U.S. Department of Commerce Volume 114 Number 4 October 2016 Fishery Bulletin U.S. Department of Commerce Penny S. Pritzker Secretary National Oceanic and Atmospheric Administration Kathryn D. Sullivan NOAA Administrator National Marine Fisheries Service Eileen Sobeck Assistant Administrator for Fisheries The Fishery Bulletin (ISSN 0090-0656) is published quarterly by the Scientific Publications Office, National Marine Fisheries Service, NOAA, 7600 Sand Point Way NE, Seattle, WA 98115-0070. 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; $32.00 domestic and $44.80 foreign. Cost per single issue: $19.00 domestic and $26.60 foreign. See back for order form. Scientific Editor Richard Langton National Marine Fisheries Service Northeast Fisheries Science Center Maine Field Station 17 Godfrey Drive, Suite 1 Orono, ME 04473 Managing Editor Sharyn Matriotti National Marine Fisheries Service Scientific Publications Office 7600 Sand Point Way NE Seattle, Washington 98115-0070 Associate Editor Kathryn Dennis National Marine Fisheries Service Office of Science and Technology 1845 Wasp Blvd., Bldg. 176 Honolulu, Haw/aii 96818 Editorial Committee Richard Brodeur National John Carlson Kevin Craig John Graves Rich McBride Rick Methot Bruce Mundy David Sampson Michael Simpkins Dave Somerton Mary Yoklavich National National Virginia National National National Oregon National National National Marine Fisheries Service, Newport, Oregon Marine Fisheries Service, Panama City, Florida Marine Fisheries Service, Beaufort, North Carolina Institute of Marine Science, Gloucester Point, Virginia Marine Fisheries Service, Woods Hole, Massachusetts Marine Fisheries Service, Seattle, Washington Marine Fisheries Service, Honolulu, Hawaii State University, Newport, Oregon Marine Fisheries Service, Woods Hole, Massachusetts Marine Fisheries Service, Seattle, Washington Marine Fisheries Service, Santa Cruz, California Fishery Bulletin web site: www.fisherybulletm.noaa.gov The Fishery Bulletin carries original research reports on investigations in fishery sci- ence, engineering, and economics. It began as the Bulletin of the United States Fish Commis- sion in 1881; it became the Bulletin of the Bureau of Fisheries in 1904 and the Fishery Bulle- tin of the Fish and Wildlife Service in 1941. Separates were issued as documents through vol- ume 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. Beginning with volume 70, number 1, January 1972, Fishery Bulletin became a periodical, issued quarterly. In this form, it is available by subscription from the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402. It is also available free in limited numbers to libraries, research institutions, state and federal agencies, and in exchange for other scientific publications. I U.S. Department of Commerce Seattle, Washington Volume 114 Number 4 October 2016 Rshery Bulletin Contents Short contribution 377-385 Elzey, Scott P., and Kimberly J. Trull Identification of a nonlethal method for aging tautog iTautoga onitis) Articles 386-396 Laidig, Thomas E., and Mary M. Yoklavich A comparison of density and length of Pacific groundfishes observed from 2 survey vehicles: a manned submersible and a remotely operated vehicle 397-408 Macchi, Gustavo J., Marina V. Diaz, Ezequiel Leonarduzzi, Maria Ines Militelli, and Karina Ridrigues Skipped spawning in the Patagonian stock of Argentine hake iMerlucdus hubbsi) The National Marine Fisheries Service (NMFS) 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, rec- ommends, or endorses any propri- etary product or proprietary mate- rial mentioned herein, or which has as its purpose an intent to cause directly or indirectly the advertised product to be used or purchased be- cause of this NMFS publication. The NMFS Scientific Publications Office is not responsible for the con- tents of the articles. 409-425 Trianni, Michael S. Life history characteristics and stock status of the thumbprint emperor (.Lethrinus harak) in Saipan Lagoon Short contributions 426-434 Whitney, Nicholas M., Marc Taquet, Richard W. Brill, Charlotte Girard, Gail D. Schwieterman, Laurent Dagorn, and Kim N. Holland Swimming depth of dolphinfish (Coryphaena hippurus) associated and unassociated with fish aggregating devices 435-444 Lopez, Miranda, Gavino Puggioni, and David A. Bengtson First assessment of the field ecology of larval Atlantic silverside (Menidia menidia) Fishery Bulletin 114(4) Articles 445-460 Bizzarro, Joseph J., Ashley N. Peterson, Jennifer M. Blaine, Jordan P. Balaban, H. Gary Greene, and Adam P. Summers Burrowing behavior, habitat, and functional morphology of the Pacific sand lance iAmmodytes personatus) 461-475 Paulakis, Gregg R., Philip W. Stevens, Amy A. Timmers, Christopher J. Stafford, Demian D. Chapman, Kevin A. Feldheim, Michelle R. Heupel, and Caitlin Curtis Long-term site fidelity of endangered small-tooth sawfish iPristis pectinata) from different mothers 476-489 Blaylock, Jessica, and Gary R. Shepherd Evaluating the vulnerability of an atypical protogynous hermaphrodite to fishery exploitation: results from a population model for black sea bass (Centropristis striata) 490-502 Mcliwain, Jennifer L., Aisha Ambu-ali, Nasr Al Jardani, Andrew R. Halford, Hamed S. Al-Oufi, and David A. Feary Demographic profile of an overexploited serranid, the orange-spotted grouper iEpinephelus coioides), from northern Oman Short contribution 503-512 Mickle, Paul F., Jacob F. Schaefer, Susan B. Adams, Brian R. Kreiser, and William T. Slack Environmental conditions of 2 river drainages into the northern Gulf of Mexico during successful hatching of Alabama shad (Atosa alabamae) 513-514 Acknowledgment of reviewers 515-518 Guidelines for authors 377 NOAA National Marine Fisheries Service Abstract-Aging of tautog (Tautoga onitis) has historically required sac- rificing fish to obtain opercula and otoliths. Use of these structures for age determination has hindered re- searchers from obtaining samples from fish that were to be released alive, as well as from commercially collected fish that are commonly sold whole. In this study we evaluated the use of scales, dorsal-fin spines, pelvic-fin spines, opercula, whole sagittal otoliths, and sectioned sag- ittal otoliths as structures for age determination of tautog. Our results indicate that pelvic-fin spines pro- vide high-precision age estimates without bias. Dorsal-fin spines had well-defined annuli, but vasculariza- tion near the core prevented consis- tent identification of the first an- nulus and led to biased ages. Scales were difficult to read and provided highly biased ages in older (>age 7) fish. The precision of age estima- tions derived from pelvic-fin spines was better than the precision of age estimations derived from the other structures. Pelvic-fin spines provide suitable age estimates for tautog, and these structures can be collected easily from a wider variety of sam- ple sources than can the structures currently being collected for age de- termination of this species. Manuscript submitted 7 December 2015. Manuscript accepted 16 June 2016. Fish. Bull. 114:377-385 (2016). Online publication date: 7 July 2016. doi: 10.7755/FB.114.4.1 The views and opinions expressed or implied in this article are those of the author (or authors) and do not necessarily reflect the position of the National Marine Fisheries Service, NOAA. Fishery Bulletin fb^ established 1881 -ds Spencer F. Baird First U.S. Commissioner of Fisheries and founder of Fishery Bulletin Identification of a nonlethal method for aging taytog iTautoga onitis} Scott P. Elzey (contact author) Kimberly J. Trull Email address for contact author: sc0tt.el2ey@state.ma. us Fish Biology Program Massachusetts Division of Marine Fisheries 30 Emerson Avenue Gloucester, Massachusetts 01930 The tautog {Tautoga onitis) is a spe- cies of fish from the family Labridae that ranges from Nova Scotia (Scott and Scott, 1988) to South Carolina (Grimes et ah, 1982). It is a com- mercially and recreationally impor- tant species from Massachusetts to Virginia (ASMFC^). Tautog grow to approximately 90 cm in total length (TL) and 10.2 kg in weight (Bigelow and Schroeder, 1953). It is a slow growing and long-lived species that reaches maturity at age 3 (Hostet- ter and Munroe, 1993) and has been estimated to live up to age 34 (Coo- per, 1965, 1967; Hostetter and Mun- roe, 1993). Cooper (1965, 1967) and Hostetter and Munroe (1993) were able to estimate the age of tautog from marks on their opercula. Hostet- ter and Munroe (1993) were also able to justify the assumption that marks on the opercula were deposited an- nually and, therefore, justify the use of those marks for age determination through marginal increment analy- sis. For these reasons, opercula have been the primary and recommended structure for estimating the age of tautog (ASMFC^). In 2012, representatives from 10 1 ASMFC (Atlantic States Marine Fisher- ies Commission). 2015. Tautog bench- mark stock assessment and peer review reports, 283 p. AFMFC, Arlington, VA. [Available at website.] different laboratories attended a workshop on aging tautog (ASMFC^). Although staff at the majority of the laboratories had considerable expe- rience aging tautog, with the use of the operculum as the structure for determining age, and staff at a few of the laboratories had experience with otoliths of tautog, the precision of age estimates between laborato- ries was similar for both structures. The results from that workshop indi- cated that, with increased experience by the staff, the use of sectioned oto- liths from tautog may yield age es- timates of higher precision than the use of opercula. After that workshop, sectioned otoliths have been used as a supplementary method for age de- termination (ASMFC^). Although current methods for age determination of tautog are based on opercula and otoliths, multiple structures have been used to age other fish species, including oper- cula, otoliths, vertebrae, fin rays, fin spines, and scales (Beamish and Mc- Farlane, 1987; Panfili et al., 2002). Fin rays, fin spines, and scales have the distinct advantage in that their collection is nonlethal. Phelps et 2 ASMFC (Atlantic States Marine Fish- eries Commission). 2012. Proceed- ings of the tautog ageing workshop, 88 p. AFMFC, Arlington, VA. [Available at website.] 378 Fishery Bulletin 114(4) al. (2007) and Watkins et al. (2015) were able to suc- cessfully determine the age of common carp iCypri- nus carpio) using cross sections of fin rays. Carbines (2004) compared ages of blue cod (Parapercis colias) derived from otoliths and fin spines and determined that spines yielded precise estimates. In comparisons of otoliths, dorsal-fin spines, and teeth of the leopard coralgrouper {Plectropomus leopardus), Hobbs et al. (2014) found that the most cost- and time-efficient structure for age determination was the dorsal-fin spine. Fin rays or spines also have been found to be useful by Sylvester and Berry (2006) for white sucker (Catostomus commersonii), by Zymonas and McMahon (2009) for bull trout iSalvelinus confluentus), by Bur- ton et al. (2015) for gray triggerfish (Balistes capris- cus), by Keller Kopf et al. (2010) for billfishes {Kajikia spp.), and by Murie et al. (2009) for Atlantic goliath grouper (Epinephelus itajara). Management recommendations for tautog stocks are a product of the stock assessment process, which cur- rently is based on values derived from an age struc- tured assessment model. The benchmark stock assess- ment for tautog, conducted in 2015 by the Atlantic States Marine Fisheries Commission (ASMFC) and incorporating an external peer review, gave evidence that tautog in all management areas (Southern New England, which includes Massachusetts and Connecti- cut; New York-New Jersey; and Delaware, Maryland, and Virginia) were overfished and that overfishing was occurring for the stock in Southern New England (ASMFCD. The assessment came with several research recommendations that included 1) examination of dif- ferences in tautog growth rates by using data that are representative of the full size-age structure of the species, 2) expanded biological sampling of the com- mercial catch (including collection of structures for age estimates), 3) enhanced collection of age informa- tion for smaller fish (<20 cm TL), and 4) maintaining and improving the precision of age readings between state agencies that are estimating the ages of tautog (ASMFCD. To address the aforementioned research recommen- dations from this stock assessment is difficult with current aging methods, primarily because removal of opercula and otoliths from tautog require sacrific- ing and disfiguring the fish. In Massachusetts, many of the commercially captured fish are sold whole, both alive and dead. Many of the commercial dealers do not want their fish damaged by the removal of opercula or otoliths; therefore, the collection of age samples from the commercial harvest is not feasible without the ex- pense of purchasing fish. Identification of a structure that could be used for age determination without the need of sacrificing or altering the marketability of the fish would enable more samples to be collected across a variety of sources. In this study, the precision of age estimates generated from multiple structures was ex- amined to establish an alternative to the current use of opercula and sectioned otoliths as the primary aging structures. Materials and methods Tautog were collected by rod and reel, as well as from the trawl survey of the Massachusetts Division of Marine Fisheries in the waters of Buzzards Bay, Massachusetts, in May, September, and October 2014. Specimens were transported frozen or on ice to the An- nisquam River Marine Fisheries Station of the Mas- sachusetts Division of Marine Fisheries in Gloucester, Massachusetts, for further processing. Total length in millimeters, weight in grams, and sex were recorded. Scales were removed from the side of each fish just posterior to the pectoral fin and placed in an enve- lope. The fourth dorsal-fin spine and the first pelvic-fin spines were removed with wire cutters as close to the body of the fish as possible and stored frozen in plas- tic bags. Both opercula were removed with a knife and stored frozen in plastic bags. Sagittal otoliths were re- moved with a serrated knife and fine forceps and then rinsed, dabbed dry on a paper towel, and stored dry in microcentrifuge tubes. Opercula were placed into boiling water for 2 min and a small brush was used to remove any flesh still adhering to the bone. Opercula were allowed to air dry for a minimum of 24 h before being examined without magnification by using a combination of reflected and transmitted light. Annuli were defined as alternating pairs of translucent and opaque growth zones. Both left and right opercula were examined together to aid in discriminating between annuli and checks. As noted by both Cooper (1967) and Hostetter and Munroe (1993), the thickness of the bone in some opercula obscured the area of earliest growth, occasionally hiding the first annulus. Dorsal- and pelvic-fin spines were placed into boil- ing water for 2 min, and a small brush was used to remove any flesh still adhering to the spines. Spines were allowed to air dry for a minimum of 24 h before being placed in bullet molds and embedded in epoxy. The epoxy block with the embedded spine was sec- tioned with an IsoMet Low Speed Saw^ (Buehler, Lake Bluff, ID affixed with 4 blades and a 0.75-mm-thick spacer between each blade. Sections were affixed se- quentially (from the spine base to the tip) to labeled glass microscope slides with Flo-Texx liquid coverslip (Thermo Fisher Scientific, Waltham, MA). Sections of pelvic- and dorsal-fin spines were examined through a compound microscope with transmitted light at lOOx magnification. Each section from each spine was exam- ined to determine the age of the structure. Annuli were considered to consist of alternating pairs of opaque and translucent growth zones. Whole sagittal otoliths were cleaned with water as needed before being placed in a black dish filled with mineral oil and were viewed through a dissecting mi- croscope with reflected light at 30-40x magnification. 3 Mention of trade names or commercial companies is for iden- tification purposes only and does not imply endorsement by the National Marine Fisheries Service, NOAA. Elzey and Trull: Identification of a nonlethal method for aging Tautoga onitis 379 Left and right otoliths were examined side by side to aid in discerning between annuli and checks. The distal surface of the otoliths provided the clearest view of the annuli, which were identified as alternating bands of opaque and translucent growth zones. One sagittal otolith from each fish was randomly se- lected for sectioning. Otoliths that showed any sign of malformation were not sectioned. Otoliths were placed on a porcelain tray that had been heated to 400°C for approximately 30 s or until they reached a caramel brown color. Otoliths were then embedded in epoxy in silicon bullet molds, and each epoxy block was marked with a pencil through the core of the otolith perpen- dicular to the sulcal groove. This mark was used as a guide to cut a section with an IsoMet Low Speed Saw affixed with 2 blades and a 0.5-mm-thick spacer between them. The resulting section was affixed to a microscope slide with a Flo-Texx liquid coverslip and labeled. If the first otolith produced an undesirable section, the second otolith was cut. Otolith sections were examined through a compound microscope with transmitted light at lOOx magnification. Annuli were counted as alternating bands of opaque and translu- cent growth zones. Scales were briefly soaked in water to soften any attached tissue before being rubbed clean with a paper towel. Impressions of the scales were made by pressing them into acetate sheets for 3 min with a heat press set at 100°C and with 5.5 metric tons of pressure. Im- pressions were made for scales in which the anterior portion of the scale appeared to make a “v” shape, sig- naling a nonregenerated scale. Regenerated and nonre- generated scales were counted to create an estimation of the percentage of regenerated scales. Scale impres- sions were examined under a microfiche reader at 25x magnification. Breakages in the circuli that continued around the anterior portion of the scale were counted as the outer margin of annual growth zones. True an- nuli were differentiated from false annuli by confirm- ing that the circuli breakages continued through the transitional area between the anterior and posterior portion of the scale. All ages were assigned on the basis of year class. Fish captured in May, before annulus deposition was complete, had the edge counted as the final annulus. Fish captured in September and October had growth past the final annulus; therefore, the edge was not counted. In all structures, the outside edge of the winter growth was treated as the end of one annual growth zone and the beginning of the next. All struc- tures were independently assigned an age by 2 read- ers and each individual read each structure twice. All ages were assigned without knowledge of fish size, sex, or previously assigned ages. When ages assigned for a structure of a fish did not agree between the 2 read- ers, both readers examined the structure together and reached a consensus-based age. A final age for each fish was reached by the 2 readers considering consensus- based ages of all structures, as well as the quality of each structure. For example, a fish determined to be age 7 with the use of opercula, age 8 with the use of whole otolith, age 8 with the use of sectioned otolith, age 7 with the use of dorsal-fin spine, age 8 with the use of pelvic-fin spine, and age 6 with the use of a scale would be assigned a final age of 8 years if the opercula was thick at the base, the dorsal-fin spine was vascu- larized, and the scale was of poor quality. Precision of readings was measured by using per- cent agreement and coefficient of variation (CV) (Chang, 1982). Estimates of precision were gener- ated for comparisons 1) within reader for each struc- ture, 2) between readers for the first reading of each structure, 3) between readers for the second reading of each structure, 4) between consensus-based ages for sectioned otoliths and consensus-based ages for each other structure, and 5) between consensus-based ages for each structure and final ages assigned to a fish. The following equation was used to calculate CV, as shown in Campana (2001): CVj = 100% X R-1 This equation gives the CV for the yth fish, where = the ith age estimate of the yth fish; Xj = the average age estimate of the jth fish; and R = the number of times that that a fish was read. For all CV analyses, consensus-based ages were treated as a single reading. Coefficients of variations listed in this article were averaged across all fish aged. Tests of symmetry (Bowker, 1948; Evans and Hoe- nig, 1998) were used to examine bias between consen- sus-based ages for each structure and the final age as- signed to the fish, as well as between consensus-based ages for sectioned otoliths and consensus-based ages for each other structure. McBride (2015) suggested that Bowker’s test (Bowker, 1948) has a lower type-1 error rate at high levels of precision than the type-1 error rate with Evans and Hoenig’s test (Evans and Hoenig, 1998). We, therefore, used Bowker’s test when the CV was less than 5% and Evans and Hoenig’s test when the CV was above 5%. Results In this study, 119 tautog were collected and analyzed (52 female, 51 male, and 16 of unknown sex). Fish ranged from 35 mm TL to 506 mm TL (average: 313 mm TL). Males were slightly larger on average (346 mm TL) than females (335 mm TL), and fish of both sexes were larger than fish of unknown sex (140 mm TL). All structures examined contained annuli (Fig. 1). Scales yielded ages 0-10, opercula yielded ages 0-11, and all ages estimated for other structures ranged from 0 to 12 years. For all structures, average consen- sus-based ages of structures agreed within 1 year of 380 Fishery Bulletin 1 14(4) Figure 1 Multiple structures were used in this study for age determination of tautog (Tautoga onitis) captured in Buzzards Bay Massachusetts, in 2014. This figure shows (A) a whole sagittal otolith, (B) a sec- tioned sagittal otolith, (C) a scale, (D) an operculum, (E) a dorsal-fin spine, and (F) a pelvic-fin spine from an age-5 tautog captured in June 2014. Black dots represent the location of annuli. the final ages of fish up to final age of 7 years. Beyond age 7, ages derived from scales diverged by more than a year, and ages derived from whole otoliths diverged at the final age of 12 years (Table 1). The percentage of regenerated scales ranged from an average of 59% at age 1 to an average of 91% at age 11 (average estimate: 74.3% for all fish examined). Tests of symmetry against ages derived from sec- tioned otoliths yielded no bias for ages from whole otoliths, dorsal-fin spines, or pelvic-fin spines. Oper- cula and scales produced biased ages compared with ages from sectioned otoliths. The bias seen in ages from scales increased with the age of the fish, where- as the bias observed in ages from opercula appeared to be systematic because the age for a portion of the fish was underestimated by 1 year (Fig. 2). Compari- sons between final ages and consensus-based ages for structures showed that only pelvic-fin spines and whole otoliths yielded ages that were not biased in compari- son with the final ages determined for fish (Fig. 2). All other structures yielded ages that were biased younger than the final ages (Fig. 2). The bias in age estimates from scales became greater with fish age, whereas the other structures appeared to produce ages with a sys- tematic bias of underestimating a portion of the fish by 1 year. Within-reader comparisons showed reader 1 had the best precision (CV=1.51%, 89.9% agreement) with pel- vic-fin spines (Table 2). Reader 2 had the best precision (CV=2.39%, 80.7% agreement) with opercula, followed closely by that for pelvic-fin spines (CV=2.69%, 79.0% agreement). Between-reader precision was best in the second reading of whole otoliths (CV=1.96%, 84.0% agreement). For 4 of the 6 structures, between-reader precision increased between the first and second read- ings. Precision decreased only between readings for scales (CV changed from 6.99% to 8.97%) and pelvic-fin spines (CV changed from 4.68% to 5.28%). Elzey and Trull: Identification of a nonlethal method for aging Tautoga onitis 381 Table 1 Average age and sample size for each type of structure examined for each final age assigned for tautog {Tautoga onitis) col- lected in Buzzards Bay, Massachusetts, in 2014. Final ages were assigned by 2 age readers taking into account ages assigned, and quality of all structures examined for each fish. Ages for each structure are ages based on the consensus of 2 readers, each performing 2 readings. Standard errors of the mean appear in parentheses after the average ages. Final age n Scale Opercula Dorsal-fin spine Pelvic-fin spine Whole otolith Sectioned otolith 0 6 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 1 6 1.17 (0.17) 1.50 (0.22) 1.00 (0.00) 1.17 (0.17) 1.00 (0.00) 1.00 (0.00) 2 3 2.33 (0.33) 2.00 (0.00) 2.33 (0.33) 2.00 (0.00) 2.00 (0.00) 1.67 (0.33) 3 3 2.67 (0.33) 2.33 (0.33) 2.67 (0.33) 3.33 (0.33) 3.00 (0.00) 2.67 (0.33) 4 5 4.00 (0.32) 4.00 (0.00) 4.00 (0.00) 4.20 (0.20) 4.00 (0.00) 4.00 (0.00) 5 11 4.91 (0.16) 4.82 (0.12) 4.82 (0.12) 5.09 (0.09) 5.00 (0.00) 5.00 (0.00) 6 14 5.71 (0.13) 5.71 (0.16) 5.79 (0.11) 6.00 (0.00) 5.93 (0.07) 6.00 (0.00) 7 32 6.53 (0.13) 6.69 (0.08) 6.94 (0.04) 6.97 (0.03) 7.00 (0.04) 6.97 (0.05) 8 20 6.70 (0.16) 7.45 (0.14) 7.60 (0.15) 8.00 (0.07) 7.75 (0.12) 7.70 (0.11) 9 6 8.17 (0.31) 8.33 (0.21) 8.67 (0.21) 9.00 (0.00) 9.00 (0.00) 8.50 (0.22) 10 9 7.89 (0.42) 9.22 (0.22) 9.56 (0.18) 10.00 (0.17) 9.78 (0.15) 9.56 (0.29) 11 2 5.50 (0.50) 11.00 (0.00) 10.00 (0.00) 11.00 (0.00) 10.50 (0.50) 11.00 (0.00) 12 2 8.50(1.50) 11.00 (0.00) 11.50 (0.50) 12.00 (0.00) 10.50 (0.50) 11.00 (0.00) Consensus-based ages from sectioned otoliths had the best precision when compared with consensus-based ages from whole otolith ages (CV=2.11%, 87.4% agree- ment), followed by consensus-based ages from pelvic-fin spines (CV=2.99%, 79.0% agreement). Consensus-based ages from pelvic-fin spines had better precision for fi- nal ages than any other structure (CV=1.17%, 92.4% agreement). The CV estimate for comparisons between final ages and consensus-based ages from whole oto- liths was identical to that for consensus-based ages from pelvic-fin spines, but the agreement was not as high (CV=1.17%, 87.4% agreement) (Fig. 2). Discussion In this study, we made age determinations based on the examination of multiple structures from tautog in an effort to find a nonlethal aging method for this spe- cies and, thereby, increase the availability of sample sources. Current standard aging techniques for tautog are based on opercula and otoliths, which necessitates sacrificing and damaging the fish. Scales were evalu- ated as possible aging structures for tautog by Cooper (1967) and Hostetter and Munroe (1993). Before this study, no one had evaluated fin spines for aging tautog. Fin spines have proven useful for aging a variety of fish species (e.g., Carbines, 2004; Hobbs et al., 2014; Burton et al., 2015; Watkins et al., 2015). Among the structures that could be removed in a nonlethal way, we found pelvic-fin spines to be the best structure for aging tautog. Pelvic-fin spines had strong annular marks (Fig. 1), yielded high-precision age estimates (Table 2) and a lack of bias (Fig. 2). The precision of age estimates was higher for pelvic-fin spines in com- parison with final ages than from any other structure tested. Furthermore, with the pelvic-fin spines used in this study, we did not find the core to be obscured by vascularization as has been seen in fin spines of other species (e.g., Keller Kopf et al., 2010; Kopf and Davie, 2011; Landa et al., 2015). Although the tautog in our study reached only age 12, we believe that the growth bands in the spines would be discernible in tautog of considerably older ages. To support this assumption, we examined a whole otolith, sectioned otolith, oper- cula, and a pelvic-fin spine from an 895-mm-TL tautog. Using the opercula and pelvic-fin spine, we determined that the fish was age 20. The whole and sectioned otoliths, however, indicated that the fish was age 21 (senior author, unpubl. data). Annuli on all structures examined were clear all the way to the edge. The dis- crepancy in ages between structures is presumed to be related to difficulty in finding the first annulus. Dorsal-fin spines had strong annular marks that were very similar to those on pelvic-fin spines (Fig. 1), but the dorsal-fin spines had more vascularization near the core than the pelvic-fin spines. The vascularized core left the readers unsure at times whether the first visible annulus was the age-1 or the age-2 annulus. This uncertainty led to decreased precision and sys- tematic bias in the age estimates. Scales have been a primary structure for nonlethal age determination in many other fish species (e.g., Penttila and Dery, 1988; Welch et al., 1993; Secor et al., 1995; Elzey et al.^), but our data support the find- Elzey, S. P., K. J. Trull, and K. A. Rogers. 2015. Massa- chusetts Division of Marine Fisheries Age and Growth Labo- ratory: fish aging protocols. Massachusetts Div. Mar. Fish. Tech. Rep. TR-58, 43 p. [Available at website.] Consensus-based sectioned otolith age (yr) Consensus-based sectioned otolith age (yr) Final age (yr) Final age (yr) 382 Fishery Bulletin 114(4) Consensus-based opercula age (yr) Consensus-based whole otolith age (yr) 0123456789 10 11 12 Consensus-based dorsal-fin spine age (yr) Consensus-based pelvic-fin spine age (yr) Consensus-based scale age | Consensus-based sectioned otolith age (yr) Consensus-based opercula age (yr) 3 4 5 6 7 8 9 10 11 12 Consensus-based whole otolith age (yr) Figure 2 Age bias plots for tautog (Tautoga onitis) collected in Buzzards Bay, Massachusetts. Final age versus consensus-based age for each structure, as well as sectioned otolith consensus-based age versus consensus-based age for each other struc- ture are presented. Numbers within each plot represent the number of fish assigned each age. P-values are the results of Bowker’s (1948) (whole otoliths, dorsal-fin spine, pelvic-fin spine, and sectioned otoliths) or Evans and Hoenig’s (1998) (opercula and scales) test of symmetry. Elzey and Trull: Identification of a nonlethal method for aging Tautoga onitis 383 Table 2 Percent coefficient of variation (CV%) and percent agreement were used to examine the preci- sion of age estimates for tautog (Tautoga onitis) collected in 2014 in Buzzards Bay, Massachu- setts. Comparisons were made within reader for each structure, between readers for each of 2 readings of each structure, between consensus-based readings from sectioned otoliths and consensus-based readings from other structures, and between consensus-based readings from each structure and final readings. Comparison Aging structure CV% Agreement (%) Between readings, reader 1 Opercula 4.02 82.4 Whole otolith 1.58 84.9 Dorsal-fin spine 3.02 79.0 Pelvic-fin spine 1.51 89.9 Scale 5.94 65.5 Sectioned otolith 1.93 84.0 Between readings, reader 2 Opercula 2.39 80.7 Whole otolith 3.05 76.5 Dorsal-fin spine 3.12 76.5 Pelvic-fin spine 2.69 79.0 Scale 6.56 56.3 Sectioned otolith 7.55 69.5 Between readers, reading 1 Opercula 4.91 79.8 Whole otolith 2.60 79.0 Dorsal-fin spine 4.36 71.4 Pelvic-fin spine 4.68 64.7 Scale 6.99 54.6 Sectioned otolith 8.53 66.1 Between readers, reading 2 Opercula 3.85 79.0 Whole otolith 1.96 84.0 Dorsal-fin spine 3.32 76.5 Pelvic-fin spine 5.28 65.5 Scale 8.97 45.4 Sectioned otolith 3.92 73.1 Consensus-based readings from Opercula 5.55 64.7 sectioned otolith vs. consensus-based Whole otolith 2.11 81.5 readings from other structures Dorsal-fin spine 3.52 73.9 Pelvic-fin spine 2.99 79.0 Scale 8.61 48.7 Consensus-based readings from each Opercula 5.01 64.7 structure vs. final readings Whole otolith 1.17 87.4 Dorsal-fin spine 2.60 77.3 Pelvic-fin spine 1.17 92.4 Scale 8.79 47.1 Sectioned otolith 2.06 82.4 ings of Cooper (1967) and Hostetter and Munroe (1993) that indicate that scales are not suitable for use with tautog. The percentage of regenerated scales ranged from an average of 59% at age 1 to an average of 91% at age 11 (overall average: 74.3%). The large amount of regenerated scales in this species made it difficult to attain an adequate sample for age determination. Addi- tionally, annuli on scales were not well defined (Fig. 1), and discrimination between true and false annuli was problematic, all of which led to poor precision and bias toward underestimating ages of fish older than age 7 (Table 1, Fig. 2). Currently, the majority of age data for tautog is gathered from the examination of opercula (ASMFCl). Cooper (1967) and Hostetter and Munroe (1993) were both able to use marginal increment analysis to jus- tify that growth marks on opercula were deposited an- nually. We found that growth marks on opercula were distinct (Fig. 1), and we were able to achieve good pre- cision with age estimates (CV<5%) between and within 384 Fishery Bulletin 1 14(4) readers (Table 2). However, in the comparison between final ages of fish and consensus-based ages for oper- cula, the CV was slightly higher (5.01%) and we found significant bias. Furthermore, we found a similar CV (5.55%) and bias between consensus-based ages from sectioned otoliths and consensus-based ages from oper- cula. The bias we observed in our data appears to be systematic and most prevalent in fish age 4 and older (Fig. 2). The most likely explanation for such a bias would be a failure to correctly identify the first annu- lus because of the thickness of the bone in older fish. As was found by both Cooper (1967) and Hostetter and Munroe (1993), this thickening can obscure the first annulus. Otoliths can be viewed whole or cross-sectioned and have been the most reliable structure for age determi- nation in many fish species (e.g., Barnes and Power, 1984; Boxrucker, 1986; Welch et al., 1993; Secor et ah, 1995; Sipe and Chittenden, 2001; Robillard et al., 2009; Zymonas and McMahon, 2009; Stolarski and Sutton, 2013; Elzey et al., 2015). In this study, we examined whole and sectioned sagittal otoliths. Ages derived from whole otoliths provided good precision within readers, between readers, and between consensus-based ages for structures and final ages of fish (Table 2). No evi- dence of bias was observed for the consensus-based ages from whole otoliths and the consensus-based ages from sectioned otoliths or final ages (Fig. 2). Hostet- ter and Munroe (1993) found that whole otoliths were useful only in young fish because, as the otolith grew thicker, the annuli near the core of the otolith became obscured; however, we did not often encounter this problem. As the age of the fish increased and growth increments decreased, we found it increasingly more difficult to distinguish between annuli near the edge of the otolith. The oldest age assigned as a final age in this study was age 12, but ages from whole otoliths were assigned to age 11. A larger sample size that in- cludes older fish would give us the ability to determine where ages from whole otoliths diverge from ages de- termined from other, more accurate structures. Precision of the consensus-based ages from sectioned otoliths in comparison with final ages of fish was good (CV=2.06%). Although bias was detected, the percent agreement was more than 80%, indicating that the bias may have been less severe than the bias seen from oth- er structures. Because otoliths of tautog are small (~5- mm in length), cutting a section exactly through the origin and getting the sectioning plane correct is dif- ficult. If the cut is not made correctly through the core, the first annulus can be missed. Because the sections are aged with transmitted light, a section that is not perpendicular to the annual growth can lead to difficul- ties in interpreting annuli close to the edge. Both of these problems that can be encountered with sectioned otoliths can introduce bias. Before this study, age determination of tautog was based on methods that require sacrificing fish to har- vest the structures used for aging. Removal of these structures alters the appearance of the fish, thereby affecting the marketability of a species that is largely sold as whole fish. The need to kill and alter fish to obtain age data negatively affects the sample sources available and the costs associated with collecting ad- equate samples from juvenile and commercially cap- tured fish. The use of pelvic-fin spines for age deter- mination should allow samples to be taken from more diverse sources covering a wider selection of the stock of tautog. Strong annuli on the pelvic-fin spines also lead to high-precision age estimations. The availability of high-precision age data will strengthen the stock as- sessment of this species. The use of pelvic-fin spines as a primary structure for age determination would al- low researchers to more easily gather the information that is necessary to successfully manage tautog with an aged-based assessment model. Acknowledgments We thank the employees of the Massachusetts Division of Marine Fisheries that helped sample tautog in Buz- zards Bay. We also appreciate the thorough comments for revisions at the manuscript stage provided by M. Armstrong, G. Nelson, and 3 anonymous reviewers. Funding for this project was provided, in part, through the U.S. Fish and Wildlife Service Wildlife and Sport Fish Restoration Program (Massachusetts Sport Fish Restoration Grant F-68-R). This article is contribu- tion no. 66 of the Massachusetts Division of Marine Fisheries. Literature cited Barnes, M. A., and G. Power. 1984. A comparison of otolith and scale ages for western Labrador lake whitefish, Coregonus clupeaformis. Envi- ron. Biol. Fish. 10:297—299. Beamish, R. J., and G. A. McFarlane. 1987. Current trends in age determination methodol- ogy. In The age and growth of fish (R. C. Summerfelt and G. E. Hall, eds.), p. 15-42. Iowa State Univ. Press, Ames, lA. Bigelow, H. B., and W. C. Schroeder. 1953. Fishes of the Gulf of Maine. Fish. Bull. 53:1-577. Bowker, A. H. 1948. A test for symmetry in contingency tables. J. Am. Stat. Assoc. 43:572-574. Boxrucker, J. 1986. A comparison of the otolith and scale methods for aging white crappies in Oklahoma. North Am. J. Fish. Manage. 6:122-125. Burton, M. L., J. C. Potts, D. R. Carr, M. Cooper, and J. Lewis. 2015. Age, growth, and mortality of gray triggerfish {Bal- istes capriscus) from the southeastern United States. Fish. Bull. 113:27-39. Campana, S. E. 2001. Accuracy, precision and quality control in age deter- mination, including a review of the use and abuse of age validation methods. J. Fish Biol. 59:197-242. Elzey and Trul!: Identification of a nonlethal method for aging Tautoga onitis 385 Carbines, G. 2004. Age determination, validation, and growth of blue cod Parapercis colias, in Foveaux Strait, New Zea- land. N. Z. J. Mar. Freshw. Res. 38:201-214. Chang, W. Y. B. 1982. A statistical method for evaluating the reproduc- ibility of age determination. Can. J. Fish. Aquat. Sci. 39:1208-1210. Cooper, R. A. 1965. Life history of the tautog, Tautoga onitis (Linaeus), from Rhode Island. Ph.D. diss., 153 p. Univ. Rhode Is- land, Kingston, RI. 1967. Age and growth of the tautog, Tautoga onitis (Lin- naeus), from Rhode Island. Trans. Am. Fish. Soc. 96; 134-142. Elzey, S. R, K. A. Rogers, and K. J. Trull. 2015. Comparison of 4 aging structures in the American shad {Alosa sapidissima). Fish. Bull. 113:47-54. Evans, G. T., and J. M. Hoenig. 1998. Testing and viewing symmetry in contingency ta- bles, with application to readers of fish ages. Biometrics 54:620-629. Grimes, C. B., C. S. Manooch, and G. R. Huntsman. 1982. Reef and rock outcropping fishes of the outer conti- nental shelf of North Carolina and South Carolina, and ecological notes on the red porgy and vermilion snapper. Bull. Mar. Sci. 32:277-289. Hobbs, J. P. A., A. J. Frisch, S. Mutz, and B. M. Ford. 2014. Evaluating the effectiveness of teeth and dorsal fin spines for non-lethal age estimation of a tropical reef fish, coral trout Plectropomus leopardus. J. Fish Biol. 84:328-338. Hostetter, E. B., and T. A. Munroe. 1993. Age, growth, and reproduction of tautog Tautoga onitis (Labridae: Perciformes) from coastal waters of Virginia. Fish. Bull. 91:45-64. Keller Kopf, R., K. Drew, and R. L. Humphreys Jr. 2010. Age estimation of billfishes {Kajikia spp.) using fin spine cross-sections; the need for an international code of practice. Aquat. Living Resour. 23:13-23. Kopf, R. K., and P. S. Davie. 2011. Fin-spine selection and section level influence po- tential age estimates of striped marlin {Kajikia audax). Copeia 2011:153-160. Landa, J., E. Rodriguez-Marin, P. L. Luque, M. Ruiz, and P. Quelle. 2015. Growth of bluefin tuna {Thunnus thynnus) in the North-eastern Atlantic and Mediterranean based on back-calculation of dorsal fin spine annuli. Fish. Res. 170:190-198. McBride, R. S. 2015. Diagnosis of paired age agreero.ent: a simulation of accuracy and precision effects. ICES J. Mar. Sci. 72:2419-2167. Murie, D. J., D. C. Parkyn, C. C. Koenig, F. C. Coleman, J. Schull, and S. Frias-Torres. 2009. Evaluation of finrays as a non-lethal ageing method for protected goliath grouper Epinephelus itajara. En- danger. Species Res. 7:213-220. Panfili, J., H. De Pontual, H. Troadec, and P. J. Wright (eds.). 2002. Manual of fish sclerochronology, 464 p. Ifremer- IRD coedition, Brest, France. Penttila, J., and L. M. Dery. 1988. Age determination methods for northwest Atlantic species. NOAA Tech. Rep. NMFS 72, 132 p. [Available at website.] Phelps, Q. E., K. R. Edv/ards, and D. W. V/illis. 2007. Precision of five structures for estimating age of common carp. North Am. J. Fish. Manage. 27:103-105. Eobillard, E., C. S. Reiss, and C. M. Jones. 2009. Age-validation and growth of bluefish (Poma- tomus saltatrix) along the East Coast of the United States. Fish. Res. 95:65-75. Scott, W. B. and M. G. Scott. 1988. Atlantic fishes of Canada. Can. Bull. Fish. Aquat. Sci. 219, 731 p. Secor, D. H., T. M. Trice, and H. T. Hornick. 1995. Validation of otolith-based ageing and a compari- son of otolith and scale-based ageing in mark-recaptured Chesapeake Bay striped bass, Morone saxatilis. Fish. Bull. 93:186-190. Sipe, A. M., and M. E. Chittenden Jr. 2001. A comparison of calcified structures for aging sum- mer flounder, Paralichthys dentatus. Fish. Bull. 99:628- 640. Stolarski, J. T., and T. M. Sutton. 2013. Precision analysis of three aging structures for amphidromous Dolly Varden from Alaskan arctic riv- ers. North Am. J. Fish. Manage. 33:732-740. Sylvester, R. M., and C. R. Berry Jr. 2006. Comparison of white sucker age estimates from scales, pectoral fin rays, and otoliths. North Am. J. Fish. Manage. 26:24-31. Watkins, C. J., Z. B. Klein, M. M. Terrazas, and M. C. Quist. 2015. Influence of sectioning location on age estimates from common carp dorsal spines. North Am. J. Fish. Manage. 35:690-697. Welch, T. J., M. J. van den Avyle, R. K. Betsill, and E. M. Driebe. 1993. Precision and relative accuracy of striped bass age estimates from otoliths, scales, and anal fin rays and spines. North Am. J. Fish. Manage. 13:616-620. Zymonas, N. D., and T. E. McMahon. 2009. Comparison of pelvic fin rays, scales and otoliths for estiro.ating age and growth of bull trout, Salvelinus confluentus. Fish. Manage. Ecol. 16:155-164. 386 NOAA Spencer F. Baird National Marine Fishery Bulletin First U.S. Commissioner Jj Fisheries Service fy- established 1881 of Fisheries and founder of Fishery Bulletin wV, A comparison of density and length of Pacific groundfishes observed from 2 survey vehicles: a manned submersible and a remotely operated vehicle Email address for contact author: tom.laidig@noaa.gov Abstract — Visual surveys of seafloor communities in deep water are be- coming more common and provide fishery-independent abundance esti- mates that could improve stock as- sessments for some groundfish spe- cies. However, limitations of the sur- vey vehicle must be considered when developing methods. To that end, we estimated densities of demersal fish- es from 28 paired strip-transect sur- veys, using a manned submersible (a human-occupied vehicle, HOV) and a remotely operated vehicle (ROV) in 3 types of habitats (high-relief rock, low-relief mixed rock, and soft sedi- ments) at water depths of 75-315 m off central California. Differenc- es in fish detection, identification, and measurements were observed between vehicles (e.g., densities of unidentified fishes, unidentified rockfishes, and unidentified species of Sebastomus were significantly higher in ROV surveys). Species most closely associated with the sea- floor were observed at higher densi- ties in HOV surveys than in ROV surveys — a result possibly due to the greater reactions of fish to the ROV. The percentage of fish for which we could not estimate size was greater from video images collected with the ROV than from in situ observations made from the HOV. Results of our study will be useful for evaluation of the limitations and biases of these survey vehicles in assessments of demersal fishes. Manuscript submitted 8 October 2015. Manuscript accepted 16 June 2016. Fish. Bull. 114:386-396 (2016). Online publication date: 12 July 2016. doi: 10.7755/FB.114.4.2 The views and opinions expressed or implied in this article are those of the author (or authors) and do not necessarily reflect the position of the National Marine Fisheries Service, NOAA. Thomas E. Laidig (contact author) Mary M. Yoklavich Southwest Fisheries Science Center National Marine Fisheries Service, NOAA 1 10 Shaffer Road Santa Cruz, California 95060 Visual surveys of seafloor communi- ties in deep water (depths >50 m) are becoming more common, and the results are used to provide fishery- independent estimates of abundance and to improve stock assessments for some demersal fish species. All survey vehicles are associated with assumptions, biases, and limitations, which must be considered when se- lecting the type of vehicle and de- veloping a survey design. Biases in underwater visual surveys can result from the influence of illumination, noise, and movement of a vehicle on fish behavior (Stoner et ah, 2008). Changes in fish behavior caused by these influences can affect the sam- pling efficiency of survey vehicles, thereby leading to over- or under- estimation of fish abundance. Various types of vehicles have been used to conduct visual surveys of demersal fish abundance in both trawlable and untrawlable habitats. Adams et al. (1995) used a relative- ly large remotely operated vehicle (ROV) to estimate population size of several groundfish species on soft sediments, and Krieger (1993) and Krieger and Sigler (1996) estimated density of rockfishes (Sebastes spp.) surveyed with a bottom trawl and a human-occupied vehicle (HOV), also known as a manned submersible, in low-relief habitats. Abundance of rockfishes that live in high-relief rock habitats has been determined by us- ing an HOV (O’Connell and Carlile, 1993; Yoklavich et ah, 2007), an ROV (Stierhoff et ah, 2013), and a combi- nation of hydroacoustics and obser- vations from an ROV (Demer, 2012) as well as by using an ROV, towed stereo-camera sled, and catch compo- sition from a bottom trawl (Jones et ah, 2012). The capabilities and limitations of visual survey vehicles need to be considered when interpreting in- formation obtained from them for management purposes. Estimates of fish abundance from different visual survey vehicles have been compared in only a few studies. O’Connell and Carlile (1994) conducted sur- veys in Alaska, using an HOV and a MiniROVER MKH ROV (Teledyne Benthos, North Falmouth, MA); how- ever, the ROV was effective only in low-relief areas and not useful for quantitatively surveying their target ' Mention of trade names or commercial companies is for identification purposes only and does not imply endorsement by the National Marine Fisheries Service, NOAA. Laidig and Yoklavich: Density and length of Pacific groundfish.es 387 36°50'0"; 36°40'0''N- 36°30'0"N- species, the yelloweye rockfish {Sehastes ru- berrimus) in its primary habitat of high-re- lief rock. Uzmann et al. (1977) used a towed camera sled (TCS) and HOV to estimate fish densities on Georges Bank. Higher densities of flounders, hakes, and dogfish were found in the HOV surveys, and density of goose- fish was greater i.ti the TCS surveys. Cail- liet et al. (1999) used a TCS and an HOV to characterize fish species from, 3 areas off California and found that species observed in deep surveys were similar between type of vehicle, but 4 species were observed from the HOV that were not seen from the TCS at shallower depths (<2200 m). More rock- fish species v/ere identified by using an ROV then by using a TCS in Alaska (Rooper et al., 2012); however, a larger number of the measurements of fish length were estimated with the stereo cameras on the TCS than with a single camera with paired lasers on the ROV. Understanding the limitations of survey vehicles in providing accurate spe- cies detection, identification, and length measurements can help to design effective surveys with a consideration of the specific capabilities of the vehicle and to improve abundance and biomass assessments of the target species. In this study, we evaluated habitat-spe- cific capabilities of 2 visual survey vehicles (i.e., an HOV and an ROV) to detect, iden- tify, and determine the length of a variety of demersal fishes. Fish density was estimated by using each vehicle in 3 different types of seafloor habitats: Mgh-relief hard rock, low- relief mixed rock, and soft sediments. In an earlier study, we examined the reactions of fishes to these same vehicles (Laidig et al., 2013), concluding that avoidance reactions to the ROV were greater than those to the HOV and that avoidance of both vehicles was greater by fishes above the seafloor than by fishes on the seafloor. Surveying fishes that display large avoidance reactions to either vehicle could result in inaccurate estimates of density. Here we expand our work to consider the differences in fish densities in relation to reactions of the fish to each vehicle. Information from this study, and that from our previously published work, can be used to evaluate po- tential limitations and biases of these underwater ve- hicles and will be useful in developing sampling strate- gies for surveying demersal fishes. Materials and methods Fish surveys were conducted off the coast of central California, from Monterey Bay to south of Carmel Bay (Fig. 1), with a 2-person HOV and an ROV. Sampling was conducted in the fall of 2007 at depths ranging 1 Figure 1 Location of 28 paired transects of surveys conducted with a hu- man-occupied vehicle and a remotely operated vehicle to survey groundfishes off central California in 2007. Bottom inset is an en- largement of 2 sets of paired transects; the dark dashed line rep- resents the HOV track and the light solid line is the ROV track. from 75 to 315 m for both vehicles. Sampling took place from 12 October through 4 November for the HOV and from 18 November through 23 November for the ROV. All surveys were conducted during daylight hours (0800-1700), when our species of interest are known to be active (Love et al., 2002). The 2-person Delta HOV was launched from the FV Velero IV and was operated by experienced pilots from Delta Oceanographies (Torrance, CA). An experienced scientific observer accompanied 1 pilot inside the un- tethered HOV. The yellow-orange HOV measured 1.8 m tali, 4.6 m long, and from 0.4 m wide at its forward- most point to 1.1 m wide at mid-vehicle. A single 24- volt motor and propeller provided thrust. An external color video camera (custom-built by DeepSea Power and Light, San Diego, CA), with 400 lines of resolu- tion and an illumination range of 2-100,000 lux, was mounted on the starboard side above the observer’s viewport. The HOV was equipped with ten 150-watt halogen lights, 4 of which were used during surveys (1 388 Fishery Bulletin 114(4) forward light by which the pilot navigated and 3 that illuminated the starboard-side survey area). Two paral- lel lasers were mounted 20 cm apart on either side of the camera. The position of the HOV was tracked from the sup- port vessel with WinFrog integrated navigation soft- ware (Fugro Pelagos, San Diego, CA) and an ORE Trackpoint II ultra-short baseline (USED acoustic tracking system (EdgeTech, West Wareham, MA). The distance traveled was estimated with a MiniRLG2 measurement unit based on ring laser gyro technol- ogy (Teledyne CDL, Houston, TX) and a NavQuest 600 Micro Doppler Velocity Log (DVL; LinkQuest Inc., San Diego, CA) mounted externally on the HOV. An unmanned Phantom DS4 ROV (Deep Ocean En- gineering Inc., San Jose, CA) was operated by experi- enced pilots from the Southwest Fisheries Science Cen- ter (La Jolla, CA), National Marine Fisheries Service, aboard the NOAA Ship David Starr Jordan. The ROV had a yellow body and black frame and measured 1 m tall, 2 m long, and 1.4 m wide. Six electric thrusters (2 angled and 4 that were perpendicular to the sea- floor) provided propulsion for the ROV. The ROV was equipped with a forward-facing, color video camera with 470 lines of horizontal resolution and an 18x opti- cal zoom (FCB-IX47C, Sony Corp., Tokyo) and a Coolpix 995 digital still camera with 3.2 megapixel resolution and 4x optical zoom (Nikon Corp., Tokyo). Illumina- tion was provided by 2 forward-facing 250-watt Multi SeaLite halogen lights (DeepSea Power and Light) mounted on the ROV camera tray. Two sets of parallel lasers (20 and 61 cm apart) and 1 crossing laser were mounted to the front of the ROV and used to determine depth of field. The position of the ROV also was deter- mined with WinFrog software and an ORE Trackpoint H plus USBL tracking system. Strip-transect surveys were conducted with each vehicle, and the resultant video footage was recorded onto MiniDV (HOV) or DVCAM (ROV) tapes. During 10-min surveys of transects, the HOV traveled along a depth contour at an average speed of 0.5 m/s (stan- dard error [SE] 0.04) and an average altitude of 1.1 m (SE 0.04) above the seafloor. The ROV transect surveys were conducted at an average of 1.2 m (SE 0.08) above the seafloor and at an average speed of 0.6 m/s (SE 0.03), and covered a path comparable to that of the HOV. During these transect surveys, the observer inside the HOV identified and counted all fish that occurred within 2 m on the starboard side. The observer also es- timated size of fish to the nearest 5 cm in total length (TL), using the paired lasers. The observer’s comments were captured on an audio channel of the video footage for later review. Video footage collected from the HOV and ROV and digital still images collected from the ROV were re- viewed by one person in the laboratory after the sur- veys. Fish were identified to the lowest possible taxon and counted. Some species were not considered in our analyses. For instance, pelagic schooling fishes, such as the northern anchovy (Engraulis mordax), jack mack- erel {Trachurus symmetricus), and Pacific chub mack- erel (Scomber japonicus), swam around the vehicles for extended periods of time, increasing the possibility that these fish would be counted more than once. From the ROV images, fish size was estimated to the nearest 5 cm TL by using the parallel lasers. An effort was made to estimate the size of all fish in both surveys, with the exception of fish in holes, partially obscured by objects, perpendicular to the plane of the laser spots, or in mid- water (providing no target for the lasers). We also used the laser spots in the images from the ROV to estimate width of transects. The space between the paired 20- cm lasers (Measured laser width) was measured with a ruler on the video screen once during each minute of a transect survey. Transect width was estimated with the following equation: Transect width = Measured screen width Measured laser width X Laser width, where Laser width = 20 cm; and Measured screen width = the horizontal width of the video screen through the laser spots. We examined only video footage that was collected while the vehicles were traveling forward in survey mode (i.e., the vehicle was considered to be on transect). Fish were counted when they were observed within 2 m of the starboard side of the HOV (as estimated with a handheld dive sonar inside the HOV) or 2 m in front of the ROV. Brief stops by the ROV to capture still im- ages for identification of species (at a rate of about 1 image/min) and to avoid obstacles were considered to occur on transect. No fish observations were counted in instances where the seafloor was not seen in the video footage for >5 s (for example, when a vehicle transited over small canyons or the ROV was pulled backwards by the ship). Information on identification and size of fish was augmented by comments from the observer inside the HOV, and data from the ROV surveys were derived only from video and still images. We determined the amount and type of primary and secondary seafloor habitat from the video footage taken along each transect. Primary habitat covered >50% of the seafloor, and secondary habitat covered >20% of the remaining seafloor. We used 4 main habitat types: bed- rock (R; large in-place rock), boulders (B; unattached rock >25 cm), cobble (C; unattached rock <25 cm), and mud (M). We reduced the 16 habitat combinations to 3 categories: hard was any combination of rock, boulder, and cobble in both the primary and secondary group- ings (RR, RB, RC, BR, BB, BC, CR, CB, CC), mixed was a combination of mud as the primary or secondary habitat and one of the other habitat types (MR, MB, MC, RM, BM, CM), and soft was entirely mud (MM). Habitat categories were assigned to patches of distinct substratum types, delineated by time. Length of each patch was then estimated from time-specific cumulative distance (as measured with the DVL [HOV] or USBL [ROV]) along each transect. The areas of all habitat Laidig and Yoklavich; Density and length of Pacific groundfishes 389 10 m apart. If sections were discarded, the remaining sections of the survey track were combined for each vehicle to form the sample transect used for analyses. Total area for each of the 3 habitat categories (average transect widthxtotal length of each habitat category) was es- timated per transect. All 3 habitats were not encoun- tered on every transect, and some habitats were pres- ent in only small amounts. To further refine compari- sons between the HOV and ROV transect surveys, only total areas >20 m^ for each habitat category on each of the paired transects were considered in the analyses. Using this method, we used 21 of the 28 paired tran- sects for the hard habitat category, 22 transects for the mixed habitat category, and 14 transects for the soft habitat category for our comparisons. Fish densities were determined for each habitat category (hard, mixed, and soft) within each transect. Density was estimated as the total number of fish of a particular taxon per total area of a habitat category on each transect. A pairwise (by transect) unbalanced analysis of variance (ANOVA), with unequal sample size, was used to compare mean densities (of each fish taxon and of all taxa combined) between the 2 vehicles, among the 3 habitat categories, and to compare inter- actions between vehicles and habitats. This work was done with the statistical program SAS 9.4 (SAS Insti- tute Inc., Cary, NC). The distribution of residuals did not differ significantly from normal for these analyses. We used a Tukey-Kramer post- hoc test to determine which taxa accounted for significant differences. We used a Kol- mogorov-Smirnov 2-sample test to examine differences in length distributions of fishes between vehicles for each habitat. Further, we examined the differences in mean length for each habitat, using a Student’s Ctest. Results We analyzed 28 pairs of transect surveys conducted with the HOV and the ROV. The estimated width of an HOV transect was 2 m, and the average width of an ROV tran- sect was 1.9 m (SE 0.07). The total survey area (transect widthxtotal length of all transects) was 12,710 m^ for the HOV and 14,068 m^ for the ROV. For HOV surveys, hard habitat was the category with the greatest amount of survey area (56%), fol- lowed by mixed and soft habitats (26% and 18%, respectively) (Fig. 2). For the ROV, mixed habitat was most abundant category (39%), and hard and soft habitats were found in similar proportions (31% and 29 %, respectively) (Fig. 2). During HOV surveys, 4489 fish were counted, and 6480 fish were counted from the ROV. For the analy- ses, we included only 23 common taxa (i.e., taxa that accounted for at least 1% of the total number of fish observed from either the HOV or ROV), which included 4235 fish from the HOV surveys and 6253 from the ROV surveys. Total fish density was not significantly different for the 2 vehicles in surveys of any of the 3 habitat categories: hard (Tukey-Kramer; P>0.05), mixed (Tukey-Kramer; P>0.05), and soft (Tukey-Kram- er; P>0.05). Most fish occupied hard habitat (73% and 57% of all fishes observed in the HOV and ROV surveys, respec- tively; Table lA). Average density of all fish in hard habitat was 42.1 individuals/100 m^ (HOV) and 53.8 in- dividuals/100 m^ (ROV). The pygmy rockfish iSebastes wilsoni), the halfbanded rockfish {S. semicinctus), and young-of-the-year (YOY) rockfishes (Sebastes spp., 5 cm TL or less) were the 3 most abundant taxa in the HOV surveys. The most abundant taxa on ROV tran- sects were the pygmy rockfish, the halfbanded rockfish, unidentified rockfishes {Sebastes spp. >5 cm TL), and unidentified species of Sebastomus (a subgenus of 11 similar-looking species of rockfishes [Love et al., 2002], 4 members of which are shown in Figure 3). Densities of YOY, rosy iS. rosaceus), and rosethorn {S. helvomacu- latus) rockfish in hard habitat were significantly great- er on HOV transects than on ROV transects (Tukey- Kramer; P<0.05). Densities of unidentified rockfishes and unidentified Sebastomus were significantly greater in the ROV surveys than in the HOV surveys (Tukey- Kramer: P<0.01). 