006 


0.017 


Fis 


0.021 




0.013 


0.005 


Lea 64 








Prs260 










;? 


197 


286 


377 


n 


189 




283 


376 


No. of A 


12 


14 


13 


No. of A 


4 




5 


7 


An 


7.19 


7.54 


6.97 


A„ 


3.45 




3.39 


3.50 


H, 


0.777 


0.778 


0.764 


He 


0.361 




0.390 


0.339 


P 


0.239 


0.684 


0.028 


Phw 


0.507 




0.285 


0.185 


F,s 


0.027 


-0.025 


0.014 


Fis 


-0.011 




-0.005 


-0.045 
continued 



146 



Fishery Bulletin 104(1) 









Appendix 


Table 1 (continued) 








Locus 


Texas 


Louisiana 


Alabama 


Locus 


Texas 


Louisiana 


Alabama 


Lco91 








Prs275 








n 


201 


285 


375 


/) 


199 


286 


374 


No. of A 


6 


7 


7 


No. of A 


9 


10 


9 


Ar 


4.49 


4.16 


4.43 


Ar 


5.46 


4.91 


5.25 


He 


0.608 


0.559 


0.580 


He 


0.635 


0.595 


0.590 


^HW 


0.002 


0.957 


0.162 


^HW 


0.000 


0.604 


0.183 


F,s 


0.002 


-0.066 


-0.021 


F,s 


0.105 


-0.014 


0.011 


Prs 55 








Prs303 








n 


184 


277 


377 


n 


200 


285 


374 


No. of A 


8 


7 


9 


No. of A 


7 


13 


11 


Ar 


3.30 


3.82 


3.60 


Ar 


5.02 


5.77 


5.29 


He 


0.158 


0.228 


0.209 


He 


0.365 


0.416 


0.375 


^HW 


0.353 


0.326 


0.102 


^HW 


0.785 


0.559 


0.007 


F,s 


-0.030 


0.001 


0.073 


F,s 


-0.029 


-0.012 


-0.026 


Lea 107 








Prs282 








;; 


189 


286 


375 


n 


202 


285 


377 


No. of A 


11 


12 


11 


No. of A 


14 


14 


14 


Ar 


8.67 


8.59 


7.90 


Ar 


8.57 


8.62 


8.06 


He 


0.809 


0.806 


0.796 


He 


0.664 


0.669 


0.623 


^HW 


0.871 


0.451 


0.776 


"hw 


0.311 


0.039 


0.066 


F,s 


0.013 


0.006 


-0.031 


F,s 


-0.006 


0.072 


-0.035 


Prsl37 








Prs328 








n 


201 


286 


376 


n 


200 


286 


377 


No. of A 


13 


13 


17 


No. of A 


6 


8 


6 


Ar 


7.83 


7.92 


8.33 


Ar 


3.70 


4.06 


3.53 


He 


0.706 


0.700 


0.711 


He 


0.555 


0.557 


0.557 


^HW 


0.103 


0.331 


0.000 


^HW 


0.008 


0.002 


0.034 


F,s 


0.049 


0.071 


0.125 


F„ 


0.072 


-0.086 


0.020 


Prs221 








Prs333 








n 


197 


282 


376 


n 


202 


283 


371 


No. of A 


16 


20 


19 


No. of A 


8 


6 


8 


Ar 


9.78 


10.26 


9.73 


Ar 


4.22 


3.98 


4.70 


He 


0.791 


0.802 


0.792 


He 


0.288 


0.294 


0.371 


'l]\V 


0.043 


0.037 


0.102 


■P//IV 


0.008 


0.638 


0.002 


F,s 


0.018 


0.050 


0.053 


F,s 


0.055 


0.013 


0.128 


Prs229 
















n 


201 


285 


376 










No. of A 


8 


7 


8 










Ar 


5.96 


5.44 


5.39 










He 


0.470 


0.508 


0.486 










P 

' HV, 


0.689 


0.392 


0.782 










F„ 


0.038 


0.088 


0.005 











Saillant and Gold; Population structure and variance effective size of Lut/anus campechanus in ttie northern Gulf of Mexico 



147 



Appendix Table 2 

Summary statistics at 19 nuclear-encoded microsatellite loci tor tfie 1997 cohort of red snapper iLutjanus campechanus) sampled 
at three localities in the northern Gulf of Mexico, ii is sample size. No. of A is number of alleles, A,j is allelic richness, Hg is gene 
diversity lexpected heterozygosity). P,,u is probability of conforming to expected Hardy-Weinberg genotypic proportions, and 
F,j; is an inbreeding coefficient measured as Weir and Cockerhani's 1 1984 1 f. Boldface indicates significant departures from HW 
equilibrium following (sequential) Bonferroni correction. 


Locus 


Texas 


Louisiana 


Alabama 


Locus 


Texas 


Louisiana 


Alabama 


Lca20 








Prs240 








n 


211 


272 


269 


n 


188 


234 


240 


No. of A 


6 


6 


6 


No. of A 


20 


20 


22 


A« 


3.72 


3.35 


3.78 


^H 


14.22 


15.37 


14.34 


He 


0.238 


0.184 


0.206 


He 


0.897 


0.898 


0.885 


^HW 


0.121 


0.561 


0.009 


^HW 


0.001 


0.012 


0.317 


F,s 


0.042 


0.043 


0.082 


F,s 


0.092 


0.043 


-0.021 


Lca22 








Prs248 








n 


208 


244 


266 


n 


211 


271 


272 


No. of A 


16 


14 


15 


No. of A 


21 


22 


21 


An 


9.88 


9.36 


9.45 


Afl 


12.58 


12.91 


12.67 


liE 


0.769 


0.757 


0.771 


He 


0.872 


0.867 


0.882 


^HW 


0.000 


0.892 


0.086 


Phw 


0.176 


0.616 


0.334 


F,s 


0.106 


0.004 


-0.009 


F,s 


0.001 


0.030 


0.017 


Lca43 








Prs257 








n 


210 


272 


272 


n 


206 


266 


246 


No. of A 


8 


12 


11 


No. of A 


17 


17 


18 


Ar 


6.22 


6.48 


6.19 


A« 


13.24 


12.86 


13.51 


He 


0.587 


0.536 


0.528 


He 


0.908 


0.898 


0.915 


^HW 


0.325 


0.669 


0.981 


^HW 


0.282 


0.113 


0.464 


F,s 


0.010 


-0.049 


-0.003 


Pis 


0.011 


0.008 


0.005 


Lca64 








Prs260 








n 


211 


271 


271 


n 


211 


272 


272 


No. of A 


11 


13 


11 


No. of A 


6 


6 


6 


Ar 


7.33 


6.93 


6.90 


A« 


3.70 


3.41 


3.88 


He 


0.784 


0.765 


0.769 


He 


0.367 


0.344 


0.429 


^HW 


0.086 


0.749 


0.495 


Phw 


0.275 


0.780 


0.311 


Fjs 


0.027 


0.020 


0.012 


Fis 


-0.019 


0.026 


-0.002 


Lca9l 








Prs275 








n 


202 


268 


262 


n 


211 


272 


273 


No. of A 


7 


8 


8 


No. of A 


7 


9 


8 


Ah 


4.22 


4.38 


4.43 


A« 


5.00 


5.07 


4.61 


He 


0.560 


0.575 


0.570 


He 


0.608 


0.612 


0.579 


^HW 


0.895 


0.927 


0.005 


^HW 


0.711 


0.334 


0.441 


^,.S 


-0.070 


0.039 


0.030 


F,s 


0.034 


0.015 


0.031 


Lea 107 








Prs282 








n 


211 


264 


269 


n 


211 


272 


273 


No. of A 


10 


11 


11 


No. of A 


13 


12 


12 


Afl 


7.90 


8.07 


8.15 


Aa 


8.46 


8.40 


7.89 


He 


0.799 


0.798 


0.775 


He 


0.636 


0.639 


0.614 


^H\V 


0.249 


0.669 


0.346 


^HW 


0.886 


0.556 


0.141 


F,s 


-0.104 


-0.015 


-0.045 


F,s 


-0.051 


0.028 


0.022 
continued 



148 



Fishery Bulletin 104(1) 









Appendix 


Table 2 (continued) 








Locus 


Texas 


Louisiana 


Alabama 


Locus 


Texas 


Louisiana 


Alabama 


Prs55 








Prs303 








n 








n 


211 


272 


270 


No. of A 


7 


6 


6 


No. of A 


10 


9 


12 


Aft 


4.34 


3.62 


3.52 


Ar 


5.33 


5.17 


6.05 


He 


0.266 


0.210 


0.221 


He 


0.375 


0.400 


0.400 


^HW 


0.100 


0.525 


0.199 


'HW 


0.527 


0.344 


0.781 


F,s 


-0.017 


-0.052 


0.051 


Pis 


-0.010 


-0.011 


-0.055 


Prsl37 








Pre328 








n 


211 


272 


271 


11 


211 


272 


273 


No. of A 


13 


13 


12 


No. of A 


6 


6 


5 


-^ft 


7.88 


7.43 


8.00 


Aft 


3.54 


3.45 


3.71 


He 


0.721 


0.694 


0.715 


He 


0.542 


0.545 


0.568 


^HW 


0.127 


0.001 


0.051 


^HW 


0.323 


0.108 


0.191 


F,s 


0.008 


0.105 


0.019 


F,s 


0.090 


-0.018 


0.007 


Prs221 








Prs333 








n 


211 


271 


270 


;i 


211 


272 


272 


No. of A 


19 


18 


17 


No. of A 


6 


7 


6 


Aft 


10.06 


9.54 


10.32 


Aft 


3.84 


4.52 


4.20 


He 


0.800 


0.792 


0.802 


He 


0.342 


0.320 


0.323 


Phw 


0.108 


0.006 


0.797 


^HW 


0.571 


0.819 


0.412 


F,s 


0.016 


0.100 


-0.025 


F,s 


0.029 


-0.022 


0.032 


Prs229 
















/! 


211 


271 


269 










No. of A 


7 


9 


9 










Aft 


5.05 


5.08 


5.51 










ȣ 


0.495 


0.464 


0.527 










^HW 


0.001 


0.000 


0.020 










^/.S 


0.081 


0.181 


0.118 











149 



The relationship between smolt and postsmolt 
growth for Atlantic salmon iSalmo salar) in the 
Gulf of St. Lawrence 



Kevin D. Friedland 

UMass/NOAA Cooperative Marine Education and Research Program 

Blalsdell House 

University of Massachusetts 

Amherst, Massachusetts 01003 

Present address: National Marine Fisheries Service 

28 Tarzwell Dr 

Narragansett, Rhode Island 02882 
E-mail address: kevinfriedlandiginoaagov 

Lora M. Clarke 

Department of Natural Resources Conservation 
Holdsworth Hall 
University of Massachusetts 
Amherst, Massachusetts 01003 

Jean-Denis Dutil 

Ministere des Peche et des Oceans 
Institut Maurice-Lamontagne 
850 route de la Mer, Mont-Joli 
Quebec, G5H 3Z4, Canada 

Matti Salminen 

Finnish Game and Fisheries Research Institute 

Viikinkaari 4 

P.O. Box 2, FIN-00791 

Helsinki, Finland 



The interaction of ocean climate 
and growth conditions during the 
postsmolt phase is emerging as the 
primary hypothesis to explain pat- 
terns of adult recruitment for indi- 
vidual stocks and stock complexes 
of Atlantic salmon (Sabno salar). 
Friedland et al. (1993) first reported 
that contrast in sea surface tempera- 
ture (SST) conditions during spring 
appeared to be related to recruitment 
of the European stock complex. This 
hypothesis was further supported 
by the relationship between cohort 
specific patterns of recruitment for 
two index stocks and regional scale 
SST (Friedland et al., 1998). One of 
the index stocks, the North Esk of 
Scotland, was shown to have a pat- 
tern of postsmolt growth that was 
positively correlated with survival. 



indicating that growth during the 
postsmolt year controls survival and 
recruitment (Friedland et al., 2000). 
A similar scenario is emerging for 
the North American stock complex 
where contrast in ocean conditions 
during spring in the postsmolt migra- 
tion corridors was associated with 
the recruitment pattern of the stock 
complex (Friedland et al., 2003a, 
2003b). The accumulation of addi- 
tional data on the postsmolt growth 
response of both stock complexes will 
contribute to a better understanding 
of the recruitment process in Atlantic 
salmon. 

Anadromous salmonids produce co- 
horts of juvenile smolts that migrate 
over a short period of time at a nearly 
uniform size; however, the variability 
that occurs in migration timing and 



the size spectra of smolts is of poten- 
tial interest in the study of recruit- 
ment (Hoar, 1976; Wedemeyer et al., 
1980). Enhancement and restoration 
programs provide data on the effect 
of smolt size on return rate where 
hatchery practices both intentionally 
and unintentionally produce fish of 
varying size and quality. There are 
many case studies that show positive 
correlations between smolt size and 
return rate in salmonids, but non- 
significant relationships have also 
been observed, which underscore 
the fact that the contrast in size 
that can be achieved in hatcheries 
is often outside ecologically relevant 
limits (Farmer, 1994; Salminen et 
al., 1994). The functional relation- 
ship between smolt size and return 
rate is probably more accurately de- 
scribed as nonlinear and as having 
some optimality within the range of 
hatchery releases (Bilton et al., 1982; 
Henderson and Cass, 1991). 

Researchers have also considered 
the effect of smolt size at ocean entry 
and the effect of smolt size on the en- 
suing growth patterns of postsmolts. 
Ward and Slaney (1988) described a 
positive relationship between return 
rate and smolt size for rainbow trout 
iSaliJio gairdneri). suggesting that 
the recruitment rate of a year class 
was mediated by the mortality that 
occurred when the fish migrated to 
sea. When new data were added to 
that relationship, the original con- 
clusions were no longer supported, 
indicating that factors other than 
smolt size at ocean entry contribute 
to recruitment (Ward, 2000). Al- 
though size at ocean entry may con- 
tribute to mortality risk, postsmolt 
growth, and the biotic and abiotic 
factors affecting postsmolt growth, 
often play a dominant role (Salmin- 
en et al., 1995). However, what has 
remained obscure is whether size 
at ocean entry influences postsmolt 
growth. Skilbrei (1989) and Nicieza 
and Branfia (1993) reported negative 
relationships between smolt size and 



Manuscript submitted 4 February 2004 to 
the Scientific Editor's Office. 

Manuscript approved for publication 
10 April 2005 by the Scientific Editor. 

Fish. Bull. 104:149-155 (2006). 



150 



Fishery Bulletin 104(1) 



marine growth in Atlantic salmon, whereas Lundquist 
et al. (1988) and Salminen (1997) reported positive 
relationships. Einum et al. (2002) reported an inverse 
relationship between pre- and postsmolt growth of At- 
lantic salmon and suggested that phenotypic charac- 
teristics favoring growth in one environment may not 
necessarily favor growth in another. The body of mixed 
results sheds little light on whether smolt size confers a 
growth advantage to postsmolts. It also leaves in doubt 
the importance of freshwater experience on postsmolt 
survival. Common to this body of work is the depen- 
dence on samples coming from river returns and fishery 
catches that may not be representative of the full range 
of growth signatures in postsmolt populations because 
they do not include data from mortalities and often do 
not include data from some return age groups. 

In this study we report on an analysis of scale growth 
indices from a collection of postsmolts from the Gulf of 
St. Lawrence. We measured freshwater growth signa- 
tures representative of smolt size and freshwater growth 
prior to migration and postsmolt growth was partitioned 
by season. These are samples of a life stage in an in- 
termediate phase of marine life that is infrequently 
sampled and may be free of some of the assumed biases 
associated with samples from spawning fish. 




Figure 1 

Postsmolt scale of Atlantic salmon iSalmo 
aalar) in the Gulf of St. Lawrence show- 
ing focus, freshwater growth zone, and 
postsmolt growth zone. 



Material and methods 

We collected data on scale circuli spacing that was rep- 
resentative of the freshwater and postsmolt growth for 
juvenile salmon captured in the Gulf of St. Lawrence. 
These postsmolt salmon were collected in 1982-84 and 
were originally reported in Dutil and Coutu (1988). 
They were captured in experimental gill nets along 
the northwest shore of the Gulf during the months of 
August to October. 

Freshwater and postsmolt growth descriptors were in- 
terpreted from circuli spacing patterns deposited on the 
scale. Scales were cleaned and mounted between glass 
slides and the spacings of scale circuli were measured 
with a Bioscan Optimas (Media Cybernatics, Inc., Sil- 
ver Spring, MD) image processing system. Freshwater 
zone length (FZL) was taken as the total distance from 
the center of the scale focus to the transition between 
freshwater and postsmolt growth (Fig. 1). This tran- 
sition zone is defined by the appearance of the first 
marine intercirculus spacing, i.e., as a wider spacing 
(according to the reader's judgment) than the progres- 
sion of spacings in the freshwater zone. The FZL ends 
at the last freshwater circulus. FZL was interpreted as 
a proxy for smolt length at migration. The mean of the 
last five circuli spacings during the freshwater phase 
was computed for each sample (CSLF, circuli spacing 
during last freshwater period). This circuli spacing 
index was interpreted as an indication of freshwater 
finishing growth (the final phases of freshwater growth 
prior to migration). Two circuli spacing indices were 
extracted from the postsmolt zone, the mean spacing 
between intercirculi 2 through 6 — an interval which 



was interpreted as the growth index for the early ma- 
rine period (CSFM, circuli spacing during first marine 
period) — and the mean intercirculi spacing of the next 
five pairs occurring later in the postsmolt growth — an 
interval interpreted as an index of summer growth 
(CSSM, circuli spacing during summer marine period). 
The total length of the postsmolt growth zone was also 
extracted (MZL, marine zone length), which was the 
distance from the freshwater-marine transition, starting 
at the last freshwater circulus to the outer edge of the 
scale. All measurements were made on a single scale 
from each specimen along the 360° axis of the scale. 

We tested the functional relationship between fresh- 
water and marine growth, early and late marine 
growth, and seasonal progression of the growth zones 
using linear regression. We tested the significance of 
the slope parameters for each relationship as an indica- 
tion of variable dependency. 

We also considered whether these data might contrib- 
ute to our understanding of size selective mortality by 
constructing a time series plot of FZL and MZL against 
date of capture. An anticipated positive trend in MZL 
would reflect growth during the marine phase, but any 
trend in FZL would reflect size-specific removals at- 
tributable to size at ocean entry. 



Results 

Measurements were taken from the scales of 587 posts- 
molts collected during 1982-84. FZL averaged 0.67 mm 
(SD = 0.163) for all samples. CSLF averaged 0.022 mm 
(SD = 0.0049), reflecting the slower growth and more 
narrowly spaced circuli of the freshwater zone. The two 



NOTE Fnedland el al Relationship between smoll and postsmolt growth in Salmo solar 



circuli spacing indices from the postsmolt growth 
zone were larger in magnitude than the index from 
the freshwater zone. CSFM averaged 0.061 mm 
(SD = 0.0109) and reflected a threefold increase in 
circuli spacing between the freshwater and marine 
zones. CSSM averaged 0.064 mm (SD = 0.0100), 
reflecting the increased growth occurring within 
the duration of marine residency'. MZL increased 
over time from approximately 0.6 mm in early 
August to 1.2 mm by mid-October. 

The data we examined to test the relation- 
ship between smolt size and freshwater finishing 
growth and postsmolt growth indicate an absence 
of any linkage between the two growth environ- 
ments. We found no significant relationships be- 
tween smolt sizes as represented by FZL and ei- 
ther spring (CSFM) or summer (CSSM) postsmolt 
growth as represented by the circuli spacing indi- 
ces. The scatter between CSFM and FZL suggests 
there were no linear trends for any of the study 
years (Fig. 2A). None of the three regressions had 
slopes significantly different from zero (Table 1). 
There was also an absence of any relationship 
between CSSM and FZL as evidenced by a similar 
pattern of scatter for the coordinates (Fig. 2B) and 
by an absence of significant slopes. 

The relationship was further tested by consid- 
ering the growth that occurs just prior to smolt 
migration as an indication of smolt condition at 
entry into the marine environment. As was the 
case with the bivariate relationships with FZL, 
CSLF was not a significant predictor of either 
CSFM or CSSM (Fig. 3). None of the slopes were 
significantly different from zero (Table 1). 

Growth in the early and later part of the post- 
smolt season were correlated. There were signifi- 
cant positive linear relationships between CSSM 
and CSFM for all three years (Fig. 4), indicating that 
the postsmolts began their marine residence with rapid 
growth and continued to grow at a rapid rate through 
summer. The slopes of all three regressions were sig- 
nificantly different from zero (Table 1). 

A time series plot of MZL, plotted by date of capture, 
shows an increasing trend reflecting the growth that 
postsmolts experience during early marine residence 
(Fig. 5). For the same fish, we also plotted FZL by date 
of capture and found a negative trend over time, indi- 
cating that FZL decreased during the period (Table 1), 
which would not be consistent with selective mortality 
of smolts of smaller initial size. 



Discussion 

Our main findings indicate that marine growth of post- 
smolt Atlantic salmon sampled from August to October 
in the Gulf of St. Lawrence was independent of fresh- 
water growth history. Neither smolt size nor the last 
freshwater growth period of smolts prior to migration 
was related to postsmolt growth patterns. Furthermore, 



0,12 



10' 



08- 



0,06  



£■ 0,04. 
E 



9- 12- 



O 0,10' 



0.08- 



0.06 



04 



1982 
1983 
1984 

■1982 
1983 

•1984 



o -o „-,=•*- 



V^'-.-Jf 



_i_.'-- — qiij ^y^T* , ■• sf-®' - 



B 



* » ° ooo.\s«' ce o ■* O 






0,2 



— I — 
0,6 



— I — 
1,0 



0,4 0,6 0.8 1,0 1,2 

Freshwater zone length (mm) 



1,4 



Figure 2 

Relationships between circuli spacing during the early marine 
period (A) and circuli spacing during the summer marine 
period (B) and freshwater zone length for Atlantic salmon 
tSalmo salar) in the Gulf of St. Lawrence. 



Table 1 

Linear associations between scale growth indices and 
date of capture for Gulf of St. Lawrence Atlantic salmon 
iSalnw salar) postmolts. FZL = freshwater zone length; 
CSLF=mean of the last five freshwater intercirculi spac- 
ings; CSFM=mean spacing of intercirculi spacings 2 
through 6; CSSM = mean spacing of intercirculi spacings 
7 through 11; DOC = date of capture. 



Linear 
associations 



Probability Hg 
slope = 



Predictor iX) Dependent (7) 1982 1983 1984 



FZL 

FZL 

CSLF 

CSLF 

CSFM 

DOC 

DOC 

Sample size 



CSFM 

CSSM 

CSFM 

CSSM 

CSSM 

FZL 

MZL 



0.53 

0.50 

0.27 

0.33 

<0.01 

<0.01 

<0.01 

374 



0.87 
0.61 
0.70 
0.57 
<0.01 



158 



0.57 
0.22 
0.22 
0.27 
<0.01 



55 



152 



Fishery Bulletin 104(1) 



there was no evidence that size at the smolt stage of 
postsmolts captured later in the season was larger than 
those captured earlier in the season, indicating that 
there was no size-selective attrition on smaller smolts 
during August to October. The strength of these data is 
that the samples were derived from collections of posts- 
molts, not river returns or fishery catches, and thus it is 
less biased and more representative of natural mortality 
factors. It will take considerably more data to fully test 
this hypothesis, but the highest priority should be given 
to similar sample collections, especially those collected 
earlier in the year (Holm et al., 2000). These data will 
be pivotal in interpreting the relative impact of size 
at ocean entry versus postsmolt growth for determing 
postsmolt survival. 

Understanding survival during the postsmolt period 
is of great importance in the face of declining stock 
abundance and possible stock extirpations over por- 
tions of the range of Atlantic salmon (Anderson et al., 
2000). With evidence showing the success of conserva- 
tion efforts to increase freshwater populations (Swans- 
burg et al., 2002), the range of potential life stages 



0.12- 



0,10' 



0.08- 



0.06  



0.04 



0.12 



O 0.10 



0.08- 



0.06 



0.04 




1982 
1983 
1984 
1982 
1983 
1984 



B 



„«oe 



o**.>-?' 



^^^■■^ 



if-- 



0.01 



0.05 



0.02 0.03 0.04 

Circuli spacing, late freshwater period (mm) 

Figure 3 

Relationships between circuli spacing during the early marine 
period (A) and circuli spacing during the summer marine period 
(B) and circuli spacing during the late freshwater period for 
Atlantic salmon (Salmo salar) in the Gulf of St. Lawrence. 



and habitats causing recruitment failure is reduced. 
Evidence at the population level showing coherence be- 
tween early marine growth and survival at sea is still 
a long way from providing sufficient evidence to fully 
describe survival mechanisms or climate linkages. The 
hypothesized climate forcing appears to be related to 
a mismatch between ocean entry time for smolts and 
the thermal regime they find during their transition 
to marine life (Friedland et al., 2003b). The concept is 
supported by the growth response of postsmolts over a 
range of temperatures that they would likely encounter 
during that period of time. Atlantic salmon postsmolts 
follow a nonlinear growth response to temperature 
and optimum growth occurs at 13°C (Handeland et al., 
2003). Variations in migration timing could result in 
fish encountering ocean conditions that are too warm 
or cool; these variations could explain the different 
associations of North American and European stocks 
to ocean temperatures. It would also be interesting to 
know if the thermal growth optimum is robust to vary- 
ing feeding rations, which may be an issue for regional 
stock groups in the North Atlantic. 

The transition period from freshwater to ma- 
rine life is characterized by a number of dy- 
namic changes in growth and predation. The 
highest mortality rates for postsmolts occur 
during the first weeks at sea (Eriksson, 1994). 
As a consequence, even a minimal variation in 
these rates will have a large impact on adult 
recruitment. Size-selective mortality on juve- 
niles has been demonstrated for other salmo- 
nids such as sockeye salmon {Ojjcorhynchiis 
nerka) and has been shown to be seasonally 
concentrated (West and Larkin, 1987). Size- 
specific predation often occurs but predators, 
such as adult bird species, are not likely to 
grow at corresponding rates to those of the fish 
(Dieperink et al., 2002). Thus, size at ocean 
entry and the ability to grow out of the size 
range vulnerable to specific predators must 
affect the dynamics of predation rate. The pe- 
riod of predation vulnerability is mediated by 
growth rate; therefore the argument returns 
to what confers faster growth on postsmolts in 
some years? 

One dynamic that challenges all stocks is the 
transition from predominantly invertebrate to 
piscivorous prey. Some authors suggest it is the 
success in making this transition, or the time 
it takes to make the transition, that dictates 
the amount of time the postsmolts take to grow 
out of the predation-vulnerable size range (Du- 
til and Coutu, 1988; Andreassen et al., 2001; 
Salminen et al., 2001). Depending on prevail- 
ing foraging conditions, this transition may be 
promoted by increased smolt size, as was the 
case in the northern Baltic Sea in 1990-93 
(Salminen et al., 2001). One potential expla- 
nation for the contradictory views on the rela- 
tionship between pre- and postsmolt growth in 



NOTE Fnedland et al : Relationship between smolt and poslsmolt growth in Salmo solar 



153 



Atlantic salmon may arise from geographical dif- 
ferences or year-to-year changes (or both) in prey 
regimes during the critical diet-transition period. 
In lake-run brown trout, Niva and Jokela (2000) 
found a positive correlation between hatchery 
growth rate and postsmolt growth in a lake with 
small fish as the main prey. In contrast, hatchery 
growth did not predict growth in a lake where the 
fish had to prey on bottom-dwelling invertebrates. 
This finding may indicate that assessment of in- 
dividual performance may be highly specific to 
environments in migratory salmonids. 

Size of postsmolts at ocean entry in salmonids 
also manifests itself in predictable patterns of life 
history variation. In sea trout (Salmo triitta) there 
is a close correspondence between size of juvenile 
stage and adult size, which is believed to have an 
impact on recruitment through size variation at 
critical life stages (Elliott, 1985). Hutchings and 
Jones (1998) found that smolt age and size had 
a small effect on the growth-rate threshold for 
maturity in Atlantic salmon. Smolt-size variation 
in sea trout has been associated with latitude, 
indicating that for this species recruitment strate- 
gies are likely adapted to local physical conditions, 
predators, and prey regimes (L' Abee-Lund et al., 
1989). Wild smolts are often smaller than hatch- 
ery smolts for the same or allied strains, yet they 
consistently survive at higher rates (Poole et al., 
2003). Obviously there are other cofactors that can 
explain why the wild fish overcome any advantage 
conferred on hatchery fish by their larger size at 
migration. The wild versus hatchery comparison 
is a relatively weak piece of evidence to use to 
dismiss the idea that smolt size is an important 
survival factor. However, a recent study of size at 
ocean entry in a Quebec stock of wild fish may 
provide more useful evidence (Dodson^ Caron-). 
The investigators found that the size spectra of 
smolts were not different from the size spectra of 
back-calculated smolt sizes for returning adults, 
suggesting some factor other than size at ocean 
entry was controlling survival. It should be noted 
that postsmolts collected in the present study were 
of unknown origin, but the scale pattern for the 
freshwater zone clearly indicates that they were 
of wild origin. 

The growth response of postsmolts may be gov- 
erned by a simple response to temperature optima 
occurring in coastal marine waters, but it may al- 
so be more complex than that. The surface waters 
in the Gulf of St. Lawrence circulate in wide-scale 



Dodson, J. 2004. Personal comniun. Departement 
de Biologie Pavilion Alexandre-Vachon, local .30.58-B 
Universite Laval, Quebec, GIK 7P4. Canada. 
Caron. F. 2004. Personal commun. Direction de 
la Recherche Faune et Pares Quebec, 67.5 est, Boul. 
Rene-Levesque Boite 92, lie etage, Quebec, GIR .5V7, 
Canada. 



012- 



0-10- 



0.08- 



0.06- 



0,04- 



1982 
1983 
1984 
-1982 
1983 
1984 




0.02-1 1 1 1 1 1 1 1 1 1 r 

0.02 0.04 0,06 0.08 010 12 

Circuli spacing, early marine period (mm) 

Figure 4 

Relationships between circuli spacing during the summer 
marine period and circuli spacing during the early marine 
period for Atlantic salmon (Salinn salar) in the Gulf of St. 
Lawrence. 



e 9 



03- 



Freshwater 
Marine 




•»•••• . . • ' 



« z: 






Aug 6 



Aug 20 



Sep 3 Sep 17 

Date of capture 



— r- 
Oct • 



Figure 5 

Freshwater zone length and marine zone length, by date of 
capture during the 1982 field season for Atlantic salmon t Salmo 
salar) in the Gulf of St. Lawrence. 



154 



Fishery Bulletin 104(1) 



gyres over the extremely cold waters of the cold inter- 
mediate layer (CIL), thus creating a thermally complex 
habitat. Strong winds periodically force cold waters 
from the CIL to the surface (Ouellet, 1997). Postsmolts 
appear to be opportunistic feeders in the ocean, and 
despite their predominant use of surface waters, they 
have been known to use benthic habitats and water 
column features as well (Levings, 1994; Sturlaugsson, 
1994). These behaviors would distribute fish in a man- 
ner independent of surface temperatures because forage 
behaviors are not necessarily controlled by temperature 
preferences. Growth in postsmolts is also affected by 
other physical parameters, such as photoperiod (Fors- 
berg, 1995), and possibly by variation in sea surface 
salinity as shown for chum salmon (Oncorhynchus keta) 
(Morita et al., 2001). We can predict migration timing 
and initial migration trajectories for postsmolts. but we 
are hard pressed to predict the physical nature of their 
habitats over time and the physiological effect of these 
habitats on postsmolt growth. 



Acknowledgments 

We thank T. Sadusky for help with the initial phases of 
this study and two anonymous reviewers for productive 
comments. 



Literature cited 

Anderson, J. M., F. G. Whoriskey. and A. Goode. 

2000. Atlantic salmon on the brink. Endangered Spe- 
cies Update 17:15-21. 

Andreassen, P. M. R.. M. B. Martinussen, N. A. Hvidsten, and 
S. O. Stefansson. 

2001. Feeding and pre-selection of wild Atlantic salmon 
post-smolts. J. Fish Biol. 58:1667-1679. 

Bilton, H. T., D. F. Alderdice, and J. T. Schnute. 

1982. Influence of time and size at release of juvenile 
coho salmon [Oncorhynchus kisutch) on returns at 
maturity. Can. J. Fish. Aquat. Sci 39:426-447. 
Dieperink, C., B. D. Bak, L. F. Pedersen, M. I. Pedersen, and 
S. Pedersen. 

2002. Predation on Atlantic salmon and sea trout during 
their first days at sea as postsmolts. J. Fish Biol. 
61:848-852. 

Dutil, J.-D., and J.-M. Coutu. 

1988. Early marine life of Atlantic salmon, Salmo salar, 
postsmolts in the northern Gulf of St. Lawrence. Fish. 
Bull. 86:197-212. 
Einum, S., E. B. Thorstad, and T. F. Nssje. 

2002. Growth rate correlations across life stages in female 
Atlantic salmon. J. Fish Biol. 60:780-784. 
Elliott, J. M. 

1985. Growth, size, biomass, and production for dif- 
ferent life stages of migratory trout Salmo trutta in 
a Lake District stream, 1966-83. J. Animal Ecol. 
54:985-1001. 
Eriksson, T. 

1994. Mortality risks of Baltic salmon during downstream 
migration and early sea-phase: effects of body size and 
season. Nord. J. Freshw. Res. 69:100. 



Farmer, G. J. 

1994. Some factors which influence the survival of hatch- 
ery Atlantic salmon [Salmo salar) smolts utilized for 
enhancement purposes. Aquaculture 121:223-233. 

Forsberg, O. I. 

1995. Empirical investigation on growth of post-smolt 
Atlantic salmon iSalmo salar L.) in land-based farms. 
Evidence of photoperiodic influence. Aquaculture 
133:235-248. 

Friedland, K. D., L. P. Hansen, and D. A. Dunkley. 

1998. Marine temperatures experienced by postsmolts 
and the survival of Atlantic salmon (Salmo salar L.) in 
the North Sea area. Fish. Oceanogr. 7:22-34. 
Friedland, K. D., L. P. Hansen, D. A. Dunkley, and J. C. 
MacLean. 

2000. Linkage between ocean climate, post-smolt growth, 
and survival of Atlantic salmon (Salmo salar L. I in the 
North Sea area. ICES J. Mar. Sci. 57:419-429. 
Friedland, K. D., D. G. Reddin, D. G., and M. Castonguay. 

2003a. Ocean thermal conditions in the post-smolt nurs- 
ery of North American Atlantic salmon. ICES J. Mar. 
Sci. 60:343-355. 
Friedland, K. D., D. G. Reddin, and J. F. Kocik. 

1993. Marine survival of North American and Euro- 
pean Atlantic salmon: effects of growth and environ- 
ment. ICES J. Mar. Sci. 50:481-492. 

Friedland, K. D., D. G. Reddin, J. R. McMenemy, and K. F. 
Drinkwater. 
2003b. Multideeadal trends in North American Atlantic 
salmon (Salmo salar) stocks and climate trends rel- 
evant to juvenile survival. Can. J. Fish. Aquat. Sci. 
60:563-583. 
Handeland, S. O., B. T. Bjbrnsson, A. M. Arnesen, and S. O. 
Stefansson. 
2003. Seawater adaptation and growth of post-smolt 
Atlantic salmon (Salmo salar) of wild and farmed 
strains. Aquaculture 220:367-384. 
Henderson, M. A., and A. J. Cass. 

1991. Effect of smolt size on smolt-to-adult survival for 
Chilko Lake sockeye salmon (Oncorhynchus nerka ). Can. 
J. Fish. Aquat. Sci. 48:988-994. 
Hoar. W. S. 

1976. Smolt transformation: evolution, behaviour, and 
physiology. J. Fish. Res. Board Can. 33:1233-1252. 
Holm, M., J. C. Hoist, and L. P. Hansen. 

2000. Spatial and temporal distribution of post-smolts of 
Atlantic salmon (Salmo salar L.) in the Norwegian Sea 
and adjacent areas. ICES J. Mar. Sci. 57:955-964. 
Hutchings, J. A., and M. E. B. Jones. 

1998. Life history variation and growth rate thresholds 
for maturity in Atlantic salmon, Salmo salar. Can. J. 
Fish. Aquat. Sci. 55(suppl. l):22-47. 
Levings, C. D. 

1994. Feeding behavior of juvenile salmon and significance 
of habitat during estuary and early sea phase. Nord. 
J. Freshwater Res. 69:7-16. 

L Abee-Lund, J. H., B. Jopnsson, A. J. Jensen, L. M. Saettem, 
T. G. Heggberget, B. . Johnson, T. F. Ntesje. 

1989. Latitudinal variation in life-history characteris- 
tics of sea-run migrant brown trout Salmo trutta. J. 
Animal Ecol. 58:525-542. 
Lundquist, H., W. C. Clarke, and H. Johansson. 

1988. The influence of sexual maturation on survival to 
adulthood of river stocked Baltic salmon, Salmo salar, 
smolts. Holarct. Ecol. 11:60-69. 



NOTE Fnedland et a\ Relationship between smolt and postsniolt growth in Salmo salai 



155 



Merita, S. H., K. Morita, and H. Sakano. 

2001. Growth of chum salmon [Oncorhynchiis keta) 
correlated with sea-surface salinity in the North 
Pacific. ICES J. Mar. Sci. 58:1335-1339. 

Nicieza, A. G, and F. Braniia. 

1993. Relationships among smolt size, marine growth, 
and sea age at maturity of Atlantic salmon iSahno 
salar) in northern Spain. Can. J. Fish. Aquat. Sci. 
50:1632-1640. 

Niva, T., and J. Jokela. 

2000. Phenotypic correlation of juvenile growth rate 
between different consecutive foraging environments 
in a salmonid fish: a field experiment. Evol. Ecol. 
14:111-126. 

Ouellet, P. 

1997. Characteristics and vertical distribution of Atlantic 
cod {Gadus morhua) eggs in the northern Gulf of St. 
Lawrence, and the possible effect of cold water tem- 
perature on recruitment. Can. J. Fish. Aquat. Sci. 
54:198-210. 
Poole, W. R., D. T. Nolan, T. Wevers, M. Dillane, D. Cotter, and 
O. Tully. 

2003. An ecophysiological comparison of wild and hatch- 
ery-raised Atlantic salmon (Salmo salar L.) smolt from 
the Burrishoole system, western Ireland. Aquaculture 
222:301-314. 
Salminen, M. 

1997. Relationships between smolt size, postmolt growth 
and sea age at maturity in Atlantic salmon ranched in 
the Baltic Sea. J. Appl. Ichthyol. 13:121-130. 
Salminen, M., E. Erkamo, and J. Salmi. 

2001. Diet of post-smolt and one-sea-winter Atlantic 
salmon in the Bothnian Sea, Northern Baltic. J. Fish 
Biol. 58:16-35. 



Salminen, M., S. Kuikka, and E. Erkamo. 

1994. Divergence in the feeding migration of Baltic salmon 
i Sal mo salar L.I; the significance of smolt size. Nord. 
J. Freshw. Res. 69:32-42. 

1995. Annual variability in survival of sea-ranched Baltic 
salmon, Salmo salar L.: Significance of smolt size and 
marine conditions. Fish. Manag. Ecol. 2:171-184. 

Skilbrei, O. T. 

1989. Relationships between smolt lengths and growth 
and maturation in the sea of individually tagged Atlantic 
salmon iSalmo salar). Aquaculture 83:95-108. 
Sturlaugsson, J. 

1994. Food of ranched Atlantic salmon iSalmo salar L. I 
postsmolts in coastal waters. West Iceland. Nord. J. 
Freshw. Res. 69:43-57. 
Swansburg, E, G. Chaput, D. Moore, D. Cassie, and N. El-Jabi. 
2002. Size variability of juvenile Atlantic salmon: links to 
environmental conditions. J. Fish Biol. 61:661-683. 
Ward, B. R. 

2000. Declivity in steelhead ( Oncorhyncbus mykiss ) recruit- 
ment at the Keogh River over the past decade. Can. J. 
Fish. Aquat. Sci. 45:298-306. 
Ward, B. R., and P. A. Slaney. 

1988. Life history and smolt-to-adult survival of Keogh 
River steelhead trout (So/mo gairdneri) and the rela- 
tionship to smolt size. Can. J. Fish. Aquat. Sci. 45: 
1110-1122. 
Wedemeyer G.A., R. L. Saunders, and W. C. Clarke. 

1980. Environmental factors affecting smoltification and 
early marine survival of anadromous salmon. Mar. 
Fish. Rev. 42:1-14. 
West, C. J., and P. A. Larkin. 

1987. Evidence for size-selective mortality of juvenile sock- 
eye salmon ( Oncorhynchus nerka I in Babine Lake, British 
Columbia. Can. J. Fish. Aquat. Sci. 44:712-721. 



156 



Fishery Bulletin 104(1) 



Erratum 



Erratum: Fishery Bulletin 102(4): p. 568. 

Calambokidis, John, Gretchen H. Steiger, David K. Ellifrit, Barry L. Troutman, 
and C. Edward Bowlby 

Distribution and abundance of humpback whales (Megaptera nouaeangliae) and other marine 
mammals off the northern Washington coast 



Correction: Table 3 should read as follows 



be 



bo 



a. 
O 



too 
c 








to 


>. 


en 














w 




M 




C 












JS 




so 








CO 








ro 




g 


a> 


c 


03 



s "3 
s ^ 

= c 
cc .5 



2 be 



c ra 
-: E 



2 c 






o to 



Z be 



2 be 



fM r- -^T CO 

O CO 



C^) lO "* 50 

CO (>] 

CO 






CD T-H 

00 



^ CO 0) 



1 Ji 

a. CO 

3 > 

X 



" ^ c S 



O § 



CQ 



CO O^ CO CD CT; Tj- OJ 

O C30 CO I> CO Ol 



CO OJ 



CD CO 



CO 
CJl 



00 00 



X ^ G^ CO CO CD 

Tt ^ CD CO 



Tji I> Tf ^ ^ CO 



<M CO CO 






Q a: 0. 



■= X -C 



CO 



_. CO 



o 



£ td 



c -a 



CDO '^CO OICO -^t^ 

^ CO CO ^ CD 



ICO ''1'^ ODCO COCO 

■-I CO CO i-i 00 



1-1 CO CO 'H ^ 



^ CO CO >-H ^ 



a 
o 

"ffl s 



. . at flJ 
1- a; m S 



T3 .g 



^ CO 



m 2: 



(X o 



o 

CO 



^H ic O^ CO -cf^ C^ 





br 




r 






o, 








c 


br 


c 


Cfl 


a 


'-- 



157 



Fishery Bulletin 

Guidelines for authors 

Content of manuscripts 

Contributions published in Fishery Bulletin describe 
original research in marine fishery science, fishery engi- 
neering and economics, as well as the areas of marine 
environmental and ecological sciences (including model- 
ing). Although all contributions are subject to peer review, 
responsibility for the contents of papers rests upon the 
authors and not upon the editor or publisher. Submission 
of an article implies that the article is original and is not 
being considered for publication elsewhere. Manuscripts 
must be written in English. Authors whose native lan- 
guage is not English are strongly advised to have their 
manuscripts checked by English-speaking colleagues 
prior to submission. Articles may range from relatively 
short contributions (10-15 typed and double-spaced pages, 
tables and figures not included) to extensive contributions 
(20-30 typed pages). Notes are reports of 5 to 10 pages 
without an abstract and describe methods or results not 
supported by a large body of data. 

Manuscript preparation 

Title page should include authors' full names and mailing 
addresses and the senior author's telephone, fax number, 
and e-mail add/-ess, and a list of key words to describe the 
contents of the manuscript. Abstract should be limited to 
150 words (one-half typed page), state the main scope of 
the research, and emphasize the author's conclusions and 
relevant findings. Do not review the methods of the study 
or list the contents of the paper. Because abstracts are cir- 
culated by abstracting agencies, it is important that they 
represent the research clearly and concisely. Text must 
be typed in 12 point Times New Roman font throughout. 
A brief introduction should convey the broad significance 
of the paper; the remainder of the paper should be divided 
into the following sections; Materials and methods, 
Results, Discussion (or Conclusions), and Acknowl- 
edgments. Headings within each section must be short, 
reflect a logical sequence, and follow the rules of multiple 
subdivision (i.e., there can be no subdivision without at 
least two items). The entire text should be intelligible to 
interdisciplinary readers; therefore, all acronyms, abbre- 
viations, and technical terms should be written out in full 
the first time they are used. Include FAO common names 
for species in the list of keywords and in the introduction. 
Regional common names may be used throughout the rest 
of the text if they are different from FAO common names 
which can be found at http;//www.fishbase.org/search. 
html. Follow the U.S. Government Printing Office Style 
Manual ( 1984 ed. ) and the CBE Style Manual ( 6th ed. ) for 
editorial style, and the most current issue of the Ameri- 
can Fisheries Society's Common and Scientific Names of 



Fishes from the United States and Canada for fish nomen- 
clature. Dates should be written as follows: 11 November 
2000. Measurements should be expressed in metric units, 
e.g., 58 metric tons (t); if other units of measurement are 
used, please make this fact explicit to the reader. Write 
out the numbers zero through nine unless they form part 
of measurement units (e.g., nine fish but 9 mm). 

Text footnotes should be inserted in 9-point font at 
the bottom of the page that displays the first citation of 
the footnote. Footnotes should be formatted in the same 
manner as citations. Footnote all personal communica- 
tions, unpublished data, and unpublished manuscripts 
with full address of the communicator or author, or, as in 
the case of unpublished data, where the data are on file. 
Authors are advised to avoid references to nonstandard 
(gray) literature (such as internal, project, processed, 
or administrative reports, ICES Council Minutes, IWC 
Minutes or Working Papers, any "research" or "working" 
documents, laboratory or contract reports. Management 
Council reports, and manuscripts in review) wherever 
possible. If these references are used, present them as 
footnotes and list whether they are available from NTIS 
(National Technical Information Service) or from some 
other public depository. Cite all software and special 
equipment or chemical solutions used in the study, not in 
a footnote but within parentheses in the text (e.g., SAS, 
vers. 6.03, SAS Inst., Inc., Cary, NO. 

Literature cited comprises published works and 
those accepted for publication in peer-reviewed litera- 
ture (in press). Follow the name and year system for 
citation format. If there is a sequence of citations in the 
text, list chronologically; (Smith, 1932; Green, 1947; 
Smith and Jones, 1985). Abbreviations of serials should 
conform to abbreviations given in the Serial Sources 
for the BIOSIS Previews Database. Authors are respon- 
sible for the accuracy and completeness of all citations. 
Literature citation format; Author (last name, followed 
by first-name initials). Year. Title of report or manu- 
script. Abbreviated title of the series to which it belongs. 
Always include number of pages. If there is a sequence 
of citations by the same first author, list the works 
alphabetically according to the last name of following 
authors (e.g.. Smith G. P., L. C. Brown, 1982; Smith, 
G. P., and T. P. Stuart, 1982). If the authorship is identi- 
cal, list works chronologically. 

Tables and figures — general format 

• Zeros should precede all decimal points for values less 
than one. 

• Sample size, n, should be italicized. 

• Capitalize the first letter of the first word in all labels 
within figures. 

• Do not use overly large font sizes in maps and for units 
of measurements along axes in figures. 

• Do not use bold fonts or bold lines in figures. 

• Do not place outline rules around graphs. 

• Do not use horizontal lines in graphs to indicate mea- 
surement units on axes. 



158 



Fishery Bulletin 104(1) 



• Use a comma in numbers of five digits or more (e.g. 
13,000 but 3000). 

• Maps should have a North arrow (or compass sign) and 
degrees latitude-longitude (e.g., 170 E) 

Tables are often overused in scientific papers; it is 
seldom necessary or even desirable to present all the data 
associated with a study. Tables should not be excessive 
in size and must be cited in numerical order in the text. 
Headings should be short but ample enough to allow the 
table to be intelligible on its own. All unusual symbols 
must be explained in the table legend. Other incidental 
comments may be footnoted with italic footnote markers. 
Use asterisks to indicate probability in statistical data. Do 
not type table legends on a separate page; place them on 
the same page as the table data. Do not submit tables in 
photo mode. 

Figures include line illustrations, photographs (or 
slides), and computer-generated graphs and must be cited 
in numerical order in the text. Graphics will aid in the 
comprehension of the text, but they should be limited 
to presenting patterns rather than raw data. Figures 
are costly to print and should be limited to six. Figures 
must be labeled with author's name and number of figure. 
Avoid placing labels vertically (except on y-axis). Figure 
legends should explain all symbols and abbreviations and 
should be double-spaced on a separate page at the end of 
the manuscript. Please note that we do not print graphs 
in color. 

FAILURE TO FOLLOW 

THESE GUIDELINES 

WILL DELAY 

PUBLICATION 

OF A MANUSCRIPT 

Copyright law does not apply to Fishery Bulletin. 
which falls within the public domain. However, if an 



author reproduces any part of an article from Fishery Bul- 
letin in his or her work, reference to source is considered 
correct form (e.g.. Source: Fish. Bull 97:105). 

