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 Fairbanli :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): l0 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 .0 J: Q ii - i - . l .., - :^^ ~-^^ ! •— --*- 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 •''' __^ *-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: *' = - 2/3 logio (a), where = 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

— >^ -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 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') in terms of growth in length with other estimates obtained for the same or a similar species (Gayanilo and Pauly, 1997). Values of

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)cl0)-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 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 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