390 Fishery Bulletin 1 14(4) Table 1 The number and mean density of fish taxa or groups observed during visual surveys conducted with a human-occupied vehicle (HOV) and a remotely occupied vehicle (ROV) in 2007 off central California. Data are reported as (A) individual taxa and (B) combined fish groups per habitat category (hard, mixed, and soft). Common names in bold text are considered benthic taxa. Densities are the number of fish per 100 m^. An asterisk (*) indicates significantly greater fish density for that vehicle compared with the other. A superscript^ indicates taxa that are in the Sebastomus group. SD=standard deviation; YOY=young-of-the-year; TL=total length. HOV Common name Scientific name Total fish No. Hard density SD No. Mixed density SD No. Soft density SD A Individual taxa Bank rockfish Sebastes rufus 124 114 1.3 5.3 8 0.3 0.7 2 0.1 0.2 Blackeye goby Rhinogobiops nicholsii 235 119 1.9 3.9 81 4.3* 6.6 35 1.8 3.4 Bocaccio Sebastes paucispinis 97 94 1.6 5.3 3 0.2 0.7 0 0.0 0.0 Flatfishes Pleuronectiformes 108 22 0.3 0.7 26 0.8 1.5 60 3.0 3.0 Greenspotted rockfish^ Sebastes chlorostictus 89 58 0.7 1.9 27 1.0 1.8 4 0.2 0.3 Greenstriped rockfish Sebastes elongatus 47 29 0.4 1.8 9 0.4 1.7 9 0.5 1.3 Hagfishes Eptatretus spp. 157 3 0.0 0.1 5 0.1 0.3 149 4.3 15.6 Halfbanded rockfish Sebastes semicinctus 779 538 8.1 14.0 207 4.7 8.6 34 1.8 5.3 Pink seaperch Zalembius rosaceus 16 8 0.1 0.2 6 0.3 0.6 2 0.1 0.3 Poachers Agonidae 77 9 0.1 0.3 39 0.9 1.9 29 1.6 2.2 Pygmy rockfish Sebastes wilsoni 971 919 11.6 15.4 50 1.0 2.2 2 0.1 0.2 Rosethorn rockfish' Sebastes helvomaculatus 99 58 0.9* 1.4 38 0.9* 1.3 3 0.2 0.1 Rosy rockfish' Sebastes rosaceus 128 111 1.7* 2.1 15 0.6* 0.9 2 0.1 0.6 Splitnose rockfish Sebastes diploproa 145 43 1.8 5.6 91 1.7 5.2 11 0.5 1.2 Squarespot rockfish Sebastes hopkinsi 171 156 2.1 2.8 14 0.5 1.5 1 0.0 0.1 Starry rockfish' Sebastes constellatus 49 44 0.6 1.1 5 0.2 0.5 0 0.0 0.0 Thornyheads Sebastolobus spp. 77 8 0.2 0.8 43 0.8 1.6 26 1.4 3.6 Unidentified fishes Osteichthyes 24 13 0.2 0.5 5 0.1 0.4 6 0.6 1.6 Unidentified rockfishes Sebastes spp. 60 44 0.5 0.6 14 0.2 0.5 2 0.1 0.5 Unidentified Sebastomus^ Sebastes spp. 101 79 1.1 1.0 19 0.6 0.9 3 0.2 0.5 Widow rockfish Sebastes entomelas 73 73 0.7 2.7 0 0.0 0.0 0 0.0 0.0 Yellowtail rockfish Sebastes flavidus 79 78 1.0 1.9 1 0.1 0.3 0 0.0 0.0 YOY rockfishes Sebastes spp. (YOY) 529 473 5.2* 9.6 34 1.7 6.3 22 1.0 1.2 Total 4235 3093 42.1 740 21.4 402 17.6 B Combined fish groups All benthic fish 1167 540 8.0 6.6 307 10.4* 6.9 320 14.8 17.6 All large fish (>30 cm TL) 496 436 5.8* 9.6 59 1.6 2.7 1 0.0 0.1 All Sebastomus Sebastes spp. 466 350 4.7 3.4 104 3.3 1.9 12 0.7 0.7 Table continued In mixed habitats, 740 fish were observed and counted in the HOV surveys (an average of 21.4 in- dividuals/100 m^) and 2208 fish were counted in the ROV surveys (27.5 individuals/100 m^; Table lA). The halfbanded rockfish was the most abundant taxon iden- tified from each vehicle in mixed habitats (an average of 4.7 individuals/100 m^ on HOV transects and 11.4 individuals/100 m^ on ROV transects). The densities of blackeye goby (Rhinogobiops nicholsii) and rosethorn and rosy rockfish in mixed habitats were significantly greater in HOV surveys than in ROV surveys (Tukey- Kramer: P<0.05). Densities of unidentified fishes (Os- teichthyes), unidentified Sebastomus, and unidentified rockfishes were significantly greater on ROV transects than on HOV transects (Tukey-Kramer: P<0.05). The lowest density of fish in surveys from both vehi- cles occurred in soft habitats (17.6 individuals/100 m^ for the HOV surveys and 11.8 individuals/100 m^ for the ROV surveys; Table lA). Hagfishes {Eptatretus spp., most likely Pacific hagfish [E. stoutii]), flatfishes (Pleu- ronectiformes), halfbanded rockfish, blackeye goby, and poachers (Agonidae) were relatively abundant in soft habitat on both HOV and ROV surveys. Unidentified Sebastomus was the only group having significantly greater density (Tukey-Kramer; P<0.05) on the ROV surveys than on HOV surveys over soft habitats. Densities of rosy rockfish and unidentified rockfish differed significantly between the 2 survey vehicles in the 3 habitat categories (i.e., there was a significant interaction between survey vehicle and habitat; un- Laidig and Yoklavich: Density and length of Pacific groundfishes 391 Table 1 continued Common name Scientific name Total fish No. Hard density SD No. ROV Mixed density SD No. Soft density SD A Individual taxa Bank rockfish Sebastes rufus 48 4 0.3 1.3 42 0.7 3.3 2 0.1 0.2 Blackeye goby Rhinogobiops nicholsii 158 42 0.8 1.3 59 1.1 1.8 57 1.4 2.8 Bocaccio Sebastes paucispinis 32 28 0.7 2.5 4 0.1 0.3 0 0.0 0.0 Flatfishes Pleuronectiformes 81 2 0.0 0.2 14 0.4 0.6 65 1.9 1.8 Greenspotted rockfish* Sebastes chlorostictus 73 24 0.5 0.8 37 0.8 1.3 12 0.4 0.7 Greenstriped rockfish Sebastes elongatus 45 6 0.2 0.5 23 0.3 0.8 16 0.3 0.5 Hagfishes Eptatretus spp. 44 0 0.0 0.0 5 0.2 0.7 39 1.0 4.9 Halfbanded rockfish Sebastes semicinctus 1906 729 15.2 38.7 1080 11.4 24.1 97 1.0 2.1 Pink seaperch Zalembius rosaceus 67 5 0.1 0.1 57 1.3 5.6 5 0.2 0.4 Poachers Agonidae 72 2 0.0 0.3 20 0.6 1.3 50 1.3 1.6 Pygmy rockfish Sebastes wilsoni 2090 1928 20.8 60.0 159 3.0 5.8 3 0.1 0.2 Rosethorn rockfish* Sebastes helvomaculatus 12 6 0.2 0.6 6 0.1 0.3 0 0.0 0.0 Rosy rockfish* Sebastes rosaceus 42 27 0.4 0.6 13 0.2 0.4 2 0.1 0.2 Splitnose rockfish Sebastes diploproa 88 11 1.0 3.9 32 0.8 2.4 45 1.2 4.3 Squarespot rockfish Sebastes hopkinsi 104 59 1.0 2.0 44 0.9 3.2 1 0.0 0.1 Starry rockfish* Sebastes constellatus 22 16 0.3 0.3 6 0.1 0.3 0 0.0 0.0 Thornyheads Sebastolobus spp. 33 4 0.2 0.6 16 0.4 0.9 13 0.3 0.9 Unidentified fishes Osteichthyes 195 133 1.8 4.3 38 0.6* 0.9 24 0.5 0.7 Unidentified rockfishes Sebastes spp. 320 164 3.9* 3.7 102 1.8* 2.8 54 1.1 1.9 Unidentified Sebastomus^ Sebastes spp. 307 162 3.9* 4.0 122 2.1* 2.6 23 0.8* 1.2 Widow rockfish Sebastes entomelas 14 14 0.3 0.6 0 0.0 0.0 0 0.0 0.0 Yellowtail rockfish Sebastes flavidus 71 55 1.0 2.0 16 0.3 0.8 0 0.0 0.0 YOY rockfishes Sebastes spp. (YOY) 429 113 1.2 4.2 313 0.3 1.2 3 0.1 0.2 Total 6253 3534 53.8 2208 27.5 511 11.8 B Combined fish groups All benthic fish 889 291 6.4 4.2 321 6.4 4.8 277 7.8 5.2 All large fish (>30 cm TL) 211 81 1.9 2.4 94 1.7 2.1 36 1.0 1.9 All Sebastomus Sebastes spp. 456 235 5.0 4.3 184 3.4 3.4 37 1.1 1.6 balanced ANOVA: P<0.05). The interaction effect was based on the observations of greater densities in the HOV surveys for rosy rockfish and in the ROV sur- veys for unidentified rockfish in both hard and mixed habitats and based on the observations of low densi- ties of each of these taxa in soft sediments in surveys conducted with both vehicles. There was no significant interaction among vehicles and habitat types in our comparisons of densities of the other taxa. We grouped taxa typically occurring on the seafloor (Love et al., 2002; Love, 2011) into a category called benthic fish (bolded common names in Table lA). We grouped them in this way because, in a related study, we found that fishes living on the seafloor reacted least to both vehicles than Ashes that occurred above the sea- floor (Laidig et al., 2013). Benthic fish represented 25% of the total number of fish observed in the HOV survey and about 14% of all fish seen in the ROV survey (Table IB). Densities of benthic fish were similar for both the vehicles in surveys of hard and soft habitats, but there were significantly more benthic fish in mixed habitat in the HOV survey (Tukey-Kramer: P<0.01). We grouped the 4 identified species of Sebastomus (i.e., the greenspotted [S. chlorostictus], starry [S. con- stellatus], rosethorn, and rosy rockfish) with the un- identified Sebastomus to investigate the degree of dif- ficulty in identification of these similar-looking species (Fig. 3). Densities for the category “all Sebastomus” were not significantly different from densities for any of the 3 habitat categories in either the HOV or ROV surveys (Table IB). Most fish in the HOV and ROV surveys were small (<15 cm TL), a size group that represented about 73% of all fish with length estimates observed from the HOV (3027 of 4146 fish) and 85% of fish seen from the ROV (4706 of 5537 fish; Fig. 4). The most abundant groups with fish <15 cm TL in surveys from both ve- hicles were pygmy, halfbanded, and YOY rockfishes. Pygmy rockfish accounted for 70% of the 5-cm-TL fish and 51% of 10-cm-TL fish from the ROV transects but for only 26% of 5-cm-TL fish and 40% of 10-cm-TL fish from the HOV transects. Most small fish were observed in hard habitats (79% and 62% of all fish with length estimates on HOV and ROV transects, respectively). 392 Fishery Bulletin 114(4) Figure 3 Underwater images of 4 similar-looking species of rockfish in the subgenus Sebastomus: (A) rosethorn rockfish (Sebastes helvomaculatus) identified by green pigmentation along dorsal surface (photograph by J. Butler); (B) greenspotted rockfish {S. chlorostictus), identified by green spots along the dorsal surface (photograph by J. Field); (C) rosy rockfish (S. rosaceus), identified by purple on the head (photograph by L. Snook); and (D) starry rockfish (S. constellatus), identified by the numerous tiny white speckles on the body (photograph by R. Starr). The lowest number of small fish (mostly blackeye goby and halfbanded rockfish) occurred on soft sediments (5% of all fish with length estimates on both HOV and ROV transects). The group called large fish (>30 cm TL) accounted for 12% and 4% of all fish observed from the HOV and the ROV (Table IB). Bocaccio (Sebastes paucispinis; n=97), as well as bank (S. rufus; /t=108), yellowtail (S. flavidus; n=67), and widow (S. entomelas; n=63) rock- fish were abundant large fish in the HOV surveys, and the most abundant large fish in the ROV surveys were unidentified Sebastomus {n=54), as well as yellowtail (n=47), greenspotted (n.=:34), and splitnose (Sebastes diploproa; n=29) rockfish. The density of large fish in hard habitats was 3 times greater on HOV transects than on ROV transects (Tukey-Kramer: F<0.05). Fish length could not be estimated for 89 fish rep- resenting 10 taxa (2% of all fish) in the HOV surveys and for 716 fish in 22 taxa (11% of all fish) in the ROV surveys (Table 2). The most abundant taxa among fish of unknown size in the HOV surveys were hagfishes (n=56), unidentified rockfishes (/i=ll), and unidenti- fied Sebastomus (n=7). For the ROV surveys, the most abundant taxa without size estimates were unidenti- fied rockfishes (n.=191), halfbanded rockfish (n=180), and unidentified Sebastomus (n-lO). The greatest per- centage of fish without size estimates was from surveys with both vehicles in soft habitat (14% of fish from the HOV surveys and 23% of fish from the ROV surveys). This finding was mainly a result of hagfishes that could not be measured because they were observed in partial view (i.e., in holes or under rocks). However, disregarding hagfishes, lengths of 15% of other fish in soft habitats (e.g., unidentified rockfishes and other fish species, and halfbanded rockfish hovering above the seafloor) could not be estimated from the ROV im- ages compared with <1% of fish from the HOV surveys. Length distributions of fish were significantly differ- ent for both vehicles for each habitat category (Kol- mogorov-Smirnov test: P<0.005). The HOV and ROV surveys yielded a different average fish size for hard habitats (Student’s i-test: P<0.03; 17.0 cm TL for HOV transects and 18.0 cm TL for ROV transects) and for soft habitats (Student’s ^-test: P<0.005; 15.1 cm TL for HOV surveys and 16.7 cm TL for ROV surveys) but av- erage fish size was not significantly different for mixed habitats (16.4 cm TL for HOV transects and 17.0 cm TL for ROV transects). Discussion The ability to accurately identify fishes is a necessity when conducting meaningful visual surveys underwa- ter. Without the fish in hand, some taxa are difficult to identify to species (e.g., 11 similar-looking Sebastomus rockfishes co-occur on the central California coast) and others are practically impossible to discern (e.g., the small cryptic species of poachers [particularly the 4 species of the deep-dwelling genus Xeneretmus]). Densi- Laidig and YokSavich: Density and length of Pacific groundfishes 393 ties of unidentified rockfishes, unidentified Sebastomus, and unidentified fishes were all significantly greater in the ROV surveys than in the HOV surveys. In a study off southern California v/ith the use of the same ROV as that used in this study (Demer, 2012), 12% of all rockfishes could not be identified to species. Rockfishes also were difficult to identify by using an ROV off Brit- ish Columbia (8% unidentified; Du Preez and Tunni- cliffe, 2011) and Alaska (9% unidentified; Rooper et al., 2012), areas where there are far fewer species of rock- fishes than off the central California coast (Love et aL, 2002). Our ability to identify species from video images should improve with the availability of advanced cam- era technology (e.g., light-field cameras in which an ar- ray of lenses can collect information, such as distance to, and size of, targets) and increased video resolution (moving from high definition with 1440 lines of resolu- tion to ultra-high definition with 2000, 4000, 8000, or 16,000 lines). Nonetheless, not all fish can be identified to species by observers inside an HOV. Unidentified rockfish (not including YOY rockfishes) varied from <1% to 4% of all fish observed by using HOVs in benthic surveys off Oregon and California (Pearcy et aL, 1989; Stein et al., 1992; Yoklavich et aL, 2002); unidentified rockfishes ac- counted for 5% of all fishes in the HOV surveys in our study. Although HOV surveys have relatively low num- bers of unidentified rockfishes, identification to species can be extremely difficult, no matter what visual survey vehicle is used, in areas of high rockfish diversity and high numbers of small individuals. For example, Love et al. (2009) counted more than 700,000 fish from a minimum of 137 species (with at least 50 Sebastes spe- cies) using an HOV off southern California. Because of the high diversity and small size of the fish (over 60% of fish were <15 cm TL) in that region, many rockfish could not be identified to species (unidentified rockfish and unidentified Sebastomus composed 15% and 3% of all fish, respectively). Our ability to accurately estimate the length of fish was limited when the ROV was used. In particular, without a reference surface, lasers on the ROV were not helpful as a measurement tool for fish that hov- ered or swam above the seafloor and for fish that were oriented perpendicular to the laser spots. Rochet et al. (2006) examined accuracy of lasers to measure fish and concluded that the major difficulty in measuring fish lengths was caused by the orientation and posi- tion of fish. Measuring fish with laser arrays also has proved problematic in other ROV surveys (Johnson et al., 2003; Rooper et al., 2012). However, the increasing use of stereo camera systems on visual survey vehicles is improving the accuracy of underwater measurements of fish, including fish in the water column that cannot be measured by using lasers alone. In a study of rock- fishes in Alaska, researchers were able to measure 35% of all fish by using a towed stereo-camera sled, com- pared with 10% of the fish observed from an ROV with paired lasers and a single camera (Rooper et al., 2012). The presence of an in situ observer, whether in an HOV in deep water or with scuba at shallow depths, is more effective than using only video footage when de- tecting, identifying, and measuring fish species. The in situ observer in an HOV has a 3-dimensional view and a wider depth of field than the depth of the view from a 2-dimensional video monitor. The in situ view allows the observer to distinguish a cryptic fish from the back- ground more easily than a fish in the same background seen in a video image. As the HOV passes an area, the observer can look in multiple directions in contrast to 394 Fishery Bulletin 114(4) Table 2 Number of all fish observed, number of unmeasured fishes observed, and proportion of all observed fish that were unmea- sured from surveys of groundfish conducted over hard, mixed, and soft habitats with a human-occupied vehicle (HOV) and a remotely operated vehicle (ROV) off central California in 2007. HOV ROV All habitats Hard Mixed Soft All habitats Hard Mixed Soft Total unmeasured 89 23 10 56 716 283 315 118 Total fish 4235 3093 740 402 6253 3534 2208 511 Proportion unmeasured (%) 2 1 1 14 11 8 14 23 the single, fixed direction of a video camera. This prac- tice creates greater opportunity to detect and identify fishes within the transect area. For example, Marliave and Challenger (2009) conducted paired strip-transect surveys with scuba off British Columbia, in which one diver surveyed by eye and the other used a video cam- era. They found that fish counts per dive hour esti- mated in situ by the divers (i.e., by eye) were as much as double the counts estimated per dive hour from the diver’s video footage. They attributed this difference to the ability of divers to survey in multiple directions (forward, left, and right) and thereby could detect fish more easily within the transect area. Further, Stein et al. (1992) and O’Connell and Carlile (1994) suggested that video and still images were not as effective as the human eye for accurate fish identifications. Estimating size of fish also can be improved with an in situ observer in the water (by using either scuba or an HOV). Length of fish in midwater is nearly impos- sible to estimate from a video image, but a human can use natural stereo vision and the laser reference dots to help estimate size. In our study, lengths were unknown for only 2% of fish in the HOV surveys, compared with 11% in the ROV surveys. Advances in video technol- ogy have increased the ability to detect, identify, and estimate the size of fish. Interestingly, most of these advances, including stereo cameras with increased field of view and high definition in 3-dimensional space, imi- tate attributes of the human eye. Difficulty in species identification and fish mea- surement during ROV surveys occurred in all habitat types, indicating that these issues are not habitat spe- cific. When using the ROV, we were less able to iden- tify or estimate the size of fish on soft sediment than we were in mixed or hard rock habitats. However, the greater number of fish of unknown size is partially the result of hagfishes viewed in holes. Other studies in which fish densities have been compared between ve- hicles have focused effort on a single seafloor habitat type, such as mud (Uzmann et al., 1977; Krieger, 1993; Adams et al., 1995) or rocky outcrops (O’Connell and Carlile, 1994). Our study is unique in its comparison of fish densities estimated by using an HOV and ROV in 3 different habitats. Small fish, in particular, can be difficult to detect and identify with visual survey vehicles. In a study in which fish abundance near 3 gas platforms was esti- mated with an ROV and compared with fish abundance estimated by scuba divers, Andaloro et al. (2013) found that 9 taxa of small benthic species were reported in the diver surveys but none of those taxa were observed with the ROV. Small rockfishes <20 cm TL could not be identified to species during nearshore ROV surveys in the waters of southeast Alaska (Johnson et al., 2003), and Love et al. (2009) suggested that densities of small fish taxa, such as the bluebanded goby (Lythrypnus dalli), could be underestimated in their HOV surveys. In our study, densities of small benthic fishes (such as blackeye goby, poachers, hagfishes, and YOY rockfishes) were greater from HOV surveys than from ROV sur- veys in all habitats. This difference is also reflected in the significantly smaller sizes of fishes in HOV surveys conducted over hard and soft habitats, compared with sizes of fishes in ROV surveys. The observer inside the HOV was able to see, identify, and measure many small species that otherwise would have been difficult to de- tect in video footage alone. Consequently, an ROV may not be the vehicle of choice to assess the importance of nursery grounds, predator-prey interactions, or ecosys- tem functions, all of which require an ability to detect and identify small fish species. When selecting a survey vehicle for visual assess- ments, associated assumptions, biases, and limitations must be considered, along with logistic variables, such as cost and availability of the tools and optimal survey design (Yoklavich et al., 2015). For example, bias in es- timating abundance can result from fish avoidance or attraction to particular survey vehicles (Adams et al., 1995; Trenkel et al., 2004; Lorance and Trenkel, 2006). Laidig et al. (2013), using the same HOV and ROV as those used in our study, assessed the reactions of fish- es and found that more fishes reacted to the tethered ROV than to the autonomous HOV (57% versus 11% avoidance, respectively). They also found that the pro- portion of fishes residing on the seafloor that reacted negatively to the vehicle was 2% for the HOV compared with 18% for the ROV. Laidig and Yoklavich: Density and length of Pacific groundfishes 395 Considering this result, we expected higher fish densities in the HOV surveys. Indeed, blackeyed goby and YOY, rosethorn, and rosy rockfish had significantly higher densities in the HOV surveys, and only the un- identified taxa had significantly higher densities in the ROV surveys. Additionally, benthic species (as a group) had higher densities in the HOV surveys. Results from our study, and those from our companion work on fish reactions (Laidig et aL, 2013), indicate that abundance of at least some demersal fish species will be under- estimated depending on which vehicle is used for a survey. This type of information improves understand- ing of the limitations and biases associated with visual survey methods used to identify and count fish species underwater. Acknowledgments We thank R. Starr, a co-principal investigator of the HOV cruise, J. Butler and S. Mau for operating the ROV, the pilots and crew of the Delta HOV, and the crews of the FV Velero IV and the NOAA Ship David Starr Jordan. We thank M. Love, M. Nishimoto, T. O’Connell, L. Krigsman, and D. Watters for help with data collection; D. Watters also assisted with data analysis. We thank C. Rooper and K. Stierhoff for their thoughtful reviews of this manuscript. Underwater im- ages of Sebastomus were taken by J. Butler, J. Field, L. Snook, and R. Starr. This study was funded in part by a grant from the California Ocean Protection Council to R. Starr and M. Yoklavich. Literature cited Adams, P. B., J. L. Butler, C. H. Baxter, T. E. Laidig, K. A. Dahlin, and W. W. Wakefield. 1995. Population estimates of Pacific coast groundfishes from video transects and swept-area trawls. Fish. Bull. 93:446-455. Andaloro, F., M. Ferraro, E. Mostarda, T. Romeo, and P. Consoli. 2013. Assessing the suitability of a remotely operated ve- hicle (ROV) to study the fish community associated with offshore gas platforms in the Ionian Sea: a comparative analysis with underwater visual censuses (UVCs). Hel- gol. Mar. Res. 67:241-250. Cailliet, G. M., A. H. Andrews, W. W. Wakefield, G. Moreno, and K. L. Rhodes. 1999. Fish faunal and habitat analyses using trawls, cam- era sleds and submersibles in benthic deep-sea habitats off central California. Oceanol. Acta 22:579-592. Demer, D. A. 2012. 2007 survey of rockfishes in the Southern Califor- nia Bight using the collaborative optical-acoustic survey technique. NOAA Tech. Memo. NMFS-SWFSC-498, 110 p. Du Preez, C., and V. Tunnicliffe. 2011. Shortspine thornyhead and rockfish (Scorpaeni- dae) distribution in response to substratum, biogenic structures and trawling. Mar. Ecol. Prog. Ser. 425:217- 231. Johnson, S. W., M. L. Murphy, and D. J. Csepp. 2003. Distribution, habitat, and behavior of rockfishes, Se- bastes spp., in nearshore waters of southeastern Alaska: observations from a remotely operated vehicle. Environ. Biol. Fish. 66:259-270. Jones, D. T., C. D. Wilson, A. De Robertis, C. N. Rooper, T. C. Weber, and J. L. Butler. 2012. Evaluation of rockfish abundance in untrawlable habitat: combining acoustic and complementary sam- pling tools. Fish. Bull. 110:332-343. Krieger, K. J. 1993. Distribution and abundance of rockfish determined from a submersible and by bottom trawling. Fish. Bull. 91:87-96. Krieger, K. J., and M. F. Sigler. 1996. Catchability coefficient for rockfish estimated from trawl and submersible surveys. Fish. Bull. 94:282-288. Laidig, T. E., L. M. Krigsman, and M. M. Yoklavich. 2013. Reactions of fishes to two underwater survey tools, a manned submersible and a remotely operated vehicle. Fish. Bull. 111:54-67. Lorance, P., and V. M. Trenkel. 2006. Variability in natural behaviour, and observed reac- tions to an ROV, by mid-slope fish species. J. Exp. Mar. Biol. Ecol. 332:106-119. Love, M. S., M. Yoklavich, and L. Thorsteinson. 2002. The rockfishes of the northeast Pacific, 405 p. Univ. Calif. Press, Berkeley, CA. Love, M. S., M. Yoklavich, and D. M. Schroeder. 2009. Demersal fish assemblages in the Southern Califor- nia Bight based on visual surveys in deep water. Envi- ron. Biol. Fish. 84:55-68. Love, M. S. 2011. Certainly more than you want to know about the fishes of the Pacific coast, 672 p. Really Big Press, San- ta Barbara, CA. Marliave, J., and W. Challenger. 2009. Monitoring and evaluating rockfish conservation areas in British Columbia. Can. J. Fish. Aquat. Sci. 66:995-1006. O’Connell, V. M., and D. W. Carlile. 1993. Habitat-specific density of adult yelloweye rockfish Sebastes ruberrimus in the eastern Gulf of Alaska. Fish. Bull. 91:304-309. 1994. Comparison of a remotely operated vehicle and a submersible for estimating abundance of demersal shelf rockfishes in the eastern Gulf of Alaska. North Am. J. Fish. Manage. 14:196-201. Pearcy, W. G., D. L. Stein, M. A. Hixon, E. K. Pikitch, W. H. Barss, and R. M. Starr. 1989. Submersible observations of deep-reef fishes of Heceta Bank, Oregon. Fish. Bull. 87:955-965. Rochet, M.-J., J.-F. Cadiou, and V. M. Trenkel. 2006. Precision and accuracy of fish length measurements obtained with two visual underwater methods. Fish. Bull. 104:1-9. Rooper, C. N., M. H. Martin, J. L. Butler, D. T. Jones, and M. Zimmermann. 2012. Estimating species and size composition of rockfish- es to verify targets in acoustic surveys of untrawlable areas. Fish. Bull. 110:317-331. 396 Fishery Bulletin 114(4) Stein, D. L., B. N. Tissot, M. A. Hixon, and W. Barss. 1992. Fish-habitat associations on a deep reef at the edge of the Oregon continental shelf. Fish. Bull. 90:540-551. Stierhoff, K. L., J. L. Butler, S. A. Mau, and D. W. Murfin. 2013. Abundance and biomass estimates of demersal fish- es at The Footprint and Piggy Bank from optical surveys using a remotely operated vehicle (ROV). NOAA Tech. Memo. NMFS-SWFSC-521, 45 p. Stoner, A. W., C. H. Ryer, S. J. Parker, P. J. Auster, and W. W. Wakefield. 2008. Evaluating the role of fish behavior in surveys con- ducted with underwater vehicles. Can. J. Fish. Aquat. Sci. 65:1230-1243. Trenkel, V. M., R. I. C. C. Francis, P. Lorance, S. Mahevas, M.- J. Rochet, and D. M. Tracey. 2004. Availability of deep-water fish to trawling and visual observation from a remotely operated vehicle (ROV). Mar. Ecol. Prog. Ser. 284:293-303. Uzmann, J. R., R. A., Cooper, R. B. Theroux, and R. L. Wigley. 1977. Synoptic comparison of three sampling techniques for estimating abundance and distribution of select- ed megafauna: submersible vs camera sled vs otter trawl. Mar. Fish. Rev. 39(12): 11-19. Yoklavich, M., G. Cailliet, R. N. Lea, H. G. Greene, R. Starr, J. de Marignac, and J. Field. 2002. Deepwater habitat and fish resources associated with the Big Creek Ecological Reserve. CalCOFI. Rep. 43:120-140. Yoklavich, M. M., M. S Love, and K. A. Forney. 2007. A fishery-independent assessment of an overfished rockfish stock, cowcod (Sebastes levis), using direct ob- servations from an occupied submersible. Can. J. Fish. Aquat. Sci. 64:1795-1804. Yoklavich, M., J. Reynolds, and D. Rosen. 2015. A comparative assessment of underwater visual survey tools: results of a workshop and user question- naire. NOAA Tech. Memo. NMFS-TM-SWFSC-547, 44 p. 397 NOAA Spencer F. Baird irj'i National Marine Fisheries Service Fishery Bulletin fy- established 1881 ■90 m) and colder waters (7-8°C). Individuals that had skipped spawning had a more in- tense feeding activity and a better nutritional condition (7^=0.68-0.75) than females collected in the main spawning area according to the Ful- ton’s condition index. In contrast, postspawning females showed the poorest condition (7^=0.62-0.68) be- cause of the energy cost involved with reproduction. Females that had skipped spawning were mostly young individuals with a modal age of 3 years and a modal size of 38 cm TL. These results indicate that a sig- nificant proportion of females that had completed their first annual spawning could skip the next spawn- ing event and stay on the periphery of the reproductive area to feed. Manuscript submitted 2 March 2016. Manuscript accepted 29 June 2016. Fish. Bull. 114:397-408 (2016). Online publication date: 26 July 2016. doi: 10.7755/FB.114.4.3 The views and opinions expressed or implied in this article are those of the author (or authors) and do not necessarily reflect the position of the National Marine Fisheries Service, NOAA. Gustavo J. Macchi (contact author) Marina V. Diaz^^ Ezequiel Leonarduzzi^ Maria Ines Militelli^'^ Karina Rodrigues^'^ Recruitment (i.e., the number of progeny that will survive each year to be incorporated into the popula- tion) depends (in most traditional fishery assessment models) directly on the abundance of the parental stock, or spawning stock biomass (SSB), calculated from the maturity ogive (Mace and Sissenwine, 1993; Rodgveller et ah, 2016). In iteropar- ous fish species that spawn repeat- edly, it is assumed that once they have reached sexual maturity, all individuals reproduce on an annual cycle, i.e., the total adult fraction of the population. In recent years, evidence from many fish species in- dicates that this pattern does not occur for at least part of the spawn- ing stock (Rideout and Tomkiewicz, 2011). This phenomenon, known as skipped spawning (SS), implies that during the breeding season a pro- portion of the adult population does not spawn during that year. This “irregularity” in the reproductive cycle would have consequences for the estimate of the SSB because not every adult would contribute to the reproductive potential of the popula- tion. This potential issue with SSB estimates could introduce greater variability in the stock-recruitment relationship. Including the nonre- productive fish in calculations of the SSB could lead to an overestimation of the number of spawning fish and, therefore, would lead to an overesti- mation of expected recruitment. The degree of effect would depend on the fraction of the population affected by skipped spawning, the magnitude of which may vary between years. Different hypotheses explain the origin of skipped spawning, but most of them are associated with food de- ficiencies that affect energy storage before spawning activity or with un- favorable physical conditions that prevent or delay ovaries from ripen- ing (Holmgren, 2003; Jprgensen et ah, 2006; Rideout et ah, 2006; Ride- out and Tomkiewicz, 2011). It has been suggested that, in general, poor nutritional conditions may not allow fish to accumulate enough energy to support egg production in con- secutive years (Kennedy, 1953; Dutil, 1986). However, the main problem in determining the origin of this 398 Fishery Bulletin 114(4) “anomaly” in the annual spawning cycle is the lack of reliable historical information about a stock (i.e., data from a histological examination of gonads, as well as condition indices during the reproductive cycle) and on the oceanographic or physical features associated with the population data. Moreover, to be able to es- timate a reliable index that represents the fraction of the population that does not reproduce during the year, data are needed for the entire distribution of the spe- cies, collected at different times during the reproduc- tive cycle. For this reason, despite evidence of skipped spawning for several species, there are very few cases where this information has been incorporated into as- sessment models (Rideout et ah, 2011). Rideout et ah, 2005, on the basis of the morphological examination of the ovaries, have suggested that there are 3 categories of skipped spawning depending on when maturation is interrupted: retaining, reabsorbing, and resting. Skipped spawning has been associated primarily with species that have determinate annual fecundity (Rideout and Tomkiewicz, 2011), which is characterized by a fixed fecundity at the onset of the spawning sea- son, when recruitment of previtellogenic oocytes to the secondary growth stage ceases (Hunter et ah, 1992). On the other hand, in species with indeterminate an- nual fecundity, unyolked oocytes mature continuously and are spawned throughout an extended reproductive season. This pattern of oocyte development and the ex- tended breeding season of such species make it difficult to evaluate the reproductive history of a fish and to determine whether spawning has been skipped. Nev- ertheless, the identification of adult fish with ovaries in the regenerating stage during the reproductive sea- son and the presence of adults outside of the spawning area were considered evidence of skipped spawning in fish with indeterminate annual fecundity (Rideout and Tomkiewicz, 2011). The Argentine hake is a batch spawner with inde- terminate annual fecundity (Macchi et ah, 2004). In the Argentine sea, the Patagonian stock, located from 41°S to 55°S and with a biomass of about 1,000,000 metric tons, is economically the most important fishery resource according to a virtual population analysis in 2013 (Villarino and Santos^). The reproductive activ- ity of this stock occurs during austral spring and sum- mer (from November through March) peaks in January (Macchi et ah, 2004, Pajaro et al., 2005). Although on several occasions the presence of nonreproductive adult individuals has been noted during the spawning season of Argentine hake (Macchi et ah, 2004), information on the incidence of skipped spawning has not been report- ed for this species. The main goal of this study was to analyze the phenomenon of skipped spawning in the ' Villarino, M. F., and B. A. Santos. 2014. Evaluacion del es- tado de explotacion del efectivo sur de 41° S de la merluza {Merluccius hubbsi) y estimation de las captura biologica- mente aceptable para 2015. INIDEP Inf. Tec. Of 30, 39 p. Institute Nacional de Investigacion y Desarrollo Pesque- ro, Mar del Plata, Argentina. Patagonian stock of Argentine hake, from macroscopic and histological analyses of samples collected during research surveys during the main reproductive season of this stock over a period of 10 years. Specific objec- tives were 1) to determine the dominant category of skipped spawning in Argentine hake, 2) to estimate the incidence of this phenomenon in the population, and 3) to study the relationship of skipped spawning with the size and nutritional status of fish, as well as its pos- sible relationship to environmental factors. Materials and methods Sample collection and laboratory processing During 10 research surveys conducted by the Insti- tute Nacional de Investigacion y Desarrollo Pesquero (INIDEP), carried out in the north Patagonian area in Argentina between 2005 and 2014, samples of Argen- tine hake were obtained from hauls of bottom trawls. These surveys were performed during the time of peak spawning of the Patagonian stock (January) in the main reproductive area of Argentine hake and in the nursery ground of this species in San Jorge Gulf (Fig. 1). Although the area sampled during these surveys is smaller than the overall area of the distribution of the Patagonian stock of Argentine hake, we are confident that most of the adult population was sampled because individuals congregated in shallow coastal waters of the north Patagonia area during the peak spawning period (Macchi et ah, 2007). Trawling was conducted along transects regularly separated by about 37 km (20 nmi) and was oriented perpendicularly to the coastline. The same stations were visited annually. Specimens of Ar- gentine hake were captured at depths between 50 and 120 m by employing a bottom trawl with a mouth width of about 20 m, a height of about 4 m, and a net with 20-mm mesh at the inner lining of the codend. At each sampling station, salinity and temperature data were collected with an SBE 19 SeaCAT Profiler CTD^ (Sea- Bird Electronics Inc., Bellevue, WA). Data series were filtered and reduced to values of temperature and salinity to a 1 meter interval approximately. Samples of Argentine hake were weighed, and total lengths (TLs in centimeters), sex, and maturity stage (Table 1) were recorded for each fish. A visual maturity key of 5 stages was used: 1) immature, 2) developing, 3) spawning, 4) postspawning or spent, and 5) resting or recovering (Macchi and Pajaro, 2003). Because the phe- nomenon of skipped spawning involves adult females, only data from fish in maturity stages 2-5 were ana- lyzed (Table 2). In addition, complementary biological data collected from subsamples of Argentine hake were also used. These data included information on the in- dividual total weight (TW) in grams and on the degree ^ Mention of trade names or commercial companies is for iden- tification purposes only and does not imply endorsement by the National Marine Fisheries Service, NOAA. Macchi et al.: Skipped spawning in the Patagonian stock of Argentine hake (Merluccius hubbsi) 399 68“ W 67“ 66“ ^ 50 m 100 m Atlantic Ocean 43“S 44“ 45“ 46“ Figure 1 Locations where Argentine hake {Merluccius hubbsi) were collected during January from 2005 through 2014 in waters of the north Pata- gonian area off Argentina. Open circles indicate sampling locations in the area of Argentine hake reproduction, and black squares indicate sampling locations in the nursery area for juvenile Argentine hake in San Jorge Gulf. of stomach fullness (Table 2). Moreover, from theses subsamples, we collected sag- ittal otoliths for age determination (Table 2). Ages were determined by the methods described by Renzi and Perez (1992). The macroscopic maturity scale was validated by histological analysis of 11,494 gonads of adult females collected from dif- ferent trawl hauls and preserved in 10% formalin during the research cruises (Ta- ble 2). Ovaries were weighed to the near- est 0.1 g to obtain gonad weight (GW), and a portion (about 2.0 g) of each gonad was removed, dehydrated in ethanol, cleared in xylol, and embedded in paraffin. Sec- tions were cut at a 5-pm thickness and stained with Harris’s hematoxylin, fol- lowed by eosin counterstain. Histological staging of ovaries was based on the stage of oocyte development and on the occur- rence of postovulatory follicles and atre- sia, described by Macchi et al. (2004) and Brown-Peterson et al. (2011). Data analysis To estimate the incidence of skipped spawning from the macroscopic informa- tion, v/e calculated the percentage of rest- Table 1 Visual maturity scale (1-5) and microscopic characteristics of the different stages in the reproductive cycle of female Argen- tine hake {Merluccius hubbsi) collected from the Patagonian stock during 2005-2014. Comparison with the phases of repro- ductive development described by Brown-Peterson et. al. (2011). The last column shows which females are in reproduction during the season. POFs=postovulatory follicles. Brown-Peterson Reproduction Maturity stage Main histological features et al. (2011) during season 1 Immature Only oogonias and oocytes in primary growing stage. Some little cortical alveoli can be observed. Thin and transparent ovarian tunica. Immature No reproduction 2 Developing Oocytes in secondary vitellogenesis present. Atresia can be observed. Developing Reproduction Yolked oocytes, with or without POFs. Atresia can be present Spawning capable (mature) Reproduction 3 Spawning Hydrated oocjrtes, with or without POFs. Atresia can be present. Spawning capable (spawning) Reproduction 4 Postspawning Flaccid ovaries with abundant atresia of oocytes in vitellogenesis. POFs and residual hydrated oocytes may be present. Regressing Reproduction 5 Resting Only oocytes in primary growing stage. Thick ovarian tunica. Regenerating Skipped spawning 400 Fishery Bulletin 1 14(4) Table 2 Number of samples and subsamples of adult female Argentine hake (Merluccius hubbsi), as well as the number of ovaries that were analyzed for this study. Fish were obtained from the Patagonian stock during research surveys conducted off Argentina in January between 2005 and 2014. Ovaries for histological Year Samples Subsamples examination 2005 4879 1253 1263 2006 3110 1766 1388 2007 5067 1507 1253 2008 5154 1163 1203 2009 6412 1163 1223 2010 7558 1434 1192 2011 4446 1036 731 2012 7076 2112 1261 2013 7763 1947 1093 2014 5810 1944 887 ing females (Table 1, maturity stage 5) within the total number of adult females (Table 1, maturity stages 2-5) for each sample, and that percentage was weighted by the abundance of Argentine hake (number of individu- als/km^) estimated for each trawl haul. This informa- tion was used to determine the spatial distribution of females that had skipped spawning and to establish a percentage of this phenomenon in the studied area during peak spawning. These data were contrasted with the spatial distribution of females in the spawn- ing stage (with hydrated oocytes) obtained from each survey by using the same method. Size distributions of individuals that had skipped spawning and reproductive females (i.e., developing, spawning, and postspawning) obtained from the visual and histological diagnoses were analyzed. Data collect- ed during each survey were grouped in 2 categories, SS and reproductive females, which were compared by using a Kolmogorov-Smirnov test. The age data obtained from the subsamples of Ar- gentine hake were used to determine the age structure of females that had skipped spawning and reproductive females for each survey. These age distributions were compared in the same way as that used to compare to- tal lengths, but only for the period 2005-2013, because 2014 data were not available. Data on stomach contents were used to estimate Argentine hake feeding intensity associated with size, condition, and maturity of females. The percentage of stomachs with contents (SC) was estimated for females at different maturity stages, according to the visual scale, and for each survey. Individuals with everted stomachs as a consequence of pressure changes during capture were not used in this analysis. To assess female nutritional condition and its re- lationship with maturity stages (histologically deter- mined), we used the gonadosomatic index (GSI) and Fulton’s condition factor (E^ for the samples collected during each survey. Because gutted weight data were not available, female TW (without ovaries) was used, according to the following equations: GSI = (GW / {TW - GW)) X 100 (1) and K = (TW - GW) / (TLS), (2) where GW = gonad weight; and TW = total weight. The mean values of GSI and K obtained for the dif- ferent maturity stages were compared by using the Kruskal-Wallis test, after analysis for normality of the data. Generalized linear modeling was used to de- termine if year, TL, K, or SC significantly influenced the probability of skipped spawning during the season. Skipped spawning was analyzed as a variable with a binomial distribution: females in the resting stage were nonreproductive (probability of skipped spawn- ing [SS] = 1), and the other stages (developing, spawn- ing, and postspawning) were considered representa- tive of reproductive individuals (probability of SS=0). Stomach contents were analyzed as a binary variable: with (1) or without (0) content. Therefore, year and SC were modeled as categorical variables, and TL and K were modeled as continuous variables. In this study, to avoid problems of multicollinearity, age was excluded from the model because it was correlated with length. Analyses were restricted to adult fish greater than 32 cm TL because very few smaller fish were available; the length at maturity for the Patagonian stock is esti- mated to fall within 32-33 cm TL (Macchi et al., 2007). The model had a logit link function and a binomial error structure. Pseudo values of the coefficient of de- termination (r^) were calculated to compare the pro- portions of the deviation by year, TL, K, and SC. This pseudo was constructed by expressing the deviance of the model as a proportion of deviance for the null model. Pseudo was estimated as the deviance in the model individually containing only the intercept (null deviance) minus the deviance after adding the mains factors, divided by the null deviance of the model con- taining only the intercept. To analyze the effect of different variables in the model, the probability of skipped spawning was plotted against TL (by using a mean K of 0.67) and against K (by using a mean TL of 51.14 cm) for females with or without SC. This analysis was performed with data from all sampled trawl hauls, but because the pattern was similar between years, we provide the results only from 2012, because that year was the year with the most sampling information (Table 2). All statistical analyses were conducted with the statistical software R, vers. 3.2.3 (R Core Team, 2015). A principal component analysis was used to deter- mine the possible relationship between the incidence of skipped spawning and some physical environmen- Macchi et al.: Skipped spawning in the Patagonian stock of Argentine hake (Merlucdus hubbsi) 401 Figure 2 Histological sections of ovaries of Argentine hake (Merlucdus hubbsi) at different maturity stages: (A) resting or skipped spawning, (B) postspawning, (C) developing, and (D) spawning. Oocytes at the primary grov/ing stage (P), in vitellogenesis (V), and in a hydrated stage (H) are shown. Also labeled are the ovarian tunica (T), and the conditions of alpha atresia of yolked oocyte (a), and beta atresia (P). tal characteristics in the reproductive area. This study also included information on incidence of spawning, for comparison with data from females that had skipped spawning. The variables used were the percentages of females that had skipped spawning and the percent- ages of females in the spawning stage (with hydrated oocytes) and depth, temperature, and salinity estimat- ed for the surface of the water and at the seafloor at sampling stations. Data collected in the nursery area of San Jorge Gulf (see Fig. 1) were excluded from this study. This analysis was performed with statistical software InfoStat, vers. 2009 (Grupo InfoStat, Facul- tad de Ciencias Agropecuarias, Universidad Nacional de Cordoba, Cordoba, Argentina). Results Histological analysis and designation of maturity stages Histological analysis of 11,494 adult females, sampled during the peak spawning period (January) of the Pa- tagonian stock between 2005 and 2014, confirmed that of the 3 categories of skipped spawning (retaining, reabsorbing, and resting), only the resting stage was observed in Argentine hake. The main characteristic of this stage is that only oocytes in the primary growth stage can be observed in the ovaries, and there is no evidence of recent maturation (Fig. 2A). Unlike that in juveniles, the ovarian tunic in adults is markedly thick because of the completion of previous maturation cycles (Table 1). The other category of skipped spawning that has been described often for fish in the wild, reabsorb- ing or massive atresia, was not observed in Argentine hake. However, it was very common to see ovaries with oocytes in atresia, both in cortical alveoli or vitellogen- esis stages, with different degrees of incidence, but all such cases corresponded with females with evidence of postspawning (Fig. 2B). The rest of the observed matu- rity stages included developing ovaries with oocytes in the second growth phase, with or without postovulatory follicles and low atresia (Fig. 2C), and active spawning, with hydrated oocytes, with or without postovulatory follicles (Fig. 2D). 402 Fishery Bulletin 1 14(4) Rawson k 2U05 * .■■'SOm N -s' Bahia 4- ♦+ ^ . Camarones • ' , + . • . + 4- ■ • • • .