Reprints are available free of charge to the senior 
author (50 copies) and to his or her laboratory (50 copies). 
Additional copies may be purchased in lots of 100 when 
the author receives page proofs. 

Submission 

The Scientific Editorial Office encourages authors to 
submit their manuscripts as a single PDF (preferred), 
Word (zipped), or WordPerfect (zipped) document by e- 
mail to Fishery.Bulletin@noaa.gov. Please use the sub- 
ject heading "Fishery Bulletin manuscript submission." 
Do not send encrypted files. For further details on elec- 
tronic submission, please contact the Scientific Editorial 
Office directly (see address below). Or you may send your 
manuscript on compact disc in one of the above formats 
along with four printed copies (one original plus three 
copies [stapled]) to the Scientific Editor, at the address 
shown below. Send photocopies only of figures with initial 
submission of manuscript; do not send original figures. 
Original figures and electronic copies of figures will be 
requested later when the manuscript has been accepted 
for publication. 

Dr. Adam Moles 

Scientific Editor, Fishery Bulletin 

11305 Glacier Hwy 

Juneau, AK 99801-8626 

Once the manuscript has been accepted for publication, 
you will be asked to submit a final software copy of your 
manuscript. When requested, the text and tables should 
be submitted in Word or Word Rich Text Format. Figures 
should be sent as PDF files, Windows metafiles, tiff files, 
or as EPS files. Send a copy of figures in original software 
if conversion yields a degraded version. 



[^^Sftsu/ mMsam9Xftst2/ mps 






U.S. Department 
of Commerce 

Volume 104 
Number 2 
April 2006 




Fishery 
Bulletin 




U.S. Department 
of Commerce 

Carios M. Gutierrez 

Secretary 



National Oceanic 
and Atmospiieric 
Administration 

Vice Admiral 

Conrad C. Lautenbacher Jr., 

USN (ret) 

Under Secretary for 
Oceans anci Atmosphere 



National Marine 
Fisheries Service 

William T. Hogarth 

Assistant Administrator 
for Fislneries 




Scientific Editor 

Adam Moles, Ph.D. 

Technical Editor 

Elizabeth Calvert 

National Marine Fisheries Service, NOAA 
11305 Glacier Highway 
Juneau, Alaska 99801-8626 



^OFCo, 




^ATBS 0» " 



Managing Editor 

Sharyn Matriotti 

National Marine Fisheries Service 
Scientific Publications Office 
7600 Sand Point Way NE, BIN C15700 
Seattle, Washington 98115-0070 



The Fishery Bulletin (ISSN 0090-0656) 
is published quarterly by the Scientific 
Puijlications Office. National Marine 
Fisheries Service, NOAA, 7600 Sand 
Point Way NE, BIN C15700, Seattle, WA 
98115-0070. Periodicals postage is paid 
at Seattle, WA. POSTMASTER: Send 
address changes for subscriptions to Fish- 
ery Bulletin, Superintendent of Docu- 
ments, Attn.: Chief, Mail List Branch, 
Mail Stop SSOM, Washington, DC 20402- 
9373. 

Although the contents of this publica- 
tion have not been copyrighted and may 
be reprinted entirely, reference to source 
is appreciated. 

The Secretary of Commerce has deter- 
mined that the publication of this peri- 
odical is necessary according to law for 
the transaction of public business of this 
Department. Use of funds for printing of 
this periodical has been approved by the 
Director of the Office of Management and 
Budget. 

For sale by the Superintendent of 
Documents, U.S. Government Printing 
Office, Washington, DC 20402. Subscrip- 
tion price per year: $55.00 domestic and 
$68.75 foreign. Cost per single issue: 
$28.00 domestic and $35.00 foreign. See 
back for order form. 



Editorial Committee 

Harlyn O. Halvorson, Ph.D. 
Ronald W. Hardy, Ph.D. 
Richard D. fVlethot, Ph.D. 
Theodore W. Pietsch, Ph.D. 
Joseph E. Powers, Ph.D. 
Harald Rosenthal, Ph.D. 
Fredric M. Serchuk, Ph.D. 
George Walters, Ph.D. 



University of Massachusetts, Boston 
University of Idaho, Hagerman 
National Marine Fisheries Service 
University of Washington, Seattle 
National Marine Fisheries Service 
Universitat Kiel, Germany 
National Marine Fisheries Service 
National Marine Fisheries Service 



Fishery Bulletin web site: www.fishbulI.noaa.gov 



The Fishery Bulletin carries original research reports and technical notes on investigations in 
fishery science, engineering, and economics. It began as the Bulletin of the United States Fish 
Commission 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. Beginning with volume 70, number 1, January 1972, the Fishery Bulletin became a 
periodical, issued quarterly. In this form, it is available by subscription from the Superintendent 
of Documents, U.S. Government Printing Office. Washington, DC 20402. It is also available free in 
limited numbers to libraries, research institutions. State and Federal agencies, and in exchange 
for other scientific publications. 



U.S. Department 
of Commerce 

Seattle, Washington 

Volume 104 
Number 2 
April 2006 



Fishery 
Bulletin 



MBLWHOI Library 

MAY 2 2006 

WOODS HOLE 

-:3achuoetts 02542 



Contents 



Articles 



159—166 Garci'a-Diaz, Mercedes, Jose A. Gonzalez, Marfa J. Lorente, 
and Victor M. Tuset 

Spawning season, maturity sizes, and fecundity in 
blacktail comber (Serranus atricauda) (Serranidae) from 
tfie eastern-central Atlantic 



167-181 Tissot, Brian N., Mary M. Yoklavich, Milton S. Love, Keri York, 
and Mark Amend 

Benthic invertebrates that form habitat on deep banks off 
southern California, with special reference to deep sea coral 



The conclusions and opinions ex- 
pressed in Fishery Bulletin are 
solely those of the authors and do 
not represent the official position of 
the National Marine Fisher-ies Ser- 
vice ( NOAAl or any other agency or 
institution. 

The National Marine Fisheries 
Service (NMFS) does not approve, 
recommend, or endorse any proprie- 
tai*y product or proprietai"y 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 approves, recom- 
mends, or endorses any proprietary 
product or proprietary material 
mentioned herein, or which has 
as its purpose an intent to cause 
directly or indirectly the advertised 
product to be used or purchased 
because of this NMFS publication. 



182—196 Li, Zhuozhuo, Andrew K. Gray, Milton S. Love, Akira Goto, 
Takashi Asahida, and Anthony J. Gharrett 

A key to selected rockfishes (Sebastes spp ) based on 
mitochondrial DNA restriction fragment analysis 



197—203 Torres-Orozco, Ernesto, Arturo Muhlia-Melo, 
Armando Trasviiia, and Sofia Ortega-Garcfa 

Variation in yellowfin tuna iThunnus a/bacares) catches related to 
El Niho-Southern Oscillation events at the entrance to 
the Gulf of California 



204-214 Pepin, Pierre 

Estimating the encounter rate of Atlantic capelin 
iMallotus villosus) with fish eggs, based on stomach 
content analysis 



215—225 Davis, Michelle L., Jim Berkson, and Marcella Kelly 

A production modeling approach to the assessment of 
the horseshoe crab (Limulus polyphemus) population 
in Delaware Bay 



Fishery Bulletin 104(2) 



226-237 Hoff, Gerald R. 

Biodiversity as an index of regime shift in the eastern Bering Sea 



238-246 Pietsch, Theodore W., and James W. Orr 

Tnglops dorothy, a new species of sculpin (Teleostei: Scorpaeniformes: Cottidae) from the 
southern Sea of Okhotsk 



247—255 Chen, Yong, Sally Sherman, Car! Wilson, John Sowles, and Minora Kanaiwa 

A comparison of two fishery-independent survey programs used to define the population structure of 
American lobster (Homarus amencanus) in the Gulf of Maine 



256—277 Walsh, Harvey J., Katrin E. Marancik, and Jonathan A. Hare 

Juvenile fish assemblages collected on unconsolidated sediments of the southeast 
United States continental shelf 



278-292 Goldman, Kenneth J., and John A. Musick 

Growth and maturity of salmon Sharks (Lamna ditropis) in the eastern and western North Pacific, 
and comments on back-calculation methods 



Notes 



293-298 Hurton, Lenka, and Jim Berkson 

Potential causes of mortality for horseshoe crabs (Limulus polyphemus) during the biomedical 
bleeding process 



299-302 Kirshenbaum, Sheril, Scott Feindel, and Yong Chen 

A study of tagging methods for the sea cucumber Cucumana frondosa in the waters 
off Maine 



303-305 Kimura, Daniel K., and Martin W. Dorn 

Parameterizing probabilities for estimating age-composition distributions for mixture models 

306-310 May-Ku, Marco A., Uriel Ordonez-Lopez, and Omar Defeo 

Morphometric differentiation in small luveniles of the pink spotted shrimp (Farfantepeneaus brasiliensis) 
and the southern pink shrimp (F. notialis) in the Yucatan Peninsula, Mexico 



311-320 Cope, Jason M. 

Exploring intraspecific life history patterns in sharks 



321 Guidelines for authors 



159 



Abstract — Blacktail comber iSer- 
ranus atricauda Gunther) is a com- 
mercially important species in the 
Canary Islands fisheries. A total of 
425 individuals were collected and 
histological techniques were used to 
investigate the reproductive data. 
Results indicated that the spawning 
season occurred throughout the year, 
peaking between March and July. 
Individuals reached 50'^'v maturity 
at 19.3 cm TL and 95'~, at 33.1 cm 
TL. Batch fecundity estimates ranged 
from 21,774 to 369,578 oocytes per 
spawning event for specimens from 
22.2 to 39.8 cm TL. Blacktail comber 
was estimated to spawn 42 times/year 
and 26. 5^^^ of individuals spawned, on 
average, every 3.8 days. Estimates 
of potential annual fecundity for the 
species ranged from 0.91 to 15.5 mil- 
lion oocytes, and an average of 5.1 
±4.1 million. 



Spawning season, maturity sizes, 
and fecundity in blacktail comber 
iSerranus atricauda) (Serranidae) 
from the eastern-central Atlantic 



Mercedes Garcia-Diaz 

Jose A. Gonzalez 

Institute Canarlo de Clencias Marinas 
Departamento de Blologia Pesquera 
PO Box 56. E-35200 
Telde (Las Palmas), Spam 

Maria J. Lorente 

Departamento de Biologia Animal (U Morfologia Microscopica) 

Universidad de Valencia 

Moliner 50 

Buriassot (Valencia), Spain 

Victor M. Tuset 

Institute Canarie de Ciencias Marinas 

Departamento de Bielegia Pesquera 

PO Bex 56, E-35200 

Telde (Las Palmas), Spam 

E-mail address (fer V M Tuset, contact author); vlcterta@iccm.rcanaria.es 

Present address (fer V M Tuset) Institute Canane de Ciencias Mannas 

PO Box 56, E-35200 Telde 

Las Palmas, Spain 



Manuscript submitted 20 November 2003 
to the Scientific Editor's Office. 

Manuscript approved for publication 
23 June 2005 by the Scientific Editor. 

Fish. Bull. 104:159-166 (2006). 



Fish of the genus Serranus are syn- 
chronous hermaphrodites; male and 
female tissues are simultaneously 
functional (Smith, 1965). The most 
common style of reproduction is serial 
monogamy, in which individuals are 
solitary during the day before pair- 
ing up and spawning in the late 
afternoon (Fischer, 1986). During 
spawning one fish in each pair func- 
tions as a male and the other as a 
female and cross-fertilization occurs. 
This special characteristic, and the 
possibility of self-fertilization, have 
encouraged detailed studies of the 
gonad structure of the genus Serranus 
(Atz, 1965; Fishelson, 1970; Reinboth, 
1970; Febvre et al., 1975; Zanuy, 1977; 
Brusle, 1983; Abd-el-Aziz and Rama- 
dan, 1990; Garci'a-Diaz et al., 1997, 
2002), although knowledge of other 
reproductive features is scarce. 

Models of dynamic population used 
in the management of fishery re- 
sources and in biological studies of 



fish require knowledge of the repro- 
ductive system (Koslow et al., 1995). 
This includes gonad morphology (ex- 
ternal and cellular description of the 
ovary and testis), reproductive pat- 
tern (hermaphoditism or gonocorism), 
reproductive behavior, reproductive 
cycle, spawning season duration, size 
at maturity, sex ratio, size at sexual 
transition, and fecundity. Of all these 
reproductive features, fecundity is the 
most difficult biological parameter to 
obtain, although it is of interest to 
fishery scientists both as a critical pa- 
rameter for stock assessments based 
on egg production methods and as a 
basic aspect of fish biology and popu- 
lation dynamics. To calculate fecun- 
dity, fish are divided into determinate 
and indeterminate spawners (Hunter 
and Macewicz, 1985a; Hunter et al., 
1992). With indeterminate spawners, 
multiple or serial batches are noted, 
and spawning may take place many 
times during a protracted spawning 



160 



Fishery Bulletin 104(2) 



•40* 



JO 



-20' 



^. . 



season. Therefore fecundity is very dif- 
ficult to determine before the onset of 
the spawning season because 1) there 
is no way to differentiate between the 
oocytes that are going to be shed dur- 
ing the next season from those which 
will remain for future seasons, and 2) 
the number of reserve oocytes that will 
mature during the next spawning season 
cannot be predetermined (Hunter and 
Goldberg, 1980; Hunter et al., 1985). To 
estimate indeterminate annual fecun- 
dity, the mean number of eggs per batch 
and spawning frequency throughout the 
spawning season must be calculated. Ow- 
ing to the complexity of obtaining these 
data, fecundity studies are normally lim- 
ited to determinate spawners or pelagic 
species of substantial economical inter- 
est (Hunter and Goldberg, 1980; Hunter 
and Macewicz, 1985a; Karlou-Riga and 
Economidis, 1997; Murua et al., 1998). 

The blacktail comber (Serranus atricau- 
da Gunther, 1874), is a littoral (3-150 m) 
benthic species, ranging throughout the 
eastern central Atlantic (from the Bay of 
Biscay to Mauritania, the Azores, Madeira, 
and the Canary Islands) and in the west- 
ern Mediterranean. It is, therefore, a spe- 
cies of wide distribution and commercial 
interest in many regions (Bauchot, 1987; Smith, 1990). 
In the Canary Islands, it is an economically important 
species for small-scale inshore fisheries (Perez-Barroso 
et al., 1993). Garcia-Di'az et al. (1999, 2002) determined 
that this species is a functional simultaneous hermaphro- 
dite. The ovary is classified as asynchronous, i.e., various 
stages of oocyte development occur simultaneously (pri- 
mary growth stage, yolk vesicle formation, vitellogenesis, 
oocyte maturation, and mature egg). The histological 
structure of the gonad and of the sperm indicates that 
this species is an externally fertilizing teleost. 

The objective of this article is to increase our under- 
standing of the reproductive biology of the blacktail 
comber in the Canary Islands by estimating its spawn- 
ing season duration, size at maturity, and fecundity. 



Materials and methods 



0' 15" 

Europe 



Atlantic Ocean 




Canary Islands 



500 km 



Location o 
[Serranus 
Atlantic). 



Figure 1 

f sampling areas (shown with dark shading) for blacktail comber 
atricauda) specimens from the Canary Islands (eastern-central 



buffered formaldehyde. After 24-48 h, they were dehy- 
drated, embedded in paraffin wax, sectioned in longi- 
tudinal or cross-sections (4 to 5 .urn thick), and stained 
with Harris "hematoxylin-Puttis" eosin. Oocyte stages 
and spermatogenic cells were classified according to 
Garcia-Diaz et al. (2002). Postovulatory follicles and 
atresia were also characterized according to the crite- 
ria used for Engraulis mordax by Hunter and Goldberg 
(1980) and Hunter and Macewicz (1985b). Maturity 
stages (MS) were determined from histological obser- 
vations (the development of the ovary and testis and 
also the presence and absence of different types of the 
oocytes and spermatocytes) according to Garcia-Diaz 
et al. (1997, 2002) (Table 1). The developmental stages 
of the oocytes were categorized according to Selman 
et al. (1993): primary growth (stage I); cortical alveoli 
formation (stage II); vitellogenesis (stage III); oocyte 
maturation (stage IV); and mature egg (stage V). 



Sampling 

A total of 425 individuals of S. atricauda. ranging from 
15.7 to 43.2 cm total length were sampled monthly from 
commercial catches off the Canary Islands between Sep- 
tember 1992 and November 1994 (Fig. 1). 

Total length (TL), to the nearest centimeter, was 
measured for each specimen, together with the total 
and gutted weight (TW and GW, respectively), gonad 
weight (GNW), and liver weight (LW), with an accu- 
racy of 0.1 g. Gonads were removed and fixed in 10% 



Seasonality of gonad development 

Monthly changes in the five following variables were 
analyzed to determine the spawning season of blacktail 
comber (Garcia-Diaz et al., 1997): 

1 Percent frequency of the maturity stages; this indi- 
cates population changes in gonad development. 

2 Gonadosomatic index (GSI=(gonad weight /gutted 
weight)xlQO): this shows differences in development 
of the gonads with respect to gutted body weight; 



Garcia-Diaz et al ; Spawning season, maturity sizes, and fecundity of Seiranus atncauda 



161 



Table 1 

Description of gonad stages according to Garci'a-Diaz et al. (1997, 2002). 



Maturity stage 



Histological appearance 



I. Immature 

II. Developing virgin, 
or recovering-spent 

III. Developing, maturing 

IV. Ripe 

V. Spent 



Ovary contains oogoniaand oocytes from primary growth (stage I). Testis formed mainly by sperma- 
togonia and primary spermatocytes. Ovary and testis joined by connective tissue. 

Ovary begins to acquire ovarian lamellae. Contains stage-I oocytes and yolk-vesicle-formation- 
stage oocytes (stage II) in advanced phases. Testis arranged in tubules with spermatogonia and 
primary and secondary spermatocytes. 

Ovary with oocytes at all previous stages and oocytes in vitellogenic stage (stage III). Seminiferous 
tubules contain all spermatogenic cells. 

Ovary with oocytes at all previous stages and with maturation oocytes (stage IV), mature and 
hydrated eggs (stage V). Atresic oocytes and postovulatory follicles appear. Testis completely 
mature, tubules filled with spermatozoa (which accumulate in deferent duct next to gonadal wall) 
to be expelled. 

Oocytes undergoing regression and reabsorption. Numerous atresic oocytes. Testis in regression; 
cells appear fused, form semicontinuous mass. 



3 Hepatosomatic index (HSl={Uver weight I gutted 
weight )y.lQO): this estimates the relative size of the 
liver to body weight: 

4 Condition factor (Kn={total weight/ TL '•)x 1000): 
this is as an overall measurement of robustness of 
the fish, b being the exponential of the regression 
TW = aTL^, which is 3.25 according to Tuset et al. 
(2004); 

5 Oocyte diameter (DO): this gives information on the 
cell development of each individual. To obtain this 
value, often fish randomly chosen from the monthly 
samples, the diameters of the first 50 oocytes encoun- 
tered were measured with an ocular micrometer. 
Measurements were taken only of oocytes sectioned 
through the nucleus. 

Length at sexual maturity 

Total length of all individuals was used to estimate the 
size at first maturity and size at mass maturity. These 
are defined as the sizes (TL) at which 50% (TLj^y,^) and 
95% (TLggtj) of all fish sampled are at the relevant 
maturity stage (developing MS III, ripe MS IV, or spent 
MS V) (Garcia-Diaz et al., 1997). The proportions were 
estimated at length classes of 1 cm, and the data were 
fitted to the logistic curve (Pope et al., 1983): 

p = WO/(l + exp{a+bTL)), 

where p = percentage of mature individuals as a func- 
tion of size class (TL); and 

a and b = specific parameters that can change during 
the life cycle. 

A logarithmic transformation was applied to the equa- 
tion to calculate the parameters a and b by means of 
linear regression. 



Reproductive potential 

The pattern of annual fecundity (indeterminate or deter- 
minate) was assessed by oocyte size-frequency distribu- 
tion (Hunter and Macewicz, 1985a). Ten fish ranging 
between 18.1 and 31.5 cm TL in ripe stage were selected 
for analysis of oocyte size-frequency distribution — five 
fish in March and another five in July — to represent 
early and late gonad development (White et al., 2003). 
Because the mean size (^-test, P>0.05) was not signifi- 
cantly different between both samples, one frequency 
distribution was obtained for each month. The diameters 
of the first 500 oocytes encountered were measured with 
an ocular micrometer for each fish. Measurements were 
taken only of oocytes sectioned through the nucleus. 

Gonads of individuals in MS III (developing) and 
MS IV (ripe) were collected to estimate fecundity. One 
lobule was fixed in 10% buffered formaldehyde (24 h) 
for histological study, and the other was weighed and 
preserved in Gilson's fluid to estimate fecundity. Gonads 
with postovulatory follicles were omitted from batch 
fecundity estimates (Hunter et al., 1985). 

The oocyte size-frequency method was applied to de- 
termine the batch fecundity (Hunter et al., 1985). This 
entails counting the number of oocytes in the most de- 
veloped modal group of oocytes and plotting the size-fre- 
quency. In each lobule 200 oocytes were measured with 
the Nikon digital counter SC-112 and data processor DP- 
201 connected to the micrometer stage of a profile projec- 
tor (V-12A, Nikon, Melville, NY). The routine NORMSEP 
(normal distribution separator) of FAO-ICLARM Stock 
Assessment Tools (FISAT II, vers. 1.0.0, FAO, Rome, 
Italy) program was used to separate the most developed 
modal group of oocytes based on Hasselblad's maximum 
likelihood method (Hasselblad, 1966). 

Batch fecundity was estimated gravimetrically by 
using the oocyte size-frequency method (Hunter et 



162 



Fishery Bullelm 104(2) 



al., 1985; Yoneda et al., 2001). Samples were filtered 
through a 100-mm mesh sieve (approximately the mini- 
mum diameter of vitellogenic oocytes, Garcia-Diaz et 
al., 2002) and were carefully washed under running 
distilled water to eliminate tissue remains and fixer. 
We used a sieve made from a piece of nylon plankton 
net inserted between two sections of PVC pipe, 15 cm 
in diameter and 10 cm in depth (Lowerre-Barbieri and 
Barbieri, 1993). A compact mass of oocytes was dried 
for 24 hours on filter paper and weighed precisely to 
0.001 g. Afterwards, a subsample of 0.2 g was selected 
and placed in the count dish, covered with 3-4 drops 
of glycerin, and the number of oocytes in the sample 
was tallied with a hand counter. Batch fecundity was 
considered to be the total number of oocytes within 
the most developed modal group of oocytes (Hunter et 
al., 1985). 

Spawning frequency (the number of spawnings per 
year by an individual) was estimated by dividing the 
duration of the spawning season by the average number 
of days between spawning for all individuals (Hunter 
and Macewicz, 1985a). The duration of the spawning 
season was the number of days between the first (7"^ 
February) and last (16''^ July) occurrence of hydrated 
oocyte or postovulatory follicles. The average number of 
days between spawning was the inverse of the percent 
frequency of hydrated individuals, multiplied by 100 
(Collins et al., 1998). 

The potential annual fecundity estimates (PAFEs) 
were obtained by multiplying batch fecundity by spawn- 
ing frequency, and relative fecundity was defined as 
PAFE divided by individual weight (Hunter et al., 
1992). The relationships between PAFE and TL, PAFE 
and TW, and PAFE and gonad weight (GW) were calcu- 
lated by the following compound equation; 



where X = TL, TW, or GW; and 

a and h = specific parameters. 

A logarithmic transformation was applied to the equa- 
tion to calculate parameters a and b by using linear 
regression (Zar, 1996). 



Results 

Spawning season and maturity sizes 

The specimens revealed the presence of maturing and 
ripe individuals between December and October, exclud- 
ing January because only one individual was sampled 
then. This finding may indicate that the population 
spawns throughout this period, peaking between Febru- 
ary (37.5'7f ) and June (89.8%) (Fig. 2). 

The highest values for GSI (Fig. 3A) occurred between 
March (2.06) and July (3.37) and a peak of maximum 
activity occurred in June (3.93), decreasing over the 
following months. The HSI (Fig. 3B) and Kn (Fig. 3C) 
presented irregular values during the annual cycle. 
Consequently, no correlation was observed between 
either ovarian and liver growth with ovarian and fish 
growth. Oocyte diameter (DO) (Fig. 3D) values varied 
similarly to those from the GSI; highest values occurred 
from March (261.37 .iim) to July (275.07 ^<m) and peaked 
in June (365.85 fim). However, the variability of oocyte 
diameter showed the presence of mature individuals 
(MS III + IV) in the months of September, October, and 
December, according to gonad classification. 

The smallest individual with mature gonads was 
16 cm TL. The maturity curve established a TLgg,; of 
19.3 cm and TLgj;,; of 33.1 cm (Fig. 4). 



PAFE = a ib^). 




May 



I Immature I I Developing virgin or recovenng-speni | Maturing ^| Ripe wg Spent 



Figure 2 

Monthly variation (7c) in the maturity stages in the blacktnil 
comber iSerranus atricauda). 



Reproductive potential 

From the oocyte size-frequency distributions, 
we concluded that the type of fecundity in this 
species is indeterminate evidenced by the lack 
of hiatus between advanced yolked oocytes and 
less mature oocytes (Fig. 5). 

Batch fecundity estimates (7?=28) varied 
between 21,774 and 369,578 oocytes. These 
estimates came from blacktail comber rang- 
ing in total length from 22.2 to 39.8 cm. total 
weight from 127.2 to 896.6 g, gutted weight 
from 122.9 to 834.6 g, and gonadal weight 
from 1.3 to 41.3 g. 

The spawning-frequency estimate for in- 
dividuals from 17.2 to 43.2 cm TL was 42 
times/year, and 26.5% (45/170) of individuals 
spawned at an average of every 3.8 days. 

Potential annual fecundity estimates ranged 
from 0.91 to 15.5 million oocytes, and an av- 
erage of 5.1 ±4.1 million (Fig. 6). Relative 
fecundity varied between 5062 and 20,869 
oocytes per gram of individual, and mean 
relative fecundity was 10,547 ±4148 oocytes. 



Garcia-Diaz et aL; Spawning season, maturity sizes, and fecundity of Serranus alncoudo 



163 



Q 
CO 



If) 



a 
w 




Jan Mar May Jul Sep No/ 

Feb Apr Jun Ago Oct Dec 




Jan Mar May Jul Sep Nov 

Feb Apr Jun Ago Oct Dec 



Q 



CO 

I 



Q 
if) 

CM 

+ 

O 
Q 




Jan Mar May Jul Sep Nov 
Feb Apr Jun Ago Oct Dec 




Jan Mar May Jul Sep Nov 
Feb Apr Jun Ago Oct Dec 



Figure 3 

Monthly variation of the mean and standard deviation (SD) of the gonadosomatic index iGSI) 
(Al, hepatosomatic index (HIS) (B), and condition factor (kn) (C) and oocyte diameter ( DOi 
iDl in the blacktail comber ^Sei-ranus atricaudat from the Canarv Islands. 



Regression analysis showed significant positive correla- 
tion between annual fecundity and total length, total 
weight, and gutted weight. Total weight was the best 
predictor of annual fecundity (PAF£; = 1, 053, 504. 23x1.0 
0^"'; r- = 0.764, P<0.01, «=28); followed by total length 
CPAF£=63.648x4351xl380TL;r2=0.752,P<0.01, ;!=28l; 
and gutted weight (PAP£=l,055,580x 1.00<''^; r'=0.723, 
P<0.01, n=28) (Fig. 6, A-C). 



Discussion 

The histological description of gonad structure is 
fundamental to understanding reproduction. In most 
reproductive fish studies the histological technique is 
omitted because it is too expensive and time-consuming. 
Consequently, in many cases knowledge of reproduc- 
tion is limited or biased (or is both) (Garcia-Diaz et 
al., 2001). 

The spawning period in teleosts is determined from 
changes occurring within the gonad throughout the 
year. The macroscopic or histological observations of the 



100 -| 
90 


  ^ — 


. y^\ TL95,.= 33 1 cm 


80 


/  


70 

— 60 
>, 

o 

§ 50 

g' 40 
it 

30 


 7  i 

-/jL'^o%= 19,3 cmi 

/"i i 


20 


/ 1 


10 


^^ \ \ 


1 5 9 13 17 21 25 29 33 37 41 


Total length (cm) 


Figure 4 


Sexual maturity curve and size-s at maturity for blacktail 


comber iSerraus atricauda) from the Canary Islands, 



164 



Fishery Bulletin 104(2) 



gonad (qualitative methods), somatic indices (e.g., GSI, 
HIS, and Kn), or evolution of oocyte diameter (quan- 
titative methods) (or all four methods) are commonly 
applied. Many authors have disagreed about their ap- 
plications and biological interpretation (De Vlaming et 
al., 1982; Wootton, 1990; Shapiro et al., 1993; Sadovy, 
1996; Karlou-Riga and Economidis, 1997). Our results 
demonstrate that the GSI indicates only the spawning 
peak and not the presence of batch or multiple spawn- 
ing within the species; and the HSI and Kn do not 
exhibit a clear trend throughout the year. Normally, 
variations of these indices (HSI and Kn) imply energy 
storage for reproduction (Hoar et al., 1983; N'Da and 
Deniel, 1993). However, S. atricauda does not require 
such storage because feeding activity is not altered 
during the reproductive period (Morato et al., 2000). 
Oocyte diameter presents a trend similar to that of GSI, 
although its biological interpretation is different. The 
oocyte may begin the vitellogenesis phase and there 



□ March ■July 




Oocyte diameter (microns) 

Figure 5 

Oocyte size-frequency distributions l25-iim intervals) for blacktail 
comber tSerraus atricauda) ovaries from the Canary Islands. 



may be no significant increase in gonad weight. Conse- 
quently. GSI values do not necessarily change signifi- 
cantly when the spawning period has begun. Thus, the 
quantitative indices (GSI, HSI, Kn) show the general 
trend of the population, whereas the oocyte diameter 
provides information about the population as a whole, 
as well as at the individual level. 

The analysis of qualitative and quantitative data in 
S. atricauda seems to indicate that spawning takes 
place throughout the year, although the general popu- 
lation spawns between March and July. In the Canary 
Islands, other members of the genus spawn over long 
periods: eight months in S. cabrilla and nine months in 
S. scriba (Garcia-Diaz et al., 1997; Garcia-Diaz 2003). 
These three species have the longest spawning season of 
all fish studied in the Canaries, although they are also 
the only species whose reproductive patterns have been 
analyzed with histological procedures. Nevertheless, 
environmental or biological factors (or both) must be 
related to this extended spawning period. 
To estimate spawning frequency, the hy- 
drated oocyte method is less time consum- 
ing but the POP method is better because 
the spatial and temporal distribution of the 
postovulatory follicle structures is not con- 
tinuous (Hunter and Goldberg, 1980). Re- 
productive potential measured as potential 
annual fecundity has not been addressed 
in any species of the genus Serranus with 
the POP method to date. It is true that this 
fecundity must be considered as a rough 
estimation because the spawning frequency 
was a preliminary estimation because of 
the scarce number of individuals in sam- 
pling months. Siau and Bouian (1994) cal- 
culated the total fecundity in S. cabrilla 
and S. scriba, considering all the vitel- 
logenesis oocytes. Their method has been 
widely rejected because it is valid only for 



16 
14 - 
12 
10 

8- 

6 

4 

2 





20 22 24 26 28 30 32 34 36 38 40 42 

Total length (cm) 



100 200 300 400 500 600 700 800 900 1000 

Total weight (g) 




100 200 300 400 500 600 700 800 900 

Gutted weight (g) 



Figure 6 

Relationships between potential annual fecundity and total length (A), total weight (B), and gutted weight (C) for black- 
tail comber (^Serranus atricauda) from the Canary Islands. 



Garcia-Diaz et al Spawning season, maturity sizes, and fecundity of Serronus atncauda 



165 



determinate spawners, and both the above-mentioned 
serranids are indeterminate spawners (Garci'a-Diaz et 
al., 1997; Garcia-Diaz, 2003). The fact that the method 
was used for these two indeterminate spawners could 
explain why the values of maximum fecundity obtained 
by these authors for S. cabrllla (441,502 oocytes) and 
S. scriba (54,649 oocytes) were lower than that for S. 
atricauda (15,5 millions oocytes). Consequently, the 
present study enables us to provide new reproductive 
data for blacktail comber, where potential annual fe- 
cundity is essential for future egg production models 
and estimates of spawning stock biomass. 



Acknowledgments 

We wish to thank Jose Ignacio Santana for the collection 
of samples, Antonio Valencia for his help in the building 
of the mechanism to separate the vitellogenic oocytes, 
Gordon Hamilton for proofreading the submission copy 
of the manuscript, and the reviewers for their comments 
and suggestions. 



Literature cited 

Abd-el-Aziz, S. H., and A. A. Ramadan. 

1990. Sexuality and hermaphroditism in fishes. I. Syn- 
chronous functional hermaphroditism in the serranid 
fish Serranus scriba L. Folia Morph. 38:86-103. 
Atz, J. W. 

1965. Hermaphroditic fish. Science 150:789-797. 
Bauchot, M. L. 

1987. Serranidae. /;; Fiches FAO d'identification des 
especes pour les besoins de la peche (revision 1), Medi- 
terranee et Mer Noire, zones de peche 37, II (vertebres) 
(W. Fischer, M. L. Bauchot, and M. Schneider, eds.), p. 
1301-1319. FAO-CEE, Rome. 
Brusle, S. 

1983. Contribution to the sexuality of a hermaphro- 
ditic teleost, Serranus hepatus L. J. Fish Biol. 22: 
283-292. 
Collins, L. A., A. G. Johnson, C. C. Koenig, and M. S. Baker Jr. 
1998. Reproductive patterns, sex ratio, and fecundity in 
gag, Mycteroperca microlepis (Serranidae), a protogynous 
grouper from the northeastern Gulf of Mexico. Fish. 
Bull. 96:415-427. 
De Vlaming, V., G. Grossman, and F. Chapman. 

1982. On the use of the gonadosomatic index. Comp. 
Biochem. Physiol. 73A (11:31-39. 
Febvre, M., M. Michele, and M. Lafaurie. 

1975. Comparaison de la sequence ovogenetique chez des 
Teleosteens (sic) ovipares gonochoriques et hermaph- 
rodites (Mullus, Serranus, Boops). Pubbl. Staz. Zool. 
Napoli 39:140-152. 
Fischer, E. A. 

1986. Mating systems of simultaneously hermaphroditic 
serranid fishes. In Proceedings of the second Indo- 
Pacific fish conference (T. Uyeno, R. Arai, T. Taniuchi, 
and K. Matsuura, eds.), p. 776-784. Ichthol. Soc. Japan, 
Tokyo, Japan. 
Fishelson, L. 

1970. Protogynous sex reversal in the fish Anthias squa- 



mipmnis regulated by the presence or absence of a male 
fish. Nature 227:90-91. 
Garcia-Diaz, M. M. 

2003. Applicacion de tecnicas histologicas para la deter- 
minacion de aspectos reproductores en tres teleosteos 
marinos (Serranidae, Serranus), 228 p. Servicio 
de Publicaciones de la Universidad de La Laguna, Ten- 
erife. 
Garcia-Diaz, M. M., V. M. Tuset, J. A. Gonzalez, and J. Socorro. 
1997. Sex and reproductive aspects in Serranus cabrilla 
(Osteichthyes: Serranidae): macroscopic and histological 
approaches. Mar. Biol. 127:379-386. 
Garcia-Diaz, M. M., M. J. Lorente, J. A. Gonzalez, and V. M. 
Tuset. 

1999. Comparative ultrastructure of spermatozoa of three 
marine teleosts of the genus Serranus: S. atricauda, S. 
cabrilla and S. scriba. J. Submicrosc. Cytol. Pathol. 
31:503-508. 

2001. Application of histology in reproductive biology 
of fishery studies. In Perspective in comparative 
endocrinology: unity and diversity (H. J. T. H. Goos, 
R. K. Rastogi, H. Vaudry. and R. Pierantoni, eds.), p. 
1095-1100. Monduzzi Editore, Bologna, Italy. 

2002. Morphology of the ovotestis of Serranus atricauda 
(Teleostei, Serranidae). Aquat. Sci. 64:87-96. 

Hasselblad, V. 

1966. Estimation of parameters for a mixture of normal 
distribution. Technometrics 8: 431-444. 
Hoar, W. S., D. J. Randall, and E. M. Donaldson (eds.). 

1983. Fish physiology, vol. IX (part B), 477 p. Academic 
Press, New York, NY. 
Hunter, J. R., and R. Goldberg. 

1980. Spawning incidence and batch fecundity in Northern 
anchovy Engraulis mordax. Fish. Bull. 77:641-652. 
Hunter, J. R., and B. J. Macewicz. 

1985a. Measurement of spawning frequency in multiple 
spawning fishes. In An egg production method for esti- 
mating spawning biomass of pelagic fish: application 
to the northern anchovy, Engraulis mordax (R. Lasker, 
ed.), p. 79-94. NOAA Tech. Rep. NMFS 36. 
1985b. Rates of atresia in the ovary of captive and wild 
anchovy, Engraulis mordax. Fish. Bull. 83: 119-136. 
Hunter, J. R., N. C. R. Lo, and R. J. H. Arnold. 

1985. Batch fecundity in multiple spawning. /;; An egg 
production method for estimating spawning biomass 
of pelagic fish: application to the northern anchovy, 
Engraulis mordax (R. Lasker, ed.), p. 66-77. NOAA 
Tech. Rep. NMFS 36. 
Hunter, J. R., B. J. Macewicz, N. C. H. Lo. and C. A. Kimbrell. 

1992. Fecundity, spawning, and maturity of female Dover 
sole Microstomas pacificus, with an evaluation of assump- 
tions and precision. Fish. Bull. 90:101-128. 

Lowerre-Barbieri, S. K., and L. R. Barbieri. 

1993. A new method of oocyte separation and preservation 
for fish reproduction studies. Fish. Bull. 91:165-170. 

Karlou-Riga, C, and P. S. Economidis. 

1997. Spawning frequency and batch fecundity of horse 
mackerel, Trackurus trachurus (L.), in the Saranikos 
Gulf (Greece). J. Appl. Ichthyol. 13:97-104. 
Koslow, J. A., J. Bell, P. Virtue, and D. C. Smith. 

1995. Fecundity and its variability in orange roughy: 
effects of population density, condition, egg size, and 
senescence. J. Fish Biol. 47:1063-1080. 
Morato, T., R. S. Santos, and J. P. Andrade. 

2000. Feeding habits, seasonal and ontogenic diet shift 
of blacktail comber, Serranus atricauda (Pisces: Serra- 



166 



Fishery Bulletin 104(2) 



nidae). from the Azores, north-eastern Atlantic. Fish. 
Res. 49:51-59. 
Murua, H., L. Motos, and P. Lucio. 

1998. Reproductive modality and batch fecundity of the 
European hake [Merluccius meiiucciiin L.) in the Bay 
of Biscay. CalCOFI Rep. 39:196-203. 
N'Da, K., and C. Deniel. 

1993. Sexual cycle and seasonal changes in the ovary of 
the red mullet, MiiUus surmuletus, from the southern 
coast of Brittany. J. Fish Biol. 43:229-244. 
Perez-Barroso, F. I., S. Jimenez, I. J. Lozano, J. A. Gonzalez, and 
J. I. Santana. 

1993. Estudio preliminar de la pesqueria artesanal 
de Serranus atru rnida Gunther, 1974 y de Serranus 
cabrilla (Linnaeus. 1758) (Osteichthyes, Serranidae) en 
el archipielago canario. Publ. Esp. Inst. Esp. Oceanogr. 
11:365-373. 
Pope, J. A., A. R. Marggets, J. M. Hamley. and E. F. .\kyuz. 

1983. Manual of methods for fish stock assessment. Pt. 
3. Selectivity of fishing gear. FAO Fish.Tech. Pap. 41 
(l):l-65. 
Reinboth, R. 

1970. Interse.xuality in fishes. Men. Soc. Endocrinol. 
18:515-543. 
Selman, K., R. A. Wallace, A. Sarka, and X. Qi. 

1993. Stages of oocyte development in the zebrafish, 
Brachydanio rerio. J. Morphol. 218:203-224. 
Sadovy. Y. J. 

1996. Reproduction of reef fishery species. In Reef fish- 
eries, vol. 2 (N. V. C Polunin, and C. M. Roberts, eds.), 
p. 15-59. Chapman and Hall, London. 
Shapiro, D. Y., Y. Sadovy, and A. McGehee. 

1993. Size, composition, and spatial structure of the 
annual spawning aggregation of the red hind, Epi- 
nephelus guttatus (Pisces: Serranidae). Copeia 1993: 
367-374. 



Siau, Y., and A. Bouain. 

1994. Variations in spawning of two species of coastal 
hermaphrodite fishes, genus Serranus, related to their 
bathymetric distribution. Oebalia 10:1-20. 
Smith, C. L. 

1965. The patterns of sexuality and classification of 

serranid fishes. Am. Mus. Novit. 2207:1-20. 
1990. Serranidae. In Check-list of the fishes of the 
eastern tropical Atlantic, vol. 2 (J. C. Quero, J. C. 
Hureau, C. Karrer, A. Post, and L. Saldanha, eds. I, p. 
695-706. UNESCO-SEI-JNICT, Lisbon, Portugal. 
Tuset, V. M., J. A. Gonzalez, I. J. Lozano, and M. M. Garcia-Diaz. 
2004. Age and growth of the blacktail comber, Serranus 
atricauda (Serranidae), off the Canary Islands (central- 
eastern Atlantic). Bull. Mar. Sci. 74:53-68. 
White, G. G., T.A. Munroe, and H. M. Austin. 

2003. Reproductive seasonality, fecundity, and spawn- 
ing frequency of tautog iTautoga onitis) in the lower 
Chesapeake Bay and coastal waters of Virginia. Fish. 
Bull. 101:424-442 
Wootton, R. J. 

1990. Ecology of teleost fishes, 404 p. Chapman and 
May, London. 
Yoneda, M., M. Tokymura, H. Fujita, N. Takeshita, K. Takeshita. 
M. Matusyama and, S. Matsuura. 

2001. Reproductive cycle, fecundity, and seasonal dis- 
tributions of the anglerfish Lophius Utulon in the East 
China and Yellow seas. Fish. Bull. 99:56-370. 
Zanuy, S. 

1977. Induccion a la puesta y estudio de la ovogenesis en 
un teleosteo marine: Paracentropristis cabrilla L. Inv. 
Pesq. 41:337-384. 
Zar, J. H. 

1996. Biostatistical analysis, 3^'^ ed., 662 p. Prentice- 
Hall International, LTpper Saddle River, New Jersey. 



167 



Abstract — There is increasing inter- 
est in the potential impacts that fish- 
ing activities have on megafaunal 
benthic invertebrates occurring in 
continental shelf and slope ecosys- 
tems. We examined how the structure, 
size, and high-density aggregations 
of invertebrates provided structural 
relief for fishes in continental shelf 
and slope ecosystems off southern 
California. We made 112 dives in 
a submersible at 32-320 m water 
depth, surveying a variety of habi- 
tats from high-relief rock to flat sand 
and mud. Using quantitative video 
transect methods, we made 12,360 
observations of 15 structure-form- 
ing invertebrate taxa and 521,898 
individuals. We estimated size and 
incidence of epizoic animals on 9105 
sponges, black corals, and gorgonians. 
Size variation among structure-form- 
ing invertebrates was significant and 
909c of the individuals were <0.5 m 
high. Less than 1% of the observa- 
tions of organisms actually shelter- 
ing in or located on invertebrates 
involved fishes. From the analysis of 
spatial associations between fishes 
and large invertebrates, six of 108 
fish species were found more often 
adjacent to invertebrate colonies than 
the number of fish predicted by the 
fish-density data from transects. This 
finding indicates that there may be 
spatial associations that do not neces- 
sarily include physical contact with 
the sponges and corals. However, the 
median distances between these six 
fish species and the invertebrates were 
not particularly small (1.0-5.5 m). 
Thus, it is likely that these fishes and 
invertebrates are present together in 
the same habitats but that there is not 
necessarily a functional relationship 
between these groups of organisms. 
Regardless of their associations with 
fishes, these invertebrates provide 
structure and diversity for continental 
shelf ecosystems off southern Califor- 
nia and certainly deserve the atten- 
tion of scientists undertaking future 
conservation efforts. 



Benthic invertebrates that form habitat 
on deep banks off southern California, 
with special reference to deep sea coral 



Brian N. Tissot 

Program in Environmental Science and Regional Planning 
14204 NE Salmon Creek Ave. 
Washington State University 
Vancouver, Washinton 98686-9600 
E-mail address tissot'a Vancouver wsu edu 



Mary M. Yoklavich 

Fisheries Ecology Division 

Southwest Fisheries Science Center 

National Marine Fisheries Service, NOAA 

Santa Cruz Laboratory 

110 Shaffer Road 

Santa Cruz, California 95060 

Milton S. Love 

Marine Science Institute 

University of California 

Santa Barbara, California 93106 

Keri York 

Program in Environmental Science and Regional Planning 
14204 NE Salmon Creek Ave. 
Washington State University 
Vancouver, Washington 98686-9600 

Mark Amend 

Southwest Fisheries Science Center 

National Marine Fisheries Service, NOAA 

Santa Cruz Laboratory 

110 Shaffer Road 

Santa Cruz, California 95060 



Manuscript submitted 8 December 2004 
to the Scientific Editor's Office. 
Manuscript approved for publication 
21 July 2005 by the Scientific Editor. 
Fish. Bull. 104:167-181 (2006). 



Science and conservation communi- 
ties are increasingly interested in the 
potential impacts that fishing activi- 
ties have on megafaunal benthic inver- 
tebrates found in continental shelf and 
slope ecosystems (Dayton et al., 2002; 
NRC, 2002; Malakoff, 2004; Roberts 
and Hirshfield, 2004; Rogers, 2004). 
Megafaunal invertebrates (>5 cm in 
height) contribute significantly to bio- 
diversity, play important functional 
ecological roles, and can be indica- 
tors of long-term environmental condi- 
tions (e.g., Riedl, 1971; Palumbi, 1986; 
Brusca and Brusca, 1990). Moreover, 
because large invertebrates, such 
as sponges and corals, enhance the 



diversity and structural component 
of fish habitat and are vulnerable to 
impacts by at least some fisheries, 
they also may signify habitat areas 
of particular concern (HAPC) and as 
such would be protected under the 
Magnuson-Stevens Fishery Conser- 
vation and Management Act (Freese, 
2001; Etnoyer and Morgan^). 



' Etnoyer, P., and L. Morgan. 2003. Oc- 
currences of habitat-forming deep 
sea corals in the Northeast Pacific 
Ocean. Technical Report, NOAA Office 
of Habitat Conservation, 31 p. Marine 
Biologv Conservation Institute, 15806 
NE 47'h Ct., Redmond, WA 98052. 



168 



Fishery Bulletin 104(2) 



Deep-sea corals, such as gorgonians (sea fans), antipa- 
ththarians (black corals), scleractinians (stony corals), 
and hydrocorals, are of particular interesc because they 
are often long-lived and slow-growing (Andrews et al., 
2002; Heifetz, 2002), poorly studied (Etnoyer and Mor- 
gan'), and in certain situations vulnerable to human 
activities (e.g., mobile fishing gear) (Watling and Norse 
1998; Freese et al., 1999; Krieger, 2001; Dayton et al., 
2002: Fossa et al., 2002; NRC, 2002;). Other megafau- 
nal invertebrates, such as crinoids, basket stars, and 
sponges also may enhance the structural components 
offish habitat (Puniwai, 2002) and may be disturbed 
or destroyed by some fishing activities 'Freese, 2001; 
Krieger, 2001). 

The potential for invertebrates to add functional 
structure to benthic communities has centeicd large- 
ly around their size and complex morphology. A size 
threshold of 1 m has often been used as an indicator 
of structure-forming species because marked changes 
in benthic community structure have been observed 
in areas where rocky substrata exceed 1 m (Lissner 
and Benech, 1993). The complex structure of deep-sea 
corals also has been discussed as an important factor 
that contributes to microhabitat diversity (Krieger and 
Wing, 2002; Etnoyer and Morgan'). In this article, be- 
sides forming complex structure and large size, we also 
believe that megafaunal invertebrates form structure 
if they aggregate in high numbers, especially in areas 
of low relief For example, aggregations of sea urchins 
and sea pens may provide significant structural relief 
for fishes in mud- and sand-dominated habitats (Bro- 
deur, 2001). 