•' * « 100 m 2010 + • • • • 2008 • + ' - , * :f \S' • . * • • ■ v« References 0.01 • 0.5 + • ' + eS^W 66° 64“ 62“ Figure 3 Spatial distribution of skipped spawning (gray circles) and spawning females (plus signs) of Argentine hake (Merluccius hubbsi) sampled in the north Patagonian area off Argentina during January from 2005 to 2014. The size of the symbols is proportional to the percentage of each stage weighted by the abundance of Argentine hake. On the basis of our histological findings, we grouped adult female ovaries of Argentine hake into 4 main cat- egories that represented the reproductive cycle: devel- oping, spawning (with hydrated oocytes), postspawning, and resting. Table 1 shows how the features used in our histological analysis are related to the 5 stages of the visual maturity scale used in our macroscopic anal- ysis and that the resting stage corresponds to skipped spawning. Spatial variation with skipped spawning and with spawning Macroscopic information collected during research surveys was used to analyze the annual spatial vari- ation of females that had skipped spawning during the sampling period. This analysis also incorporated the percentages of spawning females (with hydrated oocytes) for comparison of the spatial variation of fe- males that had skipped spawning and females that are in reproduction in the areas of spawning aggregation (Fig. 3). The maps show that females that had skipped spawning predominated in the external sector outside the spawning area of the Patagonian stock, with par- ticular abundance near the 100-m isobath and within the San Jorge Gulf. In January, when the main repro- ductive peak occurs, this spatial pattern remained rela- tively constant over the years. The percentages of females that had skipped spawn- ing for each survey within the area of reproductive ac- tivity of Argentine hake (see Fig. 1) ranged from 4% to 10%, but when the data obtained in the San Jorge Gulf were incorporated, the percentages of females that had skipped spawning increased, ranging in most cases between 10% and 15% (Fig. 4). In January 2006 and 2014, the percentage of such females for the whole area studied reached a value close to 22% — a level that is probably due to the very large catches of Argentine hake that were recorded in the San Jorge Gulf during those years. Size and age of females that had skipped spawning Figure 5A shows the length distributions, grouped for all years of the study that correspond with females that had skipped spawning and with females in reproductive condition (developing, spawning, and postspawning) de- termined from visual analysis. In comparisons of both Macchi et al.: Skipped spawning in the Patagonian stock of Argentine hake (Merluccius hubbsi) 403 25 O) 20 - c I 15 CO Q. (/) ■o Q> CL a 03 10 i Reproductive area O Total area O o O O 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Year Figure 4 Percentages of female Argentine hake {Merluccius hubbsi) that had skipped spawning among all females of this species collected from the Patagonian stock, es- timated for the reproductive area of this stock and for the total area sampled during the surveys conducted in January from 2005 through 2014. distributions, highly significant differences (P<0.001) were observed, and females that had skipped spawn- ing were mostly composed of smaller specimens with a modal size of about 38 cm TL. On the other hand, reproductive females showed higher frequencies of in- dividuals larger than 50 cm TL. To corroborate these results, the length distributions of females based on our histological results were also analyzed by grouping all surveys. The result was similar to that obtained by using the visual information: highly significant differ- ences were observed in comparisons of the 2 distribu- tions (P<0.001) and nonreproductive females generally smaller than 50 cm TL. The analysis of the age distribution obtained from subsamples of Argentine hake between 2005 and 2013 confirmed that females that had skipped spawning were primarily younger specimens with a modal age of 3 years and that the reproductive females were mainly represented by older individuals (Fig. 5B). Condition indices and feeding activity The mean GSI estimated for females that had skipped spawning was the lowest value given for females in a given maturity stage, followed by the GSI estimated for postspawning females, for developing females, and finally for spawning females (Table 3). The interan- nual variation in GSI values for different stages was relatively stable, with the widest range observed for spawning females. When the nutritional condition, represented by K, was considered, females that had skipped spawning had mean values significantly higher (P<0.0001) than those estimated for other maturity stages (Table 3). In comparison, female Argentine hake that had recently completed spawning showed the poorest condition. Results of the analysis of stomach fullness in adult female Argentine hake during the reproductive peak indicated that both females that had skipped spawn- ing and females that recently had completed spawn- ing (postspawning) had a higher frequency of stomachs with contents (Fig. 6) and that females in the spawning stage had the lowest percentages of feeding activity. We observed an increasing trend in the frequency of stom- achs with contents toward the end of the study period (2005-2014), particularly in the case of females that had skipped spawning. Results from the generalized linear model confirmed that year, TL, K, and SC influenced the probability of skipped spawning during the reproductive season of the Patagonian stock of Argentine hake (Table 4). However, although these variables were statistically significant, the amount of variation explained by these factors was very low. The model explained only 10% of the variability for the probability of skipped spawning. Total length had a negative effect in the model, which provided evidence of an increase of the probability of skipped spawning in small (<60 cm TL) adult females (Fig. 7A). In contrast, the effect of K on the probability of skipped spawning trended positively, indicating that females that had skipped spawning were characterized by higher values and, therefore, better condition (Fig. 7B). The effect of SC, as a binary variable, indicated that the probability of skipped spawning is higher when females have food in their stomachs, than in its relationship with both K and TL. Nevertheless, in the last case, the model results tended to merge when fe- males were larger than 60 cm TL (Fig. 7A). Skipped spawning and physical variables The first 2 principal components of the principal com- ponent analysis explained 64% of total variance (Table 5). Principal component 1 explained 42% of total vari- ance, and the variables most closely related to this component were the percentage of females that had skipped spawning and depth. Principal component 2 explained 22% of total variance, primarily the result of the percentage of spawning females and bottom tem- perature. This analysis indicated an opposite trend with the remaining studied variables (Table 5). An association between the highest proportion of females that had skipped spawning and deeper (>90 m) and colder (7-8°C) waters was observed. Conversely, spawn- ing females were primarily associated with shallower coastal waters, where temperatures near the bottom were higher (9-13°C). Surface temperature and both bottom and surface salinity seem to have had no in- fluence on the spatial distribution of female Argentine hake. Discussion By the year 2000, some authors mentioned the im- portance of considering the effect of skipped spawning 404 Fishery Bulletin 1 14(4) Table 3 Mean values and confidence intervals (CIs) of the gonadosomatic index (GSI) and Fulton’s condition factor {K) estimated for each maturity stage, assessed by visual analysis, in adult female Argentine hake {Merluccius hubbsi) collected from the Patagonian stock between 2005 and 2014. Rest- ing stage=skipped spawning. Visual maturity stage Mean GSI (Cl) Mean a: (Cl) Developing Spawning Postspawning Resting 6.55 (6.48-6.62) 17.34(19.94-17.74) 2.70 (2.64-2.76) 1.18 (1.15-1.21) 0.6724 (0.6705-0.6743) 0.6540 (0.6501-0.6579) 0.6517 (0.6481-0.6553) 0.6937 (0.6322-0.6418) . Skipped spawning Reproductive A Total length (cm) B Figure S Grouped for all years analyzed, (A) size distribution and (B) age distribution of females of Argentine hake (Merluccius hubhsi) that had skipped spawning (gray bars) and reproduc- tive (black line) Argentine hake sampled from the Patagonian stock off Argentina in January during 2005-2014. in estimates of the reproductive potential of fish stocks (Livingston et ah, 1997; Trippel, 1999; Rideout et al., 2000). However, the analysis of this phenomenon in general has been focused largely on experimental studies, under con- trolled conditions, and on the morphological de- scription of fish during this process (Rideout et al., 2005; Skjaeraasen et al., 2009). In a recent review about skipped spawning, the presence of this phenomenon was reported for at least 31 species, including marine teleosts (demersal and pelagic), freshwater fish, and anadromous and catadromous species (Rideout and Tomkiewicz, 2011). On the basis of when maturation is inter- rupted, 3 categories of skipped spawning have been suggested (Rideout et al., 2005): retaining, reabsorbing, and resting. In the first category, ovaries have completed the maturation process but eggs are not released because ovulation is stopped. This interruption usually happens in specimens kept in captivity and may be due to unfavorable physical or chemical conditions in the environment or to changes in the sex ratio or sexual behavior during the reproductive pro- cess. The second category includes those cases in which all oocytes in the growing phase are reabsorbed through massive follicular atresia after oogenesis is interrupted during vitellogen- esis. In the resting category, the ovaries of adult specimens capable of spawning during the re- productive season remain in a nonreproductive condition. Among the 3 categories described for the pro- cess of skipped spawning by Rideout et al. (2005), only the resting stage was observed in Argentine hake from the Patagonian stock in our study. Ovaries in the resting stage are characterized by the presence of oocytes in the primary growth stage and show no evidence of maturation or re- cent spawning. This diagnostic observed in adult specimens during the breeding peak in the repro- Macchi et al.: Skipped spawning in the Patagonian stock of Argentine hake (Merluccius hubbsi) 405 ^ 80 ■£ 70 a> c 60 o “ 50 'i 40 7 m average depth) sandy area interspersed with consolidated coral reef structure (deep patch-reeD, seagrass beds {Halodule spp. and Enhalus spp.), mid- lagoon zones of coral rubble, and backreef. Depths in Saipan Lagoon range from <1 m in nearshore areas to about 15 m in the main shipping channel. The lagoon serves as the primary area for a variety of recreational activities that do not involve the extraction of resourc- es, as well as the primary grounds for noncommer- cial and commercial fishing for the island of Saipan. The principal fishing gears include hook and line, free-dive fishing with a spear gun, and cast net or tala- ya. Spearfishing with scuba gear is prohibited through- out the CNMI by public law (CNMI Admin. Code § 85-30.1-401). The restrictions on net use issued by DLNR allow a total annual catch of reef fish of about 907 kg (2000 lb). The no-take Managaha Marine Con- Trianni: Life history characteristics and stock status of Lethrinus harak in Saipan Lagoon 411 Table 1 Macroscopic sex-specific gonad criteria used to assign maturity class to thumbprint emperor (Lethrinus harak) collected from Saipan Lagoon during 2005-2006. Stage Male Female Maturity class 1 Immature Gonads long and slender, threadlike, and translucent. Gonads long and slender, transparent or pinkish in color. Oocytes not discernible. Inactive 2 Developing Grayish-white in color, beginning to swell in girth. Ovaries elongated, some beginning to swell in girth. Individual oocytes not discernible. Some blood vessels apparent. Color pink to light pink. Active 3 Mature or ripe Thickened, white to beige in color, blood vessels visible, milt exudes when gonad is squeezed. For gonads with fur- ther thickening, milt exudes on slight pressure. Ovaries long and swollen with large oocytes visible through thin ovary wall. Blood vessels easily visible or disappear- ing. Color dark pink to red. Ova easily exuded upon application of pressure. Blood vessels disappearing. Color yellow to orange. Active 4 Spent Flaccid in appearance, thickened gonad wall loose and furrowed. Little to no milt exudes upon application of pressure. Ovaries flaccid, partially empty in ap- pearance, and color faded from dark yel- low to brown. Ovary wall thick. Active servation Area (MMCA) surrounds Managaha Island, a cay within Saipan Lagoon, and includes fringing reef, coral and rubble, deep patch-reef, and sand-dominated habitats. Data sources: fishery independent Sampling During 2005-2006, staff working in the DFW life history program obtained monthly samples of thumbprint emperor, using hook and line, to develop estimates of reproduction, age, and growth. Specimens were measured to the nearest 1 cm in fork length (FL) and weighed to the nearest 0.1 g. Sex of fish was de- termined and fish were assigned to a status of either immature or sexually active (mature) on the basis of macroscopic examination (Table 1). Excised whole go- nads were weighed to the nearest 0.01 g with an elec- tronic scale. To process them for reading, sagittal oto- liths were embedded in epoxy resin, cut transversely into 300-500 pm sections through the otolith core by using a low-speed saw, and were smoothed with abra- sive paper of 600-1200 grit until the annuli were visi- ble. These sections were then mounted on a glass slide with thermoplastic cement and hand polished with 0.3- pm alumina powder (Choat et al., 1996). With the use of a dissecting microscope with trans- mitted light, each sectioned and mounted otolith was read randomly 3 times for counts of annual increments between the core and outer otolith edge. The final age was decided when 2 of the readings agreed (Choat and Axe, 1996). The margins of otoliths were assessed and assigned to one of the 3 classifica- tions based on the relative stage of marginal increment formation. Margins were classified as thin if opaque material was visible but not necessarily continuous around the otolith margin; medium wide if there was a continuous increment of translucent material that was less than two-thirds complete (based on the width of the prior translucent increment) and visible on the outermost margin of the opaque increment; and wide if the marginal translucent increment was more than two-thirds complete. A classification of thin was indica- tive of a new annual mark and counted as a year. Oto- lith processing, age determination, and marginal edge analysis were conducted by the fish aging laboratory at James Cook University, Townsville, Australia. Biomass surveys The DFW conducted underwater vi- sual census (UVC) surveys in 2004, 2007, and 2011 in the southern portion of Saipan Lagoon to mea- sure the influence of restrictions on net use imposed in December 2003. In these surveys, 4 major lagoon habitat units were surveyed on the basis of dominant substrate by using belt transects of 25x5 m (counts of 10-15 min) for fish <20 cm FL and by using station- ary point counts with a 10-m radius (8-min duration) for fish >20 cm FL. For these surveys, standard DFW fishery-independent UVC techniques were used and the surveys were conducted by using a proportional al- location design (Cochran, 1977). Additionally, the same UVC survey techniques were used to estimate biomass of thumbprint emperor for the deep patch-reef habitat in the MMCA. Results from these surveys were used to obtain estimates of total biomass of thumbprint em- peror from Saipan Lagoon for 2007 and 2011 by using the following equations (Cochran, 1977): 412 Fishery Bulletin 1 14(4) EfxhxAand (1) = X^xH, (2) where = overall habitat mean (g/m^); = habitat unit mean (g/m^); H = total size all habitats (m^); h = habitat unit size (m^); and Y = estimated total (g). Biomass estimates of thumbprint emperor were not available from 2007 because most individuals of Lethri- nus were recorded only at the genus level. As a result, although there is uncertainty in the amount of error associated with this approach, the biomass of species of Lethrinus estimated for 2007 was multiplied by the proportion of species of Lethrinus identified as thumb- print emperor in surveys in which UVC methods were used during 2011. Data sources: fishery dependent Inshore creel survey The DFW currently operates an inshore creel survey (ICS) that targets Saipan Lagoon. The ICS collects data on nonvessel fishing activities that involve gear that includes hook and line, free-dive with a spear gun (pachinko and cast net [talaya]). For this study, length-frequency data and estimated an- nual landings of the thumbprint emperor for the years 2005-2011 were obtained from the ICS database. Exemptions from net-use restrictions In December 2003, the DLNR established regulations that prohibit the use of gill, drag, and surround nets, except as exempted by the DLNR for special cultural events, such as an- nual fiestas. Catch quotas of 68-136 kg (from 150 to 300 lb) with a 2-to-3-day time limit were established for exemptions (net-use exemptions, NUE, under an annual limit of 907 kg [2000 lb.]) For all fishing activi- ties exempted from DLNR regulations through 2011, surround nets (chenchulun umesugon) were used and the landings taken under these exemptions were moni- tored by the DFW. In this study, landings taken under the NUE were used in estimating total annual land- ings and recording length-frequency m.easurements for thumbprint emperor. Tracking commercial landings In the CNMI, the pri- mary means of tracking commercial landings of coral reef fishes is the DFW Commercial Purchase Database System (CPDS), a database that tracks commercial purchases and sales recorded on trip tickets provided by DFW and voluntarily submitted by vendors. Ad- ditionally, in 2011, a biosampling program (BSP) for coral reef fish commenced in the CNMI. Several times a week, the BSP sampled reef fish for species identi- fication, length frequency in centimeters in FL, and body weight in grams from randomly selected catches of fishermen before commercial sale. The BSP also col- lected daily data on vendor purchases from fishermen, in an effort similar to that of the CPDS. In 2011, the BSP sampled 19% of total reef fish landings from the coral reef spear fishery. The proportion of thumbprint emperor among the total biomass of coral reef fishes recorded during the BSP sampling in 2011 was used to estimate total landings of thumbprint emperor by multiplying the proportion of thumbprint emperor by the estimated CPDS annual landings of lethrinids dur- ing 2006-2011. Data analyses Size at maturity Data from the macroscopic determina- tion of sex and reproductive status were partitioned into 1-cm-FL classes for the analysis of female length at 50% maturity (L50), female age at 50% maturity (A50), and length (Lx) and age (Ax) at 50% sex transi- tion from female to male. A logistic model was used for these estimations: P = l/(l + e'“‘^^^^), (3) where a and b = the fitted model constants; P = the percent mature or transitioned in length or age class, Xf = length or age class; and L50 or A50 = -f. This model was weighted by the square root of the numbers of fish in each length class, and the regres- sion parameters were estimated by using maximum likelihood with a binomial logistic link function. Bias- corrected percentile confidence intervals were produced around all estimates from 1000 resamples drawn with replacement (R Core Team, 2015). Age and growth A nonlinear least squares model was used to fit the standard von Bertalanffy growth func- tion (VBGF) (von Bertalanffy, 1938) to length-at-age data. The VBGF is defined with the following equation: L, = L^(l-e-*''-''>’), (4) where = the length at time t; = the asymptotic length; k = the Brody growth coefficient; and fg = the theoretical age at which length is equal to 0. Data were fitted both by not constraining ^o ^-nd by constraining it to an estimated age at settlement from Japan (Nakamura, et al. 2010). Growth curves generat- ed from the VBGF fitting procedure were subsequently tested for equality by using the analysis of residual sum of squares (ARSS) method developed by Chen et al. (1992) for male and female comparisons. Parameters of the length-weight relationship were obtained by fit- ting the following power function W = a X FLb (5) to length and weight data, Trianni: Life history characteristics and stock status of Lethhnus harak in Saipan Lagoon 413 where W = the total wet weight; and a and b = the associated equation constants (R Core Team, 2015). Total mortality The expectation-maximization algo- rithm approach developed by Hoenig and Heisey (1987) was used to create length-age keys from the 2005 and 2006 life history data, and those keys were used to es- timate age groups from length-frequency data collected in subsequent years from the ICS and NUE data. The annual instantaneous rate of Z was subsequently de- termined with the age-based catch-curve (CC) method (Beverton and Holt, 1957) in its linearized form: ln(Ct) = ln(Aro)-Zi;, (6) where Ct = the catch in year t; Nq = recruitment into cohort; and Zt = the instantaneous rate of Z at age t. The descending limb of the plot of the natural loga- rithm of relative abundance versus age (time) was used to estimate Z with exclusion of derived age groups be- yond the first age class where one or fewer individuals were observed (Dunn et ah, 2002). Additionally, Z was estimated by using the maximum likelihood estimator for catch curves developed by Chapman and Robson (Chapman and Robson, 1960; Robson and Chapman, 1961): where n = the total number of fish observed on the de- scending limb of the catch curve, defined as fully recruited; and T = the fish ages on the descending limb of the catch curve, where the first fully recruited age is set to 0. Estimates of Z from the use of both methods, and both age-length keys, were generated yearly as permitted by the availability of data from the ICS, as well as the availability of the NUE data. Natural mortality Three models were used to estimate the instantaneous rate of M. For one model, the pre- dictive equation developed by Hoenig (1983) was used; the equation relates the maximum age observed in a stock to M: ln(M) = 1.46 X 1.01 X Inl^max ). (8) where = the maximum age observed in the population. Another model was Lorenzen’s (1996) power function of weight for natural populations; M = 3W-0-288, (9) where W is weight (g). Finally, the third model was the mortality estimator based on the empirical equation by Gislason et al. (2010): LogM = 0.55 - 1.611ogL + 1.44L^ log^, (10) where L = the fish fork length; = the asymptotic length; and k = the instantaneous growth coefficient. Spawning potential ratios Goodyear (1993) summarized the concept of spawning potential ratio (SPR) and eval- uated it against other traditional fisheries reference points, defining SPR as a ratio of the potential recruit fecundity (P) in an exploited stock to the potential re- cruit fecundity in an unexploited stock: SPR- , (11) ■^unfished where P = * F'n ) * exp' ’ , (12) n = the number of female length classes; E = the mean fecundity of females at length i; Fr^ = the fraction of females at length i that were mature; Fi = the fishing mortality rate of females at age t; and M = the natural mortality rate of females at age i. The SPR was estimated by calculation of P in the ab- sence of fishing mortality (P=0) and with various lev- els of fishing mortality (F>0). Length classes included the smallest mature female and the largest measured female observed during sampling by the DFW during 2005-2006. Mean fecundity per size class for thumbprint em- peror was estimated by using the fecundity relation- ship (batch fecundity=1.58x O’^xFL^^) provided by Ebisawa (1990) for the spangled emperor (Lethrinus nebulosus), a phylogenetically related species (Lo Galbo et ah, 2002) with similar trophic-level characteristics (Carpenter, 2001). The fraction of females that were mature for a specific FL was estimated from collected length-at-maturity data. The fishing mortality rate was multiplied by a selection coefficient that was as- sumed to increase linearly with length from 0.0 at 1 cm less than the smallest length sampled to 1.0 at 1 cm greater than the modal length in the length dis- tribution. Through the use of growth parameter esti- mates derived from the VBGF, M was estimated from the empirical equation of Gislason et al. (2010) and ap- plied as estimated for each length class. The ratio of estimated total annual landings of thumbprint emperor from fishery-dependent data sources over estimated biomass from the fishery-independent surveys conduct- ed in 2007 and 2011 approximated estimates of fishing mortality (Beddington and Kirkwood, 2005), and those estimates of fishing mortality were then used to assess SPR for the period 2005-2011. Exploitation The exploitation ratio (E) was computed as E=F / Z, where Z=M +F (Gulland, 1971) and F is instantaneous fishing mortality. An E value at or near 0.5 is considered indicative of full exploitation, al- 414 Fishery Bulletin 114(4) Table 2 Number of male and female fish sampled, percentage of fish that were male, and sex ratio by length bin for thumbprint emperor {Lethrinus harak) collected during 2005-2006 from Saipan Lagoon. /i=sample size. Fork length (cm) ratio(M:F) n Male Female Male (%) Sex 10.0-11.9 6 3 3 0.50 1.0:1.0 12.0-13.9 9 1 8 0.11 1.0:8.0 14.0-15.9 37 8 29 0.22 1.0:3.6 16.0-17.9 88 19 69 0.22 1.0:3.6 18.0-19.9 109 19 90 0.17 1.0:4.7 20.0-21.9 98 16 82 0.16 1.0:5. 1 22.0-23.9 116 38 78 0.33 1.0:2. 1 24.0-25.9 80 36 44 0.45 1.0:1.2 26.0-27.9 18 5 13 0.28 1.0:2.6 28.0-29.9 5 0 5 0.00 - 30.0-31.9 2 1 1 0.50 1.0:1.0 32.0-33.9 2 1 1 0.50 1.0:1.0 though Pauly (1984) suggested a more conservative value of 0.2 for lower-trophic-level, small-size fish with high recruitment variability. Although the thumbprint emperor is a higher-trophic-level species, the value of 0.2 could serve as a lower bound for E for this spe- cies. To obtain uncertainties around estimates of E, in this study, uncertainties around Z and M were first obtained by sampling with replacement (1000 resam- ples). For Z, obtaining uncertainties was accomplished by resampling within the ranges of the estimated an- nual values from both the CC and the Chapman-Rob- son (CR) methods. Estimates of M uncertainty were derived by resampling within boundaries of applicable model parameters: within the uncertainties (confidence intervals) of estimated VBGF age and growth param- eters for thumbprint emperor from this study with the Gislason model; within the range of weights (Lorenzen model) or lengths (Gislason model) observed from sam- pling thumbprint emperor from this study; within the maximum ages (Hoenig equation) of thumbprint em- peror estimated from this study and from Taylor and Mcllwain (2010) for Guam. Estimates of F were subsequently derived by resa- mpling (1000 resamples) the equivalency Z=M+F, rear- ranged as F-Z-M, by using results from the resampled M and Z estimators. All analyses were conducted with R statistical software (R Core Team, 2015). Results Reproduction The overall observed male-to-female sex ratio for the thumbprint emperor from the CNMI for the period 2005-2006 was 1.0:2. 9, and females were predominant in most size classes (Table 2). The estimated L50 and A50 for females were 19.6 cm FL and 2.6 years, respec- tively (Table 3). Values of the gonadosomatic index for female thumbprint emperor began to increase concomi- tant with a decrease in the percentage of immature in- dividuals at size, elevating noticeably at around 19-20 cm FL (Fig. 2), with fewer samples from size classes greater than 27 cm FL (Table 2). Estimates of age and length at the transition from females to males were not obtained because the expected change in sex ratio — a change to more males and fewer females that would indicate sex transition — was not observed (Table 2). Age and growth Results from the marginal edge analysis completed in 2005 and 2006 indicated that otolith accretion with a margin classification of new (opaque material visible on otolith margin), occurred from June through Octo- ber (Fig. 3, A and B). The oldest individuals collected were an age-9 male at 27.6 cm FL and an age-9 female at 29.3 cm FL. The estimates of unconstrained overall, sex-specific, and constrained VBGF parameters are pro- vided in Table 4. Growth curves for the combined un- constrained, sex-specific, and constrained model fits are displayed in Figure 4. Estimates of and k for male and female thumbprint emperor from Saipan differed substantially; females were estimated to have a larger L„ and lower k. The ARSS model used to analyze the sex-specific growth curves revealed significant differ- ences between the sexes (P=0.016) and years (P=0.004). Length frequency, biomass, and landings Percent contributions of measured thumbprint emper- or per length class generated from ICS and NUE data were graphically compared with data from female and male size histograms developed from DFW sampling in Trianni: Life history characteristics and stock status of Lethrinus harak in Saipan Lagoon 415 Table 3 Reproductive life history parameters for thumbprint emperor {Lethrinus harak) as estimated from locations in the Indo-Pacific: female size at 50% maturity (L50) with confidence intervals (CIs), female age at 50% maturity (A50 ) with CIs, length at transition from female to male (Lj), age at transition from female to male (Aj), and sex ratio (M:F). The number of male and female thumbprint emperor sampled appear in parenthesis below each location. All lengths are fork lengths measured to the nearest centimeter. Location Female L50 (Cl) Female A50 (Cl) L'y M:F Saipan (150,420) 19.6(19.1-19.9) 2.6 (2.4-2.8) - - 1.0:2.9 Guami (85,328) 20.8 3.8 24.1 5.38 1.0:3.8 Ryukyu Islands (194,254) 19.52 1.0-2.03 25.52 — 1.0:1.32 Kenya"* (386,426) 24.3 1.0:1. 1 ^Taylor and Mcllwain, 2010. ^Ebisawa, 2006. '^Ebisawa and Ozawa, 2009. ^Kulmiye et. al., 2002. Saipan Lagoon (Fig. 5). There was a distinct difference in the length classes targeted by the ICS and NUE fishermen. The ICS landed mostly smaller thumbprint emperor in the length classes from 12 to 15 cm FL, and NUE fishermen targeted fish in length classes from 22 to 25 cm FL. When length-class percentages from the different data sources were overlain, NUE data cov- ered nearly the entire male length distribution derived from DFW sampling, whereas ICS landings consisted primarily of smaller (Fig. 5) females and males. The highest percentages of thumbprint emperor in annual landings corresponded with hook-and-line landings, as estimated from the ICS. In contrast, landings of thumbprint emperor in 2011 from the BSP came from free-dive spearfishing, composing about 0.8% of the es- timated total annual BSP landings. Table 5 provides 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Figure 2 Gonadosomatic index (GSI) values, indicated with black dots, and percentag- es of individuals that were immature at a given fork length (cm), indicated by gray bars, for thumbprint emperor (Lethrinus harak) collected from Saipan Lagoon during 2005-2006. Error bars on the black dots indicate standard errors of the mean. 416 Fishery Bulletin 1 14(4) Table 4 Age and growth parameters, with standard errors of the mean (SEs), for thumbprint emperor (Lethrinus harak), as estimated for Saipan Lagoon with models that used the von Bertalanffy growth function: asymptotic length (L^), Brody growth coefficient (k), the theoretical age at which length is equal to 0 (to), and maximum age estimated (imax)- The combined, constrained model was constrained to an estimated length at settlement of 15.8 mm in fork length. All values of L„ are fork lengths given in centimeters. CI=confidence interval. n=sample size. Estimated categories n L„ (SE) k (SE) h t ‘'max Male 127 27.3 (1.6) 0.377(0.035) -1.11 9 Female 369 37.2(5.7) 0.139(0.049) -2.92 9 Combined 522 30.1 (1.4) 0.259 (0.041) -1.65 9 Combined CIs 27.2-32.9 0.180-0.340 Combined constrained 522 24.5 (0.288) 0.688 (0.024) 9 A B Figure 3 Monthly classification of otolith margins in thumbprint emperor (Lethrinus harak) collected from Saipan Lagoon during (A) 2005 (n=281) and (B) 2006 (n=337). The 3 classifications were 1) thin, opaque material visible but not necessarily continuous around otolith margin, 2) medium wide, continuous increment of translucent material visible on the outermost mar- gin of the opaque increment that was less than two-thirds complete; and 3) wide, marginal translucent increment more than two-thirds complete. Trianni: Life history characteristics and stock status of Lethrinus harak in Saipan Lagoon 417 Spawning potential ratios Approximated baseline values of F for 2005-2011 from the ratio of estimated total annual landings of thumbprint emperor from fishery-dependent data sources over estimated biomass from the 2007 and 2011 fishery-independent surveys are given in Table 5. Estimated SPR values at vary- ing levels of F along with approximated values of F and corresponding estimated SPR values from 2005 through 2011 are listed in Table 6. Estimat- ed SPR values for the period 2005-2011 ranged from 0.91 to 0.98 (Table 6). Mortality Age Figure 4 The von Bertalanffy growth curves fitted to observed fork length over age for thumbprint emperor (Lethrinus harak) collected from Saipan Lagoon during 2005-2006 (A) by sex (females, n=369; males, «=127), (B) by both sexes combined (fi=522) with confidence intervals (CIs), and (C) by both sexes combined and with the model constrained to length at settle- ment. In, graph A, open circles represent female length at age and X symbols represent male length at age. In graphs B and C, open circles represent length at age. biomass values from fishery-independent surveys in 2007 and 2011 and the estimated annual landings for the period 2005-2011 that were derived from pooling estimated landings from the ICS, NUE, and CPDS and BSP data. Estimates of Z with the CC and CR methods had similar ranges: 0.326-0.867 and 0.473-0.871, respectively. The CR model generally produced higher estimates of Z than those produced with the CC model for any given combination of source data, age-length key, and year (Table 7). The es- timates of the models were significantly different (Astat=6.57, df=17, P<0.00). In Figure 6, age fre- quencies derived from DFW samples, as well as from age-length keys for the period 2005-2006, are overlain with estimated M derived from aver- age length and average weight for the Gislason and Lorenzen estimators. The M estimates from the Gislason model were lower than the estimates derived from the Lorenzen model for all corre- sponding ages (Fig. 6). Exploitation rate Variability between the 3 estimators of M result- ed in a wide range of estimates of exploitation rate (Fig. 7). The E estimates from the use of M from the Lorenzen model were negative in 4 years when Z from the CC method was used and nega- tive for 2 years when Z from the CR method was used (Fig. 7). The E estimates were positive when Z from the CC and CR methods and M from the Gislason et al. (2010) and Hoenig (1983) methods were used. Average values of E from the 3 year- ly estimates are depicted as circles in Figure 7. These estimates of E ranged from -0.082 in 2009 (Fig. 7A) to 0.341 in 2011 (Fig. 7B). Estimates of M from the Gislason model in which VBGF parameter was constrained to size at settlement from Japan (15.8 mm FL; Na- kamura et al., 2010) produced high estimates of k, which led to high estimates of M, and no posi- tive E values for estimates derived from Z with the CC and CR methods. Discussion Fishing pressure has been found to result in truncated length and age frequencies (Friedlander and DeMar- 418 Fishery Bulletin 1 14(4) Table 5 Estimated landings (in kilograms) of thumbprint emperor (Lethrinus harak) from Saipan Lagoon for the period 2005-2011 from the Commonwealth of the Northern Mariana Islands Division of Fish and Wildlife inshore creel survey (ICS), commer- cial reef fish biosampling program (BSP), commercial purchase database system (CPDS), and records of net-use exemptions (NUE), as well as estimated fishing mortality (F) and estimated biomass from fishery-independent surveys (FIS) conducted in 2007 and 2011. Activities covered in the ICS include free-dive spearfishing (Spear) and fishing with hook and line (H&L) and cast nets. Total landings are given in kilograms and metric tons. The BSP did not begin until 2011. Year Reef fishes L. harak ICS Spear H&L Cast net BSP CPDS NUE Total (kg) Total (t) F FIS 2005 23,139 1310 595 711 4 _ 648 108 2066 2.06 0.050 41.52 2006 25,360 2493 680 1808 5 - 940 119 3552 3.55 0.085 41.52 2007 23,349 2285 536 1748 1 - 665 231 3181 3.18 0.085 37.51 2008 18,523 1915 197 1710 9 - 623 69 2607 2.61 0.063 41.52 2009 27,201 2904 236 2653 15 - 535 33 3472 3.47 0.084 41.52 2010 11,267 471 65 405 1 - 444 5 920 0.92 0.022 41.52 2011 9342 477 0 472 5 408 3 64 948 0.95 0.021 45.41 ^Estimated biomass from FIS. ^Averaged biomass from FIS estimates. ■^BPS used. tini, 2002; Graham et al., 2005; Lewin et ah, 2006) and in skewed sex ratios of hermaphroditic species (Birke- land and Dayton, 2005; Fenberg and Roy, 2008; Shep- herd et ah, 2010), and potential effects on stock recruit- ment (Alonzo and Mangel, 2004; Alonzo et ah, 2008) and ecosystem functioning (Zhou et ah, 2010; Garcia et ah, 2012). The CNMI employs a complete ban on spearfishing with scuba gear, as well as restrictions on the use of gill, drag, and surround nets. Fishermen use nets toward the central parts of Saipan Lagoon in areas dominated by soft sediments, whereas hook-and- line fishing occurs mostly from shore near seagrass Fork length (cm) I J Female I I Male — — ICS •••••• NUE Figure 5 Length-frequency histograms for female and male thumbprint emperor {Leth- rinus harak) from Saipan Lagoon during 2005-2006, measured by staff work- ing in the life history program of the Commonwealth of the Northern Mari- ana Islands Division of Fish and Wildlife (DFW), overlain with percentages of length classes from the DFW inshore creel survey (ICS) and recorded net-use exemptions (NUE) for the period 2005-2011. Trianni: Life history characteristics and stock status of Lethrinus harak in Saipan Lagoon 419 Table 6 Approximate annual values of fishing mortality (F) and estimated spawning potential ratio (SPR) for thumbprint emperor (Lethrinus harak) collected from Saipan Lagoon during 2005-2011. Values were derived from estimated total landings from fishery-depen- dent data sources and estimated biomass from fishery-indepen- dent survey and were compared with estimated SPRs at differing levels of F. Estimated from annual landings Derived Year F SPR F SPR 2005 0.05 0.95 0.00 1.00 2006 0.09 0.91 0.25 0.78 2007 0.08 0.91 0.50 0.61 2008 0.06 0.94 0.75 0.47 2009 0.08 0.92 1.00 0.37 2010 0.02 0.98 1.25 0.29 2011 0.02 0.98 1.50 0.22 1.75 0.17 2.00 0.14 2.25 0.11 beds, which have been shown to be a preferred habi- tat for juvenile thumbprint emperor in southern Japan (Nakamura et ah, 2009). The difference in length fre- quencies derived from the Saipan Lagoon ICS and in those derived from NUE data indicate that shoreline fishing targets smaller (Fig. 5) thumbprint emperors and that fishing pressure on larger thumbprint emperor was probably reduced as a result of regulations on net use. Length-frequency data in published stud- ies from Kenya (Kulmiye, et al. 2002), Guam (Taylor and Mcllwain, 2010), Ryukyu Islands (Ebisawa and Ozawa, 2009), as well as from this study in Saipan Lagoon, show that fe- males were prominent in the largest size classes and, except for females in Saipan La- goon, were the largest recorded individuals (Table 2). Kulmiye et al. (2002) concluded that identification of hermaphroditism in thumb- print emperor would require further study, but Ebisawa (2006) considered the thumb- print emperor a protogynous hermaphrodite. Sadovy de Mitcheson and Liu (2008) did not designate the thumbprint emperor as a pro- togynous hermaphrodite, and, although the seemingly complex sexuality of this species may require further inquiry, data from Guam provide evidence for protogyny in that popu- lation (Taylor and Mcllwain, 2010). Ebisawa (2006) stated that the observed difference in sex ratio in the largest size classes in the Ryukyu Is- lands, where females outnumbered males, was a result of females attaining a larger maximum size than that of males, although Ebisawa and Ozawa (2009) did not distinguish L„ between the sexes in their study. The Table 7 Estimates of total instantaneous mortality (Z) and standard errors of the mean (SEs) from analysis with the linearized catch-curve (CC) method and the Chapman-Robson (CR) method for thumbprint emperor (Lethrinus harak) from Saipan Lagoon. Values are listed by fish- ery source; inshore creel survey (ICS) or net-use exemption (NUE), year of age-length key (2005-2006), and year of data collection (2005-2011) Source Year of age- length key Year of data collection Z(CC) SE Z(CR) SE ICS 2005 2006 0.549 0.051 0.598 0.036 ICS 2005 2007 0.454 0.110 0.509 0.041 ICS 2005 2008 0.521 0.056 0.573 0.037 ICS 2005 2009 0.326 0.046 0.473 0.033 ICS 2006 2006 0.725 0.071 0.845 0.064 ICS 2006 2007 0.674 0.094 0.760 0.072 ICS 2006 2008 0.720 0.074 0.806 0.064 ICS 2006 2009 0.544 0.127 0.626 0.051 NUE 2005 2005 0.431 0.150 0.664 0.036 NUE 2005 2006 0.575 0.101 0.753 0.040 NUE 2005 2007 0.389 0.083 0.663 0.047 NUE 2005 2009 0.368 0.127 0.639 0.790 NUE 2005 2011 0.446 0.086 0.691 0.049 NUE 2006 2006 0.867 0.120 0.871 0.046 NUE 2006 2007 0.659 0.142 0.756 0.052 NUE 2006 2009 0.415 0.069 0.671 0.084 NUE 2006 2011 0.493 0.182 0.758 0.055 420 Fishery Bulletin 1 14(4) Age > CQ CD CD n c CD 3 o '< 15.0 18.4 21.1 23.3 24.6 25.8 28.3 Mean fork length (cm) 25.5 28.5 68.7 127.5 192.9 249.0 294.7 339.9 417.2 Mean weight (g) 321.0 485.1 I inFW lh ••••■•• Gislason M I 1 A-L key 2005 i 1 A-L Key 2006 — *- Lorenzen M Figure 6 Estimates of natural mortality (M) for thumbprint emperor (Lethrinus harak) from Saipan Lagoon during the period 2005-2006, derived from average weight- per-age class in the Lorenzen model and from the average length-per-age class in the Gislason model, expressed against age frequencies determined from data provided by the life history (LH) program of the Commonwealth of the Northern Mariana Islands Division of Fish and Wildlife (DFW) and age-length keys for 2005 and 2006. issue of female maximum size in the thumbprint em- peror is important given that, in general, larger female fish are exponentially more fecund and potentially more beneficial to population recruitment (Birkeland and Dayton, 2005; Fenberg and Roy, 2008). The L50 value for thumbprint emperor in Saipan Lagoon (19.6 cm FL) was lower than L50 values for thumbprint emperor from east Kenya, 23.4 cm FL (Kulmiye et al, 2002) and Guam, 20.8 cm FL (Taylor and Mcllwain, 2010). Estimated L50 from the Ryuku Islands, 19.5 cm FL (Ebisawa, 2006), was similar to the estimate from this study in Saipan Lagoon. In as- sessing regional differences in female A50, the value of A50 for thumbprint emperor in Saipan (2.6 years) was intermediate between the estimate for Guam (Taylor and Mcllwain, 2010) and the estimate for the Ryukyu Islands (Ebisawa and Ozawa 2009); see Table 2. Al- though Lt and At were not estimated for thumbprint emperor in Saipan, Taylor and Mcllwain (2010) esti- mated Lt at 24.1 cm FL and Aj at 5.38 years for this species in Guam and Ebisawa (2006) estimated Lx at 25.5 cm FL for thumbprint emperor from the Ryukyu Islands. Variability in estimates of L50 could be introduced by the differing approaches to evaluation of the matu- rity status of gonads (e.g., gonad histology vs. macro- scopic evaluation), by differences in the criteria used to classify maturity status, and by differing analytical approaches to estimation of L50 (e.g., graphical [Kul- miye et al., 2002] vs. statistical [Ebisawa, 2006; Taylor and Mcllwain, 2010]). In this study of thumbprint em- peror in Saipan Lagoon, gonad maturation was evalu- ated macroscopically and was found to be similar to the macroscopic stages documented by Kulmiye et al. (2002). Taylor and Mcllwain (2010) compared macro- scopic and microscopic gonad evaluation, and they found that macroscopic staging was adequate 92% of the time in determining sex of thumbprint emperor and adequate 76% of the time in determining maturity status. The macroscopic approach provides approximate estimates of L50 that need to be calibrated with gonad histological examination to evaluate the accuracy of maturity status classifications, particularly for species that undergo sex change. Such studies would lend sup- port toward the goal of refining macroscopic techniques for use in circumstances in which microscopic staging is not possible, as well as toward the goal of under- standing the uncertainties associated with macroscopic staging among families of reef fishes. Otolith annuli formation in thumbprint emperor from Saipan Lagoon during June through October (Fig. 3, A and B) was similar to that observed by Ebisawa and Ozawa (2009) in thumbprint emperor from the Ryuku Islands. The similarity in estimates of L50 and the timing of otolith annuli formation between thumb- print emperor from Saipan and the Ryukyu Islands seems to indicate some level of demographic correspon- dence in the western Pacific. However, differences in Trianni: Life history characteristics and stock status of Lethrinus harak in Saipan Lagoon 421 A B 0.6 n UJ -0.1 - lU -0.2 - i -0.3 2005 2006 2007 2008 2009 2010 2011 Year oLorenzen oGislason cHoenig ©Average Figure 7 Estimated exploitation ratios for thumbprint emperor {Le- thrinus harak) in Saipan Lagoon for the period 2005-2011, derived from estimates of natural mortality (M) generated from the Lorenzen and Gislason models and from the Hoenig equation, estimates of total mortality based on results from the use of the linearized catch curve and Chapman-Robson methods, and estimates of fishing mortality {F) derived as F - Z - M. Exploitation ratios estimated by using total mortal- ity (A) with the linearized catch curve method and (B) with the Chapman-Robson method are shown. The bars depict the means of exploitation estimates, from 1000 simulations, with the use of the stated natural mortality model, and error bars indicate standard deviation. The circles are the averaged val- ues from the listed exploitation rate estimates for that year. other life history parameters (e.g., A50) among thumb- print emperor from Saipan, Guam, and the Ryukyu Islands support the need for standardized methods to study reproductive maturity, as well as age and growth for coral reef species. Demographic plasticity for age and growth and in reproductive parameters pertaining to maturity, sex change, or both (Adams et ah, 2000; Gust, 2004; Munday et ah, 2006; Mariani et ah, 2013) has been documented in both gonochoristic (Choat and Robertson, 2002; Robertson et ah, 2005) and hermaph- roditic (Gust et ah, 2002; Donovan et ah, 2012) species of coral reef fish. In addition to altering sex ratios, fishing intensity can alter life history traits, such as length and age at maturity (Heino and Dieckmann, 2008; Sharpe and Hendry, 2009), the timing of sex transition in hermaphrodites (Hamilton et ah, 2007; Gotz et ah, 2008), size and biomass spectra (Friedlander and DeMartini, 2002; Graham et ah, 2005) and, in some cases, coral reef fish community struc- ture through compensatory effects from decreased predation (Dulvy et ah, 2004; Ruttenberg et ah, 2011). The governments of Guam and Saipan have implemented marine protected areas as means to enhance fishery resources within and outside of those areas, and the CNMI has insti- tuted a ban on spearfishing with scuba gear and has implemented restrictions on the use of gill, drag, and surround nets. The NUE data for 2005-2011 show that thumbprint emperor composed about 35% of NUE landings in the CNMI (Tenorio^), but historical landings of thumbprint emperor in Saipan La- goon by net fishing were not available for com- parison. The commencement of the life history project in Saipan Lagoon shortly after implemen- tation of net-use regulations, coupled with the 22.2% decline of the Saipan population observed in the CNMI census from 2000 to 2010, suggests that a life history study of thumbprint emperor should be repeated for comparative purposes. Ad- ditionally, it would be useful to obtain life his- tory data from an unexploited or lightly exploited population of thumbprint emperor in the CNMI, although acquisition of such data would be logis- tically challenging within the populated islands of the CNMI where unfished reef flats or embay- ments are limited. Such a study would provide estimates of Z directly from natural mortality models. The estimation of k with the VBGF was im- portant in the estimation of M in one of the 3 M models used in this study and, therefore, in the subsequent estimation of E. Constraining the VBGF equation to the settlement size for this species produced a high estimate of k and, con- sequently, a lower estimate of in comparison with the estimates from the unconstrained equa- tion (Table 4). Constraining the VBGF model to an estimated settlement size resulted in larger estimates of M and in estimates of E that were subse- quently negative, indicating that the current levels of fishing mortality were not having a detrimental influ- ence on the thumbprint emperor. In anchoring the fit- ted VBGF models, in the absence of smaller specimens in the sample, a lack of variability was assumed in the length-age relationship below the smallest fish aged. As shown in this study, constriction had the effect of altering the curve from the fitted VBGF model. Obtain- ing samples from individuals after settlement, as well ^ Tenorio, M. 2012. Personal commun. Division of Fish and Wildlife, Department of Land and Natural Resources, Com- monwealth of the Northern Mariana Islands, Saipan, MP 96950. 422 Fishery Bulletin 114(4) as more samples from fish in larger size classes to en- able more accurate estimation of variability of length at age for these population segments, should remain a goal of life history research, although obtaining fish of those sizes remains a challenging task. The ratio of estimated annual landings to fishery- independent biomass estimates from 2005 through 2011 resulted in low values of F and subsequently to high estimates of SPR. Also, these estimates of F form what can be considered a “rough” baseline, and the high SPR values tend to reflect the nature of fishing for this species. The thumbprint emperor was (and is) not a primary target of the nighttime commercial free- dive spear fishery in Saipan, and this species appeared to be primarily harvested by hook and line. Data pre- sented here indicate that larger thumbprint emperors were harvested during fishing activities under net-use exemption — activities that were employed frequently in the past and that have declined since implemen- tation of net-use regulations. One of the purposes of the Magnuson-Stevens Fishery Conservation and Man- agement Act is to achieve and maintain the optimum yield from each fishery through the implementation of fishery management plans, in which the optimum yield is prescribed under certain circumstances on the basis of maximum sustainable yield (MSY). Because MSY is often difficult to estimate for data-moderate and data- poor fisheries, U.S. Fishery Management Councils have used SPR as a proxy for MSY-based precautionary con- trol rules (Restrepo et ah, 1998). The exploration of new approaches in the use of SPR for evaluating the status of data-poor fisheries has begun (Hordyk et ah, 2015). Estimates of E were variable because they were derived by sampling within the range of annually de- rived estimates of Z from 2 common estimation meth- ods (Table 7): by resampling within the M model pa- rameter uncertainty, and through the use of derived F estimates. Results indicate that thumbprint emperor from Saipan Lagoon, with estimated values of average annual E falling below 0.4, were likely not fully ex- ploited. Still, the use of the more conservative value of 0.2 for E, as suggested by Pauly (1984), for small fish at lower trophic levels with high recruitment vari- ability, would have resulted in values above 0.2 for 4 of the 6 estimates of E determined with the CR method to estimate Z. In simulations where the CR method and a variety of approaches to CC analysis were evaluated in the presence of stochastic error, Dunn et al. (2002) found that the CR estimator generally was more precise in estimating true Z values, although any advantage of the CR method over the CC method decreased with increasing Z estimates and stochastic error. The val- ues of Z generated in this study with the CR model were greater than those derived with the the CC ap- proach, and those higher values of Z resulted in larger estimates of F and subsequently higher estimates of E. The use of 3 equations for M and 2 approaches for Z resulted in a range of E values, with uncertainty, for any given year as opposed to the singular values gener- ally produced when E is estimated. All M models come with caveats (Kenchington, 2014), and modifications and improvements in M estimation continue (Gislason et al., 2010). The use of a few M models that resulted in a range of E estimates provides ample uncertainty that may be useful in ascertaining a relative under- standing of species status, given a data-poor scenario as described here. One aspect of the results of this study that should be highlighted is the use of size limits as a fish mortal- ity control for protogynous species in coral reef fisher- ies. As noted, larger thumbprint emperor are captured in nets than those captured with the fishing methods used in the ICS. If a size limit for thumbprint emperor coincident with L50 (e.g., 20 cm FL) is implemented, about 65% of the fish landed in the ICS and 17% of the fish landed as NUE would be prohibited from harvest. Conversely, more than 58% of males and more than 43% of females would be harvested, including those fe- males with the greatest fecundity and greatest impor- tance to population recruitment. Additionally, with respect to estimated ages, lengths, and weights of thumbprint emperor, size limitations would force increased fishing mortality on older and larger size classes of females that have lower esti- mated rates of M and greater reproductive potential and would eliminate fishing mortality on classes with the highest rates of M and, therefore, the best ability to absorb fishing mortality in the population. Heppell et al. (2006) concluded that catch and effort controls with spatial closures should be enacted to enhance the sustainability of hermaphroditic species, and Fenberg and Roy (2008) noted the potential negative biologi- cal effects of size-selective harvesting. The existence of a large marine protected area in Saipan Lagoon, the MMCA, and restrictions on the use of nets appear to be beneficial for thumbprint emperor and may be for other coral reef fish species in Saipan Lagoon. Standardized methods for estimating life history pa- rameters should be implemented for the macroscopic staging of gonads and be calibrated against histological staging (Brown-Peterson et al., 2011) because dervied life history parameters directly influence estimates of L50 and Lx, as well as age estimation from sectioned otoliths (Marriott et al., 2010). A standardized ap- proach for estimation of life history parameters will make jurisdictional and regional comparisons practi- cal. In the case of thumbprint emperor from Saipan Lagoon, deficiencies in sample sizes existed for indi- viduals less than 1 year old and for those fish beyond age 6, in comparison with other age classes. The recent implementation of a BSP, as well as similar programs in other U.S. jurisdictions of the central and western Pacific, provides a useful platform for representative sampling of the fisheries and for the standardization of life history parameters. The thumbprint emperor in Saipan Lagoon, unlike many other coral reef species in the CNMI, is a da- ta-poor species for which relatively considerable data Trianni: Life history characteristics and stock status of Lethrinus harak in Saipan Lagoon 423 were available or estimable from a number of sources. More species-specific or species-complex data will be required before formal stock assessments can be en- gaged for coral reef fishes in the CNMI. Acknowledgments Data collection efforts were funded by the Wildlife and Sport Fish Restoration Program of the U.S. Fish and Wildlife Service. Saipan-based Micronesian Environ- mental Services provided vendor logs for estimations of nonsampled and sampled landings. The Western Pacific Fishery Information Network of the NOAA Pa- cific Islands Fisheries Science Center (PIFSC) provided data from the DFW’s inshore creel survey program and commercial purchase database system. Various person- nel from the DFW participated in field and labora- tory worked for this project. A preliminary draft was reviewed by R. Humphreys and J. O’Malley of PIFSC. An earlier version of the manuscript benefited substan- tially from the comments of 3 reviewers. Literature cited Adams, S., B. D. Mapstone, G. R. Russ, and C. R. Davies. 2000. Geographic variation in the sex ratio, sex spe- cific size, and age structure of Plectropomus leopardus (Serranidae) between reefs open and closed to fishing on the Great Barrier Reef. Can. J. Fish. Aquat. Sci. 57:1448-1458. Alonzo, S. H., and M. Mangel. 2004. The effects of size selective fisheries on the stock dynamics of and sperm limitation in sex changing fish. Fish. Bull. 102:1-13. Alonzo, S. H., T. Ish, M. Key, A. D. MacCall, and M. Mangel. 2008. The importance of incorporating protogynous sex change into stock assessments. Bull. Mar. Sci. 83:163-179. Beddington, J. R., and G. P. Kirkwood. 2005. The estimation of potential yield and stock status using life-history parameters. Philos. Trans. R. Soc. B 360:163-170. Beverton, R. J. H., and S. J. Holt. 1957. On the dynamics of exploited fish populations. Fish- eries Investigation Series 2, vol. 19, 533 p. Ministry of Agriculture, Fisheries, and Food, London. Birkeland, C., and P. K. Dayton. 2005 The importance in fishery management of leaving the big ones. Trends Ecol. Evol. 20:356-358. Brown, I. W., and W. D. Sumpton. 1998. Age, growth and mortality of redthroat emperor Le- thrinus miniatus (Pisces: Lethrinidae) from the southern Great Barrier Reef, Queensland, Australia. Bull. Mar. Sci. 62:905-917. Brown-Peterson, N. J., D. M. Wyanski, F. Saborido-Rey, B. J. Macewicz, and S. K. Lowerre-Barbieri. 2011. A standardized terminology for describing reproduc- tive development in fishes. Mar. Coast. Fish. 3:52-70. Carpenter, K. E. 2001. Lethrinidae: Emperor (emperor snappers). In The living marine resources of the Western Central Pacific, vol. 5. Bony fishes, part 3 (Menidae to Pomacentridae) (K. E. Carpenter and V. H. Niem, eds.), p. 3004-3050. FAO, Rome. Chapman, D. G., and D. S. Robson. 1960. The analysis of a catch curve. Biometrics 16:354-368. Chen, Y., D. A. Jackson, and H. H. Harvey. 1992. A comparison of von Bertalanffy and polynomial functions in modelling fish growth data. Can. J. Fish. Aquat. Sci. 49:1228-1235. Choat, J. H., and L. M. Axe. 1996. Growth and longevity in acanthurid fishes; an analysis of otolith increments. Mar. Ecol. Prog. Ser. 134:15-26. Choat, J. H., and D. R. Robertson. 2002. Age-based studies. In Coral reef fishes: dynamics and diversity in a complex ecosystem (P. F. Sale, ed.), p 57-80. Academic Press, New York. Choat, J. H., L. M. Axe, and D. C. Lou. 1996. Growth and longevity in fishes of the family Scari- dae. Mar. Ecol. Prog. Ser. 145:33—41. Cochran, W. G. 1977. Sampling techniques, 3rd ed., 428 p. John Wiley & Sons Inc., New York. Currey, L. M., A. J. Williams, B. D. Mapstone, C. R. Davies, G. Carlos, D. J. Welch, C. A. Simpfendorfer, A. C. Ballagh, A. L. Penny, E. M Grandcourt, et al. 2013. Comparative biology of tropical Lethrinus species (Lethrinidae): challenges for multi-species manage- ment. J. Fish Biol. 82:764-788. Dalzell, P., S. Sharma, and G. Nath. 1992. Estimation of exploitation rates in a multispecies emperor (Pisces: Lethrinidae) fishery in Fiji, based on length frequency data. In Papers on fisheries science from the Pacific Islands, vol. 1 (P. Dalzell, ed.), p. 43- 50. South Pacific Commission, Noumea, New Caledonia. Dalzell, P., T. J. H. Adams, and N. V. C. Polunin. 1996. Coastal fisheries in the Pacific Islands. Oceanogr. Mar. Biol., Annu. Rev. 34:395-531. Dick, E. J., and A. D. MacCall. 2011. Depletion-based stock reduction analysis: a catch- based method for determining sustainable yields for data-poor fish stocks. Fish. Res. 110:331-341. Donovan, M. K., A. M. Friedlander, E. E. DeMartini, M. J. Do- nahue, and 1. D. Williams. 2012. Demographic patterns in the peacock grouper (Cephalopholis argus), an introduced Hawaiian reef fish. Environ. Biol. Fish. 96:981-994. Dulvy, N. K., N. V. Polunin, A.C. Mill, and N. A. Graham. 2004. Size structural change in lightly exploited coral reef fish communities: evidence for weak indirect ef- fects. Can. J. Fish. Aquat. Sci. 61:466-475. Dunn, A R. 1. C. C. Francis, and 1. J. Doonan. 2002. Comparison of the Chapman-Robson and regression estimators of Z from catch-curve data when non-sam- pling stochastic error is present. Fish. Res. 59:149-159. Ebisawa, A. 1990. Reproductive biology of Lethrinus nebulosus (Pisces: Lethrinidae) around the Okinawan waters. Nippon Su- isan Gakkaishi 56:1941-1954. 2006. Reproductive and sexual characteristics in five Le- thrinus species in waters off the Ryukyu Islands. Ich- thyol. Res. 53:269-280. 424 Fishery Bulletin 1 14(4) Ebisawa, A, and T. Ozawa. 2009. Life-history traits of eight Lethrinus species from two local populations in waters off the Ryukyu Is- lands. Fish. Sci. 75:553-566. Fenberg, P. B., and K. Roy. 2008. Ecological and evolutionary consequences of size- selective harvesting: how much do we know? Mol. Ecol. 17:209-220. Friedlander, A. M., and E. E. DeMartini. 2002. Contrasts in density, size, and biomass of reef fishes between the northwestern and the main Hawaiian Is- lands: the effects of fishing down apex predators. Mar. Ecol. Prog. Sen 230: 253-264. Garcia, S. M., J. Holding, J. Rice, M. J. Rochet, S. Zhou, T. Arimoto, J. E. Beyer, L. Borges, A. Bundy, D. Dunn, et al. 2012. Reconsidering the consequences of selective fisher- ies. Science 335:1045-1047. Gislason, H., N. Daan, J. C. Rice, and J. G. Pope. 2010. Size, growth, temperature and the natural mortal- ity of marine fish. Fish Fish. 11:149-158. Goodyear, C. P. 1993. Spawning stock biomass per recruit in fisheries management: foundation and current use. In Risk eval- uation and biological reference points for fisheries man- agement (S. J. Smith, J. J. Hunt, and D. Rivard, eds.), p. 67-81. Can. Spec. Pub. Fish. Aquat. Sci. 120. Gotz, A., S. E. Kerwath, C. G. Attwood, and W. H. H. Sauer. 2008. Effects of fishing on papulation structure and life history of roman Chrysoblephus laticeps (Spari- dae). Mar. Ecol. Prog. Ser. 362:245-259. Graham, N. A. J., N. K. Dulvy, S. Jennings, and N. V. C. Polunin. 2005. Size-spectra as indicators of the effects of fishing on coral reef fish assemblages. Coral Reefs 24:118-124. Grandcourt, E. M. 2002. Demographic characteristics of a selection of exploit- ed reef fish from the Seychelles: preliminary study. Mar. Freshw. Res. 53:123-130. Grandcourt, E. M., T. Z. Al Abdessalaam, F. Francis, and A. T. AI Shamsi. 2010. Reproductive biology and implications for manage- ment of the spangled emperor Lethrinus nebulosus in the southern Arabian Gulf. J. Fish Biol. 77:2229-2247. Gulland, J. A. 1970. The fish resources of the ocean. FAO Fish. Tech. Pap. 97, 425 p. Gust, N. 2004. Variation in the population biology of protogynous coral reef fishes over tens of kilometres. Can. J. Fish. Aquat. Sci. 61:205-218. Gust, N., J. Choat, and J. Ackerman. 2002. Demographic plasticity in tropical reef fishes. Mar. Biol. 140:1039-1051. Hamilton, S. L., J. E. Caselle, J. D. Standish, D. M. Schroeder, M. S. Love, J. A. Rosales-Casian, and O. Sosa-Nishizaki. 2007. Size-selective harvesting alters life histories of a temperate sex-changing fish. Ecol. Appl. 17:2268-2280. Heino, M, and U. Dieckmann. 2008. Detecting fisheries-induced life-history evolution: an overview of the reaction-norm approach. Bull. Mar. Sci. 83:69-93. Heppell, S. S., S. A. Heppell, F. C. Coleman, and C. C. Koenig. 2006. Models to compare management options for a pro- togynous fish. Ecol. Appl. 16:238-249. Hoenig, J. M. 1983. Empirical use of longevity data to estimate mortal- ity rates. Fish. Bull. 81:898-903. Hoenig, J. M., and D. M. Heisey. 1987. Use of a log-linear model with the EM algorithm to correct estimates of stock composition and to convert length to age. Trans. Am. Fish. Soc. 116:232-243. Hordyk, A., K. Ono, K. Sainsbury, N. Loneragan, and J. Prince. 2015. Some explorations of the life history ratios to de- scribe length composition, spawning-per-recruit, and the spawning potential ratio. ICES J. Mar. Sci. 72:204-216. Jiao, Y, E. Cortes, K. Andrews, and F. Guo. 2011. Poor-data and data-poor species stock assessment using a Bayesian hierarchical approach. Ecol. Appl. 21:2691-2708. Kenchington, T. J. 2014. Natural mortality estimators for information lim- ited fisheries. Fish Fish. 15:533-562. Kulmiye, A. J., M. J. Ntiba, and S. M. Kisia. 2002. Some aspects of the reproductive biology of the thumbprint emperor, Lethrinus harak (Forsskal, 1775), in Kenyan coastal waters. West. Indian Ocean J. Mar. Sci. 1:135-144. Laursen, T, G. R. Russ, S. J. Newman, and J. B. Higgs. 1999. Age, growth and mortality of Gymnocranius audleyi (Pisces: Lethrinidae). Asian Fish. Sci. 12:187-200. Lewin, W. C., R. Arlinghaus, and T. Mehner. 2006. Documented and potential biological impacts of rec- reational fishing: insights for management and conserva- tion. Rev. Fish. Sci. 14:305-367. Lo Galbo, A. M., K. E. Carpenter, and D. L. Reed. 2002. Evolution of trophic types in emperor fishes {Lethri- nus, Lethrinidae, Percoidei) based on cytochrome b gene sequence variation. J. Mol. Evol. 54:754-762. Lorenzen, K. 1996. The relationship between body weight and natu- ral mortality in juvenile and adult fish: a comparison of natural ecosystems and aquaculture. J. Fish Biol. 49:627-647. MacCall, A. D. 2009. Depletion-corrected average catch: a simple formu- la for estimating sustainable yields in data-poor situa- tions. ICES J. Mar. Sci. 66:2267—2271. Mariani, S., M. Sala-Bozano, J. Chopelet, and C. Benvenuto. 2013. Spatial and temporal patterns of size-at-sex-change in two exploited coastal fish. Environ. Biol. Fish. 96:535-541. Marriott, R. J., B. D. Mapstone, A. C. Ballagh, L. M. Currey, A. Penny, A. J. Williams, G. Jackson, D. Lou, A. J. Mapleston, N. D. C. Jarvis, et al. 2010. Accepting final counts from repeat readings of oto- liths: should a common criterion apply to the age estima- tion of fish? Mar. Freshw. Res. 61:1171-1184. McGilliard, C. R., R. Hilborn, A. MacCall, A. E. Punt, and J. C. Field. 2011. Can information from marine protected areas be used to inform control-rule-based management of small- scale, data-poor stocks? ICES J. Mar. Sci, 68:201-211. Munday, P. L., P. M. Buston, and R. R. Warner. 2006. Diversity and flexibility of sex-change strategies in animals. Trends Ecol. Evol. 21:89-95. Myers, R. F. 1999. Micronesian reef fishes: a comprehensive guide to the coral reef fishes of Micronesia, S*''* ed., 330 p. Coral Graphics, Barrigada, Guam. Trianni: Life history characteristics and stock status of Lethrinus harak in Saipan Lagoon 425 Nakamura, Y., M. Horinouchi, M. Sano, and T. Shibuno. 2009. The effects of distance from coral reefs on seagrass nursery use by 5 emperor fishes at the southern Ryukyu Islands, Japan. Fish. Sci. 75:1401-1408. Nakamura, Y., T. Shibuno, N. Suzuki, J. Nakamori, K. Kanashi- ro, and Y. Watanabe. 2010. Interspecific variations in age and size at settlement of 8 emperor fishes (Lethrinidae) at the southern Ryukyu Islands, Japan. Fish. Sci. 76:503-510. Nanami, A., and H. Yamada. 2009. Site fidelity, size, and spatial arrangement of day- time home range of thumbprint emperor Lethrinus harak (Lethrinidae). Fish. Sci. 75:1109-1116. Pauly, D. 1984. Fish population dynamics in tropical waters: a man- ual for use with programmable calculators. ICLARM Stud. Rev. 8, 325 p. International Center for Living Aquatic Resources Management, Manila, Philippines. Punt, A. E., D. C. Smith, and A. D. M. Smith. 2011. Among-stock comparisons for improving stock as- sessments of data-poor stocks: the “Robin Hood” ap- proach. ICES J. Mar. Sci: 68:972-981. R Core Team. 2015. R: a language and environment for statistical com- puting. R Foundation for Statistical Computing, Vienna, Austria. [Available at website, accessed May 2015.] 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, et al. 1998. Technical guidance on the use of precautionary ap- proaches to implementing National Standard 1 of the Magnuson-Stevens Fishery Conservation and Manage- ment Act. NOAA Tech. Memo, NMFS-F/SPO-31, 48 p. Robertson, D. R., J. L. Ackerman, J. H. Choat, J. M. Posada, and J. Pitt. 2005. Ocean surgeonfish Acathurus bahianus. I. The ge- ography of demography. Mar. Ecol. Prog. Sen 295:229- 244. Robson, D. S., and D. G. Chapman. 1961. Catch curves and mortality rates. Trans. Am. Fish. Soc. 90:181-189. Ruttenberg, B. I., S. L. Hamilton, S. M. Walsh, M. K. Donovan, A. Friedlander, E. E. DeMartini, E. Sala, and S. A. Sandin. 2011. Predator-induced demographic shifts in coral reef fish assemblages. PLoS ONE 6(6):e21062. Sadovy de Mitcheson, Y, and M. Liu. 2008. Functional hermaphroditism in teleosts. Fish Fish. 9:1-43. Sharpe, D. M. T, and A. P. Hendry. 2009. Life history change in commercially exploited fish stocks: an analysis of trends across studies. Evol. Appl. 2:260-275. Shepherd, S. A., J. B. Brook, and Y. Xiao. 2010. Environmental and fishing effects on the abundance, size and sex ratio of the blue throated wrasse, Notola- brus tetricus, on South Australian coastal reefs. Fish. Manage. Ecol. 17:209-220. Shibuno, T., Y. Nakamura, M. Horinouchi, and M. Sano. 2008. Habitat use patterns of fishes across the mangrove- seagrass-coral reef seascape at Ishigaki Island, southern Japan. Ichthyol. Res. 55:218-237. Sumpton, W., and I. Brown. 2004. Reproductive biology of the red throat emperor Le- thrinus miniatus (Pisces: Lethrinidae) from the south- ern Great Barrier Reef, Australia. Bull. Mar. Sci. 74:423-432. Taylor, B. M., and J. L. Mcllwain. 2010. Beyond abundance and biomass: effects of marine protected areas on the demography of a highly exploited reef fish. Mar. Ecol. Prog. Sen 411: 243-258. Trianni, M. S. 2011. Biological characteristics of the spotcheek emperor, Lethrinus rubrioperculatus, in the Northern Mariana Is- lands. Pac. Sci. 65:345-363. von Bertalanffy, L. 1938. A quantitative theory of organic growth (inquiries on growth laws. II). Hum. Biol. 10:181-213. Williams, A. J., C. R. Davies, B. D. Mapstone, and G. R. Russ. 2003. Scales of spatial variation in demography of a large coral reef fish — an exception to the typical model? Fish. Bull. 101:673-683. Williams, A. J., B. D. Mapstone, and C. R. Davies. 2007. Spatial and interannual patterns in growth of an exploited coral-reef fish. J. Fish Biol. 71:970-992. Zhou, S., A. D. M Smith, A. E. Punt, A. J. Richardson, M. Gibbs, E. A. Fulton, S. Pascoe, C. Bulman, P. Bayliss, and K. Sainsbury. 2010. Ecosystem-based fisheries management requires a change to the selective fishing philosophy. Proc. Natl. Acad. Sci. USA 107: 9485-9489. 426 NOAA National Marine Fisheries Service Fishery Bulletin ^ established 1881 Spencer F. Baird First U.S. Commissioner of Fisheries and founder of Fishery Bulletin SiMimming depth of dolphinfish iCoryphaena hippurm} associated and unassociated mith fish aggregating dewices Abstract— Dolphinfish (Coryphaena hippurus), large pelagic predators and important fishery targets, fre- quently associate with floating debris or manmade fish aggregating devices (FADs). We tagged 8 dolphinfish with pressure-sensitive ultrasonic trans- mitters and actively tracked indi- viduals continuously for up to 40 h to elucidate the vertical movement pat- terns and differences between FAD- associated (FAD-A) and FAD-unasso- ciated (FAD-U) fish. Four additional fish were equipped with acoustic transmitters and passively monitored for several days with receivers at- tached to FADs. When not associated with FADs, dolphinfish used the up- per 75-100 m of the water column during the day and made descents up to 160 m during the night. In con- trast, FAD-A fish generally stayed within the upper 10 m of the water column and tended to make deeper excursions during the day rather than at night. Water temperature data from expendable bathythermo- graphs deployed during active track- ing showed that fish only descended to depths where temperatures were <3°C cooler than the uniform-temper- ature surface layer. The use of verti- cal behavior to determine whether a dolphinfish is associated or not with a floating object opens the possibility for new, large-scale research aimed at investigating the role of floating objects in the ecosystem inhabited by this species and at assessing the im- pacts of FADs on its ecology. Manuscript submitted 2 June 2015. Manuscript accepted: 14 July 2016. Fish. Bull: 426-434 (2016). Online publication date: 8 August 2016. doi: 10.7755/FB.114.4.5 The views and opinions expressed or implied in this article are those of the author (or authors) and do not necessarily reflect the position of the National Marine Fisheries Service, NOAA. Nicholas M. Whitney (contact author)’ Marc Taqiiet^ Richard W. Brill^ Charlotte Girard^ ' Behavioral Ecology and Physiology Program Mote Marine Laboratory 1600 Ken Thompson Parkway Sarasota, Florida 34236 2 Unites Mixtes de Recherche (UMR) Ecosystemes Insulaires Oceaniens (ElO) Institut de Recherche pour le Developpement BP 529 98713 Papeete, Tahiti, Polynesie fran^iaise 2 Pacific Islands Fisheries Science Center National Marine Fisheries Service, NOAA 2570 Dole Street Honolulu, Hawaii 96822-2396 Dolphinfish {Coryphaena hippurus) are large pelagic fish that are com- mon globally in tropical and warm temperate seas (Palko et ah, 1982). They have diverse diets composed of floating debris-associated organisms, such as portunids (crabs) and epipe- iagic cephalopods (Olson and Galvan- Magana, 2002; Rudershausen et al., 2010), as well as neritic and demer- sal fish (Tripp-Valdez et al., 2010). The fast growth rate (Schwenke and Buckle, 2008; Furukawa et ah, 2012), early sexual maturity (less than one year; Trippel, 1995; Furukawa et al., 2012), and high food value (Beardsley, 1967; Rodriguez-Ferrer et al., 2004) of this species makes it an increasingly important target of commercial, sport, and artisanal fisheries (Rodriguez- Ferrer et al., 2004). The tendency of dolphinfish to aggregate around float- ing objects and flora, such as sargas- sum (Oxenford, 1999; Rooker et al., Gail D. Schwieterman’ Laurent Dagorn^ Kim N. Holland® Unites Mixtes de Recherche (UMR) Marine Biodiversity, Exploitation, and Conservation (MARBEC) Institut de Recherche pour le Developpement Avenue Jean Monnet CS 30171 34203 Sete Cedex, France ^ Hawaii Institute of Marine Biology University of Hawaii at Manoa P.O. Box 1346 Kaneohe, Hawaii 96744 2006; Casazza and Ross, 2008; Farrell et ah, 2014; Merten et ah, 2014a), logs (Baughman, 1941), and manmade fi.sh aggregating devices (FADs; Rose and Hassler, 1974; Massutf et al., 1998, Oxenford and Hunte, 1999; Dempster, 2004; Dagorn et al., 2007; Taquet et al., 2007), facilitates their exploitation by both commercial and recreational fisheries throughout the world. Recent estimates put annual catch of dolphinfish at more than 102,986 metric tons (t) (FAO, 2013), which excludes unreported bycatch (com- mon in longline tuna fisheries) and underreported artisanal landings. A recent study estimated that 5382 t of illegal or unreported dolphinfish were imported from Ecuador to the United States in 2011 alone (Pramod et al., 2014). The high exploitation rate and commercial importance of this species justify a better foundational knowl- edge of the life history and behavioral Email address for contact author: nwhitney@mote.org Whitney et al.: Swimming depth of Coryphaena hippurus 427 Table 1 Characteristics of the tracking or monitoring of dolphinfish, including which method was used: active (boat-based) or passive (receiver attached to a fish aggregating device [FAD]). Behavior categories are day (D) or night (N) and either FAD-unassociated (U) or FAD-associated (A). FAD types (drifting or anchored) are listed only for animals that were actually associated with a FAD for a part of the time they were tracked. Method Fish Fork length (cm) Tracking start date Duration of tracking (h) Behavior category FAD type Location Active 1 125 4-Mar-86 40.1 DU/NU Hawaii 2 108 3-Feb-87 6.0 DU Hawaii 3 103 7-Mar-87 22.5 DU/NU Hawaii 4 104 24-Mar-87 27.2 DU/NA drifting Hawaii 5 116 21-Jul-03 8.0 DA anchored Reunion Is. 6 124 6-Aug-03 8.0 DU Reunion Is. 7 58 12-Apr-05 4.3 DU^ Seychelles 8 69 21-Apr-05 7.5 DU^ Seychelles Passive 9 110 13-Dec-Ol 38.4 DA/NA drifting S. Indian 10 77 12-Oct-03 81.6 DA/NA drifting S. Indian 11 86 8-Feb-04 87.8 DA/NA drifting Seychelles 12 102 16-Oct-04 84.5 DA/NA drifting Seychelles ^Fish tracked in shallow areas (depths of 50—60 m) were not included in statistical analyses. patterns of dolphinfish in order to ensure development of appropriate measures for fishery management and resource conservation. Previous studies of the movements of dolphinfish have focused either on broadscale migrations or on movements in highly specific contexts. For example, Merten et al. (2014a) used mark-recapture data to describe broadscale migrations along the East Coast of the United States, from the Florida Keys to Long Island, New York. Merten et al. (2014b) used plastic dart tags and pop-off satellite tags to determine regional patterns of horizontal move- ment, and Norton (1999) described a global poleward shift of dolphinfish that corresponded with an increase in ocean temperatures. Individual dolphinfish have been shown to home to a specific moored FAD (Girard et ah, 2007) and to re- main associated with an anchored FAD for a mean du- ration of 3.98 days (Dagorn et ah, 2007) and with drift- ing FADs for a mean duration of 6.25 days (Taquet et al., 2007). Dolphinfish in the western Atlantic generally maintain swimming depths in the uniform-temperature surface layer and descend beyond the surface layer (0- 10 m) longer at night than during the day, possibly to forage (Merten et al., 2014c). FAD-U fish in the north- ern East China Sea often remain above the thermocline (Furukawa et al., 2011), although Merten et al. (2014c) found that the thermocline was not a barrier to vertical movements for dolphinfish in the Atlantic. In general, dolphinfish have the shallowest vertical distribution of other sympatric, mid-trophic level predators, such as the striped marlin (Kajikia audax; Brill et al., 1993) and sailfish (Istiophorus platypteru; Hoolihan, 2005). To date, the movement patterns between FAD-A and FAD-U dolphinfish have not been compared. We ana- lyzed the vertical swimming behavior of both FAD-A and FAD-U dolphinfish, tracked by using active and passive telemetry, with particular regard to diurnal patterns. Understanding the vertical movement patterns of this commercially important species is increasingly essential for sustainable fisheries management, especially in the face of climate change. The use of a comprehensive ethol- ogy will enable managers to predict changes in distribu- tion of dolphinfish as seawaters warm and to implement bycatch reduction regulations, such as a mandate that would limit the depths at which gear can be set to depths outside the range preferred by dolphinfish. Materials and methods Study sites A total of 12 dolphinfish were tagged and released at lo- cations 1) in the central Pacific (near the main Hawaiian Islands, n=A) in 1986 and 1987, 2) in the southwestern Indian Ocean (near Reunion Island; n=4) in 2005, and 3) in the Western Indian Ocean (near Seychelles, n=4) in 2003 (Table 1). In the central Pacific, active tracking of 4 fish was conducted off the northeast (windward) coast of Oahu near an array of FADs that consisted of sur- face buoys anchored at depths from 670 to 3660 m. The tracks of 4 more fish were recorded with active telemetry around FADs made of about 10 surface buoys, a metal cable, and strap bands located on the cable at a depth of ~20 m, anchored with concrete blocks on the Seychelles Plateau in relatively shallow (50-60 m) water or around drifting FADs of similar construction. In the open sea off Reunion Island, 2 fish were passively monitored in 428 Fishery Bulletin 1 14(4) depths of 2000-5000 m, one fish in 2001 and another in 2003, by using drifting FADs, made of bamboo rafts with trailing netting panels, that were deployed specifically to study pelagic fish aggregations (Taquet, 2004). Off the Seychelles Plateau in 2004, 2 more fish were passively monitored near FADs made of surface buoys anchored at depths of 500-1500 m (Girard et al., 2007). Fish from all locations were caught by trolling and pole techniques and were brought aboard the vessel for tagging. Active telemetry A detailed account of the tracking methods employed in Hawaii are described in Holland et al. (1985), and those used off Reunion Island and in the Seychelles are de- tailed in Girard et al. (2007). In Hawaii, pressure-sensi- tive V16 ultrasonic transmitters (16.0 mm in diameter, 27.7 g in air and 11.7 g in seawater; Vemco Inc.^, Bed- ford, Nova Scotia, Canada) were attached externally to a fish by passing 2 nylon cable ties through the dorsal or ventral pterygiophores and trunk musculature adjacent to the second dorsal fin (we switched to the latter proce- dure after one fish was observed swimming at the sur- face with its dorsal tag carried above the surface of the water). The transmitters were equipped with pressure sensors that modulated the rate of pulses transmitted in response to changes in water pressure (depth). There- fore, vertical movements of the fish were determined by measuring the time between signal pulses (e.g., Holland et al., 1990b). Individuals were held out of the water for approximately 1 min during the tagging procedure. For tracking, we used a 12V DC amplifier and receivers (CR- 40, Vemco Inc.) and a directional hydrophone (Vemco Inc.), which was mounted to a pole extending about 1.5 m below the surface. Attempts were made to keep the tracking vessel 200 m away from the tagged individual to maintain maximum signal amplitude and not disturb natural behavior. The fish actively tracked off Reunion Island were caught at FADs and held in a container of flowing sea- water, and the fish caught in the Seychelles were placed in a padded cradle for a few minutes and provided with flow of water to oxygenate their gills. Fish tracked in the Seychelles and off Reunion Island were moved dis- tances of 70-1720 m from the FAD during tagging, as part of a study on dolphinfish homing abilities around FADs (Girard et al., 2007). For fish tracked in these 2 areas, V16P-4H transmitters (16.0 mm in diameter, 25 g in air and 11 g in seawater, with a random delay be- tween 40 and 120 s, depth resolutions from 0.3 to 1.5 m; Vemco Inc.) were attached by means of a hook that passed through the pterygiophores of the anal fin. Four hydrophones were towed below the tracking vessel on a V-fin depressor, and the signal emitted by a tag was processed with a VR28 receiver system (Vemco Inc.) con- nected to a laptop computer to record the data that were 1 Mention of trade names or commercial companies is for iden- tification purposes only and does not imply endorsement by the National Marine Fisheries Service, NOAA. automatically stored in the receiver. Whenever possible, a distance of at least 100 m was maintained between the tagged fish and the boat to reduce the effect of the tracking vessel on fish behavior. During active tracking, data for seawater temperature were acquired through periodic deployments of an ex- pendable bathythermograph system (Lockheed Martin Sippican, Marion, MA.). We defined association with a FAD as a fish remaining within 360-655 m of a FAD or floating debris for at least 30 min, as determined bf direct observation. Passive telemetry Fish were passively monitored with V13P-1H acoustic transmitters that would transmit on a random delay of 40-120 s (fish 9, 11, and 12; 13 mm in diameter, 11 g in air and 6 g in seawater; Vemco Inc.). Tags were inserted into the peritoneal cavity through an incision of 1-2 cm made to the side of the ventral centerline of the fish and 2-3 cm anterior to the cloaca. The wound was closed by using sterile needles and suture material; the entire pro- cedure was completed in less than 3 min. Fish 10 was equipped with a V16P-4H acoustic transmitter on a de- lay of 10-30 s (18 mm in diameter, 36 g in air and 16 g in seawater; Vemco Inc.) attached above their anal fin with a hook. To monitor tagged fish around FADs, VR2 receivers (Vemco Inc.) were suspended 5 m below a buoy that was tied to the FAD. Following Girard et al. (2007), we defined association of passively monitored fish as fish remaining within the detection range of the transmitters on the FAD (distances of 360 and 655m for V13P-4H and V16P-4H receivers, respectively). After a few days of observation, VR2 receivers were removed from the FADs so that data could be downloaded. Analyses Pooled time-at-depth histograms were constructed with 5- or 10-m bins as described by Holland et al. (1990b). These data subsequently were expressed as a fraction of the total time each fish was followed, and the frac- tional data bins were averaged across all fish. Fish 7 and 8 were excluded from statistical analyses because their vertical movements were limited by a bottom depth of 50-60 m. A linear mixed-effects model was used to examine the effects of time of day and FAD association on swimming depth. Our model included fixed effects for tagging method, FAD association (associated or unassoci- ated), and time of day (day or night). Additionally, indi- vidual fish and location of study site were random effects because it was not our aim to study intraspecific vari- ability or differences that resulted from study site. Our model did not include interaction terms. Models were ran with the Imer function in the lme4 package (Bates et al., 2014) in R, vers. 2.15.3 (R Core Team, 2013). The candidate model was selected by using the Akaike infor- mation criterion, with factor significance determined by pairwise analysis of variance (ANOVA) tests (a=0.05). P- values for pairwise comparisons were computed by using Whitney et al.: Swimming depth of Coryphaena hippurus 429 20 H IliiJ 40 -jH 60 -H 80 -L a 1 CD Q 100- 1 120-1 140 - 160 - B D. (D Q 100 120 ^ Markov chain Monte Carlo methods from the LMERConvenienceFunction package (Tremblay and Ransijn, 2013) in R. Results A total of 8 dolphinfish (mean fork length:; 100.9 cm [standard deviation (SD) 24.7]) were actively tracked, and 4 fish (mean fork length: 93.8 cm [SD 15.0]) were monitored with passive acoustic receivers (Table 1). The duration of active tracking of individual fish in=8) ranged from 4.3 to 40.1 h (mean duration of tracking: 15.5 h). Actively tracked fish were unassociat- ed with FADs for the majority of the time that they were tracked, with the exception of fish 4 and 5 (Table 1). Passively moni- tored fish {n=4), and therefore FAD-A fish, were monitored for periods of 38.4-87.8 h (mean duration of tracking: 73.1 h). Four dolphinfish (fish 5, 6, 9, and 11) immediately after release made quick de- scents that lasted less than 1 h before they resumed shallower swimming behavior for the remainder of the time that they were tracked. No other response to tagging was observed. Our analysis revealed that study loca- tion and monitoring method did not have a significant effect on swimming depth (variance approached 0 for both effects (P>0.05), and these effects were removed from the model. We did not find significant interaction between time of day (day or night) and FAD association (associated or unassociated) (ANOVA: F>0.Q5); therefore, we present values from a reduced model in which depth was dependent on the fixed effects of time of day and FAD association, with individual (fish) as a random effect. We found mean daytime depth of FAD-A fish to be 1.8 m (SD 8.1), mean nighttime depth of FAD-A fish to be 0.8 m (SD 1.7), mean daytime depth of FAD-U fish to be 49.5 m (SD 0.3), and mean nighttime depth of FAD-U fish to be 28.3 m (SD 2.9). During the day when not associated with FADs or floating debris, dolphinfish remained within the uniform-temperature surface layer (above the thermocline) and made only limited vertical excursions to depths of 75-100 m. At night, FAD-U fish swam at depths between 30 and 160 m and ventured into cooler water. However, fish reached depths that were no more than 3°C cooler than the uniform-temper- ature surface layer (Fig. 1). Although FAD-U fish spent an average of 29.4% of their time in the upper 10 m of the water column during both day and night, there was Q. (U Q 100 - 120 - 140 - 160 00:00 12:00 00:00 06:00 Time Figure 1 Swimming depth of dolphinfish {Coryphaena hippurus) actively tracked off Oahu, Hawaii, in March 1986 and March 1997: (A) fish 1 during track- ing for a period of 40.1 h, (B) fish 3 during tracking for a 22.5-h period, and (C) fish 4 during tracking for a period of 27.2 h. Shades of gray distinguish temperature isotherms. Bold black lines above the A:-axes in- dicate nighttime. The arrow in graph C indicates the time at which fish 4 became associated with drifting debris and numerous other (untagged) dolphinfish. The isotherms are based on temperature recordings from the active acoustic tracking and extend to the last recorded temperature reading. another peak of time-at-depth at depths between 30 and 40 m, their overall depth distribution during nighttime was broader and deeper than their daytime distribution (P<0.001; Fig. 2). When associated with FADs or floating debris, dol- phinfish spent an average of 94.8% of their time in the 430 Fishery Bulletin 114(4) Percent time Figure 2 Percent time at depth (A) for all dolphinfish (Coryphaena hip- purus) unassociated with a fish aggregating device (FAD) during day (white bars: n=7) and night (gray bars: n=2), (B) for FAD- associated fish during day (white bars: n=5) and night (gray bars: n=5), and (C) for fish unassociated with FADs (white bars: n=7) and fish associated with FADs (gray bars: n=6). Sample sizes dif- fer because not all individuals were actively tracked or monitored during full 24-h cycles (see Table 1.). Dolphinfish were actively tracked in Hawaii, Reunion Island, and the Seychelles Plateau between 1986 and 2005 and were passively monitored in the Seychelles Plateau and the Indian Ocean between 2001 and 2004. upper 10 m of the water column. During the night, fish typically remained within 5 m of the surface, venturing into depths that were only slightly deeper during the day (Fig 3); this difference was not significant. Only 1 fish (fish 11) exceeded a depth of 30 m while associated with a FAD (Fig 3). Fish 4 became associated with an abandoned net and other drifting debris while being tracked (Fig. 1). It swam at depths between 25 and 75 m for the first 4 h after it was tagged and while it was unassociated with a FAD or debris. After be- coming associated with floating debris, this fish remained within the upper 20 m of the water col- umn for the final 17 h. that it was tracked, and it did not have the deeper nighttime swimming behavior observed in FAD-U fish. These results from active telemetry were confirmed by direct ob- servations of the tagged fish near the surface on several occasions. By the end of the active tracking of fish 4, ap- proximately 300 dolphinfish, a large school of small yellowfin tuna (Thunnus albacares), and at least 1 marlin ilstiophoridae) were observed to also be associated with debris. No fish were ob- served to leave a FAD once it became associated with one, and passive monitoring ended when tracked fish were recaptured or the receivers were retrieved. At the end of its passive monitoring, fish 11 showed a rapid descent to more than 200 m and then exceeded the range of the receiver (not shown in Fig. 3), likely as a result of shedding its tag or possibly mortality. Actively tracked fish 1 and 3, which were un- associated with floating objects during crepuscu- lar periods, interrupted their vertical movement patterns and remained within the upper 10 m of the water column for several minutes during the dawn or dusk periods (fi=4). These shallow periods lasted from 39 to 169 min. Discussion We present a first examination of the effect of FAD association on diel movement patterns of dolphinfish. Overall, individuals remained in the uniform-temperature surface layer in congruence with results of other studies on tracking dolphin- fish (Furukawa et aL, 2011, 2014; Merten et al., 2014c), as well as with results from studies on tracking wahoo (Acanthocybium solandri; Sepulve- da et al., 2011) and sailfish (Chiang et al., 2011). However, we found differences in the vertical movement patterns between FAJ)-A fish and FAD- U fish. The distributions of the former were ex- tremely shallow, with deeper excursions observed during the day. In contrast, the latter ranged throughout the upper 100 m of the water column and made deeper descents at night. The differ- ences in behavior of FAD-A and FAD-U fish can Whitney et al.: Swimming depth of Coryphaena hippurus 431 be noted particularly in the behavior of fish 4, which became associated partway through the period in which it was actively tracked (Fig. 1). These results are similar to those from studies of bigeye tuna {Thunnus obesus): deeper diving occurred when fish were not associated with surface objects and shallow distribution was generally uniform during day and night when fish were associated with surface objects (Hol- land et al., 1990b; Musyl et al., 2003). The difference in depth distribution between FAD-A and FAD-U dolphinfish may be due to differences in feeding strategy. Although Ta- quet (2004) found that only 27% of the diet of FAD-A dolphinfish comes from FAD-associated organisms, dolphinfish are primarily visual predators (Massuti et al., 1998) and would be expected to forage primarily during the day- time. For FAD-A fish, deeper daytime dives may represent attempts to forage for prey spe- cies distributed throughout the uniform-tem- perature surface layer. If this is the case, the motivation to make deeper descents would be lessened at night because reduced light levels make foraging difficult. The broad patterns of vertical movement that we observed in FAD-U dolphinfish are similar to the behaviors observed in other large pelagic fishes, such as tropical tunas and marlin, unassociated with a FAD or debris (e.g., Holland et al., 1990a, 1990b; Brill et al., 1993, Musyl et al., 2003). Such movement patterns may allow fish to explore the water column for prey. During the day, FAD-U dolphinfish may target prey that occupy the uniform-tempera- ture surface layer — prey species that likely are not part of the deep scattering layer. However, at night, dolphinfish may forage on prey of the deep scattering layer that rise to occupy the uniform-temperature surface layer. Although we did not observe the dissociation of any fish from a FAD, Taquet et al. (2007) suggested the need to forage may prompt dolphinfish to leave FADs or other floating objects. The rapid, deep descents of 4 fish immedi- ately after release were likely a response to tagging because these fish quickly resumed vertical movement patterns that were main- tained for extended periods and, therefore, pre- sumably represent natural behavior (Girard et al., 2007). The remaining 8 individuals showed no such initial response. Hoolihan et al. (2011) suggested that stress from capture and han- dling affects the behavior of large pelagic fishes for periods that span from days to weeks after release. However, a comparison of vertical movement behaviors from our study with those from other acoustic telemetry studies (with tracking conducted for -24-48 h after release) and with those from studies that employ im- C 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 Figure 3 Swimming depths of 4 dolphinfish (Coryphaena hippurus) associated with a fish aggregating device and monitored passively with V13P- IH or V16-4H transmitters over periods of several days in the open sea in the southwestern Indian Ocean or near the Seychelles Pla- teau. (A-C) Fish 9, 11, and 12 were tracked in 2001 and 2004. (D) Fish 10 was tracked in 2003. Individual dots represent detections by the acoustic receiver. Bold gray lines above the x-axes indicate nighttime. 432 Fishery Bulletin 114(4) planted archival tags and pop-up satellite archival tags (with tags often containing months long data records) showed remarkable similarity and indicates that 1) re- covery after release for pelagic teleosts, such as tunas and billfishes, requires only 2-6 h (Holland et ah, 1990a, 1990b; Brill et al., 1993) and 2) differences in vertical movement patterns can be explained largely by differ- ences in oceanographic conditions and prey distributions (e.g., Dewar et al., 2011; Schaefer et al., 2007, 2011). We observed 2 fish that interrupted their regular pat- terns of vertical movement during crepuscular periods to remain instead in the upper 10 m of the water col- umn. These shallow periods may enhance the ability of fish to use sunrise and sunset as zeitgebers to entrain a circadian clock (Aschoff, 1965; Takahashi and Zatz, 1982; Neilson and Perry, 1990) or may represent optimal times for feeding on shallow-water prey, such as fiying fish, a main portion of the diet of dolphinfish (Olson and Galvan-Magana, 2002). Merten et al. (2014c) also posited that dawn and dusk were periods of transition that rep- resent the end and start, respectively, to more extensive vertical movements. Bluefin tuna {Thunnus thynnus) have been observed to occasionally interrupt their reg- ular day or night dive behavior at crepuscular periods (Gunn and Block, 2001). However, rather than remain- ing at shallow depths, these fish make “spike dives” that may have a navigational role through the detection of polarized light patterns or magnetic anomalies (Willis et al., 2009). The results presented here should be examined fur- ther in future studies. Because of resource constraints, our sample size is relatively small and does not encom- pass different life stages. Although location was deter- mined to be an insignificant factor in our model, there may be differences between populations that were un- detectable here. To help determine how vulnerable dol- phinfish are to fishing pressure, more work is needed to investigate the original drivers behind FAD association if they are not a source of prey and to determine sexual or life-stage preference for FAD associations. Further studies may incorporate the use of accelerometry (e.g., Furukawa et al., 2011) or further vertical movement profiles from time series data (e.g., Merten et al., 2014c) in order to further elucidate proximate drivers of the dif- ference in behavior between FAD-A and FAD-U fish that was observed in our study. Our results show that, as with results reported for bigeye tuna (Holland et al., 1990b, Schaefer and Fuller, 2005, 2010), it may be possible to use information on vertical behavior to assess when an individual dolphin- fish is associated with a floating object. Combined with the long-term data-recording capabilities of archival tag studies (Gunn and Block, 2001), information on the im- pact of FAD association on swimming depth could help elucidate the dependence of dolphinfish on floating ob- jects. This information is key in assessing the effects of fisheries on populations of dolphinfish given the increas- ing use of FADs by the purse-seine fisheries that target tunas and given the substantial bycatch of dolphinfish (Dagorn et al., 2013, Fonteneau et al., 2013, Leroy et al.. 2013). The findings presented here regarding the verti- cal distributions of FAD-A and FAD-U dolphinfish may help to develop sustainable fishing techniques and man- agement regulations of FAD-based fisheries. Acknowledgments Tagging in the Indian Ocean was cofunded by the European Union Fish Aggregating Devices as Instru- mental Observaties of pelagic ecosystems (FADIO) project (DG Research, QLRI-CT-2002-02773) and the European Union Dynamique et Organisation des Res- sources Associees aux Dispositifs Epipelagiques (DO- RADE) project (DIRED-Ifremer No. 31008/DIRED/JPP/ rp), with additional support from the Regional Council of Reunion Island. We are grateful to the crews of the MV Indian Ocean Explorer and FV Cap Morgan, to the French and Spanish skippers of the purse-seine fishing fleet in the Indian Ocean for their collaboration, and to C. White for analytical advice. C. Girard benefited from a grant provided by the Region Reunion. Literature cited Aschoff, J. 1965. The phase-angle difference in circadian periodic- ity. In Circadian clocks (J. Aschoff, ed.), p. 262-278. North Holland Press, Amsterdam. Bates, D., M. Machler, B. M. Bolker, and S. C. Walker. 2015. Fitting linear mixed-effects models using lme4. J. Stat. Softw. 67(l):l-47. Baughman, J. L. 1941. On a heavy run of dolphin, Coryphaena hippurus, off the Texas coast. Copeia 1941:117. Beardsley, G. L., Jr. 1967. Age, growth and reproduction of dolphin, Coryphaena hippurus, in the Straits of Florida. Copeia 1967:441-451. Boggs, C. H. 1992. Depth, capture time, and hooked longevity of longline- caught pelagic fish; timing bites of fish with chips. Fish. Bull. 90:642-658. Brill, R. W., D. B. Holts, R. K. C. Chang, S. Sullivan, H. Dewar, and F. G. Carey. 1993. Vertical and horizontal movements of striped marlin (Tetrapturus audax) near the Hawaiian Islands, deter- mined by ultrasonic telemetry, with simultaneaous mea- surement of ocean currents. Mar. Biol. 117:567-574. Casazza, T. L., and S. W. Ross. 2008. Fishes associated with pelagic Sargassum and open water lacking Sargassum in the Gulf Stream off North Carolina. Fish. Bull. 106:348-363. Chiang, W.-C., M. K. Muysl, C.-L. Sun, S.-Y. Chen S. Y, W.-Y. Chen, D.-C. Liu, W.-C. Su, S.-Z. Yeh, S.-C. Fu, and T. L. Huang. 2011. Vertical and horizontal movements of sailfish (/s- tiophorus platypterus) near Taiwan determined using pop- up satellite tags. J. Exp. Mar. Biol. Ecol. 387:129—135. Dagorn, L., D. Pincock D., C. Girard, K. Holland, M. Taquet, G. Sancho G., D. Itano, and R. Aumeeruddy. 2007. Satellite-linked acoustic receivers to observe behav- ior of fish in remote areas. Aquat. Living Resour. 20:307- 312. Whitney et al.: Swimming depth of Coryphaena hippurus 433 Dagorn, L., K. N. Holland, V. Restrepo, and G. Moreno. 2013. Is it good or bad to fish with FADs? What are the real impacts of the use of drifting FADs on pelagic marine ecosystems? Fish Fish. 14:391—415. Dempster, T., and M. Taquet. 2004. Fish aggregation device (FAD) research: gaps in cur- rent knowledge and future directions for ecological stud- ies. Rev. Fish Biol. Fish. 14:21-42. Dewar, H., E. D. Prince, M. K. Muysl, R. W. Brill, C. Sepulveda, J. Luo, D. Foley, E. S. Orbesen, M. L. Domeier, N. Nasby-Lu- cas, et al. 2011. Movements and behaviors of swordfish in the Atlan- tic and Pacific Oceans examined using pop-up satellite ar- chival tags. Fish. Oceanogr. 20:219-241. FAO (Food and Agriculture Organization of the United Nations). 2013. Species fact sheets: Coryphaena hippurus (Linnae- us, 1758). FAO, Rome. [Available at website, accessed March 2015.] Farrell, E. R., A. M. Boustany, P. N. Halpin, and D. L. Hammond. 2014. Dolphinfish {Coryphaena hippurus) distribution in relation to biophysical ocean conditions in the northwest Atlantic. Fish. Res. 151:177-190. Fonteneau, A., E. Chassot, and N. Bodin. 2013. Global spatio-temporal patterns in tropical tuna puse seine fisheries on drifting fish aggregating devices (DFADs): taking a historical perspective to inform current challenges. Aquat. Living Resour. 26:37-48. Furukawa, S., R. Kawabe, S. Ohshimo, K. Fujioka, G. N. Nishi- hara, Y. Tsuda, T. Aoshima, H. Kanehara H., and H. Nakata. 2011. Vertical movement of dolphinfish Coryphaena hippu- rus as recorded by acceleration data-loggers in the north- ern East China Sea. Environ. Biol. Fish. 92:89-99. Furukawa, S., S. Ohshimo, S. Tomoe, T. Shiraishi, N. Nakatsuka, and R. Kawabe. 2012. Age, growth, and reproductive characteristics of dol- phinfish Coryphaena hippurus in the waters off west Ky- ushu, northern East China Sea. Fish. Sci. 78:1153-1162. Article Furukawa, S., Y. Tsuda, G. N. Nishihara, K. Fujioka, S. Ohshi- mo, S. Tomoe, N. Nakatsuka, H. Kimura, T. Aoshima, H. Kane- hara, et al. 2014. Vertical movements of Pacific bluefin tuna {Thunniis orientalis) and dolphinfish {Coryphaena hippurus) relative to the thermocline in the northern East China Sea. Fish. Res. 149:86-91. Girard, C., L. Dagorn, M. Taquet, R. Aumeeruddy, C. Peignon, and S. Benhamou. 2007. Homing abilities of dolphinfish {Coryphaena hippu- rus) displaced from fish aggregating devices (FADs) deter- mined using ultrasonic telemetry. Aquat. Living Resour. 20:313-321. Gunn, J. S., and B. Block. 2001. Advances in acoustic, archival, and satellite tagging of tunas. Fish Physiol. 19:167-224. Holland, K., R. Brill, S. Ferguson, R. Chang, and R. Yost. 1985. A small vessel technique for tracking pelagic fish. Mar. Fish. Rev. 47:26-32. Holland, K. N., R. W. Brill, and R. K. C. Chang. 1990a. Horizontal and vertical movements of Pacific blue marlin captured and released using sportfishing gear. Fish. Bull. 88:397-402. 1990b. Horizontal and vertical movements of yellowfin and bigeye tuna associated with fish aggregation devic- es. Fish. Bull. 88:493-507. Hoolihan, J. P. 2005. Horizontal and vertical movements of sailfish {Is- tiophorus platypterus) in the Arabian Gulf, determined by ultrasonic and pop-up satellite tagging. Mar. Biol. 146:1015-1029. Hoolihan, J. R, J. Luo, F. J. Abascal, S. E. Campana, G. De Me- trio, H. Dewar, M. L. Domeier, L. A. Howey, M. E. Lutcavage, M. K. Musyl et al. 2011. Evaluating post-release behaviour modification in large pelagic fish deployed with pop-up satellite archival tags. ICES J. Mar. Sci. 68:880-889. Leroy, B., J. S. Phillips, S. Nicol, G. M. Pilling, S. Harley, D. Bromhead, S. Hoyle, S. Caillot, V. Allain, and J. Hampton. 2013. A critique of the ecosystem impacts of drifting and anchored FADs use by purse-seine tuna fisheries in the western and central Pacific Ocean. Aquat. Living Resour. 26:49-61. Massuti, E., Deudero S., Sanchez R, and Morales-Nin B. 1998. Diet and feeding of dolphin {Coryphaena hippu- rus) in western Mediterranean waters. Bull. Mar. Sci. 63:329-341. Merten, W., R. Appeldoorn, and D. Hammond. 2014a. Movements of dolphinfish {Coryphaena hippurus) along the U.S. East Coast as determined through mark and recapture data. Fish. Res. 151:114. 2014b. Spatial differentiation of dolphinfish {Coryphaena hippurus) movements relative to the Bahamian archipela- go. Bull. Mar. Sci. 90:849-864. Merten, W., R. Appeldoorn, R. Rivera, and D. Hammond. 2014c. Diel vertical movements of adult male dolphinfish {Coryphaena hippurus) in the western central Atlantic as determined by use of pop-up satellite archival transmit- ters. Mar. Biol. 161:1823-1834. Musyl, M. K., R. W. Brill, C. H. Boggs, D. S. Curran, T. K. Ka- zama, and M. R Seki. 2003. Vertical movements of bigeye tuna {Thunnus obesus) associated with islands, buoys, and seamounts near the main Hawaiian Islandsfrom archival tagging data. Fish. Oceanogr. 12:152-169. Neilson, J. D., and R. 1. Perry. 1990. Diel vertical migrations of marine fishes: an obligate or facultative process? In Advances in marine biology, vol. 26 (J. H. Blaxter and A. J. Southward, eds.), p. 115- 168. Academic Press, San Diego. Norton, J. G. 1999. Apparent habitat extensions of dolphinfish {Cory- phaena hippurus) in response to climate transients in the California current. Sci. Mar. 63:239-260. Olson, R. J., and F. Galvan-Magana. 2002. Food habits and consumption rates of common dol- phinfish {Coryphaena hippurus) in the eastern Pacific Ocean. Fish. Bull. 100:279-298. Oxenford, H. A. 1999. Biology of the dolphinfish {Coryphaena hippurus) in the western central Atlantic: a review. Sci. Mar. 63:277-301. Oxenford, H. A., and W. Hunte. 1999. Feeding habits of the dolphinfish {Coryphaena hip- purus) in the eastern Caribbean. Sci. Mar. 63:303- 315. Palko, B. J., G. L. Beardsley, and W. J. Richards. 1982. Synopsis of the biological data on dolphin-fishes, Coryphaena hippurus Linnaeus and Coryphaena equise- lis Linnaeus. FAO Fish. Synop. 130. NOAA Tech. Rep. NMFS Circ. 443, 28 p. 434 Fishery Bulletin 114(4) Rodriguez-Ferrer, G., Y. Rodriguez-Ferrer, D. Matos-Caraballo, and C. L. Yestrom. 2004. Comparison of dolphinfish {Coryphaena hippurus) commercial and recreational fisheries in Puerto Rico dur- ing 2000-2003. Proc. Gulf Caribb. Fish. Inst. 57:297-316. R Core Team. 2013. R: a language and environment for statistical com- puting. R Foundation for Statistical Computing, Vienna, Austria. [Available from website, accessed March 2015.] Rooker, J. R., J. P. Turner, and S. A. Holt. 2006. Trophic ecology of Sargassum-associated fishes in the Gulf of Mexico determined from stable isotopes and fatty acids. Mar. Ecol. Prog. Ser. 313:249-259. Rose, C. D., and W. W. Hassler. 1974. Food habits and sex ratios of dolphin Coryphaena hippurus captured in the western Atlantic Ocean off Hat- teras. North Carolina. Trans. Am. Fish. Soc. 103:94-100. Rudershausen, P. J., J. A. Buckel, J. Edwards J., D. P. Gannon, C. M. Butler, and T. W. Averett. 2010. Feeding ecology of blue marlins, dolphinfish, yellow- fin tuna, and wahoos from the North Atlantic Ocean and comparisons with other oceans. Trans. Am. Fish. Soc. 139:1335-1359. Schaefer, K. M., and D. W. Fuller. 2005. Behavior of bigeye (Thunnus obesus) and skipjack (Katsuwonus pelamis) tunas within aggregations associ- ated with floating objects in the equatorial eastern Pacific. Mar. Biol. 146:781-792. 2010. Vertical movements, behavior, and habitat of bigeye tuna (Thunnus obesus) in the equatorial eastern Pacific Ocean, ascertained from archival tag data. Mar. Biol. 157:2625-2642. Schaefer, K. M., D. W. Fuller, and B. A. Block. 2007. Movements, behavior, and habitat utilization of yel- lowfin tunas (Thunnus albacares) in the northeastern Pa- cific Ocean, ascertained through archival tag data. Mar. Biol. 152:503-525. 2011. Movements, behavior, and habitat utilization of yel- lowfin tunas (Thunnus albacares) in the Pacific Ocean off Baja California, Mexico, determined from archival tag data analysis, including unscented Kalman filter- ing. Fish. Res. 112:22-37. Schwenke, K. L., and J. A. Buckel. 2008. Age, growth, and reproduction of dolphinfish (Co- ryphaena hippurus) caught off the coast of North Caro- lina. Fish. Bull. 106:82-92. Sepulveda, C. A., S. A. Aalbers, S. Ortega-Garcia, N. C. Wegner, and D. Bernal. 2011. Depth distribution and temperature preferences of wahoo (Acanthocybium solandri) off Baja California Sur, Mexico. Mar. Biol. 158:917-926. Takahashi, J. S., and M. Zatz. 1982. Reculation of circadian rhythmicity. Science 217: 1104-1111. Taquet, M. 2004. Le comportement agregatif de la dorade coryphene (Coryphaena hippurus) autour des objets flottants. Ph.D. thesis, 168 p. Univ. Paris, Paris. Taquet, M., G. Sancho, L. Dagorn, J.-C. Gaertner, D. Itano, R. Aumeeruddy, B. Wendling, and C. Peignon. 2007. Characterizing fish communities associated with drifting fish aggregating devices (FADs) in the Western Indian Ocean using underwater visual surveys. Aquat. Living Resour. 20:331-341. Tremblay, A., and J. Ransijn. 2013. LMERConvenienceFunctions: a suite of functions to back-fit sized effects and forward-fit random effects, as well as other miscellaneaous functions. R package, vers. 2.5. Tripp-Valdez, A., F. Galvan-Magana F., and S. Ortega-Garci'a. 2010. Feeding habits of dolphinfish (Coryphaena hippurus) in the southeastern Gulf of California, Mexico. J. Appl. Ichthyol. 26:578-582. Trippel, E. A. 1995. Age at maturity as a stress indicator in fisher- ies. BioScience 45:759-771. Willis, J., J. Phillips, R. Muheim, F. J. Diego-Rasilla, and A. J. Hobday. 2009. Spike dives of juvenile southern bluefin tuna (Thun- nus maccoyii): a navigational role? Behav. Ecol. Socio- biol. 64:57-68. 435 NOAA National Marine Fisheries Service Fishery Bulletin ^ established 1881 -(f. Spencer F. Baird First U.S. Commissioner of Fisheries and founder of Fishery Bulletin First assessment of the field ecology of larval Atlantic silverside iMenidia menidia} Email address for contact author: dbengtson@uri.edu ' Department of Fisheries, Animal, and Veterinary Sciences University of Rhode Island 113 Woodward Hail 9 East Alumni Avenue Kingston, Rhode Island 02881 ^ Department of Computer Science and Statistics University of Rhode Island 246 Tyler Hall 9 Greenhouse Road, Suite 2 Kingston, Rhode Island 02881 Abstract — The Atlantic silverside {Menidia menidia) is extremely abundant in estuaries in eastern North America, is a significant com- ponent of food webs, and is the sub- ject of many laboratory studies; how- ever, the ecology of the larvae of this species in estuaries is poorly known. Using 4 simple collecting gears, we sampled Atlantic silverside larvae in 2 estuaries in Rhode Island that differ in anthropogenic inputs, Pet- taquamscutt River estuary and Point Judith Pond, to assess the distribu- tion and abundance of larvae of this species. These larvae occur predomi- nantly in waters less than 1 m deep and are patchily distributed. Larvae collected at depths of 0.2-0. 6 m were significantly shorter than those col- lected at depths of 0.6-0. 