An important question is the extent to which struc- 
ture-forming invertebrates are ecologically important to 
fishes, especially those of economical value. Most studies 
have focused on "associations" between structure-form- 
ing invertebrates and fishes as a measure of ecological 
importance at several spatial scales. Fishes have been 
considered to be associated with invertebrates if they 
are found in the same trawl sample (Heifetz, 2002), 
if fishing is higher in areas with corals than without 
corals (Husebo et al., 2002), if they are found together 
within similar habitats observed from a submersible 
(Hixon et al.-), or if they are observed "among or within 
1 m" from corals (Krieger and Wing, 2002). In this 
article we investigated association at three different 
levels: 1) fishes that are physically touching large in- 
vertebrates; 2) fishes that are found statistically more 
frequently near large invertebrates in relation to their 
overall abundance patterns; and 3) fishes that are found 
as nearest neighbors to large invertebrates. 

The goal of this study was to describe patterns in 
the density, distribution, and size of structure-forming 
megafaunal invertebrates on deep rocky banks and 



outcrops off southern California. Given the recent inter- 
est in these organisms as potentially important habitat 
for groundfishes, and thus targets for protection from 
fishing activities, these organisms deserve a critical 
examination of their potential to contribute structure 
to continental shelf and slope ecosystems and an ex- 
amination of their associations with fishes and other 
marine organisms. Accordingly, our specific objectives 
were the following: 

• Identify structure-forming invertebrates based on cri- 
teria of size, morphological complexity, and density; 

• Quantify the density and size distributions of these 
invertebrates according to depth and substratum 
types; 

• Quantify associations between large, structure-form- 
ing invertebrates and other organisms, particularly 
fishes; and 

• Assess the health of these organisms in terms of 
obvious physical damage. 



Materials and methods 

Underwater surveys were conducted off southern Califor- 
nia by using nonextractive video-transect methods and 
direct observations from an occupied research submers- 
ible (Delta) from 8 October to 6 November 2002. These 
surveys were conducted as part of a larger investigation 
into the abundance, size, and distribution of cowcod 
(Sebastes levis) and associated benthic fishes and habi- 
tats inside and around the newly established Cowcod 
Conservation Areas (CCAs) off southern California (Fig. 
1). The CCAs, which encompass 14,750 km- and are 
closed to groundfish harvest in water depth >37 m, were 
established in 2001 to assist in rebuilding the depleted 
cowcod population off southern California. 

Digital, georeferenced maps of seafloor substratum 
types, interpreted from side-scan sonar, multibeam ba- 
thymetry, seismic reflection, and other past geophysical 
surveys, were used to identify and select sites of rocky 
habitats (Greene et al.'). We attempted to restrict the 
substratum types to mixed sediment and rock and to 
30-330 m depth (i.e., likely cowcod habitat). 

The Delta submersible was tracked by using an ORE 
Trackpoint II plus (ORE Offshore, West Wareham, MA) 
USBL system and WINPROG (vers. 3.1, FUGRO. San 
Diego, CA) software. We linked the tracking system to 
our ArcView"^ GIS (vers. 3.2, ESRI Corp., Redlands, CA) 
seafloor mapping project and tracked the submersible 
real-time in relationship to depth and seafloor habitat 
maps. 



2 Hixon, M. A., B. N. Tissot, and W. G. Pearcy. 1991. Fish 
assemblages of rocky banks of the Pacific northwest, Heceta, 
Coquille, and Daisy Banks. OCS Study MMS 91-0052. 
410 p. U.S. D.I. Minerals Management Service 770 Paseo 
Camarillo. 2"'' Floor, Camarillo, CA 93010. 



3 Greene, H. G., J. J. Bizzarro, D. M. Erdey. H. Lopez, L. 
Murai, S Watt, and J. Tilden. 2003. Essential fish habi- 
tat characterization and mapping of California continental 
margin. Moss Landing Marine Laboratories Technical Pub- 
lication Series No. 2003-01, 29 p., 2 CDs. Moss Landing 
Marine Laboratories, 8272 Moss Landing Rd., Moss Landing, 
CA 9.5039. 



Tissot et al Invertebrates on deep banks off souttnern California 



169 



120"30'W 



120=W 



119 'SOW 



119°W 



118'30'W 



118°W 



117°30'W 117-W 



34°30'N 



34 N 



33°30'N 



33°N 



32°30'N 



Point Conception 



Saci Miguel Is. Santa Cruz Is 





Anacapa Is ^ 

"Harrison'iS^^i^ 
'- " Reef 



A 



\, Hidden Reef 



/?T n V ^ ' 'ft J—i TJ»~^v wcv"' 

(lx "^ S Potato ©imk. • S^N1«pIas.ls7 



Kidney BanH<- ~ ^ , — 
Santa-Bjirbir jTs^ .^i 
^■^ Ost)Orn^ank 




25 






;-fV 






..,., 




<> 








V.\ :■■ 


% , 



^^pheri-y Bank 



50 

I L. 



100 km 



120°30W 



120°W 



119"30W 



119''W 



118'30'W 



118°W 



117°30'W 




34°30'N 



34°N 



33 30N 



33°N 



32'>30N 



117°W 



Figure 1 

Locations of survey sites and dives (black dots) conducted inside and outside the cowcod conservation areas 
(delineated by outlined boxes) off southern California in 2002 to ascertain patterns in the density, distribution. 
and size of habitat-forming invertebrates. 



We documented dives continuously during daytime 
hours with an externally mounted high-8 video camera 
positioned above the middle viewing-porthole on the 
starboard side of the submersible. The observer ver- 
bally annotated all tapes with observations on fishes, 
invertebrates, and physical habitats. Two parallel lasers 
were installed 20 cm apart on either side of the external 
video camera for estimating fish and invertebrate sizes 
and delineating a 2 m-wide belt transect for counting 
fishes and invertebrates. We used personal dive sonar 
from inside the submersible to verify the width of swath 
for the belt transects. Digital still and video cameras 
were used inside the submersible to help document 
fishes, invertebrates, and habitats. 

We defined "habitat" using a combination of nine dif- 
ferent categories of substratum and standard geologi- 
cal definitions (see Stein et al., 1992; Yoklavich et al., 
2000). In order of increasing particle size or relief, these 
substrata were the following: mud (code M), sand (S). 
gravel (G), pebble (P), cobble (C), boulder (B), continu- 
ous flat rock (F), rock ridge (R), and pinnacles (T). A 
two-character code was assigned each time a distinct 
change in substratum type was noted along the tran- 
sect, thus delineating habitat patches of uniform type. 
The first character in the code represented the substra- 



tum that accounted for at least 50% of the patch, and 
the second character represented the substratum ac- 
counting for at least 20% of the patch (e.g., "BC" repre- 
sented a patch with at least 50% cover by boulders and 
at least 20% cover by cobble). Each habitat patch also 
was assigned a code based on the degree of its three- 
dimensional structure as defined by the vertical relief of 
the physical substrata from the seafloor. Habitats were 
coded as l=low (<1 m), 2=moderate (1-5 m), or 3=high 
relief (>5 m). Patches less than 10 seconds in duration 
were not recorded. The area of each habitat patch was 
determined by calculating the distance between the 
beginning and end of habitat patches with ArcGISC© and 
multiplying by the width of the transect (2 m). 

A total of 58 different types of habitat patches were 
observed across all dives. These data were analyzed by 
a cluster analysis (Euclidean distance, group average 
method) by using the abundances of the 20 most com- 
mon invertebrate species, and the resulting dendrogram 
was used to pool the number of codes into the 17 most 
distinct habitat types exhibiting a similarity of >50%. 

Direct counts of megafaunal invertebrates were made 
from videotapes within each habitat patch; patch areas 
varied from 12-1472 m-. Densities of invertebrates were 
estimated by dividing the total number of individuals 



170 



Fishery Bulletin 104(2) 



Table 1 






Number of submersible dives and habitat 


patches sur- 


veyed for structure-forming 


invertebrates 


on rocky out- 


crops inside and outside the cowcod conservation areas in | 


the southern CaUfornia borderland. 












No. of 


Study site No 


. of dives 


habitat patches 


Inside the cowcod 








conservation areas 








43-fathom 


4 




142 


Cherry Bank 


8 




266 


Hidden Reef 


5 




156 


Kidney Bank 


21 




474 


Osborn Bank 


5 




199 


Potato Bank 


4 




113 


San Nicolas Island 


23 




785 


Santa Barbara Island 


9 




203 


Tanner and Cortes banks 


21 




587 


Outside cowcod 








conservation areas 








Harrison's Reef 


9 




198 


Dume Canyon 


3 




66 


Total 


112 




3189 



in each group, identified to the lowest taxonomic level, 
by the area of their associated habitat patches. For the 
larger invertebrates we recorded the geographic position 
and estimated maximum height of each solitary sponge, 
gorgonian. and black coral. We also noted the color of 
black corals. We made observations on the occurrence 
of animals (i.e., epizoids) found directly on sponges, gor- 
gonians, and black corals, and also noted any damaged 
or dead individuals. Voucher specimens were collected 
to assist in taxonomic identification. 

To quantify fish-invertebrate associations we used 
ArcGIS - to estimate the distance between each sponge, 
gorgonian, and black coral invertebrate and the nearest 
fish. These data were compared to the total number 
of fishes counted in habitats that contained sponges, 
gorgonians, and black corals by using a chi-square 
test to look for significant differences in the frequency 
of fish observed near corals in relation to overall fish 
abundance. 



Results 

We completed 112 dives and surveyed 3189 habitat 
patches (Table 1), covering 26.1 hectares at 32-320 m 
depths (median depth of 110 m). The distributions of 
number of patches and of surface area of habitats were 
similar except for sand (SS), sand-gravel (SG). and 
mud (MM) habitats, all of which had greater surface 
areas in relation to number of patches (Fig. 2, A and B). 
Overall, cobble-sand (CS), sand, mud, and sand-gravel 




TT RR RM BB BC BS CB MB SB CS SC SP MG SG MM MS SS 



CD 2 - 



26 1 hectares 



-l?l 



m 



ra 



kl 



FU 



I 



a 



■4 

TT RR RM BB BC BS CB MB SB CS SC SP MG SG MM MS SS 



^, 2 



n n 



TT RR RM BB BC BS CB MB SB CS SC SP MG SG MM MS SS 

Habitat type 

Figure 2 

Characteristics of habitat patches surveyed on southern 
California rocky banks. (A) Number of patches in each 
substratum type. (B) Total area of each substratum 
type. (C) Mean relief (±1 standard error). See text for 
description of method and habitat codes. See page 169 
for definitions for the substrate abbreviations along 
the .V axis. 



constituted the largest habitat areas, but cobble-sand 
and sand were the most frequent habitat types. Verti- 
cal physical structure varied among habitat types; the 
highest structure was found in high-relief rock areas 
(TT to BS) and lower structure was found in low-relief 
mixed rock (CB to SP) and mixed sediment areas (MG 
to SS; Fig. 2C). The frequency of patches of each sub- 
stratum type varied by depth (Fig. 3). The incidence of 
most high-relief rock categories (TT, RR, BB, and BS) 
decreased with depth, and the occurrence of mud-domi- 
nated habitat patches (MB, MG, MS, MM) increased 
with depth. 

Overall, 12,360 observations were made on 521,898 
individuals from 15 taxa of megafaunal, structure-form- 
ing invertebrates; during these observations, estimates 
of size and incidence of epizoic animals on 9105 sponges, 
black corals, and gorgonians were made (Table 2). The 
most common structure-forming invertebrates (98% 
of total) included the crinoid Florometra serratissima 
(40%), the brittle star iOphiacantha spp. [33%]), bra- 
chiopods (order Terebratulida [11%]), the white sea ur- 
chin {Lytechinus anamesus [9%]), the fragile sea urchin 



Tissot et al : Invertebrates on deep banks off southern California 



171 



(Allocentrotus fragilis [4%]), and sea pens 
(suborder Subselliflorae; [2%]; Fig. 4). 

The density of common structure-forming 
invertebrates was variable across habitat 
types; some species were found over a wide 
range of habitats. Crinoids and basket stars 
were found on all 17 habitat types but were 
most dense on either high-relief rock or low- 
relief mixed rock (Fig. 5). In contrast, brittle 
stars and brachiopods were dense in low-re- 
lief mixed rock but rare or absent in low-re- 
lief mixed sediment. White sea urchins were 
most dense in habitats with sand, whereas 
fragile sea urchins were most dense in habi- 
tats with mud. White-plumed anemones were 
most dense in mud-gravel habitats, and sea 
pens were most dense in low-relief mixed- 
sediments (Fig. 5). 

Deep sea corals and sponges were the larg- 
est structure-forming invertebrates but were 
relatively uncommon (27^ of total) (Table 2). 
Gorgonians were difficult to distinguish and 
were categorized into one group (order Gor- 
gonacea). The black coral is a new species 
that recently has been described and named 
the Christmas tree coral (Antipathes den- 
drochristos) (Opresko, 2005). Sponges were 
categorized into five groups based on their 
structure and shape: flat, barrel, shelf, vase, 
and foliose sponges (Fig. 6). 

Gorgonians and black corals were most 
dense on low-relief mixed rock areas (Fig. 7). 
However, gorgonians were found in only four 
habitat types at 144-163 m depth, whereas 
black corals were found on 12 habitat types 
at 100-225 m depth, including pinnacle, boulder, and 
sand areas. These differences may be due to the un- 
equal number of observations (i.e., 27 gorgonian vs. 135 
black coral colonies). 

The five morphological groups of sponges displayed 
broad distributions across habitat types but were espe- 
cially dense on high-relief rock and low-relief mixed rock 
(Fig. 7). Flat, barrel, vase, and foliose sponges were found 
in all habitats; shelf sponges were found in all habitats 
except MM, MS, and SS. Foliose sponges were found 
at significantly deeper depths (mean=191 m; SE = 53; 
n=1259) than were other sponge groups (pooled mean=152 
m; SE = 0.6; n=7545), which were not significantly dif- 
ferent from each other (Kruskal-Wallis //=594; df=4; 
P<0.01). Generally sponge size increased with increasing 
depth, although the correlation was low (/■=0.07; P<0.001; 
n = 6551). Although sponges were found throughout the 
study area, gorgonians and black corals were restricted in 
their distribution to a small number of sites (Fig. 8). 

Structure-forming invertebrates displayed wide varia- 
tion in size; maximum height ranged from 4 cm for 
brachiopods to 2.5 m for black corals (Table 2). There 
was no significant correlation between size of the in- 
vertebrate taxa and structural relief of the substratum 
types (r=0.28, P=0.30, /z = 15). 



20 



10 



< 90 m. n =770 patcties 



IJ 



nH^^n 



r-1 n 



20 - 


91-125 m. n = 


1087 patcties 
















10  


n 


, , 


p 


















in rrn ^ 






n 




n 




n 




n 







20 - 



10 - 



126-175 m. n=751 patcties 



n 



CL 



- n 



20 - 



10 



>175m, n =570 patches 



_a 



n^rin 



LX 




TT RR RM BB BC BS CB IV1B SB CS SC SP IVIG SG IVIIV1 fVIS SS 
Habitat code 

Figure 3 

The frequency of habitat patches of each substratum type, stratified 
by depth. See page 169 for definitions for the substrate abbrevia- 
tions along the .r axis. 



Gorgonians and black corals had different size distri- 
butions (Fig. 9). Black corals ranged from 10-250 cm 
in height (mean=33.6; SE=1.9; n=195) and most indi- 
viduals were in the 10-50 cm size category. Individuals 
were found in three color forms: gray-to-white (SO'/f ), 
rusty-brown-to-red (47%), and gold (3%). Gorgonians 
ranged in size from 10 to 40 cm (mean=21.7; SE = 1.2; 
n=27) and were found in multiple morphological forms 
from elongate to fan-like (Fig. 6). 

Sponges displayed similar size distributions to those 
of gorgonian corals, but had different mean and max- 
imum sizes (Fig. 9). Mean sizes of flat, barrel, and 
foliose sponges were not significantly different from 
each other (pooled mean=19.8 cm; SE = 0.1; ;!=7373) but 
were significantly smaller than vase and shelf sponges 
(pooled mean = 20.9 cm; SE = 0.2; ?! = 1289), which were 
not different from each other (one-way ANOVA; F=4.52; 
df=4,8657; P=0.001). The maximum observed height 
was 50 cm for shelf sponges, 60 cm for foliose sponges, 
and 100 cm for barrel, flat, and vase sponges (Fig. 9). 

Most (98.2% by number) sponges, gorgonians, and 
black corals by number did not have any other organ- 
isms living on them (Table 3). Overall, crinoids (1.4%) 
were most commonly associated with these large in- 
vertebrates, followed by sponges (0.1%), and nine other 



172 



Fishery Bulletin 104(2) 











Table 2 
















Significant criteria of structure-forming invertebrates on rocky outcrops off southern California. "X" indicates taxa that exhibit 
large size, complex morphological shape, and high-density aggregations. SE is standard error and n is the total number of obser- 
vations per taxon. The sample sizes for density and depth calculations were reduced because of incomplete observations on all 
individuals. Organisms are ordered from largest to smallest group. 


Taxa 


n 




Criteria 




Density 
(no. /hectare) 


Maximum 
height (cm) 


Depth 


(m) 


Mean physical 
structural relief 


Size 


Complex 
morphology 


High 
density 


Mean SE 


Mean 


SE 


Black coral 13.5 
tAntipathes dendrochristos ) 


X 


X 




10 


4 




250 


183 


24 


1.7 


Flat sponge 


4240 


X 






312 


53 




100 


149 


43 


1.9 


Barrel sponge 


2138 


X 






163 


28 




100 


151 


46 


2.0 


Vase sponge 


1167 


X 






90 


16 




100 


163 


44 


1.9 


Sea pen 

(Subselliflorae) 


9726 


X 




X 


819 


153 




100 


157 


61 


1.4 


Basket star 

iGorgonocephalus eucriemis) 733 


X 


X 




58 


12 




90 


162 


44 


1.7 


White-plumed anemone 
iMetridium farcuneni 


57 


X 






5 


2 




80 


161 


46 


1.9 


Foliose sponge 


1259 


X 






96 


25 




60 


191 


53 


2.0 


Shelf sponge 


139 


X 






10 


3 




50 


158 


47 


2.1 


Gorgonian 
1 Gorgonacea I 


27 




X 




3 


3 




40 


159 


4 


1.2 


Crinoid 

[Florometra serratissima 


174,231 

) 




X 


X 22,045 


8466 




35 


146 


46 


1.6 


Brittle star (Ophiuridae 


207,667 






X 17,786 


4301 




15 


148 


47 


1.8 


Fragile sea urchin 
(Allocentrntus fragilif:) 


18,363 






X 


2031 


668 




8 


185 


45 


1.8 


White sea urchin 
iLytechinu.s anamesui^i 


45,092 






X 


4631 


1388 




4 


85 


20 


1.2 


Brachiopod 

1 Order Terebratulida ) 


56,924 






X 


3623 


1705 




4 


100 


17 


1.8 















Table 3 
















Associations of 


organisms 


with 


large structure-form 


ng invertebrates on rocky banks off southern California. 














Associated organisms {% of total observations) 




















Egg 




Basket 


Brittle 


Sea 








Taxa 


u 


None 


C 


inoid 


s Sponges 


Fishes 


cases 


Crabs 


stars 


stars 


stars 


Anemone 


Algae 


Salps 


Foliose sponge 


1262 


99.0 




0.6 


0.3 





























Flat sponge 


4043 


99.5 




0.5 


<0.1 











<0.1 

















Barrel sponge 


2068 


98.0 




1.9 

















0.1 


0.1 











Shelf sponge 


134 


99.3 




0.7 
































Vase sponge 


1155 


96.1 




3.6 





0.1 


0.1 





0.1 

















Gorgonian 


27 


100 





































Black coral 


194 


84.7 




6.8 


3.1 


1.3 





1.8 


0.4 


0.4 





0.4 


0.4 


0.4 


Total 


8883 


98.2 




1.4 


0.1 


<0.1 


<0.1 


<0.1 


<0.1 


<0,1 


<0.1 


<0.1 


<0.1 


<0.1 



Tissot et al : Invertebrates on deep banks off soutfiern California 



173 






Figure 4 

Some major structure-forming invertebrates on southern California rocky banks: (A) crinoids {Flo- 
rometra fierratissima), (B) basket stars (Gorgonocephalus eucnemis), (C) brittle stars (Ophiacantha 
spp.), (D) brachiopods (Terebratulida i, lEl white sea urchins [Lytechinus anamesus), (F) white-plumed 
anemones (Metridium farcimen), (G) fragile sea urchins (AUocentrotus fragilis). and IH) sea pens 
iSubselliflorae). 



174 



Fishery Bulletin 104(2) 



taxa fall <0.1'7(). Black corals had the largest incidence 
of associated animals (15.3% of individuals), followed by 
vase (3.9%), barrel (2%), foliose (1%), shelf (0.7%), and 
flat (0.5%) sponges. Fish were most commonly observed 
on black corals (1.3%) but were also observed on vase 
sponges, including an attached egg case. No organisms 
were observed living on gorgonians. 

The frequency of fish species near sponges, gorgoni- 
ans, and black corals was significantly different from 
the frequency of those same species found elsewhere 
along transects (chi-square, all P<0.01; Table 4). Of the 
108 species adjacent to these large invertebrates, six 
species were found at significantly higher frequencies 
than predicted by their density along transects: cowcod, 
bank rockfish iSebastes rufus). swordspine rockfish 
iSebastes ensifer). shortbelly rockfish iSebastes jordani), 
pinkrose rockfish iSebastes simulator), and members of 
the rockfish subgenus Sebastomus (Table 4). 



600 



Cnnoids 



nn^nnnrir^nn 



■^^-^L. 



The distribution of mean nearest-neighbor distances 
between fishes and large invertebrates varied from 0.1 
to 9.9 m (Fig. 10). Overall median distances varied from 
0.9 m (shelf sponges) to 1.8 m (black corals). However, 
there were no statistical differences between the me- 
dian distances for each group (Kruskal-Wallis, H=WA; 
df=6; P=0.11). For the six fishes that were found more 
frequently near large invertebrates than on transects, 
the overall median distances to the invertebrates were 
5.5 m (cowcod). 1.0 m (bank rockfish), 1.3 m (swordspine 
rockfish), 1.5 m (shortbelly rockfish), 1.7 m (pinkrose 
rockfish), and 1.4 m (Sebastomus). 

The overall incidence of damaged and dead sponges, 
gorgonians, and black corals was low (0.3% of total num- 
ber observed). Black corals were more commonly damaged 
(1.7%) or dead (1.1%), followed by vase sponges (0.6% and 
0.1%, respectively), barrel sponges (0.5% and 0%), and 
foliose sponges (0% and 0.1%). No dead or damaged flat 
or shelf sponges or gorgonians were observed. 
Damage in black corals included portions of the 
colony that appeared discolored; dead black cor- 
als lacked polyps and were discolored. Among 
sponges, the most common damage was individu- 
als that had broken from the substratum and 
were lying on their side or broken colonies. 



Basket stars 



I 



r^ r^ .J. r^ n r;-! 



fi 



ri r^ Pn .=,,—, Pn Fl 



900 - 



Brittle stars 



^r^^OUO 



' I ' ' t ' ' I ' ' 1 



n 



JL 



^n^^ 



500 



E 
o 



Bractiiopods 



^^^^ 



^1 1^ 



>, 


200 - 


Wtiite sea urchins 




T T 


n=45,092 


c 
Q 


- 




-n- 1 r- 


rn n 


fl , ^ 





2 n 



White plumed anemone 



i 



rJ^ 



^ , ^ , 



200 - Fragile sea urchins 



1 



.n. 



o ^. 



-Oa 



50 



Sea pens 



rjl ^ -^ -, p O rp [7] 



n=9,726 

X 



n_ 



TT RR RIVl BB BC BS CB MB SB CS SC SP MG SG MM MS SS 
Habitat code 

Figure 5 

Mean density of structure-forming invertebrates in habitat patches 
of each substratum type. Vertical bars are + one standard error. 
See page 169 for definitions for the substrate abbreviations along 
the X axis. 



Discussion 

Several groups of invertebrates were distin- 
guished by their large size, such as black corals, 
sponges, crinoids, basket stars, anemones, and 
sea pens. Organism size is an important aspect 
of structural habitat because it contributes to 
vertical relief and increases the availability of 
microhabitats. For example, yelloweye iSebastes 
rube?Ti?nus) rockfish may use the large gorgo- 
nian coral Primnoa as a vantage point to prey 
upon small fishes (Krieger and Wing, 2002). 
Size variation among structure-forming inver- 
tebrates was significant. Individual black corals, 
sea pens, and sponges greater than 1 m in height 
represented only 0.1% of all organisms, and 90% 
of the individuals were <0.5 m high. 

Similarly, the complex structures of crinoids, 
gorgonians, black corals, and basket stars may 
increase the availability of microhabitats and 
create a high surface area for settlement or 
retention of other organisms. Fish egg cases 
have been observed attached to both gorgonians 
(Etnoyer and Morgan') and vase sponges (pres- 
ent study. Table 3). 

High-density aggregations have not been used 
as a criterion for defining structure-forming 
invertebrates, but there are several examples 
that illustrate the potential importance of these 
aggregations. High density "forests" of crinoids 
provide refuge and substrata for a wide variety 
of small fishes and invertebrates in rocky areas 
(Lissner and Benech, 1993; Puniwai, 2002). 



Tissot et al Invertebrates on deep banks off southern California 



175 







Figure 6 

Structure-forming invertebrates; (A) gorgonians (Gorgonacea), (B) black coral iAntipathes dendrochristos), 
(Cl flat, (D) barrel, (E) shelf, (F) vase, and (G) foliose sponges on southern California rocky banks. 



176 



Fishery Bulletin 104(2) 



Similarly, high-density aggregations of brittle stars 
and brachiopods in boulder-cobble areas and fields 
of sea pens and sea urchins in sand and mud habi- 
tats also may provide space and structure for other 
organisms (e.g., Brodeur, 2001). 

With the exception of black corals and sea pens, 
the largest structure-forming invertebrates, such as 
sponges, Metridium spp., and crinoids, were most 
common on high- to moderate-relief rocky habi- 
tats. These long-lived organisms are likely to be 
favored in stable habitats that are more insulated 
from sediment transport and high particle loads 
than low-relief, mud-dominated areas (Lissner and 
Benech, 1993). Large invertebrates add structure 
and micro-scale complexity to these rocky habitats 
that already contain high-to-moderate amounts of 
relief Sponges, with their broad distributions, may 
also provide structure for flatfishes in low-relief 
mud habitats (Ryer et al., 2004) 

Black corals and gorgonians, in contrast, are 
more commonly found in current-swept areas near 
drop-offs and under ledges (Grigg, 1974; Parrish, 
2004). In our study, these invertebrates were found 
in low-relief mixed cobble-boulder-sand habitats at 
100-225 m depths, providing significant vertical 
structure for potential use by a wide variety of 
organisms. 

Aggregations of sea pens and sea urchins may 
provide important structure in low-relief sand and 
mud habitats where there is little physical habi- 
tat complexity. In addition, these organisms may 
provide refuge for small planktonic and benthic 
invertebrates, which in turn may be preyed upon 
by fishes. They also may alter water current fiow, 
thereby retaining nutrients and entraining plank- 
ton near the sediment. Urchins rapidly respond 
to patches of drift kelp (Harrold and Reed, 1985), 
which provide organic material to deep sea habitats 
(Harrold et al., 1998). 

One of the central issues currently relevant to 
structure-forming invertebrates is the degree to 
which these species contribute to the spawning, 
breeding, feeding, or growth-to-maturity of eco- 
nomically important fishes. Although there are several 
studies that report fish-invertebrate associations within 
common habitats (Hixon et al.^), or make anecdotal or 
general observations on fish-invertebrate associations 
(e.g., Krieger and Wing, 2002), few studies have sys- 
temically quantified these relationships. In our study. 
for 9105 observations on the larger invertebrates found 
on southern California rocky banks only 1.8% of indi- 
viduals had other organisms lying on or attached to 
them. Moreover, the vast majority of these organisms 
were other invertebrates, including crinoids, sponges, 
crabs, basket stars, brittle stars, seastars, anemones 
and salps. Less than 1% of the observations of organ- 
isms actually sheltering in or located on invertebrates 
involved fishes (a total of five individuals and one egg 
case), and most were observed on large black corals 
(Table 3). This result implies that fishes are not strong- 



10 



05  



00 



0,5 



0,0 



10 



0,6 - 



0,0 
4 




10 



Gorgonians 



n=27 



Blacl< corals 



1 



A 



^^ 



-ia 



Flat sponges 



n 



nll^nHn^ 



n^ 



Barrel sponges 



on 



A 



r;i r^ r^-i n ^ ,J^ r^ 



Stielf sponges 



^^^n 



1 



Habitat Code 



n=139 



i 



htJx 



M- 



Vase sponges 

1 



A 



I, 



m 



Oii 



r^ ,^ ^ r-p 



Foliose sponges 



1 



M 



n r;n ri rp I I r-1 



i 



XU- 



T l X -r -I- T 



TT RR RM BB EC BS CB MB SB CS SC SP I^G SG MM MS SS 

Habitat code 

Figure 7 

Mean density of gorgonians, black corals, and sponges in 
habitat patches of each substratum type. Vertical bars are 
± one standard error. See page 169 for definitions for the 
substrate abbreviations along the .v axis. 



ly associated with structure-forming invertebrates in 
the areas we surveyed off southern California. 

However, we should note that our observations were 
limited to daylight hours and that the viewing angle 
from the submersible generally precluded seeing inside 
some of the sponges (especially vase and barrel types). 
Moreover, our analyses focused on associations between 
fishes and individual solitary invertebrates, most of 
which were <0.5 m in height. We did not examine as- 
sociations between all structure-forming invertebrates, 
nor did we examine associations between invertebrates 
and assemblages of fishes at the level of discrete habitat 
patches (100-1000 m scale) (e.g., Tissot et al.^) 



■• Tissot, B. N., M. A. Hixon, D. L Stein. Unpubl. manu- 
script. Habitat-base submersible assessment of groundfish 
assemblages at Heceta Bank, Oregon from 1988-1990. 



Tissot et al , Invertebrates on deep banks off southern California 



177 



120 W 



119 30'W 



34°30'N 



34 N 



33°30N 



33°N 



32°30N 



Point Conception 




San Miguel Is. Santa Cruz Is 



^^o ^... 



Santa Rosa Is. 



^^nacapa Is 

Harrison's Reef Dume Canyon 
Reef 



1 1 7 30 W 




117 


No. of Colonies 


• 


1- 


2 


• 


3- 


4 


• 


5- 


7 


• 


8- 


23 


# 


24 


-57 



Hidden 



Osborn Bank-c^/ 



. \\V.. v. 

^< (Cv'^'l Potato Bank. Saa^icdtes Is. - — .V 




'-^ .TV\ 



f — 



■^^^ X/ 



-^y^t 



25 50 100 km 



\^ Cherry Bank 

■A r-'.Tannel/Cortes[6ank 

\fc:"' -V. • 






»\'> 



:).4 rtO'N 




33 30 N 






^^^ ¥A" 



32'30N 



120 30' W 



120 W 



1 1 9°30'W 



117°30W 



119 'SOW 



119"W 



118 30W 



33'30'N 



32°30N 



B 



Point Conception ^""'^ 



u 



San H/liguel Is. Santa Cruz Is. ^ .^ * 

- --■ _ Anacapals. "~- 

•~», , ' -— ' --^. •n:^~-'^ - "^ 

^^ vr _ V tSrrison's Reef Dumfe Canyon 

^- -" ' <;nntfl RAfin l«. - 




No. of Colonies 



• 1 

• 2 



X-^-v , Santa Rosa Is. 



% 



%. 



Hidden] Reef 
"*v _rv v^-- V \ Kidney BaTm' < 



I ..^. . ^ Santa Barljara lb 

., , 5,. ^J \ '<i' ' I ; ' Osborn Bank."'" , 

^,J .M,^\ Potato ?!«;>'< 'sa^liit'olasls. -—X 

X W'CH^ryBank 






tx 



^^- V. '^ 



12 5 25 50 km 

1 1 I 1 III 






 "" ^■S5i^;U:-/Tar)nel/CortesBank\ 






'.-;> 



;:i: 




r- g>i|th^ t\ ■^ 






r--N 









34 30'N 



34 N 



33"N 



120-30W 



120 W 



119"30'W 



11 9° W 



117°30'W 



Figure 8 

Distribution and number of (A) black coral and (B) gorgonian colonies inside and outside the 
cowcod conservation areas (delineated as outlined boxes) of southern California in 2002. Dive 
sites are indicated by black dots. 



178 



Fishery Bulletin 104(2) 



From the analysis of spatial associations between 
fishes and large, individual structure-forming inver- 
tebrates, six of 108 species were found more often ad- 
jacent to colonies than predicted by their abundance 
along transects. This result indicates that there may 
be spatial associations that do not necessarily include 
physical contact with the sponges and corals. However, 
the median distances between these six fish species and 
large invertebrates (1.0-5.5 m) were not particularly 
small. Thus, it is likely that these fishes and inverte- 
brates are present in the same types of habitats and 
that there is not necessarily a functional relationship 
between these two groups of organisms. 

Parrish (2004) reached similar conclusions on stud- 
ies of black coral in Hawaii. Although fish densities 
were higher in areas that included corals, when bot- 
tom relief and depth were accounted for these densi- 
ties were not higher than those for surrounding areas 
without corals. Thus, there was no clear evidence that 
corals served to aggregate fish. Rather, fishes and cor- 
als co-occurred in areas with similar physical relief and 
unique flow regime (Parrish, 2004). Auster (2005) also 
reached similar conclusions by finding no significant 
difference in the density of a common rockfish species 
iSebastes fasiatus) between areas of rock and boulders 
with dense coral cover and similar areas having dense 
epifaunal cover (i.e., without coral). Auster concluded 
that although dense coral and dense epifaunal habitats 
were functionally equivalent, the epifaunal habitat was 
more widespread in his study area, making that habitat 
perhaps more important to the greater rockfish popu- 
lation. Finally, Syms and Jones (2001) demonstrated 
that removal of high densities of soft corals caused no 
significant changes in the associated fish communities 
and that the heterogeneity of habitat generated by soft 
corals was indistinguishable from equivalent habitat 
formed by rock alone. Thus, fish-invertebrate associa- 
tions, by themselves, do not necessarily demonstrate 
the functional importance of invertebrates as habitat 
to benthic fish populations. 

One possible conclusion from our study is that ob- 
served fish-invertebrate associations, like those reported 
for many cold-water corals, can be overstated. In the 
absence of quantitative information, observers may re- 
member the few positive associations between fishes and 
structure-forming invertebrates but ignore (or forget) 
the more numerous observations of large invertebrates 
with no associated fishes. Indeed, the general impression 
of the authors after making submersible observations 
was that there were higher numbers of fishes associated 
with large invertebrates when in reality only five fishes 
were observed lying directly on a large invertebrate in 
the video transects. A more likely conclusion, however, 
is that the continental shelf communities of southern 
California are unique and that large black corals do 
not have the high number of commensals as seen, for 
example, on Primnoa in Alaska. An additional consider- 
ation is the relatively low number and size of individual 
sponges, gorgonians, and black corals observed in this 
study. Primnoa in Alaska can form massive stands 3 m 



20 

10 



200 

100 



3000 

1500 



CO 

■g 

> 1000 

"D 

g 500 
o 

d 
z 

100 

50 





400 


800 

400 





Gorgonians 



X. 



15 20 25 30 40 

Black corals 





50 


100 150 200 


250 


n=4,043 


Flat sponges 




r^ 




n 


i 



10 


20 


30 


40 50 60 70 


80 


90 


100 


n=2,068 


1— 




Barrel sponges 








n 




n 








t 



10 20 30 40 50 60 70 80 90 100 
"=134 Shelf sponges 



10 


20 


30 40 50 60 70 


80 


90 


100 


n=1,155 




Vase sponges 








rn 




n „ 






i 



10 20 30 40 50 60 70 80 90 100 
n=i.262 Foliose sponges 



10 20 30 40 50 60 70 80 90 100 
Height (cm) 

Figure 9 

Size distributions of large, structure-forming inver- 
tebrates. Arrows indicate ma.ximum sizes observed in 
each group. 



tall and 7 m wide in some areas (Krieger and Wing, 
2002). Moreover, the majority of fishes were observed 
on the largest individuals in their study (>15 m^ in 
volume) Most of the corals in our study were <0.5 m in 
height. Given the importance of this issue, we argue for 
more rigorous quantitative studies on fish-invertebrate 
associations that would include densities of fishes and 
sizes of both fishes and invertebrates. 

Regardless of their associations with fishes, the struc- 
ture-forming invertebrates described in this study are 
very likely to be ecologically important on continental 
shelf ecosystems and are certainly significant in their 
own right. Observation on the health of the larger in- 
vertebrates indicates few damaged (0.1%) or dead (0.2%) 
individuals and a low incidence of fishing gear in the 
areas surveyed (Tissot et al., unpubl, data). These ob- 
servations are consistent with the absence of a signifi- 
cant commercial bottom trawling fishery in our survey 
area, which has been associated with negative impacts 
on large invertebrates in other locations (Watling and 
Norse, 1998; Freese et al. 1999: Krieger. 2001). Thus, 
this study affords a unique view of what appears to be 



Tissot et dl Invertebrates on deep banks off southern California 



179 



6 

3 



-a 



a -a 
^ c 
a ■- 



Si ^ 

a to 

«} -c 

. a 

-2 >. 

o -a 

v. 0) 

*^ a 

^ c 



_o 

CO CO 

c a; 

HI /■ ^ 






QJ 



•^ -C 



".I 

CO 

CO ^ 

c S 

c ^ 
M CO 

bo *^ 

, c/l 
M g 

S o. 



o 


o 


a 


r: 












u 


T. 


crt 


y= 






0) 




> 




T3 






c 


!B 


q; 


an 


cl; 


c 

o 


0- 


CO 



m 



o 



o 



H e-s 



H C bfl 



— J= a; 

CO "S tin 

E-> CO Q. 
c 



H CO 



-2 2^ 



— j5 0) 

to =C be 

o '- C 

[2 CO o 

c ■-« 



V ~ 



o 

CO 



CO •-^ 





Vi 




























CO 














^ 










CO 


X 








C 








■^ 


r; 




■^ 










en 


CJ 


CO 
0) 








CO 


r* 


CO 

Si 










-2 

c 




1 

5 


3 




t>5 


0^ 


^ 

s 


•n 

CC 

J:! 


p 


8 


3 
^ 


o 
o 
;~i 


1 


£ 
o 




CO 


■J: 

"5, 


o 
;- 


_Cj 


CC 
P 


a 
a 

CO 


tC 


a 
a 

CO 


cc 


^ 
■^ 


8 

■a 

CD 

cc 


Q. 

a. 

CO 






Li 


o 












O 


"j; 


S-H 




n 


C>", 




Q 


o 


■-A 


a; 


cc 






£! 




>, 




n 








a 


■j; 


y= o: 














s 




o. 










T3 








Ql 


^,) 






^ 


O) 




^^■• 






S 


a 
-a 

o 


■j; 




to 


CC 


en 
01 


en 






CO 


C 




= 




OJ 








'iJ 




T3 

O 
CJ 


•*^ 








S CO 




o 


C^ 


>. 


CI 




Cl 


CI 








n 


to 


o 




n 




M 




















■A 
CO 


c 
c 


cj 

CO 


bB 






CO 


C3 

rr 




o 
r! 


1 


c 

CO 


-o 

CO 


"to 


-a 
CO 


c 
c 




c 

> 

3 


CD 

CO 


o 


CO 


S 


1 


CO 


CO 


OJ 


CO 




D 


w 




a. 




tl. 




C/2 




C/^ 




pa 




E 




:u 






m 




^ 




u 




c/: 




O 




Di 



180 



Fishery Bulletin 104(2) 









-u 
















15 




t-H 














^ 


o i=! 


V 


V 


V 


V 


V 




a 




Xj 














;- 


















o 


















u 


















^ 




en 














s 




? 
































m 


s 
^ 


4J 


o 




o 


o 


o 








o c 

CO 


i-H 














e2 


V 


V 




V 






c 




-D 














.2 




































'S 


















o 




en 














bo 




a 














a 


"3 

o 


V, .2 














o 


ts c 
a; o 


o 


o 


^ 


o 


o 






H 


C bo 


















o c 


CO 


^ 


CO 


^ 


^ 




y: 


o 




V 




V 






Ci 


H 


ca 
































o 


















a 


















« 




Cm 














;<-- 




1 s 

^ o 












^ 




"n 


^ 


T-K 


^ 


T-H 


I-l 


u 


CO 


o 






V 






3 
C 




H 


M Q. 












'S 


















c 

























jj 












LI 




^^ 


u 
















to 

o 


c; 












V 




c 2 
o c 


CO 


.—1 


Tj- 


V 


^ 


V 




^ 


2 










£ 


















."I 


















1- 




"5 


















-tj 


c^ 




o 


o 


CM 




X 


^ 


« 1 














OJ 


















b£) 




QJ 














C 




"3 














o 




O ? 














CI. 




 x; oj 














'11 


CO 


s f 




















>— t 




!— ! 


r-H 




o 


^ 


2 1 


V 


V 


V 








S 




c 














o 


















O 




^ s 
































CQ 


tc bo 


















•„ c 


c^ 




IN 










H 


01 o 
c <" 


















t. a) 
















CC 


n! bo 


















^ S 
ta Q. 

c 






CO 


CM 


.—1 










'^. 


rX o 














3 


















■C C 


O 


Q 


f^ 










en 2 


■3 o 
o "^ 


c; 


2? 


CO 










■5 ^ 

2 e 




^1 




3 

a. 










-21 


ij CO 


tlD o 


5 












o a; 


OJ ^ 


■a 


.2 « 










t. cc 


>^ O 


-a o 










C- 


M e 


OJ c 


s-i 


CJ c 










t, -o 


? -o 


-^ ';:: 


a -o 








ro 


CO "iJ 


01 ^ 


u "-^ 


bo a. 


CO ^ 








X 

ra 


f « 


t ^ 


M tt; 


c O 


g^ 








H 


Q 


O 


3 


J 


CQ 




30 




20 




15 




20 




20 





-1 


I-^ 


r-n 


Gorgonians median=1.4m 




— 1 


— 


— 


Black corals median=l 7 m 

n 


l— 1 




Flat sponges median=1.4m 




— 1 


r-l 


Barrel sponges median=i 3m 




— 


Shelf sponges median=0.9m 




(— 1 




Vase sponges median=1.4 m 




— 




Foliose sponges median=i 3 m 



01 23456789 10 
Mean distance to nearest fish (m) 

Figure 10 

Mean nearest neighbor distances of fish species to 
gorgonians, black corals, and sponges. 



relatively undisturbed megafaunal invertebrate commu- 
nities in the southern California Borderlands, and sup- 
ports the continued protection of these animals within 
the Cowcod Conservation Areas. 



Acknowledgments 

We thank the dedicated crews of the RV Velero and Delta 
for support in the field. D. Schroeder, M. Nishimoto, L. 
Snook, T. Laidig, T.Anderson, R. Starr, B. Lea, and C. 
Wahle assisted with the survey work. K. Greenwood 
and S. Green assisted with habitat and invertebrate 
analyses. Funding was provided by NOAA Fisheries, 
SWFSC Fisheries Ecology Division, Offices of Habitat 
Conservation and Protected Resources; the National 
Undersea Research Program; NOAA Marine Protected 
Area Science Center; the David and Lucile Packard 
Foundation; Washington State University Vancouver; 
and the Monterey Bay Sanctuary Foundation. 



Tissot et al Invertebrates on deep banks off southern California 



181 



Literature cited 

Andrews, A. H., E. E. Cordes, M. M. Mahoney, K. Munk, 
K. H.Coal, G. M. Calliet, and J. Heifetz. 
2002. Age, growth, and radiometricage validation of 
a deep-sea, habitat-forming gorgonia iPrimnoa reife- 
daeformin) from the Gulf of Alaska. Hydrobiologia 
471:101-110. 
Auster, P. J. 

2005. Are deep water corals important habitat for fishes? 
Cold-water corals and ecosystems (A, Freiwald and J. 
M. Roberts (eds.l, 1244 p. Springer, New York, NY. 
Brodeur, R. D. 

2001. Habitat-specific distribution of Pacific ocean perch 
iSebastes ahitiis) in Pribilof Canyon, Bering Sea. Cont. 
Shelf Res. 21:207-224. 

Brusca, R. C, and G. J. Brusca. 

1990. Invertebrates, 922 p. Sinauer Associates, Inc., 
Sunderland, MA. 
Dayton, P. K., S. Thrush, and F. C, Coleman. 

2002. Ecological effects of fishing in marine ecosystems 
of the United States, 52 p. Pew Oceans Commission, 
Arlington, VA. 

Fossa, J. H., P. B. Mortensen. and D. M. Furevik. 

2002. The deep-sea coral Lophclia pertufa in Norwegian 
waters: distribution and fishery impacts. Hydrobiologia 
471:1-12. 
Freese, J. L. 

2001. Trawl-induced damage to sponges observed from a 
research submersible. Mar. Fish. Rev. 63(3):7-13. 

Freese, J. L., P. J. Auster, J. Heifetz, and B. L. Wing. 

1999. Effects of trawling on seafloor habitat and associ- 
ated invertebrate taxa in the Gulf of Alaska. Marine 
Ecol. Prog. Ser. 182:119-126. 
Grigg, R. W. 

1974. Distribution and abundance of precious corals in 
Hawaii. In Proc. second international coral reef sym- 
posium (A. M. Cameron et al., eds.), p. 235-340. Great 
Barrier Reef Committee, Brisbane, Queensland, Aus- 
tralia. 
Harrold, C, and D. Reed. 

1985. Food availability, sea urchin grazing, and kelp 
forest community structure. Ecology 66:1160-1169. 
Harrold, C, K. Light, and S. Lisin. 

1998. Organic enrichment of submarine-canyon and 
continental-shelf benthic communities by macroalgal 
drift imported from nearshore kelp forests. Limnol. 
Oceangr. 43:669-678. 
Heifetz, J. 

2002. Coral in Alaska: distribution, abundance, and 
species associations. Hydrobiologia 471:19-28. 

Husebo, A. L. Nottestad, J. H. Fossa, D. M. Furevik, and S. B. 
Jorgensen. 

2002. Distribution and abundance of fish in deep-sea 
coral habitats. Hydrobiologia 471: 91-99. 
Krieger, K. J. 

2001. Coral \Pniunoa) impacted by fishing gear in the 
Gulf of Alaska. In Proceedings of the first international 
symposium on deep-sea corals (Willison et al., eds.), 
106-116. Ecology Action Centre, Halifax, Canada. 

Krieger. K. J., and B. L. Wing. 

2002. Megafaunal associations with deepwater corals 
iPrimnoa sp.) in the Gulf of Alaska. Hydrobiologia 
471:83-90. 



Lissner, A., and S. Benech. 

1993. Benthic hard-substrate community ecology of 
the Santa Maria basis and western Santa Barbara 
channel. In Taxonomic atlas of the benthic fauna of the 
Santa Maria Basin and western Santa Barbara Channel 
(J. A. Blake and A. L. Lissner, eds.), p. 33-45. Santa 
Barbara Museum of Natural History, Santa Barbara, 
CA. 

Malakoff, D. 

2004. U.S. asked to act immediately to protect deep-sea 
coi-als. Science 304(5667):31. 

NRC (National Research Council). 

2002. Effects of trawling and dredging on seafloor 
habitat. Committee on Ecosystem Effects of Fishing: 
Phase 1-effects of bottom trawling on seafloor habitats, 
136 p. National Research Council, National Academy 
Press, Washington, D.C. 

Opresko, D. M. 

2005. A new species of antipatharian coral (Cnidaria: 
Anthozoa: Antipatharia) from the southern California 
Bight. Zootaxa 852:1-10. 

Palumbi, S. R. 

1986. How body plans limit acclimation: responses of a 
demosponge to wave force. Ecology 67( 1):208-214. 
Parrish, F. A. 

2004. Foraging landscape of the Hawaiian monk 
seal. Ph.D. diss., 146 p. Univ. Hawaii, Manoa. 
Puniwai, N. P.F. 

2002. Spatial and temporal distribution of the crinoid 
Florometra serratissma on the Oregon continental 
shelf. M.S. thesis, 34 p. Washington State Univ., 
Vancouver, WA. 
Roberts, S., and M. Hirshfield. 

2004. Deep-sea corals: out of sight, but no longer out of 
mind. Front. Ecol. Environ. 2(3):123-130. 
Rogers. A. 

2004. The biology, ecology and vulnerability of deep-sea 
coral reefs. lUCN publication:l-13. 
Riedl, R. 

1971. Water movement: animals. Marine ecology: a com- 
prehensive, integrated treatise on life in oceans and 
coastal waters (O. Kinne, ed.), 1(2):1123-1156. Wiley- 
Inter Science, London. 
Ryer, C. H. A. W. Stoner and R. H. Titgen. 

2004. Behavioral mechanisms underlying the refuge 
value of benthic habitat structure for two flatfishes with 
differing anti-predator strategies. Mar. Ecol. Prog. 
Ser. 268:231-243. 
Stein, D. L., B. N. Tissot, M. A. Hixon and W. Barss. 

1992. Fish-habitat associations on a deep reef at the 
edge of the Oregon continental shelf. Fish. Bull. 
90:540-551. 
Syms, C, and G. P. Jones. 

2001. Soft corals exert no direct effects on coral reef 
assemblages. Oecologia 127:560-571. 
Watling, L., and E. A. Norse. 

1998. Disturbance of the seabed by mobile fishing 
gear: a comparison to forest clearcutting. Cons. Biol. 
12(6):1180-1197. 
Yoklavich, M. M., H. G. Greene, G. M. Cailliet, D. E. Sullivan, 
R. N. Lea, and M. S. Love. 

2000. Habitat associations of deep-water rockfishes in a 
submarine canyon: an example of a natural refuge. Fish. 
Bull 98:625-641. 



182 



Abstract — Larval and juvenile rock- 
fishes iSebastes spp.) are difficult to 
identify using morphological charac- 
ters. We developed a key based on 
sizes of restriction endonuclease frag- 
ments of the NADH dehydrogenase-3 
and -4 (ND3/ND4) and 12S and 16S 
ribosomal RNA (12S/16SI mitochon- 
drial regions. The key makes use of 
variation in the ND3/ND4 region. 
Restriction endonuclease Dde I varia- 
tion can corroborate identifications, as 
can 12S/16S variation. The key, based 
on 71 species, includes most North 
American taxa, several Asian species, 
and Sehastolnbus alascanus and Heli- 
colenus hilgendnrfi that are closely 
related to rockfishes. Fifty-eight of 
71 rockfish species in our database 
can be distinguished unequivocally, 
using one to five restriction enzymes; 
identities of the remaining species 
are narrowed to small groups: 11 S. 
polyspinis. S. crameri, and S. ciliatus 
or variabilis (the two species could 
not be distinguished and were consid- 
ered as a single species) ; 2) S. chlo- 
rostictus, S. eos, and S. rosenblatti: 
3) S. entomelas and S. mystinus; 4) 
S. emphaeus, S. variegatus, and S. 
wilsoni; and 5) S. carnatus and S. 
chrysomelas. 



A key to selected rockfishes iSebastes spp.) 
based on mitochondrial DNA restriction 
fragment analysis 

Zhuozhuo Li' 
Andrew K. Gray' 
Milton S, Love' 
Akira Goto^ 
Takashi Asahida^ 
Anthony J. Gharrett' 

' Fisheries Division, School of Fisheries and Ocean Sciences 
University of Alaska Fairbanks 
11120 Glacier Highway 
Juneau, Alaska 99801 

Email address (for A J Gharrett, contact author) ffaigia'uafedu 

^Marine Science Institute 
University of California 
Santa Barbara, California 93106-6150 

3 Graduate School of Fisheries Sciences 
Hokkaido University 
3-1-1 Minato-cho, Hakodate 
Hokkaido 041-8611, Japan 

'' School of Fisheries Sciences 
Kitasato University 
Sanriku, Ofunato 
Iwate 022-0101, Japan 



Manuscript submitted 11 October 2004 
to the Scientific Editor's Office. 

Manuscript accepted for publication 
1 August 2005 by the Scientific Editor. 

Fish. Bull. 104:182-196 (2006). 



More than 100 species of rockfishes 
(genus Sebastes) are found worldwide 
(Kendall, 2000), and the majority are 
distributed along the Pacific coast of 
North America and in the northwest 
Pacific from the western Bering Sea 
south to Japan and Korea (Love et 
al., 2002). Sixty-five rockfish species 
are found along the California coast 
(Moser, 1996). Within the genus, there 
is a high degree of similarity of the 
morphological characters among many 
species. These similarities are in part 
due to recent divergence but also may 
have resulted from convergence of con- 
geners occupying similar habitats. 

Historically, species identification of 
Sebastes has been based on morphol- 
ogy; however, this approach is often 
insufficient, especially for identify- 
ing sympatric species, where there 
is considerable similarity in size and 
physical features and in pigmenta- 
tion, as well as overlap in meristic 
characters. The difficulty in identify- 
ing Sebastes to species level exists for 



all developmental stages, but larvae 
are especially difficult to identify be- 
cause they lack diagnostic characters 
(Kendall, 1991), Currently, only 15 
species at the larval stage can be sep- 
arated by physical characters such as 
body shape, pigmentation patterns, 
and head spine development (Love et 
al., 2002). 

Rockfishes are important ecologi- 
cally, and most species are economi- 
cally valuable. Sebastes larvae form 
a large portion of ichthyoplankton 
collections and rank third or fourth 
in abundance among all fish larvae 
taken during California Cooperative 
Fisheries Investigations (CalCOFI) 
surveys that have covered the entire 
length of the California and Baja Cal- 
ifornia coast and now cover waters 
off southern California. The ability 
to identify Sebastes accurately and 
efficiently at all developmental stages 
will greatly improve both our ability 
to learn about their life histories and 
our management and conservation ef- 



Li et a\. A key to selected Sebastes spp based on mitochondrial DNA restriction fragment analysis 



183 



forts. An increased understanding of life history varia- 
tion can improve the systematic descriptions of Sebastes 
species, which has mostly been based on morphology. 

Several methods have been developed to obtain spe- 
cies-specific information that supplements the use 
of morphological and meristic characters for species 
identification. Otolith microstructure and other hard 
structures have been used to distinguish the late- 
stage larvae and young-of-the-year pelagic juveniles of 
some Sebastes species (Laidig and Adams'; Laidig and 
Ralston, 1994; Laidig et al., 1996). Biochemical genetic 
methods have been used to identify adults (Barrett et 
al.. 1966; Seeb, 1986), and may also be used to identify 
juveniles of some species (Seeb and Kendall, 1991). 
LeClair and Buckley (2001) used allozyme variation at 
37 loci to positively identify 155 individuals of juvenile 
Sebastes diploproa and Sebastes melatwps; but they 
were unable to identify another 29 individuals, partly 
because of a limited database. Despite some success 
with their use, allozymes often have low resolution, 
and there are complexes of species that cannot be dis- 
tinguished by using allozymes alone (Seeb and Kendall, 
1991). 

DNA sequence-based methods have some advantages 
over the use of allozymes for the identification of juve- 
nile fish species. They often have greater resolution, due 
in part to the fact that expression of some metabolic 
enzymes changes during development, whereas DNA 
sequences generally do not. DNA is less susceptible to 
degradation than the enzymes used in allozyme stud- 
ies. Also, DNA sequence-based methods require a small 
amount of tissue sample, which is especially suited for 
work with samples from early life stages (e.g.. Gray 
et al., 2006). Because it has a relatively high rate of 
substitution (Moritz et al., 1987), mitochondrial DNA 
(mtDNA) can be useful for distinguishing closely re- 
lated species. Mitochondrial DNA is usually inherited 
maternally in vertebrates and does not recombine with 
paternal mtDNA if it is leaked into the zygote (Gyllen- 
sten et al., 1991; but see Rokas et al., 2003). Thus, the 
mtDNA sequence represents a matrilineal phylogeny 
and is often useful in delineating phylogenetic rela- 
tionships of closely related species. Sequences of the 
mitochondrial cytochrome b gene were used to identify 
the pelagic young of Sebastes constellatus and Sebastes 
ensifer (Rocha-Olivares et al., 2000). Multiplex PCR 
of haplotype-specific regions of mtDNA has also been 
used to identify early stages of Sebastes species (Rocha- 
Olivares, 1998). 

The objective of this study was to devise a key for 
species identification using mtDNA restriction fragment 
data from both the ND3/ND4 and 12S/16S regions. 
In a previous study, data for species-specific mtDNA 
restriction site variation in the ND3/ND4 region were 



' Laidig, T. E., and P. B. Adams (eds. ). 199L Methods used 
to identify pelagic juvenile rockfish (genus Sebastes) occurring 
along the coast of central California. NOAA-TM-NMFS- 
SWFSC-166. 180 p. NMFS Southwest Fisheries Science 
Center, 110 Shaffer Rd., Santa Cruz, CA 95060. 



presented for 15 Sebastes species (Gharrett et al., 2000). 
We have included data for 56 additional species using 
the ND3/ND4 and 12S/16S regions of the mtDNA as the 
target region of PCR-amplification. We have included 
specimens of Helicolenus hilgendorfi and Sebastolobus 
alascanus for contrast. 



Materials and methods 

Adult specimens of 71 species of rockfishes were collected 
from the Gulf of Alaska, the coastal waters of Califor- 
nia and Baja California, and from the coast of Japan. 
Samples of H. hilgendorfi and Sebastolobus alascanus, 
species from two sister genera, were collected from Japa- 
nese coastal waters and the Gulf of Alaska, respectively, 
to provide outgroup comparisons. A sample of heart 
tissue from each specimen was preserved in either 95% 
ethanol or a solution that was 80% 0.25M ethylenediami- 
netetraacetic (EDTA) acid at pH 8 and saturated with 
NaCl (Seutin et al., 1991) and 20% dimethyl sulfoxide 
(DMSO). At least five individuals of each species were 
analyzed, except for a few species for which less than 
five specimens were available. We did not distinguish 
between S. ciliatus and S. variabilis, which had not yet 
been described (Orr and Blackburn, 2004), when we 
collected samples. 

Total genomic DNA was isolated by using a Purgene 
DNA^M isolation kit (Gentra Systems, Inc., Minneapolis, 
MN). Two target regions of mtDNA were amplified by 
using the polymerase chain reaction. The ND3/ND4 
region begins in the glycyl tRNA gene and spans the 
NADH-dehydrogenase subunit-3, arginyl tRNA, NADH- 
dehydrogenase subunit-4L, and NADH-dehydrogenase 
subunit-4 genes, ending in the histidyl tRNA gene. The 
12S/16S region starts near the phenylalanyl tRNA end 
of the 12S rRNA gene, and runs through the valyl tRNA 
gene to near the leucyl tRNA end of the 16S rRNA gene. 
Primers for target regions have been used to amplify 
these regions in northern Pacific rockfish (Gharrett et 
al., 2001). The lengths for the amplified ND3/ND4 and 
12S/16S regions are about 2385 and 2430 base pairs, 
respectively, based on the aggregate restriction maps. 

Subsamples of the PCR products of each individual 
were digested with one or more of the restriction endo- 
nucleases BstN I, BstU I, Cfo I, Dde I, Hind II, Hinf I, 
Mbo I, Msp I, Rsa I, and Sty I. Fragments were sepa- 
rated electrophoretically in 1.5% agarose gels (one part 
agarose [Sigma-Aldrich, St. Louis, MO] and two parts 
SynergeF'"^' [Diversified Biotech Inc., Boston MA]) in 
0.5xTBE buffer (TBE is 90mM Tris-boric acid, and 
2 mm EDTA, pH 7.5). A 100 base-pair ladder provided 
molecular weight markers to estimate restriction frag- 
ment sizes. Gels were stained with ethidium bromide 
and photographed on an ultraviolet light transillumi- 
nator. Restriction fragments that could not be accu- 
rately measured from agarose gels were separated on 
8% polyacrylamide gels and stained with SYBR Green 
I Nucleic Acid Stain'^^' (Molecular Probes, Eugene, OR) 
using a 25-bp ladder for a molecular weight standard. 



184 



Fishery Bulletin 104(2) 



A restriction site map was cre- 
ated for each endonuclease by 
using all observed restriction 
fragment patterns and neces- 
sary double digests. 



Results 
Baseline data 



2000' 



1000- 



_ 500- 

Q. 

S 300- 
200- 



100- 



50- 



Interspecific variation was 
observed for all enzymes in 
both mitochondrial regions, 
except for Hind II in the 
12S/16S region. Intraspecific 
variation was observed for sev- 
eral enzymes. A total of 215 
restriction sites were detected 
in the ND3/ND4 and 12S/16S 
regions (Appendix 1). Of the 
215 sites, 97 were unique to 
Sebastes species, 21 were 
unique to Sebastolobus alas- 
carius, seven were unique to 
H. hilgendorfi, and one was 
shared by Sebastolobus alas- 
caniis and H. hilgendorfi. The 
ND3/ND4 region had 141 sites, 
and the 12S/16S region had 74 

sites. In the ND3/ND4 region, 36 sites were common 
in all haplotypes. whereas 46 occurred in only a single 
haplotype. 

A total of 132 composite haplotypes resulted from 
site differences in the two mtDNA regions (Table 1). 
The individuals of Sebastes species had 127 haplotypes; 
Sebastolobus alascanus had four haplotypes; and H. 
hilgendorfi had a single haplotype. Thirty-four of the 
71 species displayed intraspecific variation and were 
represented by more than one composite haplotype; 
the remaining 37 species had a single composite hap- 
lotype. 

Two pairs and three triplets of Sebastes species had 
identical composite haplotypes and could not be sepa- 
rated at the species level with this restriction site in- 
formation (see identification key). These groups were 
1) S. carnatus and S. chrysomelas; 2) S. chlorostictus, 
S. eos, and S. rosenblatti; 3) S. ciliatus or variabilis, 
S. crameri, and S. polyspinis; 4) S. emphaeus, S. var- 
iegatus, and S. wilsoni\ and 5) S. entomelas and S. 
mystinus. Several pairs of haplotypes between variable 
Sebastes species were separated by a single restriction 
site. These included 1) S. hopkinsi and S. ovalis — Dde 
I; 2) S. zacentrus and the S. emphaeus-S. variegatus-S. 
wilsoni complex — Rsa I; and 3) S. reedi and the S. cilia- 
tus or variabilis-S. crameri-S. polyspinis complex — Mbo 
I. The most variable species was S. mystinus, with five 
haplotypes. Five Sebastes species, S. dalli, S. hubbsi, 
S. polyspinis, S. trivittatus, and S. zacentrus, as well 
as Sebastolobus alascanus, had four haplotypes each. 



T I I I I I — I — I — I — I — I — I — I — I — I — I — I — r — 1 — I — I — I — I — I — I — I — I — I — 1 — I — I — I — I — r 
OPABIpMqmRVgcCJfiDeHNhEGKnaFrd I ikb 

Mbo I haplotype 

Figure 1 

A mock gel showing expected fragment patterns for rockfish (Sebastes spp.) hap- 
lotypes as a result of digestion of the mitochondrial ND3/ND4 PCR product by 
restriction endonuclease Mbo I. The mobilities are the logarithm of the fragment 
sizes and separation was assumed to have taken place in l.S'S agarose gels (one 
part agarose and two parts SynergeF™). The haplotypes correspond to the fragment 
sizes in Appendix 1 and to the haplotypes in Table 1 and the identification key. A 
haplotype may occur in more than one species. Most of the fragments that occurred 
in the shaded region would probably not be well resolved in an agarose gel. 



The remaining 58 Sebastes species have species-specific 
markers that allow unambiguous identification. 

Use of the key 

The key we developed for Sebastes species was based 
exclusively on variation in the ND3/ND4 region because 
it required only a single PCR amplification and variation 
in the 12S/16S region contributed no additional resolu- 
tion between species to that provided by the ND3/ND4 
region. The key is not a dichotomous key, but it is applied 
in the same way that taxonomic keys are applied. The 
first step is to digest PCR-amplified DNA from the 
ND3/ND4 region with restriction endonuclease Mbo I 
and to estimate the sizes of the fragments produced by 
separating the fragments on an agarose-SynergeF*^' gel. 
The best results will be achieved by using molecular 
weight markers, digitally photographing the gels, and 
estimating the fragment sizes with appropriate software. 
Alternatively, visual recognition of the fragment pat- 
tern can be accomplished by constructing a graphic key 
(e.g., Fig. 1 for Mbo I fragments) from fragment sizes 
predicted by the restriction site maps in Appendix 1. 
Note that in the figure, the logarithm of the size of the 
fragment (in base pairs) is used to estimate the mobility 
of a fragment (Sambrook et al., 1989). When the Mbo I 
haplotype has been identified, proceed to the next step. 
For example if the Mbo I digest results in haplotypes 
D or e, proceed to step g. in the key, which specifies 
digestion of another subsample of the ND3/ND4 PCR 



Li et al, A key to selected Sebastes spp based on mitochondrial DNA restriction fragment analysis 



185 





















Table 1 






















Composite haplotypes for Seba 
mtDNA regions. The haplotype 


ites spp., Hclicniciuis hilgcndnrfi. and Seba. 
codes refer to haplotypes in Appendix 1. 


■tolobu 


s ala 


scaiuis in 


the ND3/ND4 and 12S/16S 


Species 










ND3/ND4 


















12S/16S 










Bs(N I 


BslVl 


Cfol 


Dde\ 


Htnill Hin(\ 


Mho I 


Msp 


Rsttl 


Sly I 


SsfNI BrfUI 


Cfol 


Dde\ HindU HmU 


Mho I 


Msp I 


Rsa I 


S(vl 


aleutianus A 


F 


C 


B 


N 


A 


E 


K 


D 


D 


C 


A 


A 


A 


D 


A 


B 


C 


B 


A 


A 


aleiitianiis B 


F 


C 


B 


N 


A 


F 


K 


D 


D 


c 


A 


A 


A 


D 


A 


B 


c 


B 


A 


A 


alutiia A 


B 


C 


A 


J 


A 


A 


B 


B 


C 


c 


A 


A 


A 


D 


A 


B 


c 


B 


A 


A 


alutus B 


B 


C 


A 


J 


A 


A 


B 


B 


C 


c 


A 


A 


A 


D 


A 


A 


c 


B 


A 


A 


atrovircnti 


O 


B 


D 


b 


C 


G 


.1 


C 


B 


I 


A 


A 


A 


D 


A 


B 


A 


A 


A 


A 


aiii-iciilatiia 


F 


B 


D 


V 


A 


G 


k 


C 


B 


A 


A 


A 


A 


D 


A 


B 


A 


A 


A 


A 


aurora 


B 


C 


D 


m 


A 


A 


b 


L 


A 


e 


A 


A 


A 


D 


A 


A 


C 


B 


A 


A 


bahcocki A 


F 


C 


D 


G 


A 


A 


G 


B 


B 


C 


A 


A 


A 


C 


A 


B 


c 


B 


B 


A 


babcocki B 


F 


c 


D 


F 


A 


A 


H 


B 


B 


C 


A 


A 


A 


C 


A 


B 


c 


B 


B 


A 


borealis 


F 


c 


D 


D 


A 


A 


F 


B 


B 


C 


B 


A 


A 


A 


A 


B 


c 


A 


C 


A 


breiispinis 


G 


c 


D 


C 


A 


A 


K 


D 


B 


D 


A 


A 


A 


D 


A 


A 


c 


A 


C 


A 


capensis A 


F 


c 


D 


G 


B 


A 


E 


B 


D 





A 


A 


A 


C 


A 


A 


c 


A 


D 


A 


capensis B 


F 


c 


D 


F 


B 


A 


E 


B 


D 





A 


A 


A 


C 


A 


A 


A 


A 


D 


A 


capensis C 


F 


c 


D 


G 


B 


A 


E 


B 


D 





A 


A 


A 


D 


A 


A 


C 


A 


D 


A 


carnatiis A 


F 


a 


A 


I 


C 


G 


C 


a 


B 


A 


A 


A 


A 


D 


A 


B 


A 


A 


A 


A 


carnatus B 


F 


a 


A 


I 


C 


A 


C 


a 


B 


A 


A 


A 


A 


D 


A 


B 


A 


A 


A 


A 


carnatus C 


F 


a 


a 


I 


C 


G 


C 


a 


B 


A 


A 


A 


A 


D 


A 


B 


A 


A 


A 


A 


caurinus A 


F 


B 


C 


L 


A 


G 


c 


C 


B 


A 


A 


A 


A 


D 


A 


B 


A 


A 


A 


A 


caurinus B 


F 


B 


C 


L 


A 


G 


J 


C 


B 


A 


A 


A 


A 


D 


A 


B 


A 


A 


A 


A 


chlorosticus 


F 


B 


D 


G 


B 


A 


E 


B 


C 


C 


A 


A 


A 


C 


A 


B 


C 


A 


D 


A 


chrysomelas A 


F 


a 


A 


I 


C 


G 


C 


a 


B 


A 


A 


A 


A 


D 


A 


B 


A 


A 


A 


A 


chrysometas B 


F 


a 


A 


I 


C 


G 


c 


a 


B 


A 


A 


A 


A 


D 


A 


B 


A 


A 


C 


A 


ciliatus variabili 


s A 


D 


A 


M 


A 


A 


F 


B 


F 


C 


A 


A 


A 


D 


A 


B 


C 


A 


A 


A 


constellatus A 


F 


C 


D 


G 


F 


A 


E 


E 


D 


M 


G 


A 


A 


C 


A 


A 


C 


E 


D 


A 


constellatus B 


F 


C 


D 


G 


F 


A 


E 


E 


D 


M 


A 


A 


A 


C 


A 


A 


C 


A 


D 


A 


era;?!?;'! 


A 


D 


A 


M 


A 


A 


F 


B 


F 


C 


A 


A 


A 


D 


A 


B 


C 


A 


A 


A 


da«/ A 


F 


C 


D 


bb 


b 


G 


P 


C 


B 


g 


A 


A 


A 


D 


A 


H 


A 


A 


A 


A 


rfa//; B 


F 


C 


D 


bb 


b 


G 


1 


C 


B 


g 


A 


A 


A 


D 


A 


H 


A 


A 


A 


A 


da//; C 


F 


C 


D 


bb 


b 


G 


P 


c 


a 


b 


A 


A 


A 


D 


A 


H 


A 


A 


A 


A 


da/// D 


F 


C 


D 


bb 


b 


G 


P 


c 


B 


g 


A 


A 


A 


D 


A 


B 


A 


A 


A 


A 


diploproa 


g 


B 


C 


aa 


A 


A 


C 


M 


C 


C 


A 


A 


A 


C 


A 


B 


G 


F 


A 


A 


elongatus A 


W 


C 


I 


h 


A 


A 


c 


M 


P 


C 


A 


A 


A 


D 


A 


B 


I 


A 


A 


A 


elongatus B 


W 


C 


I 


h 


A 


A 


A 


M 


P 


c 


A 


A 


A 


D 


A 


B 


I 


A 


A 


A 


emphaeus A 


F 


C 


D 


E 


A 


A 


D 


B 


D 


c 


A 


A 


A 


B 


A 


B 


C 


A 


C 


A 


emphaeus B 


F 


c 


D 


E 


A 


A 


D 


B 


B 


c 


A 


A 


A 


B 


A 


B 


C 


A 


c 


A 


ensifer A 


F 


c 


D 


F 


B 


Q 


E 


B 


C 


c 


A 


A 


A 


C 


A 


B 


A 


A 


c 


A 


ensifer B 


F 


c 


D 


F 


B 


A 


E 


B 


C 


c 


A 


A 


A 


C 


A 


B 


C 


A 


c 


A 


entomelas 


F 


c 


J 


1 


A 


A 


E 


a 


B 


c 


A 


A 


A 


D 


A 


A 


A 


F 


B 


A 


eos 


F 


B 


D 


G 


B 


A 


E 


B 


C 


c 


A 


A 


A 


C 


A 


B 


C 


A 


D 


A 


exsul 


F 


C 


D 


D 


B 


A 


i 


B 


D 


M 


A 


A 


A 


C 


A 


B 


C 


A 


B 


A 


flavidus 


E 


C 


D 


H 


A 


C 


A 


B 


A 


c 


A 


A 


A 


A 


A 


B 


C 


B 


D 


A 


giUi A 


b 


C 


D 


X 


A 


P 


m 


A 


R 


N 


A 


E 


D 


D 


A 


B 


C 


B 


A 


A 


g//// B 


b 


C 


D 


X 


A 


P 


K 


A 


R 


N 


A 


E 


D 


D 


A 


B 


C 


B 


A 


A 


^oode/ 


Y 


C 


D 


P 


A 


O 


a 


B 


D 


D 


A 


A 


A 


D 


A 


I 


C 


A 


A 


A 


helvoinaculatus A F 


C 


D 


K 


B 


A 


A 


B 


C 


C 


A 


A 


A 


C 


A 


B 


B 


A 


D 


A 








































conti 


nued 



186 



Fishery Bulletin 104(2) 



Table 1 (continued) 


Species 










ND3/ND4 


















12S/16S 










Bs^N I 


Bs/UI 


Cfo I 


Ddel 


Hinill HinSl 


Mho I 


Msp I 


Rsa I 


S(v I 


BrfN I BstV 1 


Cfo I 


DdelHmill Hmtl 


Mho I 


Msp 1 


Rsal 


Sh\ 


helvnmactilatus 


B F 


C 


D 


K 


B 


A 


A 


B 


E 


C 


A 


A 


A 


C 


A 


B 


C 


A 


D 


A 


hopkinsi A 


Z 


c 


A 


n 


A 


g 


K 


B 


C 


e 


A 


A 


A 


c 


A 


B 


C 


A 


B 


A 


hopkinsi B 


Z 


c 


A 


n 


A 


g 


K 


A 


C 


e 


A 


A 


A 


c 


A 


B 


c 


A 


B 


A 


hopkinsi C 


Z 


c 


A 


u 


A 


g 


g 


B 


C 


e 


A 


A 


A 


c 


A 


A 


c 


A 


B 


A 


hiihbsi A 


L 




D 


R 


A 


I 





A 


C 


G 


A 


C 


A 


D 


A 


A 


c 


A 


A 


A 


huhhui B 


M 




D 


S 


A 


I 





A 


C 


G 


A 


C 


A 


D 


A 


A 


c 


A 


A 


A 


hiibhsi C 


M 




D 


T 


A 


I 


P 


A 


c 


G 


A 


c 


A 


D 


A 


B 


c 


A 


A 


A 


huhhsi D 


M 




D 


S 


A 


F 





A 


c 


G 


A 


c 


A 


D 


A 


A 


c 


A 


A 


A 


inermis A 


F 




D 


X 


A 


A 


D 


e 


J 


I 


A 


A 


A 


A 


A 


B 


c 


A 


C 


A 


inermis B 


F 


H 


D 


X 


A 


A 


D 


e 


J 


I 


A 


A 


A 


A 


A 


B 


c 


A 


C 


A 


inermis C 


F 


I 


D 


cc 


A 


I 


D 


e 


J 


I 


A 


A 


A 


A 


A 


B 


c 


A 


c 


A 


jordani A 


a 


c 


D 


c 


E 


A 


c 


A 


p 


C 


F 


A 


A 


H 


A 


B 


c 


A 


D 


A 


jordani B 


a 


c 


D 


c 


E 


A 


d 


A 


p 


c 


F 


A 


A 


H 


A 


B 


c 


A 


D 


A 


jordani C 


a 


c 


D 


q 


E 


A 


c 


A 


p 


c 


F 


A 


A 


H 


A 


B 


c 


A 


D 


A 


joyneri A 





H 


D 


a 


A 


A 


D 


e 


L 


A 


A 


A 


A 


A 


A 


B 


c 


A 


C 


A 


joyneri B 


O 


H 


D 


a 


A 


I 


D 


e 


L 


A 


A 


A 


A 


A 


A 


B 


c 


A 


C 


A 


/(?;!/(gnosus 


F 


C 


D 


F 


B 


A 


n 


B 


D 


M 


A 


A 


A 


C 


A 


B 


c 


A 


B 


A 


/et'i's 


B 


A 


cr 

to 


d 


A 


A 


q 


d 


B 


C 


A 


A 


A 


D 


A 


B 


c 


A 


D 


A 


mafrfona/c/( 


P 


b 


A 


e 


A 


b 


c 


g 


D 


C 


B 


A 


A 


G 


A 


G 


c 


A 


A 


D 


»!a/;g^(?/- 


F 


B 


D 


L 


C 


D 


c 


C 


B 


A 


A 


A 


A 


D 


A 


B 


A 


A 


B 


A 


melanops A 


D 


C 


D 


F 


A 


B 


E 


A 


A 


C 


A 


A 


A 


D 


A 


B 


c 


B 


D 


A 


melanops B 


D 


C 


D 


F 


A 


B 


E 


B 


A 


C 


A 


A 


A 


D 


A 


B 


c 


B 


D 


A 


me/anostomou.s 


A 


C 


D 


g 


A 


A 


K 


A 





C 


A 


A 


A 


D 


A 


B 


c 


B 


C 


A 


me/a?ioston!0!/.s 


B 


c 


D 


g 


A 


A 


K 


A 


B 


C 


A 


A 


A 


D 


A 


B 


c 


B 


C 


A 


miniatus A 


F 


c 


K 


r 


C 


A 


e 


B 


A 


C 


A 


A 


A 


C 


A 


A 


c 


A 


A 


A 


miniatus B 


F 


c 


K 


r 


c 


N 


e 


B 


A 


C 


A 


A 


A 


C 


A 


A 


c 


A 


A 


A 


mystinus A 


F 


c 


J 




A 


A 


E 


a 


B 


C 


A 


A 


A 


D 


A 


A 


A 


B 


B 


A 


mystinus B 


B 


c 


J 




A 


C 


E 


D 


B 


c 


A 


A 


A 


D 


A 


A 


A 


A 


B 


A 


mystinus C 


F 


c 


J 




A 


A 


E 


a 


B 


e 


A 


A 


A 


D 


A 


A 


A 


B 


B 


A 


mystinus D 


B 


c 


J 




A 


A 


E 


D 


B 


C 


A 


A 


A 


D 


A 


A 


A 


B 


B 


A 


mystinus E 


B 


c 


J 




A 


C 


E 


D 


B 


C 


A 


A 


A 


D 


A 


A 


A 


B 


B 


A 


nebulosus A 


F 


G 


A 


Z 


A 


G 


C 


C 


B 


A 


A 


A 


A 


D 


A 


B 


A 


A 


A 


A 


nehulosus B 


F 


G 


A 


Z 


B 


G 


C 


C 


B 


A 


A 


A 


A 


D 


A 


B 


A 


A 


A 


A 


n!groc(/ic<us 


F 


C 


D 


c 


A 


A 


G 


B 


B 


C 


A 


A 


A 


C 


A 


B 


C 


A 


B 


A 


ni'yosus 


N 


A 


D 


U 


A 


I 


C 


B 


I 


A 


A 


A 


A 


G 


A 


A 


C 


A 


A 


C 


oca/Zs 


Z 


C 


A 





A 


g 


K 


B 


C 


e 


A 


A 


A 


C 


A 


B 


c 


A 


B 


A 


paucispinis A 


F 


A 


b 


c 


A 


A 


V 


A 


b 


d 


A 


A 


A 


D 


A 


B 


c 


A 


B 


A 


paucispinis B 


F 


A 


b 


c 


g 


A 


V 


A 


b 


d 


A 


A 


A 


D 


A 


B 


c 


A 


B 


A 


p/(;7/(p,s; 


B 


C 


D 


m 


A 


A 


b 


B 


B 


e 


A 


A 


A 


D 


A 


B 


c 


B 


A 


A 


p»?/i(,ger 


F 


C 


D 


s 


A 


A 


C 


B 


B 


A 


A 


A 


A 


D 


A 


A 


c 


A 


A 


A 


polyspinis A 


A 


D 


A 


M 


A 


A 


F 


B 


F 


C 


A 


A 


A 


D 


A 


B 


c 


A 


A 


A 


poly spin is B 


A 


J 


A 


f 


A 


A 


F 


A 


F 


C 


A 


A 


A 


D 


A 


B 


c 


A 


A 


A 


polyspinis C 


A 


D 


A 


M 


A 


A 


F 


B 


F 


c 


A 


A 


A 


I 


A 


B 


c 


A 


A 


A 


polyspinis D 


V 


D 


A 


M 


A 


A 


F 


B 


N 


c 


A 


A 


A 


D 


A 


B 


c 


A 


A 


A 


pronger 


F 


C 


D 


E 


A 


A 


K 


E 


B 


c 


A 


A 


A 


D 


A 


B 


c 


A 


C 


A 


ra.s?re//jger 


c 


c 


C 


V 


A 


a 


A 


C 


B 


A 


A 


A 


A 


D 


A 


B 


A 


A 


A 


A 


reed; 


A 


D 


A 


M 


A 


A 


r 


B 


F 


C 


A 


A 


A 


D 


A 


B 


c 


A 


A 

con 


A 
tinned 



Li et al A key to selected Sebastes spp based on mitoctiondrial DNA restriction fragment analysis 



187 



Table 1 (continued) 


Species 










ND3/ND4 


















12S/16S 










Bs(NI 


Ss/UI 


C/bl 


flrffi 


ffindll Hint I 


Mho I 


)hp 1 


ffsol 


Sly\ 


Bs(\I BrfUI 


Cfn I 


Dde\ ffindll ffinfl 


Mbo 1 


Msp I 


Rstt I 


Sty I 


rosace us 


F 


c 


D 


k 


B 


A 


E 


B 


B 


C 


A 


A 


A 


C 


A 


B 


C 


A 


B 


A 


rosenblatti 


F 


B 


D 


G 


B 


A 


E 


B 


C 


C 


A 


A 


A 


c 


A 


B 


C 


A 


D 


A 


ruhcrrimiis A 


C 


C 


A 


B 


A 


D 


I 


B 


D 


c 


A 


A 


A 


D 


A 


C 


c 


A 


A 


A 


ruberrimus B 


C 


C 


D 


B 


A 


D 


I 


B 


D 


c 


A 


A 


A 


D 


A 


C 


c 


A 


A 


A 


rubrivinctus 


g 


C 


D 


D 


A 


A 


E 


b 


B 


c 


A 


A 


A 


C 


A 


B 


c 


B 


B 


A 


rufus A 


F 


C 


D 


aa 


A 


A 


F 


B 


B 


e 


E 


A 


A 


C 


A 


B 


c 


D 


D 


A 


rufus B 


F 


c 


D 


aa 


A 


A 


F 


B 


B 


e 


E 


A 


A 


c 


A 


B 


c 


D 


B 


A 


rufus C 


F 


c 


D 


aa 


A 


A 


C 


B 


B 


e 


E 


A 


A 


c 


A 


B 


c 


D 


B 


A 


saxicola 


O 


c 


D 


t 


A 


G 


f 


B 


C 


C 


G 


F 


C 


D 


A 


B 


c 


A 


A 


A 


semicinctus 


X 


c 


D 


J 


A 


C 


P 


A 


Q 


L 


A 


D 


A 


A 


A 


A 


H 


A 


A 


A 


serrenoides A 


D 


c 


D 


D 


B 


C 


f 


B 


P 


C 


A 


A 


A 


D 


A 


B 


C 


B 


D 


A 


serrenoides B 


D 


c 


C 


F 


B 


C 


f 


B 


P 


C 


A 


A 


A 


D 


A 


B 


C 


B 


D 


A 


serriceps A 


F 


c 


D 


y 


A 


g 


E 


B 


B 


C 


A 


A 


A 


C 


A 


B 


C 


A 


B 


A 


serriceps B 


F 


c 


D 


z 


A 


g 


E 


B 


B 


C 


A 


A 


A 


C 


A 


B 


c 


A 


B 


A 


simulator 


F 


K 


D 


F 


A 


A 


E 


B 


C 


C 


A 


A 


A 


C 


A 


B 


c 


A 


D 


A 


spmorbis 





C 


D 


D 


B 


A 


E 


B 


D 


M 


A 


A 


A 


C 


A 


B 


c 


A 


B 


A 


taczanowski A 


P 


A 


D 


W 


A 


I 


R 


H 


C 


I 


A 


A 


A 


A 


A 


D 


c 


B 


A 


C 


taczanowski B 


P 


A 


D 


w 


C 


I 


R 


H 


C 


I 


A 


A 


A 


A 


A 


D 


c 


B 


A 


C 


taczanowski C 


P 


A 


D 


w 


C 


I 


R 


H 


C 


I 


A 


A 


A 


A 


A 


D 


c 


A 


A 


c 


thompsoni 





H 


D 


Y 


A 


A 


D 


e 


L 


I 


A 


A 


A 


A 


A 


B 


c 


A 


C 


A 


trivittatus A 


F 


A 


D 


V 


A 


F 


K 


E 


C 


H 


A 


A 


A 


A 


A 


E 


B 


B 


A 


c 


trivittatus B 


F 


A 


D 


V 


A 


D 


K 


E 


C 


H 


E 


A 


A 


A 


A 


E 


B 


C 


A 


c 


trivittatus C 


F 


A 


D 


V 


B 


D 


K 


E 


B 


H 


E 


A 


A 


A 


A 


E 


B 


C 


A 


c 


trivittatus D 


F 


A 


D 


V 


B 


D 


K 


E 


C 


H 


E 


A 


A 


A 


A 


E 


B 


C 


A 


c 


umbrosus 


F 


C 


D 


w 


B 


A 


n 


B 


D 


M 


A 


A 


A 


C 


A 


B 


A 


A 


B 


A 


variegatus 


F 


C 


D 


E 


A 


A 


D 


B 


B 


C 


A 


A 


A 


B 


A 


B 


C 


A 


C 


A 


vulpes A 


B 


D 


D 


V 


A 


A 


h 


B 


C 


G 


A 


C 


A 


A 


A 


E 


B 


B 


A 


B 


vulpes B 


B 


D 


D 


V 


A 


A 


h 


B 


C 


H 


A 


C 


A 


A 


A 


E 


B 


B 


A 


B 


vulpes C 


B 


D 


D 


i 


A 


A 


h 


B 


C 


H 


A 


c 


A 


A 


A 


E 


B 


B 


A 


B 


wilsoni A 


F 


C 


D 


E 


A 


A 


D 


B 


B 


C 


A 


A 


A 


B 


A 


B 


C 


A 


C 


A 


wilsoni B 


F 


C 


A 


i 


A 


A 


D 


B 


B 


C 


A 


A 


A 


B 


A 


B 


C 


A 


C 


A 


zacentrus A 


F 


A 


D 


E 


A 


A 


D 


B 


C 


c 


A 


A 


A 


B 


A 


B 


C 


A 


c 


A 


zacentrus B 


F 


A 


D 


A 


A 


A 


D 


B 


C 


c 


A 


A 


A 


B 


A 


B 


C 


A 


c 


A 


zacentrus C 


F 


C 


D 


E 


A 


A 


D 


B 


C 


c 


A 


A 


A 


B 


A 


B 


C 


A 


c 


A 


zacentrus D 


F 


A 


D 


E 


A 


A 


D 


B 


C 


B 


A 


A 


A 


B 


A 


B 


C 


A 


c 


A 


Helicolenus 










































hilgendorfi 


J 


G 


F 


Q 


A 


A 


N 


G 


H 


A 


D 


B 


B 


F 


B 


B 


F 


B 


A 


A 


Sebastolobus 










































alascanus A 


H 


E 


E 





A 


H 


M 


F 


G 


F 


C 


B 


B 


E 


A 


D 


E 


B 


E 


B 


Sebastolobus 










































alascanus B 


I 


E 


E 





A 


H 


M 


F 


G 


F 


C 


B 


B 


E 


A 


D 


E 


B 


E 


B 


Sebastolobus 










































alascanus C 


H 


F 


E 





A 


H 


M 


F 


G 


E 


C 


B 


B 


E 


A 


D 


E 


B 


E 


B 


Sebastolobus 










































alascanus D 


H 


E 


E 


P 


A 


H 


M 


F 


G 


F 


C 


B 


B 


E 


A 


D 


E 


B 


E 


B 



188 



Fishery Bulletin 104(2) 



Identification key 


Key for distinguishing Sebastes spp. using ND3/ND4 restriction site information. This is one of many possible schemes to identify 


unknown specimens to species level or to small groups of species. Because new intraspecific variation may be encountered, it is 


wise to confirm the identification by cutting the DNA usi 


ng an additional one or two enzymes. 


1. Digest mtDNA with M6o I. 


ii. 0. 


2. Place resulting restriction fragment patterns 


Digest mtDNA with Dde I. 


ihaplotypes) in one of the following categories 


1. Y — S. tlwmpsoni 


isee Appendices 1 and 2 for details). 


2. a — S.Joyneri 


a. 0— S. hubbsi A. B, D. 


h. c — S.jordani A 


b. P—S.hubb.siC. 


i. n—S. babcocki B 


c. AorB. 


j. g — S. hopkinsi A 


Digest mtDNA with S.9/N I. 


k. h — S. vulpes 


i. B — S. alutus 


1. E. 


ii. W — S. elongatiis A 


Digest mtDNA with Ss/N I. 


iii. E — S. flavidus 


i. B — S. mystinus A 


iv. F — S. helvomaculatus 


ii. — S. spinorbis 


V. C — S. rastrelliger 


iii. D — S. melanops 


d. p or 1. 


iv. F or g. 


Digest mtDNA with B.s/N I. 


Digest mtDNA with Dde I. 


i. X — S. semicinctus 


1. 1. 


ii. F—S.dalli 


Digest mtDNA with BstN I. 


e. R — S. taczanowski 


a. F — S. entomelas/ S. mystinus A, C 


f. q,V, C,J,f,j. 


b. B—S. mystinus B, D 


Digest with Bsm I. 


2. D — S. ruhrivinctus 


i. B— S. levis 


3. G. 


ii. N — S. nivosus 


Digest mtDNA with BstV I. 


lii. O. 


a. C — S. constellatus 


Digest mtDNA with BstV I. 


b. B — S. chlorostictus, S. eos, S. rnsenblatti 


1. B — S. atrovirens 


(identical) 


2. C—S. saxicola 


4. z — S. serriceps A 


iv. W — S. elongatus 


5. k or y. 


V. g — S. diplnproa 


Digest mtDNA with Hind II. 


vi. F. 


a. A — S. serriceps B 


Digest mtDNA with BstV I. 


6. B — S. I'osaceus 


1. A — S. paucispinis 


m. G — S. babcocki A 


2. C. 


n. K. 


Digest mtDNA with Dde I. 


Digest mtDNA with BstN I. 


a. s — S. pinniger 


i. Z — S. hopkinsi B 


h. aa — S. rufus C 


ii. — S. melanostomus 


3. a — S. carnatus, S. civynomelas 


iii. F. 


4. B. 


Digest mtDNA with H(;!f I. 


Digest mtDNA with Cfo I. 


1. A — S. proriger 


a. D — S. inaliger 


2. D— S. trivittatus 


b. C — S. caurinus 


3. E — S. aleutianus A 


5. G — S. nebulosus 


4. BorF. 


vii. P — S. macdonaldi 


Digest mtDNA with BstV I. 


g. D or e. 


a. A — S. trivittatus 


Digest mtDNA with B.s?N I. 


b. C — S. aleutianus B 


i. F. 


iv. G — S. brevispinis 


Digest mtDNA with Dde I. 


0. a — S. goodei 


1. A — S. zacentrus B 


p. F. 


2. E. 


Digest mtDNA with BstN I. 


Digest mtDNA with Rsa I. 


i. A — S. crameri, S. ciliatus, S. polyspinis A, B (first 2 spp. 


a. C—S. zacentrus A,C.D 


and S. polyspinis A identical) 


b. B — S. emphaeus B, S. 


Digest mtDNA with BstV I. 


variegatus, S. wilsoni A 


1. D — S. crameri, S. ciliatus. S. polyspinis A 


(identical) 


2. J — S. polyspinis B 


r. D — S. emphaeus A 


ii. F. 


3. cc — S. inermis C 


Digest mtDNA with Dde I. 


4. Xorr. 


1. D— S. borealis 


Digest mtDNA with BstXi I. 


2. aa — S. rufus A, B 


a. C — S. miniatus 


iii. V — S. polyspinis C 


b. H — S. inermis A 


q. d — S.jordani B 


c. I — S. inermis B 


r. I — S. ruberrimus 


5. i — S. wilsoni B 


s. i — S. e.xsut 




t. b — S. aurora 




u. k — S. auriculatus 




v. r — S. reedi 



Li et al a key to selected Sebostes spp, based on mitochondrial DNA restriction fragment analysis 



189 



product with eiidonuclease BstN I. Continue digesting 
with the specified restriction endonucleases, identifying 
the resultant haplotype, and continue to the next step 
until the species has been identified. Between one and 
five restriction enzymes will be needed to achieve the 
separation. Examining the fragment patterns of addi- 
tional restriction enzymes can increase confidence in the 
identifications (see key). 

Although variation observed in the 12S/16S region 
was not used in the key, it does provide alternatives to 
species identification that may be useful for resolving 
the identification of some species or for corroborating 
identifications, especially when there is intraspecific 
variation that has not been previously observed for the 
enzymes applied in resolving species. In particular, Rsa 
I, Dde I, and Mbo I exhibited substantial interspecific 
variation in the 12S/16S region. In addition, if one has 
reduced the possible species by other means, a single 
12S/16S digest can be used in some instances. For ex- 
ample, S. aleutianus and S. borealis differ in fragments 
produced by Bs^N I and by Rsa I, and S. caurinus and 
S. maliger also differ in fragments produced by Rsa I. 



Discussion 

Restriction site analysis of mtDNA is a simple and effec- 
tive tool for identifying juveniles of Sebastes species. 
Because inexpensive equipment is used to obtain data 
from restriction fragment patterns, the analyses can 
be conducted in most laboratories, including many high 
school laboratories. 

Although the restriction fragment key identified most 
(58 or 81.7%) of the 71 different rockfish species we 
evaluated, it failed to identify 13 species. The identities 
of those 13, however, were narrowed to five small groups 
of species. One group, S. carnatus and S. chrysomelas, 
are very closely related; they differ obviously only in 
body coloration as adults: S. carnatus has flesh-col- 
ored blotches on an olive-brown background, and S. 
chrysomelas has yellow blotches on a black background 
(Love et al., 2002). They are included in the subgenus 
Pteropodus (Kendall, 2000). Another group includes S. 
chlorostictus, S. eos, and S. rosenblatti, which are also 
morphologically similar, occur sympatrically, and are 
closely related (Chen, 1971; Love, 1996; Rocha-Olivares 
et al., 1999; Love et al., 2002); all are members of the 
subgenus Sebastomus. Estimates of divergence times 
of the subgenus Sebastomus suggest that the three 
species are the result of the most recent speciation 
events within the subgenus, which may have begun less 
than 140,000 years ago (Rocha-Olivares et al., 1999). 
Members of a third group, S. emphaeus-S. variegatus- 
S. wilsoiii, are assigned to the subgenus Allosebastes 
(Kendall, 2000). The relationships between species in 
the other two unresolved groups are not as clear be- 
cause the subgenera of the members differ. In one of 
those groups, S. entomelas is in subgenus Acutomen- 
tum and S. mystinus is in subgenus Sebastosomus. In 
the other subgroup, S. polyspinis remains unassigned 



to a subgenus, S. crameri is in Eosebastes, and iS. cil- 
iatus (subgenus Sebastosomus) has only recently been 
separated from S. variabilis (Orr and Blackburn, 2004). 
The similarities observed may actually reflect genetic 
relationships because the subgenera assignments are 
probably inaccurate reflections of phylogeny (Kendall, 
2000). Our inability to resolve within those five groups 
of rockfish to species indicates the need for additional 
markers. Approaches for obtaining such markers include 
screening additional regions of the mtDNA and appli- 
cation of additional restriction enzymes. If additional 
mtDNA regions and restriction enzymes do not provide 
species-specific information, other molecular techniques 
such as microsatellites should be considered. 

We applied the baseline data from which our key was 
developed to identification of recently extruded rock- 
fish larvae in Southeast Alaskan waters (Gray et al., 
2006). That application, which was made while the key 
presented in this article was being developed, evolved 
over subsequent years of application. The key was also 
used to delineate juvenile rockfishes from the southern 
California Bight (Li et al., in press). We were able to 
identify all the specimens to species or to narrow iden- 
tification to one of the five small groups of genetically 
similar species. The identifications of juvenile rockfish 
were concordant with genetic and morphological crite- 
ria, except that the genetic key did resolve species in 
the closely related subgenus Sebastotyius, which could 
not be resolved with morphological criteria. The larval 
specimens, in contrast, could not be identified to spe- 
cies with morphological criteria. We detected previously 
unobserved intraspecific variation in both studies as 
well. Neither the variation observed in the larval and 
juvenile studies, the intraspecific variation observed in 
developing the key, nor intraspecific variation observed 
among large numbers of individuals of several species 
examined as part of a population genetic analysis (Li, 
2004) obscured species detection. 

Molecular genetic keys can remove much of the guess- 
work for species determination in ichthyoplankton and 
juvenile surveys. Because DNA-based characters re- 
main constant throughout the life of an individual, 
genetic divergence can provide unequivocal markers 
for species delineation. For specific questions involving 
the identification of small numbers of species, such as 
distinguishing between the two sibling species included 
in S. aleutianus (Gharrett et al., 2005), it is possible 
to develop assay methods (e.g., single nucleotide poly- 
morphism — referred to as SNP (Collins et al., 1996), 
that can rapidly identify large numbers of specimens. 
Finally, such a key can be used to identify larval and 
juvenile species of Sebastes so that the variation in the 
morphological characters can be determined. 