8 m — a dif- ference in mean total length of ~2 mm. We also compared diets and growth rates of larvae in the 2 estu- aries, using gut content analysis and otolith analysis, respectively. Cope- pod eggs made up 76% of the diet of larval Atlantic silverside in Pet- taquamscutt River, whereas copepod nauplii made up 73% of their diet in Point Judith Pond. Growth rates of the larvae did not differ between estuaries. Manuscript submitted 19 November 2015. Manuscript accepted 28 July 2016. Fish. Bull. 114:435-444 (2016). Online publication date: 23 August 2016. doi: 10.7755/FB.114.4.6 The views and opinions expressed or implied in this article are those of the author (or authors) and do not necessarily reflect the position of the National Marine Fisheries Service, NOAA. Miranda Lopez^ Gavino Puggioni^ David A. Bengtson (contact author)' The Atlantic silverside {Menidia me- nidia) is one of the most abundant estuarine species along the east coast of North America from Nova Scotia, Canada, to Florida (Middaugh et ah, 1981). Although this species has no commercial fishery value, it serves as a forage species for commercially important fish, such as bluefish (Po- matomus saltatrix), striped bass {Mo- rone saxatilis), and Atlantic mackerel {Scomber scombrus) (Fay et ah, 1983). It is perhaps best known scientifically because of the important laboratory experiments of Conover and Kynard (1981), whose work demonstrated that sex was determined by environ- mental factors, and of Conover and Munch (2002), whose investigation showed multigenerational reductions in fish size after size-selective “fish- ing” in experimental tanks. Because the adults of this species are easy to spawn in captivity (Barkman and Beck, 1976; Middaugh and Lempesis, 1976), larvae have been the subject of laboratory studies for decades. How- ever, surprisingly, the larval ecology of this species in the field is poorly known. In Rhode Island, Atlantic silver- side occupy estuaries from March through December and migrate to open water during the winter months. Adults return after winter in an emaciated condition, then feed heavily on zooplankton in March and April to become reproductively mature, which occurs between May and early July (Huber and Bengt- son, 1999). Spawning throughout the species range occurs with semilunar periodicity, and eggs are deposited in discrete areas in the upper inter- tidal reaches of salt marshes (Mid- daugh, 1981; Middaugh et ah, 1981; Conover and Kynard, 1984; Conover, 1985), the latter of which reduces egg predation by open-water predators (Tewksbury and Conover, 1987). In the upper reaches of 2 estuar- ies in Rhode Island, the upper Pet- taquamscutt River (UPR) and upper Point Judith Pond (UPJP), zooplank- ton communities are quite different in early spring when Atlantic silver- side adults return to feed and pre- pare for spawning (Bengtson, 1982; Huber, 1995; Volson, 2012). The zoo- plankton community in the UPR is 436 Fishery Bulletin 1 14(4) 71"35'0"W 7r30’0"W 7r25'0'W Figure 1 Map of the upper portions of the Pettaquamscutt River es- tuary and Point Judith Pond, the 2 estuaries in Rhode Island where larvae of Atlantic silverside {Menidia menid- ia) were sampled in 2012 for this study. The rectangle on the smaller map indicates the location of both estuaries in southern New England. The dots on the larger map show the approximate sampling locations in each estuary. dominated by crustaceans during early spring; their presence indicates a fairly pristine environment. The UPJP is dominated by polychaete larvae, which require a eutrophic environment. Given the general propensity of marine larval fish to feed on copepods, an a priori assumption might be made that Atlantic silverside in the UPR feed on higher quality prey than those in the UPJP. Volson (2012) examined effects of the nutrition- al quality of zooplankton prey from these 2 estuaries on adult Atlantic silverside and on their eggs, along with the hatching length of their larvae after incuba- tion in the laboratory. Surprisingly, in each of 2 years, length at hatching was greater for fish from the UPJP than for fish from the UPR. It has remained unclear whether a greater length at hatching translates into different larval growth rates during the first 2 weeks of life in the field. Therefore, in this study, we examined whether it does. Very little is known about the habitat ecology of Atlantic silverside larvae during their first 2-3 weeks of life in an estuary. Because newly hatched lar- vae in the laboratory are attracted to the interface between the water surface and the tank edge (D. Bengtson, personal observ.) and because adults spawn (and embryos hatch) in the upper inter- tidal, we suspected that larvae might be found in extremely shallow water. The feeding ecology of lar- val Atlantic silverside in the field is undocumented (Fay et al., 1983). Therefore, knowledge of critical elements of the field ecology of this important spe- cies during the larval period is lacking. The goals of our study, therefore, were 1) to determine the depth distribution of Atlantic silverside larvae, 2) to compare abundance and distribution of Atlantic silverside larvae between estuaries, 3) to compare feeding habits of the larvae in the 2 estuaries by analyzing gut contents, and 4) to compare growth of larvae in the 2 estuaries through age-length re- lationships based on otolith analysis. Materials ancd methods Study sites Field collections took place in the UPJP and UPR (Fig. 1). These estuaries are approximately 5 km apart, located in Washington County, Rhode Island, and have different physical characteristics (Table 1). Point Judith Pond is a shallow coastal lagoon, 1 of 7 along the southern coast of Rhode Island, connected to Block Island Sound by a breachway (LeeM. The Pettaquamscutt River is an annual fiooded river valley that drains into Narragansett Bay (Gaines, 1975). Abundance and distribution To determine distribution patterns and densities (abundance per cubic meters) of Atlantic silverside larvae in the field, 4 sampling devices were used: 1) a cylindrical polycarbonate quadrat (with a diameter of 0.5 m to sample the land-water interface); 2) an aquar- ium net, 19.05x26.03 cm with 500-pm mesh, to collect larval samples in water that was 0.05-0.80 m deep; 3) a plankton net with a diameter of 0.2 m, length of 0.6 m, and 200-pm mesh to collect samples in water 0.15-0.91 m deep, and 4) a second plankton net with a diameter of 0.5 m, length of 1.8 m, and 100-pm mesh to collect samples in water with depths slightly greater than 1 m. At both estuaries, collections occurred 7 days af- ter the new moon of 20 May 2012 and continued for 2 weeks. A second 2-week period of sampling occurred 7 days after the full moon of 4 June 2012. These dates were chosen to collect larvae hatched during those pre- 1 Lee, V. 1980. An elusive compromise: Rhode Island coastal ponds and their people. Univ. Rhode Island, Coast. Resour. Cent., Mar. Tech. Rep. 73, 65 p. [Available at website.] Lopez et al.: Field ecology of Menidia menidia 437 Table 1 Physical parameters of 2 estuaries, the upper Pettaquamscutt River (UPR) and upper Point Judith Pond (UPJP), from May through July 2012 documented within the program of the University of Rhode Island Watershed Watch (website). Parameters Time Depth (m) UPR Depth (m) UPJP Temperature May 2012 0.1 19.5 0.5 19.8 (°C) June 2012 24.0 21.0 July 2012 27.0 25.3 Average: 24 22 Practical salinity May 2012 0.1 16.5 0.5 - June 2012 15.0 27.5 July 2012 16.0 30.5 Average: 16 29 Fecal coliform May 2012 <10 478 (per 100 mL) June 2012 <10 189 July 2012 <10 84 Enterococci May 2012 <10 124 (per 100 mL) June 2012 <10 20 July 2012 124 30 Dissolved phosphorus May 2012 0.5 5 0.5 7 (pg/L) June 2012 <3 8 July 2012 4 29 Ammonium-nitrogen May 2012 0.5 45 0.5 60 (pg/L) June 2012 40 45 July 2012 25 75 Total phosphorus May 2012 0.5 16 0.5 42 (pg/L) June 2012 23 72 July 2012 35 107 sumed semilunar spawning periods Sampling began at 0630, an important time for determining foraging habits because this is the time when Atlantic silver- side begin feeding for the day. Initially, each sampling device was used at 4 locations within each estuary to try to identify microhabitat differences. Collections on and after 14 June in the UPJP were sampled from one location only because we determined that this estuary had undifferentiated microhabitat structure. At each of the locations within the UPR and UPJP, the quadrat, aquarium net and 2 plankton nets were each used 4 times (replicates) daily. Once sampling began, it contin- ued throughout the day until all tows and plots were complete or until weather conditions prohibited further sampling. The quadrat was haphazardly tossed at the land- water interface. Both plankton nets were pulled along 10-m transects by means of a rope; the senior author deployed the nets in the water, walked around the tran- sect with the rope, waited for any disturbed sediment to settle out of the water column, and quickly pulled the net over the 10-m distance. Finally, the aquarium net was pushed along a 10-m transect. However, on and after June 14, the aquarium net was pushed along a 1-m transect. Immediately before each sampling event, the water depth at the sampling point was measured in meters with a yard stick. In between each tow with the plankton nets and aquarium net, a 15-min waiting period allowed suspended sediment from the previous tow to settle. The quadrat on average sampled 0.010 of water per sample at the land-water interface. For field collections made before 14 June, the volume of water filtered by the aquarium net was 0.496 per sample. For field collections made on and after 14 June, the aquarium net filtered 0.049 m® of water per sample because of the shorter sampling transect. The small plankton net filtered 0.324 of water per sam- ple, whereas the large plankton net filtered 1.980 of water per sample. Although it was not possible to de- termine the sampling efficiency of each device, we as- sumed that the swimming speed of Atlantic silverside larvae is insufficient for them to avoid these sampling devices in any meaningful way. Larvae that were collected for laboratory analysis 438 Fishery Bulletin 114(4) Table 2 Catch data for each sampling device used in this study of Atlantic silverside {Menidia menidia) larvae in the upper Pettaquamscutt River (UPR) and upper Point Judith Pond (UPJP), Rhode Island, before 14 June 2012 (A), as well as on and after 14 June 2012 (B). Also included in each table are descriptions of the vol- ume of water filtered by each sampling device. Average densities of larvae, measured in number of fish per cubic meter, are provided with standard error (SE) values in parentheses. Total number Average density of of larvae collected larvae (fish/m®) (SE) Volume of water Sampling device sampled per tow (m®) UPR UPJP UPR UPJP A Quadrat 0.01 33 2 0.18 (SE 0.12) 0.01 (SE 0.01) Aquarium net 0.49 152 311 2.55 (SE 0.55) 10.11 (SE 3.21) Small plankton net 0.32 35 3 0.90 (SE 0.38) 0.11 (SE 0.08) Large plankton net 1.98 0 1 0 0.01 (SE 0.01) B Quadrat 0.01 8 1 0.05 (SE 0.03) 0.02 (SE 0.02) Aquarium net 0.049 25 24 6.00 (SE 3.70) 20.16 (SE 9.71) Small plankton net 0.32 2 0 0.07 (SE 0.05) 0 Large plankton net 1.98 0 0 0 0 were euthanized with tricaine methanesulfonate (MS- 222) mixed in seawater (90 g/mL) and preserved in ei- ther 95% ethanol (for otolith analysis) or 10% formalin (for gut content analysis). Each larva collected in the field was measured to the nearest 0.01 mm for total length (TL) with a dial caliper. Gut content analysis Foraging habits of Atlantic silverside larvae were de- termined by gut content analysis of preserved larvae collected from the field. In the laboratory, the gut was gently pulled apart and examined in a 50-mm Sedge- wick-Rafter counting cell under a compound microscope. Each prey item was tallied and identified to the lowest possible taxon. From fish collected from the UPJP, 58 guts were examined. From fish sampled from the UPR, 51 guts were examined. All larvae dissected were be- tween 4.18 and 9.36 mm TL for both estuaries. The num- ber method was used to show food type as a percentage of the total gut contents of each larva (Hyslop, 1980). Each taxon was represented as a percentage of the total gut contents for all the larvae dissected for each estuary. Otolith analysis In the laboratory, one sagittal otolith was extracted from each larva, placed on a microscope slide with one drop of immersion oil (Grade A, Cargille-Sacher Labo- ratories^, Cedar Grove, NJ), photographed by using a ^ Mention of trade names or commercial companies is for iden- tification purposes only and does not imply endorsement by the National Marine Fisheries Service, NOAA. light microscope camera (at lOOx or 400x magnifica- tion), and its rings were counted by using the methods of Barkman (1978). Otoliths were first examined by using light microscopy (400x magnification) to count daily growth rings and in turn to determine age (in days). Measurements of the diameters of the sagittae were taken as a proxy for growth. At a later time, a second reading was completed by the same observer from the photographs taken of the sagittal otoliths. Six prehatching rings were subtracted from the total num- ber of daily rings on each sagittal otolith because Bark- man determined that the rings start to be laid down 6 days before hatching (Barkman, 1978). The sagittae did not require additional processing because the core was visible. The relation of the number of daily rings (age in. days) to larval length was determined for fish from each estuary. The slopes of these linear relation- ships provide an estimate of growth (in millimeters per day) of larvae in each of the estuaries. All measure- ments were made in micrometers with the computer software program Imaged (Abramoff et al., 2004). We examined 17 larvae from the UPJP and 19 larvae from the UPR. Statistical analysis The relationship between age and length of larvae was determined for each estuary and the slopes of these re- gressions were analyzed with an analysis of covariance (ANCOVA). A chi-square analysis was applied to the gut content data to determine a significant difference between the feeding habits of Atlantic silverside larvae from UPJP and the feeding habits of Atlantic silverside larvae from UPR. Lopez et ai.: Field ecology of Menidia menidia 439 n cs n o ’O o '“D 0 ”u C 0 O O” 0 •D d 0 £ 0 CD E ° o X Observed frequencies + ZINB estimated probabilities rTTTTTTTrrrmTT 0 4 8 13 nTrnn ii'iii mil I III 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 19 25 31 37 43 49 Counts 1 1 1 1 ! I nin rmn 55 61 67 Observed frequencies Hurdle-NB estimated probabilities 0 1 2 3 4 5 6 7 8 Counts 15 n m n o T3 0 'U @ a. 'm q 'o c 0 3 O” © “O 0 £ 0 m n O B X Observed frequencies + ZiNB estimated probabilities 10 — I — 20 — \ 1 — 30 40 Counts 50 60 o d D X Observed frequencies + Hurdle-NB estimated probabilities 8 Counts —\ — 10 'k — I — 12 * — I 14 Pigure 2 The frequency of occurrence of the total number of larvae of Atlantic silverside {Menidia menidia) collected from both the upper Point Judith Pond and the upper Pettaquamscutt River (A and B) before 14 June 2012 and (C and D) on and after 14 June 2012. For all graphs, the gray symbols represent the probability estimates following the zero-inflated negative binomial (ZINB) or the hurdle negative binomial (hurdle-NB) models used for estimation. The black symbols in each graph represent the observed relative frequencies of the total number of larvae collected from the field. Both models take into account the probability of ob- serving a zero (absence of Atlantic silverside larvae), as well as observing a positive value (presence and abundance of Atlantic silverside larvae). Panels B and D differ from panels A and C in that the former lack the zero probability estimates and observed values. The components of both models depend on covariates, and that is why occupancy and abundance are modeled jointly. All graphs also show how many Atlantic silverside larvae were collected from the field per tow. 70 Given the large presence of zeros in the abundance data collected from the field, we worked with 2 differ- ent classes of generalized linear models that explicitly take into account this feature: zero-inflated count mod- els and hurdle count models. A zero-inflated model is a mixture of 2 distributions: a binary distribution for structural zeros and a distribution for the counts (typi- cally Poisson or negative binomial) that can be zero or positive. A hurdle model is similar, except that all the zeros are modeled with the binary distribution and the distribution for the counts models only the positive values. For each class of models, we fitted 2 different distributions: the Poisson and the negative binomial (White and Bennetts, 1996). The Poisson distribution is appropriate when mean and variance of the data are similar, and the negative binomial features an addi- tional parameter to measure overdispersion (variance larger than the mean) in the data. All these models allow specifications of covariates that can be different for the zero-inflated part and the count part. Possible covariates that could be predic- tors of abundance were depth of the measurements, date, site effect (UPJP versus UPR), type of collec- tion gear or net effect (quadrat, aquarium net, small plankton net, and large plankton net). Estimation was performed by using maximum likelihood methods and implemented in the package pscl, vers. 1.4.9 (Jackman, 2015; Zeileis et al., 2008) in the statistical software R, vers. 3.2.3 (R Core Team, 2015). Statistical significance for the parameters was assessed by using Z-tests. Giv- en the relatively small amount of nonzero data, we had to choose the predictors carefully, using Akaike infor- 440 Fishery Bulletin 1 14(4) Table 3 Parameter estimates from the zero-inflated negative minomial (ZINB) model for field samples of larval Atlantic silverside {Menidia menidia) collected before 14 June 2012. The top part of this table includes parameters from the ZINB model that analyzes all data values greater than zero (i.e., when a larva was collected). Site refers to the upper Pettaquamscutt River and upper Point Judith Pond in Rhode Island. Date represents the duration of sampling, 30 May 2012 through 13 June 2012. The Z statistic tests whether the probability of collecting a larva is significantly influenced by site, depth, the quadrat, the aquarium net, and the small plankton net. The bottom table includes the parameters influencing the probability of having a count of zero in the data; the Z statistic tests whether the probability of not collecting a larva is significantly affected by 2 of the sampling devices. SE=standard error. Estimate SE Z-value P-value Significance Count model coefficients (negative binomial with log link) Date -0.0405 0.0506 -0.802 0.4226 NS Depth 0.5182 1.2352 0.420 0.6748 NS Aquarium net 1.6155 0.7151 2.259 0.0239 ** Quadrat -1.9364 1.0629 -1.822 0.0684 * Small plankton net -2.0110 1.2911 -1.558 0.1193 NS Site [2 sites] -1.2817 0.4146 -3.092 0.0020 *** Quadrat:Site[2 sties] 3.8550 1.1643 3.311 0.0009 Small plankton net:Site[2 sites] 3.5268 0.9957 3.542 0.00040 Log( theta) -1.6682 0.1635 -10.206 <0.0001 Zero-inflated model coefficients (binomial with logit link) Aquarium net -10.8903 381.4482 -0.029 0.9772 NS Quadrat 1.4663 0.5074 2.89 0.0039 Small plankton net 1.0366 0.5092 2.036 0.0418 ** Significance codes: **** (P<0.001); *** (P<0.01); ** (P<0.05); * (P<0.1); NS=not significant. mation criterion (AIC). We analyzed the data collected before 14 June separately from data collected on and after that date. Finally, to determine whether the sizes of the lar- vae changed with depth of capture, we used analysis of variance (ANOVA) to compare the lengths of larvae collected from the following 5 depth strata: 0.0-0. 2 m; 0.2-0. 4 m; 0.4-0. 6 m; 0.6-0. 8 m; and >0.8 m. Results Abundance and distribution in the field 2 Average density and the number of larvae collected by each sampling device (Table 2) indicated that only one Atlantic silverside larva was collected with the large plankton net; therefore, this device was excluded from the remainder of the analysis. The quadrat, aquarium net, and small plankton net all collected more larvae than the large plankton net. This finding indicates that Atlantic silverside larvae are generally not found in waters greater than 1 m deep but can be found in wa- ters less than 1 m deep in the littoral zone. The quad- rat, the aquarium net, and small plankton net can all be used to collect Atlantic silverside larvae from the littoral zone of estuaries. Distribution and abundance of Atlantic silverside larvae have a discrete distribution with a high fre- quency of zeros on the basis of our analysis of field collections (Fig. 2), indicating that Atlantic silverside larvae have a patchy distribution in the littoral zone. In terms of maximum numbers per tow, for field collec- tions made before 14 June, up to 69 larvae were collect- ed per tow (Fig. 2, A and B). For field collections made on and after 14 June, up to 15 larvae were collected per tow (Fig. 2, C and D). The results of model analysis provided the predictors that influenced the number of larvae collected, as well as the predictors that influenced the structural zeros in the data (Tables 3 and 4). For the data collected before 14 June, the use of AIC indicated that the chosen model was a zero-inflated negative binomial. The presence of larvae in the littoral zone in both estuaries correlated with date, depth, and all sampling devices, as well as with interaction terms (Table 3). Date was found to be Lopez et al.: Field ecology of Menidia menidia 441 Table 4 Parameter estimates for the negative binomial hurdle model obtained from field samples of larval Atlantic silverside (Menidia menidia) collected on and after 14 June 2012. The top table includes parameters from the model that analyzes all data values greater than zero (i.e., when a larva was collected). Date repre- sents the duration of sampling, from 14 June 2012 through 25 June 2012. The Z statistic tests whether the probability of collecting a larva is significantly influenced by depth and site. The bottom table includes the parameters influencing the probability of having a count of zero in the data. Site l=Upper Point Judith Pond; site 2=Upper Pettaquamscutt River. SE=standard error. Estimate Standard error (SE) Z-value F-value Significance Count model coefficients (truncated negative binomial with log link) Depth 13.272 5.614 2.364 0.0181 Site 1 -14.943 64.922 -0.230 0.8180 NS Site 2 -9.885 64.827 -0.152 0.8788 NS Log( theta) -9.077 64.822 -0.140 0.8886 NS Zero-inflated hurdle model coefficients (binomial with logit link) Depth 6.7017 2.0211 3.316 0.0009 Aquarium net -3.4220 0.6359 -5.381 <0.0001 Quadrat -3.9519 1.0082 -3.920 <0.0001 Small plankton net 1.5323 2.2454 0.682 0.4950 NS Depth : Quadrat 1.4099 9.3280 0.151 0.8799 NS Depth : Small plankton net -22.8333 7.4331 -3.072 0.0021 Significance codes; ***(P<0.01); **(P<0.05); NS=not significant. nonsignificant (date="0-04, Z=-0.802, P=0.422; Table 3). The number of larvae collected did not increase with depth in the littoral zone (depth^O-SlS, Z=0.420, P=0.674; Table 3). The different nets all had significant effects for the count part of the model. The quadrat and the small plankton net influenced the presence of zeros in the data to a larger extent than that of the aquarium net. Notice that all the estimated coefficients (Table 3) need to be interpreted, considering that the model is in the log scale (the negative binomial part) and in the logit scale (the zero-inflated part). For field collections made on and after 14 June, giv- en the smaller number of positive counts, we needed to implement a more parsimonious model, with fewer predictors. We tried several different specifications. For simplicity, we report that the negative binomial hurdle model offered the best performance in terms of AIC. Depth influenced the counts; depth and net type in- fluenced the number of structural zeros. The number of larvae collected increased significantly with depth in the littoral zone (depth“13-272, Z=2.364, P=0,0181; Table 4). One of the goals of our project was to determine whether the density of Atlantic silverside larvae dif- fered between the 2 estuaries. The results of the zero- inflated Poisson analysis for site indicated that, for the entire sampling period, the UPR had a higher density of Atlantic silverside larvae than the UPJP. Results of the AN OVA of larval lengths in the depth strata from the shoreline to depths >0.8 m indicated that, at the UPR, larvae captured at the depth stra- tum 0.2-0. 4 m were significantly shorter (mean: 10.1 mm TL [standard deviation (SD) 3.4]) than larvae cap- tured at the depth stratum 0.6-0. 8 m (mean: 14.2 mm TL [SD 2.7]). At the UPJP, larvae captured at both the depth strata of 0.2-0. 4 m and 0.4-0. 6 m were signifi- cantly shorter (means: 7.7 mm TL [SD_2.5] and 7.8 mm TL [SD 2.2], respectively) than larvae captured at the depth stratum 0.6-0. 8 m (9.5 mm TL [SD 3.1]). Gut content analysis Feeding habits of Atlantic silverside larvae between estuaries were significantly different (x^=622.7, F<0.0001). In the UPJP, copepod nauplii made up 72.5% of total gut contents (Fig. 3). In the UPR, At- lantic silverside larvae consumed mostly copepod eggs, which made up 76.2% of the total gut contents (Fig. 3). 442 Fishery Bulletin 114(4) Copepod eggs □ Copepod nauplii Q Barnacle nauplii [1] Rotifera B Cladocerans 11 Polychaete larvae I Unknown nauplii B Unknown H Actinula larvae B Veliger larvae H Cyprids B Spiricules ^ Pieces of sponge m Figure 3 Gut contents of larvae of Atlantic silverside (Menidia menidia) collected in 2012 (A) in the upper Pettaquamscutt River (n=51) and (B) in the upper Point Judith Pond (n=58). Each taxon is represented as a percentage of the total gut contents for all larvae collected in each of these 2 estuaries in Rhode Island. Results from the chi-square analysis indicate a significant dif- ference in feeding habits of Atlantic silverside larvae (P<0.0001). Otolith analysis Results from the ANCOVA showed a significant re- lationship between length of larvae and age for fish from both estuaries (P<0.0001; Fig. 4) The age-length regressions indicated that larvae grow 0.65 mm/d in the UPR and 0.66 mm/d in the UPJP (Fig. 4). The re- sults from the ANCOVA showed no significant differ- ence (P=0.8147) between estuaries in the age-length relationship of larvae (Fig. 4). Discussion This article provides the first detailed report of the field ecology of Atlantic silverside larvae, although laboratory studies on this larval species have been conducted for decades (e.g., Austin et ah, 1975; Mid- daugh and Lempesis, 1976; Morgan and Prince, 1977; Deacutis, 1978; Bengtson, 1985; Lankford et ah, 2001). Field collections from the littoral zone of the UPR and UPJP during the summer of 2012 in- dicate that this larval fish can be captured at the shoreline interface to waters 1 m deep, can be collected with a variety of sampling devices, and has a patchy distribution. The 2 estuaries sampled differed in abundance of Atlantic sil- verside larvae and in the prey consumed by the larvae, but the larvae grew at the same rates regardless of those differences. The quadrat, aquarium net, and small plank- ton net collected more larvae than the large plankton net; however, the aquarium net col- lected the most Atlantic silverside larvae both in absolute and per-volume-sampled terms. We acknowledge that our use of different sam- pling devices, chosen by necessity because of the shallow depths, and the methods we used for deploying those devices add a degree of uncertainty to our results. Nevertheless, it is clear that Atlantic silverside larvae can be col- lected from the shallowest water out to about a 1-m depth. Middaugh (1981), Middaugh et al. (1981), Conover and Kynard (1984), and Conover (1985) documented Atlantic silverside adults depositing eggs en masse at very dis- crete spawning sites in the upper intertidal zone of salt marshes. The results of our study indicate that Atlantic silverside larvae stay in the very shallow littoral zone after hatching. The distribution of Atlantic silverside larvae collected from the field in this study followed a zero-inflated Poisson model, indicating that these larvae are not distributed evenly in the littoral zone and have a patchy distribution. The patchiness that we found in the distri- bution of Atlantic silverside larvae may be re- lated to the patchiness of the egg deposition sites, although we did not specifically try to test that idea. Hewitt (1981) proposed that fish larvae have a patchy distribution because it benefits schooling. Shaw (1960, 1961) reported that Atlantic silverside begin to school at a size around 11-12 mm standard length. Lindsay et al. (1978) sampled ichthyo- plankton in the Indian River, Delaware, in deeper wa- ters than those that we sampled, and noted that the low abundance of atherinid larvae was not representa- tive of their high abundance as juveniles and adults. Occupation of very shallow waters by Atlantic silver- side larvae likely provides them with protection from predators. The fact that we found shorter larvae in the shallower waters and larger larvae in slightly deeper waters within the depth stratum of <1 m within both estuaries indicates that these larvae are avoiding pred- ators, or just diffusing very slowly from the hatching sites, or doing both. Gut content data indicated that Atlantic silver- side larvae in the UPR consume mostly copepod eggs, whereas larvae in the UPJP consume mostly copepod nauplii. Volson (2012) found a high abundance of cal- Lopez et al.: Field ecology of Menidia menidia 443 18 -] 16 - 14 - E E 12 - 10 - 05 C m 8 - “5 1 6 - 4 - oUpper Point Judith Pond ♦Upper Pettaquamscutt River ANCOVA Age: P<0.0001 Site: P=0.8147 5 10 Age (days) 15 20 Figyre 4 Growth of larvae of Atlantic silverside {Menidia menidia) collected in 2012 from the upper Point Judith Pond (UPJP) and upper Pettaquams- cutt River (UPR) in Rhode Island. Linear regressions represent the age-length relationship of Atlantic silverside larvae from the UPJP (dashed line), 3/=0.66x+2.98, and the UPR (solid line), y=0.65x+3.06. In the linear equations, y is total length in millimeters and x is age in days. Results from the analysis of covariance (ANCOVA) indicate no significant differences in the slopes of the age-length relationship of larvae between estuaries (P=0.8147). However, age is a significant indicator of the size of larvae (P<0.0001). anoid copepods in the UPR and a varying zooplank- ton community in the UPJP, where from early spring (April through early May) through late spring (June), the dominant zooplankton present switches from poly- chaete larvae to copepods. Most of the sampling for our study took place in late spring, when the polychaete larvae had already settled and were not available to, or not preferred by, the larvae. The exact species of cope- pod from which the eggs came from, for our study, was not determined. The significant age-length relationship for Atlantic silverside larvae in our study has been shown previ- ously in work by Barkman (1978). Between estuaries, there was no significant difference in the age-length relationship of Atlantic silverside laiwae. According to the regression coefficients in the age-length equations, the larvae in our study grew at 0.65-0.66 mm/d. Bark- man et al. (1981) found that over a length range of about 12-90 mm TL Atlantic silverside grew at 0.84 mm/d, on the basis of an age-length relationship deter- mined with otolith analysis, whereas Mulkana (1966) estimated a growth rate of 7-11 mm/month (0.23-0.37 mm/d) on the basis of length-frequency analyses of a cohort. Volson (2012) found that larval length at hatch- ing was significantly greater for Atlantic silverside lar- vae in the UPJP than for larvae in the UPR. The results from our study indicate that a greater length at hatching does not translate into faster growth for larval Atlantic silver- side. Temperature influences the growth of fish. Water temperatures in the UPJP are cooler than the water temperatures in the UPR, even during the summer months (Volson, 2012) when sampling occurred for our study (Table 1). Despite a greater length at hatching for Atlantic silverside in the UPJP, the cooler water temperature in this estuary may have resulted in a slower growth rate for the larvae. As a result, larval growth v/as not greater in the UPJP than larval growth in the UPR. The larval life stage is important for recruitment of adult populations. Studies on Atlantic silverside are important not only because of the abundance of this spe- cies but also because of its role as a for- age fish for fisheries species and the way in which these fish influence the energet- ics of estuaries. We hope that the initial information presented here will stimulate researchers of estuaries to further exam- ine the larval ecology of this important species of estuarine ecosystems. Acknowledgments We thank the College of the Environment and Life Sci- ences at the University of Rhode Island for a diver- sity teaching assistantship to M. Lopez. We also thank M. Rice for use of a microscope and image analysis software. Literature cited Abramoff, M. D., P. J. Magalhaes, and S. J. Ram. 2004. Image processing with Imaged. Biophotonics Int. 11:36-42. Austin, H. M., A. D. Sosnow, and C. R. Hickey Jr. 1975. The effects of temperature on the development and survival of the eggs and larvae of the Atlantic silverside, Menidia menidia. Trans. Am. Fish. Soc. 104:762-765.115 Barkman, R. C. 1978. The use of otolith growth rings to age young Atlan- tic silversides, Menidia menidia. Trans. Am. Fish. Soc. 107:790-792. Barkman, R. C., and A. D. Beck. 1976. Incubating eggs of the Atlantic silverside on nylon screen. Prog. Fish-Cult. 38:148-150. Barkman, R. C., D. A. Bengtson, and A. D. Beck. 1981. Daily growth of the juvenile fish {Menidia menidia) in the natural habitat compared with juveniles reared in the laboratory. Rapp. p.-v. Reun. - Cons. Int. Explor. Mer 178:324-326. 444 Fishery Bulletin 114(4) Bengtson, D. A. 1982. Resource partitioning by Menidia menidia (Lin- naeus) and Menidia beryllina (Cope) in two Rhode Is- land estuaries. Ph.D. diss., 214 p. Univ. Rhode Island, Kingston, RI. 1985. Laboratory experiments on mechanisms of competi- tion and resource partitioning between Menidia menidia (L.) and Menidia beryllina (Cope) (Osteichthyes: Atherin- idae). J. Exp. Mar. Biol. Ecol. 92:1-18. Conover, D. O. 1985. Field and laboratory assessment of patterns in fe- cundity of a multiple spawning fish: the Atlantic silver- side Menidia menidia. Fish. Bull. 83:331-341. Conover, D. O., and B. E. Kynard. 1981. Environmental sex determination: interaction of temperature and genotype in a fish. Science 213:577- 579. 1984. Field and laboratory observations of spawning peri- odicity and behavior of a northern population of the At- lantic silverside, Menidia menidia (Pisces: Atherinidae). Environ. Biol. Fish. 11:161-171. Conover, D. O., and S. B. Munch. 2002. Sustaining fisheries yields over evolutionary time scales. Science 297:94-96. Deacutis, C. F. 1978. Effect of thermal shock on predator avoidance by larvae of two fish species. Trans. Am. Fish. Soc. 107:632- 635. Fay, C. W., R. J. Neves, and G. B. Pardue. 1983. Species profiles: life histories and environmental re- quirements of coastal fishes and invertebrates (Mid-At- lantic)— Atlantic silverside. U.S. Fish Wildl. Serv., Div. Biol. Serv., FWS/OBS-82/11.10. U.S. Army Corps Eng., TR EL-82-4, 15 p. [Available at website.] Gaines, A. G. 1975. Papers on the geomorphology, hydrography and geo- chemistry of the Pettaquamscutt River estuary. Ph.D. diss., 278 p. Univ. Rhode Island, Kingston, RI. [Avail- able at website. [ Hewitt, R. 1981. The value of pattern in the distribution of young fish. Rapp, p.— V. Reun. Cons. Int. Explor. Mer 178:229-236. Huber, M. 1995. Environmental control of reproduction in an es- tuarine fish, Menidia beryllina (Cope) in Rhode Is- land. Ph.D. diss., 167 p. Univ. Rhode Island, Kingston, RI. Huber, M., and D. A. Bengtson. 1999. Interspecific differences in growth of somatic and re- productive tissues during the breeding season in Menid- ia menidia and M. beryllina. J. Fish Biol. 55:274-287. Hyslop, E. J. 1980. Stomach contents analysis — a review of methods and their application. J. Fish Biol. 17:411-429. Jackman, S. 2015. pscl: classes and methods for R developed in the Po- litical Science Computational Laboratory, Dep. Political Sci., Stanford Univ., Stanford, CA. Manual for R pack- age vers. 1.4.9. [Available at website.] Lankford, T. E., Jr., J. M. Billerbeck, and D. O. Conover. 2001. Evolution of intrinsic growth and energy acquisi- tion rates. II. Trade-offs with vulnerability to predation in Menidia menidia. Evolution 55:1873-1881. Lindsay, J. A., E. R. Radle, and J. C. S. Wang. 1978. A supplemental sampling method for estuarine ich- thyoplankton with emphasis on the Atherinidae. Estu- aries 1:61-64. Middaugh, D. P. 1981. Reproductive ecology and spawning periodicity of the Atlantic silverside, Menidia menidia (Pisces: Atherinidae). Copeia 1981:766-776. Middaugh, D. R, and P. W. Lempesis. 1976. Laboratory spawning and rearing of a marine fish, the silverside Menidia menidia menidia. Mar. Biol. 35:295- 300. Middaugh, D. R, G. 1. Scott, and J. M. Dean. 1981. Reproductive behavior of the Atlantic silverside, Menidia menidia (Risces, Atherinidae). Environ. Biol. Fish. 6:269-276. Morgan, R. R, II, and R. D. Rrince. 1977. Chlorine toxicity to eggs and larvae of five Chesa- peake Bay fishes. Trans. Am. Fish. Soc. 106:380-385. Mulkana, M. S. 1966. The growth and feeding habits of juvenile fishes in two Rhode Island estuaries. Gulf Res. Rep. 2:97-167. R Core Team. 2015. R: a language and environment for statistical com- puting. R Foundation for Statistical Computing, Vien- na, Austria. [Available at website, accessed December 2015.] Shaw, E. 1960. The development of schooling behavior in fishes. Physiol. Zool. 33:79-86. 1961. The development of schooling behavior in fishes. H. Physiol. Zool. 34:263-272. Tewksbury, H. T., II, and D. O. Conover. 1987. Adaptive significance of intertidal egg deposition in the Atlantic silverside Menidia Men jdfa. Copeia 1987:76- 83. Volson, B. 2012. Effects of zooplankton community structure and nu- tritional value on the eggs and larvae of two estuarine fish species. Ph.D. diss., 288 p. Univ. Rhode Island, Kingston, RI. [Available at website.] White, G. C., and R E. Bennetts. 1996. Analysis of frequency count data using the negative binomial distribution. Ecology 77:2549-2557. Zeileis, A., C. Kleiber, and S. Jackman. 2008. Regression models for count data in R. J. Stat. Software 27(8): 1-25. 445 NOAA National Marine Fisheries Service Fishery Su/teftn established 1881 Spencer F. Baird First U.S. Commissioner of Fisheries and founder of Fishery Bulletin Byrrowing behavior, habitat, and functional morphology of the Pacific sand lance UkmmodytBS personatus} Email address for contact author: ibizzarro@mlml.calstate.edu ' School of Aquatic and Fishery Sciences University of Washington Box 355020 Seattle, Washington 98195-5020 Present address: Center for Habitat Studies Moss Landing Marine Laboratories 8272 Moss Landing Road Moss Landing, California 95039 ^ Department of Ecology and Evolutionary Biology University of California, Irvine 321 Steinhaus Hall Irvine, California 92697 Washington Department of Fish and Wildlife 600 Capitol Way N. Olympia, Washington 98501 Friday Harbor Laboratories University of Washington 620 University Road Friday Harbor, Washington 98250 Kingston, Rhode Island 02881 Abstract— -The Pacific sand lance (Ammodytes personatus) is a small, elongate forage fish that spends much of its life buried in the sea- floor. We determined that the Pa- cific sand lance can burrow in a wide variety of sediments from silt to gravel, but it prefers coarse sand (0.50-1.00 mm grain size). In the absence of coarse sand, the Pacific sand lance chooses larger grain sizes over smaller ones. These preferences are independent of light or the com- paction of sediment, and therefore indicate that visual cues and ease of entry are not primary means of choosing burial substrate. Instead, we speculate that the Pacific sand lance is morphologically adapted for rapid mobility in coarse sand and that coarse sand has enough interstitial spaces to enable respira- tion during protracted immersion. As an obligate borrower in specific sediments, the Pacific sand lance is a good candidate for habitat-based management. Substrate maps of 3 fishing grounds in southeast Alas- ka where the Pacific sand lance is abundant and where habitat-based management is practiced were used to create potential habitat maps. Different geologic histories have re- sulted in variable amounts of pre- ferred (sand-gravel), suitable (sand mixed with silt, cobble-boulder, or rock outcrop), and unsuitable (mud, pebble-boulder) habitat for this spe- cies among regions. Manuscript submitted 11 November 2015. Manuscript accepted 3 August 2016. Fish. Bull: 114:445-460 (2016). Online publication date: 25 August 2016. doi: 10.7755/FB.114.4.7 The views and opinions expressed or implied in this article are those of the author (or authors) and do not necessarily reflect the position of the National Marine Fisheries Service, NOAA. Joseph J. Bizzarro (contact author)^ Ashley N. Peterson^ Jennifer IVl. Blaine® Jordan P. Balaban^ H. Gary Greene^ Adam P. Summers^ Burrowing presents biomechanical and ecological challenges but enables access to expanded trophic oppor- tunities and protection from many predators. The selective pressures surrounding burrowing are expressed in the morphological features of bur- rowing animals, as in the stout fore- arms of moles and armadillos and in the heavily reinforced skull of caeci- lians and dibamids (Kleinteich et ah, 2012; Rose et ah, 2013). A subterra- nean lifestyle is uncommon for aquat- ic vertebrates but has been observed in a taxonomically diverse group of marine fish taxa. Flounders and skates routinely cover themselves in substrate; jawfishes, tilefishes, and garden eels excavate permanent bur- rows; and some fishes (e.g., Pacific sandfish, Trichodon trichodon; sand lances, Ammodytes spp.) spend a ma- jority of their life beneath the sub- strate after creating a tunnel that collapses behind them. The terres- trial equivalent of this behavior is seen in the “sand-swimming” skink species (Mushinsky and Gans, 1992; Maladen et ah, 2011). Penetrating friable substrates in a completely aqueous environment is fundamen- tally different from terrestrial sand swimming by virtue of the density of water and its potential contribution to the excavation process. The Pacific sand lance {Ammodytes personatus) is an elongate, burrowing forage fish with a wide distribution in the eastern North Pacific and a his- tory of taxonomic confusion. Only A. hexapterus, formerly the Pacific sand lance and now assigned the common name of Arctic sand lance (Orr et ah, 2015), and A. personatus were con- sidered historically as valid North Pacific species, but the number of Ammodytes species in the North Pa- cific region and the extent of their 446 Fishery Bulletin 1 14(4) distributions have long been debated (Ohshima, 1950; Han et ah, 2012; Turanov and Kartavtsev, 2014). Recent genetic and morphological evidence has resolved this is- sue and indicates the presence of 2 additional congeners (Orr et ah, 2015). Furthermore, the only species with an expansive eastern North Pacific distribution, formerly considered to be A. hexapterus, was redescribed as A. personatus (Orr et ah, 2015). The range of A. personatus was established from southern California to the western Aleutian Islands and may extend to the Sea of Okhotsk in the western Pacific (Mecklenberg et al., 2011; Orr et al., 2015). Ammodytes personatus mainly occurs in coastal intertidal and subtidal waters but has been re- ported at depths of 172 m (Love et ah, 2005). The Pacific sand lance is a source of energy trans- fer between secondary producers and upper-trophic- level species and pelagic and benthic regions because it grazes on zooplankton in the water column and has an obligate affiliation with sediments. The structure and dynamics of nearshore ecosystems are heavily in- fluenced by the biomass of forage fishes (Gaichas et al., 2010) , and species oi Ammodytes, including the Pacific sand lance, are of vital importance for the energetics and breeding success of a variety of marine mammals (Weinrich et ah, 1997), seabirds (Curry et al., 2011), and fishes (Arnott et al., 2002). For instance, hump- back whales {Megaptera novaeangliae) at Stellwagon Bank, Massachusetts, excavate bottom sediments at night to forage on dense aggregations of buried sand lances [Ammodytes spp.) (Friedlander et al., 2009). Off British Columbia, growth rates of rhinoceros auklet (Cerorhinca moncerata) chicks are positively corre- lated with abundance of Pacific sand lance (Bertram and Kaiser, 1993). Several groundfishes (e.g., starry flounder [Platichthys stellatus]; great sculpin [Myoxo- cephalus polyacanthocephalus] have been reported to feed on schools of Pacific sand lance as they move from foraging to burial grounds off southeast Alaska (Hob- son, 1986). Where species of Amtnodytes are exploited in com- mercial fisheries, the associated loss of forage bio- mass can have ecosystem-level effects. In the North Sea, overfishing of the sand eel (A. marinus) has been linked to poor breeding success of several seabird spe- cies (Arnott et ah, 2002), and the prosecution of a fishery for the sand eel has negatively impacted the breeding productivity of the population of black-legged kittiwake {Rissa tridactyla) (Frederiksen et al., 2008). Conversely, an abundance of Ammodytes species can enhance the productivity and efficiency of groundfish fisheries. For instance, when biomass of Ammodytes species is relatively high, Atlantic cod (Gadus morhua) form feeding aggregations in small, predictable areas (i.e., over burial habitat) where they can be targeted easily (Richardson et al., 2014). Given the importance of the Pacific sand lance to the trophic dynamics of nearshore systems (Beacham, 1986; Borstad et al., 2011) and current concern over ecosystem-level effects of exploiting forage fishes (Smith et al., 2011; Essing- ton et al., 2015), determining the specific habitat and burrowing requirements of the Pacific sand lance are necessary steps toward the development of ecosystem approaches to the management of this species. Like the sandfish [Scincus scincus), a lizard found in sandy habitats in North Africa and Southwestern Asia (Maladen et al., 2009), the Pacific sand lance is able to burrow rapidly into the substrate (Gidmark et al., 2011). It is tempting to suppose that this fish takes advantage of the viscosity and density of water to stir, or fluidize, the sand before burrowing; however, this behavior has not been observed. Instead, high-speed video of Pacific sand lance burrowing in the laboratory indicates that this fish dives headfirst into the sand, beating its tail and driving the head and anterior two- thirds of its body underground. At this point, the bur- ied part of its body undulates and draws the remaining part of the fish beneath the sediment (Gidmark et ah, 2011). Models show that the sandfish uses substan- tial force to burrow into dry sand. It is not possible to extrapolate this type of movement to the Pacific sand lance because there are no data to indicate the relative ease of penetrability through dry sand and sand inun- dated with water. Results from isolated laboratory and field studies, however, indicate that burial preferences range from fine to very coarse sands (Pinto et ah, 1984; Haynes et al., 2007; Robinson et ah, 2013). If these burial preferences can be refined further through more complete testing, this information could 1) be used in habitat-based management plans for Pacific sand lance and 2) may reveal morphological and behavioral adap- tations that explain the preference of Pacific sand lance for a particular grain size or sizes. The Pacific sand lance is an accomplished burrower that spends a large percentage of its time in sediment of unknown characteristics. Because it is an impor- tant forage fish with a strong benthic association, the habitat preferences of this species has direct implica- tions for the development of ecosystem approaches to management and conservation. This fish is also an ex- cellent model for studying locomotion by an elongate, anguilliform burrower in an aquatic environment. The goals of our study were fivefold: 1) to assess the grain sizes, ranging from silt to very fine gravel, that are po- tential burial habitats for Pacific sand lance; 2) to de- termine whether Pacific sand lance prefer substrates of a particular size; 3) to use lighting and sediment com- paction to gain insight into the factors that favor the selection of a burial substrate; 4) to use field sampling and habitat mapping to link sediment preferences of the Pacific sand lance in the laboratory with substrate associations in the field; and 5) to reveal specialized morphological features for burrowing. Materials and methods Data collection Pacific sand lance were collected at Jackson Beach, San Juan Island, Washington, (48.520°N, 123.011°W). Fish Bizzarre et al.: Burrowing behavior, habitat, and functional morphology of Ammodytes personatus 447 Tabie 1 Size range, Wentworth (1922) grade, and phi ((j)) scale of uniform sediment types used in laboratory experiments to determine the preferred habitat of Pacific sand lance [Ammodytes personatus). Experiments were conducted in 2010 and 2012. Size range Wentworth grade Phi scale 2. 0-4.0 mm Very fine gravel -1 to -2 1. 0-2.0 mm Very coarse sand 0 to -1 0. 5-1.0 mm Coarse sand 1 to 0 0.25-0.5 mm Medium sand 2 to 1 125-250 gm Fine sand 3 to 2 62.5—125 gm Very fine sand 4 to 3 3.9-62.5 gm Silt 5 to 4 were captured in a boat-deployed bag seine in subtidal and intertidal waters (depths <10 m) during the sum- mers of 2010 and 2012. Within an hour of capture, indi- vidual fish in good condition were taken to the Univer- sity of Washington’s Friday Harbor Laboratories, where they were maintained in an 1136-L (300-gallon) aquar- ium with running seawater and a mixture of sediments to enable burrowing. The entire size range of sampled Pacific sand lance (5.0-14.5 cm in total length [TL]), that corresponded to juveniles, subadults, and adults (Wyllie-Echeverria^’, was used to determine the bur- rowing capabilities of this species. However, specimens used in sediment preference trials were restricted to individuals corresponding to subadult sizes (8.5-11.0 cm TL) to account for possible scaling effects and be- cause they were the dominant size class in catches. All laboratory experiments were conducted within a month after fish were collected. Approximately 300-500 indi- viduals were maintained throughout the experiments and periodically fed with locally caught mysids and co- pepods. Only fish that appeared to be in good physical condition (i.e., were active, had no obvious abrasions or injuries, and their fins were intact) were used in experimental trials. Marine sediments were collected from local beach- es throughout San Juan Island to obtain a variety of grain sizes. Sediments were dried and sorted into 7 uniform grain sizes ranging from silt (0.4 mm) to very fine gravel (4.0 mm) by using a Ro-Tap sediment analyzer (W. S. Tyler, Mentor, Ohio) (Table 1). For all laboratory trials, sediments were placed in paired alu- minum trays with a volume of 0.014 m^ and a depth of approximately 15 cm that corresponded to the max- imum depth at which Pacific sand lance have been observed locally. A dark, plastic divider measuring 7.5 cm in width separated the trays to minimize an arbi- trary sediment choice. ' Wyllie-Echeverria, T. 2010. Personal commun. Friday Harbor Laboratories, Univ. Wash., Friday Harbor, WA 98250. Laboratory experiments To determine the grain sizes that represent potential burial habitats, 10 fish were placed in a 76-L (20-gal- lon) aquarium with running seawater and a uniform substrate consisting of 1 of the 7 grain sizes (described in Table 1). Experimental fish were netted out of the larger holding tank after the sediment was stirred to mobilize burrowed individuals. Seawater was fed di- rectly through a screen placed over the top of the tank so that water flow did not create bottom currents that could influence burrowing. Fish were not introduced into the tank until the water was no longer turbid. Individuals spanning the entire observed size range of the species were used in each trial, periodically ob- served, and continuously filmed for 6 consecutive day- light hours. The Pacific sand lance typically shelters within an hour of exposure to a suitable sediment type (Pinto et al., 1984). If at least one individual burrowed into sediment of a particular grain size during the trial period, sediment suitability was verified and the grain size was advanced for use in experiments of preferred sediment types. For experiments on sediment preference, 50 indi- viduals were introduced into a 76-L tank with paired sediment types configured as previously indicated. Fish were allowed to acclimate for 4 h, after which covers were placed over the sediment trays, all mobile indi- viduals were removed, and the tank was drained. The number of buried fish in each sediment type was then recorded and all fish in good condition were returned to the holding tank. Daytime trials (n=41) were conducted between 1100 and 1700 h, and illumination was provid- ed simultaneously by sunlight and ambient room light- ing. Nighttime trials (?z=16) occurred between 2300 and 0300 h in complete darkness. Eight replicates were planned for all trials; however, diurnal trials involving medium sand — a grain size that was largely avoided by fish — consisted only of either 5 or 6 replicates because of time and logistical constraints (e.g., condition of cap- tive Pacific sand lance). Preferred sediment was determined by using repli- cated G-tests of goodness of fit from an expected 1:1 ratio (Connallon and Jabukowski, 2009; McDonald, 2009). A G-test of goodness of fit was first conducted for each replicate in a comparison of paired sediments (individual test). To determine whether all of the data from the different experiments fit the expected 1:1 ra- tio, the individual G-values from these replicates were then added to assess the significance of the aggregate G-value (total test). A pooled G-test was then conducted by adding all the observations among experiments and by testing the resulting G-value (pooled test). Finally, a G-test of independence was used to determine whether the individual trials had significantly different relation- ships from one another (heterogeneity test). The totality of these results was synthesized to determine burrowing preference and to explain burrowing behavior. Laboratory experiments were combined with the testing of compaction by using resin models of Pacific 448 Fishery Bulletin 1 14(4) sand lance to determine whether burrowing preference and penetration force of the Pacific sand lance differed in uncompacted and compacted sediments. Full compac- tion was achieved by vibrating a plexiglass plate over one of the paired sediment trays until readings, made with a penetrometer (Forestry Suppliers Inc., Jackson, MS), peaked. In experiments of preferred sediments, compacted and uncompacted sediments consisting of the 4 largest grain sizes were paired, as described in Table 1, and were conducted (only) during daylight hours (Table 1). Models of subadult Pacific sand lance were created by making dental wax (President light body^; Coltene, Altstatten, Switzerland) molds of eutha- nized fish and then filling the molds with Spurr resin. These models were pressed into inundated sediment until a third of the body was covered, corresponding with the penetration stage of burial (Gidmark et al., 2011). Force was calculated with a force gauge (MTS Systems Corp., Eden Prairie, MN) at increments of 1.0 mm, and 5 replicates were conducted for each tested grain size and compaction level. The force necessary to penetrate uncompacted and compacted sediments of different grain sizes was plotted against penetration depth, and the data were log-transformed to achieve linearity. Slopes of the relationship between force and penetration depth were compared by using a 2-way analysis of variance (ANOVA), with compaction and sediment size as in- dependent variables. Where ANOVAs were significant, a Tukey’s honestly significant difference test was run to determine the groups that contributed to these differences. Field collections A comprehensive fish and sediment sampling effort was conducted at the central San Juan Channel sand wave field, a region where bottom currents have shaped the seafloor geomorphology into a series of successive crests and troughs at depths of 60-80 m to determine sediment associations of Pacific sand lance in offshore waters. Twenty-one target sites with a minimum sepa- ration distance of 70 m were chosen randomly through- out the sand wave field, excluding areas near cable crossings (Fig. 1). By using the RV Centennial, fish and sediment were collected with a Van Veen bottom grab, a clamshell-type sampler with long lever arms and sharp cutting edges that enable deep (up to 22 cm) penetration into seafloor sediment. The Van Veen grab has a rapid, powerful closing mechanism, with over- lapping flaps, that allows the jaws to excavate 0.12 m^ of relatively undisturbed sediment while prevent- ing the loss of sediment or fish. This sampling method has been extremely successful for obtaining significant numbers of live sand eel and representative seafloor sediment (Freeman et al., 2004). 