Acknowledgments 

This work represents, in part, the master's thesis work 
of Z. Li at the University of Alaska Fairbanks. The pro- 
ject was supported by funding from the U.S. Geological 



190 



Fishery Bulletin 104(2) 



Survey (Biological Resources Division), Western Regional 
Office in Seattle, WA (R.W.O. 32) to AJG. AJG was at 
Hokkaido University and Kitasato University and was 
supported by the Japan Society for Promotion of Science 
when he contributed some of the effort to this study. 



Literature cited 

Barrett, I., J. Joseph, and H. G. Moser. 

1966. Electrophoretic analysis of hemoglobins of Cali- 
fornia rockfish (genus St'6as?rif/es). Copeia 1966:489-494. 
Chen, L.-C. 

1971. Systematics, variation, distribution, and biol- 
ogy of rockfishes of the subgenus Sebastotnus (Picses, 
Scorpaenidae, Sebastes). Bull. Scripps. Inst. Oceaogr. 
18:1-115. 
Collins, F. S., M. S. Guyer, and A. Chakravarti. 

1996. Variations on a theme: cataloging human DNA 
sequence variation. Science 278:1580-1581. 
Gharrett, A. J., A. K. Gray, and J. Heifetz. 

2001. Identification of rockfish iSehastes spp.) from 

restriction site analysis of the mitochondrial ND-3/ 

ND-4 and 12S/16S rRNA gene regions. Fish. Bull. 

99:49-62. 

Gharrett, A. J., A. P. Matala. E. L. Peterson, A. K. Gray, Z. Li, 

and J. Heifetz. 

2005. Two genetically distinct forms of rougheye rock- 
fish are different species. Trans. Am. Fish. Soc. 
134:242-260. 
Gray, A. K., A. W. Kendall Jr, B. L. Wing, M. G. Carls, 
J. Heifetz, Z. Li, and A. J. Gharrett. 

In press. Identification of larval rockfishes i Sebastes 
spp.) in Southeast Alaskan waters was possible 
using mitochondrial markers but not pigmentation 
patterns. Trans. Am. Fish. Soc. 134. 
Gyllensten, U., D. Wharton, A. Josefsson, and A. C. Wilson. 

1991. Paternal inheritance of mitochondrial DNA in 
mice. Nature 352:255-257. 
Kendall, A. W., Jr 

1991. Systematics and identification of larvae and 
juveniles of the genus Sebastes. Environ. Biol. Fish. 
30:173-190. 

2000. An historic review of Sebastes taxonomy and 
systematics. Mar. Fish. Rev. 62:1-23. 

Laidig, T. E., and S. Ralston. 

1994. The potential use of otolith characters in iden- 
tifying larval rockfish (Sebastes spp.). Fish. Bull. 
93:166-171. 
Laidig, T. E., K. M. Sakuma. and M. M. Nishimoto. 

1996. Description of pelagic larval and juvenile Sebastes 
saxicola (Family Scorpaenidae), with an examination 
of larval growth. Fish. Bull. 94:289-299. 
LeClair, L. L., and R. M. Buckley. 

2001. Electrophoretic identification of juvenile rockfish 
(Genus Sebastes) recruiting to drifting algae and sea- 
grass habitats off the Washington coast. Northwest 
Sci. 75:53-60. 

Li, Z. 

2004. Phylogenetic relationship and identification of 



juveniles of the genus Sebastes. M.S. thesis, 146 p. 

Univ. Alaska Fairbanks, Fairbanks AK. 
Li, Z., M. M. Nishimoto, M. S. Love, and A. J. Gharrett. 

In press. Comparing the identification of southern 

California juvenile rockfishes (genus Sebastes spp.i by 

restriction site analysis of the mitochondrial ND3/ND4 

region and by morphological characteristics. Fish. Bull. 

104(3). 
Love, M. S. 

1996. Probably more than you want to know about the 

fishes of the Pacific coast, 386 p. Really Big Press. 

Santa Barbara, CA. 
Love, M. S., M. Yoklavich, and L. Thorsteinson (eds. ). 

2002. The rockfishes of the Northeast Pacific. 405 
p. Univ. California Press, Berkeley, CA. 

Moritz, C, T. E. Dowling, and W. M. Brown. 

1987. Evolution of animal mitochondrial DNA: relevance 
for population biology and systematics. Ann. Rev. Ecol. 
Syst. 18:269-292. 

Moser, H. G. 

1996. Scorpaenidae: scorpionfishes and rockfishes. //; 
The early stages of fishes in the California Current 
region (H. G. Moser, ed.), p. 793-795. CalCOFI Atlas 
33. 

Orr J. W., and J. E. Blackburn. 

2004. The dusky rockfishes iTeleostei: Scorpaeniformes) 
of the North Pacific Ocean: resurrection of Sebastes 
variabilis (Pallas, 1814) and a redescription of Sc- 
bastes ciliatus (Tilesius, 1813). Fish. Bull. 102:328-348. 

Rocha-Olivares, A. 

1998. Multiplexing haplotype-specific PCR: a new 
approach for species identification of the early life 
stages of rockfishes of the species-rich genus Sebastes 
Cuvier J. Exp. Mar Biol. Ecol. 231:279-290. 

Rocha-Olivares, A., H. G. Moser, and J. Stannard. 

2000. Molecular identification and description of pelagic 

young of the rockfishes Sebastes constellatus and Sebastes 

ensifer. Fish. Bull. 98:353-363. 
Rocha-Olivares, A., R. H. Rosenblatt, and R. D. Vetter 

1999. Molecular evolution, systematics, and zoogeography 
of the rockfish subgenus Sebastoiniis iSebastes. Scor- 
paenidae I based on mitochondrial cytochrome 6 and con- 
trol region sequences. Mol. Phy. Evol. 11:441-458. 

Rokas, A., E. Ladoukakis. and E. Zouros. 

2003. Animal mitochondrial DNA recombination 
revisited. Trends Ecol. Evol. 18:411-417. 

Sambrook, J., E. F. Fritsch, and T. Maniatis. 

1989. Molecular cloning: a laboratory manual, 2""^ 
ed. Cold Spring Harbor Laboratory Press, Cold Spring 
Harbor, NY. 
Seeb, L. W. 

1986. Biochemical systematics and evolution of the Scor- 
paenidae genus Sebastes. Ph.D. diss., 176 p. Univ. 
Washington, Seattle, WA. 
Seeb, L. W., and A. W. Kendall Jr 

1991. Allozyme polymorphisms permit the identifi- 
cation of larval and juvenile rockfishes of the genus 
Sebastes. Environ. Biol. Fish. 30:191-201. 
Seutin, G., B. N. White, and P T. Boag. 

1991. Preservation of avian blood and tissue samples 
for DNA analvsis. Can. J. Zool. 69:82-90. 



Li et a\ A key to selected Sebastes spp based on mitocliondnal DNA restriction fragment analysis 



191 























Appendix 


1 
























Restriction sit 


c locations foi 


Sebastes 


spp., 


Seha 


stnli 


hiis a 


lascanus. 


and Helii 


olcinis III 


IgendOrfi 


n the ND3/ND4 and 12S/16S 1 


mtDNA region 


s. Letters denote hapioti 


•pes 


resti 


iction dige 


st patterns). "X 


" denotes 


presence 


of site. "O' 


denotes absence of site. 


ND3/ND4 
























haplotypes 






















BstN I 
sites 


















































A 


B 


c 


D 


E 


F 


G 


H 


I 


J 


L 


M 


N 





P 


V 


W 


X 


Y 


Z 


a 


b 


c 


g 


112 





X 


X 


X 


X 


X 


X 


X 


X 


X 





X 


X 


X 


X 








X 


X 





X 


X 


X 


X 


291 


O 


O 


o 





X 


O 


O 











O 





X 








O 


O 








o 


O 


X 


X 


O 


551 














O 


O 


O 


X 


X 




















O 








o 


o 











O 


829 


X 


X 





X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


o 











X 


X 


X 


1035 


O 











O 


o 











X 














o 


O 








o 


o 











O 


1214 


X 


X 


X 


o 


O 


X 


X 








o 


o 





X 


o 


X 





X 


o 


X 


o 


X 


X 


X 


X 


1261 





o 


o 














o 


X 


o 


X 


X 


X 





o 


O 


O 




















o 


1495 
































X 


X 


o 





o 





o 


o 


X 











o 





1694 








X 


X 


X 


X 


X 


o 





o 











X 


X 





X 


X 





X 





o 


X 


X 


1982 


O 








X 


X 


o 


X 


o 








o 


O 


o 


o 














X 











o 


o 


2010 











o 


O 


o 














o 











o 





























2121 





o 


o 














o 


o 














o 


X 











o 








X 





o 


2255 





























o 


o 


O 











o 

















o 








2325 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 





X 


o 


ND3/ND4 












haplotypes 


































BstVl 












































































sites 


A 


B 


c 


D 


E 


F 


G 


H 


I 


J 


K 


a 


b 
























4' 


(Xi 


(Xi 


(Xi 


(Xi 


(Xi 


(Xi 


(Xi 


iX) 


(X) 


iXi 


iX) 


(Xi 


(X) 
























344 








o 








X 


o 


o 







































668 











o 


o 


o 





X 





o 

































710 








o 








o 


o 





X 








o 


X 
























1499 














X 


o 

















o 



























1854 





o 


o 


X 

















X 


X 


o 


o 
























2025 





X 


X 


o 











X 


o 





X 


X 


X 
























2109 








o 

















X 


X 





X 



























2306 





X 








X 


X 


X 








o 


o 


X 



























1 The site at 4 


is in 


the 


primer region. 
















































haplotypes 




































C/bl 










































































sites 


A 


B 


C 


D 


E 


F 


I 


J 


K 


a 


b 


g 


























709 


X 


O 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


























1221 








o 


O 


X 


X 


X 


o 


o 



































1436 





o 








X 

















o 


o 


























1464 














o 








X 


o 


X 
































1472 





o 


o 














o 








X 


X 


























1741 





X 


X 


X 








X 


X 


X 


o 


X 


X 


























1762 








o 














o 


X 



































1813 








X 

















o 


o 


X 





























ND3/ND4 
Ddel 
























haplotyp 


es 






































































sites 


A 


B 


c 


D 


E 


F 


G 


H 


I 


J 


K 


L 


M 


N 


o 


p 


Q 


R 


s 


T 


u 


V 


w 




65 














O 








o 





o 


o 

















X 






















195 


X 





o 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 














o 


o 


X 


X 







266 

















o 


X 





X 








X 








X 


X 


X 






















298 


X 


X 





X 


X 


X 


X 


o 


X 


X 


X 


X 


X 


X 








X 











o 


X 


X 




336 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 




374 


X 


X 


X 





X 








o 





X 








X 


o 


o 








X 


X 


X 





X 





















































continued 



192 



Fishery Bulletin 104(2) 





















Appendix 


1 (continued) 






















ND3/ND4 (continued 


















haplotypes 






















Ddel 






























































































sites 


A 


B 


c 


D 


E 


F 


G 


H 


I 


J 


K 


L 


M 


N 





p 


Q 


R 


s 


T 


u 


V 


w 


590 





X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 








X 


X 


X 


X 


X 


X 


X 


624 








X 






























































663 


















































X 




















695 


























X 








X 



































754 


























X 








X 


X 











X 











X 


X 


X 


1144 







































































1190 





















































X 


X 


X 











1257 


















































X 




















1409 
































X 






































1560 


























X 








X 

















X 

















1577 












































X 


X 


X 




















1587 







































































1601 


X 





X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 











X 


X 


X 


X 


X 


X 


1696 


















































X 




















1722 





























X 











X 


X 


X 


X 




















1779 








X 

















X 








X 


X 














X 


X 


X 








X 


1795 

















X 


X 


X 


X 


X 


X 


X 


X 











X 











X 


X 


X 


1912 


X 


X 


X 


X 


X 


X 


X 


X 





X 


X 


X 


X 


X 








X 


X 


X 





X 


X 


X 


2091 












































X 

















X 








2122 







































































2188 












































X 


X 























2374-' 


iXi 


iX) 


iX) 


iX) 


iXl 


1X1 


(Xl 


(Xl 


iXi 


(Xl 


iXi 


iXi 


iX) 


iXi 


(Xl 


(X) 


(Xl 


(Xl 


(Xl 


(X) 


(Xl 


iXi 


iX) 


ND3/ND4 




















haplotypes 


continued) 




















Ddel 






























































































sites 


X 


Y 


z 


a 


b 


c 


d 


e 


f 


g 


h 


i 


J 


k 


1 


m 


n 





p 


q 


r 


s 


t 


65 







































































195 








X 


X 





X 


X 





X 


X 


X 


X 


X 


X 


X 


X 


X 


X 





X 


X 


X 


X 


266 








X 





X 























X 

















X 











X 


298 


X 





X 





X 


X 


X 


X 


X 


X 





X 


X 


X 





X 


X 


X 


X 


X 





X 





336 


X 





X 





X 


X 


X 


X 


X 


X 


X 


X 


X 


X 





X 








X 


X 


X 


X 


X 


374 


X 


X 





X 











X 


X 


X 





X 


























X 


X 





590 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


624 







































































663 







































































695 








X 





X 
























































754 


X 





X 








X 








X 








X 





X 





X 


X 


X 





X 











1144 



























































X 











1190 







































































1257 







































































1409 


























X 












































1560 


X 


X 


X 


X 


X 
























































1577 







































































1587 

































































X 





1601 


X 


X 


X 


X 


X 





X 





X 


X 


X 


X 





X 


X 


X 


X 


X 








X 


X 


X 


1696 







































































1722 







































































1779 


X 


X 


X 


X 


X 











X 





X 





X 





























X 


1795 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 





X 





X 


X 





X 


X 


X 


X 


X 


X 


1912 


X 


X 


X 


X 





X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


2091 




















X 


















































2122 








X 






























































2188 







































































23742 


(X) 


(X) 


iXi 


(X) 


IX) 


iXi 


(Xl 


1X1 


(X) 


iX) 


IX) 


iX) 


(Xl 


(X) 


iX) 


(Xl 


(Xl 


(X) 


iX) 


iX) 


(Xl 


(X) 


(X) 














































continued 



Li et al A key to selected Sebastes spp based on mitochondrial DNA restriction fragment analysis 



193 

















Appendix 


1 (cont 


nued) 














ND3/ND4 
Ddel 

sites 






h 


iplotypes (continued 










Hind II 

sites 






haplotype 


s 




u 


V 


w 


X 


y 


z 


aa 


bb 


cc 




A 


B 


C 


E 


F 


b g 


65 


O 











o 
















308 





o 











X 


195 


X 


X 


X 


O 


X 


X 


X 


X 


X 




320 


o 





O 





X 


O 


•266 





X 


O 


o 








O 


X 







324 


o 














X 


298 


X 


X 


X 





X 


X 


X 


X 







978 








O 





o 


X 


336 





X 


X 


X 


X 


X 


X 


X 


o 




1007 


o 


X 








X 


O 


374 


o 








o 


o 











X 




1071 


o 


o 





X 


o 





590 





X 


X 


X 


X 


X 


X 


X 


X 




1091 








X 








o o 


624 











X 











o 


o 


















663 





o 











o 








o 


















695 


o 


X 


o 











o 


X 





















754 


X 














o 


X 





o 


















1144 











o 





o 


o 
























1190 





o 








o 






























1257 





o 


o 

















o 


















1409 














X 


X 


o 
























1560 


o 


X 








o 











o 


















1577 











o 








o 
























1587 











o 








o 


o 





















1601 


X 


X 





X 


X 


o 


X 


X 


X 


















1696 











o 








o 
























1722 


o 








o 

































1779 





X 














X 


X 


X 


















1795 


X 


X 


X 


X 








X 


X 


X 


















1912 


X 


X 


X 


X 


X 


X 


X 


X 


X 


















2091 





o 




















o 


















2122 




















o 
























2188 











o 








o 
























23742 


(X) 


(X) 


(X) 


(Xi 


IX) 


(X) 


(X) 


(X) 


iX) 


















2 The site at 2374 


is in 


the primer region. 


























ND3/ND4 
Hirdl 

sites 
















haplotypes 


















A 


B 


C 


D 


E 


F 


Cx 


H 


I 


N 





p 


Q 


a 


b 


g 




43 








O 

















o 








X 
















130 





X 


O 








X 








X 


o 


X 


X 


o 


O 










183 




















o 


O 





X 


O 


o 








O 







389 





X 


X 








o 





X 


o 











o 













472 




















o 


























X 




494 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 




621 








O 








o 











o 











X 










853 


o 

















o 


X 








O 








O 










999 





o 

















O 


o 








X 











o 




1017 








O 








o 











o 











O 


X 







1342 




















o 











O 








O 


X 







1448 

















o 





X 


o 




















o 




1537 











o 








X 








o 


X 








X 










1755 





o 





o 


X 


X 


o 




















o 


o 







1888 


o 








X 











X 


o 

























2011 





o 














o 








o 








X 





o 







2232 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


continued 



194 



Fishery Bulletin 104(2) 



Appendix 1 (continued) 


ND3/ND4 






















haplotype 


s 




















Mho I 


























































































sites 


A 


B 


c 


D 


E 


F 


G 


H 


I 


J 


K 


M 


N 





p 


R 


V 


a 


b 


c 


d 


e 


150 


O 


o 


o 








o 




















o 


o 


o 


o 




















190 




















o 


o 


o 











o 





o 








o 


o 





o 





198 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 








X 


X 


X 


X 


X 


X 


X 


X 


X 


261 




















o 




















o 


o 


o 














o 





272 



































X 


o 


o 


o 





o 











o 





302 





o 








o 








o 


o 











X 











o 


X 














632 

















o 


o 




















o 


o 


o 

















X 


648 











X 














o 











o 





























758 


X 


X 


X 


X 


X 


X 


X 


o 


X 


X 


X 


X 





o 


o 


o 





X 


X 


X 


X 


X 


797 























o 


o 








X 


o 























o 





861 


o 

















X 


X 


o 


o 




















o 








X 


X 





886 





X 






































o 


o 


o 


o 














900 





o 


o 











o 


o 


o 


X 
































o 





940 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


o 


o 


X 


X 


X 


X 


X 


X 


X 


971 


o 


X 











X 





o 


o 


o 











o 


o 























1029 


























X 

















o 


o 


o 


o 














1270 






































o 


o 


o 


o 


o 


o 


X 











1481 


o 


o 


o 














o 


o 


o 








o 


o 

















X 


X 





1534 














X 


X 


X 


X 


X 





X 


X 


X 





X 





X 


X 


X 


X 


X 





1647 


o 


o 


X 


X 





X 





o 


X 


X 


X 





X 


X 


X 


X 


X 





X 


X 


X 


X 


1658 












































o 


o 


o 


o 








o 





1748 





o 


o 











o 


o 


o 











o 





o 























1979 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 





X 


X 


X 


X 


X 


X 


X 





X 


X 


2016 


o 

















o 


o 


o 











X 


o 


o 























2339 




































































ND3/ND4 










haplotypes 


(continued) 






























Mbol 






































































sites 


f 


g 


h 


i 


j 


k 


1 


m 


n 


p 


q 


r 






















150 





o 


X 


o 


o 











































190 














X 











































198 


X 


X 


o 


X 


X 


X 


X 


o 


X 


X 


X 


X 






















261 


o 


o 


o 


o 


o 


o 








X 































272 


























































302 





o 


o 


o 


o 


o 











o 




























632 


o 


o 


o 


o 














































648 


























































758 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 






















797 


























































861 





o 


o 


o 


o 


o 


o 


o 


o 


o 


o 

























886 


























































900 











o 


o 


o 


o 


o 


o 































940 


X 


X 


X 


X 


X 


X 





X 


X 








X 






















971 








o 


o 


o 


o 





o 


o 








X 






















1029 



































o 






















1270 











X 





X 











o 




























1481 





o 





o 

















o 


o 


o 






















1534 





X 


X 


X 








o 


X 


X 


o 


X 


X 






















1647 


X 


X 


X 


o 


X 


X 


X 


X 





X 


X 


X 






















1658 


X 














o 


o 


o 


o 


o 




























1748 


o 





o 


o 


o 





X 








o 


o 


o 






















1979 


X 





X 


X 


X 


X 


X 


X 


X 


X 


X 


X 






















2016 


























o 








o 






















2339 


o 


o 


o 


o 























X 
































































continued 



Li et al A key to selected Sebastes spp based on mitochondrial DNA restriction fragment analysis 



195 



Appendix 1 (continued) 


ND3/ND4 












haplotypes 




















Msp I 






































sites 


A 


B 


c 


D 


E 


F 


G 


H 


L 


M 


a 


b 


d 


e 


g 








30 


X 


X 


X 


X 


X 


X 


X 


X 


X 





X 


X 


X 


X 


X 








933 

















X 


O 











o 


o 








o 








1199 














o 


X 


O 











o 




















1230 


O 





o 























o 





X 














1261 








o 


O 


o 


o 


X 























o 








1541 








o 


o 











X 











X 








X 








1624 











o 











o 


X 





o 








X 











1738 


X 


X 


X 


X 


X 


X 


X 





X 


X 


X 


X 


X 


X 


X 








1826 


X 


X 





X 


X 





O 


X 


X 


X 





X 


X 


X 


X 








1844 











X 


X 


o 


O 

















X 


o 


o 








2073 


O 





o 








X 


o 
































2110 





X 


X 





X 





o 


o 





o 








X 


X 


X 


























laplotypes 


















Rsa 1 






































sites 


A 


B 


c 


D 


E 


F 


G 


H 


I 


J 


L 


N 





p 


Q 


R 


a 


b 


346 


O 


o 





X 


X 





X 


X 


o 


o 


X 


X 





X 














561 














o 





o 














o 


o 








X 





o 


742 


O 





o 











o 


X 


X 


o 

















o 








1077 














o 


X 


o 











X 


X 

















o 


1231 


O 





o 











X 


o 


o 




















o 








1321 














o 


o 




















X 














o 


1339 





























X 

















o 








1357 








o 





o 


o 
































X 


o 


1433 

















o 


o 


o 




















o 





X 





1492 











o 


o 


o 





X 














o 











o 


o 


1863 





o 


X 





X 





o 


X 


X 


X 


X 





o 


o 


o 








X 


1985 





X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 





X 


X 


X 





2063 





o 




















o 














o 


X 











ND3/ND4 


















haplotypes 


















Sty I 






































sites 


A 


B 


c 


D 


E 


F 


G 


H 


I 


L 


M 


N 


o 


b 


d 


e 


g 




49 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 




108 








o 


o 














o 


o 


X 








X 


o 


o 


X 




194 





o 





X 


X 


X 





o 


X 








X 


o 


X 













365 


X 


X 


X 


X 


o 





X 


X 


X 


X 


X 


X 





X 





X 


X 




391 





o 


o 





X 


X 








o 


X 








o 


o 













534 


X 





X 


X 


X 


X 








X 


X 


X 


X 


X 


X 


o 


X 


X 




1258 








o 


o 





X 








o 


o 











o 


o 










1570 





X 


X 


X 


X 


X 











X 


X 


o 


X 





X 


X 


o 




1725 





o 


o 











X 





o 








X 


o 


o 


o 


X 







2311 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 


X 




12S/16S 






haplotypes 




























BstN I 




















































sites 


A 


B 


C 


D 


E 


F 


G 
























326 


X 


X 


X 


X 





X 


X 
























988 








O 


X 





X 



























1105 














o 


X 



























1530 


o 


o 














X 
























1687 





o 


X 








o 



























1741 


o 


X 







































2416 


X 


X 


X 


X 


X 


X 


X 

























196 



Fishery Bulletin 104(2) 

















Appendix 


1 (continued) 














12S/16S 






haplotypes 
















haplotypes 








BstV I 




















C/bl 


































sites 


A 


B 


C 


D 


E 


F 








sites 


A 


B 


c 


D 






87 


X 


X 


X 


X 


X 


X 








537 


X 


X 





X 






600 


X 


o 


X 


X 





X 








602 


X 


O 


X 


X 






664 


X 


X 


X 


X 


X 


X 








665 


X 


X 


X 


X 






729 


X 


X 





o 


X 


X 








1630 











X 






1633 





o 





o 


X 


o 








1898 


X 


X 


X 


X 






1872 


X 


X 


X 


X 


X 


X 








2268 


X 


X 


X 


X 






2128 


X 


X 


X 


X 


X 


X 






















2191 











X 





X 






























haplotype 


s 
















haplotypes 




Ddel 




















Hmn 












































sites 


A 


B 


C 


D 


E 


F 


G 


H 


I 


sites 


A 


B 


c 


D 


E 


G H 1 


44 


X 


X 


X 


X 


X 


X 


X 


X 


X 


982 








X 








X 


55 


X 


X 


X 


X 


X 


X 


X 


X 


X 


1181 

















X 


976 


X 


X 


X 


X 


X 











X 


1291 


X 


X 


X 





o 


XXX 


1043 


X 


X 


O 


X 


X 


o 








X 


1700 

















X X 


1056 


X 


X 





X 


o 





X 





X 


2094 





X 


X 





X 


O X X 


1304 


X 


X 


X 


X 


X 





X 


X 


X 
















1339 


























X 
















1735 





X 


X 


X 


X 





X 


X 


X 
















2181 


X 





X 


X 


X 


o 





X 


X 
















2393 


X 


X 


X 


X 


X 


X 


X 


X 


X 


















haplotypes 






























Hind II 


































sites 


A 


B 






























1717 


X 

































12S/16S 








haplotypes 
















haplotypes 






Mho I 




















Mspl 










































sites 


A 


B 


c 


E 


F 


G 


H 


I 




sites 


A 


B 


C 


D 


E 


F 


201 


O 


X 


X 








X 


X 


X 




65 


X 


X 


X 


X 


X 


X 


849 


X 


X 


X 





X 


X 


X 


X 




241 

















X 


1015 


X 


X 


X 


X 








X 


X 




330 


o 


O 


X 


X 


O 





1210 


O 


o 





O 





o 


o 


X 




766 


X 


X 


X 


X 


X 


X 


1403 


X 


X 


X 


X 


X 


X 


X 


X 




1259 





X 


X 








X 


1507 


X 


o 


X 


X 


X 


X 


X 


X 




1390 


X 


X 


X 


X 


X 


X 


1637 











O 


o 





X 


o 




1535 


X 


X 


X 


X 


o 


X 


1984 


X 


X 


X 


X 


X 


X 


X 


X 




2226 


X 


X 


X 


X 


X 


X 


2059 


X 


X 


X 


X 


X 


X 


X 


X 




2403 


X 


X 


X 


X 


X 


X 


2228 


X 


X 


X 


X 


X 


X 


X 


X 


















2318 


X 


X 


X 


X 


X 


X 


X 


X 


















2388 


X 


X 


X 


X 


X 


X 


X 


X 


















Rsal 




haplotypes 












Sty I 




haplotypes 






































sites 


A 


B 


c 


D 


E 










sites 


A 


B 


c 


D 






293 











O 


X 










295 


X 


O 


X 


X 






507 


O 





X 


X 


o 










982 











X 






588 


O 





o 





X 










1748 








X 









761 


X 


X 


X 


X 


X 










2053 





X 












950 








o 





X 










2254 


X 


X 


X 


X 






1000 


X 


X 


X 


X 



























1071 


X 


X 


X 


X 



























1263 


o 


O 








X 
























1308 





X 


o 


X 


X 
























1358 


X 


X 


X 


X 


X 
























2164 


o 





o 


o 


X 

























197 



Abstract — Fishery catch data on yel- 
lowfin tuna {Thunnus albacares) were 
examined to study the effects of El 
Nino events between 1990 and 1999 
for an area in the northeastern tropi- 
cal Pacific (18-24°N, 112-104'W). The 
data were extracted from a database 
of logbook records from the Mexican 
tuna purse-seine fleet. Latitudinal 
distribution of the catches increased 
from south to north for the 10-year 
period. Highest catches and effort 
were concentrated between 22°N 
and 23°N. This area accumulated 
48% of the total catch over the 10- 
year period. It was strongly correlated 
with El Nino-Southern Oscillation 
(ENSO) events. At least two periods 
of exceptionally high catches occurred 
following El Niiio events in 1991 and 
1997. Peaks of catches were triggered 
by the arrival of positive anomalies of 
sea surface temperature (SST) to the 
area. A delay of two to four months 
was observed between the occur- 
rence of maximum SST anomalies 
at the equator and peaks of catch. 
Prior to these two events, negative 
SST anomalies were the dominant 
feature in the study area and catch 
was extremely low. This trend of nega- 
tive SST anomalies with low catches 
followed by positive SST anomalies 
and high catches may be attributed 
to northward yellowfin tuna migra- 
tion patterns driven by El Nifio forc- 
ing, a result that contrasts with the 
known behavior of decreasing relative 
abundance of these tuna after El Nino 
events in the eastern Pacific. How- 
ever, this decrease in relative abun- 
dance may be the result of a local or 
subregional effect. 



Variation in yellowfin tuna iThunnus albacares) 
catches related to El Niiio-Southern Oscillation 
events at the entrance to the Gulf of California 



Ernesto Torres-Orozco 

Universidad de Colima, Facultad de Ciencias Marinas 
Km 20, Carretera Manzanillo-Barra de Navidad 
C P 28860 
Manzanillo, Colima, Mexico 



Arturo Muhlia-Melo 

Centre de Investigaciones Biologicas del Noroeste, S.C. 

Mar Bermeio No. 195 

Col Playa Palo de Santa Rita 

Apdo, Postal 128 

La Paz, BCS 23090, Mexico 

Email address (for A Mulilia Melo, contact author) amuhlia04(a'cibnormx 

Armando Trasvina 

Oceanogralia Fisica, CICESE en BCS 
Miraflores 334 e/Mulege y La Paz 
Fracc. Bella Vista 
La Paz, BCS 23050, Mexico 

Sofia Ortega-Garcia 

Centre Interdiscipiinarie de Ciencias Marinas 
La Paz, BCS, Mexico, 23000, Becano COFAA 



Manuscript submitted 3 December 2003 
to the Scientific Editor's Office. 

Manuscript approved for publication 
9 August 2005 by the Scientific Editor. 
Fish. Bull. 104:197-203 (2006). 



The entrance to the Gulf of Califor- 
nia (from 18°N to 24°N and 104°W to 
112°W) is located in the convergence 
zone of the North Pacific Gyre, where 
the California Current separates from 
the coast to feed the North Equato- 
rial Current. It has a complex hydro- 
graphic structure due to the confluence 
of different water masses (Roden and 
Groves, 1959; Roden, 1972; Alva- 
rez-Borrego and Schwartzlose, 1979; 
Bray and Robles, 1991; Torres-Orozco, 
1993 ). This region is highly responsive 
to the El Niiio phenomenon. Its main 
response is characterized by positive 
sea level anomalies, warming of the 
upper layer, and a general alteration 
of water current patterns (Baumgart- 
ner and Christensen, 1985; Robles and 
Marinone, 1987; Torres-Orozco, 1993; 
Lavin et al., 1997; Ortega-Garcia et 
al, 1999; Trasvifia et al, 1999; Castro 
et al., 2000; Bernal et al., 2001). 

High abundance of yellowfin tuna 
(Thunnus albacares; YFT) is reported 
in the study area (Allen and Punsly, 
1984; Castro-Ortiz and Quiiiones- 



Velasquez, 1987; Muhlia-Melo, 1993; 
Ortega-Garcia, 1998). However, stud- 
ies about the YFT interaction with the 
physical environment of the Mexican 
Pacific are lacking. Blackburn (1965, 
1969) considers that the abundance of 
the YFT correlates with sea surface 
temperature (SST) in the range of 
20°C to 30°C, but it can also be pres- 
ent in regions having temperatures 
between 18°C and 31°C. Blackburn 
(1969) considered 30°C as an optimal 
estimate of the maximum tempera- 
ture of occurrence of YFT. Similarly, 
Ortega-Garcia (1998) reported that 
YFT are distributed in regions where 
SST ranges from 17°C to 31°C and 
that they are frequently observed in 
waters close to 28°C. Other studies 
such as that of Laevastu and Rosa 
(1963) and Castro-Ortiz and Quirio- 
nes-Velasquez (1987) have indicated 
that YFT are found at concentrations 
favorable to commercial fisheries in 
regions where the SST ranges from 
20°C to 28°C. In summary, YFT 
catches are historically reported to 



198 



Fishery Bulletin 104(2) 



occur at SST values ranging from 12^ to 30°C, and 
favorable catches are reported between 17° and 25°C 
(Lehodey^). More recently, Bautista-Cortes (1997) has 
studied the relationships between varying vertical tem- 
perature structure and YFT catches in the Mexican 
Pacific to 140° longitude W, finding that the depth of the 
23°C isotherm is related to catch and that the shallower 
the isotherm, the larger the tuna catches. 

Castro-Ortiz and Quinones-Velasquez (1987) reported 
reduced catches of YFT during the 1982-83 El Niiio 
events in the northeastern Pacific, in contrast to the 
following 1984-85 seasons. The 1982-83 El Nmo event 
was reported to have caused the largest decrease in the 
availability-vulnerability index (AVI=CPUE/{1-SSR); 
where SSR is the successful-set ratio and CPUE is 
catch per unit of effort; IATTC-). This finding is con- 
sistent with the resource being less susceptible to fish- 
ing in the northeastern Pacific during an intense El 
Nino. Lehodey^ analyzed data of purse-seine and long- 
line CPUE of skipjack {Katsuwonus pelamis), yellowfin 
(Thunnus albcares), bigeye (Thuimus obesus), and alba- 
core (Thunnus alalunga) tunas during El Nino and La 
Nina events, from 1982 to 1998 for the whole tropical 
Pacific Ocean (20°S to 20°N). He concluded and sup- 
ports the hypothesis that ENSO affects the recruitment 
of skipjack, yellowfin, and bigeye tunas positively in 
the western Pacific. A direct positive effect related to 
the vertical change in the thermal structure during El 
Nino increased purse-seine catches of yellowfin in the 
western Pacific. He suggested two theoretical cases 
be considered for yellowfin and bigeye in the eastern 
Pacific: either a positive La Niiia effect on the recruit- 
ment and catchability, or conversely a positive El Nifio 
effect on recruitment and catchability. His statistical 
analysis did not identify the direct effects of ENSO 
on catchability for yellowfin and bigeye tunas. He sug- 
gested the need for additional analyses in the eastern 
Pacific region and for comparisons with independent 
results from modeling and observations to confirm the 
preliminary results. Lu et al. (2001) analyzed catches 
of YFT and bigeye tuna caught by the longline fleet of 
the tropical Pacific Ocean. They concluded that high 
hook rates for both species were mostly associated with 
regions where SST increased during El Niiio or La Niiia 
years. During La Niiia episodes, YFT populations ap- 
pear to undergo a poleward displacement thus shifting 
both north and south of the equator. During El Nifio 
events YFT populations in the equatorial Pacific are 
found where SST anomalies are higher, in the central 
and eastern equatorial Pacific. 



For the region of interest, Ortega-Garcia et al. (1999) 
studied the impact of the ENSO during the 1997-98 
El Nino events in comparison to the 1996-97 non El 
Niiio year within the eastern Pacific Mexican purse- 
seine tuna fishery. They reported that during the first 
months of the ENSO (July-December 1997) of the El 
Niiio event, the extent of the YFT catches was higher 
in oceanic waters than the observed catches during a 
non El Niiio year in these areas for the same period. A 
decline of effort was observed along the coast of Baja 
California and inside the Gulf of California. In contrast, 
during the second part of the ENSO El Niiio 1997-98 
(January- June 1998), an increase of effort was ob- 
served along the coasts of Baja California and inside 
the Gulf of California; however, a decrease of effort was 
observed in oceanic areas. 

The relative abundance of YFT in the eastern Pa- 
cific is reported to diminish during El Niiio events; 
however, as supported by the IATTC° report, the YFT 
vertical displacement during El Niiio events generates 
a diminishing fishing effort for the purse-seine fleet in 
traditional catch areas. As a consequence, this has led 
to a good recruitment and greater yield-per-recruitment. 
Large recruitments after El Niiio events are reported 
to occur in the eastern Paciflc. Joseph and Miller^ ana- 
lyzed YFT catch data from the eastern Pacific for a 
22-year period and found positive anomalies in tuna 
recruitment after El Niiio events. Large recruitments 
were observed in years 1971, 1974, 1978 and 1985, all 
of which were preceded by the El Niiio events of 1969, 
1972, 1976, and 1983, respectively. 

The objective of the present study was to examine 
the effect of the ENSO El Nifio and La Niiia events for 
a 10-year period (1990-99) on YFT catches in the en- 
trance to the Gulf of California, an area of importance 
for the industry and for the ecology of the resource in 
the eastern Pacific. 



Materials and methods 

The classification of El Niiio intensities was taken from 
the Climate Prediction Center (NOAA'*). The process 
of classification takes into account re-analyzed SSTs 
produced at the National Centers for Environmental 
Prediction (NCEP), at the Climate Prediction Center, 
and at the United Kingdom Meteorological Office. 

Environmental data in the form of monthly means of 
sea surface temperature anomalies (SSTAs) from 1990 to 
1999 were extracted from the NCEP monthly SSTA data- 



' Lehodey, P. 2000. Impacts of the El Nino Southern Oscilla- 
tion on tuna populations and fisheries in the tropical Pacific 
Ocean. Working Paper. RG-1, 32 p. Oceanic Fisheries 
Programme, Noumea, New Caledonia, Secretariat of the 
Pacific Community. http;//www.spc.org.nc/OceanFish/Html/ 
SCTB/SCTB13/rgl.pdf [accessed on 20 February 2002]. 

- lATTC ( Inter American Tropical Tuna Commission). 1989. 
Annual Report of the Inter-American Tropical Tuna Com- 
mission, 1988. Annu. Rep. lATTC, 288 p. httpV/www.iattc. 
org/PublicationsSPN.htm [accessed on 11 March 2002]. 



3 Joseph, J., and F. R. IVIiller. 1988. El Nino and the surface 
fishery for tunas in the eastern Pacific. In Proceedings 
of tuna fish. res. conf., p. 199-207. Japan Fish. Agency- 
Far Seas Fish. Res. Lab. IVIuguro Gyogyo Kyogikai Gjiroku, 
Suisancho-Enyo Suisan Kenkyusho. 

•• NOAA (National Oceanic and Atmospheric Administration). 
2001. Cold and warm episodes by season. Website: http:// 
www.cpc.ncep.noaa gov/products/analysis_monitoring/enso- 
stuff/ensoyears.html [accessed on 10 January 2002). 



Torres-Orozco et al Variation in catches of Thunnus albacores related to El Niho-Southern Oscillation events 



199 



base (IRI'). The SST fields were blended from 
ship, buoy, and bias-corrected satellite data. A 
description of this procedure can be found in 
Reynolds and Smith (1994). SSTA data for El 
Niiio region 3 (NIN03, from 150°W to 90°W 
and from 5°N to 5°S) were obtained from Cli- 
mate Prediction Center databases (CPC'). 

Yellowfin tuna catch data were obtained 
from the database of the ATUN (tuna) Project 
of CICIMAR (the Interdisciplinary Research 
Center of Marine Science of the National 
Polytechnic Institute in La Paz, Mexico). This 
database has information on daily fishing ac- 
tivities for about 80% of the Mexican purse- 
seine fleet operating in the eastern Pacific. 
The data used in the present study include 
carrying capacity, catch-by-species, location, 
and school type (YFT associated with dol- 
phins; YFT associated with floating objects; 
and free-swimming YFT), from 1990 to 1999, 
comprising 11,690 records. Total distribution 
of the sets made by the Mexican purse-seine 
fleet from 1990 to 1999 at the entrance of the 
Gulf of California is shown in Figure 1. 

Interannual variation 

In order to study annual variation, total 
catches of YFT of the purse-seine fleet within 
the study area were accumulated by year. 
A cross-correlation analysis was applied to 
obtain information on lags between SST 
anomalies and YFT catches. 

Latitudinal stratification 

In order to analyze the latitudinal catch 
variation of YFT within the study area, 
catch data for the period 1990-99 were 
accumulated in six latitudinal bands of one 
degree from 18°N to 24°N at the entrance of 
the Gulf of California. Longitudinal limits 
extended from the coastal line of the Mexi- 
can continent (from 104°W to 112°W). 



24' N I 



23' N 



22"N 



2r'N 



20"N 



19' N 




t. . . - . .......Csivv. J.'.: ■• i 




114"W 



108°W 



•■:i;;ii^iC-r:^ 



\rd--^ ■'■'■-••:••', .•■•'!••: ' .Itf- v-fftlH 



i 



112 W now 108 W 106 W 104°W 

Figure 1 

Total distribution of the sets made by the Mexican purse-seine fleet 
during 1990 to 1999 at the entrance to the Gulf of California. Catch 
data for yellowfin tuna {Thunnus albacares) were obtained from the 
CICIMAR (the Interdisciplinary Research Center of Marine Science 
of the National Polytechnic Institute in La Paz, Mexico). 



« 20 




1994 1995 

Year 



1996 



Total 
1990- 



Figure 2 

catch for yellowfin tuna [Thunnus albacares) by year during 
-99 in the study area. 



Results 

Interannual variation in YFT catch 

The catch of YFT in the study area showed high inter- 
annual variation (Fig. 2). This high variation showed 



■■' IRI (International Research Institute for climate prediction). 
2002. Sea surface temperature anomaly data. Website: 
http://ingrid.ldgo.columbia.edU/SOURCES/.NOAA/.NCEP/. 
EMC/.CMB/.GLOBAL/.Reyn_SmithOIvl/.monthly/.ssta/ 
laccessed on 9 February 2002]. 

" CPC (Climate Prediction Center). 2002. Monthly atmo- 
spheric and SST indices. Website: http://www.cpc.ncep.noaa. 
gov/data/indices/index.html [accessed on 9 February 2002J. 



substantial increments, 85% from 1991 to 1992, 54% 
from 1992 to 1993, 102% from 1995 to 1996, and 102% 
from 1997 to 1998. In particular, the sum of the two El 
Niiio years in the record (1991-93, 1997-98) represents 
59.5% of the total catches over the 10-year period. 

Latitudinal variation in YFT catches 

Figure 3 shows the variation of catch and effort (in sets) 
from 1990 to 1999. Vertical bars include catches from dif- 
ferent one-degree latitudinal range areas between 18°N 
and 24°N and meridional ranges between 104°W to 112°W 
(Fig. IJ. With the exception of the southern 18-20°N area, 



200 



Fishery Bulletin 104(2) 



there is a general linear increase of catch with latitude. 
Fishing effort was concentrated in the 22-23°N and 
23-24°N areas. Both account for 48% of the total capture 
for the 10-year period in this region. From 18° to 21°N, 
catch values averaged 12.9% of the total catch. 



Yellowfin tuna catches affected by ENSO events 

Sea surface temperature anomalies (SSTA) were used to 
investigate the effect of interannual warming or cooling 
events on the variability of YFT catch. Anomalies for 



D Sets 
 Catch 



? 20- 

mm 



18-19 19-20 20-21 21-22 22-23 23-24 

Latitude ('N) 

Figure 3 

Latitudinal distribution of catches of yellowfin tuna ^Thunnun albacares] 
for the period 1990-99. 




p 
w 

CO 



2 


-2 

4 

2 



-2 

2000 

c 

2 1000 



-1000 



SW 



MW Cold 



SW 



Cold 



^4 ^. 



B 


I ,si 


1 
1 
1 

L 






1 
1 
1 
1 1 1 1 



1 I i " 

I ^1 



1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 



Figure 4 

Sea surface temperature anomalies (SSTAs) in the El Nino region 3 (A), 
SSTAs in the study area (B) and catches of yellowfin tuna tThunnus alba- 
cares) (C). Bars at the top show the equatorial central Pacific episodes as 
compiled by the Climate Prediction Center (see "Methods" section): Strong 
warm (SW) peaks of catch are indicated by thin vertical lines. Thick verti- 
cal lines indicate maximum SST anomalies in NIN03 (A). Dashed vertical 
lines indicate peaks of SSTA in the area of study (B). Moderate warm (MW) 
and cold episode. The maximum intensity of the warm episodes is marked 
with dashed lines (B and C). 



Torres-Orozco et al : Variation in catches of Thunnus albacares related to El Nino-Southern Oscillation events 



201 



the entrance to the Gulf of California were calculated 
between 22°N and 23°N and 103°W to 112°W. These 
anomalies were extracted from the Reynolds and Smith 
(1994) monthly 1-degree SSTA climatology. This variable 
was then compared to the catch of YFT. We did not exam- 
ine SSTAs south of 21'N because they fall within the 
Mexican Warm Pool (Trasvina et al., 1999), where persis- 
tently high SSTs (above 27°C all year in the warm pool 
region) mask the propagation of warm and cold signals 
from the Equator. The 10-year time series (1990 to 1999) 
was "detrended" and filtered to eliminate periods shorter 
than three months, as described in Godin (1972). 

SSTAs in the NIN03 (Fig. 4A) showed cold and warm 
conditions associated with five identifiable events. These 
events were the following: 1) warm El Nifio events in 
1991-92 (strong warm, SW), 2) in 1994-95 (moderate 
warm, MW), and 3) in 1997-98 (strong warm, SW); 
4) moderately cold La Nifia events that took place in 
1995-96 and 5) in 1998-99. Strong warm events are 
marked with a black line located at each respective 
maximum SSTA (Fig. 4). The 1997-98 was the stron- 
gest warm episode in the 1990-99 decade. 

The 1991-92 and 1997-98 SSTAs in El Nino region 3 
were higher, by more than 1°C, than the SSTAs in the 
study area (Fig. 4, A and B). Within the 10-year period, 
near-normal conditions occurred from 1993 to 1996. The 
coldest and most persistent event of this decade took 
place in the study area from July 1998 to December 
1999, i.e., during a moderate La Nina event. 

Significant interannual variability of YFT catches 
was observed during the 10-year period in the study 
area (Fig. 4C). This graph shows the time series of 
monthly values of YFT catch anomalies related to the 
average for the 10-year period. Five exceptionally high 
catch events in years 1992, 1993, 1996, 1998, and 1999 
were observed. These years account for 70'7f of the total 
capture during the 10-year period. 

Several qualitative and quantitative features were 
apparent from these results: 

1 Catch peaks in 1992 and 1998 occurred 3 (r=0.73) 
and 4 (r=0.64) months, respectively, after the onset of 
an El Niiio event at the equator (time span between 
thin and thick lines) (Fig. 4, A and C). 

2 High catches of YFT at the entrance of the Gulf of 
California in 1992 and 1998 occurred 2 (/—0.71) and 3 
(r=0.73) months, respectively, after the SST anomaly 
signal reached the study area (time span between 
dashed and thin lines) (Fig. 4, B and C). 

3 Higher catches of YFT occurred during the spring 
following an El Nifio winter. These were observed in 
the spring of 1992 and 1998 (thin lines; Fig. 4C). 

4 Peaks of YFT catch were also observed in 1993 and 
1999. These occurred one year after the El Niiio 
event. The 1993 peak took place in nearly normal 
SST conditions, whereas the 1999 peak occurred 
during negative SST anomalies (La Nifia event. Fig. 
4, B and C). These peaks of YFT catch were higher 
than the ones recorded in the previous year during 
the El Nino event 1991-92 and during 1997-98. 




1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 
Year 

Figure 5 

Annual catches (in tons) of yellowfin tuna (Thunnus 
albacares) in latitudinal bands within the area of study 
for the period 1990-99. 



5 The 1996 peak seemed not to be related to the vari- 
ability of SST anomalies alone. 

6 There was a delay from one to two months between 
the mature phase of the El Niiio event at the equator 
and the presence of high SSTAs in the study area 
(time span between thick and dashed lines in Fig. 
4, A and B). 

Annual YFT catches of stratified latitudinal bands 
within the study area are shown in Figure 5. A progres- 
sive increase in YFT catches from southern to northern 
latitudes was observed in years 1993 and 1999. This 
progressive increase may be indicative of a northward 
movement of the resource. This phenomenon has been 
observed in the western tropical Pacific Ocean during 
La Nifia events where YFT have dominated the longline 
catch (Lu et al., 2001). 



Discussion 

Catch affected by the onset El Nino 

Cross-correlation between the onset of El Nifio at the 
equator in 1991-92 and 1997-98 with catch peaks of 
YFT in the study area showed a delay of three and 
four months respectively. Catch peaks of YFT at the 
entrance of the Gulf of California in 1992 and 1998 
occurred two to three months after the SSTA signal 
reached the study area. These results are similar to 
those found by Lehodey' in the western and central 
regions of the Pacific Ocean, where rising (deepen- 
ing) of the mixed-layer depth related to El Nifio (La 
Nifia) was associated with an increase in the pole-and- 
line and purse-seine CPUE of YFT and where there 



202 



Fishery Bulletin 104(2) 



was a concomitant delay of two to three months in 
this increase. 

Catch of YFT in relation to recruitment 

Recruitment explains most of the YFT catch fluctuations 
in relation to El Nino events at the entrance of the Gulf of 
California. A recurrent pattern in the time-series of the 
catch revealed a peak of catch fourteen ( 1993 1 to twelve 
(1999) months following an El Nino event. A similar 
result has been observed in a cross-correlation between 
the Southern Oscillation index (SOI) and the long-line 
series of catch data for yellowfin tuna; a positive correla- 
tion was found in the eastern Pacific region after a lag 
of fourteen months (LehodeyM. This finding supports the 
hypothesis of a recruitment base (age-effect) for yellowfin 
because at 14 months YFT reach 75 cm, corresponding to 
the first age class recruited to the long-line fishery. These 
results are also consistent with those from previous 
analyses of time-series of catches of YFT in the eastern 
Pacific, namely by Joseph and Miller^ and the IATTC-. 
Therefore, recruitment can be the major explanation of 
the YFT catch fluctuations in relation to El Niiio events 
at the entrance of the Gulf of California. 



been observed in the central Western Pacific for the 
long-line fishery (Lu et al., 2001). 

Forecasting catch within the study area may give us 
the ability to reliably develop predictions of catch fol- 
lowing El Nifio or La Niiia events. 



Acknowledgments 

The first author was a CONACYT grant holder for the 
duration of his doctoral studies at CIBNOR (Reg. 60997). 
The Universidad de Colima and PROMEP also supported 
our research effort. The research project was supported 
by a CONACYT grant (990107004). We are grateful to 
the ATUN project (CICIMAR-IPN) for providing the 
capture data. The third author is a SNI grant holder 
and wishes to acknowledge the support of the Oceanol- 
ogy Division of CICESE and of CICESE's campus at Baja 
California Sur. We are grateful for the critical analysis 
and suggestions from two anonymous reviewers and the 
scientific editor. 



Literature cited 



Catch distribution of YFT related to ENSO episodes 

For the long-line fishery in the tropical Pacific Ocean, 
areas with significant higher hook rates of YFT during 
El Nino years are located east of 150'"W within tropical 
waters of the central eastern Pacific (Lu et al., 2001). 
Conversely, higher hook rates occur during La Nina epi- 
sodes in areas where SSTs rise during El Nifio events. 
These two findings may provide two possible reasons 
for the change in hook rates; the expansion of optimal 
habitat and the change in vulnerability of the resource 
to the fishing gear during the ENSO episodes. 



Conclusions 

At the entrance of the Gulf of California, an El Nino 
event is associated with an increase in the purse-seine 
catch of YFT after two to three months, when SSTAs 
reach the study area. Similarly, this increase is delayed 
three to four months after the onset of El Niiio at the 
equator. 

A positive correlation of El Nifio on YFT catch at 
the entrance of the Gulf of California after a twelve 
to fourteen month delay supports the hypothesis that 
the ENSO affects recruitment of YFT. This correlation 
seems to be independent of the thermal structure dur- 
ing the recruitment phase. In 1993, normal conditions 
in SST were present, whileas in 1999 La Niiia was 
observed. 

A northward displacement of YFT seems to occur at 
the entrance of the Gulf of California twelve to fourteen 
months after El Nino events. Lower catches in southern 
latitudes were recorded, whereas higher catches were 
recorded in northern latitudes. Similar results have 



Alvarez-Borrego, S., and R. Schwartzlose. 

1979. Water masses of the Gulf of California. Ciencias 
Marinas 6:43-63. 
Allen, R., and R. Punsly. 

1984. Catch rates as indices of abundance of yellowfin 
tuna, Thunnus albacares, in the Eastern Pacific Ocean. 
Bull. Inter-Amer. Trop. Tuna Comm. 18(41:303-79. 

Baumgartner, T. R., and N. Christensen Jr. 

1985. Coupling of the Gulf of California to large- 
scale interannual climatic variability. J. Mar. Res. 
43:825-848. 

Bautista-Cortes, L. F. 

1997. Analysis of the Mexican tuna fishery with emphasis 
on three tuna school types. M.S. thesis, 69 p. CIC- 
IMAR National Polytechnic Institute, CICIMAR, La 
Paz, B.C.S., Mexico. 
Bernal, G.. P. Ripa, and J. C. Huerguera. 

2001. Oceanographic and climatic variability of the lower 
Gulf of California: influences from the tropic and the 
North-Pacific. Ciencias Marinas 274:595-617. 
Blackburn, M. 

1965. Oceanography and the ecology of tunas. Oceanogr. 

Mar. Biol. Ann. Rev. 3:299-322. 
1969. Conditions related to upwellings which deter- 
mine distribution of tropical tunas off western Baja 
California. Fish. Bull. 68:147-176. 
Bray, N. A., and J. M. Robles. 

1991. Physical oceanography of the Gulf of California. 
Invited review paper, part V, chap. 25. In The Gulf 
and peninsular provinces of the Californias, AAPG 
Memoir 47 (J. P. Dauphin and B.R. Simoneit, eds.) p. 
511-553. Am. Assoc. Petrol Geol. Mem., Tulsa, OK. 
Castro, R., A. S. Mascarenas, R. Durazo, and C. A. Collins. 

2000. Seasonal variation of the temperature and salinity 
at the entrance to the Gulf of California, Mexico. Cien- 
cias Marinas 26(4):561-583. 
Castro-Ortiz, J. L., and C. Quinones-Vclasquez. 

1987. Geographic distribution and abundance of yellow- 



Torres Orozco et al : Variation in catches of Thunnus albacares related to El Niiio-Southern Oscillation events 



203 



fin {Thuninm albacares) related to sea surface tem- 
perature in the Northeast Pacific (1983-1985). In 
Symposium of investigations in biology and fisher- 
ies oceanography in Mexico iM. Ramirez, ed.), p. 
21-23. CICIMAR. La Paz, B.C.S., Mexico. 

Godin, G. 

1972. The analysis of tides, 264 p. Univ. Toronto Press. 
Toronto, Canada. 

Laevastu, T., and H. Rosa. 

1963. Distribution and relative abundance of tunas 
in relation to their environment. FAO Fish. Rep. 
6(3):1835-51. 

Lavin, M. F., E. Beier, and A. Badan. 

1997. Hydrographic structure and water circulation of the 
Gulf of California: seasonal and inter-annual scales. In 
Contribution to the physical oceanography in Mexico (F. 
Lavin, ed.), p. 141-171. Mexican Geophysical Organi- 
zation, Monograph No. 3. Mexican Geophysical Org., 
Ensenada, Baja California. 

Lu, H. J., K. T. Lee, H. L. Lin, and C. H. Liao. 

2001. Spatio-temporal distribution of yellowfin tuna 
Thunnus albacares and bigeye tuna Thunnus obesus 
in the Tropical Pacific Ocean in relation to large-scale 
temperature fluctuation during ENSO episodes. Fish. 
Sci. 67:1046-1052. 

Muhlia-Melo, A. F. 

1993. Overview of the yellowfin tuna Thunnus albacares 
(Bonaterre 1788). fishery, resource and management 
in Mexico. Ph.D. diss., 80 p. Facultad de Ciencias, 
National Univ. Mexico, Mexico D. F. 

Ortega-Garcia, S. 

1998. Analysis of the spatial and seasonal fluctuations 
of tuna abundance in the eastern Pacific Ocean. Ph.D. 



diss., 67 p. Facultad de Ciencias, National Univ. Mexico, 
Mexico D.F. 
Ortega-Garcia, S., D. Lluch-Belda, A. Zarate-Villafranco, 
R. Rodriguez-Sanchez, R. Beltran-Pineda, and S. Hernandez- 
Vazquez. 

1999. The ENSO 1997-1998 effect on the Mexican Tuna 
Fishery. In Proceedings of 50th Annu. Tuna Conf., 
Lake Arrowhead, California, May 24-27. 1999. Inter. 
Am. Trop. Tuna Comm., Lake Arrowhead, CA. 
Reynolds. R. W.. and T. M. Smith. 

1994. Improved global sea surface temperature analy- 
ses using Optimum Interpolation. J. Climate 7: 
929-948. 
Roden. G. I. 

1972. Thermohaline structure and baroclinic flows across 
the Gulf of California entrance and in the Revillagigedo 
Island region. J. Phys. Ocean. 2(2):177-183. 
Roden, G. I., and G. W. Groves. 

1959. Recent oceanographic investigations in the Gulf 
of California. J. Mar. Res., 18:10-35. 
Robles, J. M., and S. C. Marinone. 

1987. Seasonal and interannual thermohaline variability 
in the Guaymas basin of the Gulf of California. Cont. 
Shelf Res. 7(7):715-733. 
Torres-Orozco, E. 

1993. Volumetric analysis of the water masses of the Gulf 
of California. M.S. thesis, 80 p. Centre de Investig- 
acion Cientifica y de Educacion Superior de Ensenada, 
Ensenada, Mexico. 
Trasviiia, A., D. Lluch Cota, A. E. Filonov, and A. Gallegos. 

1999. El Niiio and oceanography. Chapter 3 in El 
Nino impacts in Mexico (Victor O. Magana R., ed.), 
p. 69-102. Secretariat of the Interior, Mexico. D. F. 



204 



Abstract — The number of pelagic 
fish eggs (cod and cunner) found in 
stomachs of capelin tMallntus villosiis) 
sampled in coastal Newfoundland was 
used to estimate the encounter rates 
between capelin and prey, and thus 
the effective volume swept by capelin. 
Fish eggs were found in 4-8'7f of cap- 
elin stomachs, represented an average 
of 1% of prey by numbers, and their 
abundance increased as relative stom- 
ach fullness decreased. The average 
number of eggs per stomach doubled 
for each 5-cm increase in length of 
capelin. The effective volume swept 
for eggs by capelin ranged from 0.04 
to 0.84 m-Vh — a rate that implies 
either very slow capelin swimming 
speeds (<1 cm/s) or that fish eggs are 
not strongly selected as prey. The pre- 
dation rate estimated from stomach 
contents was higher than that pre- 
dicted from laboratory studies of feed- 
ing pelagic fish and lower than that 
predicted by a simple foraging model. 
It remains uncertain whether capelin 
play an important regulatory role in 
the dynamics of early life stages of 
other fish. 



Estimating the encounter rate of Atlantic capelin 
(Mallotus villosus) with fish eggs, 
based on stomach content analysis 



Pierre Pepin 

Fisheries and Oceans Canada 

P.O. Box 5667 

St Johns, NL 

Canada A1C 5X1 

E-mail address Pepinpia^dfo-mpo gc.ca 



Capelin (Mallotiis villosus (Miiller, 
1776)) is a key species in Arctic eco- 
systems, serving as a major prey 
for top predators, from fish to birds 
and whales (Akenhead et al., 1982; 
Gj0saeter, 1998; Vilhjalmsson, 2002). 
Off the coast of Newfoundland and 
Labrador (eastern Canada), capelin 
are also the dominant consumers of 
secondary production (O'Driscoll et 
al., 2001), as they are in other ecosys- 
tems (Skjoldal and Rey, 1989; Hassel 
et al., 1991). They feed on a wide 
variety of zooplankton taxa, copepods 
and euphausiids being the dominant 
prey (Vesin et al., 1981; Panasenko'; 
Huse and Toresen, 1996; Assthors- 
son and Gislason, 1997; O'Driscoll et 
al., 2001). The choice of prey shows 
some degree of size-dependency; 
copepod size increases with increas- 
ing size of capelin, and euphausiids 
become more frequent prey items as 
fish size increases. In most studies 
of the feeding habits of capelin, fish 
eggs and larvae have been found to 
have been eaten by capelin, although 
these prey types generally represent 
a minor portion of the diet, occurring 
in less than 5% of stomachs (Huse 
and Toresen, 1996; O'Driscoll et al., 
2001). Because capelin are dominant 
consumers of zooplankton in many of 
the ecosystems in which they are pres- 
ent, their overall impact on the sur- 
vival of pelagic fish eggs and larvae 
may be significant depending on their 
encounter rates with these fish eggs 
and larvae. 



Manuscript submitted 2 November 2004 
to the Scientific Editor's Office. 

Manuscript approved for publication 
10 August 200.5 by the Scientific Editor. 

Fish. Bull. 104:204-214(20061. 



' Panasenko, L. D. 1984. Feeding of 
Barents Sea capelin. ICES CM 1984/ 
H:6, 16 p. 



There is evidence that capelin can 
have a significant impact on the 
survival of ichthyoplankton through 
predation on pelagic fish eggs and 
larvae. In a series of patch studies, 
Pepin et al. (20021 found that the 
mortality rate of radiated shanny 
(Ulvaria subbifurcata, Storer 1839) 
larvae increased in direct relation 
to hydroacoustic estimates of capelin 
abundance; this loss could be attribut- 
ed to predation by capelin on fish lar- 
vae. In a study of two coastal areas, 
differences in the regional patterns 
of selective loss of radiated shanny 
larvae, derived from sequential obser- 
vations of the distribution of growth 
rates, were strongly associated with 
differences in the spatial distribution 
and abundance of capelin (Baumann 
et al., 2003; Pepin et al., 2003). How- 
ever, the overall impact of capelin on 
fish eggs and larvae has been difficult 
to estimate (Paradis and Pepin, 2001). 
Much of the basic foraging ecology is 
unknown (e.g., predator's reactive dis- 
tance, behavior of both predator and 
prey, and selection of prey) and labo- 
ratory experiments can only provide 
a rough approximation of encounter 
rates between capelin and their prey 
(Paradis et al., 1996). 

There have been few studies to pro- 
vide an estimate of the contribution 
of predation by planktivorous fish 
on the overall mortality rates of fish 
eggs or larvae (Hunter and Kimbrell, 
1980; Cowan et al., 1999; Koster and 
MoUmann, 2000; Munk, 2002; Pepin 
et al., 2002). Population studies have 
shown that planktivorous fish can 
play an important role in regulating 
the stock-recruitment relationship of 
other species (Swain and Sinclair, 



Pepin: The encounter rate of Mallotus villosus with fish eggs 



205 









Table 1 








Location, average zooplankton dens 


ty (all organisms), average fish 


egg density 


and number of capelin iMallotus villosini) stom- | 


achs 


collected at each sampling site 
















Average 


Average 




No. of stomachs colleected at each site 


















zooplankton 


fish egg 


00:01- 


04:01- 08:01- 12:01- 16:01- 


20:01- 


Site 


Latitude Longitude 


density l/m-'i 


density (/m') 


04:00 


08:00 12:00 16:00 20:00 


24:00 


1 


47 44N 53 40'W 


1255 


0.094 


50 


45 — — — 


_ 


2 


48'18'N 53=16'W 


319 


0.91 


35 


50 17 30 18 


48 


3 


47"44'N 53='38'W 


2766 


0.23 


— 


100 22 — — 


— 


4 


48°02'N 53 36'W 


210 


0.29 


99 


50 99 50 9 


49 


5 


48°19'N 53'15'W 


3096 


0..58 


51 


49 44 49 49 


47 



2000; Koster et al., 2001; Garrison et aL, 2002). Indi- 
vidual-based models have often served as a tool to infer 
the potential effect of predators on early life stages 
(Cowan et al., 1999; Paradis et al., 1999; Werner et al., 
2001) where the likelihood of a prey being eaten by a 
predator is the product of the probabilities of encounter, 
attack, and capture. Although the latter two components 
of the predation process have been estimated in the 
laboratory for a number of species (see Paradis et al., 
1996, for a review), encounter rates are often modeled 
as the product of predator's swimming speed and reac- 
tive distance with limited substantive data to support 
the overall estimate. Furthermore, encounter rates are 
assumed to occur randomly and thus follow a Poisson 
distribution; yet there have been few attempts to vali- 
date this assumption (but see Huse and Toresen, 2000). 
Given that encounter rates represent a key parameter 
required to estimate the potential impact of predators 
on early life stages, it is surprising that knowledge of 
this element in the field is so limited. 

The goal of the present study was to estimate effective 
encounter rates of adult capelin with fish eggs and larvae 
in coastal Newfoundland. The approach was to obtain 
a high number of specimens from a series of intensive 
diurnal collections to provide an accurate estimate of 
the probability distribution of prey numbers from which 
encounter rates could be derived. The primary focus of 
this analysis was on the presence and frequency of occur- 
rence of fish eggs as prey for capelin, because they per- 
sist longer in stomachs than do fish larvae (Hunter and 
Kimbrell, 1980; Folkvord, 1993). I used estimates of prey 
availability from plankton samples combined with the 
frequency of occurrence of eggs in stomachs, in relation 
to the size of capelin and time of day, to estimate effec- 
tive encounter rates (volume searched) by capelin. Fish 
eggs are well suited to this objective: they are completely 
passive and thus the probabilities of their avoidance of 
both plankton nets and predators can be considered as 
nil. It is thus possible to measure accurately their abun- 
dance in the field, given appropriate choice of mesh size, 
and their numbers in predator stomachs can provide a 
measure of the effective volume searched, which is the 
product of the probabilities of encounter and attack. 



Materials and methods 

The study was conducted in Trinity Bay, Newfound- 
land, Canada, between 17 and 31 July 2000, as part 
of a larger program dealing with the dynamics of ich- 
thyoplankton (Baumann et al., 2003). Five sites were 
selected at which to conduct diurnal sampling of capelin 
based on the observation of high juvenile and adult 
capelin densities from hydroacoustic returns. Sampling 
was conducted within a 3-km radius of each site and at 
4-hour intervals over a single twenty-four hour interval. 
Each site was treated as an independent observation of 
the diurnal feeding patterns of capelin. As a result of 
poor catches, sampling was stopped after two sampling 
periods at the two sites located at the head of the bay 
(Table 1). 

Capelin were sampled using an international young 
gadoids pelagic trawl (lYGPT) mid-water-trawl towed 
in a single oblique haul from the surface to a 100-m 
depth. Each tow lasted approximately 30 minutes. Up 
to 50 juvenile and adult capelin were taken at random 
from the catch. When capelin were highly numerous, 
sample size was increased to 100. For each specimen, 
total length (TL) was measured to the nearest mm, 
the entire intestinal tract was removed by cutting the 
oesophagus as close to the mouth as possible, and the 
tract was then preserved in individual numbered vials 
containing 5% formaldehyde. A total of 1052 specimens 
were obtained from these collections. After each trawl 
haul, a plankton sample was taken by using a 60-cm 
bongo net (fitted with 333-im mesh net and flowmeters) 
towed obliquely from the surface to 100 m along the 
same path as the mid-water-trawl, and the sample was 
preserved in 2% formaldehyde solution. 

Stomachs were processed by slitting them open and 
scraping out the contents onto preweighed aluminium 
weight boat. The contents were blotted dry to remove 
all excess liquid and weighed to the nearest 0.01 mg 
by using a Cahn microbalance. Stomach contents from 
each capelin specimen were placed into individual petri 
dishes and examined for the presence, number, and 
taxonomic identification of pelagic fish eggs only. An egg 
was counted only when the specimen was complete and 



206 



Fishery Bulletin 104(2) 



the chorion was relatively intact. The presence of empty 
chorions was noted but these were not included in the 
analysis because of their potentially longer residence 
time in the stomach of predators (Hunter and Kimbrell, 
1980). As it turned out, empty chorions occurred in less 
than 2'7f of capelin with fish eggs present in their gut. 
For every fifth capelin specimen, all prey items were 
identified according to broad taxonomic groups: Co- 
pepoda, Amphipoda. Euphausia, Pteropoda, molluscan 
veligers, brachyuran zoeae, Larvaecea. Cladocera, Chae- 
tognatha, Isopoda, and fish eggs and larvae. The num- 
ber of individuals was determined for each taxonomic 
group from the entire stomach content and the total 
weight for each group was determined to the nearest /.ig 
by using a Cahn microbalance. This protocol was fol- 
lowed to provide a complete assessment of the presence 
of fish eggs in the gut and to provide a representative 
assessment of the patterns of prey composition from 
each site to confirm that general feeding patterns were 
consistent with results from previous studies. O'Driscoll 
et al. (2001) had performed a detailed analysis of feed- 
ing patterns of capelin in Newfoundland waters and I 
perceived no need or advantage in repeating the work. 
The complete analysis of stomach contents was per- 
formed on 218 capelin specimens. 

All fish eggs and larvae were sorted from plankton 
samples and identified to the lowest taxonomic level 
possible. Because of difficulties in differentiating early 
developmental stages of two species groups, eggs were 
lumped into either of two groups: 1) CHW (namely, cod 
(Gadus morhiia L. ), haddock iMelanogrammus aeglefi- 
nus L.) and witch flounder (Glyptocephalus cynoglossus 
L.) or 2) CYT (cunner (Taiitogolabrus adspersus L.), 
yellowtail flounder [Limanda ferriiginea, Storer 1839)). 
Eggs were most likely those of cod, for the CHW cat- 
egory, and cunner, for the CYT category because only 
larvae of those species were found in other elements of 
the research program. A subsample of other zooplank- 
ton, yielding 200-300 specimens, was identified to the 
lowest taxonimic level possible. 

For each capelin, the total gut fullness index (GFI) 
was estimated as GFI=WJW, where W,=the weight of 
stomach contents (g); and W=is the weight of the fish 
(g), based on the local bias-corrected length-weight rela- 
tionship (log,|j W=-6.44+3.38 logjo TL, where W=weight 
in g and T'=total length in mm; r^ = 0.98; and the stan- 
dard error of the intercept and slope are 0.049 and 
0.023; Mowbray2). 



Analysis 



was determined. These data, arcsine-square-root trans- 
formed, were contrasted using principal component anal- 
ysis in which the proportion of prey type, log-transformed 
capelin length and GFI were included to determine if the 
observations from this study were consistent with those 
from previous ones. 

Frequency of fish eggs 

The probability distribution of egg numbers per gut 
was contrasted with the expectations based on a Pois- 
son process. Encounters between predator and prey can 
be described by using a modification of Gerritsen and 
Strickler's (1977) equations where 



■^■R~ -^pr,; VU~ + V'^, 



(1) 



where 7? = the reactive distance (m); 

A .,,^, = the density of prey (/m'^); and 
u and V = the swimming speeds of predator and prey 
(m/h), which yields an encounter rate A (per h). 

Huse and Toresen (2000) used laboratory estimates 
of the various parameters to estimate encounter rates 
between juvenile herring iClupea harengus L.) and cap- 
elin larvae. Instead of using their approach, I chose to 
reduce Equation 1 to its simplest form, such that 



X = K A„ 



(2) 



where K = the effective volume swept (m^/h) during 
the period during which eggs will remain 
discernible in the gut of capelin; and 
A . - the observed density of eggs (/m^) in the 
water column from plankton samples. 

The resulting probability of finding A'^ prey in the gut of 
capelin then becomes 



P(Ar) = -.e-\ 



where A = the mean encounter rate. 



(3) 



Stomach content composition 



Mean number of eggs per stomach (A''), taken as a mea- 
sure of encounter rates for each four-hour sampling 
interval, was estimated by fitting a generalized linear 
model with Poisson error structure using a log link 
function by maximum likelihood (GENMOD procedure, 
vers. 8, SAS Inc., Gary, NC) to the data from all 4-hour 
sampling intervals (T) from all locations, with TL of 
individual capelin as a linear covariate. 



For each capelin on which the full stomach contents 
were analyzed, the number of each major prey taxon 



2 Mowbray, F. 2005. Personal commun. Fisheries and 
Oceans, St. John's, Newfoundland, Canada. 



f(N)^h^,+h{r + h.{rL. 



(4) 



The value oi N reported in the "Results" section is the 
least squares adjusted mean, which represents the value 
of the average length of capelin from field collections. 
We attempted to include site into the analysis but the 



Pepin The encounter rate of Mallotus villosus with fish eggs 



207 



algorithm failed to converge because of missing observa- 
tions. Effective volume swept (K) was calculated from 
the estimated mean number of eggs per stomach (as a 
measure of encounter rates) and the estimated density 
of eggs based on plankton samples by re-arranging 
Equation 2 and assuming that encounters between 
prey and predators were random over the top 100 m 
of the water column. I define K as the effective volume 
swept because it implicitly includes some measure of the 
probabilities of attack and capture. Values of the esti- 
mated volume swept were corrected for evacuation rates. 
Hunter and Kimbrell (1980) found that the evacuation 
rate of northern anchovy {EngrauUs mordax, Girard 
1854) feeding on fish eggs at 15°C was -SO'/r/h based 
on an exponential model (exponential rate of 0.7/h), 
which is appropriate for microphagic fish (Temming et 
al., 2002). Based on the duration of our tow and the 
handling time to process the samples, the observations 
of stomach content from this study represent 50% of 
the eggs consumed during the last hour before capture. 
Hunter and Kimbrell's (1980) estimate of the evacuation 
rate is approximately twice that obtained in field studies 
by Arrhenius and Hansson (1994) and Darbyson et al. 
(2003) for herring (C. harengus L.) feeding primarily on 
copepods. However, the estimates from the latter two 
studies were based on total gut fullness and not on the 
disappearance of specific food items, such as fish eggs. 
As a result. Hunter and Kimbrell (1980) provided the 
only current measure of the rate of disappearance of 
eggs in the stomachs of planktivorous fish. Increases or 
decreases in evacuation rates because of inaccuracies in 
our knowledge of the rate of disappearance of eggs from 
stomach contents will result in corresponding changes 
in the estimate of the effective volume swept. By using 
the mean number of eggs per stomach as an estimator 
of the effective volume swept, the probabilities of attack 
and capture within the reactive distance are assumed 
to be 100%. The resulting value of the effective volume 
swept represents an average over the top 100 m of the 
water column, the depth range over which we sampled 
both capelin and zooplankton and over which I assume 
that the encounters are random. Ideally, one would 
want to obtain a depth-stratified measure of stomach 
contents, and prey availability and overlap between prey 
and predator, such as has been inferred in the Baltic 
Sea (Kbster and Mbllmann 2000; Koster et al. 2001), 
but this was not possible in the present study because 
there were insufficient data on the vertical distribution 
of capelin. 



Results 

Capelin used in this analysis ranged from 53 to 195 
mm TL (mean = 122 mm; median = 130 mm). Average 
zooplankton concentrations ranged from 210 to 3096 
organisms/m'. In most samples, the copepods Calanus 
finmarchicus and Pseudocalanus sp. and the larva- 
cean Fritillaria borealis (Lohmann 1896) represented 
60-80% of the zooplankton by number, based on samples 



? 0.01 



0.0001 



10000 



1000 



o 




70 



90 110 130 150 170 190 

Length (mm) 






••• Vj* 

• • •* 



70 



90 



110 130 
Length (mm 



^ — »A« — * 
150 170 190 



-1,4 


r 














-1.6 


- 




-1.8 


- 










-2.0 


- ' 










 


-2.2 


- 


' 










[ 


•2.4 


-  












-26 


J 










-2.8 


' ' ' 1 




' 



8 12 16 20 

Time of day (h) 



24 



Figure 1 

Weight of stomach contents (upper graph) 
and number of prey per stomach (center 
graph) in relation to capelin {Mallotus 
villosus) length, for specimens with com- 
plete analysis of contents, and average 
and standard deviation of the log-trans- 
formed gut fullness index (GFI) in rela- 
tion to time of day ( lower graph), based 
on all fish sampled. Statistics were cal- 
culated for all samples collected within 
each 4-hour time interval. 



from bongo nets. In both species of copepods, copepodite 
stages II and III were most abundant. A finer mesh 
net would have revealed that Oithoina similis was also 



208 



Fishery Bulletin 104(2) 





Table 2 




Frequency of occurrence (number of capelin [Mallo- 
tiis I'illosus] for which complete stomach analysis was 
performed) and average relative proportion of prey (by 
number) for the dominant prey categories found in July 
2000. The total number of capelin examined for complete 
stomach analysis was 218. There were 12 fish with empty 
stomachs. 


Prey type 


Frequency 
of occurrence 


Relative 
proportion 


Copepoda 


176 


0.77 


Euphausia 


50 


0.066 


Amphipoda 


43 


0.088 


Pteropoda 


22 


0.010 


Larvacea 


21 


0.032 


Cladocera 


17 


0.022 


Fish eggs 


16 


0.010 


Chaetognatha 


4 


<0.001 


Mollusca veliger 


3 


0.001 


Brachyura zoea 


2 


0.005 


Isopoda 


1 


<0.001 



numerically important in the area (Pepin an(i Maillet''), 
but O'Driscoll et al. (2001) did not find this species in 
any of their stomach samples despite its importance in 
the regional plankton community. Fish eggs represented 
an average of 0.13% of the total zooplankton community 
among all sites sampled (Table 1). 

Only 12 of 218 capelin stomachs, for which com- 
plete stomach analysis was performed, were found 
to be empty. The weight of prey in capelin stomachs 
increased with increasing length but the number of 
prey decreased with increasing length, indicating that 
the average size of prey increased with capelin length 
(Fig. 1). Stomach fullness increased at sunset, peaked 
shortly after midnight, and then gradually decreased 
during the remainder of the night and day (Fig. 1). 
Copepods were the only prey in 73 of the remaining 
206 capelin. Euphausiids and amphipods were the next 
most abundant prey item, found in approximately 25% 
of capelin stomachs and representing on average 6.5% 
and 8.8% of prey by number (Table 2). Pteropods, larva- 
ceans, cladocerans and fish eggs were present in approx- 
imately 10% of capelin stomachs and represented -1-3% 
of prey by numbers (Table 2). Other prey taxa were 
found in a few capelin stomachs. Principal components 
analysis showed that the proportion of euphausiids 



^ Pepin, P.. and G. L. Maillet. 2002. Biological and chemi- 
cal oceanographic conditions on the Newfoundland Shelf 
during 2001 with comparisons with earlier observations. 
Canadian Science Advisory Secretariat, Research Document 
2001/073, 60 p. [Available from http;//www.dfo-mpo.gc.ca/ 
csas/ or Canadian Science Advisory Secretariat, Depart- 
ment of Fisheries and Oceans, 200 Kent Street, stn. 12032, 
Ottawa, Ontario, Canada KIA 0E6.J 







Table 3 








Results of the 


principle 


components (PC) analysis 


of pro- 


portion of prey 


categories from capelin stomachs 


Values 


in parentheses 


along the topmost 


row represent the pro- 


portion of the 


variance 


explained by each pr 


nciple com- 


ponent. Colum 


n values 


represent 


the average loading of | 


each variable i 


ncluded 


n the ana! 


ysis. GFI= 


SUt 


'uUness 


index. 
















PCI 


PC 2 




PC 3 


Variables 




(17<7f) 


(12'7f) 




(10'7r) 


Ln icapclin len 


gth\ 


0.494 


-0.068 




-0.144 


LnlGF/l 




0.166 


-0.575 




0.325 


Copepod 




-0.598 


-0.276 




-0.081 


Cladocera 




-0.035 


0.315 




0.593 


Pteropoda 




0.044 


-0.131 




-0.050 


Amphipoda 




0.490 


-0.017 




-0.026 


Euphausia 




0.330 


-0.153 




0.122 


Cheatognatha 




-0.066 


-0.083 




-0.110 


Mollusca 




-0.017 


0.260 




0.550 


Isopoda 




-0.023 


0.003 




-0.173 


Larvacea 




-0.011 


0.440 




-0.065 


Brachyura 




0.038 


0.280 




-0.310 


Fish eggs 




0.106 


0.340 




-0.232 



and amphipods in stomachs increased with capelin 
length, whereas copepods were more frequent in small- 
er fish, as indicated by the first principal component 
(Table 3). Other prey categories, particularly fish eggs 
and larvaceans, were generally more likely to occur in 
capelin stomachs when they were relatively empty, as 
indicated by the second principle component (Table 3). 
Finally, the third principal component indicated that 
fish eggs appeared to be less important in areas where 
mollusks and Cladocera were more frequent in the diet 
of capelin — a finding that possibly reflects the influence 
of water masses in which coastal rather than shelf zoo- 
plankton were dominant (Table 3). These observations 
of prey composition were generally consistent with those 
of O'Driscoll et al. (2001). 

When we consider all stomachs analyzed for fish eggs, 
the overall frequency of occurrence of fish eggs was 
4.3%- (46 of 1052 stomachs). This is less than the value 
(7.9% of stomachs with prey) obtained from those stom- 
achs with complete analysis of prey composition. The 
frequency distribution of fish eggs was well described 
by a Poisson distribution (Fig. 2). The majority offish 
eggs (89%) were of the CHW complex and the remainder 
(11%) were of the CYT complex. The generalized linear 
model revealed that there was a significant effect of 
time of day (Table 4), and the greatest number of fish 
eggs in capelin stomachs were found between 20:00 
and 24:00, following sunset, and the lowest numbers 
were found between 00:00-04:00 (Fig. 3). There was 
approximately a 20-fold difference in the least squares 



Pepin: The encounter rate of Mallotus villosus with fish eggs 



209 





Table 4 


Analysis of deviance for the effects of time of day and 
capelin TL with Poisson error based on a generalized 
linear model. The effect of time of day was based on group- 
ing data into 4-hour classification periods. Probability 
values were calculated by using a x^ function. 


Difference 


Residual 
degrees of 
Variable freedom (df) 


Residual 

deviance df Deviance P 


1052 


365.9 


Time of day 1046 


321.7 6 44.2 <0.001 


Length 1045 


317.2 1 4.5 <0.05 



mean estimate of the number of eggs per stomach over a 
diurnal cycle. There was a significant and positive effect 
due to capelin length (x~=4.5, P<0.05; Table 4) which 
would yield a twofold increase in the mean number of 
eggs per stomachs for each 5-cm increase in capelin 
length. This finding may indicate that smaller capelin 
are more effective predators of fish eggs, in terms of 
body mass, than larger individuals. There were insuf- 
ficient data to obtain reliable estimates of the diurnal 
feeding periodicity at each site. 

The overall unweighted average density of fish eggs 
among all sites was 0.52/m3 (median = 0.48/m^; mean= 
0.53/m*; SD = 0.43/m*; range = 0.09-2.2/m-^). There were 
significant differences in the densities of fish eggs 
among sites (ANOVA F_, .^j=4.56, P<0.01); a Bonferoni 
post hoc comparison revealed that site 2 had significant- 
ly higher densities than other sites, with the exception 
of site 5 (Table 1). However, there was no clear evidence 
of a diurnal pattern to the density estimates derived 
from plankton collections. Because there were insuf- 
ficient data to obtain reliable estimates of the diurnal 
feeding periodicity at each site and because there was 
no clear evidence of a diurnal pattern in egg densities 
estimated from the plankton samples, I chose to use the 
overall unweighted average density from all collections 
as the estimate oi A , for Equation 2. 

Because it was not possible to obtain a site-specific 
estimate of diurnal patterns in egg consumption by cap- 
elin, the combined data on egg densities were used to 
estimate the effective volume swept (K). Each estimate 
of the mean number of eggs per stomach was taken to 
represent 50% of the eggs ingested in the last hour dur- 
ing each sampling interval (4 h) (see "Methods"). The 
overall daily average effective volume swept based on 
mean encounter rates was 0.27 m^/h, which ranged from 
0.04 m'Vh at night {00:00-04:00) to 0.84 m'Vh following 
sunset (20:00-24:00) (Fig. 4). Based on a reactive dis- 
tance of 0.15 m (~1 body length) used by Huse and Tore- 
sen (2000) (for herring larvae from Utne-Palm, 2000), 
the effective volume swept estimated from the observa- 
tions reported in the present study would imply that the 



1 - 



01 



0,001 



00:01-04:00 
04:01-08:00 
08:01-12:00 
12:01-16:00 
16:01-20:00 
20:01-24:00 



12 3 

Number of fish eggs per stomach 

Figure 2 

Bar chart of the frequency offish eggs in capelin (Mallo- 
tus villosus) stomachs for each 4-hour time interval. 



1 




o 

CD 

E 

o 
To 


' 


g. 0.1 


T 1 






0) 


< 


' 1 


< 




o 






-'- 




QJ 

1 001 


1 


> 




c 






c 

a> 

5 






0-001 


^ 




4 8 12 16 20 24 


Time of day 


Figure 3 


Least squares estimates of the mean number of fish 


eggs per stomach in relation to time of day. Error bars 


represent 1 standard error. 



swimming speed of capelin would be very low (si cm/s) 
or that fish eggs are not strongly selected as prey dur- 
ing feeding. If on the other hand we consider modal 
capelin (0.13 m) swimming at 0.5-1 body length/s, 
the reactive distance would range from 0.024-0.034 m 
at the time of peak foraging activity, assuming 100% 
probability of attack within the reactive area. To de- 
termine the effect of variations in the density of eggs 
encountered by capelin, I also calculated the variability 



210 



Fishery Bulletin 104(2) 



10 



01 



0.01 



0.001 




8 12 16 

Time of day (h) 



20 



24 



Figure 4 

Estimated effective volume swept (A'), with 
±1 standard error (solid symbols and thick 
error bars) based on mean number of eggs 
per stomach and mean density of eggs from 
plankton collections, in relation to time of day 
(offset by +1 hour from midpoint of interval). 
Confidence intervals are based on the maxi- 
mum likelihood estimates of mean numbers of 
eggs per capelin stomach. Box plots show the 
25"^, 50"', and 75"' percentiles of the estimated 
effective volume swept (K) based on the mean 
number of eggs per stomach and each estimate 
of egg density based on plankton collections; 
whiskers show the 10"' and 90''' percentiles 
of the distribution and open symbols show the 
outlying points. 



in estimated K using each observation of egg density 
from the plankton samples. This calculation generally 
resulted in less variability in the estimate of effective 
volume swept than the uncertainty in eggs per stomach 
(Fig. 4, box plots) because of the skewed distribution 
of egg densities based on plankton tows. The latter 
reflects the skewed distribution of egg densities based 
on plankton tows. 



Discussion 

Capelin prey composition observed in this study is con- 
sistent with that observed in previous studies, which 
show that copepods, euphausiids, and amphipods repre- 
sent the principal prey items; the overall results of this 
study may therefore have general applicability. Fish eggs 
represent a relatively minor part of the diet of capelin. 
Neither Huse and Toresen (1996) nor Assthorsson and 
Gislason (1997) reported any occurrence of pelagic fish 
eggs in the stomachs of capelin from the Barents Sea 
and waters north of Iceland. O'Driscoll et al. (2001) 
reported an occurrence of less than 1% in capelin sam- 



pled during May-June in inshore waters and during 
August-September in offshore areas off Newfoundland. 
In the present study, an occurrence of 4*7^ to 8%, which 
contributed an average of 1% of prey by numbers, was 
found for collections in coastal Newfoundland waters 
during July. There is little basis for explaining the 
differences between studies because only this analysis 
provides an estimate of the density of eggs in the water 
column where capelin were sampled. Fish eggs were 
approximately 1% of the prey found in the stomachs of 
capelin but they represented slightly more than 0.1% 
of the organisms sampled by the plankton nets used in 
this study. Huse and Toresen (1996) collected plankton 
data with different gear than that used in the present 
study, but they did not report the occurrence of pelagic 
fish eggs in their samples, possibly because pelagic fish 
eggs were in very low abundance or were absent during 
the collection periods of their study. 

In general, fish eggs were more likely to be found in 
capelin stomachs as the length of individual fish in- 
creased. These fish tend to feed more heavily on larger 
prey types (e.g., euphausiids and amphipods) but the 
occurrence of fish eggs in individual fish tended to in- 
crease as the relative stomach fullness decreased. When 
considered with the general pattern in feeding periodic- 
ity, fish eggs were more likely to be eaten at the onset 
of feeding activity shortly after dusk, but as individual 
fish filled their guts with larger or more numerous prey, 
feeding on fish eggs tended to decrease. Fish eggs may 
represent an attractive prey for capelin capable of feed- 
ing on larger prey types (e.g., stage V of C. finmarchi- 
cus, euphausiids, and amphipods) because their weight 
is comparable to that of large copepods (Darbyson et al., 
2003). However, as larger prey types are encountered or 
as hunger decreases, capelin may become more selective 
feeders and focus their foraging on energetically more 
profitable prey. The diurnal pattern of occurrence of fish 
eggs in capelin stomachs may also reflect differences 
in the visibility of eggs as a result of changing light 
levels, as well as predator hunger. Alternatively, the 
greater number of eggs in stomachs at dusk may reflect 
increased spatial overlap between predator and prey as 
capelin move toward the surface in the evening to feed 
(O'Driscoll et al.^). 

The effective volume swept by capelin estimated in 
the present study indicates that the probability that 
a capelin will attempt to eat a pelagic fish egg is low, 
because it is highly unlikely that fish will swim at 
a very slow speed based on an assumed reactive dis- 
tance. The predation rates based on average densi- 
ties and encounter rates range from 0.04 to 0.86 egg/h 



'O'Driscoll R. L., J. T. Anderson, and F. K. Mow- 
bray. 2001b. Abundance and distribution of capelin from 
an acoustic survey in conjunction with the 1999 pelagic 
juvenile survey. Capelin in SA2 + Div. 3KL during 1999. 
Canadian Science Advisory Secretariat Research Document 
2001/161. [Available from http://www.dfo-mpo.gc.ca/csas/ 
or Canadian Science Advisory Secretariat, Department of 
Fisheries and Oceans, 200 Kent Street, stn. 12032, Ottawa, 
Ontario, Canada KIA 0E6.] 



Pepin: The encounter rate of Mallotus villosus witli fish eggs 



211 



based on an evacuation rate of 5Q'7< of eggs per hour. 
A low predation rate on fish eggs is not entirely unex- 
pected. The eggs of cunner and cod from our samples 
had an average diameter of 0.86 and 1.3 mm, whereas 
the modal length of capelin from our collections was 
approximately 130 mm, making the eggs -1% of the 
predator's body length. Based on Paradis et al.'s (1996) 
analysis, maximum predation rates by fish feeding on 
ichthyoplankton occur when the prey are 10% of the 
predator's length, and the minimum predation rates 
were observed when prey were -0.5-1% of the predator's 
length. Paradis et al.'s (1996) summary equations indi- 
cate that an average predation rate of 0.007-0.03 eggs/ 
h would not be inconsistent with laboratory estimates 
of predation rates for a range of evacuation times from 
1 to 4 hours and would be considerably lower than the 
predation rates estimated in the present study. How- 
ever, estimates based on a simple foraging model (e.g., 
Paradis et al., 1999) for a modal capelin swimming at 
1 body length/s and with a reactive distance of 1 body 
length yield a mean of 12 eggs/h, and halving these 
parameter values yields a predation rate of 1.6 eggs/h. 
These alternative methods of estimating predation rates 
indicate that capelin are somewhat more likely to feed 
on fish eggs in the field than laboratory studies would 
indicate but less likely than would be anticipated from 
a simple foraging model. The results from this study do 
not provide an exhaustive estimate of encounter rates 
between capelin and pelagic fish eggs, but the range 
of effective encounter rates derived from the different 
approaches highlights the lack of understanding of this 
key parameter essential for any model of multispecies 
interactions. Such a broad range of expected predation 
(and encounter) rates indicates that in order to develop 
predictive skills in estimating the potential impact of 
planktivores on fish eggs, a better understanding of the 
factors that prompt a predator to feed on a fish egg is 
needed. Most laboratory studies dealing with predation 
by fish on fish eggs and larvae have used single prey 
and few have provided a measure of the probability of 
attack (Bailey and Houde, 1989; Paradis et al., 1996). 
Feeding patterns can be affected by the complexity of 
the available prey community (Kean-Howie et al., 1988; 
Gotceitas and Brown, 1993; Pepin and Shears, 1995), 
but methods to relate prey consumption with prey avail- 
ability in the field are still limited. 

Uncertainty in encounter and predation rates are due 
partly to errors in the estimation of prey availability 
and stomach content and partly due to our general lack 
of knowledge concerning the evacuation rate of fish 
eggs from predator stomachs. For this study I relied on 
Hunter and Kimbrell's (1980) observations of evacuation 
rates for northern anchovy from experiments conducted 
at 15°C. Water temperatures in Trinity Bay in July 
2000 showed considerable variation with depth, rang- 
ing from ~13°C at the surface to -1°C at 100 m, with 
the overall average temperature over the water column 
being ~3-4°C. Given that most metabolic processes 
increase by 1.3-2 times over ten degrees Celsius (e.g., 
Brett and Groves, 1979), evacuation rates in the pres- 



ent study could have been half of those measured by 
Hunter and Kimbrell (1980) and thus would imply that 
our observations of stomach contents could represent 
70% of the eggs consumed in the last hour rather than 
the 50% assumed in the analysis. The result would be 
that estimates of volumes swept by capelin would have 
to be reduced by -30%, further reducing their potential 
impact on ichthyoplankton populations. The general un- 
certainty in our knowledge of evacuation rates for fish 
eggs from the stomachs of capelin, and from most other 
planktivorous fish, therefore limits the inferences that 
can be derived about their impact on ichthyoplankton 
survival. 

A second source of uncertainty is due to the inher- 
ent sampling variability associated with the study of 
elements that form a small fraction of the prey of a 
predator. This study was based on more than 1000 
specimens from a small region in order to provide the 
greatest accuracy possible within the confines of the 
study area. Stomach sampling is often restricted to 
fewer specimens sampled over a much broader geo- 
graphic range. Without a large number of observations 
from a local environmental setting (i.e., site), estimates 
of mean numbers of rare prey types would likely be 
unable to reflect the effect of site-specific differences 
in environmental conditions on predator feeding and 
would thereby increase the uncertainty around the 
estimated effective encounter rates. The situation is 
well illustrated by the doubling in the frequency of 
occurrence of fish eggs in the specimens that were ran- 
domly selected for complete gut analysis versus all the 
specimens collected for analysis. Further uncertainty 
comes from the variability in estimates of egg density 
encountered by capelin — a variability that tended to 
result in slightly tighter confidence intervals in esti- 
mates of effective volume swept than the uncertainty 
due to variability in the number of prey per predator 
stomach. Variations in egg densities from plankton 
samples did result in lower estimates of effective vol- 
ume swept, partly as a result of the rare high densities 
(typical of highly skewed distribution) that characterize 
the variability in plankton catches (Power and Moser, 
1999). Although this study is not intended to provide 
a general estimate of the effective encounter rate of 
capelin feeding on fish eggs, the approach provides a 
useful example of the sampling requirements for the 
study of feeding on prey that are a minor part of the 
diet of a predator but which may be greatly affected by 
the predator's impact on the prey population. 

Greater efforts must be directed at understanding 
patterns of predation in the field, in relation to prey 
availability, if the role of planktivorous fish in ichthyo- 
plankton dynamics is to be understood. The assump- 
tions in the estimation approach could affect the effec- 
tive volume swept: estimates of the effective volume 
swept are based on average prey densities integrated 
over the water column, and the simplification of the 
encounter model (Eq. 1) incorporates possible variations 
in the probability of attack (e.g., due to diurnal varia- 
tions in visibility of eggs) into the value of K. Pepin et 



212 



Fishery Bulletin 104(2) 



al. (2005) noted that in the study region, pelagic fish 
eggs are generally more abundant near the surface and 
decrease exponentially in abundance with increasing 
depth. As a result, if feeding by capelin occurs predomi- 
nantly in surface waters, the effective density of prey 
encountered would be greater than the average density 
over the water column, which would in turn decrease 
the estimate of the effective volume swept based on 
our estimates of encounter rates. However, the pat- 
terns of vertical distribution of both eggs (Pepin et al., 
2005) and capelin (O'Driscoll et al., 2001) are highly 
variable among sites and without accurate estimates 
of these elements from the current study, inferences 
about the impact on estimates of the effective volume 
swept would be speculative. In contrast, I implicitly 
incorporated the probability of attack into the estimate 
of the effective volume search, essentially giving it a 
value of 100%. However, Wieland and Koster (1996) 
found that the visibility of eggs increased with devel- 
opmental stage and hence may alter the probability 
of attack by a predator. Lower visibility of early stage 
eggs could lead to an increase in the estimate of the 
effective volume swept, but because we could not stage 
most eggs sampled from the stomachs, the net effect on 
the estimated volume swept is unclear. 