2 Mention of trade names or commercial companies is for iden- tification purposes only and does not imply endorsement by the National Marine Fisheries Service, NOAA. Samples were collected at each site during both nighttime and daytime low-tide periods; nighttime sampling was conducted during 2000-2400 on 6 No- vember 2006, and daytime sampling was conducted during 1030-1430 on 7 November 2006. The night- time samples were collected as close to the target sites as possible given drift and current conditions, and the actual vessel coordinates were recorded when the grab hit bottom. These coordinates then became the target sites for the daytime samples. After each sample was retrieved, all fish were removed and fro- zen, and a 400-600-mL subsample of sediment was collected. Fish were later counted, measured, and dissected to determine sex and maturity stage (after Macer, 1966; Nelson and Ross, 1991). Sediment was dried and sorted by using a Ro-Tap sediment analyz- er, as previously described, and the total weight and relative proportion of each grain size was recorded (according to the method of Wentworth, 1922). Data were evaluated for normality and homoscedasticity, and a paired Gtest was conducted to investigate the difference in the mean number of Pacific sand lance collected during day and night grabs that occurred at the same target locations. Field results of habi- tat preference were used for comparison with results from laboratory experiments. Habitat mapping Sediment preferences determined from laboratory ex- periments and field collections were applied to previ- ously constructed seafloor-substrate maps of common fishing grounds for groundfish (especially for rockfish) off southeast Alaska (Fairweather Ground, Cape Om- maney, and Edgecumbe). Substrate maps were created during 1998-2004 by 2 of the authors (H. Greene and senior author) for use in habitat-based fishery man- agement by the Alaska Department of Fish and Game, Sitka office. Source data included a combination of side-scan sonar and multibeam imagery. Map interpre- tations were verified with data from dives of a human- occupied submersible. The mapped depths of the Fairweather Ground and Edgecumbe fishing grounds were within the known depth range of Pacific sand lance (Love et al., 2005). The substrate map for the Cape Ommaney fishing ground, however, extended to a depth of 305 m, well beyond the approximate maximum depth of known occurrence of Pacific sand lance (Love et al., 2005). A 175-m depth contour, therefore, was extracted from the multibeam imagery for this region and used to create a deepwater boundary. All mapping and spatial analysis were conducted in ArcMap, vers. 10.2.2 (Esri, Redlands, CA). Ostrand et al. (2005) determined that depth was the primary factor associated with offshore distribu- tion of Pacific sand lance and that the population in Prince William Sound was largely restricted to depths <60 m. However, the Pacific sand lance is extremely common to depths of at least 80 m off the San Juan Is- lands, and our predictions are of potential habitat dis- Bizzarre et al.: Burrowing behavior, habitat, and functional morphology of Ammodytes personatus 449 123°5'W 123°0'W 122”55'W 122°50’W Figure 1 Location of field effort to sample Pacific sand lance (Ammodytes personatus) in the San Juan Channel, Washington, in 2006. (A) The inset map in the top left corner shows the general location of the study site among the San Juan Islands (black box) at the Pacific border between the United States and Canada (dashed line). The main map highlights the main onshore and offshore features of the region, includ- ing the location of the San Juan Channel and the sand wave field (black box) where field collections were focused. The other (B) inset map provides a multibeam image of the San Juan Channel sand wave field; black dots indicate the sites where Van Veen grabs (n=42) were made. tribution, not fish distribution. Habitat quality, there- fore, was considered to be consistent across the entire known depth range of occurrence of Pacific sand lance. Substrate data were converted into 3 potential habitat categories (preferred, suitable, unsuitable) on the basis of a synthesis of laboratory and field results. Morphology A neotype of A. personatus in the University of Wash- ington Burke Museum collection was scanned with 5.6-pm resolution on a SkyScan 1173 micro-CT scan- ner (Bruker AXS GmbH, Karlsruhe, Germany) at the Karel F. Liem Bioimaging Facility at Friday Harbor Laboratories. Cross-sectional 2-dimensional images (i.e., slice data), generated across the 3-dimensional volume of the fish, were reconstructed and rendered in Amira software (FEI Co., Hillsboro, OR), and the image stack was uploaded to MorphoSource (website) as an open access resource (Godersky and Summers, 2016). A 5-mm-by-lO-mm section of skin from the lateral aspect of the body, just behind the opercular cover, was ex- cised, dehydrated in ETOH, and dried according to the critical point drying method. The skin sample was then sputter-coated with gold-palladium and visualized with a JCM-5000 NeoScope scanning electron microscope (SEM) (JEOL, Ltd., Tokyo). Digital images were made from the combined data of the secondary electron and backscatter electron detectors. To identify structural and morphological characteristics that may contribute to burrowing success, CT- and SEM-generated images were inspected. 450 Fishery Bulletin 1 14(4) Table 2 Results of experiments with paired sediment types in relation to the burrowing activity of Pacific sand lance (Am- modytes personatiis). Experiments were conducted in 2010 and 2012. Data were calculated by using replicated G-tests of goodness of fit. Comparisons were made between medium sand (MS), coarse sand (CS), very coarse sand (VCS), and very fine gravel (VEG). In all comparisons, the grain size that contained the greater number of buried fish is listed first. The number of experimental trials (N) and the number of trials in which dominant grain size had significantly more buried fish (N*) are given. The power of the test for each nonsignificant result is provided in parentheses. Comparison N N* Total Pooled Heterogeneity G P G P G P Diurnal CS-MS 6 5 94.31 <0.001 75.46 <0.001 18.85 0.002 VCS-MS 5 5 166.48 <0.001 152.39 <0.001 14.09 0.007 VFG-MS 6 4 29.28 <0.001 13.94 <0.001 15.34 0.009 CS-VCS 8 5 75.01 <0.001 11.78 <0.001 62.23 <0.001 CS-VFG 8 6 126.73 <0.001 77.54 <0.001 49.19 <0.001 VCS-VFG 8 2 69.33 <0.001 <0.01 0.957 (5) 69.33 <0.001 Nocturnal CS-MS 8 6 79.31 <0.001 73.76 <0.001 5.55 0.593 CS-VCS 8 0 7.98 0.435 (48) 0.22 0.636 (8) 7.76 0.354 (49) Results Laboratory experiments Pacific sand lance were capable of burrowing into all 7 provided sediment types (see Table 1). However, in the smaller grain sizes (silt-fine sand), motility was con- siderably reduced and fish soon reoriented themselves so that their heads were exposed. Large individuals, corresponding to adult sizes, were more likely to bur- row into the larger grain sediments, whereas more small individuals (juveniles) penetrated the smaller grain sediments. During daylight hours. Pacific sand lance preferred coarse sand to all other sediment types (total G-value, pooled G-value; Table 2). Very coarse sand was selected significantly more than medium sand and significantly more than very fine gravel on the basis of the total, but not pooled, G-test. Correspondingly, only 2 individual tests yielded significant results and supported a prefer- ence for very coarse sand over very fine gravel, and the total number of buried individuals was nearly identical (7z=174 and n=173, respectively). A significantly greater number of fish burrowed into very fine gravel than into medium sand (total G-value, pooled G-value; Table 2), which was largely avoided in all trials. Individual G- tests generally were consistent with the results from total and pooled tests; however, heterogeneity G-tests for all paired-preference trials indicated significant variation among individual trials in the observed ratio of burrowed individuals (Table 2). At night. Pacific sand lance maintained a strong preference for coarse over medium sand, but no prefer- ence was found for either coarse sand or very coarse sand (Table 2). The total number of fish buried in coarse (n=147) and very coarse (n=139) sand was sim- ilar at night, and no individual tests revealed a sig- nificant preference for either grain size. The power to detect a significant difference was, however, low for these comparisons (Table 2). For both paired trials, re- sults of the heterogeneity G-test indicated that results of individual experiments were consistent (Table 2). Night experiments were restricted to sediment types that directly bounded the preferred grain size (coarse sand) determined from daytime experiments (Table 1). Therefore, very fine gravel was not included in night experiments. The number of buried individuals among (pooled) paired-sediment preference trials further supports a preference for coarse sand over other sediment types but indicates that results of individual trials are vari- able. The median number of buried individuals and intertrial variability were similar for comparisons of very coarse sand and very fine gravel (daytime) and of coarse sand and very coarse sand (nighttime) (Fig. 2). Correspondingly, results of pooled G-tests were insig- nificant for these comparisons (Table 2). The greatest median number of buried fish was observed for very coarse sand, from trials in which it was paired with medium sand during the day. The most highly vari- able result was also that of very coarse sand (range of nearly 40 individuals), from its paired trials with very fine gravel during daytime hours (Fig. 2). The most pro- nounced grain-size preferences were observed between coarse sand and very fine gravel (daytime), very coarse sand and medium sand (daytime), and coarse sand and medium sand (daytime and nighttime) (Fig. 2). Obvious preferences of grain size, therefore, were evident when Bizzarro et al.: Burrowing behavior, habitat, and functional morphology of Ammodytes personatus 451 ^ Diurnal trial ® Nocturnal trial • Medium sand • Coarse sand ® Very coarse sand • Very fine gravel Box-and-whisker diagrams displaying results of pooled G-tests for each paired sediment preference trial conducted with Pacific sand lance (Ammodytes personatus) in 2010 and 2012. Displayed are the minimum and maximum values (whiskers), first and third quartiles (box), and median values (bold horizontal line). all trials were considered for a comparison, but prefer- ences were not generally consistent among trials. The average proportion of Pacific sand lance bur- ied among pooled experimental trials was greatest for coarse sand (Fig. 3). Among diurnal trials, the propor- tion of individuals buried in coarse sand was nearly 20% greater than expected by chance. Very coarse sand and very fine gravel were selected at ratios that were very close to the expected 1:1, whereas medium sand was selected in only about 20% of pooled trials (Fig. 3). Coarse sand also had the greatest average proportion of individuals buried among nocturnal trials, but the preference for this grain size was somewhat reduced compared with results from diurnal trials (Fig. 3). A similar situation was observed for very coarse sand, whereas the opposite trend was observed for medium sand (Fig. 3). Variability in the average proportion of buried sand lance among experimental trials was pro- nounced during diurnal trials but reduced in relation to medium sand and very coarse sand at night (Fig. 3). Compacted sediments of all grain sizes required significantly more force to penetrate than did uncom- pacted sediments, as indicated by values for the mean slope of penetration depth (measured in millimeters) in relation to log force (measured in Newtons): compact- ed=0.071 (standard deviation [SD] 0.015), uncompact- ed=0.061 (SD 0.014) (Table 3). However, the mean per- centage of buried fish in uncompacted (20.8% [SD 4.8]) and compacted (21.5% [SD 5.2]) coarse sand was simi- lar. Correspondingly, no significant differences were de- tected in the selectivity for these substrates (total test: G=8.19, P=0.22; pooled test: G=8.12, F=0.80; heteroge- neity test: G=0.06, P=0.15, N=6). The power to detect a difference in selectivity was low, however, especially for the heterogeneity test (total=58, pooled=57, hetero- geneity=5). In comparisons of burrowing force among grain sizes, smaller grain sizes generally required more force to be penetrated regardless of compaction level (mean slope: very fine gravel=0.052 [SD 0.007], very coarse sand=0.059 [SD 0.001], coarse sand=0.075 [SD 0.008], medium sand=0.078 [SD 0.017]; Table 3). The interaction between sediment size and compaction was not significant (P=0.17). Field collections Field collections in the San Juan Channel yielded 421 Pacific sand lance during 21 daytime and 21 nighttime samples collected with a Van Veen bottom grab. Fish abundance, and therefore density, varied considerably among grab samples collected during both time peri- ods. During the day. Pacific sand lance were caught in every grab, with a range of 2-62 individuals/grab and a median value of 13 individuals/grab (quartile 1 [Ql]=6, Q3=18). By contrast, Pacific sand lance were caught less frequently (66.7% of grabs, n=14) and in lower abundance (range=0-26 individuals, median=3, Q1=0, Q3=ll) at night. Between paired samples collect- ed at the same locations, a significantly greater mean number of individuals was collected during the day than during the night (day=14.4 individuals [SD 13.2]; night=5.7 individuals [SD 7.0]; t=2.47, df=20, P=0.023). Consequently, the density of buried fish among these grabs also was much greater during the daytime (119.8 individuals/m^ [SD 110.0]) than during nighttime (47.7 individuals/m^ [SD 58.7]). The deepwater population of Pacific sand lance in San Juan Channel and of subtidal individuals used in laboratory experiments were consistent in size and maturity stage. The mean size of fish captured at San Juan Channel was 8.5 cm TL (SD 0.6) (range=6.5-12.3 cm TL, n=415), and 92.7% of all sampled fish (n=356) were determined to be immature on the basis of in- ternal inspection. Mean sizes and maturity stage, the determination of which requires sacrificing individu- als, were not recorded for laboratory specimens col- 452 Fishery Bulletin 114(4) Figure 3 Average proportion of Pacific sand lance {Ammodytes per- sonatiis) buried in each grain size (medium sand [MS], coarse sand [CS], very coarse sand [VCS], and very fine gravel [VFG]) in relation to all grain sizes for (A) diur- nal and (B) nocturnal trials conducted in 2010 and 2012. Horizontal bars depict the average deviation from the expected 1:1 ratio, and the zero line, therefore, is equiva- lent to a proportion of 0.50. The number of replicates for each grain size is indicated in parentheses. Error bars represent first and third quartiles. lected from intertidal and subtidal depths at Jackson Beach; however, the size range (8.5-11.0 cm TL) that was used was skewed toward smaller specimens. This size range was chosen to correspond to large subadult specimens — a supposition that was validated through field sampling. The sediment composition of the sampled sand wave field consisted primarily of coarse sand (51.4%) and similar proportions of very coarse sand (14.5%), gravel (16.2%), and medium sand (17.7%) (n=42 grab samples). Very coarse sand and medium sand never contributed >33% to sample weight, but 8 samples were composed of at least one-third gravel (Fig. 4). Fine sand (0.2%) and silt (0.1%) contributed trivial amounts to overall sample weight. No grain sizes larger than that of grav- el were encountered. Linear regression indicated no re- lationship between fish abundance and the proportions Table 3 Mean slope of the relationship between log force (mea- sured in Newtons) and penetration depth (measured in millimeters) determined from models of different sediment types used by Pacific sand lance {Ammodytes personatus), including uncompacted (U), compacted (C), very fine gravel (VFG), very coarse sand (VCS), coarse sand (CS), and medium sand (MS). The number of ex- perimental trials (N) and standard deviation of the mean (SD) are given. Sediment N Slope SD U 20 0.061 0.014 C 20 0.071 0.015 VFG 10 0.052 0.007 VCS 10 0.059 0.001 CS 10 0.075 0.008 MS 10 0.078 0.017 of medium sand (t=0.718, P=0.477, coefficient of corre- lation [r2]=0.01, df=40), coarse sand (t=0.365, P=0.717, r^=0.02, df=40), very coarse sand {t= -0.079, P=0.717, r2=0.02, df=40), or gravel it= -0.553, P=0.583, r2=0.01, df=40) (Fig. 4). However, limited variability in sample composition may have influenced results (Fig. 4). Habitat mapping The relative amount of preferred habitat for Pacific sand lance (i.e., sand; coarse sand, very coarse sand, and gravel; and sand wave fields), varied among 3 important commercial fishing grounds of southeast Alaska (Fig. 5). Preferred habitat types were predict- ed to be extremely rare at Cape Ommaney (1.9% of the total mapped seafloor, 168.9 km^) and uncommon at Fairweather Ground (13.8% of the total mapped seafloor, 288.0 km^) but common at Edgecumbe (36.1% of the total mapped seafloor, 538.2 km^). The total amount of potential habitat, consisting of habi- tat types that were preferred and suitable (i.e., sand mixed with mud, pebble, boulder, or rock outcrops) was greatest at Fairweather Ground (66.9%) and similar at Cape Ommaney (39.4%) and Edgecumbe (39.0%). Nearly all unsuitable habitats (i.e., grain sizes <3.9 pm [silt] or >4.0 mm [pebble]) in all re- gions were composed of rocky substrates or sediment of large grain sizes (Edgecumbe: 100% of 328.4 km^; Cape Ommaney: 95.7% of 102.3 km^; and Fairweather Ground: 100% of 95.4 km^). Description of morphological features Images of the cranial and axial skeleton were produced with CT scans of the Pacific sand lance (Fig. 6, A-C). The dermatocranial elements are thin and lightly min- eralized. The parasphenoid and occipital bones are well Bizzarro et al.: Burrowing behavior, habitat, and functional morphology of Ammodytes personatus 453 B 0 -rO 0-r ,Q* ^ Coarse sand (%) -40 10% silt or clay, and variation in its abundance has been negatively associated with the amount of sedi- mentary silt (Wright et ah, 2000). Grain sizes smaller than medium sand generally are too poorly oxygenated to enable protracted burial by the Pacific sand lances and other Ammodytes species; therefore, respiratory tolerances influence sediment choice and limit the re- alized burial habitat of species of Ammodytes. An upper boundary of grain size for burial of Pacific sand lance is more difficult to establish but appears to fall within the size range of gravel. A positive associa- tion between length and grain size was reported for the sand eel (Holland et al., 2005) and is supported by the results of our experiments on burial capability of Pacific sand lance. Some possible explanations for this asso- ciation are that larger subadult and adult specimens, compared with smaller juvenile specimens, 1) can more easily generate the force necessary to penetrate grav- els, 2) have greater respiration demands and require more well-drained sediments, and 3) are less susceptible to abrasion. The largest grain size used in this study. 456 Fishery Bulletin 114(4) very fine gravel, was not commonly selected; however, the Pacific sand lance is able to use uniform very fine gravel (this study) and a mixture of very fine-to-medium gravel for long-term burial (Pinto et ah, 1984). The force required or damage incurred by burial in increasingly coarse sediment will outweigh the benefit of increased water flow at some tolerance point, above which larger grain sizes will be avoided. Coarse sand with pebbles and fine-medium gravel have been reported as buri- al habitat for Ammodytes species (Robards and Piatt, 1999; Holland et ah, 2005), but sediments dominated by coarse gravel, pebble, or cobble may represent such a boundary. More research is needed to determine the largest grain size or sizes of sediment that Pacific sand lance can use as burial habitat and how this tolerance varies with life stage and oxygen level. Although clear sediment preferences were estab- lished, there was substantial variability in results among experimental trials that appears to be related to behavior. Even in trials where grain preference was marked, there was considerable inter-trial variability in the total number of fish that buried themselves. Additionally, the relative proportion of buried fish was highly variable among trials involving increasing grain sizes from coarse sand-very fine gravel. Inter- trial variability was greatest between coarse sand and very coarse sand at night, possibly because it is more difficult for Pacific sand lance to distinguish between similar sediment types without visual cues. As reported by Pinto et al. (1984), Pacific sand lance did not burrow en masse. Instead, individuals buried themselves at different times and sometimes switched sediment types during a trial. Variability in burial preference among trials therefore seems to be related to the individual differences in behavior. This variability appears to be enhanced when sediment types of comparable prefer- ence are available. Results of day and night experiments and field col- lections generally ran counter to the established diur- nal burial pattern of Ammodytes species. Ammodytes species seek refuge in the seafloor at night to avoid predators, and they are typically active during the day (Robards and Piatt, 1999). However, when food is scarce (Winslade, 1974) or predators are active (Hobson, 1986), species of Ammodtyes also will burrow during the day- time. No food was provided to Pacific sand lance during trials, and the tank enclosure did not represent typical daytime feeding habitat, which is open water. A high proportion of Pacific sand lance may, therefore, have buried themselves during daylight hours to conserve energy or to reduce stress. Field collections indicated that substantially more Pacific sand lance were buried during the day than at night. This result may be an artifact of limited temporal and spatial sampling. Dis- tribution of Pacific sand lance is known to be patchy (van der Kooij et al., 2008), and perhaps too few samples were collected to capture this spatial variability. There also may have been additional considerations, such as current flux or predator dynamics that influence the relative number of day and night burials. It also is possible that the diurnal pattern of habitat use that has been established for Ammodytes species is more com- plex or not directly attributable to Pacific sand lance. More research is needed to determine day and night use of deepwater habitats by Pacific sand lance. Because of the importance of Pacific sand lance to regional trophic dynamics and the potential for anthro- pogenic disturbance of important burial grounds, de- termining and conserving preferred habitats of Pacific sand lance may be a future management consideration. There is no active fishery for Pacific sand lance in U.S. waters, and the main drivers of abundance of Ammod- tyes species are environmental conditions and density dependence during early life stages (Arnott et al., 2002). However, fishing activities (especially bottom trawling) can change substrate composition, primarily by decreas- ing complexity and fluidizing sediment, generally lead- ing to a proliferation of smaller grains sizes (e.g., muds instead of sands) (Auster et al., 1996). Other anthropo- genic disturbance (e.g., dredging and sediment mining) also may alter sediment composition, reduce the amount of preferred sediment types, or cause direct mortality to populations of Pacific sand lance (Eleftheriou and Robertson, 1992). The field techniques described in our study can be used with data that have been previously collected, newly accumulated from seafloor imagery, or bottom- sampled in areas of interest to create and evaluate maps of potential habitat. Crucial habitats of Pacific sand lance can then be identified and conserved through establishment of marine protected areas or other no- take zones. This type of applied research is currently being conducted on populations of Pacific sand lance in British Columbia (Robinson et al., 2013) and has impor- tant considerations for American populations of Pacific sand lance off Washington and in the Gulf of Alaska. Different geological characteristics explained the variable amount of potential preferred and suitable habitat among fishing grounds in southeast Alaska. Preferred habitat for Pacific sand lance was predict- ed to be most abundant at Edgecumbe, where sand surrounds a recent lava flow in deep water (>120 m; Greene et al., 2007a). Small pockets of sand and gravel, including sand waves, occur on Fairweather Ground, as do a majority of sandy sediments (Greene et al., 2007b). These sands and gravels have been eroded from the ex- tensive sandstone on the Fairweather Ground, and they account for the greatest amount of predicted preferred and suitable habitat among locations. Some of the preferred habitat of Pacific sand lance at Fairweather Ground is associated with a sand wave field that has been formed by an underlying syncline (Greene et al., 2007b). Dense aggregations of Pacific sand lance have been observed on this feature (Greene^), lending sup- port to our predictions of preferred habitat. Much of the seafloor at Cape Ommaney consists of either rock, a mixture of granite and sand, or large, unconsolidated ^ Greene, H. G. 2010. Personal commun. Friday Harbor Laboratories, Univ. Wash., Friday Harbor, WA 98250. Bizzarro et al.: Burrowing behavior, habitat, and functional morphology of Ammodytes personatus 457 grain sizes (gravel-cobble). Therefore, preferred habitat was predicted to be scarce at Cape Ommaney. Overall, preferred habitat of Pacific sand lance was limited on the fishing grounds we examined off southeast Alaska. This result is not surprising because rock-associated species (e.g., tiger rockfish [Sebastes nigrocinctus]; yelloweye rockfish [Sebastes ruberrimus]; lingcod [Ophiodon elongates]) are the main targets of regional fisheries (Greene et al., 2011). However, Robin- son et al. (2013), using a habitat suitability model that incorporated several physical variables, indicated that only 6% of the study region in the Strait of Georgia was suitable habitat for Pacific sand lance. Our predic- tions, therefore, probably overestimate the amount of preferred and suitable habitat types on fishing grounds of southeast Alaska. The amount of preferred and suit- able burial grounds would probably be reduced if more factors, such as depth and current speed, were consid- ered in addition to grain size. Our day and night experiments indicated that the Pacific sand lance does not rely exclusively on visual cues to assess habitat quality and that this species chooses sediments that are more difficult to penetrate in order to access preferred grain sizes. Some visual assessment of the substrate is supported by the equiva- lence of coarse and very coarse sand in night trials, but we have no explanation for the decrease in the variability in choice of sediment at night. Fish were often observed “nosing” the sediment before burrowing, and they may use tactile cues to determine grain size. We suppose that coarse sand provides a desired com- bination of ease of penetration and interstitial space that represents the best means for reducing the ener- getic costs of burrowing while maintaining access to oxygenated waters. This supposition was supported by our anecdotal observation that in medium sand and smaller grain sizes, fish often reoriented themselves after burrowing so that their entire head was exposed. Larger grain sizes may be too difficult to penetrate for subadult Pacific sand lance or may abrade the skin. Burrowing in terrestrial environments usually is associated with elongation of the body, reduced eyes and limbs, and well-ossified skulls with fused elements (i.e.. Summers and O’Reilly, 1997; Lee, 1998). Although the Pacific sand lance does have an elongate body, a small cross section, and fins that lie flat, it otherwise has little in common with typical burrowers. The Pa- cific sand lance has prominent, large eyes, its skull is complex, and no mineralization is evident in several areas, including the dorsal region caudal to the max- illa. The lower jaw is well mineralized, but no more so than in fish species that do not burrow. Nevertheless, individuals of this species are able to bury themselves rapidly and in a wide variety of sediments. They can burrow for several body lengths and spend consider- able periods of their lives under the substrate. There is evidence that they do not fluidize the sand in advance of penetration — a behavior that would have explained what appears to be poorly adapted morphological fea- tures for burrowing (Gidmark et al., 2011). We are left to remark upon the sole morphological oddity for which we can imagine a burrowing func- tion. The scales of the Pacific sand lance, and those of all species of Ammodytes, are unusual with respect to other fishes in that they are fused into bands that run roughly dorsoventrally. These bands are of taxonomic interest, although no function has ever been ascribed to them (Orr et al., 2015). We propose that the spacing of the scale bands interacts with the grain size of the substrate to produce a reduced frictional interaction between the fish and the sand. Indeed, movement was observed to be considerably more rapid for subadult Pacific sand lance in saturated coarse sand than in oth- er grain sizes. Our supposition should be testable with nanoscale accurate replicas of the Pacific sand lance and may be modeled with contact mechanics in an un- usual low-force regime (Persson and Scaraggi, 2014). Because of the variation in scale spacing on the body in different Ammodytes species and the change in spacing as an individual grows to maturity, the implication of a friction reducing system is that habitat preference of sand lances should vary with spacing on inter- and intraspecific bases. Other habitat characteristics (e.g., depth, current velocity, oxygen content, particle sorting) are impor- tant, but none is a more fundamental driver of dis- tribution and abundance of Pacific sand lance than grain size. Considerable laboratory and field evidence indicates that coarse sand sediments and mixtures of sediment that are dominated by coarse sand are pre- ferred burial habitats for Pacific sand lance in the sub- adult and adult life stages from intertidal depths to at least 80 m. Use of deepwater habitats, especially sand wave fields, are poorly understood but should be further evaluated because seafloor structure, current regimes, and other physical factors play an unknown role in habitat selection of Pacific sand lance. Determi- nation of sediment preferences of the Pacific sand lance will be important for potential habitat-based manage- ment of this important forage species because trawling, dredging, and other anthropogenic disturbances may modify its preferred habitat. A link between form and function has been established for Pacific sand lance, in that this species appears to have a scale pattern that is adapted for rapid burial in coarse sand. Future research should focus on investigating the spacing of scale bands on the body for different sizes and species of Ammodytes and the role of such spacing in reducing friction during burial. Acknowledgments We thank T. Wyllie-Echeverria, P. Bourdillon, and the students of the 2010 and 2012 Functional Ecology and Morphology of Marine Fishes courses at Friday Har- bor Laboratories for their assistance with this proj- ect. We also thank T. O’Connell, C. Brylinsky, and K. Green of the Alaska Department of Fish and Game for their prior collaborations with H. Greene and the 458 Fishery Bulletin 1 14(4) senior author that generated the substrate maps that we used to predict habitat quality of Pacific sand lance. A. Stone provided inspiration for the success- ful completion of this project. The North Pacific Re- search Board (Project 1207), National Science Founda- tion (IOS-1256602), Friday Harbor Laboratories, and Northwest Straits Commission provided financial sup- port for our research. Literature cited Arnott, S. A., G. D. Ruxton, and E. S. Poloczanska. 2002. Stochastic dynamic population model of North Sea sandeels, and its application to precautionary manage- ment procedures. Mar. Ecol. Prog. Ser. 235:223-234. Auster, P. J., R. J. Malatesta, R. W. Langton, L. Watling, P. C. Valentine, C. L. S. Donaldson, E. W. Langton, A. N. Shepard, and W. G. Babb. 1996. The impacts of mobile fishing gear on seafioor habi- tats in the Gulf of Maine (Northwest Atlantic): implica- tions for conservation of fish populations. Rev. Fish. Sci. 4:185-202. Barrie, J. V., K. W. Conway, K. Picard, and H. G. Greene. 2009. Large-scale sedimentary bedforms and sediment dynamics on a glaciated tectonic continental shelf: ex- amples from the Pacific margin of Canada. Cont. Shelf Res. 29:796-806. Beacham, T. D. 1986. Type, quantity, and size of food of Pacific salmon (Oncorhynchus) in the Strait of Juan de-Fuca, British Columbia, Canada. Fish. Bull. 84: 77-90. Behrens, J. W., H. J. Stahl, J. F. Steffensen, and R. N. Glud. 2007. Oxygen dynamics around buried lesser sandeels Am- modytes tobianus (Linneaus 1785): mode of ventilation and oxygen requirements. J. Exp. Biol. 210:1006-1014. Bertram, D. F., and G. W. Kaiser. 1993. Rhinoceros auklet (Cerorhinca moncerata) nestling diet may gauge Pacific sand lance (Ammodytes hexapter- us) recruitment. Can. J. Fish. Aquat. Sci. 50:1908-1915. Borstad, G., W. Crawford, J. M. Hipfner, R. Thompson, and K. Hyatt. 2011. Environmental control of breeding success of rhinoc- eros auklets at Triangle Island, British Columbia. Mar. Ecol. Prog. Ser. 424:285-302. Connallon, T, and E. Jakubowski. 2009. Association between sex ratio distortion and sexu- ally antagonistic fitness consequences of female choice. Evolution 63:2179-2183. Curry, P. M., 1. L. Boyd, S. Bonhommeau, T. Anker-Nilssen, R. J. M. Crawford, R. W. Furness, J. A. Mills, E. J. Murphy, H. Osterblom, M. Paleczny et al. 2011. Global seabird response to forage fish depletion — one third for the birds. Science 334:1703-1706. de Beer, D., F. Wenzhdfer, T. G. Ferdelman, S. E. Boehme, M. Huettel, J. E. E. van Beusekom, M. E. Bottcher, N. Musat, and N. Dubilier. 2005. Transport and mineralization rates in North Sea sandy intertidal sediments, Sylt-Rpmp Basin, Wadden Sea. Limnol. Oceanogr. 50:113-127. Eleftheriou, A., and M. R. Robertson. 1992. The effects of experimental scallop dredging on the fauna and physical environment of a shallow sandy com- munity. Netherlands J. Sea Res. 30:289-299. Essington, T. E., P. E. Moriarity, H. E. Froehlich, E. E. Hodg- son, L. E. Koehn, K. L. Oken, M. C. Siple, and C. C. Stawitz. 2015. Fishing amplifies forage fish population collaps- es. Proc. Natl. Acad. Sci. U.S.A. 112:6648-6652. Frederiksen, M., H. Jensen, F. Daunt, R. A. Mavor, and S. Wanless. 2008. Differential effects of a local industrial sand lance fishery on seabird breeding performance. Ecol. Appl. 18:701-710. Freeman, S., S. Mackinson, and R. Flatt. 2004. Diel patterns in the habitat utilisation of sandeels revealed using integrated acoustic surveys. J. Exp. Mar. Biol. Ecol. 305:141-154. Friedlaender, A. S., E. L. Hazen, D. P. Nowacek, P. N. Halpin, C. Ware, M. T. Weinrich, T. Hurst, and D. Wiley. 2009. Diel changes in humpback whale Megaptera novae- angliae feeding behavior in response to sand lance Am- modytes spp. behavior and distribution. Mar. Ecol. Prog. Ser. 395:91-100. Gaichas, S. K., K. Y. Aydin, and R. C. Francis. 2010. Using food web model results to inform stock as- sessment estimates of mortality and production for ecosystem-based fisheries management. Can. J. Fish. Aquat. Sci. 67:1490-1506. Gidmark, N. J., J. A. Strother, J. M. Horton, A. P. Summers, and E. L. Brainerd. 2011. Locomotory transition from water to sand and its effects on undulatory kinematics in sand lances (Ammo- dytidae). J. Exp. Biol. 214:657-664. Godersky, A., and A. P. Summers. 2016. Ammodytes personatus neotype slice data. Greene, H. G., V. M. O’Connell, W. W. Wakefield, and C. K. Brylinsky. 2007a. The offshore Edgecumbe lava field, southeast Alas- ka: geologic and habitat characterization of a commer- cial fishing ground. In Mapping the seafioor for habitat characterization (B. J. Todd and H. G. Greene, eds.), p. 251-270. Geol. Assoc. Canada, St. Johns, Canada. Greene, H. G., J. J. Bizzarro, V. M. O’Connell, and C. K. Brylinsky. 2007b. Construction of digital potential marine benthic habitat maps using a coded classification scheme and its application. In Mapping the seafloor for habitat char- acterization (B. J. Todd and H. G. Greene, eds.), p. 145- 160. Geol. Assoc. Canada, St. Johns, Canada. Greene, H. G., V. M. O’Connell, and C. K. Brylinsky. 2011. Tectonic and glacial related seafloor geomorphology as possible demersal shelf rockfish habitat surrogates — examples along the Alaskan convergent transform boundary. Cont. Shelf Res. 31:S39-S53. Han, Z., T. Yanagimoto, Y. Zhang, and T. Gao. 2012. Phylogeography study of Ammodytes personatus in northwestern Pacific: Pleistocene isolation, temperature and current conducted secondary contact. PLoS ONE 7(5): 637425. Haynes, T. B., R. A. Ronconi, and A. E. Burger. 2007. Habitat use and behavior of the Pacific sand lance (Ammodytes hexapterus) in the shallow subtidal region of southwestern Vancouver Island. Northwest. Nat. 88:155-167. Hobson, E. S. 1986. Predation on the Pacific sand lance, Ammodytes Bizzarro et al.: Burrowing behavior, habitat, and functional morphology of Ammodytes personatus 459 hexapterus (Pisces: Ammodytidae), during the transition between day and night in southeastern Alaska. Copeia 1:223-226. Hpines, A. S., and O. A. Bergstad. 2001. Density of wintering sand eel in the sand recorded by grab catches. Fish. Res. 49:295-301. Holland, G. J., S. P. R. Greenstreet, I. M. Gibb, H. M. Fraser, and M. R. Robertson. 2005. Identifying sandeel Ammodytes marinas sediment habitat preferences in the marine environment. Mar. Ecol. Prog. Ser. 303:269-282. Johnson, S. W., J. F. Tehdinga, and K. M. Munk. 2008. Distribution and use of shallow-water habitats by Pacific sand lances in Southeastern Alaska. Trans. Am. Fish. Soc. 137:1455-1463. Kleinteich, T., H. C. Maddin, J. Herzen, F. Beckmann, and A. P. Summers. 2012. Is solid always best? Cranial performance in solid and fenestrated caecilian skulls. J. Exp. Biol. 215:833- 844. Lee, M. S. Y. 1998. Convergent evolution and character correlation in burrowing reptiles: towards a resolution of squamate re- lationships. Biol. J. Linn. Soc. 65:369-453. Love, M. S., C. W. Mecklenburg, T. A. Mecklenburg, and L.K. Thors teinson. 2005. Resource inventory of marine and estuarine fishes of the West Coast and Alaska: a checklist of North Pa- cific and Arctic Ocean species from Baja California to the Alaska-Yukon border, 276 p. U.S. Geological Survey, Seattle. [Available at website.] Macer, C. T. 1966. Sand eels (Ammodytidae) in the south-western North Sea: their biology and fishery. Great Britain Minist. Agric. Fish. Food, Fish. Invest. Ser, H Mar. Fish. 24, 55 p. Her Majesty’s Stationary Office, London. Maladen, R. D., Y. Ding, C. Li, and D. 1. Goldman. 2009. Undulatory swimming in sand: subsurface locomo- tion of the sandfish lizard. Science 325:314-318. Maladen, R. D., Y. Ding, P. B. Umbanhowar, A. Kamor, and D. 1. Goldman. 2011. Mechanical models of sandfish locomotion reveal principles of high performance subsurface sand-swim- ming. J. R. Soc. Interface 8:1332-1345. McDonald, J. H. 2009. Handbook of biological statistics, 2nd ed., 319 p. Sparky House Publishing, Baltimor, MD. Mecklenburg, C. W., P. R. Moller, and D. Steinke. 2011. Biodiversity of arctic marine fishes: taxonomy and zoogeography. Mar. Biodivers. 41:19-140. Mushinsky, H. R., and C. Cans. 1992. The role of the tail in channel passage by the sand skink, Neoseps reynoldsi. Amphibia-Reptilia 13:393- 403. Nelson, G. A., and M. R. Ross. 1991. Biology and population change of northern sand lance {Ammodytes dubius) from the Gulf of Maine to Middle Atlantic Bight. J. Northwest Atl. Fish. Sci. 11:11-28. Orr, J. W., S. Wildes, Y. Kai, N. Raring, T. Nakabo, O. Katugin, and J. Guyon. 2015. Systematics of North Pacific sand lances of the ge- nus Ammodytes based on molecular and morphometric evidence, with the description of a new species from Ja- pan. Fish. Bull. 113:126-156. Ohshima, Y. 1950. Biological notes on the sand-eel Ammodytes per- sonatus Girard. Nippon Suisan Gakkaishi 16:99-107 Ostrand, W. D., T. A. Gotthardt, S. Howlin, and M. D. Robards. 2005. Habitat selection models for Pacific sand lance {Ammodytes hexapterus) in Prince William Sound, Alas- ka. Northwestern Naturalist 86:31-143. Parks, D., A. Shaffer, and D. Barry. 2013. Nearshore drift-cell sediment processes and ecolog- ical function for forage fish: implications for ecological restoration of impaired Pacific Northwest marine ecosys- tems. J. Coast. Res. 29:984-997. Persson, B. N. J., and M. Scaraggi. 2014. Theory of adhesion: role of surface roughness. J. Chem. Phys. 141:124701. Pinto, J. M., W. H. Pearson, and J. W. Anderson. 1984. Sediment preferences and oil contamination in the Pa- cific sand lance Ammodytes hexapterus. Mar. Biol. 83:193- 204. Richardson, D. E., M. C. Palmer, and B. E. Smith. 2014. The influence of forage fish abundance on the aggre- gation of Gulf of Maine Atlantic cod {Gadus morhua) and their catchability in the fishery. Can. J. Fish. Aquat. Sci. 71:1349-1362. Robards, M. D., and J. F. Piatt. 1999. Biology of the genus Ammodytes, the sand lanc- es. In Sand lance; a review of biology and predator relations and annotated bibliography (M. D. Robards, M. F. Wilson, R. H. Armstrong, and J. F. Piatt, eds.), p. 1-16. U.S. Dep. Agric., Forest Serv., Seattle, WA. Robinson, C. L. K., D. Hrynyk, J. V. Barrie, and J. Schweigert. 2013. Identifying subtidal burying habitat of Pacific sand lance {Ammodytes hexpaterus) in the Strait of Georgia, British Columbia, Canada. Prog. Oceanogr. 115:119- 128. Rose, J. A., M. Sandefur, S. Huskey, J. L. Dernier, and M. T. Butcher. 2013. Muscle architecture and out-force potential of the thoracic limb in the eastern mole {Scalopus aquati- cus). J. Morphol. 274:1277-1287. Smith, A. D. M., C. J. Brown, C. M. Bulman, E. A. Fulton, P. Johnson, 1. C. Kaplan, H. Lozano-Montes, S. Makinson, M. Marzloff, L. J. Shannon, et al. 2011. Impacts of fishing low-trophic level species on ma- rine ecosystems. Science 333:1147-1150. Summers, A. P, and J .C. O’Reilly. 1997. A comparative study of locomotion in the caecilians Dermophis mexicanus and Typhlonectes natans (Amphib- ia: Gymnophiona). Zool. J. Linn. Soc. 121:65-76. Turanov, S. V., and Y. P Kartavtsev. 2014. The taxonomic composition and distribution of sand lances from the genus Ammodytes (Perciformes; Ammodytidae) in the North Pacific. Russ. J. Mar. Biol. 40:447-454. van der Kooij, J., B. E. Scott, and S. Mackinson. 2008. The effects of environmental factors on daytime sandeel distribution and abundance on Dogger Bank. J. Sea Res. 60:201-209. Weinrich, M., M. Martin, R. Griffiths, J. Bove, and M. Schilling. 1997. A shift in distribution of humpback whales, Mega- ptera novaengliae, in response to prey in the southern Gulf of Maine. Fish. Bull. 95:826-836. Wentworth, C. K. 1922. A scale of grade and class terms for clastic sedi- ments. J. Geol. 30:377-392. 460 Fishery Bulletin 114(4) Winslade, P. 1974. Behavioural studies on the lesser sandeel Ammo- dytes marinus (Raitt). II. The effect of light intensity on activity. J. Fish Biol. 6:577-586. Wright, P. J., H. Jensen, and I. Tuck. 2000. The influence of sediment type on the distribution of the lesser sandeel, Ammodytes marinus. J. Sea Res. 44:243-256. 461 NOAA National Marine Fisheries Service Fishery Bulletin ftr established 1881 ■3 m near the marina. The substrate was mud in the creek and in deeper areas (>2 m) and was sand (with the presence of oysters) at the creek mouth and near shore. Genetic analysis Ten juvenile smalltooth sawfish were caught at the capture location in the Caloosahatchee River between April and August 2006. The individuals were all <1500 mm STL, indicating that they were young-of-the-year (Simpfendorfer et ah, 2008b; Scharer et ah, 2012). Fin clippings were stored in 95% reagent-grade ethanol at room temperature. Genomic DNA was extracted by us- ing a PureGene DNA extraction kit (Qiagen Inc., Va- lencia, CA). Six polymorphic microsatellite loci labeled with fluorescent markers (Feldheim et ah, 2010; Table 1) were amplified with 2 multiplexed polymerase chain reactions (PCRs). Approximately 100 ng of template DNA was used in a 12.5-pL PGR containing 0.3125 U GoTaq DNA Polymerase (Promega Corp., Madison, WI), 0.8 pM of each primer, 200 pM of each dNTP, 2.5 mM MgCl2, 0.5x bovine serum albumin (BSA), and lx col- orless GoTaq buffer. The PGR profile consisted of an initial denaturation at 94°C for 2 min, followed by 30 cycles at 94°C for 30 s, 53°C for 30 s, 72°C for 30 s, and a final extension step at 72°C for 5 min. The resulting fragments were analyzed on an au- tomated Applied Biosystems 3130 Genetic Analyzer Poulakis et al.: Long-term site fidelity of Pristis pectinata 465 Table 1 Microsatellite loci, size of loci (in base pairs), number of alleles {K, Kq), and cor- responding reference for smailtooth sawfish (Pristis pectinata) caught in 2006 in the Caloosahatchee River, Florida. K is the number of alleles per locus for all samples (n=10); Kt is the number of alleles per locus for the smailtooth sawfish containing the mitochondrial DNA control region haplotype T (n=7); Kq is the number of alleles per locus for the smailtooth sawfish containing the mitochondrial DNA control region haplotype C (n-3). Locus Size K Kt Kc Reference Ppe69 153-173 4 3 2 K. Feldheim, unpubl. data Ppell4 181-285 11 7 5 Feldheim et al., 2010 Ppel22 236-264 6 6 3 Feldheim et al., 2010 Ppel60 164-174 3 2 2 Feldheim et al., 2010 Ppel65 251-299 8 7 4 Feldheim et al., 2010 Ppel68 219-257 8 4 6 Feldheim et al., 2010 (Thermo Fisher Scientific Inc., Waltham, MA) with Ge- neScan ROX 500 Size Standard (Thermo Fisher Scien- tific Inc.). Fragment lengths were analyzed by using GeneMapper software, vers. 4.0 (Applied Biosystems, Thermo Fisher Scientific Inc.), and modified by hand when necessary. Loci were checked for Hardy- Weinberg equilibrium by using GENEPOP (Raymond and Rous- set, 1995; Rousset, 2008). The prevalence of null alleles, large allele dropout, and scoring errors was estimated by using MICRO-CHECKER, vers. 2.2.3 (van Ooster- hout et al., 2004). To address our hypotheses, we used a combination of mitochondrial and microsatellite data to conservatively estimate the number of mothers that contributed to the group of smailtooth sawfish. Microsatellite data were analyzed separately within each of the mitochondrial lineages that were identified. For mitochondrial-DNA sequencing analysis, we am- plified a portion of the control region that had approxi- mately 750 base pairs (bp), using the forward primer ProL2 (CTGCCCTTGGCTCCCAAAGC) (Pardini et al., 2001) and reverse primer CRR (atgcaaatattatgtcgagggtag) (Phil- lips et al., 2011). Approximately 100 ng of DNA was used as template in a 50-pL PCR reaction containing 1.25 U GoTaq DNA Polymerase, 0.8 pM of each primer, 200 pM of each dNTP, 2.5 mM MgCl2, 0.5x BSA, and lx colorless GoTaq buffer. The PCR profile consisted of an initial denaturation at 94°C for 2 min, followed by 30 cycles at 94°C for 45 s, 52°C for 45 s, 72°C for 45 s, and a final extension step at 72°C for 5 min. After am- plification, the PCR reaction was run for approximately 45 min at 90 amps on a 1.5% TAE agarose gel with ethidium bromide. PCR products were visualized, ex- tracted from the gel, purified by using Stratagene Gel Cleaning spin columns (Agilent Technologies, Santa Clara, CA), and eluted in sterile water. Purified PCR products were assessed for quantity and quality by using a NanoDrop 2000 microvoiume spectrophotometer (Thermo Fisher Scientific Inc.). Ap- proximately 200 ng of each PCR product was cycle-se- quenced with ABI Prism BigDye Terminator sequenc- ing mix (Applied Biosystems) by using the following reaction profile: 30 cycles at 95°C for 30 s, 55°C for 15 s, and 72°C for 4 min. Sequencing reactions were pre- cipitated with ethanol and sodium acetate, dried for at least 30 min at 37°C, and resuspended in 23mL of Hi- Di formamide (Thermo Fisher Scientific Inc.). Samples were analyzed in both directions on an Applied Biosys- tems 3130x1 DNA Analyzer (Thermo Fisher Scientific Inc.). Forward and reverse sequences were assembled into contiguous sequences by using Sequencher, vers. 4.9 (Gene Codes Corp., Ann Arbor, MI) and, when nec- essary, modified by hand. Primer sequences were iden- tified and excised. Microsatellite loci exhibited as many as 29 alleles per locus in >200 individuals surveyed in southwest Florida and showed high levels of heterozygosity (C. Curtis, unpubl. data). No evidence of null alleles, large allele dropout, or scoring errors was detected with MICRO-CHECKER among the 6 loci, although locus Ppe69 was found to be out of Hardy-Weinberg equilib- rium. With 95% confidence limits and 1000 randomiza- tions, ML-Relate software (Kalinowski et al., 2006) was used to obtain a maximum likelihood estimate of the coefficient of relatedness (r) among potential siblings to estimate the number of mothers represented in the samples. Movement analysis Analysis of acoustic data followed the method of Poula- kis et al. (2013). The position of individuals along the Caloosahatchee River was estimated by using a mean- position algorithm (Simpfendorfer et al., 2002, 2008a). Each acoustic receiver was assigned a position along the centerline of the river on the basis of the distance of the receiver from the mouth of the river (the mouth of the river was assigned the position of river kilome- 466 Fishery Bulletin 114(4) ter 0; the capture location was assigned the position of river kilometer 10.5). A geographic information system (GIS) layer of the centerline of the river was extracted from the U.S. Geological Survey National Hydrography Dataset (website) and divided into 0.2-km (0.1-nauti- cal mile) segments. The mean position estimates of the location of smalltooth sawfish, averaged in 1-h incre- ments, formed the basis of all subsequent analyses. In general, a location designation (e.g., present at the capture location) was given only to smalltooth saw- fish position estimates that exactly matched the river kilometer values assigned to each acoustic receiver. For example, to be counted as present at the non-main- stem Cape Coral capture location, data from a trans- mitter had to be recorded on at least one of the interior receivers for an entire hour (i.e., without main-stem detections). A smalltooth sawfish positioned at the en- trance of the capture location would have yielded a mean river kilometer position estimate intermediate between the capture location and a nearby main-stem receiver and, therefore, would have been placed in the main stem. Receivers at the capture location and in Glover Bight were isolated from direct acoustic connec- tion with nearby main-stem receivers to help with this determination. We analyzed movement data for smalltooth sawfish that retained their acoustic tags for at least 25 days. Position estimates for individuals were used to calcu- late daily activity space (in river kilometers), residence time, and diel activity. As with previous studies (e.g., Simpfendorfer et ah, 2011; Poulakis et ah, 2013), daily activity space was calculated for each individual as the difference between the most upriver and most down- river position estimates for each day. Main-stem and non-main-stem activity (presence or absence) were cal- culated for each individual as the percentage of hourly position estimates that corresponded with each habitat. Diel activity was estimated by adding hourly position estimates that occurred during local day (0600-1859) and night (1900-0559) in the 2 habitats for each in- dividual. A chi-square test was used to determine whether the observed day and night position estimates in the 2 habitats (pooling all individuals) differed from an even distribution. The South Florida Water Man- agement District provided data on daily mean fresh- water inflow from the Franklin Lock (located at river kilometer 48) for comparison with movement tracks of individual smalltooth sawfish. Results Genetics Two mitochondrial haplotypes were identified among the 10 smalltooth sawfish from the capture location. The haplotypes differed by one C or T transition. The capture location contained 3 individuals with haplo- type C and 7 individuals with haplotype T (Table 1), indicating that at least 2 mothers had contributed to the group of juveniles sampled because, in vertebrates, mitochondrial DNA is inherited from the mother. To further refine the number of mothers that con- tributed to the group, microsatellite data from the 2 mitochondrial groups (C and T) were analyzed sepa- rately using ML-Relate software to estimate the relat- edness of individuals within each of the 2 maternal lin- eages. Within the C group (3 individuals), 2 microsatel- lite loci (Ppell4 and Ppel68) had more than 4 alleles per locus (Table 1). In the absence of mutation, the maximum number of alleles that can be shared among full siblings at a single locus is 4. The presence of more than 4 alleles at more than 1 locus indicates that more than 2 parents contributed to the C group of juveniles because simultaneous, independent mutational events at more than 1 locus are highly unlikely. Within the T group (7 individuals), 3 loci (Ppell4, Ppel22, and Ppel65) contained more than 4 alleles per locus, and 2 of the loci had 7 alleles per locus (Table 1), indicating that more than 2 parents contributed to this group as well. An excess of alleles (i.e., >4) was not observed in loci Ppe69 and Ppel60 for either of the mitochondrial groups. This finding is not inconsistent with more than 2 parents contributing to the group because one or both contributing parents may have been homozygous for the alleles detected for these loci. The values of r among the 10 unique smalltooth sawfish caught and monitored at the capture location indicate that at least 4 pairs of individuals in the T group had high r estimates (-0.5) and that those pairs were full-sibling dyads (Table 2). There was a potential half-sibling pair (r= -0.25) in the C group, along with several unrelated individuals in each mitochondrial lineage. These estimates of relatedness indicate that at least 4 mothers contributed to the 10 individuals. These data support hypotheses 3 and 4 (not related and transient; not related and resident [see more on distinguishing between these hypotheses below]) and are inconsistent with hypotheses 1 and 2 (related and transient; related and resident). Movements Young-of-the-year individuals ranged from 775 to 1218 mm STL when they were tagged, and mean tag reten- tion was 184 days (standard error [SE] 68.5; Table 3). Half of the 10 individuals were caught on hook and line, and half were caught in gill nets. Six individuals met the analysis criterion of at least 25 days of tag retention. The mean daily activity space of these indi- viduals was 0.7 river km (SE 0.2). We documented the presence of multiple smalltooth sawfish at the capture location at the same time; on average, individuals spent 61% of their time there. Up to 4 of the tagged individuals used the capture location at the same time (sawfish 2, 4-6; 1-14 August), and 2 individuals used it at the same time for 2.5 months (sawfish 2 and 4; 1 June-14 August) (Fig. 2). The in- dividual with the longest tag retention (sawfish 4) re- sided at the capture location for long periods, ranging Poulakis et al.: Long-term site fidelity of Pristis pectinata 467 Table 2 Pairwise coefficients of relatedness (r) among the 10 smalltooth sawfish (Pristis pectinata) caught in 2006 in the Caloosahatchee River, Florida. Pairwise compari- sons are shown within each of the 2 mitochondrial DNA lineages (mtDNA haplotypes C and T). Individual pair r mtDNA haplotype FWC33: FWC36 0.2530 C FWC33: FWC38 0 C FWC36: FWC38 0 c FWC26 : FWC27 0 T FWC26: FWC28 0.0824 T FWC26: FWC29 0.0830 T FWC26: FWC31 0.0236 T FWC26: FWC32 0.6096 T FWC26: FWC37 0.5652 T FWC27: FWC28 0.5985 T FWC27: FWC29 0 T FWC27: FWC31 0.1658 T FWC27: FWC32 0 T FWC27: FWC37 0 T FWC28: FWC29 0 T FWC28: FWC31 0 T FWC28: FWC32 0 T FWC28: FWC37 0.0824 T FWC29: FWC31 0.0236 T FWC29: FWC32 0.5459 T FWC29: FWC37 0.5000 T FWC31: FWC32 0.0263 T FWC31; FWC37 0.0236 T FWC32: FWC37 0 T from a few days to as many as 86 consecutive days, without being detected on a main-stem receiver beyond the receiver located at the entrance of the canal sys- tem of the capture location. In addition, 3 smalltooth sawfish tagged with acoustic transmitters in the main- stem of the river moved into the capture location array, remaining there for as long as 14 consecutive days. Tracks of the analyzed individuals showed that dai- ly activity space was limited to a small portion of the study area when smalltooth sawfish were associated with the capture location, but they occasionally moved greater distances (Fig. 3). During September in 2006, after Tropical Storm Ernesto passed over the study area (29 August-1 September), prompting a freshwater release from the Franklin Lock that exceeded 500 m%, 4 individuals (sawfish 3-6) moved downriver away from the capture location to at least river kilometer 5. At least 3 of these individuals (sawfish 4-6) returned to the capture location after the storm passed, and 1 individual remained in the vicinity of the capture loca- tion for at least 10 more months (sawfish 4). After the storm, sawfish 5 and 6 returned to the capture location periodically but continued to make excursions down- river even without large increases in freshwater flow. Sawfish 3 may have exited the river because of the ef- fects of the storm, and sawfish 6 may have exited the river or shed its tag by the end of the year. These data, when combined with the genetic analyses, eliminate hypothesis 3 as a possibility and support hypothesis 4: young-of-the-year smalltooth sawfish that occurred at the capture location were 1) from at least 4 mothers and 2) long-term users of this location. While they were associated with the capture loca- tion, all 4 young-of-the-year smalltooth sawfish that Table 3 Summary of movement and habitat use by 6 smalltooth sawfish {Pristis pectinata) after they were tagged and monitored in 2006 and 2007 at the Cape Coral capture location in the Caloosahatchee River, Florida. Individuals are listed in order by date (month/day/year) of first detection from most recently tagged (top) to oldest. Residence time is the percentage of total hourly position estimates that occurred within the non-main-stem seawall canal capture location or in the main stem of the river. Sawfish 4 and 5 were monitored for ~1 year or more because they were recaptured and re-tagged. Four additional individuals were captured but shed their acoustic tags within a few days (their sizes were 800, 805, 842, 1200 mm stretch total length [STL]). Standard errors (SEs) of the mean are given in parentheses. The asterisk (*) indicates that although the individual was tagged at the capture location on 24 April 2006 the acoustic receivers were not deployed until 22 May 2006. Smalltooth sawfish STL (mm) Sex Date of first detection Date of last detection Days monitored Mean (SE) daily activity space (km) Residence time in canals (%) Residence time in main stem habitats (%) 6 1097 F 08/01/06 12/29/06 150 1.2 (0.2) 28.1 71.9 5 1218 F 08/01/06 07/19/07 352 0.7 (0.1) 36.7 63.3 4 950 F 06/14/06 08/20/07 432 0.4 (0.0) 80.3 19.7 3 996 F 06/07/06 08/23/06 77 1.0 (0.3) 70.5 29.5 2 1045 M 06/07/06 08/15/06 69 0.9 (0.1) 51.6 48.4 1 775* F 05/22/06 06/16/06 25 0.0 (0.0) 99.7 0.3 Mean (SE) 1013 (SE 61) 184 (SE 68.5) 0.7 (SE 0.2) 61.2 (SE 11.1) 38.8 (SE 11.1) 468 Fishery Bulletin 114(4) to (n 10.5 river km; night) leading up to the passage of Tropical Storm Ernesto that occurred 29 August-1 September. (B) Diel move- ment pattern of the same individual after passage of Tropical Storm Ernesto. Between 13 and 19 September (the only portion shown for detail), this juvenile was located in the natural red-mangrove-lined creek por- tion of the Glover Bight hotspot (river kilometer 3.6) during the day and moved into the open water portion of Glover Bight at night (river kilometer 3.1) before eventually returning to the capture location (see Fig. 3 for a longer time series). activity spaces, it seems unlikely (Simpfendorfer et al., 2011). The fact that some of these smalltooth sawfish had a high degree of site fidelity for their capture lo- cation, for months at a time, although their activity spaces were predicted to increase as the fish grew, is noteworthy. Poulakis et al.: Long-term site fidelity of Pristis pectinata 471 Table 4 Diel activity of smalltooth sawfish (Pristis pectinata) monitored at the non-main-stem Cape Coral capture location and in the main stem of the Caloosahatchee River, Flor- ida, in 2006 and 2007. The number of position estimates (all individuals pooled) was significantly greater at night in the main stem of the river (P<0.001, x^=208, df=l). STL=stretch total length. Capture location activity Main-stem activity Smalltooth sawfish STL (mm) Day Night Day Night 6 1097 176 61 114 492 5 1218 361 249 489 565 4 950 1437 1892 277 538 3 996 100 80 22 53 2 1045 262 125 60 310 1 775 207 182 1 0 Total 2543 2589 963 1958 Movements of related species associated with small portions of their nurseries have been observed in oth- er systems. Using a variety of data including acoustic tracking, Gruber et al. (1988) documented site fidelity by juvenile lemon shark (Negaprion breuirostris), as well as habitat partitioning by young-of-the-year, ju- venile, and adult lemon shark on short (up to 8 days) and long temporal scales. The “nursery zone” of their study area in Bimini, Bahamas, was used by the small- est juveniles and was characterized by the shallowest water and mangrove-fringed shorelines, similar to the habitats used by smalltooth sawfish in their natural hotspots (Poulakis et al., 2011). Using acoustic moni- toring, Heupel et al. (2003) documented movements of blacktip shark (Carcharhinus limbatus) away from and back to Terra Ceia Bay (a small bay connected to the Tampa Bay estuary in Florida) — movements that were associated with the passage of a tropical storm. They analyzed several environmental factors and at- tributed the response of the blacktip shark to changes in barometric and hydrostatic pressure. Collectively, these data are examples of patterns of general habi- tat-use and the types of environmental cues that influ- ence habitat use by elasmobranchs in localized coastal nurseries (for review, see Simpfendorfer and Heupel, 2012). Factors influencing the initial use of nursery hotspots by juvenile elasmobranchs could include selec- tion of the area as a birth site by their mothers. Long- term interdisciplinary data have indicated that adult female lemon shark and blacktip shark are philopatric and, therefore, multiple cohorts of juveniles use local- ized nursery habitats (Feldheim et al., 2004; Hueter et al., 2005; DiBattista et al., 2008). Adult female small- tooth sawfish and largetooth sawfish also have been shown to be philopatric (Phillips et al., 2011; K. Feld- heim, unpubl. data); therefore, they are returning to the same general nursery areas to give birth, findings that have implications for management, conservation, and recovery on a large spatial scale. Upon arrival in the nurseries, pregnant female smalltooth sawfish may give birth to their young in the hotspots rather than more broadly in the river system. In general, these hotspots are close to shore (<100 m), shallow (<1 m) and near deeper (at least 2 m) water. Hotspots identified in previous studies are currently dominated by shorelines with natural red mangroves and shallow water, and these characteristics have been incorporated into the official critical habitat designa- tion for juvenile smalltooth sawfish (Poulakis et al., 2011; Norton et al., 2012). Historical aerial photographs indicate that the capture location, which was the focus of our study, may have been dominated by shorelines with natural red mangrove and shallow water (<1 m) before canals were constructed. Whether this site rep- resents a relict habitat, genetically imprinted into the species, or is in fact suitable habitat that provides a fitness advantage for individual smalltooth sawfish is of interest in the context of niche theory. We document a close association of juvenile small- tooth sawfish with hotspots at a nursery-level spatial scale. Although the entire river was accessible, mul- tiple individuals used the habitats associated with the capture location for months at a time. As in previous studies, however, large-scale movements were observed when juveniles moved downriver as freshwater flows associated with a tropical storm exceeded 500 m^/s (Poulakis et al., 2013). That individuals returned to their exact capture location even with its artiflcial, sea- wall-canal habitats, after relatively large-scale move- ments of 5-7 river km away from the capture location, was somewhat unexpected. It indicates that factors supporting site fldelity (outside the period of the tropi- cal storm), such as adequate food availability, optimal temperature, and reduced predation risk, were satisfied in these artificial habitats. 