Planktivorous fish represent important forage for 
piscivores in several marine ecosystems but they can 
also play a significant role as predators of early life 
stages. In the Baltic Sea, spratt iSprattus sprattus L.) 
feed extensively on cod eggs (Koster and MoUmann, 
2000) and they may play an important role in regulat- 
ing the underlying form of the stock-recruitment rela- 
tionship (Koster et al., 2001). In upwelling systems, 
anchovies and sardines may also have a significant 
impact on egg and larval mortality rates (Smith et 
al., 1989). It has also been suggested that herring 
(C harengus L.) and mackerel (Scomber scombrus L.) 
predation on ichtyplankton may play an important 
role in fish population dynamics in the Gulf of St. 
Lawrence (Swain and Sinclair, 2000) and on Georges 
Bank (Garrison et al., 2002). However it is unclear 
whether capelin play a significant role in the fish 
community dynamics in the ecosystems they inhabit, 
despite evidence that they are likely to have a signifi- 
cant impact on zooplankton abundance (Vesin et al., 
1981; Akenhead et al., 1982; Skjoldal and Rey, 1989; 
Hassel et al., 1991). Clupeids, scombrids and engrau- 
lids can alternate between filter- and particulate-feed- 
ing modes, which may allow them to more effectively 
consume relatively small prey (~1 mm), such as fish 
eggs. There is no information on the feeding modes of 
capelin, but another osmerid, rainbow smelt (Osmerus 
mordax, Mitchill 1814), is known to be a primarily 
particulate feeder (Mills et al., 1995). Particulate feed- 
ing is generally directed toward larger prey, whereas 
filter feeding is used when prey are smaller or more 
abundant. Gill raker structure does not provide signifi- 
cant insight into the possible feeding modes of these 
various planktivorous fish. The estimate of reactive 
distance from the present study (0.024-0.034 m) is 



of the same order as the gape width of capelin (3.6% 
for TL; Pepin, unpubl. data), making filter feeding a 
distinctly possible mode of feeding. For known filter 
feeders, Uotani (1985) reported a maximum gill raker 
length of 6 mm in Engraiilis japonicus (100 mm TL), 
Gibson (1988) reported an average gill raker length of 
4 mm in Clupea harengus (150 mm TL), and Molina et 
al. (1996) reported gill raker lengths ranging from 5 to 
9 mm in Scomber japoiucus (150-250 mm SL). On the 
other hand, Lecomte and Dodson (2004) reported gill 
raker lengths of 3.3-3.9 mm in Osmerus mordctx (150 
mm TL). In the case of capelin, I found that gill rak- 
ers ranged from 3.0-4.5 mm in length (130-170 mm 
TL) (Pepin, unpubl. data). Inter-raker gaps among 
all these species ranges from 0.19 mm (£. japonicus) 
to 0.4 mm (S. Japonicus) and from 0.36 to 0.45 mm 
in capelin (Pepin, unpubl. data). Thus all these spe- 
cies, which feed extensively on zooplankton but which 
use different ingestion modes, are equipped with gill 
rakers designed to retain small particles. To better 
understand the role of capelin in Arctic ecosystems, a 
greater knowledge of feeding will be required. If filter 
feeding is a dominant feeding mode, a simple filtra- 
tion or volume-swept model may provide an accurate 
measure of the potential impact of capelin on fish eggs, 
whereas greater knowledge of behavior and selectivity 
will be required if particulate feeding is the dominant 
feeding mode because selection for or against fish eggs 
could be influenced by availability of alternate prey 
(e.g., Kean-Howie et al., 19881. 

In most of the areas where planktivores can have 
a significant impact on the mortality of fish eggs, the 
ecosystems can be characterized as being "wasp-waist- 
ed" (Cury et al., 2000), i.e, as having a relatively high 
diversity of plankton species and top predators that 
are all influenced by the abundance and productivity 
of relatively few and dominant forage fish species that 
act to transfer energy from secondary producers to up- 
per trophic levels in a manner similar to that found 
in many Arctic regions. It is clear that capelin are 
important prey for a number of top predators in the 
waters of the Barents Sea (Gjosaster, 1998), Iceland 
(Vilhjalmsson, 2002), and on the Newfoundland and 
Labrador coast (Akenhead et al., 1982). and they may 
also have an important impact on the plankton commu- 
nity (Skjoldal and Rey, 1989; Hassel et al., 1991). What 
remains uncertain is whether they play an important 
regulatory role in the dynamics of the early life stages 
of fish. Observations from this study revealed that the 
highest occurrence of fish eggs in the stomachs of cap- 
elin came from capelin near one of the few remaining 
major spawning aggregations of the northern cod stock 
(Rose, 2003). The dynamics of such circular prey-preda- 
tor interactions need to be studied further in order to 
understand their potential importance to population 
regulation (Walters and Kitchell, 2001) but to do so 
will require consideration of the spatial and temporal 
patterns of predator-prey interactions if predictive re- 
lationships are to be derived (Pepin et al., 2002, 2003; 
Baumann et al., 2003). 



Pepin The encounter rate of Mallotus villosus with fish eggs 



213 



Acknowledgments 

I would like to thank Tim Shears, Gary Maillet, Hugh 
Maclean, and Greg Redmond for their assistance in the 
field, Leah Wilson for performing the laboratory analysis 
of stomach contents, and the officers and crew of the 
CCGS Templeman for their diligence. Comments by J. 
Carscadden, J. Dower, G. Huse, F. Mowbray, and two 
anonymous referees helped to improve the manuscript. 



Literature cited 

Akenhead S. A., J. Carscadden , H. Lear, G. R. Lilly, and 
R. Wells. 

1982. Cod-capelin interactions off Northeast New- 
foundland and Labrador. Can. Spec. Pub. Fish. Aquat. 
Sci. 56:141-148. 
Arrhenius F., and S. Hansson. 

1994. In situ food consumption of young-of-the-year 
Baltic Sea herring Clupea harengus: a test of predic- 
tions from a bioenergetic model. Mar. Ecol. Prog. Ser. 
11:145-149. 
Assthorsson O. S., and A. Gislason. 

1997. On the food of capelin in the subsarctic waters 
north of Iceland. Sarsia 82:81-86. 
Bailey K. M., and E. D. Houde. 

1989. Predation on eggs and larvae of marine fishes and 
the recruitment problem. Adv. Mar. Biol. 25:1-83. 
Baumann H., P. Pepin, F. J. M. Davidson, F. Mowbray, 
D. Schnack, and J. F. Dower. 

2003. Reconstruction of environmental histories to 
investigate patterns of larval radiated iUlvaria sub- 
bifurcata) growth and selective survival in a large bay 
of Newfoundland. ICES J. Mar. Sci. 60:243-258. 
Brett J. R., and T. D. D. Groves. 

1979. Physiological energetics. In Fish physiology, vol. 
8 (W. S. Hoar, D. J. Randall, and J. R. Brett, eds.), p. 
279-352. Academic Press, New York, NY. 
Cowan J. H., K. A. Rose. E. D. Houde, S. B. Wang, and J. Young. 

1999. Modeling effects of increased larval mortality on 
bay anchovy population dynamics in the mesohaline 
Chesapeake Bay: evidence for compensatory reserve. 
Mar. Ecol. Prog. Ser. 185:134-146 

Cury P., A. Bakun, R. J. Crawford, A. Jarre, R. A. Quinones, L. 
J. Shannon, and H. M. Verheye. 

2000. Small pelagics in upwelling systems: patterns of 
interraction and structural changes in "wasp-waist" 
ecosystems. ICES J. Mar. Sci 57:603-618. 

Darbyson, E., D. P. Swain, D. Chabot, and M. Castonguay. 

2003. Diel variation in feeding rate and prey consumption 
of Atlantic herring iClupea harengun L.) and Atlantic 
mackerel (Scomber scomrbus L.) in the southern Gulf 
of St. Lawrence. J. Fish Biol. 63:1235-1257. 
Folkvord, A. 

1993. Prey recognition in stomachs of cannibalistic juve- 
nile cod iGcidus morhua L.). Sarsia 78: 97-100. 
Garrison, L. P., W. Mieheals, J. S. Link, and M. J. Fogarty. 

2002. Spatial distribution and overlap between ichthyo- 
plankton and pelagic fish and squids on the southern 
flank of Georges Bank. Fish. Oceanogr. 11:267-285. 
Gerritsen, J., and J. R. Strickler. 

1977. Encounter probabilities and community structure 
in zooplankton: a mathematical model. J, Fish. Res. 
Board Can. 34:73-82. 



Gibson, R. N. 

1988. Development, morphometry and particle retention 
capability of the gill rakes in herring, Clupea harengusi 
L. J. Fish. Biol. 32:949-962. 
GjosEBter, H. 

1998. The population biology and exploitation of cap- 
elin (Mallotus villoaus^ in the Barents Sea. Sarsia 
83: 453-496. 
Gotceitas, V., and J. A. Brown. 

1993. Risk of predation to fish larvae in the presence of 
alternative prey: effects of prey size and number. Mar. 
Ecol. Prog. Ser. 98:215-222. 
Hassel, A., H. R. Skjoldal, H. Gjosseter, H. Leong, and L. Omli. 
1991. Impact of grazing from capelin (Mallotus villosus) 
on zooplankton: a case study in the northern Barents 
Sea in August 1985. Polar Res. 10:371-388. 
Hunter, J. R., and C. A. Kimbrell. 

1980. Egg cannibalism in the northern anchovy, Eit- 
graulis mordcix. Fish. Bull. 78:811-816. 
Huse, G., and R. Toresen. 

1996. A comparative study of the feeding habits of herring 
(Clupea liarengus, clupeidae, L.) and capelin (Mallotus 
villosus. Osmeridae, MiiUer) in the Barents Sea. Sarsia 
81:143-153. 

2000. Juvenile herring prey on Barents Sea capelin 
larvae. Sarsia 85:385-391. 

Kean-Howie, J. C, S. Pearre, and L. M. Dickie. 

1988. Experimental predation by sticklebacks on 
larval mackerel and protection of fish larvae by zoo- 
plankton alternative prey. J. Exp. Mar. Biol. Ecol. 
124:239-259. 
Koster, F. W., H. H. Hinrichsen, M. A. St. John, D. Schnack, 
B. R. MacKenzie, J. Tomkiewicz, and M. Plikshs. 

2001. Developing Baltic cod recruitment models. 2. 
Incorporation of environmental variability and species 
interaction. Can. J. Fish. Aquat. Sci. 58:1534-1556. 

Koster, F. W., and C. Mollmann. 

2000. Trophodynamic control by clupeid predators on 
recruitment success in Baltic cod? ICES J. Mar. Sci 
57: 310-323. 

Lecomte, F., and J. J. Dodson. 

2004. Role of early life-history constraints and resource 
polymorphism in the segregation of sympatric populations 
of an estuarine fish. Evol. Ecol. Res. 6:631-658. 
Mills, E. L., R. O'Gorman, R. F. Roseman, C. Adams, and R. W. 
Owens. 

1995. Planktivory by alewife (Alosa pseudoharengus) 
and ranbow smelt (Osmerus mordax) on microcrustacan 
zooplankton and dreissenid (Bivalvia: Dreissenidae) 
veligers in southern Lake Ontario. Can. J. Fish. Aquat. 
Sci. 52:925-935. 

Molina, R. E., Manrique, F. A. and H. E. Velasco. 

1996. Filtering apparatus and feeding of the Pacific mack- 
erel (Scomber japonicus) in the Gulf of California. Rep. 
Cal. Coop. Ocean. Fish. Invest. 37:251-156. 

Munk, P. 

2002. Larval sand lance (Ammodytes sp.) in the diet 
of small juvenile wolffish (Anarhiclias spp.): preda- 
tory interactions in frontal water masses off western 
Greenland. Can. J. Fish. Aquat. Sci. 59:1759-1767. 

O'Driscoll, R. L., M. J. D. Parsons, and G. A. Rose. 

2001. Feeding of capelin (Mallotus villosus) in Newfound- 
land waters. Sarsia 86:165-176. 

Paradis, A. R., and P. Pepin. 

2001. Modelling changes in the frequency distributions of 



214 



Fishery Bulletin 104(2) 



fish larvae using field estimates of predator abundance 
and size distributions. Fish. Oceanogr. 10:217-234. 

Paradis, A. R., P. Pepin, and J. A. Brown. 

1996. Vulnerability to predation offish eggs and larvae: 
a review of the influence of the relative size of prey and 
predator. Can. J. Fish. Aquat. Sci. 53:1226-1235. 

Paradis, A. R., M. Pepin, and P. Pepin. 

1999. An individual-based model of the size-selective 
vulnerability offish larvae to four major predator types: 
disentangling the effect of size-dependent encounter 
rates and susceptibility to predation. Can. J. Fish. 
Aquat. Sci. 56:1562-1575. 

Pepin, P., and T. H. Shears. 

1995. The influence of body size and alternate prey abun- 
dance on the risk to predation of fish larvae. Mar. 
Ecol. Prog. Ser. 128:279-285. 

Pepin, P., J. F. Dower, J. A. Helbig, and W. C. Leggett. 

2002. Regional differences in mortality rates of a larval 
fish, the radiated shanny iUlvaria subbifurcata): estimat- 
ing the relative roles of dispersal and predation. Can. 
J. Fish. Aquat. Sci. 59:105-114. 

Pepin, P., J. F. Dower, and F. J. M. Davidson. 

2003. A spatially-explicit study of prey-predator interac- 
tions in larval fish: assessing the influence of food and 
predator abundance on growth and survival. Fish. 
Oceanogr. 12:19-33. 

Pepin, P., P. V. R. Snelgrove, and K. P. Carter. 

2005. Accuracy and precision of the continuous underway 
fish egg sampler iCUFES) and bongo nets: a comparison 
of three species of temperate fish. Fish. Oceanogr. 
14:432-447. 
Power, J. H., and E. B. Moser. 

1999. Linear model analysis of net catch data using the 
negative binomial distribution. Can. J. Fish. Aquat. 
Sci. 56:191-200. 
Rose, G. A. 

2003. Monitoring coastal northern cod: towards an opti- 
mal survey of Smith Sound, Newfoundland. ICES J. 
Mar. Sci 60:453-462. 
Skjoldal, H. R., and F. Rey. 

1989. Pelagic production and variability of the Barents 
Sea ecosystem. In Biomass yields and geography of 
large marine ecosystems (K. Sherman and L. M. Alex- 



ander, eds.l, p. 241-286. Selected Symposia Series 111, 
American Association for the Advancement of Science. 
Westview Press Inc., Boulder, CO. 

Smith, P. E., H. Santander, and J. Alheit. 

1989. Comparison of the mortality rates of Pacific sar- 
dine, Sardinops saga.x, and Peruvian anchovy, Engraulis 
ringens. eggs off Peru. Fish. Bull. 87:497-508. 

Swain, D. P., and A. F. Sinclair. 

2000. Pelagic fishes and the cod recruitment dilemma 
in the northwest Atlantic. Can. J. Fish. Aquat. Sci. 
57:1321-1325. 

Temming, A.. B. Bohle. D. W. Skagen, and F. R. Knudsen. 

2002. Gastric evacuation in mackerel: the effect of 
meal size, prey type and temperature. J. Fish Biol. 
61:50-70. 

Uotani, I. 

1985. The relation between the development of feeding 
organs and feedings modes of the anchovy. Bull. Jap. 
Soc. Sci. Fish. 51:197-204. 

Utne-Palm, A. C. 

2000. Prey visibility, activity, size and catchabilitiy's 
(evasiveness) influence on Gobisculus flavescens prey 
choice. Sarsia 85:157-165. 

Vesin, J. P., W. C. Leggett. and K. W. Able. 

1981. Feeding ecology of capelin [Mallotus villosus) in 

the estuary and western Gulf of St. Lawrence and its 

multispecies implications. Can. J. Fish. Aquat. Sci. 

38:257-267. 
Vilhjalmsson, H. 

2002. Capelin {Mallotus villosus) in the Iceland-East 

Greenland-Jan Mayen ecosystem. ICES J. Mar. Sci 

59: 870-883. 
Walters, C, and J. F. Kitchell. 

2001. Cultivation/depensation effects on juvenile sur- 
vival and recruitment: implications for the theory of 
fishing. Can. J. Fish. Aquat. Sci. 58:39-50. 

Wieland, K., and F. W. Koster. 

1996. Size and visibility of Baltic cod eggs with refer- 
ence to size-selective and stage-dependent predation 
mortality. J. Appl. Ichthyol. 12:83-89. 
Werner, F. E., J. A. Quinlan, R. G. Lough, and D. R. Lynch. 

2001. Spatially-explicit individual based modeling of 
marine populations: a review of the advances in the 
1990s. Sarsia 86:411-421. 



215 



Abstract — Horseshoe crab {Liiuuhis 
polyphemus) is harvested commer- 
cially, used by the biomedical indus- 
try, and provides food for migrating 
shorebirds, particularly in Delaware 
Bay. Recently, decreasing crab popula- 
tion trends in this region have raised 
concerns that the stock may be insuf- 
ficient to fulfill the needs of these 
diverse user groups. To assess the Del- 
aware Bay horseshoe crab population, 
we used surplus production models 
(programmed in ASPIC), which incor- 
porated data from fishery-independent 
surveys, fishery-dependent catch- 
per-unit-of-effort data, and regional 
harvest. Results showed a depleted 
population (B._,|,„3/B;^,gY=0.03-0.71 ) 
and high relative fishing mortality 
<^2002'^MSV=0-9-9-5). Future harvest 
strategies for a 15-year period were 
evaluated by using population projec- 
tions with ASPICP software. Under 
2003 harvest levels (1356 t), popula- 
tion recovery to B,,,,.,- would take at 
least four years, and four of the seven 
models predicted that the population 
would not reach Bi/sy within the 15- 
year period. Production models for 
horseshoe crab assessment provided 
management benchmarks for a spe- 
cies with limited data and no prior 
stock assessment. 



A production modeling approach to the assessment 
of the horseshoe crab iUmulus polyphemus) 
population in Delaware Bay 



Michelle L. Davis 

Department of Fisheries and Wildlife Sciences 
Virginia Polytechnic Institute and State University 
210 Cheatham Hall 
Blacksburg, Virginia 24061-0321 
Email address midavistSvtedu 



Jim Berkson 

National Marine Fisheries Service RTR Unit at Virginia Tech 
100 Cheatham Hall 
Blacksburg, Virginia 24061-0321 



Marcella Kelly 

Department of Fisheries and Wildlife Sciences 
Virginia Polytechnic Institute and State University 
210 Cheatham Hall 
Blacksburg, Virginia 24061-0321 



Manuscript submitted 25 January 2005 
to the Scientific Editor's Office. 

Manuscript approved for publication 
10 August 2005 by the Scientific Editor. 

Fish. Bull. 104:215-225 (2006). 



The horseshoe crab {Limulus poly- 
phemus) has become a source of con- 
troversy on the Atlantic coast of the 
United States (Berkson and Shuster, 
1999; Walls et al., 2002). This species 
is commercially harvested for use as 
bait, is used by the biomedical indus- 
try, and is an important source of food 
for a large number of species, includ- 
ing migrating shorebirds. However, 
population trends in the Delaware Bay 
region in recent years have indicated 
a possible decline in horseshoe crab 
abundance, raising concerns that the 
population may be unable to fulfill 
the present and future needs of these 
diverse user groups. 

Horseshoe crabs are harvested 
commercially for use as bait in the 
American eel (Anguilla rostrata), 
whelk, and conch (family Melongeni- 
dae) pot fisheries (ASMFC). Histori- 
cally, horseshoe crabs were considered 
"trash fish" of little commercial value 
and were used primarily as fertilizer 
or animal feed. When the bait fish- 
ery began, there were few restrictions 
on harvest and no harvest-reporting 
requirements. A maximum reported 
coastwide harvest of about 2 million 
crabs (3100 metric tons [t]) occurred 
in 1998 (ASMFC2). Commercial har- 



vest has decreased in recent years 
owing to the adoption of state-by- 
state quotas in 2000 (ASMFC'') and 
the increased use of bait-saving de- 
vices for the eel and conch fisheries, 
both of which have reduced the de- 
mand for crabs. 

Horseshoe crabs are also used by 
the biomedical industry. The blood 
of horseshoe crabs contains Limulus 
Amebocyte Lysate (LAL), a substance 
used to detect the presence of endo- 
toxin contamination in injectable and 
implantable drugs and devices (No- 
vitsky, 1984). The Food and Drug Ad- 
ministration estimated that 260,000 
horseshoe crabs were bled for LAL 
in 1997. After bleeding, the animals 
were released at the capture site, and 



1 ASMFC (Atlantic States Marine Fisher- 
ies Commission). 1998. Interstate fish- 
ery management plan for horseshoe crab, 
57 p. ASMFC, 1444 Eye Street, NW, 
Sixth Floor, Washington, DC 20005. 

- ASMFC. 2004. Horseshoe crab 2004 
stock assessment report, 87 p. ASMFC, 
1444 Eye Street, NW, Sixth Floor, Wash- 
ington, DC 20005. 

■' ASMFC. 2000. Addendum I to the fish- 
ery management plan for horseshoe crab, 
9 p. ASMFC, 1444 Eye Street, NW, 
Sixth Floor, Washington, DC 20005. 



216 



Fishery Bulletin 104(2) 



the mortality from the bleeding process was estimated 
to be 7.5% (Walls and Berkson, 2003). Currently, there 
is no substitute for LAL that offers comparable speed 
and sensitivity. 

Horseshoe crabs also play an important role in ma- 
rine and terrestrial food webs (Botton and Shuster, 
2003). Shorebirds migrating from South America to 
Arctic breeding grounds stop in the Delaware Bay to re- 
build depleted energy reserves (Botton et al., 1994). The 
time and place of their stop-over coincides with that of 
annual horseshoe crab breeding, when the crabs arrive 
en masse to spawn on sandy beaches during high tides 
of May and June (Botton and Harrington, 2003). An 
adult female horseshoe crab lays approximately 88,000 
eggs per year (Shuster, 1982), and a single red knot 
(Calidris canutus) can consume an estimated 18,000 
crab eggs daily (USFWS-^). 

Horseshoe crabs account for substantial economic 
value in the Delaware Bay region. Regional economic 
contribution of the eel and conch fisheries is approxi- 
mately $2.2 to $2.8 million annually. The regional eco- 
nomic value of the horseshoe crab biomedical industry 
is $26.7 to $34.9 million annually. Ecotourism related 
to migrating shorebirds has become increasingly im- 
portant to the economy of the Delaware Bay region. An 
estimated 6,000 to 10,000 recreational bird watchers 
visit the Delaware Bay in spring and contribute $6.8 
to $10.3 million to the regional economy. 

Recently, there has been concern that the horseshoe 
crab population can fulfill the current needs of these 
user groups. These concerns led the Atlantic States 
Marine Fisheries Commission (ASMFC) to develop a 
fishery management plan for horseshoe crabs. Unfortu- 
nately, very few abundance data are available for this 
species. Many state and federal trawl surveys record 
horseshoe crabs caught during sampling, but gear and 
sampling methods are not designed for horseshoe crabs 
and catches are not common. Only recently have sta- 
tistically robust horseshoe crab-specific surveys been 
initiated: a Delaware Bay spawning survey (Smith et 
al., 2002) and an offshore trawl survey (Hata and Berk- 
son, 2003). 

Because of the limited data available, previous stock 
assessments of the Delaware Bay population have been 
restricted to trend analyses to determine whether a 
single survey identifies a significant change in the pop- 
ulation or whether there is a consensus among data 
sets. However, many Delaware Bay surveys have high 
variability and low power to detect population change 
and would therefore only be able to identify dramatic 
changes in population size. Also, trend analyses do not 
provide estimates of stock status (Caddy, 1998) such as 
relative biomass (B/Syt,.y) and relative fishing mortality 
iF/fusY^- 'I'^^ ultimate goal for horseshoe crab assess- 
ment employs a stage-based catch-survey methodology 



(Collie and Sissenwine, 1983; HSC-SAS^), incorporating 
data from harvest and surveys. It will be a number of 
years before this modeling approach can be implement- 
ed, however, since stage-class data from commercial 
harvest are not currently being collected. 

The surplus production modeling approach (Prager, 
1994) used in our study is an appropriate bridge be- 
tween these two methods. Production models allow for 
the incorporation of harvest data and multiple surveys, 
improving predictive power over that of a single survey. 
This technique does not include a stage-structure in the 
model; instead it focuses on the dynamics of the popula- 
tion as a whole. Similar methods have been successfully 
applied to horseshoe crab data from Rhode Island (Gib- 
son and Olszewski'') and production models have been 
widely used for assessments of other species (Booth and 
Punt, 1998; Cadrin and Hatfield, 2002; Vaughan and 
Prager, 2002). Surplus production models assume a low 
population growth rate at small population sizes and as 
the population nears the carrying capacity (Quinn and 
Deriso, 1999). In the logistic growth form of the model 
used in the present study, maximum growth rate (and 
maximum surplus production) occurs at one-half of the 
carrying capacity. At this point, the maximum surplus 
population growth can be harvested while still main- 
taining a stable population size. Surplus production 
models provide estimates of maximum sustainable yield 
(MSY; the largest harvest that can continuously be 
removed from a stock), population biomass, and fishing 
mortality, as well as allow for the estimation of effects 
of future management. 

We fitted a regional-scale production model to Dela- 
ware Bay horseshoe crab data in order to quantify the 
current status of the Delaware Bay population and 
to estimate impacts of future management actions. 
The results from this production model will allow the 
ASMFC and member states to manage the Delaware 
Bay population of horseshoe crabs more effectively with 
the goal of providing a sustainable resource for com- 
mercial harvest, the biomedical industry, and migrating 
shorebirds. 



Methods 

Production model 

We used an age-aggregated production model with the 
Prager (1994) form of the Graham-Schaefer surplus-pro- 
duction model (i.e., logistic population growth). 



^ USFWS( U.S. Fish and Wildlife Service). 2003. Delaware 
Bay shorebird-horseshoe crab assessment report and peer 
review. Migratory Bird Publication R9-03/02, 107 p. USFWS, 
4401 N. Fairfax Dr., MBSP 4107, Arlington, VA 22203. 



^ HSC-SAS (Horseshoe crab stock assessment subcom- 
mittee). 2000. A conceptual framework for the assess- 
ment of horseshoe crab stocks in the mid-Atlantic region, 
19 p. ASMFC, 1444 Eye Street, NW, Sixth Floor, Wash- 
ington, DC 20005. 

^ Gibson, M., and S. Olszewski. 2001. Stock status of horse- 
shoe crabs in Rhode Island in 2000 with recommendations 
for management, 13 p. RI Division of Fish and Wildlife, 
4808 Tower Hill Rd., Wakefield, RI 02879. 



Davis et al Population assessment of Limulus polyphemus 



217 



3.00 



200 



1.00 



00 



rn -100 



-2 00 




MD coastal bays 

DE 16-tt trawl, 
<160-mm crabs 

DE 16-ft YOY 

DE 30-ft trawl 

NJ ocean 

NMFS spring 

NMFSfall 

DB spawning 



1 1 1 1 r 

1991 1993 1995 



1997 
Year 



1999 



2001 



2003 



Figure 1 

Fishery-independent survey indices for the Delaware Bay (DBi region from 1991 through 
2003, standardized by survey for comparison. MD = Maryland; DE = Delaware: NJ = 
New Jersey; DB = Delaware Bay. 



dB, 




rB' 


. ' 


r-F,)B, 




d, 




K 



where /• = the stock's intrinsic growth rate; 

K = the carrying capacity, both of which are 
assumed to be constant (Prager, 19941; 
and 

F, and B, = Fishing mortality and biomass, respec- 
tively, at time t. 

In addition, the harvest was not assumed to equal sur- 
plus production (i.e., the model was a dynamic or non- 
equilibrium model; Quinn and Deriso, 1999). This model 
assumes B^jy = 0.5K, where By^jy is the spawning 
biomass that would produce MSY. This form is often 
used because of its theoretical simplicity and because 
it is central among possible production model shapes. 
This production model was conditioned on catch, mean- 
ing that landings data were assumed to be more precise 
than abundance indices. By assuming that abundance 
indices are correlated measures of population abun- 
dance, the model is able to incorporate multiple indices 
by interpreting differences among indices as sampling 
error. To fit the production model, we used the ASPIC 
software (vers. 5.02) of Prager (1994), a program that 
has been used extensively in stock assessments (Cadrin 
and Hatfield, 2002; MacCall, 2002). We included data 
from fishery-independent and fishery-dependent sources 
in model runs. 

Abundance indices 

In the Delaware Bay region, there are a number of fish- 
ery-independent surveys that collect data on horseshoe 



crabs (Fig. 1). These include National Marine Fisheries 
Service (NMFS) trawl (spring, 1968-2003, and fall, 
1963-2002), New Jersey (NJ) ocean trawl (1989-2002), 
Maryland (MD) coastal bays trawl (1988-2002), Dela- 
ware (DE) 16-ft trawl (juvenile and young-of-the-year; 
1992-2002), DE 30-ft trawl (1990-2002), and Delaware 
Bay spawning survey (1999-2003). Detailed descriptions 
of these surveys can be found in ASMFC.^ We selected 
1991-2003 as the modeling timeframe because both 
harvest data and abundance index data were available 
for this period. 

In spite of the large number of surveys and long time 
series for some of these surveys, many had high vari- 
ability and low power to detect a decline. Additionally, 
many of these surveys were negatively correlated with 
each other for the years investigated (Table 1). Be- 
cause an underlying assumption for the model is that 
each survey is representative of the population being 
evaluated, total disagreement (i.e., negative correla- 
tion) among any pair of surveys cannot be reconciled 
by the model, resulting in model errors. We therefore 
used three subsets of fishery-independent surveys in 
which all pairs were positively correlated for popula- 
tion modeling (Table 2), incorporating six of the eight 
fishery-independent surveys into production model runs. 
Fishery-dependent data were also available from 1991 
to 2002 from the Delaware hand and dredge fisheries. 
Abundance indices based on catch-per-unit-of-effort 
(CPUE) for these fisheries were calculated from the 
annual number of trips and landings for each fishery 
(Fig. 2). 

Within the models, abundance indices were weighted 
by the inverse of the coefficient of variation (CV) from 
regressions, which gave more weight to surveys with less 
variability. For comparison, we also conducted model 



218 



Fishery Bulletin 104(2) 



runs where surveys were weighted equally within the 
models. Table 2 lists the three fishery-independent mod- 
els (referred to as FI) and the four fishery-dependent 



models (referred to as FD) used in production model 
applications and the CV and weighting of each survey 
within the models. 



a. 
o 



30 

25 

20- 

1 5- 

1 

05 

0.0 

-0.5- 

-1 

-1 5 



Hand harvest CPUE 
Dredge harvest CPUE 




1991 1993 1995 1997 1999 2001 

Year 

Figure 2 

Fishery-dependent abundance indices based on catch-per-unit- 
of-effort data for the Delaware hand and dredge horseshoe crab 
tLimiiliiti polyphemus) fisheries. 1991-2002. standardized by index 
for comparison. 



Harvest 

We obtained horseshoe crab harvest data for 
Virginia, Maryland, Delaware, Pennsylvania, 
and New Jersey from 1995 to 2003 (NMFS'; 
Michels*). We combined harvests among states 
for regional-scale production model runs. We 
estimated the regional harvest from 1991 to 
1994 from available Delaware landings data 
(Fig. 3). Reporting of harvest was not manda- 
tory during this time period, and harvest data 



NMFS (National Marine Fisheries Service). 
2004. Fisheries Statistics and Economics Divi- 
sion, Silver Spring, MD. Landings data (by state 
and year) obtained in August 2004 from http:// 
www.st.nmfs.gov/stl/commercial/index.html. Web 
page was last modified 24 September 2003. 
' Michels, S. Personal commun. 2004. DE Div. 
of Fish and Wildlife, 89 Kings Highway, Dover, 
DE 19901. 













Table 1 










Correlation matrix for abundance indices derived from 10 surveys, with number of pairwise 
theses for 1991-2003. Negatively correlated indices are shown in bold font. Indices used in pi 
with an asterisk (*). 


comparisons (i.e., years) in paren- 
oduction model runs are identified 














4 


5 














1 

NMFS 

fall 


2 
NMFS 
spring 


3 

DE 
30-ft 
trawl 


DE 16-ft 
trawl, 

<160-mm 
crabs 


DE 

16-ft 
trawl. 
YOY 


6 

N.J 
ocean 


7 
MD 
coast 


8 9 10 
Spawning DE DE 
survey dredge hand 


1 NMFS sail 






1.000 

(12) 
















2 NMFS spring* 






-0.409 

(12) 


1.000 
(13) 














3 DE 30-ft* trawl 






0.706 
(12) 


-0.370 

(12) 


1.000 

(12) 












4 DE 16-ft trawl, < 160- 


mm' 


crabs 


-0.261 

(11) 


0.182 
(11) 


0.215 

(11) 


1.000 

(11) 










5 DE 16-ft trawl, YOY* 






-0.153 

(10) 


0.214 
(10) 


0.341 
(10) 


0.617 
(10) 


1.000 
(10) 








6 NJ ocean* 






-0.102 

(12) 


-0.249 

(12) 


0.479 
il2) 


0.252 
(11) 


-0.132 

(10) 


1.000 
(12) 






7 MD coastal bays* 






-0.18X 

(12) 


0.086 

(12) 


-0.038 

(12) 


0.639 

(11) 


0.336 

(10) 


-0.116 

(12) 


1.000 

(12) 




8 Spawning survey 






-0.786 

(4) 


0.211 
(5) 


-0.069 

(4) 


-0.657 

(4) 


-0.781 

(3) 


0.392 

(4) 


-0.087 

(4) 


1.000 
(5) 


9 DE dredge harvest* 






0.210 

(12) 


-0.329 

(12) 


0.450 
(12) 


0.379 
(11) 


0.031 
(10) 


0.506 
(12) 


0.135 
(12) 


0.509 1.000 

(4) (12) 


10 DE hand harvest* 






-0.536 

il2i 


0.002 

(12) 


-0.214 

(12) 


0.639 

(11) 


0.084 
(10) 


0.485 

Il2i 


0.123 

il2i 


0.167 0.564 1.000 

111 112) (12) 



Davis et al Population assessment of Limulus polyphemus 



219 



from other states were unavailable or unreliable. 
We therefore expanded the Delaware harvest to 
represent the Delaware Bay region by making the 
assumption that regional landings were equal to 
ten times the landings in Delaware, the approxi- 
mate relationship from 1997 through 2003 when 
reporting was mandatory. When applicable, we 
converted numbers to metric tons (t) (Prager and 
Goodyear, 2001) using the relationship of Gibson 
and Olszewski'' (1.8182 kg/horseshoe crab). Com- 
mercial harvest of horseshoe crabs has been below 
the regional quota of 2595 t since 2000 (Fig. 3). 

Assumptions associated with 
the production models 

There were a number of general assumptions asso- 
ciated with production models (Quinn and Deriso, 
1999). We assumed that productivity (change in 
biomass over time) responded instantaneously to 
changes in population size. Changes in the biotic 
and abiotic environments were ignored, and /• (the 
intrinsic rate of population growth) and K (the carrying 
capacity) were assumed to be constant. Because produc- 
tion models combine all age classes, it was assumed that 
size or age structure of the population would not have 
major effects on population dynamics. 

Specific assumptions about population values were 
also required by the model. All starting values of 



4000 n 


Max = 3678 t 


3500  


^V y/^*~~A. 


S 3000  


/ ^'''^^ v 


to 


Quota = 2595 1 / \ 


5 2500  


. _ / \ " 


to 


/ \ 


■c 2000 - 


/ \ 


ro 


f5 'L 


o 1500  


/ ^v^,^^,^^ 


O) 


X ^tr ^^ 


S. 1000  


J 


500 . 








1991 1993 1995 1997 1999 2001 2003 


Year 


Figure 3 


Delaware Bay regional hor.seshoe crab iLimulus polyphemus) 


harvest (in t). Regional harvest, 1991-1994 (open diamonds). 


was estimated to be ten times the Delaware harvest. The current 


regional quota of 2595 t is shown by the horizontal line. 



MSY and K were based on the maximum harvest in 
the Delaware Bay from 1991 through 2003. The initial 
guess for MSY was 1850 t (half of the largest catch), 
and the initial guess for K was 37,000 t (ten times the 
largest catch). For fishery-independent model runs, 
we had the model freely estimate initial biomass in 
relation to carrying capacity (B^IK), with our start- 









Table 2 














Description 


of production model runs, includ 


ng the data sources. 


years 


, and coefficient of variation 


(CV) from regression 


analyses. The weighting of surveys within the 


model 


s is the inverse 


of the CV. The 


starting 


estimate of B,/K fo 


r each model is 


also shown 


FI refers to models that include fishery- 


independent ab 


undance indices, and FD identifies 


models 


where fisherv- 


dependent indices from CPUE data were used. 
































Relative 




Initial 


Model 


Data sources 




Years 




CV 




weighting 




Bi/K" value 


FI-1 


NMFS spring 

DE 16-ft trawl, <160-mm crabs 

DE 16-ft trawl YOY 

MD coastal bays 




1991-2003 
1992-2002 
1992-2001 
1991-2002 




0.636 
0.517 
0.928 
0.515 




0.241 
0.296 
0.165 
0.297 




0.5 


FI-2 


DE 30-ft trawl 

DE 16-ft trawl, <160-mm crabs 

DE 16-ft YOY 




1991-2002 
1992-2002 
1992-2001 




0.524 
0.517 
0.928 




0.388 
0.393 
0.219 




0.5 


FI-3 


DE 30-ft 

DE 16-ft trawl, <160-mm crabs 

NJ ocean trawl 




1991-2002 
1992-2002 
1991-2002 




0.524 
0.517 
0.345 




0.283 
0.287 
0.430 




0.5 


FD-1 


DE hand CPUE 
DE dredge CPUE 




1991-2002 
1991-2002 




0.449 
0.555 




0.553 
0.447 




Fixed at 0.1 


FD-2 


DE hand CPUE 
DE dredge CPUE 




1991-2002 
1991-2002 




0.449 
0.555 




0.553 
0.447 




Fixed at 0.2 


FD-3 


DE hand CPUE 
DE dredge CPUE 




1991-2002 
1991-2002 




0.449 
0.555 




0.553 
0.447 




Fixed at 0.3 


FD-4 


DE hand CPUE 
DE dredge CPUE 




1991-2002 
1991-2002 




0.449 
0.555 




0.553 
0.447 




Fixed at 0.4 



220 



Fishery Bulletin 104(2) 



ing estimate of this value equal to 0.5. In the absence 
of other information about starting biomass, B^IK = 
0.5 (i.e., Bj=BygY' is an appropriate default value 
(Punt, 1990; Prager, 1994). In some cases B^/K was 
poorly estimated, and we employed a common solu- 
tion of fixing the value o{ B^IK (Vaughan and Prager, 
2002). For model runs based on fishery-dependent 
indices, we fixed B^IK at 0.1, 0.2, 0.3, and 0.4 (Table 
2). The CPUE value for the Delaware hand fishery 
was relatively low in 1991; therefore we assumed that 
the population biomass was below B^jgy. ie. B^IK was 
less than 0.5. 



lo- 
g- 


 








FI-2 


 CV 




s' 

7  

J ^• 

O 

o 
U.~ 4- 


a 


a Equal 




High F 


gFI-3 


LowB 
and High 


F 


3- 
2- 

1 - 


FI-1 
C>^ FI-2 P1.3 
FI-1 °^»__^j»__£X 






Low 


B Ideal 


0. 


M 0.25 0.50 75 100 125 


^2003/6^5^ 


Figure 4 


Relative 2003 biomass (B2003''5msy> and relative 2002 fishing 
mortality (^2002^^;MSy' of Delaware Bay horseshoe crab iLimu- 


lus polyphemus) for each of the seven model runs presented. 


with abundance indices weighted equally or inversely to CV 


within the models. 



Quantities estimated by the model 

The production model estimated several benchmarks 
and status indicators useful in understanding horse- 
shoe crab biology and in improving management. These 
quantities included relative biomass (B/Sj^jgyl. relative 
fishing mortality (F/Fy^^y). population biomass (B), and 
maximum sustainable yield (MSY). Point estimates and 
80% confidence intervals for each of these quantities 
were calculated for each model run. 



Population projections 

We used the Delaware Bay population estimates 
calculated by the surplus production model to 
project the population forward in time for a period 
of 15 years to evaluate potential harvest levels. 
We conducted projections using ASPICP (Prager, 
1994), with annual landings specified for each 
year of the projections. We selected a 15-year 
time period because this is the longest projection 
period that can be programed in ASPICP, and 
confidence in projection model results decreases 
at longer time intervals. We evaluated trends in 
biomass over time for a range of harvest levels, 
including harvest at 2003 levels and proportional 
reductions of that harvest. These harvests were 
0% of 2003 catch (no harvest), 25% (339 t annu- 
ally), 50%' (678 t), 75% (1017 t), and 100% (1356 
t). We identified the number of years under each 
harvest scenario required for the population (and 
80% confidence intervals) to rebuild to B^/^y. We 
also compared relative biomass (with 80%5 confi- 
dence intervals) in the final year of projections 
(2018) for each harvest level. 



I.b  










B/Bmsy 






1.4 - 


FI-1 




1.2  


. - -  FI-3 




10 
08 
06 








— • 


— 




/ 




04 . 


X 




^ 
^ 




02 - 


T77T — ~— -~— ___^ 




_ 


n n - 










 - - - -■-.»-,_ 



1991 



1993 



1997 

Year 



Figure 5 

Production model estimates of relative biomass (B/B/^igy'i of horseshoe 
crabs (Limulus polyphemus) in the Delaware Bay region, 1991-2004. 
Results from fishery-independent model runs are shown, and the hori- 
zontal line represents B/B^,^y=l. 80% confidence intervals each model 
run are the same line pattern in gray. 



Results 

Results differed little between model sim- 
ulations for surveys weighted inversely 
to CV and for simulations with equally 
weighted surveys (Fig. 4). 

Production model runs showed that 
BIB^i^Y in the Delaware Bay region 
increased in the early 1990s and has 
declined steadily since 1995 (Figs. 5 
and 6). Slight increases since 2001 were 
evident in some model runs. Relative 
biomass in 2003 was estimated to be 
low, with point estimates ranging from 
0.03 to 0.20 for models with fishery- 
independent data and 0.20 to 0.71 for 
models with fishery-dependent data 
(Table 3). Eighty-percent confidence 
intervals for B^oos/B^gy ranged from 
0.005 to 1.16. 



Davis et a\ Population assessment of Limulus polyphemus 



221 













Table 3 












Status indicators and management benchmarks for horseshoe crabs in the Delaware Bay region, estimated from production 
model runs with fishery-independent iFIi or fishery-dependent (FDi indices weighted by the inverse of the CV. Point estimates 
and lower (L) and upper (U) 80"* confidence intervals (CIs) are shown for each model run. The objective function is a measure of 
how well the model was able to fit the data using a lognormal error structure — a lower value representing a better fit. 




FI-1 




FI-2 




FI-3 




FD-1 


FD-2 


FD-3 


FD-4 


"imrx^MSY 


0.232 




0.022 




0.030 




0.201 


0.427 


0.588 


0.710 


80'* CKL) 


0.109 




0.005 




0.008 




0.127 


0.238 


0.319 


0.388 


80^f CKU) 


0.479 




0.154 




0.041 




0.347 


0.762 


1.015 


1.157 


■'^2002''^.MSi' 


2.156 




9.501 




6.560 




1.795 


1.269 


1.034 


0.911 


BO'/rCKL) 


1.442 




6.005 




3.999 




1.258 


0.770 


0.588 


0.513 


80% CKU) 


3.299 




18.320 




21.300 




2.592 


2.069 


2.112 


2.184 


-S2003 <t) 


2197 




1084 




4098 




2579 


3653 


5035 


6604 


80% CI (L) 


1319 




524 




2295 




1423 


1853 


2955 


4340 


80% CI (U) 


4603 




4306 




9379 




4372 


5012 


7778 


12,080 


MSYit) 


3064 




5082 




7220 




3768 


2628 


2354 


2196 


80% CI (L) 


2550 




2985 




5340 




3274 


2566 


1905 


1372 


80%CI(U) 


4475 




8713 




9577 




4499 


2787 


2505 


2501 


Objective function 


16.05 




12.40 




9.57 




2.62 


2.93 


3.39 


3.85 



F/Fmsy 

Although harvest of horseshoe crabs has 
decreased in recent years (Fig. 3), fishing 
mortality remains high (Figs. 7 and 8). F.,„„.,/ 
P\isY point estimates ranged from 2.3 to 9.5 
for models with fishery-independent data and 
from 0.9 to 1.8 for models with fishery-depen- 
dent data (Table 3). 



Biomass 

Biomass of horseshoe crabs in the Delaware 
Bay region has decreased substantially since 
1995, such that the 2003 biomass was less 
than 56% of the 1995 biomass. This equates 
to an annual decline of greater than 7% 
during this period. Point estimates for 2003 
biomass ranged from 1084 t (596,000 crabs) 
to 6604 t (3,632,000 crabs). As is charac- 
teristic of production models (Prager, 1994), 
absolute biomass was estimated much less 
precisely than relative biomass (B/By^jj). The 
range of 80% confidence intervals for 2003 
biomass across all seven model runs was 524 t (288,000 
crabs) to 12,080 t (6,644,000 crabs). 

Population projections 

We used results from production model runs to project 
the horseshoe crab population forward in time to evalu- 
ate potential management options. Figure 9 shows the 
trajectory of B/B^,^y over time for model FI-1 under 
each harvest scenario. The number of years required 











1.6 T 


S/Smsj, 


FD-1 




1 4- 


\ 


— - - FD-2 




1 2- 

10- 

Cq" 08  

CD 

06- 

04- 


y^ y"^ v;>^^ 


FD-3 

._ pQ ^ 












02- 


„•''* **------ 






1991 1993 1995 1997 1999 2001 2003 


Year 


Figure 6 


Production model estimates of relative biomass iS/B,,,;..,.) of horseshoe 


crabs {Limulus polyphemus) in the Delaware Bay region, 1991-2003. 


Results from fishery-dependent model runs are shown, and the 


horizontal line represents B/Bn,gY=l- 80% confidence intervals for 


models FD-1 and FD-4 are shown in gray. 



to rebuild the population to B„sy varied substantially 
among models (Table 4). At 2003 harvest levels (i.e., 
100%), projections showed population recovery in a mini- 
mum of four years, although four of the seven models 
did not reach B^^.y in the 15-year projection period. In 
the absence of harvest (i.e., 0%), recovery could occur in 
as few as 2 years, but two models did not reach B^^^y in 
the projection period. Estimates of B/Bj^g^y in the final 
year of projections (2018) also differed among model 
applications (Table 5). 



222 



Fishery Bulletin 104(2) 



Discussion 

According to surplus production model runs and corre- 
sponding projections, the horseshoe crab population in 
the Delaware Bay region has been depleted and current 
harvest levels may be too high to allow the population to 
rebuild to S^gy within 15 years. Biomass in this region 
has decreased steadily since 1995 and is currently well 



below B 



A/.sy 



This decline was evident in models that 



incorporated regional fishery-independent surveys and 
Delaware fishery-dependent indices. Figure 4 shows 
current relative biomass and relative fishing mortality 
for each of the model applications. All model runs indi- 
cated a depleted population with high fishing mortality, 



10 -. 
9 






F/F„,5, 




FI-1 


/ 


8 






/ 


7 . 




- - - FI-3 


/ 






-. // 


s":: 


.-..-''^-^'^^^^■' 


3- 


/""^ 


2 . 

1 . 
 




-^>: - -'--''''"^"^--'■- 










~' — 






1991 1993 1995 199?' 1999' ' 2001 ' 2003' 1 


Year 


Figure 7 


Production model estimates of relative fishing mortality (F/F,,,,-) of 


horseshoe crabs (Limulus polyphemus) in the Delaware Bay region. 


1991-2003. Results from fishery-independent model runs are shown. 


and the horizontal line represents F/F„j,.y=l. SOVr confidence inter- 


vals for each model run are the same line pattern in gray. 



45 
40 
3.5 
3.0 
25 
2.0 
15 
10 
05 
0.0 



F/F, 




1993 



1995 



1997 



1999 



Year 



Figure 8 

Production model estimates of relative fishing mortality ^F/F^|^y) of 
horseshoe crabs iLimulus polyphemus) in the Delaware Bay region, 
1991-2002. Results from fishery-dependent model runs are shown, 
and the horizontal line represents FIFmsy-^- ^0% confidence inter- 
vals for models FD-1 and FD-4 are shown in gray. 



although the estimated extent differed among models 
(Fig. 4). Projections for some of the model runs predicted 
a relatively fast population recovery. In the absence of 
harvest, five of the seven model applications predicted 
rebuilding to 5^/.,y in five or fewer years. However, two 
of the model runs estimated such low population biomass 
that rebuilding the Delaware Bay population to B^,^y 
would take greater than 15 years, even with no harvest. 
A precautionary management strategy may therefore be 
appropriate for this population in the short term as more 
data are being collected. 

Population model results may have been affected by 
assumptions that we made during the modeling pro- 
cess. In order to extend the time period of the model 
back to 1991, we had to estimate regional 
harvest based on Delaware harvest data for 
the early years of the model. Because har- 
vest data from most states were not avail- 
able prior to 1995, we assumed that regional 
harvest was equal to ten times the Dela- 
ware harvest for the years 1991-94. This 
was the approximate relationship between 
Delaware and regional harvest from 1997 
through 2003 when reporting was manda- 
tory. Actual regional harvest for 1991-94 
was unknown because there were no har- 
vest-reporting requirements. We conducted 
sensitivity analyses which showed very little 
difference among runs with 1991-94 harvest 
equal to Delaware landings, 10 times Dela- 
ware landings, or 20 times the Delaware 
landings (the mean difference for Bo(fQ-^IB^,^y 
from these runs was 0,011). If the actual 
harvest was substantially greater than 20 
times Delaware landings, differences in re- 
sults may occur. 