472 Fishery Bulletin 1 14(4) At a small spatial scale, a diel activity pattern was observed for multiple individuals at the capture loca- tion and for one individual after relocation to a down- river hotspot after a tropical storm. This observation indicates that the factors eliciting these diel behaviors were present at both places. For example, predation pressure may have caused the juveniles to reside in non-main-stem portions of both hotspots during the day, whereas insufficient food may have caused them to move into open-water habitats at night. In contrast to detection of diel movements, we ob- served continuous use of the capture location. There are likely multiple explanations for these apparently contradictory behaviors. A portion of the population may favor refuge, continuously occupying habitats off the main stem of the river, such as those in the capture location canal system. The long periods of site fidel- ity documented at the capture location indicate that adequate food was available there, at least for some individuals. Another portion of the population may be driven to maximize growth and is more likely to ven- ture into the main stem of the river, especially if intra- or interspecific competition occurs in the refuge habi- tat. Those individuals exhibiting a diel pattern may fall between these 2 behaviors. Trade-offs between food and refuge have been discussed thoroughly in the literature on juvenile fish and may help explain the varied move- ment patterns of juvenile smalltooth sawfish in rela- tion to their nursery hotspots (e.g., for background, see Beck et al., 2001 and Heupel et al., 2007). There is some support for the idea that predation influences habitat use by juvenile smalltooth sawfish, especially for the smallest individuals. For example, at the Glover Bight hotspot, neonate smalltooth sawfish have been observed among red mangrove prop roots during the day while a shark (species unknown) was feeding in the central portion of the creek (Poulakis et al., 2011; Fig. 1), providing anecdotal evidence that behavior of smalltooth sawfish is influenced by preda- tor avoidance. Further support for the use of non- main-stem habitats as refuges from predation comes from research that has shown that juvenile bull shark (Carcharhinus leucas), the primary potential predator of juvenile smalltooth sawfish in most nurseries (Simp- fendorfer et al., 2005), are more common in the deeper, open-water portions of the Caloosahatchee River (Heu- pel et al., 2010), which smalltooth sawfish have been observed to often avoid during the day. Juvenile black- tip sharks have exhibited what might also be predator- avoidance behaviors whereby they aggregate during the day and disperse at night (Heupel and Simpfen- dorfer, 2005). Feeding is likely to have a major influence on habi- tat use of smalltooth sawfish because growth is fast in the nursery; smalltooth sawfish double in length dur- ing their first year of life (Simpfendorfer et al., 2008b; Scharer et al., 2012). Little is known about the diet of smalltooth sawfish beyond general characterizations and a few observations from field sampling, anglers, and necropsies to document food items, including pink shrimp (as bait) and fish, such as clupeids, carangids, mullet (Mugil spp.), pinfish (Lagodon rhomboides) , and a stingray (Dasyatis sp.) (Bigelow and Schroeder, 1953; Poulakis et al., 2013). Traditional stomach-content analyses are not possible because the smalltooth saw- fish is endangered; therefore, indirect methods such as analysis of stable isotopes (Fisk et al., 2002) have been employed, and such studies have indicated that fish make up the majority of the diet at all life stages (Poulakis et al.^). Monitoring the distribution and abundance of poten- tial prey fish may provide insights into habitat use by smalltooth sawfish. Also, given the distinct side-to-side movements of the rostrum that the smalltooth sawfish uses for feeding (Wueringer et al., 2012), research is underway that uses a novel indirect method, the at- tachment of acceleration data loggers (Whitney et al., 2012), to monitor potential feeding behavior and may help determine whether food acquisition or other fac- tors (e.g., temperature; Schlaff et al., 2014) contribute to the large-scale relocation of individuals within the nursery or the smaller-scale diel movements observed in this study. Evidence that movements of juvenile smalltooth sawfish in the nursery are cued, directly or indirectly, by changes in the volume of freshwater inflow (Pou- lakis et al., 2013) and evidence that juveniles move between nursery hotspots (this study) highlight the need for freshwater inflow management strategies that would minimize the need for these movements in flow- managed systems such as the Caloosahatchee River (Barnes, 2005). Although there appears to be little pre- dation on smalltooth sawfish in the nursery (Poulakis et al., 2011), predation risk may increase during the time it takes for smalltooth sawfish to find suitable alternate habitats. Predation risk could also increase because predators, like the bull shark, tend to relocate toward the river mouth and co-occur with smalltooth sawfish when freshwater flow increases (Heupel and Simpfendorfer, 2008). As a result, freshwater inflow management strategies that mimic the environmental variability induced by the natural dry and wet seasons may maximize survivorship of smalltooth sawfish (see Poulakis et al., 2011; Simpfendorfer et al., 2011). Our study was conducted in a highly altered nurs- ery, but the hotspot concept introduced by Poulakis et al. (2011) appears to also apply to smalltooth sawfish in other nurseries. Poulakis et al.^ showed that the num- ber of hourly acoustic detections from tagged smalltooth sawfish in upper Charlotte Harbor was greatest in a 2-river-km portion of the Peace River associated with the only documented hotspot in this river (Poulakis et al., 2011). In 2010-2013, between river kilometers 8 and 10 of the Peace River, there were more than twice the number of hourly detections than there were in ei- ther of the 2-river-km sections above and below this location, indicating that for most of their monitoring periods, individuals from multiple cohorts remained in a small region of this large, relatively unaltered river. The high site fidelity to hotspots of juveniles observed Poulakis et al.: Long-term site fidelity of Pristis pectinata 473 in the Peace and Caloosahatchee rivers, including the hotspot with artificial habitats in our study, indicates that habitat fragmentation in the highly altered Caloo- sahatchee River may not be the only factor driving pat- terns of habitat use there. Therefore, growing evidence indicates that hotspots exist in nurseries throughout the range of smalltooth sawfish, including in regions to the south of the Charlotte Harbor estuarine system, such as the relatively pristine Ten Thousand Islands and Everglades National Park (Simpfendorfer et al., 2010; Hollensead et al., 2016; O’DonnelH). In these areas, catch rates, acoustic monitoring, and acoustic tracking have all shown high interannual occurrence in hotspot-like areas, such as Faka Union and Mud bays (Simpfendorfer et al., 2010; Hollensead et al., 2016). Consistent, large-scale acoustic monitoring in these regions would test our hypothesis that hotspots exist in nurseries throughout the range of smalltooth sawfish and would help identify their boundaries. Further, mul- tiple lines of evidence indicate that hotspots also exist for larger (>2.2 m STL) size classes of smalltooth saw- fish, after they leave the nurseries, and the same tech- niques that have been applied in the nurseries are be- ing applied in northwest Florida Bay and off southeast Florida to elucidate the spatial and temporal extent of these high-use areas by smalltooth sawfish (Poulakis and Seitz, 2004; Waters et al., 2014; Papastamatiou et al., 2015). Testing the hotspot concept in a variety of localities and for multiple life stages may help focus future research, management, and conservation efforts for the Pristidae and ultimately promote recovery of these endangered elasmobranchs. Acknowledgments Our research on smalltooth sawfish is ongoing and has been supported primarily by funding from the Nation- al Marine Fisheries Service, NOAA, through Section 6 (Cooperation with the States) of the U.S. Endangered Species Act under grant awards to the Florida Fish and Wildlife Conservation Commission from both NOAA (NA06NMF4720032) and the National Fish and Wildlife Foundation (2003-0206-008 and 2004-0012-008). State- ments, findings, conclusions, and recommendations are those of the authors and do not necessarily reflect the views or policies of the DOC, NOAA, or the National Fish and Wildlife Foundation. We thank J. Darrow for assistance producing Figure 1, R. Scharer for producing Figures 4 and 5, and B. Yeiser for access to data re- corded on lower-river receivers. D. Adams, J. Adams, B. Crowder, and R. Scharer improved earlier versions of the manuscript. The senior author thanks G. Maul, J. Shenker, E. Irlandi, K. Johnson, and C. Knoll for their support of this work, which was completed to partially fulfill the requirements for his doctoral degree. This re- search was conducted under Endangered Species Per- * O’Donnell, P. 2014. Unpubl. data. Rookery Bay National Estuarine Research Reserve, Naples, FL 34113 mit number 1475 (FWC) issued by the National Marine Fisheries Service. Literature cited Barnes, T. 2005. Caloosahatchee Estuary conceptual ecological model. Wetlands 25:884-897. Beck, M. W. K. L. Heck Jr., K. W. Able, D. L. Childers, D. B. Eggleston, B. M. Gillanders, B. Halpern, C. G. Hays, K. Hoshino, T. J. Minello, et al. 2001. The identification, conservation, and management of estuarine and marine nurseries for fish and inverte- brates. BioScience 51:633-641. Bigelow, H. B., and W. C. Schroeder. 1953. Fishes of the western North Atlantic, Part 2. Saw- fishes, guitarfishes, skates, rays, and chimaeroids, 588 p. Mem. Sears Found. Mar. Res., Yale Univ., New Ha- ven, CT. Carlson, J. K., S. J. B. Gulak, C. A. Simpfendorfer, R. D. Grubbs, J. G. Romine, and G. H. Burgess. 2014. Movement patterns and habitat use of smalltooth sawfish, Pristis pectinata, determined using pop-up satel- lite archival tags. Aquat. Conserv. 24:104-117. Chamberlain, R. H., and P. H. Doering. 1998. Freshwater inflow to the Caloosahatchee Estuary and the resource-based method for evaluation. In Pro- ceedings of the Charlotte Harbor Public Conference and Technical Symposium. Charlotte Harbor Natl. Est. Pro- gram Tech. Rep. 98-02; Punta Gorda, FL, 15-16 March 1997 (S. F. Treat, ed.), p. 81-90. South Florida Water Management District, West Palm Beach, FL. Collins, A. B., M. R. Heupel, and C. A. Simpfendorfer. 2008. Spatial distribution and long-term movement pat- terns of cownose rays Rhinoptera bonasus within an es- tuarine river. Est. Coasts 31:1174-1183. DiBattista, J. D., K. A. Feldheim, X. Thibert-Plante, S. H. Gru- ber, and A. P. Hendry. 2008. A genetic assessment of polyandry and breeding- site fidelity in lemon sharks. Mol. Ecol. 17:3337-3351. Elton, C. 1927. Animal ecology, 209 p. The Macmillan Co., New York. Estok, P. 2007. Seasonal changes in the sex ratio of Nyctalus spe- cies in north-east Hungary. Acta Zool. Acad. Sci. Hung. 53:89-95. Federal Register 2009. Endangered and threatened species; critical habitat for the endangered distinct population segment of small- tooth sawfish. Fed. Reg. 74:45353-45378. Feldheim, K. A., S. H. Gruber, and M. V. Ashley. 2004. Reconstruction of parental microsatellite genotypes reveals female polyandry and philopatry in the lemon shark, Negaprion brevirostris. Evolution 58:2332-2342. Feldheim, K. A., D. D. Chapman, C. A. Simpfendorfer, V. P. Richards, M. S. Shivji, T. R. Wiley, G. R. Poulakis, J. K. Carl- son, R. Eng, and S. Sagarese. 2010. Genetic tools to support the conservation of the en- dangered smalltooth sawfish, Pristis pectinata. Conserv. Genet. Resour. 2:105-113. 474 Fishery Bulletin 114(4) Fisk, A. T., S. A. Tittlemier, J. L. Pranschke, and R. J. Norstrom. 2002. Using anthropogenic contaminants and sta- ble isotopes to assess feeding ecology of Greenland sharks. Ecology 83:2162-2172. Grinnell, J. 1917. The niche-relationships of the California thrasher. The Auk 34:427-433. Grubbs, R. D. 2010. Ontogenetic shifts in movements and habitat use. In Sharks and their relatives II: biodiversity, adaptive physiology, and conservation (J. C. Carrier, J. A. Musick, and M. R. Heithaus, eds.), p. 319-350. CRC Press, Boca Raton, FL. Gruber, S. H., D. R. Nelson, and J. F. Morrissey. 1988. Patterns of activity and space utilization of lemon sharks, Negaprion brevirostris, in a shallow Bahamian lagoon. Bull. Mar. Sci. 43:61-76. Guisan, A., and N. E. Zimmermann. 2000. Predictive habitat distribution models in ecol- ogy. Ecol. Model. 135:147-186. Heupel, M. R., and C. A. Simpfendorfer. 2005. Quantitative analysis of aggregation behavior in ju- venile blacktip sharks. Mar. Biol. 147:1239-1249. 2008. Movement and distribution of young bull sharks Carcharhiniis leucas in a variable estuarine environ- ment. Aquat. Biol. 1:277—289. Heupel, M. R., C. A. Simpfendorfer, and R. E. Hueter. 2003. Running before the storm: blacktip sharks respond to falling barometric pressure associated with Tropical Storm Gabrielle. J. Fish Biol. 63:1357-1363. Heupel, M. R., J. K. Carlson, and C. A. Simpfendorfer. 2007. Shark nursery areas: concepts, definition, char- acterization and assumptions. Mar. Ecol. Prog. Ser. 337:287-297. Heupel, M. R., B. G. Yeiser, A. B. Collins, L. Ortega, and C. A. Simpfendorfer. 2010. Long-term presence and movement patterns of ju- venile bull sharks, Carcharhiniis leucas, in an estuarine river system. Mar. Freshw. Res. 61:1-10. Hirzel, A. H., and G. Le Lay. 2008. Habitat suitability modelling and niche theory. J. Appl. Ecol. 45:1372-1381. Hollensead, L. D., R. D. Grubbs, J. K. Carlson, and D. M. Bethea. 2016. Analysis of fine-scale daily movement patterns of juvenile Pristis pectinata within a nursery habitat. Aquat. Conserv. 26:492-505. Hueter, R. E., M. R. Heupel, E. J. Heist, and D. B. Keeney. 2005. Evidence of philopatry in sharks and implications for the management of shark fisheries. J. Northwest Atl. Fish. Sci. 35:239-247. Hutchinson, G. E. 1957. Concluding remarks. Population studies: animal ecology and demography. Cold Spring Harbor Symp. Quant. Biol. 22:415-427. Jones, C. M. 2002. Age and growth. In Fishery science: the unique contributions of early life stages. (L. A. Fuiman and R. G. Werner, eds.), p. 33-63. Blackwell Science, Oxford, UK. Kalinowski, S. T., A. P. Wagner, and M. L. Taper. 2006. ML-RELATE: a computer program for maximum like- lihood estimation of relatedness and relationship. Mol. Ecol. Notes 6:576-579. Morgan, D. L., J. M. Whitty, N. M. Phillips, D. C. Thorburn, J. A. Chaplin, and R. McAuley. 2011. North-western Australia as a hotspot for endan- gered elasmobranchs with particular reference to saw- fishes and the northern river shark. J. R. Soc. West. Aust. 94:345-358. Morgan, D. L., M. G. Allen, B. C. Ebner, J. M. Whitty, and S. J. Beatty. 2015. Discovery of a pupping site and nursery for criti- cally endangered green sawfish Pristis zijsron. J. Fish Biol. 86:1658-1663. Norton, S. L.,T. R. Wiley, J. K. Carlson, A. L. Frick, G. R. Pou- lakis, and C. A. Simpfendorfer. 2012. Designating critical habitat for juvenile endangered smalltooth sawfish in the United States. Mar. Coast. Fish. 4:473-480. Papastamatiou, Y. P., R. D. Grubbs, J. Imhoff, S. J. B. Gulak, J. K. Carlson, and G. H. Burgess. 2015. A subtropical embayment serves as essential habi- tat for sub-adults and adults of the critically endangered smalltooth sawfish. Global Ecol. Conserv. 3:764-775. Pardini, A. T, C. S. Jones, L. R. Noble, B. Kreiser, H. Malcolm, B. D. Bruce, J. D. Stevens, G. Cliff, M. C. Scholl, M. Francis, et al. 2001. Sex-biased dispersal of great white sharks. Na- ture 412:139-140. Phillips, N. M., J. A. Chaplin, D. L. Morgan, and S. C. Peverell. 2011. Population genetic structure and genetic diversity of three critically endangered Pristis sawfishes in Aus- tralian waters. Mar. Biol. 158:903-915. Polovina, J., 1. Uchida, G. Balazs, E. A. Howell, D. Parker, and P. Dutton. 2006. The Kuroshio Extension Bifurcation Region: a pe- lagic hotspot for juvenile loggerhead sea turtles. Deep- Sea Res., H 53:326-339. Poulakis, G. R. 2012. Distribution, habitat use, and movements of juve- nile smalltooth sawfish, Pristis pectinata, in the Char- lotte Harbor estuarine system, Florida. Ph.D. diss., 165 p. Florida Inst. Tech., Melbourne, Florida. Poulakis, G. R., and J. C. Seitz. 2004. Recent occurrence of the smalltooth sawfish, Pristis pectinata (Elasmobranchiomorphi: Pristidae), in Florida Bay and the Florida Keys, with comments on sawfish ecology. Fla. Sci. 67:27-35. Poulakis, G. R., P. W. Stevens, A. A. Timmers, T. R. Wiley, and C. A. Simpfendorfer. 2011. Abiotic affinities and spatiotemporal distribution of endangered smalltooth sawfish, Pristis pectinata, in a south-western Florida nursery. Mar. Freshw. Res. 62:1165-1177. Poulakis, G. R., P. W. Stevens, A. A. Timmers, C. J. Stafford, and C. A. Simpfendorfer. 2013. Movements of juvenile endangered smalltooth saw- fish, Pristis pectinata, in an estuarine river system: use of non-main-stem river habitats and lagged responses to freshwater inflow-related changes. Environ. Biol. Fish. 96:763-778. Raymond, M., and F. Rousset. 1995. GENEPOP (version 1.2): population genetics software for exact tests and ecumenicism. J. Hered. 86:248-249. Rousset, F. 2008. GENEPOP ’007: a complete re-implementation of Poulakis et al.: Long-term site fidelity of Pristis pectinata 475 the GENEPOP software for Windows and Linux. Mol. Ecol. Resour. 8:103-106. Scharer, R. M., W. F. Patterson III, J. K. Carlson, and G. R. Poulakis. 2012. Age and growth of endangered smalltooth sawfish (Pristis pectinata) verified with LA-ICP-MS analysis of vertebrae. PLoS ONE 7(10):e47850. Schlaff, A. M., M. R. Heupel, and C. A. Simpfendorfer. 2014. Influence of environmental factors on shark and ray movement, behavior and habitat use: a review. Rev. Fish Biol. Fish. 24:1089-1103. Seitz, J. C., and G. R. Poulakis. 2002. Recent occurrence of sawfishes (Elasmobranchiomor- phi: Pristidae) along the southwest coast of Florida (USA). Fla. Sci. 65:256-266. Simpfendorfer, C. A., and M. R. Heupel. 2012. Assessing habitat use and movement. In Biology of sharks and their relatives, 2nd ed. (J. C. Carrier, J. A. Musick, and M. R. Heithaus, eds.), p. 579-601. CRC Press, Boca Raton, FL. Simpfendorfer, C. A., M. R. Heupel, and R. E. Hueter. 2002. Estimation of short-term centers of activity from an array of omnidirectional hydrophones and its use in studying animal movements. Can. J. Fish. Aquat. Sci. 59:23-32. Simpfendorfer, C. A., G. G. Freitas, T. R. Wiley, and M. R. Heupel. 2005. Distribution and habitat partitioning of immature bull sharks (Carcharhinus leucas) in a southwest Florida estuary. Estuaries 28:78-85. Simpfendorfer, C. A., M. R. Heupel, and A. B. Collins. 2008a. Variation in the performance of acoustic receivers and its implication for positioning algorithms in a river- ine setting. Can. J. Fish. Aquat. Sci. 65:482-492. Simpfendorfer, C. A., G. R. Poulakis, P. M. O’Donnell, and T. R. Wiley. 2008b. Growth rates of juvenile smalltooth sawfish Pristis pectinata Latham in the western Atlantic. J. Fish Biol. 72:711-723. Simpfendorfer, C. A., T. R. Wiley, and B. G. Yeiser. 2010. Improving conservation planning for an endangered sawfish using data from acoustic telemetry. Biol. Con- serv. 143:1460-1469. Simpfendorfer, C. A., B. G. Yeiser, T. R. Wiley, G. R. Poulakis, P. W. Stevens, and M. R. Heupel. 2011. Environmental influences on the spatial ecology of juvenile smalltooth sawfish (Pristis pectinata): results from acoustic monitoring. PLoS ONE 6(2):el6918. Stevens, P. W., D. A. Blewett, and G. R. Poulakis. 2007. Variable habitat use by juvenile common snook, Centropomus undecimalis (Pisces: Centropomidae): ap- plying a life-history model in a southwest Florida estu- ary. Bull. Mar. Sci. 80:93—108. Stevens, P. W., M. F. D. Greenwood, and D. A. Blewett. 2013. Fish assemblages in the oligohaline stretch of a southwest Florida river during periods of extreme fresh- water inflow variation. Trans. Am. Fish. Soc. 142:1644- 1658. Stoker, Y. E. 1992. Salinity distribution and variation with freshwater inflow and tide, and potential changes in salinity due to altered freshwater inflow in the Charlotte Harbor estua- rine system, Florida. U.S. Geological Survey, Water-Re- sour. Invest. Rep. 92-4062, 30 p. [Available at website.] van Oosterhout, C., W. F. Hutchinson, D. P. M. Wills, and P. Shipley. 2004. MICRO-CHECKER: software for identifying and correcting genotyping errors in microsatellite data. Mol. Ecol. Notes 4:535-538. Waters, J. D., R. Coelho, J. Fernandez-Carvalho, A. A. Tim- mers, T. Wiley, J. C. Seitz, M. T. McDavitt, G. H. Burgess, and G. R. Poulakis. 2014. Use of encounter data to model spatio-temporal distribution patterns of endangered smalltooth sawfish, Pristis pectinata, in the western Atlantic. Aquat. Con- serv. 24:760-776. Whitney, N. M., Y. P. Papastamatiou, and A. C. Gleiss. 2012. Integrative multisensory tagging: emerging tech- niques to link elasmobranch behavior, physiology, and ecology. In Biology of sharks and their relatives, 2nd ed. (J. C. Carrier, J. A. Musick, and M. R. Heithaus, eds.), p. 265-290. CRC Press, Boca Raton, FL. Whitty, J. M., D. L. Morgan, S. C. Peverell, D. C. Thorburn, and S. J. Beatty. 2009. Ontogenetic depth partitioning by juvenile fresh- water sawfish (Pristis microdon: Pristidae) in a riverine environment. Mar. Freshw. Res. 60:306-316. Wueringer, B. E., L. Squire Jr., S. M. Kaiiura, N. S. Hart, and S. P. Collin. 2012. The function of the sawfish’s saw. Current Biol. 22:R150-R151. 476 NOAA National Marine Fisheries Service Fishery Bulletin -%• established 1881 Spencer F. Baird First U.S. Commissioner of Fisheries and founder of Fishery Bulletin Evaluating the vulnerability of an atypical protogynous hermaphrodite to fishery exploitation; results from a population model for black sea bass iCentropristis striata} Jessica Blaylock^ Gary R. Shepherd^ Email address for contact author; jessica.blaylock@noaa.gov Abstract— Fish species that are pro- togynous hermaphrodites generally have sex ratios skewed toward males at large sizes. These skewed sex ratios at length can make a stock more vulnerable to overexploitation, particularly in fisheries where mini- mum size is regulated, resulting in the removal of reproductively active males. We developed a length-based population model to test the vulner- ability to exploitation of a population with an atypical protogynous life history, in particular the northern stock of black sea bass iCentropris- tis striata). Black sea bass north of Cape Hatteras, North Carolina, are unusual for protogynous species in that they may undergo prematura- tional transformation, remain female at large sizes, involve secondary males in spawning, and undertake seasonal migrations. The model was developed to examine the impact of participation by secondary males in population productivity, the influ- ence of size at sex transformation, and the subsequent robustness of the population under exploitation, in comparison with equivalent gonocho- ristic and typical protogynous popu- lations. Although the model does not capture all the dynamics of a popu- lation, such as density-dependent regulation of sex transformation, our results indicate that the northern stock of black sea bass may be more resilient in response to exploitation than would be expected if they were typical protogynous hermaphrodites. Manuscript submitted 21 May 2015. Manuscript accepted 19 August 2016. Fish. Bull. 114:476-489 (2016). Online publication date: 8 September 2016. doi: 10.7755/FB. 114.4.9 The views and opinions expressed or implied in this article are those of the author (or authors) and do not necessarily reflect the position of the National Marine Fisheries Service, NOAA. ' Integrated Statistics 16 Sumner Street Woods Hole, MA 02543 Present address: 3825 South Drive Fort Worth, Texas 76109 ^ Northeast Fisheries Science Center National Marine Fisheries Service, NOAA 166 Water Street Woods Hole, Massachusetts 02543 The capacity of a fish population to persist under exploitation is highly dependent on the life history of a species and its ability to effectively reproduce. In gonochoristic species, individuals remain the same sex throughout their lives, generally en- suring a constant presence of both males and females capable of repro- duction, even under relatively strong exploitation. In contrast, species with a hermaphroditic life history (either protogynous or protandrous) might not have an adequate abun- dance of sexually mature individuals across sizes and ages (Provost and Jensen^). In protogynous species, a common pathway for sex change in- volves individuals initially maturing and reproducing as female and then 1 Provost, M., and O. Jensen. 2013. Use of sex change considerations in stock assessments. In Proceedings from a workshop on modeling protogynous her- maphroditic fishes; Raleigh, NC, 29-30 August 2012 (G. Shepherd, K. Shertzer, J. Coakley, and M. Caldwell, eds.), p. 5-9. Mid-Atlantic Fishery Manage- ment Council, Dover, DE. [Available at website, accessed May 2015.] changing sex to male as needed. The result is a sex ratio dominated by fe- males at small sizes or young ages and males at large sizes or old ages (Warner, 1975; Allsop and West, 2004; Munday et ah, 2006). It is typical in a protogynous species for the large, dominant male to monopolize spawn- ing, although there is a wide range of behavioral roles for secondary males that may allow participation (Peters- en, 1991). The resulting sex ratio, highly skewed toward males at large sizes, tends to make protogynous spe- cies vulnerable to size-selective fish- eries, possibly resulting in a shortage of males — a shortage that limits the potential for egg fertilization (Hunts- man and Schaaf, 1994; Heppell et al., 2006). Characteristics of hermaphroditic species make them particularly chal- lenging to assess and manage. Man- agement of exploited fish stocks often is guided by the results of population dynamic models (Quinn and Deriso, 1999; Haddon, 2001; Brander, 2003), and basic models are predicated on certain assumptions about the life history of a species, including the ex- Blaylock and Shepherd: Vulnerability of Centropristis striata to exploitation 477 pectation that both sexes will be equally vulnerable to exploitation. Along with other possible scenarios, such as differential spatial patterns by sex or significant sex- ual dimorphism, hermaphroditic life histories often vio- late this assumption, and unique analytical approaches may be warranted. Implications of a protogynous life history for stock assessments and management have been explored previously (Alonzo and Mangel, 2004; Heppell et ah, 2006; Brooks et al., 2007; Alonzo et ah, 2008). The conclusion from simulation models, as well as empirical observations, is that populations subjected to sex-specific exploitation as a consequence of having a hermaphroditic life history are at a higher risk of overexploitation than are gonochoristic species (Alonzo and Mangel, 2004; Heppell et ah, 2006; Hamilton et ah, 2007). Thus developing management measures for a protogynous species on the basis of the assumptions implicit for gonochoristic species could possibly result in overfishing. The black sea bass {Centropristis striata) is a pro- togynous hermaphroditic species that is distributed from the Gulf of Maine to the Gulf of Mexico (Lavenda, 1949; Collette and Klein-MacPhee, 2002) and is man- aged as 3 unique stocks segregated by Cape Hatteras, North Carolina, and the Florida Keys (Bowen and Avise, 1990; Roy et al., 2012; McCartney et al., 2013). Each stock is subjected to commercial and recreation- al exploitation and is primarily caught in trawl gear and fish pots, as well as on hook-and-line gear in the coastal regions (Shepherd and Terceiro, 1994). Black sea bass in the South Atlantic and Gulf of Mexico are not migratory and are commonly associated with local structured habitat (Drohan et al., 2007). In contrast, black sea bass in the Middle Atlantic (i.e., the north- ern stock) undergo seasonal migrations of up to 300 km, moving from the continental shelf or slope edge in winter to the coastal waters in spring and summer (Musick and Mercer, 1977; Moser and Shepherd, 2009). Life history characteristics vary between the 2 southern stocks and the northern stock. Spawning in the Middle Atlantic stock occurs within a relatively lim- ited period compared with the spawning period of black sea bass in the southern stocks. In all 3 stocks, domi- nant males in spawning aggregations are characterized by a prominent nuccal hump and bright blue coloration around the hump and eyes (Lavenda, 1949); however, in the northern stock, additional males in ripe gonadal condition but lacking the secondary sex characteris- tics are also present in the spawning areas (NEFSC^). These features are typical of secondary males that may attempt to fertilize the eggs of females spawning with dominant males (Pitcher, 1993; Taborsky, 1994; Young et al., 2013). Empirical data collected during bottom trawl surveys of the Northeast Fisheries Science Cen- ter (NEFSC), National Marine Fisheries Service, in- ^ NEFSC (Northeast Fisheries Science Center). 2012. SB'"'* Northeast regional stock assessment workshop (53'''’ SAW) assessment report. Northeast Fish. Sci. Cent. Ref. Doc. 12- 05, 559 p. [Available at website, accessed May 2015.] dicate that about 30% of black sea bass at each size below 30 cm (all lengths in this manuscript refer to total length) were male and that a significant propor- tion (i.e., up to 45%) of black sea bass at sizes above 45 cm were female (NEFSC^). Additionally, growth in this stock appears to be similar for males and females. The characteristics of the black sea bass of the north- ern stock, such as a prematurational sex change, the presence of mature secondary males, and the relative abundance of large females (>45 cm ), distinguish them from typical protogynous hermaphrodites. The northern Atlantic population of black sea bass has been evaluated in stock assessments to support fishery management programs (NEFSC^). One of the main uncertainties cited in relation to the development of management measures is the implication of a pro- togynous life history. Specifically, there is concern that the removal of large males (>45 cm) would significantly affect the reproductive capability of the stock, there- by making it particularly vulnerable to exploitation. Implicit in that concern is that only the large males reproduce or, if not, the large females change sex to compensate for the reduced availability of large males. Although methods for stock assessment of protogy- nous populations have been evaluated, in several stud- ies, the implications of the possible contribution of secondary males to spawning in a population have not been explored fully. In particular, the response to ex- ploitation of a stock with functional secondary males that result from a prematurational sex change has not been tested to determine whether the stock is more or less vulnerable to overexploitation than a gonochoris- tic or typical protogynous stock with similar population attributes. Our objective was to develop a population model based on the northern stock of black sea bass to evaluate the robustness of an atypical protogynous population in response to fishery exploitation. We do not intend to provide new methods for stock assess- ments of hermaphroditic species but rather to improve the understanding of the relationship between the life history of a species and its ability to withstand exploitation. We developed a length-based population model to evaluate the impact of a range of exploitation inten- sities on population persistence and catch, under dif- ferent theoretical life history scenarios. We considered an experimental atypical protogynous (AP) population with 3 scenarios: 1) no mature secondary males con- tribute to spawning (AP-0); 2) 50% of mature secondary males participate in spawning (AP-50); and 3) 100% of mature secondary males participate in spawning (AP- 100). Note that our use of the term secondary simply implies that these males are not visibly dominant (i.e., secondary males lack the prominent nuccal hump and bright blue coloration). We did not make any assump- tions about the behavior of secondary males except that some of them are able to spawn. Results from simulations for the AP population were compared against those for equivalent gonocho- ristic (G) and typical protogynous (TP) populations. 478 Fishery Bulletin 1 14(4) Figure 1 Diagram of possible trajectories through 4 different life stages of black sea bass {Centropristis striata) (female [U], transitional [T], secondary male [S], and dominant male [D]) in the length- based model developed to study the response to exploitation by gonochoristic, atypical protogynous, and typical protogynous populations. Recruits are either female (RU) or male {RS), and there are 4 types of transitions in which a fish can transform into a different life stage by the next time period: 1) a female can become a transitional, 2) a transitional can become a sec- ondary male, 3) a transitional can become a dominant male, or 4) a secondary male can become a dominant male. Subscripts letters U, T, S, and D indicate the life stage at the previous time step. When available, parameter and other values used in the model came from data collected during bottom trawl surveys conducted by the NOAA Northeast Fisheries Science Center dur- ing 1984-2013. where spawning is monopolized by the domi- nant males. The model was structured to re- flect black sea bass life history, and param- eters were derived from empirical data when possible. Sex ratio characteristics for the AP population were based on the northern stock of black sea bass, and the TP population fol- lows a biased sex ratio expected among pro- togynous species (Provost and Jensen^) with recruits being 100% female and all fish tran- sitioning to male by the time they reach the maximum size in the model. We examined stock performance in the form of relative stock size, catch, spawning stock biomass (SSB), and number of recruits (i?) under increasing levels of fishing-induced mortality [hereafter: fishing mortality] (F), and we ran sensitivity analy- ses to evaluate the effect of different capture probabilities, female-to-male transition rates, proportions of secondary versus dominant males, and recruitment scenarios, as well as the effect of possible spawning by secondary males in the TP population. Materials and methods Model structure A length-structured model was developed to compute the abundance at length of fish in 4 life stages (i) over time: females {i=U), transi- tionals {i=T), secondary males (i=S), and domi- nant males ii=D). The model follows a monthly time step where each time period (t) is equiva- lent to the corresponding year (y) and month (m) (i.e., t=month m in year y). Recruits (<11 cm) in the model are ini- tially all female but some immediately un- dergo a change into male, according to population- specific recruit sex ratios. Female and male recruits can follow several potential trajectories over their life- time (Fig. 1). Female recruits (RU) will either remain female over their entire life or become transitional- phase fish [transitionals] for a single time step followed by transformation into either secondary or dominant males. Male recruits (RS) enter the population as sec- ondary males and either remain secondary males or become dominant males (without a transition phase). Therefore, at any time t, the abundance at length I of fish at a specific life stage is the sum of recruits of the given sex, of individuals entering the stage, and of individuals that were previously in that stage and survived: f/]t = RUn + f/uit, (1) ^it = ^uit. (2) (3) ■Dit - TJtu + Dsit + (4) where the subscripts U, T, S, and D indicate life stage at the previous time step (for example, St refers to the portion of all secondary males at time t that were transitionals in the previous month). Total population abundance over all lengths, iVt, is then calculated as + ^it + 'Sit + At*’ (5) where Z^ax = the maximum length of a fish in the population. From one time step to the next, processes affecting the population occur in the following order: mortality (including natural mortality [M] and F), transition, and growth. At each time step, SSB (calculated by using maturity-at-length and length-weight equations), re- cruitment, and catch are determined. Natural mortality at time t (M^) is assumed constant across stages, lengths, and months; therefore, Sit - -RSit + Sxit + Ssit, Blaylock and Shepherd: Vulnerability of Centropristis striata to exploitation 479 My 12 (6) where My = total annual natural mortality for year y. Fishing mortality for length I and time t is defined as (7) ^it = — 12 ' where Fy = total annual fishing mortality for year y; and 5i = the probability of capture (i.e., capture se- lectivity) at length I defined by a logistic 2-parameter model, 1 l + e (8) where a = the length of the fish with 50% probability of being retained (L50); and P = the slope of the selectivity curve 5i. It follows that catch in numbers for stage i at length I at time t is ^ilt — ■M, -(Fu+Mp ). (9) In the protogynous population scenarios, we as- sumed transitions (Fig. 1) occur only after spawning in month 11 or 12, where a fish can transform from a female to a transitional (/n=;ll), from a transitional to a secondary male (m=12), from a transitional directly to a dominant male (ot=12), or from a secondary male to a dominant male (m=12). For simplicity, fish can be in the transitional stage only during one time step. The probability of an individual of stage ii and length I in month m becoming stage ^2 iii the following month is represented by Transition rates are all zero for the G population because no sex change occurs; the sex ratio for this scenario is fixed at 1:1. Growth is assumed constant across stages and sex- es but varies annually by quarter, (g). Growth occurs at the start of each quarter and is defined by a lower triangular matrix (Gg) with dimensions /max^^maxj which each element g'L2,Li,q represents the probability that a fish of length LI at the start of quarter q will grow to length L2 by the start of the following quarter. Therefore, the abundance of individuals of stage ^2 af length L2 at time t, Ni^i,2u is calculated as the number of individuals of stage ii at length I (where 1^2) at time (^-l) that survived, that transitioned to stage 12, and that grew to length L2, summed over all values of 1: jv,,]. = (10) Given the uncertainty resulting from defining SSB as female only (Brooks et aL, 2007), recruitment levels are determined from male and female spawning stock as of September (m=9) and from recruits that entered the population in October (ni=10). Maturity is specified by using sex-specific maturity ogives defined by a 4-pa- rameter logistic equation, as follows: ■d; (11) where Oji 6j d, the proportion of fish of sex j (where j=U for females, j=V for males) and length I that are mature; the minimum asymptote parameter for sex j; the slope parameter for sex j; the inflection point (i.e., length at 50% matu- rity) parameter for sex j; and the maximum asymptote parameter for sex j. Spawning stock biomass is the sum of female and male spawning biomass (i.e., biomass of mature fish) for the G population and is the sum of female and dominant male spawning biomass for the TP scenario. For the AP population scenarios, SSB is the sum of female and dominant male spawning biomass, plus the spawning biomass of the given proportion (0%, 50%, or 100%) of secondary males. The lack of an accepted stock-recruitment model for black sea bass, combined with the desire to include the SSB sex ratio as a factor influencing recruitment, led to the development of an approach that involves re- cruitment bins. With this approach, depending on the value of SSB and the sex ratio (in numbers) within the spawning stock, recruitment abundance for year y {Ry) was randomly drawn (on the basis of a random num- ber seed) from 1 of 3 cumulative distribution functions (CDF) representing 3 productivity bins. Each CDF was created from a lognormal distribution generated by us- ing a mean recruitment and standard deviation from empirical data. As long as the proportion of females iPropF) and males (PropM) in the spawning stock remained above a sex ratio threshold (Hi) (i.e., as long as the sex ratio in the spawning stock was not overly skewed), recruit- ment was drawn according to the value of SSB in re- lation to 3 SSB breakpoints, Lmin; Bi, and B2, which delimit the productivity bins. Recruitment values as- sociated with Bmin - SSB < Bi represent the low pro- ductivity bin, recruitment values associated with Ri < SSB < B2 represent the medium productivity bin, and recruitment values associated with SSB > B2 represent the high productivity bin. If PropF or PropM dropped below Hi but remained above a minimum sex ratio threshold (H^in), recruitment was drawn from the low productivity bin. There was no recruitment (i.e., i?y=0) when SSB < Rmin or if either PropF or PropM dropped below Hmin- As mentioned above, recruits can enter the population either as females or secondary males. Re- spectively, if TO=10, the abundance of female and sec- ondary male recruits at length I at time t, RU\i and RSit, are calculated as 480 Fishery Bulletin 1 14(4) Table 1 Values and descriptions of parameters used in the length-based population model developed to study the response to exploi- tation by gonochoristic, atypical protogynous, and typical protogynous populations of black sea bass (Centropristis striata). Parameter Value Description Ri 15,000,000 Initial number of recruits (i.e., initial population) My (y =1-5) 0.01 Annual full natural mortality rate for yeary, initial years Afy(y=6-45) 0.40 Annual full natural mortality rate for yeary Fy(y=l-14) 0.00 Annual fishing mortality rate for yeary, initial years Fy{y>U) (0.00,0.10, . ..1.50) Annual fishing mortality rate for yeary a 27.09 Length (cm) of 50% probability of retention by the gear (L50) p 1.053 Slope of fishing selectivity curve 5; o ^min 690 Minimum spawning stock biomass (SSB; metric tons [t]) required for potential recruitment Bi 2,955 SSB Breakpoint 1 (t) B2 7,388 SSB Breakpoint 2 (t) ^m\n 0.02 Minimum sex ratio threshold (proportion female or male in spawning stock) Hi 0.10 Sex ratio threshold 1 (proportion female or male in spawning stock) Ou 0.000 Minimum asymptote parameter of female maturity ogive Oyi 6u 6.569 Slope (steepness) parameter of female maturity ogive Oui Cu 21.233 Inflection point parameter of female maturity ogive Oyi d\i 1.000 Maximum asymptote parameter of female maturity ogive Om ay 0.000 Minimum asymptote parameter of male maturity ogive Oyi bv 6.374 Slope (steepness) parameter of male maturity ogive Oyi Cv 18.506 Inflection point parameter of male maturity ogive Oyi 1.000 Maximum asymptote parameter of male maturity ogive Oyi 56.7 Asymptotic maximum length (cm) K 0.0585 Growth coefficient (per seasonal quarter) Qo 1.888 Theoretical age (in seasonal quarter units) at length 1=0 cm C 0.026 Oscillation amplitude coefficient for seasonal growth model Si -0.2794 Sine wave starting time coefficient (in seasonal quarter units) for the seasonal growth model Ciu = CiT -10.912 Weight-at-length coefficient for female (U) and transitional (T) stages V2V = C2T 2.9120 Weight-at-length exponent for female (U) and transitional (T) stages ClS = t'lD -10.954 Weight-at-length coefficient for secondary (S) and dominant (D) male stages = C2D 2.9094 Weight-at-length exponent for secondary (S) and dominant iD) male stages «t/lt = iJyPld - Pl), (12) RS\i = RyP\p\, (13) where p\ = the proportion of recruits at length /; and Pi = the proportion of male recruits at length 1. If m^lQ, then RU^ = RS^ = 0. Finally, weight at length for a fish of stage i and length I (W,)) was modeled with the allometric relationship; Wii = exp (Inluii) + t;2ilfi(^)). (14) where eii = the weight-at-length coefficient for stage i; V2i = the weight-at-length exponent for stage i; and Uii and f;2i are constants under the assumption of isometric growth. Parameter estimation and input values for models All parameter and input values were estimated on the basis of empirical data from the northern stock of black sea bass (an exploited stock) where appropriate. Estimated model parameters specifying M, selectivity, maturity, and weight at length were identical for all experimental populations, as were growth, initial popu- lation size, and recruitment inputs (Table 1). Transi- tion rates and size at transition were specific to each experimental population, resulting in differences in the proportion of male recruits and population sex ratio. For the first 5 years, My was set to 0.01 to initial- ize the simulated populations and for subsequent years was held constant and equal to 0.4 (Shepherd^). Ini- tially, Fy was 0.00, after which Fy was set to different levels as described later in the Simulations and sensi- tivity analyses section. Selectivity parameters were es- timated from assessment results (Shepherd^). Maturity ogives for females and males were defined on the basis of maturity data from the NEFSC spring survey cruise (1984-2013; «=3285), and weight-at-length parameters were estimated by using results from the NEFSC bot- tom trawl surveys conducted in spring and fall, 1993- 2013 (Shepherd^). ® Shepherd, G. R. 2009. Black sea bass 2009 stock assess- ment update. Northeast Fish. Sci. Cent. Ref. Doc. 09-16, 30 p. [Available at website, accessed May 2015.] Blaylock and Shepherd: Vulnerability of Centropristis striata to exploitation 481 Quarterly growth matrices were derived from aver- age length-at-age data from the NEFSC spring (1984- 2011, /z=3026), fall (1983-2010, ?z=:3381), and winter (1992-2007, n=3390) surveys, and from commercial catch data from a summer survey (2012, n=59). Quar- terly average length at age from all seasonal quarters were initially fitted (nonlinear regression, SAS 9.4^, SAS Institute Inc., Cary, NC) to a von Bertalanffy growth curve for length in quarter q, (15) Data were then fitted to a seasonal growth model by us- ing parameter values from the von Bertalanffy growth model as defined in Pitcher and MacDonald (1973): 4, = L^(l-e-^C, (16) where where K^—C sin 2TT(q — Si) 52 + K{q-qQ), (17) C = the amplitude of the oscillation around the average (nonseasonal) growth curve; and Si = the starting time for the sine wave. Expected quarterly growth by length was calculated for each centimeter from the sine wave growth model and formed the basis for deriving the quarterly growth ma- trices, which also were calibrated to empirical length distributions by quarter. Growth was fastest during the summer (quarter 3) and almost negligible during win- ter (quarter 1) — findings that are consistent with those from another study (Able and Hales, 1997). Initial recruitment in year 1 was 15 million indi- viduals, a value scaled to produce catch at equilibrium equivalent to empirical estimates at corresponding F values (Shepherd^). The length distribution of recruits was based on the average length distribution (at age zero), from NEFSC fall surveys, for lengths 4-11 cm for 2009-2013 (n=214). The 3 recruitment productiv- ity bins were structured to reflect the distribution of SSB and recruits in the most recent assessment of the northern stock of black sea bass (Shepherd^). We defined 3 overlapping lognormal distributions (1 for each bin) and randomly drew 500 values from each distribution to create a CDF for each productivity bin. The average (and standard deviation [SD]) of the 500 R values per bin were 6517.9 individuals (SD 3058.0) for the low recruitment bin, 13,912.1 individuals (SD 4057.0) for the medium recruitment bin, and 29,634.3 individuals (SD 9289.1) for the high recruitment bin. Sex ratio thresholds (if^in and Hi) were set to 0.02 and 0.10, respectively; these values were assumed reason- able because no empirical data were available. Input data specifying the proportion of male recruits and the transitions between the 4 different stages (i.e., female-transitional-secondary-dominant) differed among the 3 experimental populations. For the G case. * Mention of trade names or commercial companies is for iden- tification purposes only and does not imply endorsement by the National Marine Fisheries Service, NOAA. half of the recruits were males (pi=0.5; Fig. 2A) for a population with only females and males (i.e., transi- tion rate from female to transitional [6’^^im] and tran- sition rate from secondary to dominant male were equal to zero). For the protogynous populations, the sex ratio of recruits and transition rates were de- termined such that the sex ratios of the AP population would reflect empirical data for the northern stock of black sea bass from the NEFSC bottom trawl surveys (n=13,107) conducted in spring (1984-2013, ?i=4040), fall (1983-2013, n=5126), and winter (1992-2007, «=3941), and the sex ratio of the TP population would have a characteristic pattern with all female recruits and all fish transitioning to male by 65 cm (Fig. 2C). Transition rates were identified with F set at 0.35 to approximate the F that was likely occurring when the empirical data were collected. In the AP popula- tion, 23% of recruits were male and were equal to zero for lengths below 15 cm and followed a linear relationship (with slope 0.0039 and intercept -0.0492) for lengths 15 to 65 cm. The TP population had zero male recruits, and were zero for lengths below 12 cm, approximated a linear trend with slope 0.0108 and intercept -0.0228 for lengths 12-43 cm, and were con- stant at 0.43 for lengths 44-65 cm. Because of a lack of data to inform the split into secondary and dominant males in the protogynous populations, we assumed that half of males were dominant at 60 cm and set tran- sition rates accordingly. Transition rates from transi- tional to dominant male (S^^im) and from secondary to dominant male (0®°im) were held constant for all pro- togynous populations according to a linear relationship that was derived from the data, with slope 0.0109 and intercept -0.1522 for lengths 15 to 60 cm. A constant transition rate of 0.5 was applied for lengths greater than 60 cm. Simulations and sensitivity analyses For each experimental population, the model was run for 45 years with values of annual F ranging from 0.0 to 1.5 by increments of 0.1 (i.e., 16 constant F sce- narios). Each model configuration was run 100 times, with random draws of recruitment values (Ry). Results from year 45 were assumed to be representative of the populations at equilibrium under each configuration. For each scenario, the average and 95% confidence in- terval (Cl) for stock abundance, stock biomass, catch abundance, catch biomass, SSB, and R were calculated at each value of F. Stock results represent abundance on 1 May (m=5), and catch results represent the sum of catch over the entire year. Additional runs were performed to evaluate the sensitivity of the model results to capture selectivity, rates of sex change by females, proportion of dominant males, sex ratio thresholds of the spawning stock, and inclusion of the TP secondary males in SSB. For selec- tivity, the AP-50 configuration was run with values of ±10% (24.38 cm and 29.8 cm) for parameter a (L^q) of the selectivity curve 5] (Eq. 8). To assess the influence 482 Fishery Bulletin 114(4) A O) CQ 80% 100% O) CQ I 80% 0 §_ 60% 0) 40% _ro 1 20% D 0 0% 5 15 25 35 45 55 Length (cm TL) c 100% 0) D) CO ^ 80% 0) 5) 60% Q. > 40% 1 20% :□ O 0% F selectivity ^ S selectivity Figure 2 Cumulative percentages of life stages of black sea bass (Centro- pristis striata) by length (centimeters in total length [TL]) for (A) gonochoristic, (B) atypical protogynous, and (C) typical pro- togynous populations, which were simulated to study response to exploitation by using a length-based population model. Each shade of gray represents a different maturity state, immature (i) and mature (m), and life stage, including female (F), secondary male (S), dominant male (D), or simply male (M), in the gono- choristic case. Dashed lines represent fishing selectivity curves for each stage, as indicated in the key. Life stage characteristics were based on data collected, when available, from bottom trawl surveys conducted by the NOAA Northeast Fisheries Science Center during 1984-2013. 