Results were also influenced by the data 
sources that were included. We based the 
inclusion of fishery-independent surveys on 
positive correlations among surveys, incor- 
porating the largest number of data sources 
possible into model runs. However, two fish- 
ery-independent surveys were not included: 
the NMFS fall trawl and the Delaware Bay 
spawning survey. The NMFS fall trawl was 
negatively correlated with seven of the nine 
other data sources (Table 1), and we there- 
fore assumed that it was not a reliable index 
of horseshoe crab abundance. The spawning 
survey was negatively correlated with five of 
the nine other data sources (Table 1), This 
survey also had a very short time series be- 
cause it was redesigned in 1999 to improve 
statistical power (Smith et al,, 2002). With so 
few data points, the production model would 
be unable to distinguish population trends 
from survey variability; therefore spawning 
survey data were excluded. However, as one 
of only two horseshoe crab-specific surveys 
currently in place, the Delaware Bay spawn- 



Davis et aL: Population assessment of Limulus polyphemus 



223 



ing survey will likely prove to be a valuable 
source of information in the future. 

Negative correlations were present among 
a relatively large number of Delaware Bay 
surveys (Table 1) which are assumed to be 
sampling the same population. This is a 
common problem in fisheries stock assess- 
ments (Richards, 1991; Schnute and Hilborn, 
1993). The observed differences among these 
surveys could be attributed to a number of 
factors. Catches of horseshoe crabs are not 
common, leading to small sample sizes and 
high variability. Also, these surveys may 
differ in location, time of year, and gear 
selectivity. Future studies could employ an 
analysis of variance (ANOVA) to attempt 
to separate these factors from underlying 
horseshoe crab abundance trends. 

The data sources included in individual 
model runs also led to differing results. 
Models incorporating fishery-dependent data 
tended to present a slightly more optimis- 
tic view of the population and the fishery. 
predicting higher relative biomass and lower relative 
fishing mortality. Although harvest data often have the 
benefit of having been derived from large sample sizes 
(and have resulting low variance estimates), there is 
often a bias associated with fishery-dependent data. 
Fisheries do not sample randomly because they target 
areas of highest abundance, and thus biased indices 
are produced (Quinn and Deriso, 1999). In addition, 
the use of fishery-dependent abundance indices is often 
complicated by changes in gear, regulations, or sam- 
pling methods over time, any of which could affect catch 
rates. Fishery-independent surveys are usually more 



2 °°1 Projected B/B^sy (model FI-1) 

1 75- 
1 50 
1 25- 



CO 1 00 

HI 

75- 

50 
25- 
00 




2004 



2006 



2008 



2010 2012 

Year 



2014 



Figure 9 

Example of projection results. Projected relative biomass iB/B^/^y) 
in shown for model FI-1, with each line representing a harvest 
level applied annually in the 15-year projections. The percentage 
refers to the percent of the 2003 Delaware Bay regional landings 
of 1356 t (i.e., 50% = 678 t). 



appropriate for assessments, assuming there is high 
consistency in sampling methods among years (Chen 
et al., 2003). 

Differences also existed among fishery-independent 
surveys. Models FI-2 and FI-3 predicted a much low- 
er population size than FI-1 or the fishery-dependent 
models. Projections with these runs predicted that the 
population was too depleted to recover in 15 years, even 
in the absence of harvest. The models differed only in 
the abundance indices included. In the trend analyses 
conducted for ASMFC,- the most significant population 
declines since 1997 were identified in the NJ ocean 



Table 4 

Results of Delaware Bay population projections from production model runs from fishery-independent (FI) and fishery- 
dependent (FDl indices. Projections were conducted for 15 years, with a constant harvest (in t) applied annually. Harvest levels 
were based on 2003 harvest and are listed in the left column. The time period (and 80'7f confidence intervals) shown represent 
the number of years (starting in 2003) required for the population biomass to reach B,,^.).. "n/a" indicates that the biomass did 
not reach S.i/sy during the 15-year projection period. 


Harvest level 
in relation 
to that of 2003 








Years to rebuild tc 


Bms\ 








FI-1 


FI-2 


FI-3 


FD-1 




FD-2 


FD-3 


FD-4 


QVc (Ot) 
SOTc CI 




5 years 
(2, 12) 


n/a 
(3, n/a) 


n/a 
( n/a, n/a) 


4 years 
(3,7) 




3 years 
(1,6) 


2 years 
(0,8) 


2 years 
(0, 15) 


25'7r(339t) 
80% CI 




5 
(2, n/a) 


n/a 
(3, n/a) 


n/a 
( n/a, n/a) 


5 
(3,8) 




3 

(1,8) 


2 
(0,9) 


2 
(0, n/a) 


50'7f (678 t) 
80% CI 




6 

(2, n/a) 


n/a 
(n/a, n/a) 


n/a 
(8, n/a) 


6 

(3, 11) 




4 
(1, 10) 


3 

(0, 14) 


2 

(0, n/a) 


75% (1017 t) 
80% CI 




9 

(3, n/ai 


n/a 
(n/a, n/a) 


n/a 
(5. n/a) 


7 
(4, 15) 




4 
(1, n/a) 


3 

(0, n/a) 


3 
(0, n/a) 


100% (1356 t) 
80% CI 




n/a 
(3, n/a) 


n/a 

(n/a. n/a) 


n/a 

( n/a, n/a) 


n/a 
(5, n/a) 




6 

(2, n/a) 


5 

(0, n/a) 


4 
(O.n/a) 



224 



Fishery Bulletin 104(2) 













Table 5 












Results of Delaware Bay population projections from production model runs from fishery-independent (FI) and fishery- 
dependent ( FD 1 indices. Projections were conducted for 15 years, with a constant harvest (in t) applied annually. Harvest levels 
were based on 2003 harvest and are listed in the left column. The relative biomass (B/B„gy) in the final year of projections 
(2018J and 80^r confidence intervals are shown for each model simulation. 


Harvest level 
in relation 
to that of 2003 












^2018''^MSy 










FM 


FI-2 






FL3 


FD-1 


FD-2 


FD-3 




FD-4 


0%(Oti 


2.00 


0.41 






0.14 


2.00 


2.00 


2.00 




2.00 


80'7f CKLl 


1.49 


o.oa 






0.04 


1.91 


1.87 


1.63 




1.05 


80'7f CKU) 


2.00 


1.97 






0.28 


2.00 


2.00 


2.00 




2.00 


25<7f (339t) 


1.94 


0.00 






0.06 


1.95 


1.93 


1.92 




1.91 


SO'/f CKL) 


0.00 


0.00 






0.00 


1.95 


1.90 


1.42 




0.86 


80<7, CKU) 


1.96 


1.94 






0.42 


1.95 


1.93 


1.93 




1.93 


50% (678 t) 


1.87 


0.00 






0.00 


1.90 


1.86 


1.84 




1.82 


80% CKL) 


0.00 


0.00 






0.00 


1.89 


1.80 


1.07 




0.61 


80% CKU) 


1.93 


0.00 






1.76 


1.90 


1.86 


1.86 




1.86 


75% (1017 t) 


1.78 


0.00 






0.00 


1.83 


1.78 


1.74 




1.71 


80% CI (L) 


0.00 


0.00 






0.00 


1.66 


0.86 


0..57 




0.26 


80% CKU) 


1.85 


0.00 






1.74 


1.85 


1.78 


1.78 




1.77 


100% (13561) 


0.00 


0.00 






0.00 


0.76 


1.67 


1.62 




1.58 


80% CKL) 


0.00 


0.00 






0.00 


0.00 


0.00 


0.00 




0.00 


80% CKU) 


1.72 


0.00 






0.00 


1.77 


1.69 


1.69 




1.68 



trawl and the DE 16-ft 160-mm survey, both of which 
were included in model FI-3, and possibly explain the 
low population estimates. 

In other trend analyses conducted during the previous 
assessment, a population decline in the Delaware Bay 
was less evident. Using data from 1997 through 2003, 
we found that only four of eight fishery-independent 
surveys showed a significant decline, partially owing 
to high variability and low power. By incorporating a 
number of these surveys into a production model, we 
also found that the decreasing biomass in recent years 
becomes more apparent. These production model runs 
provide the added benefit of estimating stock status 
and management benchmarks, as well as the benefit of 
evaluating future management options. 

Although interpretation of absolute biomass or popula- 
tion size from a surplus production model can be some- 
what problematic, our estimates are roughly comparable 
to estimates from previous studies. Hata and Berkson 
(2003) calculated a mean 2001 population size of 4.4 
million adults (95% confidence intervals of 2.1 million 
and 6.8 million) in the Delaware Bay from daytime 
trawl survey data. Botton and Ropes (1987) estimated 
2.3 to 4.5 million adults in this region. In the present 
study, our estimates of the 2003 population size ranged 
from 0.6 million to 3.6 million crabs, and the mean of 
the seven model runs equaled 2.0 million crabs. Eighty 
percent confidence intervals ranged from 0.3 million to 
6.6 million crabs for all model applications. Although 
within the range of results from the other studies, these 



wide confidence intervals provide little information for 
management. It will therefore be more appropriate to 
interpret relative biomass (B/B„j;j.) and relative fishing 
mortality (i^/.F^/.sj) for use in management decisions 
(Prager, 1994). 

It is important to understand the spatial scale and 
population represented by these models and analyses. 
This regional model is a compilation of a number of 
localized fishery-independent surveys, most of which 
encompassed a relatively small spatial area. However, 
the model results should not be interpreted at a more 
localized scale because landings data are combined for 
the region. Similarly, interpretation of results should 
not be expanded to represent other Atlantic horseshoe 
crab populations outside the Delaware Bay region be- 
cause neither survey nor harvest data in this model 
extend to other regions. Nevertheless, the Delaware Bay 
is believed to be the center of abundance and spawning 
activity for Atlantic horseshoe crabs; therefore popula- 
tion trends in this region may have significant implica- 
tions for adjacent populations. 

In analyses conducted for ASMFC,- trend analyses 
identified dramatic regional differences in horseshoe 
crab population trends. Although the Delaware Bay, 
eastern Long Island Sound, and New England popula- 
tions have experienced declines in recent years, the 
southeast population and the western Long Island popu- 
lation have remained stable or have increased. Future 
production models applied to these other regions will 
hopefully clarify these trends and allow managers to 



Davis et al Population assessment of Limulus polyphemus 



225 



determine regional harvest regulations. By identify- 
ing appropriate management for each region, we will 
improve our ability to rebuild the Atlantic horseshoe 
crab population and provide a sustainable resource for 
the diverse user groups. 



Acknowledgments 

This research was funded by the National Marine Fish- 
eries Service. We thank M. Prager and M. Gibson for 
providing modeling advice, D. Smith, S. Michels, and 
B. Andres for providing horseshoe crab data and guid- 
ance, and the members of the Horseshoe Crab Stock 
Assessment and Technical Committees of the ASMFC 
for their contribution to this study. This manuscript 
was improved by the comments and suggestions from M. 
Prager, M. Millard, S. Michels, B. Murphy, D. Hata, and 
S. Klopfer, and two anonymous reviewers. We greatly 
appreciate the time and effort of all involved. 



Literature cited 

Berkson, J., and C. N. Shuster. 

1999. The horseshoe crab: the battle for a true multiple- 
use resource. Fisheries 24:6-10. 
Booth, A. J., and A. E. Punt. 

1998. Evidence for rebuilding in the panga stock on the 
Agulhas Bank. South Africa. Fish. Res. 34:103-121. 
Botton, M. L., and B. A. Harrington. 

2003. Synchronies in migration: shorebirds, horseshoe 
crabs, and Delaware Bay. In The American horseshoe 
crab (C. N. Shuster, R. B. Barlow, and H. J. Brock- 
man, eds.), p. 5-26. Harvard Univ. Press, Cambridge, 
MA. 
Botton, M. L., R. E. Loveland, and T. R. Jacobsen. 

1994. Site selection by migratory shorebirds in Dela- 
ware Bay, and its relationship to beach characteristics 
and abundance of horseshoe crab {Limulus polyphemus) 
eggs. Auk 111:605-616. 
Botton, M. L., and J. W. Ropes. 

1987. Populations of horseshoe crabs, Limulus polyphemus. 
on the northwestern Atlantic continental shelf. Fish. 
Bull. 85:805-812. 
Botton, M. L., and C. N. Shuster. 

2003. Horseshoe crabs in a food web: who eats whom? In 
The American horseshoe crab (C. N. Shuster, R. B. 
Barlow, and H. J. Brockman, eds.), p. 33-153. Har- 
vard Univ. Press. Cambridge, MA. 
Caddy, J. 

1998. A short review of precautionary reference points and 
some proposals for their use in data-poor situations. FAO 
Fish. Tech. Pap. 379, 29 p. FAO, Rome. 
Cadrin, S. X., and E. M. C. Hatfield. 

2002. Relative biomass and production of longfin inshore 
squid, Loligo pealeii. Bull. Mar. Sci. 71:1115-1116. 



Chen, Y.. L. Chen, and K. I. Stergiou. 

2003. Impacts of data quantity on fisheries stock 
assessment. Aquat. Sci. 65:92-98. 
Collie, J. S., and M. P. Sissenwine. 

1983. Estimating population size from relative abundance 
data measured with error. Can. J. Fish. Aquat. Sci. 
40:1871-1879. 

Hata, D., and J. Berkson. 

2003. Abundance of horseshoe crabs iLimulus polyphe- 
mus) in the Delaware Bay area. Fish. Bull. 101:933-938. 
MacCallA. D. 

2002. Use of known-biomass production models to de- 
termine productivity of west coast groundfish stocks. 
N. Am. J. Fish. Manag. 22:272-279. 
Novitsky, T. J. 

1984. Discovery to commercialization: the blood of the 
horseshoe crab. Oceanus 27:13-18. 

Prager, M. H. 

1994. A suite of e.xtensions to a nonequilibrium surplus- 
production model. Fish. Bull. 92:374-389. 
Prager, M. H., and C. P. Goodyear. 

2001. Effects of mixed-metric data on production model 
estimation: simulation study of a blue marlin-like 
stock. Trans. Am. Fish. Soc. 130:927-939. 

Punt. A. E. 

1990. Is B] = K an appropriate assumption when apply- 
ing an observation error production-model estimator to 
catch-effort data? S. Afr. J. Mar. Sci 9:249-259. 

Quinn, T. J., and R. B. Deriso. 

1999. Quantitative fish dynamics, 542 p. Oxford Univ. 
Press, New York, NY. 
Richards, L. J. 

1991. Use of contradictory data sources in stock 
assessments. Fish. Res. 11:225-238. 

Schnute, J. T, and R. Hilborn. 

1993. Analysis of contradictory data sources in fish stock 
assessment. Can. J. Fish. Aquat. Sci. 50:1916-1923. 
Shuster, C. N. 

1982. A pictorial review of the natural history and ecology 
of the horseshoe crab, Limulus polyphemus, with refer- 
ence to other Limulidae. In Physiology and biology of 
horseshoe crabs: studies on normal and environmentally 
stressed animals (J. Bonaventura, ed.), p. 1-52. Alan 
R. Liss, Inc. New York. 
Smith, D. R., P. S. Pooler, B. L. Swan, S. F Michels, W. R. Hall. 
P. J. Himchak. and M. J. Millard. 

2002. Spatial and temporal distribution of horseshoe 
crab {Limulus polyphemus) spawning in Delaware Bay: 
implications for monitoring. Estuaries 25:115-125. 

Vaughan. D. S., and M. H. Prager. 

2002. Severe decline in abundance of the red porgy 
{Pagrus pagrus) population off the southeastern 
United States. Fish. Bull. 100:351-375. 

Walls, E. A., and J. Berkson. 

2003. Effects of blood extraction on horseshoe crabs 
{Limulus polyphemus). Fish. Bull. 101:457-459. 

Walls, E. A., J. Berkson. and S. A. Smith. 

2002. The horseshoe crab, Limulus polyphemus: 200 
million years of existence, 100 years of study. Rev. 
Fish. Sci. 10:39-73. 



226 



Abstract — Data collected from an 
annual groundfish survey of the 
eastern Bering Sea shelf from 1975 
to 2002 were used to estimate bio- 
mass and biodiversity indexes for two 
fish guilds: flatfish and roundfish. 
Biomass estimates indicated that 
several species of flatfish (particu- 
larly rock sole, arrowtooth flounder, 
and flathead sole), several large 
sculpins {Myoxocephalus spp.), big- 
mouth [Hemitripterus bolini), and 
skates iBathyraja spp.) had increased. 
Declining species included several 
flatfish species and many smaller 
roundfish species of sculpins, eel- 
pouts (Lycodes spp.), and sablefish 
(Anoplopoma fimbria). Biodiversity 
indexes were calculated by using bio- 
mass estimates for both guilds from 
1975 through 2002 within three physi- 
cal domains on the eastern Bering 
Sea shelf. Biodiversity trends were 
found to be generally declining within 
the roundfish guild and generally 
increasing within the flatfish guild 
and varied between inner, middle, and 
outer shelf domains. The trends in 
biodiversity indexes from this study 
correlated strongly with the regime 
shift reported for the late 1970s and 
1980s. 



Biodiversity as an index of regime shift 
in the eastern Bering Sea 



Gerald R. Hoff 

Alaska Fisheries Science Center 
National Marine Fisheries Service 
7600 Sand Point Way N E 
Seattle, Washington 98115 
Email address; jerry.hofftSnoaagov 



Manuscript submitted 25 March 2004 
to the Scientific Editor's Office. 

Manuscript approved for publication 
13 August 2005 by the Scientific Editor. 

Fish. Bull. 104:226-237 (2006). 



Environmental and biological events in 
the North Pacific can indicate regime 
shifts or reorganizations of the eco- 
system at the environmental and bio- 
logical level (Hare and Mantua, 2000). 
Measurable climatic events were iden- 
tified in the mid-1970s, late 1980s 
(Anderson and Piatt, 1999; Anderson, 
2000; Hare and Mantua, 2000; Zhang 
et al., 2000; Minobe, 2002) and the 
late 1990s (Hunt and Stabeno, 2002; 
Minobe, 2002). Regime shifts in the 
eastern Bering Sea (EBS) have been 
correlated with several climatic events, 
including the Pacific Decadal Oscilla- 
tion, the El Niiio Southern Oscilla- 
tion (Hollowed et al., 2001), sea ice 
coverage (Stabeno et al., 2001; Hunt 
et al., 2002; Hunt and Stabeno, 2002), 
and summer sea surface temperatures 
(Bond and Adams, 2002; Hunt et al., 
2002; Minobe, 2002). The far reach- 
ing effects that climate change has 
on an ecosystem are not well defined, 
but many studies have shown strong 
correlations between climate change, 
fish recruitment, and plankton pro- 
duction in the North Pacific (Brodeur 
and Ware, 1992; Anderson et al., 1997; 
Anderson and Piatt, 1999; Brodeur 
et al., 1999; Clark et. al., 1999; Hare 
and Mantua, 2000; Zhang et al., 2000; 
Hollowed et al., 2001; Sugimoto et al., 
2001; Conners et al., 2002; Hunt et 
al., 2002; Wilderbuer et al., 2002). 
During periods of climatic change, 
some fishes may not be well adapted 
to dramatic changes over a short time 
scale (1-10 years), whereas others 
may proliferate in a more hospitable 
environment. Key triggers of regime 
shifts and the extent to which species 
will respond remain unclear; however, 
evidence may indicate that species 



diversity is correlated with primary 
production in many systems, and the 
two may be interdependent (Rosenz- 
weig and Abramsky, 1993; Waide et 
al., 1999). A biodiversity index can 
be useful for monitoring the stability, 
health, and productivity of an ecosys- 
tem, as well as for aiding manage- 
ment, by tracking exogenous changes 
and their far reaching effects on spe- 
cies. In the present study, I used bio- 
diversity indexes, reflecting richness 
and evenness, as indicators of species 
composition changes and related these 
changes to regime shift events for the 
eastern Bering Sea (EBS) shelf. 



Methods 

Data were synthesized from a con- 
tinuous 24-year period of standard- 
ized groundfish surveys conducted by 
the Alaska Fisheries Science Center 
(AFSC) on the EBS shelf from 1979 
through 2002; additional estimates 
from 1975 were included. Biodiversity 
indexes (species richness and even- 
ness) were calculated from biomass 
estimates for two fish guilds, flatfish 
and roundfish, in each of the inner, 
middle, and outer domains of the EBS 
shelf. Biodiversity indexes were plot- 
ted for each fish group and domains 
and examined for changes over the 
study period. The observed changes 
were correlated with regime shift and 
ocean climatic change events for the 
EBS. 

Data collection 

Gear, station location, sampling pro- 
cedures, and time of year of the AFSC 



Hoff Biodiversity as an index of regime shift in tlie eastern Bering Sea 



227 



Survey area in the 
stations from 20 
outer domains. 



eastern Bering Sea continental 
shelf survey have been standard- 
ized since 1982. Prior to 1982, a 
400-mesh eastern trawl (smaller 
than the current 83-112 trawl) was 
used (Bakkala, 1993). Although 
the biomass estimates produced 
by the two trawls are not directly 
comparable because of unknown 
catchability differences between 
the nets, survey data from 1975 
and 1979-81 were included in the 
present analysis because of the 
importance in the timing of the 
1970s regime shift. The biodi- 
versity indexes are based on the 
relative proportion of the species 
and therefore are an indicator of 
assemblage structure. 

Data used for this study were 
collected by the Resource Assess- 
ment and Conservation Engineer- 
ing (RACE) division of the AFSC, 
which surveys the EBS shelf each 
summer (May- August). The sur- 
vey area extends from the Alaska 
Peninsula north to Nunivak Is- 
land and St. Matthew Island, and 
west to the 200 m shelf break. 
Trawl hauls were conducted on a 

grid of 356 fixed stations (20 nmi by 20 nmi) (Fig. 1) 
during daylight hours. Hauls were towed for 30 min at 
3 knots and ranged in depth from 15 m in Bristol Bay 
to nearly 200 m near the shelf edge. Most trawling 
was conducted with the AFSC 83-112 eastern trawl 
(1982-2002), which is a low-opening two-seam trawl 
with a 26.5-m headrope and 34.1-m cable footrope 
wrapped with rubber striping and chain hangings that 
contact the bottom while the trawl is towed (Rose and 
Walters, 1990). 

Height and width of the net were measured with an 
acoustic SCANMAR (Scanmar, Asgardstrand, Norway) 
or NETMIND (Northstar Technical Inc., St. John's, NF, 
Canada) net mensuration system, or estimated by us- 
ing a function that relates trawl widths to tow depths 
from measured hauls. Each haul was measured with 
GPS or LORAN to record latitude and longitude data 
at the start and end of the trawl in order to determine 
distance fished. 

Processing of the catch was done entirely in the field 
at the time of capture. The entire catch was sorted to 
species, enumerated, and weighed, or a weighed sub- 
sample was used for very large catches. 

Catch per unit of area (CPUE) was determined for 
each trawl station completed during each survey. Bio- 
mass estimates for each species were then calculated 
by expanding the average CPUE for each stratum and 
then summing over all strata of the total survey area 
to obtain a biomass estimate in metric tons for each 
species. 




Figure 1 

eastern Bering Sea. Crosses represent annual trawl survey 
200 m. Bathymetry lines delineate the inner, middle and 



EBS shelf domains and study area 

The EBS is composed of three well-defined regions des- 
ignated as the inner, middle, and outer domains from 
east to west, respectively, across the shelf (Fig. 1). The 
three domains are distinct regions characterized by 
depth, water temperature, current flow, summertime 
primary production, and species composition (Bakkala, 
1993; Schumacher and Stabeno, 1998; Stabeno et al., 
2001). Briefly, the inner domain is a relatively shal- 
low (<50 m) and well-mixed warm basin with strong 
influences from a large coastal area, several major 
river systems, and a current flow northward along the 
coast from the Aleutian Islands. The middle domain is 
a relatively stagnant area of deeper water (50-100 m) 
that has strong summer water-column stratification and 
low current flows. The bottom water mass is relatively 
cold, overlaid by a layer of warmer wind-mixed water. 
Although characterized as a cold region, the water 
mass in the middle domain varies from year to year. 
The outer domain is influenced by the EBS slope region 
and by upwelling and northward current flow from the 
Aleutian Islands. This domain is relatively warm when 
compared to the middle domain (see Hunt et al., 2002, 
for review). For the purposes of this study the inner 
domain designation consisted of trawl stations less 
than 50 m in depth; the middle shelf designation was 
for trawl stations between 51 and 100 m in depth; and 
outer domain designation was for trawl stations between 
101 and 200 m in depth. 



228 



Fishery Bulletin 104(2) 







Table 1 


Flatfish guild species or species groups used for biodiversity indexes. * indicates the major proportion of the biomass for the 
group. Superscripts indicate the primary domains where each species occurs during the summertime AFSC survey. I=inner 
domain, M=middle domain, 0=outer domain. 


Group name 




Scientific name 


Pacific halibut 




Hippoglossus stenolepis"^'° 


Flathead/Bering flounder 




*Hippoglossoides elassodon and H. robustus'^"^ 


Greenland turbot 




Reinhardtius hippoglossoides'^ 


Arrowtooth/Kamchatka flounders 




'■Atheresthes stomias and A. evermaiiti'^"^ 


Starry flounder 




Platichthys stellatus'^' 


Alaska plaice 




Pleuronectes quadrituberculatus'^' 


Northern/southern rock sole 




*Lepidopsetta polyxystra and L. bilineata'''^"^ 


Longhead dab 




Limanda proboscidea' 


Yellowfin sole 




Limanda aspera"^ 


Rex sole 




Glyptocephalus zachirus^ 


Oher flatfish 




Hsopsetta isolepis'^'. 'Limanda sakhalinensis^"^, Psettichthys melanostictus, 
Microstomus pacificuti 



Fish assemblages differed noticeably among the inner, 
middle, and outer domains in the EBS; many species 
(sculpins, poachers, and eelpouts) primarily inhabited 
a single domain and many flatfishes and skates inhab- 
ited all three domains (Kaimmer'; Kinder and Schum- 
acher, 1981; Smith and Bakkala, 1982: Tables 1 and 
2). Because of the distinct domain environments and 
assemblages, the flatfish and roundfish species groups 
were analyzed by inner, middle, and outer domains 
separately and all domains were combined to detect the 
influence of climate change on each assemblage. 

Species groups 

The taxonomic level of species identification for the 
entire survey period (1975-2002) has varied because 
of an incomplete knowledge of species characters and 
because of survey time constraints at-sea. Historically, 
survey goals were to obtain fisheries data on commer- 
cially or potentially commercially important species, 
and limited effort was put forth for identification of 
other species. However, given more efficient technolo- 
gies, better identification field guides, and increased 
focus on noncommercial species, species identification 
has improved and is approaching the current extent of 
taxonomic knowledge. 

Efforts in recent years (1998-2002) have increased 
our knowledge of current species distributions in the 



Kaimmer, S. M., J. E. Reeves, D. R., Gunderson, D. R.. Smith, 
G. B., and R. A. Macintosh. 1976. Baseline information 
from the 1975 OCSEAP survey of the demersal fauna of 
the eastern Bering Sea. In Demersal fish and shellfish 
resources of the eastern Bering Sea in the baseline year 
1975 (W. T Pereyra, J. R. Reeves, and R. G. Bakkala, eds.), 
p. 157-367. NWAFC Processed Report. Alaska Fisheries 
Science Center, National Marine Fisheries Service, NOAA, 
7600 Sand Point Way N.E. Seattle WA 98115. 



EBS shelf region. To assess species-identification con- 
fidence over the study period, current species distribu- 
tions from the AFSC survey and primary literature not 
associated with AFSC survey data, as well as historical 
fish collections of selected taxonomic groups (University 
of Washington Fish Collection, Oregon State University 
Fish Collection, and Auke Bay Fish Collections) were 
examined. After this assessment, it was subjectively 
determined whether species in this study were identified 
correctly throughout the study period. Species that were 
determined to be distinguishable and correctly identified 
were examined as individual species, and species that 
were possibly misidentified were grouped at a higher 
taxonomic level for this study. The resulting list of spe- 
cies or species groups includes the greatest number of 
fish taxa (Tables 1 and 2). 

At least 15 flatfish species (within the order Pleu- 
ronectiformes) and more than 75 species of fish and 
elasmobranchs not of the order Pleuronectiformes were 
recorded during the AFSC summer surveys (1975-2002) 
on the EBS shelf (20-200 m). Although flatfish make 
up about 16% of the total number of fish species, they 
contribute approximately 50% of the biomass of all fish 
combined (2001 and 2002 AFSC EBS survey estimates). 
Walleye pollock (Theragra chalcogramma) and Pacific 
cod (Gadus macrocephalus) represent approximately 
42% of the survey biomass, and 8% of the biomass that 
included all other roundfish. Because of the extremely 
large biomass of walleye pollock and Pacific cod, the 
biodiversity indexes used are uninformative when wall- 
eye pollock and Pacific cod are included owing to the 
"swamping" effect by their relatively large biomasses 
when compared to those of other species. To account 
for this, two species guilds were defined: flatfish and 
roundfish. The flatfish guild included all Pleuronecti- 
formes recorded from the EBS survey and comprised 
11 species or species groups (Table 1). The roundfish 



Hoff Biodiversity as an index of regime shift in tfie eastern Bering Sea 



229 







Table 2 


Roundfish guild species or species groups used for biodiversity indexes. * indicates the major proportion of the biomass for the 
group. Superscripts indicate the primary domains where each species occurs during the summertime AFSC survey. I=inner 
domain, M=middle domain, 0=outer domain. IMO=inner, middle, and outer domains. 


Group name 




Scientific name 


Lamprey 




Lampetra tridentata 


Skate 




*Bathyraja parniifera"''". *B. interrupta", B. aleutica. B. taranetzi. Raja bmociilata 


Shark 




*Somniosus pacificus^, Squalus acanthiaa. Lamna ditropis 


Other poachers 




*Aspidophoroides bartoni'^, Bathyagonus ainscanus. Leptagonus leptnrhynchus. 
Pallasina barbata, Percis japonicus 


Sawback poacher 




Sarritor frenatus-'''"^ 


Sturgeon poacher 




Agonus acipenserinus"^' 


Bering poacher 




Ocellus dodecaedron' 


Bering wolffish and wolfeel 


Auarhichas nrieittalis and Anarrhichthys ocellatus 


Searcher 




Bathymaster signatus" 


Other sculpins 




Artediellus pacificus, Leptocottus armatus, Enophrys spp., 

Psychrolutes paradoxus, Blepsias bilobus, Nautichthys pribilovious, Icelinus spp. 


Staghorn sculpin 




*Gymnocanthus pistilligei-', *G. galeatus'^. G. dctrisus 


Darkfin sculpin 




Malacocottus zonurus'-' 


Yellow Irish lord 




Hem ilepidotu s jorda n i' 


Butterfly sculpin 




Hemitepidotus papilio-^' 


Triglops sculpins 




*Triglops scepticus^. *T. pingeli'^'. T. macellus. T. forficatus 


Myoxocephalus sculpt 


ns 


*Myoxocephalusjaok', *M. polyacanthocephalus^"^, M. verrucosus'^ 


Spinyhead sculpin 




Dasycottus setiger^ 


Bigmouth sculpin 




Hemitripterus holinfi 


Icelus sculpins 




^'Icelus spmiger" and */. spatula'-'^' 


Pacific sandfish 




Trichodon trichodon' 


White-spotted greenl 


ing 


Hexagrammos stelleri' 


Snailfish 




*Liparis gibbus'^'-', *Careproctus rastrinus^'^, C. cyclospilus 


Smooth lumpsucker 




*Aptocyclus ventricosus'^ , Eumicrotremus orbis 


Pricklebacks 




*Lumpenus maculata"'"^, *L. sagittae'^"^, L. fabricii, L. medius. Poroclinus rolhrocki 


Other eelpouts 




*Lycodes spp. and Gyinnelus spp. 


Marbled eelpout 




Lycodes raridens' 


Wattled eelpout 




Lycodes palearis^° 


Shortfin eelpout 




Lycodes brevipes^ 


Pacific tomcod 




Microgadus proximus 


Saffron cod 




Elegiiius gracilis' 


Arctic cod 




Boreogadus saida^ 


Pacific sand lance 




Ammodytes hexapterus 


Pacific herring 




Clupea pallasii"^' 


Eulachon 




Thaleichthys pacificus'-' 


Capelin 




Mallotus villosus' 


Rainbow smelt 




Osmerus mordax"^"^ 


Salmon 




Oncorhynchus spp. 


Sablefish 




Anoplopoma fimbria^ 


Atka mackerel 




Pleurogrammus monopterygius 


Prowfish 




Zaprora silenus'^ 


Rockfish 




*Sebastes alutus°, S. polyspinis. S. ciliafus 



230 



Fishery Bulletin 104(2) 



guild included 40 species or species groups and excluded 
walleye pollock and Pacific cod (Table 2). The two guilds 
devised were subjective; however, owing to their dis- 
tinctive biomass and life history trends, the members 
of the flatfish guild were examined separately from all 
other species. 

Biodiversity indexes 

Biodiversity measures were calculated by using Ludwig 
and Reynolds's (1988) recommendations for species rich- 
ness and evenness. Richness and evenness are consid- 
ered robust measures and allow one to use biomass 
proportions for biodiversity estimations. The richness 
index used was as follows: 

Richness index = e" 



where H' = ^ 



'iHi) 



A piecewise linear model (Neter et al., 1996) was 
applied to the biodiversity indexes using S-Plus (vers. 
6.1, Insightful Corporation, Seattle, WA) to determine 
distinct breaks or inflection points in the linear data 
trends. The piecewise model finds a "knot," or inflec- 
tion point, that breaks the data set into two periods by 
using the least sum of squared residuals and the high- 
est i?'- for the two-line model to determine the best fit 
lines to the data. Linear regression models were then 
applied to the data sets by using the recommended 
inflection point as the breaking point for the two lines. 
The years covered for each best fit line, as well as the 
accompanying linear statistics for the individual lines 
and the piecewise linear model for the best-fit model are 
presented in Table 3. The indexes calculated for survey 
year 1975 are included for reference (denoted as an X 
on the diversity index graphs) but were not included in 
the linear regression models because of the time gap 
from 1975 to 1979 and the lack of confidence in the 
standardization of the 1975 survey compared with later 
surveys (Bakkala, 1993). 



and /?, = the biomass of individuals belonging to the /th 
of S species in the sample; and 
n = the total summed biomass for the entire 
guild for a given year (Ludwig and Reynolds, 

1988). 

The evenness index is the "modified Hill's ratio" (Ala- 
talo, 1981) which is 



Evermess = 



where A=-;^py, where Pi = N and ?!, is the biomass of the 
(th species and A^ is the total biomass of all iS (species) 
in the guild (Ludwig and Reynolds, 1988). 

Biomass estimates obtained from the surveys were 
used to calculate biodiversity values for both flatfish 
and roundfish guilds on an annual basis. Biodiversity 
values calculated were used as an index of the relative 
proportion of species in each guild and allowed a direct 
comparison from year to year of the most abundant 
species. Species richness indicated the effective number 
of species that are influencing the index (Hill, 1973; 
Ludwig and Reynolds, 1988). A higher index indicates a 
larger number of predominant species in the assemblage 
(i.e., higher diversity). The evenness index determines 
the distribution of the biomass proportions of the more 
abundant species among the species in the guild. As the 
evenness index approaches zero, the biomass proportion 
of a single species dominates the guild; as the evenness 
index approaches one, there is less of a single dominant 
species and the biomass proportions are shared more 
equally among many species in the guild (Hill, 1973; 
Ludwig and Reynolds, 1988). 



Results 

Flatfish biomass 

Biomass estimates for the flatfish guild showed an 
increase from the late 1970s until the early 1990s, and 
then an overall slight decline between 1999 and 2002 for 
all domains (Fig. 2). Species that represented a large por- 
tion that showned an increase in biomass estimates over 
the study period included northern and southern rock sole 
(Lepidopsetta polyxystra and L. bilineata, spp. undeter- 
mined), arrowtooth and Kamchatka flounder (Atheresthes 
stomias and A. evermanni, spp. undetermined), flathead 
sole and Bering flounder (Hippoglossoides elassodon and 
H. robustus, spp. undetermined), and Pacific halibut 
iHippoglossus stenolepis). Flatfish species with declining 
biomass estimates included the yellowfin sole (Limanda 
aspera) and Greenland turbot {Reinhardtius hippoglos- 
soides), and a decline in lesser abundant species such as 
longhead dab {Limanda proboscidea) and Alaska plaice 
(Pleuronectes qiiadrituberciilafus). The inner and middle 
domains were dominated by northern rock sole, yellowfin 
sole, Alaska plaice, and flathead sole and Bering flounder 
(spp. undetermined), whereas main portions of the outer 
shelf comprised arrowtooth and Kamchatka flounder 
(spp. undetermined), flathead sole and Bering flounder 
(spp. undetermined). Pacific halibut, and northern and 
southern rock sole (spp. undetermined). 

Roundfish biomass 

Roundfish biomass increased from the 1970s to the 1980s 
and then declined steadily from the early 1990s through 
2002 for all domains combined (Fig. 2). Among roundfish 
species there has been a decline since the 1970s and 1980s 
in many middle domain species, such as yellow Irish lord 
(Hemilepidotus jordani), butterfly sculpin (H. papilio), 



Hoff Biodiversity as an index of regime shift in tlie eastern Bering Sea 



231 











Table 3 












Statistics of the 1 


near regression models of biodiversity indexes for flatfish and 


roundfish guilds for the 


eastern Ber 


ng Sea shelf 


domains. The years encompassing each time 


period 


and the piecewise model r- 


for the two-1 


ne model are included. 




















Standard 


r^of 


















error of 


piecewise 


Domain 


Index 


Years 


n 


Intercept 


Slope 


P-value 


r2 


estimate 


model 


Flatfish guild 




















Inner 


Richness 


1979-1993 


15 


-112.9280 


0.0579 


0.0000 


0.8797 


0.0993 


0.8850 






1994-2002 


9 


3.9168 


-0.0008 


0.9645 


0.0303 


0.1270 




Middle 


Richness 


1979-1988 


10 


-358.4240 


0.1821 


0.0001 


0.8829 


0.2130 


0.8640 






1989-2002 


14 


7.8482 


-0.0020 


0.9171 


0.0940 


0.2887 




Outer 


Richness 


1979-2002 


24 


-2.8281 


0.0030 


0.6875 


0.7493 


0.2533 




Combined 


Richness 


1979-1989 


11 


-327.9090 


0.1666 


0.0000 


0.8740 


0.2212 


0.9350 






1990-2002 


13 


6.1232 


-0.0012 


0.8866 


0.1932 


0.1150 




Inner 


Evenness 


1979-1990 


12 


-48.3712 


0.0247 


0.0000 


0.9291 


0.0258 


0.9610 






1991-2002 


12 


-0.0349 


0.0005 


0.6927 


0.0163 


0.0161 




Middle 


Evenness 


1979-1987 


9 


-61.6390 


0.0314 


0.0003 


0.8663 


0.0361 


0.8660 






1988-2002 


15 


3.7805 


-0.0015 


0.5159 


3.3171 


0.0375 




Outer 


Evenness 


1979-2002 


24 


1.8798 


-0.0006 


0.6184 


0.0115 


0.0392 




Combined 


Evenness 


1979-1988 


10 


-59.8421 


0.0305 


0.0000 


0.9318 


0.0265 


0.9630 






1989-2002 


14 


2.0678 


-0.0007 


0.4732 


0.0437 


0.0136 




Roundfish guild 




















Inner 


Richness 


1979-2002 


24 


-5.0326 


0.0040 


0.8287 


0.2175 


0.6197 




Middle 


Richness 


1979-1986 


8 


865.6880 


-0.4338 


0.0119 


0.6789 


0.7895 


0.9040 






1987-2002 


16 


203.6660 


-0.1006 


0.0000 


0.7196 


0.3095 




Outer 


Richness 


1979-1986 


8 


908.1290 


-0.4564 


0.0011 


0.8511 


0.5051 


0.8690 






1987-2002 


16 


-65.9678 


0.0340 


0.0762 


0.2075 


0.3276 




Combined 


Richness 


1979-1986 


8 


1535.3000 


-0.7713 


0.0003 


0.9072 


0.6527 


0.9410 






1987-2002 


16 


81.4226 


-0.0394 


0.1204 


0.1634 


0.4390 




Inner 


Evenness 


1979-1990 


12 


-14.9002 


0.0078 


0.3003 


0.1066 


0.0853 


0.5620 






1991-2002 


12 


-32.0753 


0.0164 


0.0057 


0.5504 


0.0561 




Middle 


Evenness 


1979-1990 


12 


51. .5981 


-0.0257 


0.0001 


0.7779 


0.0518 


0.9210 






1991-2002 


12 


20.3337 


-0.0099 


0.0002 


0.7694 


0.0206 




Outer 


Evenness 


1979-1987 


9 


66.1545 


-0.0331 


0.0024 


0.7550 


0.0552 


0.7150 






1988-2002 


15 


-16.5892 


0.0085 


0.0110 


0.4028 


0.0482 




Combined 


Evenness 


1979-1989 


11 


46.7485 


-0.0233 


0.0003 


0.7782 


0.0434 


0.8630 






1990-2002 


13 


1.1549 


-0.0003 


0.8818 


0.0021 


0.0306 





armorhead sculpin (Gymnocanthus galeatus), snailfishes 
{Liparis gibbus and Careproctus ?-astrinus). marbled eel- 
pout (Lycodes raridens), and wattled eelpout [Lycodes 
palearis), and in outer domain species such, as shortfin 
eelpout (Lycodes brevipes) and sablefish (Anoplopoma 
fimbria). There has been a increase in large sculpins, 
including the bigmouth sculpin iHemitripteriis bolini) and 
the great sculpin (Myoxocephahis polyacanthocephcdus), in 
the outer domain. The biomass of skates predominantly 
the Alaska skate (Bathyraja parmifera), increased consid- 
erably from 1979 through 1990, followed by a prolonged 
period of little change from 1990 through 2002. 



domain the major portion of the assemblage comprises 
3-5 species. Figure 3 shows the change in biodiversity 
indexes from 1975 to 2002 for each domain separately 
and all domains combined. The piecewise model shows a 
period of inflection for evenness and richness indexes for 
flatfish in the inner and middle domains that occurred 
between 1987 and 1993, with the average year of inflec- 
tion being 1990 (Fig. 3, Table 3). The outer domain for 
flatfish showed little change in evenness and richness 
throughout the 24-year study period. Three species 
dominated the entire period, with somewhat evenly 
distributed biomass proportions. 



Flatfish biodiversity 

Eleven species or species groups were used for biodi- 
versity indexes for the flatfish guild (Table 1). All 11 
species are rarely found together, and within each shelf 



Roundfish biodiversity 

Forty-one species or species groups were used in the 
roundfish biodiversity analysis (Table 2). The even- 
ness and species richness index for the roundfish group 



232 



Fishery Bulletin 104(2) 



Flatfish 
guild 



Inner domain 




Middle domain 




Flatfish guild 

B other flatfish 
□ longhead dab 
H rex sole 



Year 



starry flounder 
Pacific halibut 
Alaska plaice 




$113 g 





@ Greendland turbot 

[H flathead/Berring flounder 

n arrowtooth/Kamchatka flounder 



Year 



yellowfin sole 
northern/southern rock sole 



Figure 2 

Biomass estimates for flatfish guild and roundfish guild (left panel), and relative proportion of each species (right panel) 
for inner, middle, outer, and all three domains combined, respectively. 



Hoff: Biodiversity as an index of regime shift in the eastern Bering Sea 



233 



Roundfish 
guild 



Inner domain 








JSOQXi 
4WXD0 



Middle domain 






■IS rjis ■i^V^fe' 



s«r 






l^w 



03 <som Outer domain 








looaxK. ^11 domains combined 

9000)0 



SOOCDO 

40oa» 








Year 



Year 



Roundfish guild 

□ Atka mackeral 

□ prowfish 

□ pricklebacks 

□ salmon 

□ hagfish and lampreys 

□ sharks 

□ arctic cod 

n Pacific tomcod 

□ rainbow smelt 



□ eulachon 

□ Pacific sand lace 

□ Pacific sandfish 

□ other sculpins 

□ /ce/us sculpins 

□ spinyheaded sculpin 

□ Tr/g/ops sculpins 

□ darkfin sculpin 



□ armorhead sculpin 


Q saffron cod 


^ wattled eelpout 


n Bering wolffish and wolfeel 


 other eelpouts 


\Zl butterfly sculpin 


n Bering poacher 


^ snailfishes 


IS marbled eelpout 


n sawback poacher 


U white spotted greenling 


H sturgeon poacher 


n other poachers 


jg yellow Irish lord 


□ sablefish 


□ searcher 


B shortfin eelpout 


g Pacific herring 


n lumpsuckers 


^ rockfish 


^ Myoxocephalus sculpins 


ES capelin 


^ bigmouth sculpin 


M skates 



Figure 2 (continued) 



234 



Fishery Bulletin 104(2) 



Flatfish 



Inner domain 




1974 1979 1984 1989 



Middle domain 



1999 2004 




-^r^o- 



Oo 



« 1974 



1 984 1 989 



1999 2004 



EC 4 



Outer domain 



0,0 



1974 1979 1984 1989 1994 1999 2004 



All domains combined 




no, 



1974 1979 1984 1989 1994 

Year 



1 999 2004 



1 
0.& 

06- 
0.4- 
0.2- 



o.a 

0.6 

4. 
0.2. 



) ''' fl ^ P 



1974 1979 1984 1989 1994 1999 2004 
1 




On 



(U 1974 1979 1984 

c 

> 

0.6 
0.4 

02 



1994 1999 2004 



^puO ' ^pooooo '■'»"" — a-e-6 



1974 1979 1984 1989 1994 1999 2004 

1 
0.8- 
06 
04- 
0.? 




oJ> oOnoOoOnooo^oo 



1974 1979 1984 1989 1994 1999 2004 

Year 



Figure 3 

Biodiversity indexes (richness left, evenness right) for the flatfish guild determined with two 
linear models (or a single model) separated at the inflection point derived from the piecewise 
linear model. The graphs represent the inner, middle, outer, and all domains combined, respec- 
tively. The X represents biodiversity index for survey year 1975 which was not included in linear 
models or the piecewise model because of a gap in the time series. 



is plotted for each shelf domain and for all domains 
combined. The inner domain richness index revealed 
a steady level of approximately 2-3 dominant species 
or species groups iMyoxocephalus sculpins, skates, and 
Pacific herring) over the 24-year period. The gradual 
increase in skates and slight decline in the plain sculpin 
caused an increase in the evenness index to somewhat 
even proportions of the three dominant species in the 



inner domain, from 7 to 2 dominant species over the 
24-year period, and evenness declined by about one-half 
(Fig. 4). The increase in skates and decline of many 
inner domain species from 1975 through the early 1990s 
resulted in an assemblage dominated by two species 
groups, Myoxocephalus sculpins and skates (Fig. 2). 

The outer domain showed a dramatic decline in rich- 
ness and evenness from 1975 to the mid-1980s and 



Hoff Biodiversity as an index of regime shift in tfie eastern Bering Sea 



235 



Roundfish 



Inner domain 



'" °°o"oo°""°''»°"°°"°°' 



1 

o.s 

06 

4- 
0? 



1974 1979 1984 1989 1994 1999 2004 



1974 1979 1984 1989 1994 1999 2004 



10- 


Middle domain 








a- 










8i 


X 








7i 


~^ ° 








6' 


° >. 








5- 


^ 








4- 




"^ 






3- 




6 ct"^-^ 




-T-<-^:^ 


2- 

1- 








rrf>-^ 



1-, 

0.8- 
0.6- 
0.4. 
0.2. 




1974 1979 1984 1989 1994 1999 2004 



1974 1979 1984 1989 1994 1999 2004 



Outer domain 



10 
9-\ 
8 
7 
6 
5- 
4 
3 
2 
1 
1974 1979 1984 1989 1994 1999 2004 1974 1979 1984 1989 1994 1999 2004 




0.8- 
0.6-1 X 

0.4. 
0.2. 



All domains combined 




1- 
0.8- 
0.6- 
0.4- 
2- 

0- 



o^~a -e-r, — n P o .-■ f7< o 



P '  '  ' '■ ' u o 



1974 1979 1984 1989 1994 1999 2004 19'" 1979 1984 1989 1994 1999 2004 

Year Year 

Figure 4 

Biomass estimates of roundfish guild (left panel), and relative proportion of each species (right 
panel) for the inner, middle, outer, and all domains combined, respectively. 



then a slight increase through 2002. A decrease in the 
number of dominant species, such as sablefish, wattled 
and shortfin eelpouts from 1979 to 1986, and a large 
increase of skate biomass from 20% in 1975 to approxi- 
mately 75% of the total biomass by 1986, explain the 
index decline during this same period as the skate 
biomass dominated. After 1986 the skate populations 
stabilized somewhat to a slight decline which accounts 
for the slight increase in evenness and richness as 
other species gained a larger percentage of the total 
biomass. 



All indexes for the roundfish guild, except for inner 
domain richness, indicated two distinct linear periods. 
Inflection points ranged from 1986 to 1990 and 1988 
was an average inflection year for the roundfish 
guild. 



Discussion 

Periods of climatic change have been shown to instill 
long-term changes in the North Pacific ecosystem (Hare 



236 



Fishery Bulletin 104(2) 



and Mantua, 2000). Biodiversity indexes are a robu