5 15 25 35 45 55 Length (cm TL) of female transition rates, the AP-50 and TP configura- tions were run with values of ±10% of 0^^ . We ran 2 sets of simulations on the protogynous populations to consider sensitivity of these populations to the pro- portion of males that were dominant, using transition rates that resulted in 25% and 75% males at 60 cm being dominant, respectively. Another set of sensitivity runs evaluated the effect of adding a third sex ratio threshold for the spawning stock (H2). This additional threshold acts as a penalty for moderately skewed sex ratios in the spawning stock by directing Ry to be drawn from the medium productivity bin if PropF or PropM is between Hi (where iTi=0.20) and H2, all population simulations were rerun with 172=0.40. Finally, the effect of spawning hy secondary males in the TP population was explored with runs allowing participation in spawning by 50% and 100% of mature second- ary males. Results Simulations All simulation results showed a similar pat- tern at equilibrium (year 45) of decreasing av- erage stock size (Fig. 3) with increasing F, and of catch increasing with F, reaching a maxi- mum, and then decreasing to zero as harvest- ing pressure rose (Fig. 4). However, the specific performance of each experimental population differed. A general pattern existed where the results for G and TP populations bounded the results from the 3 AP populations, the size of which declined in an order consistent with the percentage of contributions to spawning made by secondary males. The performance of the AP-0 population was closest to the TP scenario; the AP-lOO population generally did not differ from the G case, and the AP-50 population fell in between. The size of the TP population at equilibrium remained well below levels reached by the G case at the same F, with stock size (abundance and biomass) never exceeding 81% of the size of the G population and remaining below 50% of the size of the G population for F>0.2, before approaching zero at F=:1.0 (Fig. 3). The equilib- rium stock size of the AP-0 population started at almost the same size as that of the G popu- lation (94%) under no exploitation but dropped to levels similar to those of the TP population once F increased, even dropping below the size of the TP population for F=0.2 and F=0.3. The AP-50 equilibrium stock size was equivalent to that of the G population for low values of F but declined at a faster rate as F increased (Fig. 3). In contrast, the AP-lOO stock size was almost identical to that of the G population for all values of F. The difference in response of the AP populations stemmed from the increase in participation of sec- ondary males in spawning — an increase that guarded against abrupt stock size decline under exploitation. Blaylock and Shepherd: Vulnerability of Centropristis striata to exploitation 483 B CO m 03 E g Iq o o For example, at F=0.3, stock abun- dance for the AP-0 population m’-as 31% of the abundance of the G popu- lation, the abundance of the AP-50 population was 76% of the abundance of the G population, and the AP-lOO population was 98% of the abundance of the G population (Fig. 3A); the cor- responding stock biomass results were 23%, 82%, and 100%, respectively (Fig. 3B). Stock abundance approached zero at F=l.l for the AP-0 population and at ^=1.4 for the AP-50 population, and stock abundance of the AP-lOO popula- tion collapsed above F=1.5. Contributions of secondary males to spawning altered the F associated with maximum catch. Catch for the G population was highest at F=0.3 be- fore declining more than 65% by F=0.7 and approaching zero near F=1.5 (Fig. 4). Maximum yield in the TP configu- ration was reached at a lower F (0.2) with catch number equal to 42% of the maximum catch number for the G population (Fig. 4A) and with catch biomass equal to 45% of catch biomass of the G population (Fig. 4B). Catch and F at maximum, catch in the AP scenarios increased with contributions of secondary males to spawning. The AP-0 population reached a maximum catch abundance of 34% of maximum G catch at F=0.1 (i.e., lower than the maximum of the TP population), and the AP-50 and AP-lOO populations reached 81% and 99%, respectively, of maximum G catch at F=0.3 (Fig. 4A); corresponding catch biomass amounts for the 3 populations were 41%, 83%, and 100% (Fig. 4B). Analysis of trends in SSB and B (averages and 95% CIs of the 100 sim- ulation runs) versus F provided infor- mation in relation to recruitment and indicated which productivity bin each population was drawing from at equilibrium (year 45) at different values of F (Fig. 5). These results confirmed the pattern that, in general, results for the G and TP populations represented ex- tremes in population productivity; the results of the 3 AP populations, however, occurred in a consistent order in between those extremes. Although all populations be- gan in the high productivity recruitment bin (SSB^2) at F=:0.0, the progression to the end of recruitment as F increased and SSB fell below S^in differed across populations. The G and AP-lOO populations remained in the high productivity bin until F-0.3, at which point average recruitment declined to the medium and low productivity bins when F=:0.4-0.8 and approached zero as F reached 1.5. The AP-50 population functioned un- AP-100 AP-50 1.6 Figure 3 Equilibrium (year 45) (A) relative stock abundance and (B) relative stock biomass (measured in metric tons [t]) at values of fishing mortality (F) from 0.0 to 1.5 by 0.1 increments for 5 experimental populations of black sea bass (Centropristis striata): a gonochoristic (G) population (thick, solid line), atypical protogynous (AP) populations with 100% (AP-lOO; thin, solid line), 50% (AP-50; dashed line), and 0% (AP-0; dotted line) of mature secondary males contributing to spawning, and a typical protogy- nous population (TP; dashed-dotted line). Each line indicates the average stock size (from 100 runs performed at each value of F) in relation to the average stock size of the G population at F=Q.0; vertical bars repre- sent 95% confidence intervals. These results were obtained by using a length-based population model to study the response of these populations to exploitation. der high productivity until F reached 0.2 before fol- lowing a similar decline and approaching recruitment failure at F=:1.2. The TP and AP-0 scenarios stayed in the high productivity bin only if there was zero exploi- tation and SSB and R declined more rapidly than they did for the other populations as F increased, reaching recruitment failure by F=0.9. As seen with the results of stock size discussed pre- viously, the results for SSB and R for the AP-0 popula- tion were below those of the TP population when F=0.2 and F=0.3. With increasing exploitation, females quick- ly constituted more than 90% of the spawning stock abundance for the AP-0 population (Fig. 2B), such that PropM fell between the sex ratio thresholds (i7niin=0.02 and Hi=0.10), causing recruitment to drop to the low productivity bin (Fig 5B). In contrast, the TP popula- 484 Fishery Bulletin 1 14(4) tion had a higher proportion of males in the spawning stock (Fig. 2C); there- fore, the spawning stock sex ratio did not become overly skewed and did not trigger lower recruitment. Stock-recruitment curves indicate that the TP and AP-0 populations had a lower productivity relationship than that of the G, AP-lOO, and AP-50 popu- lations (Fig. 6). A lower line indicated that R was drawn more often from the low recruitment bin because of an overly skewed spawning stock sex ra- tio (i.e., less than 10% males or less than 10% females). Sensitivity analyses The sensitivity analysis on capture selectivity for the AP-50 population produced the expected results. A 10% decrease in length at capture caused 1) lower stock size and earlier stock collapse (average decrease of 27% for stock abundance and 31% for stock biomass, with both reaching zero by F=l.l), 2) a higher (about 10% in- crease) maximum catch that occurred at a lower F (0.2) but was followed by lower catch numbers, on average 36% (46% for catch biomass) below the levels of the base AP-50 population, and 3) a general reduction in recruit- ment (46% decrease in R on average). On the other hand, a 10% increase in length at capture generated the oppo- site results: 1) stock size increased by about 53% (66% for stock biomass) and did not collapse by F=1.5; 2) maximum catch was close to the results from the base AP-50 population and still oc- curred at 7^=0. 3, and catch numbers for values of F beyond where the max- imum occurred were on average 55% (83% for catch biomass) higher than the values for the base AP-50 population; and 3) i? in- creased by about 25%. Sensitivity to changes in of^ differed for the AP- 50 and TP populations. All results indicate that stock performance for the AP-50 population had very low sensitivity to a 10% change in of^^, closely matching the results for the base AP-50 population. In contrast, the TP population was more sensitive to change in O},^. A 10% decrease in sex change transitions resulted in slightly higher stock sizes (7% increase on average for stock number and stock biomass) and higher catch (9% average increase in catch numbers and biomass), and a 10% increase in sex change transitions resulted in slightly lower stock sizes (8% decrease on average for stock number and stock biomass) and lower catch (9% 1.6 B Figure 4 Equilibrium (year 45) (A) relative catch abundance and (B) relative catch biomass (measured in metric tons [t]) at values of fishing mortality (P) from 0.0 to 1.5 by 0.1 increments for 5 experimental populations of black sea bass {Centropristis striata): a gonochoristic (G) population (thick, solid line), atypical protogynous (AP) populations with 100% (AP-lOO; thin, solid line), 50% (AP-50; dashed line), and 0% (AP-0; dotted line) of mature secondary males contributing to spawning, and a typical protogy- nous population (TP; dashed-dotted line). Each line indicates the average catch size (from 100 runs performed at each value of F) in relation to the average catch size of the G population at F=0.0; vertical bars rep- resent 95% confidence intervals. These results were obtained by using a length-based population model to study the response of these populations to exploitation. average decrease in catch numbers and 8% average decrease for catch biomass); F at maximum catch re- mained the same as it was for the base AP-50 popula- tion. Finally, drawing from recruitment bins was not affected by a 10% change in the sex transition rates. Modifying the proportion of dominant versus sec- ondary males in the protogynous populations had a greater effect on the AP-0 and TP populations than on the AP-50 configuration, and it had no effect on the AP-lOO scenario because all males already contributed to spawning in this population. For the AP-50 popula- tion, reducing the proportion of dominant males to 25% at 60 cm led to an average decrease of just over 6% for stock abundance and catch at F’>0.2 and to average drops of 8% in SSB and of 5% in R. With transition Blaylock and Shepherd: Vulnerability of Centropristis striata to exploitation 485 A B Figure 5 Equilibrium (year 45) (A) relative spawning stock biomass (SSB) and (B) relative number of recruits {R) at values of fishing mortality (F) from 0.0 to 1.5 by 0.1 increments for 5 experimental populations of black sea bass (Centropristis striata): a gonochoristic (G) population (thick, solid line), atypical protogynous (AP) populations with 100% (AP-lOO; thin, solid line), 50% (AP-50; dashed line), and 0% (AP-0; dotted line) of ma- ture secondary males contributing to spawning, and a typical protogynous population (TP; dashed-dotted line). Each line indicates the average SSB or R (from 100 runs performed at each value of F) in relation to the average SSB or R of the G population at F=0.0; vertical bars represent 95% confidence intervals. The gray reference lines indicate the position of the SSB breakpoints (Bmin> ^2) and corresponding J? delimiting the recruitment productivity bins. These results were obtained by using a length-based population model developed to study the response of these populations to exploitation. rates set for 75% of males being dominant at 60 cm, the population experienced an average increase of 4% in stock abundance and catch and a rise in SSB and i? of about 5%. In contrast, the 2 populations with no secondary- males contributing to spawning were more sensitive to a change in proportion of dominant males. With transition rates set for 25% of males being dominant at 60 cm, stock size and catch for the AP-0 popula- tion dropped by around 65% for F values below 0.3 and by just above 50% for F>0.3. The TP population stock size and catch dropped by about 40% for F val- ues below 0.3 and by close to 65% for F>0.3. Average SSB for the AP-0 population decreased by 58%, and R declined by 52%; corresponding values for the TP configuration showed an average 62% decrease in SSB and 55% decrease in R. The increase in proportion of domi- nant males for the AP-0 population re- sulted in an increase in stock size and catch up to 70% at F values below 0.3 and in an increase of around 45% for i^O.3. The increase for the TP popula- tion was just over 30% for F<0.3 and over 50% for P'SO.S. On average, SSB increased by 60% and R by 55% for both the AP-0 and TP populations. In general, the performance of the AP-0 and TP populations moved closer to that of the base G population with ad- ditional dominant males. Sensitivity analyses that consid- ered the effect of an alternative re- cruitment process indicated that most protogynous scenarios were sensitive to the addition of a third sex ratio threshold for the spawning stock, Il2, where recruitment was drawn from the medium productivity bin if PropF or PropM was between ifi=0.10 and iJ2=0.40. Because sex ratio in the G population was fixed at 1:1, this popu- lation was not affected by any of the sex ratio thresholds. However, all the protogynous populations experienced a significant drop in recruitment with the addition of H2, especially at lower F values (^<0.7), where R decreased by an average of 32% for the AP-lOO population, 28% for the AP-50 popu- lation, and 16% for the AP-0 and TP populations. At higher values of F, re- cruitment declined as well, but it did so less dramatically (from 0% to 19%) because recruitment was already low as a result of low SSB at those levels of exploitation. Recruitment for the protogynous scenarios was limited to the low and medium productivity bins, indicating that the spawning stock sex ratio of these populations was always skewed (Fig. 2), specifically that PropM was always less than 172=0.4. In parallel to the decrease in recruitment, all other measures of stock performance (SSB, stock size, and catch) decreased as well. In general, all results for the AP populations moved closer to the results for the TP population, while continuing to follow the order that the TP population was least productive, followed by the AP-0, AP-50, and AP-lOO populations, as seen in the results from the runs of the base model. In the final sensitivity analysis, the effect of spawn- ing by secondary males in the TP population was con- sidered. Allowing all secondary males to spawn brought the TP population performance close to the performance of the base G and AP-lOO populations, although it was 486 Fishery Bulletin 1 14(4) Figure 6 Equilibrium (year 45) relationship between stock and recruitment represent- ed by the relative number of recruits (R) versus the relative spawning stock biomass (SSB) for 5 experimental populations of black sea bass {Centropris- tis striata): a gonochoristic (G) population (thick, solid line), atypical pro- togynous (AP) populations with 100% (AP-lOO; thin, solid line), 50% (AP-50; dashed line), and 0% (AP-0; dotted line) of mature secondary males contrib- uting to spawning, and a typical protogynous population (TP; dashed-dotted line). Each line indicates the average R and SSB (from 100 runs performed at each value of F) in relation to the average R and SSB of the G population at F=0.0; vertical and horizontal bars represent 95% confidence intervals for relative R and relative SSB, respectively. The gray reference lines indicate the position of the SSB breakpoints (Bmin. ^i. and B^) that define the differ- ent recruitment productivity bins. These results were obtained by using a length-based population model to study the response of these populations to exploitation. still slightly (about 3%) below it. Similarly, a 50% rate of spawning by secondary males in the TP population brought stock performance closer to that of the AP-50 population; however, levels for the TP population were 12-17% lower. Discussion We developed a length-based model to simulate the dy- namics of a protogynous fish population and examined the resilience of an atypical protogynous population on the basis of the northern stock of black sea bass. The model allowed evaluation of the impact of exploitation on a population’s resilience and productivity across a range of life history attributes. The life history of the northern stock of black sea bass is atypical for a pro- togynous species because the sex ratio at length is not as skewed toward females at small sizes (<30 cm) nor toward males at large sizes (>45 cm), as would be ex- pected in a typical protogynous population. Previously published findings (Alonzo and Mangel, 2004; Heppell et ah, 2006; Hamilton et al., 2007) indicate that protog- ynous stocks are more susceptible to overexploitation than are equivalent gonochoristic stocks. Our results support that conclusion, in particular for typical pro- togynous stocks, which have highly skewed sex ratios. However, we found that atypical protogynous life histo- ries, for which a higher proportion of the small fish are males, may provide added resiliency to exploitation in comparison with typical protogynous species for which all small fish are females and males arise only as a result of subsequent sex change by females. The differ- ent transition rates that led to different proportions of females, secondary males, and dominant males in the 2 types of populations resulted in different SSB and SSB sex ratios and ultimately in higher recruitment for the AP population. Moreover, we found that the ability of the AP population to persist under exploita- tion increases as a function of contribution of second- ary males to spawning. The different response to exploitation by the G, AP, and TP populations can be explained by considering which individuals contribute to SSB because the com- position of the spawning stock defines recruitment and ultimately population persistence. In the G case, we as- sumed that all mature males and females contributed to spawning, whereas in the protogynous cases, SSB was the sum of females and the subset of dominant males and had a range of possible contributions from secondary males for the AP populations. Size-selective exploitation primarily targeting large individuals de- creased SSB, skewed protogynous population spawning stock sex ratios (Fig. 2, B and C), and made the popula- tions more likely than the G population to experience decreased fertilization because of sperm-limitation. Within the protogynous scenarios, different sex ratios of mature fish also affected the size of SSB and thus Blaylock and Shepherd: Vulnerability of Centropristis striata to exploitation 487 R, particularly if a substantial portion of the mature males were nonspawning secondary males. For exam- ple, even in the absence of fishing, the TP population was less productive than the other populations (Fig. 5) because of a smaller SSB that resulted in recruitment occasionally being drawn from the medium recruit- ment bin, whereas all other populations had SSB high enough to have R drawn consistently from the high productivity bin. Our results indicate that as the contribution to spawning by secondary males in the AP population in- creased, SSB approached that of the G species and the population became more resilient to exploitation. The addition of secondary males to the spawning stock not only increased SSB but also shifted the sex ratio and lowered the risk to the population of being sperm-lim- ited. In other words, the more secondary males partici- pated in spawning, the more an AP population would perform as if it were gonochoristic because the spawn- ing stock sex ratio approached 1:1. Consequently, the greater the contribution of secondary males, the more resilient the AP population was to fishing. Sensitivity analyses indicated that the same was true for the TP population; however, for equivalent scenarios (i.e., both having 50% or 100% of secondary males spawn) the AP population was still more resilient than the TP popu- lation, although the magnitude of the difference de- pended on the proportion of secondary males included in SSB. The presence of male recruits in the AP population was also an important characteristic. It not only damp- ened the effect of size-selective exploitation by chang- ing the sex ratios at length but also provided a con- stant supply of males to the population from a source other than postmaturational sex change. Although transition rates were held constant in our model, previ- ous research (Huntsman and Schaaf, 1994; Alonzo and Mangel, 2004; Taylor, 2013) suggests that sex change is likely to be socially controlled and transition rates may increase in response to removal of males as a way to maintain stable sex ratios. Under severe exploita- tion, such compensatory sex change would be the only avenue for a typical protogynous population to avoid extremely low abundance of males. In contrast, an atypical protogynous life history, with males available beginning at recruitment, provided an additional buffer against overharvesting of dominant males. Our sensi- tivity analyses support this interpretation with results indicating that abundance and SSB in an AP configu- ration were more robust than abundance and SSB in the TP case in response to variations in the rate of sex change. In addition, the proportion of mature domi- nant males played a significant role because a higher fraction of dominant males brought sex ratios closer to a gonochoristic situation and rendered the population more resilient to fishing. The life history characteristics of the AP population increased robustness in response to exploitation, and we expect they would also improve resistance to chang- es in capture selectivity. Results from sensitivity analy- ses on the selection pattern indicated that a reduction in size at selection had some influence on abundance for the AP-50 population. Because any changes to se- lectivity would apply across all scenarios, we would expect similar results in the other populations if we tested them. However, as seen with other factors, the TP population was generally more vulnerable to a drop in recruitment because of a decrease in SSB or an over- ly skewed SSB sex ratio; therefore, reducing the size at selectivity would likely affect this population more severely. The sensitivity tests were not performed in a multiplicative manner, and, as a result, it is possible that a response to changes to exploitation with changes to transition rates could alter the outcome. In addition, the effect of dome-shaped selectivity was not explored here because of the lack of empirical evidence support- ing it, but the influence of this type of selectivity could be explored in the future. The sex ratio at maturity, the presence of mature secondary males, and large mature females all worked toward making the AP population more stable than a population with a typical protogynous life history. Black sea bass north of Cape Hatteras undergo sea- sonal migrations to the edge of the continental shelf — movements that in some cases cover long distances (Musick and Mercer, 1977; Moser and Shepherd, 2009). Such behavior is generally not seen in tropical or sub- tropical species for which hermaphroditism is most common. The advantages of hermaphroditism evolved in natural systems where there is a selective advan- tage of size differences between sexes (Ghiselin, 1969; Warner, 1975; Kazancioglu and Alonzo, 2010). The selective advantage gained by beginning life as female then switching to male as needed depends on feedback about the existing sex ratio within the repro- ductive community; too many large males would reduce any reproductive size advantage and only increase sperm competition (Petersen, 1991). However, in a mi- gratory species adopting this life history, the feedback loop may be seasonally interrupted. If individual fish stray from the group with the sex ratio that elicited the sex change, the advantage of being a large male could be lost. In the northern stock of black sea bass, results from a tagging study indicate that the farther the migration, the lower the probability of returning to the original spawning aggregation (Moser and Shep- herd, 2009). Consequently, the possibility of returning to a group with large males already present would re- duce the advantage gained by switching sex before the migration. Under such conditions, gonochoristic traits may provide more reproductive stability than typical protogyny. Although our model inputs were based on empirical data as much as possible, we had to make assumptions about several key population processes. The relation- ship between SSB, spawning stock sex ratio, and sub- sequent recruitment for black sea bass is poorly un- derstood, and productivity might be highly influenced by environmental factors (Able and Hales, 1997). Given the absence of an accepted stock-recruitment model. 488 Fishery Bulletin 114(4) we used a binned recruitment approach on the basis of empirical data, recognizing that the intent of our model was to evaluate the contrast in response between the population life histories rather than to estimate abso- lute abundances. We included spawning stock sex ratio as bounds within the stock-recruitment relationship for cases where the sex ratio became extremely skewed. Even though our approach was rather rudimentary, it allowed us to highlight the importance of incorporat- ing spawning stock sex ratio in addition to SSB in the determination of recruitment for protogynous stocks. If a small proportion of males can successfully fertilize all mature females, the population could remain fully productive even with a skewed sex ratio. In contrast, if the optimal fertilization sex ratio is close to 1:1, re- cruitment would drop if the spawning stock sex ratio became more skewed toward females, as seen in our sensitivity analyses, which showed that recruitment was sensitive to changes in the sex ratio thresholds for the spawning stock. These findings also emphasized the need for more information on spawning behavior of black sea bass and of protogynous hermaphrodite species in general. Specifically, data on male-to-female ratios within a spawning event would help to improve understanding of when a population might be suscep- tible to sperm limitation or egg limitation and to evalu- ate the effectiveness of using the combined male and female SSB for populations with skewed sex ratios. Finally, it is unknown to what degree compensatory re- cruitment may occur and to what extent such recruit- ment might alter population resilience under exploita- tion. Given the lack of data to inform this process, we chose not to include it in our model; however it could warrant further exploration. Transition rates were determined to produce sex ra- tios at length closely resembling those of the northern stock of black sea bass, but actual transition rates, in- cluding their timing and the split between secondary and dominant males, are unknown. We assumed that sex change occurred quickly and that rates of transi- tion from secondary to dominant male were the same for all protogynous populations because our main con- cerns were to investigate the implication of the atypi- cal life history and the contribution of secondary males to spawning in the north. The duration of the transi- tion made little difference in our model as long as it did not overlap with the spawning period; as long as all transitions were completed before SSB was calculated, there was little influence from this factor. However, it is possible that true transition probabilities differ sig- nificantly from what we assumed and that transition rates may differ between an atypical versus typical hermaphrodite population. Moreover, we did not incor- porate density dependence into our model, although density-dependent transition rates (compensatory sex change) may have a significant effect on adjusting sex ratios under high fishing pressure (Ellis and Powers, 2012). Molloy et al. (2007) found that flexibility in size at sex change may make protogynous populations as resilient to fishing as gonochoristic populations; there- fore, this feature should be included in future research, if possible. We did not model social behavior, competition among males, or density-dependent changes in fertilization success for each life stage in part because of a lack of empirical data. The lack of secondary sex characteris- tics in secondary males would indicate that they are opportunistic spawners rather than direct competitors of dominant males. Although it is unlikely that males participating in spawning as secondary males would be as effective as the dominant males from a behavioral perspective, there is evidence that relative gonad size of secondary males may be larger than that of domi- nant males, resulting in increased likelihood of fertil- ization success (Knapp and Carlisle, 2011). Finally, other possible factors, such as density- or size-dependent M, movement, or spatial patterns, were not incorporated into the model. We also assumed that growth was constant across sexes; however, if sexual dimorphic growth favoring males exists, one would expect the G and AP scenarios to become even more aligned. Despite the limitations of the data and the model, the results of this research provide new insight into the response of protogynous species to exploitation. In particular, atypical sex ratios at length and possible contribution of secondary males to spawning have the potential to significantly increase the resilience of a protogynous stock to fishing. This study indicates that stocks exhibiting atypical protogynous characteristics, such as the northern stock of black sea bass, may be more resilient to exploitation than typical protogynous hermaphrodites characterized by dominant males mo- nopolizing spawning opportunities. In addition, our re- sults highlight the need for a better understanding of factors that govern key processes, such as sex change, secondary male spawning, and recruitment, in pro- togynous species in general and in black sea bass in particular. Acknowledgments The authors thank A. Seaver for his time program- ming and building a user interface for the model, E. Brooks for her helpful advice concerning the modeling approach, M. Wuenschel for fruitful discussions about reproduction and life history of black sea bass, J. Evans for help with Python, and E. Robillard and J. Dayton for helping us with questions regarding aging of black sea bass. We also are grateful to our reviewers for their constructive comments. Literature cited Able, K. W, and L. S. Hales Jr. 1997. Movements of juvenile black sea bass Centropristis striata (Linnaeus) in a southern New Jersey estuary. J. Exp. Mar. Biol. Ecol. 213:153-167. Blaylock and Shepherd: Vulnerability of Centropristis striata to exploitation 489 Allsop, D. J., and S. A. West. 2004. Sex-ratio evolution in sex changing animals. Evo- lution 58:1019-1027. Alonzo, S. H., and M. Mangel. 2004. The effects of size-selective fisheries on the stock dynamics of and sperm-limitation in sex-changing fish. Fish. Bull. 102:1-13. Alonzo, S. H., T. Ish, M. Key, A. D. MacCall, and M. Mangel. 2008. The importance of incorporating protogynous sex change into stock assessments. Bull. Mar. Sci. 83:163- 179. Bowen, B. W., and J. C. Avise. 1990. Genetic structure of Atlantic and Gulf of Mexico populations of sea bass, menhaden, and sturgeon: influ- ence of zoogeographic factors and life-history patterns. Mar. Biol. 107:371-381. Brander, K. 2003. What kinds of fish stock predictions do we need and what kinds of information will help us to make better predictions? Sci. Mar. 67 (suppl. l):21-33. Brooks, E. N., K. W. Shertzer, T. Gedamke, and D. S. Vaughan. 2007. Stock assessment of protogynous fish: evaluating measures of spawning biomass used to estimate biologi- cal reference points. Fish. Bull. 106:12-23. Collette, B. B., and G. Klein-MacPhee. 2002. Bigelow and Schroeder’s fishes of the Gulf of Maine, 3'‘‘* ed., 748 p. Smithsonian Institution Press, Washing- ton, DC. Drohan, A. R, J. P. Manderson, and D. B. Packer. 2007. Essential fish habitat source document: black sea bass, Centropristis striata, life history and habitat char- acteristics, 2"'* ed. NOAA Tech. Memo. NMFS-NE-200, 68 p. Ellis, R. D., and J. E. Powers. 2012. Gag grouper, marine reserves, and density-de- pendent sex change in the Gulf of Mexico. Fish. Res. 115-116:89-98. Ghiselin, M. T. 1969. The evolution of hermaphroditism among animals. Q. Rev. Biol. 44:189-208. Haddon, M. 2001. Modelling and quantitative methods in fisheries, 424 p. Chapman & Hall/CRC Press, Boca Raton, FL. Hamilton, S. L., J. E. Caselle, J. D. Standish, D. M. Schroeder, M. S. Love, J. A. Rosales-Casian, and O. Sosa-Nishizaki. 2007. Size-selective harvesting alters life histories of a temperate sex-changing fish. Ecol. Appl. 17:2268-2280. Heppell, S. S., S. A. Heppell, F. C. Coleman, and C. C. Koenig. 2006. Models to compare management options for a pro- togynous fish. Ecol. Appl. 16:238-249. Huntsman, G. R, and W. E. Schaaf 1994. Simulation of the impact of fishing on reproduction of a protogynous grouper, the graysby. North Am. J. Fish. Manage. 14:41-52. Kazancioglu, E., and S. H. Alonzo. 2010. A comparative analysis of sex change in Labridae supports the size advantage hypothesis. Evolution 64:2254-2264. Knapp, R., and S. L. Carlisle. 2011. Testicular function and hormonal regulation in fishes. In Hormones and reproduction of vertebrates, vol. 1: fish- es (D. O. Norris and K. H. Lopez, eds.), p. 43-63. Aca- demic Press, Burlington, MA. Lavenda, N. 1949. Sexual difference and normal protogynous her- maphroditism in the Atlantic sea bass, Centropristes striatus. Copeia 3:185-194. McCartney, M. A., M. L. Burton, and T. G. Lima. 2013. Mitochondrial DNA differentiation between popu- lations of black sea bass {Centropristis striata) across Cape Hatteras, North Carolina (USA). J. Biogeogr. 40:1386-1398. Molloy, P. R, N. B. Goodwin, 1. M. Cote, M. J. G. Gage, and J. D. Reynolds. 2007. Predicting the effects of exploitation on male-first sex-changing fish. Anim. Conserv. 10:30-38. Moser, J., and G. R. Shepherd. 2009. Seasonal distribution and movement of black sea bass {Centropristis striata) in the northwest Atlantic as determined from a mark-recapture experiment. J. Northwest Atl. Fish. Sci. 40:17-28. Munday, P. L., P. M. Buston, and R. R. Warner. 2006. Diversity and flexibility of sex-change strategies in animals. Trends Ecol. Evol. 21:89-95. Musick, J. A., and L. P. Mercer. 1977. Seasonal distribution of black sea bass, Centropris- tis striata, in the Mid-Atlantic Bight with comments on the ecology and fisheries of the species. Trans. Am. Fish. Soc. 106:12-25. Petersen, C. W. 1991. Sex allocation in hermaphroditic sea basses. Am. Nat. 138:650-667. Pitcher, T. J. (ed.) 1993. Behaviour of teleost fishes, 2"'* ed., 740 p. Chap- man & Hall, London. Pitcher, T. J., and P. D. M. MacDonald. 1973. Two models for seasonal growth in fishes. J. Appl. Ecol. 10:599-606. Quinn, T. J., II, and R. B. Deriso. 1999. Quantitative fish dynamics, 560 p. Oxford Univ. Press, New York. Roy, E. M, J. M. Quattro, and T. W. Greig. 2012. Genetic management of black sea bass: influence of biogeographic barriers on population structure. Mar. Coast. Fish. 4:391-402. Shepherd, G. R., and M. Terceiro. 1994. The summer flounder, scup, and black sea bass fish- ery of the Middle Atlantic Bight and Southern New Eng- land waters. NOAA Tech. Rep. NMFS-122, 13 p. Tabors ky, M. 1994. Sneakers, satellites, and helpers: parasitic and co- operative behavior in fish reproduction. Adv. Study Be- hav. 23:1-100. Taylor, B. M. 2013. Drivers of protogynous sex change differ across spa- tial scales. Proc. R. Soc., B 281:20132423. Warner, R. R. 1975. The adaptive significance of sequential hermaphro- ditism in animals. Am. Nat. 109:61-82. Young, B., D. V. Conti, and M. D. Dean. 2013. Sneaker “jack” males outcompete dominant “hook- nose” males under sperm competition in Chinook salmon (Oncorhynchus tshawytscha). Ecol. Evol. 3:4987-4997. 490 NOAA National Marine Fisheries Service Abstract — Age, growth, and monthly reproductive characteristics were determined for the orange-spotted grouper (Epinephelus coioides) from northern Oman. This species is char- acterized by a prevalence of females (1-11 years old), and males make up 6.5% of the total sample. Growth pa- rameters indicate a typical pattern for groupers with a low growth co- efficient (if=0.135). The trajectory of the von Bertalanffy growth function was almost linear with no evidence of asymptotic growth. Estimates of mortality revealed a low natural mortality of 0.14/year but a high fishing mortality of 0.59/year. More alarming was the high rate of exploi- tation (0.81/year), considered unsus- tainable for a slow-growing grouper. The population off southern Oman is diandric protogynous, and sex change takes place between 449 and 748 mm in total length (TL) or over a period of 4-8 years. The gonadoso- matic index for females showed a short spawning season from March through May, although -30% of fe- males were ripe for 7 months of the year. Size and age at 50% maturity for females was estimated to be 580 mm TL and 4 years, respectively. We suggest that substantial changes in the management of this species will be vital in sustaining viable popula- tions of orange-spotted grouper and other species of Epinephelidae with- in Oman. Manuscript submitted 14 July 2015. Manuscript accepted 21 July 2016. Fish. Bull. 114:490-502 (2016). Online publication date: 13 Sept. 2016. doi: 10.7755/FB.114.4.10 The views and opinions expressed or implied in this article are those of the author (or authors) and do not necessarily reflect the position of the National Marine Fisheries Service, NOAA. Fishery Bulletin established 1881 ■30 m) waters (Cheat and Robertson, 2002; Pears et al., 2006; Wakefield et al., 2013, 2015). Stocks that have such traits present challenges for fisheries managers, especially in developing countries where life history informa- tion is rarely collected (Polunin et al., 1996), mostly because of a lack of funding. The presence of these traits may also explain why conventional fisheries management practices have failed, even for sectors that are regu- lated (Coleman et al., 2000; Sadovy de Mitcheson et al., 2013). The epinepheline serranids, or groupers, form an important taxo- nomic group both from a commercial and ecological perspective (Sluka et al., 2001; Sadovy de Mitcheson et al., 2013). However, a suite of life history strategies and certain be- havioral traits make most species of grouper susceptible to overfishing. Life-history strategies include slow growth and late-onset reproduction (Manooch, 1987; Pears et al., 2006), Mcliwain et al.: Demographic profile of Epinephelus coloides in northern Oman 491 long life spans (Manooch, 1987; Bullock et al., 1992; Grandcourt, 2005; Wakefield et al., 2015), complex and variable reproductive patterns (Shapiro, 1987; Fen- nessy and Sadovy, 2002; Rhodes and Sadovy, 2002) in- cluding sex change, while traits include a propensity to aggregate during spawning (Colin, 1992; Sadovy et al., 1994; Samoilys and Squire, 1994; Robinson et al., 2015; Tuz-Sulub and Brule 2015), and a high degree of site fidelity (Zeller and Russ, 1998; Starr et al., 2007; Luckhurst, 2008). Frequently published low values of natural mor- tality (M) for grouper indicate that harvesting levels should be set at less than 10% of the total biomass (Walters and Pearse, 1996; Coleman et al., 2000). How- ever, low rates of M coupled with commonly cited low rates of growth can also lead to “growth overfishing” with resultant truncated size-distributions as fish are unable to grow to their maximum sizes (e.g. Beverton and Holt 1957). It has also been suggested that popula- tions in which sex change is predominately female to male, under a scenario of increased fishing mortality (F), lose reproductive capacity in 2 ways: through selec- tive removal of larger, mostly male individuals that in turn cause sperm limitation (Coleman et al., 1996) and a decrease in the size of females that in turn reduces total fecundity (Sadovy, 1996; Adams et al., 2000; Sa- dovy de Mitcheson and Liu, 2008). Hermaphroditism in groupers can be expressed in 1 of 2 ways: as simultaneous, when individuals are ca- pable of reproducing as both male and female, or se- quential (the most common mode), when sex change is made from female to male (protogyny) (Smith, 1965; Shapiro, 1987; DeMartini et al., 2011; Wakefield et al., 2015). Protogynous populations can be further char- acterized on the basis of the sexual pathway of male development. Monandric protogyny occurs when males develop only from females (secondary males) through a transitional phase, whereas diandric protogyny in- volves 2 pathways for male development: directly from juveniles (primary males) or through sex change from mature females (Reinboth, 1967). To date, most grou- per investigated in detail and that belong to the genus Epinephelus have been confirmed to be monandric her- maphrodites (Sadovy et al., 1994; Brule et al., 2000; Pears et al., 2006; DeMartini et al., 2011; Wakefield et al., 2015). However, increasing evidence indicates that some species are diandric hermaphrodites, such as the catface grouper {Epinephelus andersoni) (Fennessy and Sadovy, 2002), and Sadovy and Colin (1995) confirmed gonochorism, with the potential for protogynous sex change, in the Nassau grouper {Epinephelus striatus). Clearly, the genus shows sexual plasticity in reproduc- tive development, which has a bearing on population persistence and fisheries management. In Oman, grouper, or hamoor, are considered one of the most economically important groups of finfish in the artisanal fishery and are heavily targeted by fishermen with a variety of fishing gear (Siddeek et al., 1999). One species, the orange-spotted grouper {Epinephelus coioides) occurs primarily in northern Oman along the Musandam Peninsula, inhabiting deep rocky reefs to depths of 100 m (Randall, 1995). Fishermen that target this species use semicircular wire basket traps with, a radius of 2. 5-3. 5 m, and 35-mm wire-mesh. Traps are deployed either from fiberglass boats of 4-10 m in over- all length (and an average of 28 traps were deployed per trip at an average water depth of 70 m) or from large wooden dhows <25 m overall length (and an av- erage of 75 traps were deployed per trip at an average water depth of 150 m) (senior author, unpubl. data). To date, there are no restrictions on fishing effort for this fishery, and therefore no limits on the numbers or size of orange-spotted grouper landed by fishermen. The official fisheries statistics reveal that catches of grou- per from Musandam have remained steady since 1993 at an average of 370 metric tons (t) per year (MAF^). This catch represents 10.5% of the total landings of grouper for Oman from 1993 to 2009 (MAF^). On the basis of the total landings of grouper in Musandam, the orange-spotted grouper is the thirdmost important species of grouper by numbers and secondmost impor- tant by weight (senior author, unpubl. data). Despite the importance of orange-spotted grouper to both the southeast Asian aquacultural industry and the trade of live reef fish (Grandcourt, 2012), little de- mographic or reproductive information exists on wild stocks from throughout its geographic range (but see Grandcourt et al., 2005, 2009), with recent calls by the International Union for Conservation of Nature and Natural Resources for further work that would ex- amine the demography and life history of this species (Cornish and Harmelin-Vivien, 2004). The earliest re- productive study produced some evidence of sex change through social control, but it did so for only a few in- dividuals (n=6) kept in captivity (Quinitio et al., 1997). More recently, Grandcourt et al. (2005, 2009), focusing on populations within the southern Arabian Gulf, and Liu and Sadovy de Mitcheson (2009), through a con- trolled experiment of hatchery-reared juvenile orange- spotted grouper, confirmed diandric protogyny. The lack of basic biological data on orange-spotted grouper from the Oman region, the recent anecdotal evidence of declines in landings, and concerns over the spread of the trade of live reef fish into the west Indian Ocean prompted the work described here to determine the demographic makeup of orange-spotted grouper in Oman and to characterize the sexuality and spawning times of this species.. Specifically, we set out to de- scribe the age structure, growth parameters, and mor- tality schedules of the population from northern Oman. In light of the heavy fishing pressure on this species both in Oman and neighboring countries (Grandcourt et al., 2005, 2009), we also discuss potential manage- ment strategies that are appropriate for hermaphro- ditic populations. 1 MAF (Ministry of Agriculture and Fisheries). 2010. Fish- eries statistics book 2009, 130 p. [Available from Omani Fisheries Statistic Department, Ministry of Agriculture and Fisheries, P.O. Box 467, Muscat, Sultanate of Oman.] 492 Fishery Bulletin 1 14(4) Materials and methods Sampling Between May 2004 and August 2005, 214 orange-spotted grouper were pur- chased at monthly intervals either di- rectly from fishermen or agents at the fish landing site at Dibba, the only landing site on the east coast of the Musandam Peninsula that extends into the Strait of Hormuz at the mouth of the Arabian Gulf (Fig. 1). Sampling took place during the first week of each month. Fish were kept on ice until they were processed, and all fish were mea- sured to the nearest millimeter in total length (TL) and weighed for total wet weight to the nearest gram. An addi- tional 156 individual orange-spotted grouper were weighed and measured at the landing site and included in the length-weight relationship. Because of the difficulty in staging nonripe gonads of grouper macroscopically (Fennessy and Sadovy, 2002), all gonads, ovaries and testes, were removed and weighed to the nearest 0.01 g and preserved in 10% formalin solution for later his- tological preparation. Gutted whole weights (GWs) of fish were also taken. Sagittal otoliths were removed, cleaned in distilled water, and stored dry for subsequent age determination. Otolith processing and validation of otolith age For microscopic examination, one sagittal otolith was weighed (to the nearest 0.001 g), then set in Crystal- bond^ resin (Aremco Products Inc., Valley Cottage, NY) on the edge of a glass microscope slide. A modified grinding wheel, with P600 wet and dry grit paper, was used to grind the otolith as close to the nucleus as pos- sible (Cheat et ah, 2003). The otolith half was then re- positioned in the middle of the slide, with the polished side facedown, and again ground as close to the nu- cleus as possible. Sectioned otoliths were left on a hot plate at 60°C for 1-2 h until dark brown in color. The age of each individual was determined by counting the annual growth increments under a microscope, with transmitted light at 15-25x magnification. All otoliths were read twice by the senior author using the double- blind method (Russ et ah, 1998). If the 2 readings dis- agreed by more than one increment, the otolith was read a third time. The otolith was eliminated from the analysis if the third reading differed as well. The date 50“N 55”N 2 Mention of trade names or commercial companies is for iden- tification purposes only and does not imply endorsement by the authors or the National Marine Fisheries Service, NOAA. Figure 1 Map of the major landing site (Dibba) where samples of orange-spotted grouper (Epinephelus coioides) were collected between May 2004 and Au- gust 2005 for this study. The black area denotes the fishing grounds that support the fishery for orange-spotted grouper in northern Oman. of May 1 was assigned as the nominal hatching date for each individual on the basis of the peak spawning period between April and May for orange-spotted grou- per from Musandam (Grandcourt et al., 2005, 2009). Either marginal zone or edge analysis was used to validate the annual deposition of opaque zones on the otoliths of orange-spotted grouper by inferring the time of year that the increments were formed (Radebe et al., 2002). The otolith margin was recorded as either opaque or translucent for 91% of individuals. To de- termine the time when deposition took place, the fre- quency of otoliths with opaque margins was plotted by month. Determination of sex and maturity stages After preservation, gonads were subject to standard histological preparations and examination to assign a sexual and maturity stage to individual fish (West, 1990) . A thin, transverse section was taken from the central region of one gonad lobe, dehydrated in a Shan- don Citadell 1000 processor (Thermo Fisher Scientific Inc., Waltham, MA), embedded in wax, sectioned at 5-7 pm, mounted on a slide, and stained with haematoxylin and eosin stains. Each gonad was assigned to both a maturity stage (Sadovy and Colin, 1995; Adams, 2002; Fennessy and Sadovy, 2002; Pears et al., 2007) from Mcliwain et al.: Demographic profile of Epinephelus coloides in northern Oman 493 one of the following 8 classifications: 1) undetermined inactive female; 2) immature female; 3) mature, rest- ing or inactive female; 4) mature, active female; 5) spent female; 6) transitional individual; 7) immature male; 8) mature, active male. Ovaries were classified by the presence of the most advanced oocyte, regardless of its abundance (West, 1990). Maturity schedules for female fish were calculated by plotting the percentage of frequency of both ma- ture, active and mature, inactive females by 3-mm size classes and age groups for the entire sampling period. Females whose spawning history could not be determined (undetermined, inactive females) were not included. Ideally, effective maturity estimates are pre- ferred, and those estimates include only females that are sexually active during the spawning period (Pears et al., 2006). However, it was not possible to limit our estimates in this way because very few active females were sampled during the spawning months. A logistic curve was fitted to the data in the following form: P = 1/(1 + exp [-r(L - L^)]), (1) where r = the slope of the curve fitted to ln[(l-P)/P] versus TL; P = the proportion of mature female fish; and Ln, = the mean length at sexual maturity. The size (L50) and age (^50) at first maturity for females was estimated to be the intercept point at which 50% of individual fish were mature. Seasonal reproductive patterns for female fish were calculated by plotting estimates based on the gonad- osomatic index (GSI) and the frequency of reproductive stages by month. Because males were sampled rarely, the GSI was calculated only for female fish with the following equation: GSI = GonW / GW x 100, (2) where GonW = the gonad wet weight (in grams). To determine the frequency of reproductive stages each month, only mature, active (ripening, ripe, and running ripe) and mature, inactive (resting) females were in- cluded. The relationship between gonad weight and to- tal length and age was explored to determine whether there was a disproportionate increase in gonad weight above size and age at first maturity, as seen in other epinepheline groupers (Pears et al., 2006). Data analysis The relationship between TL and GW was estimated for 371 fish by using linear regression analysis. To lin- earize the power curve that best described this rela- tionship {GW=slTL'°), both variables were transformed by using the natural logarithm of x, loge x. The line of best fit for the linear relationship was described by loggGW = loggC + logePL. (3) Because of the low numbers of male fish, the von Bertalanffy growth function (VBGF) was fitted to length-at-age data for all sexes combined, by calculat- ing a nonlinear least-squares regression of TL on age: Lt = (1 - e-«^-^o)), (4) where - TL at age t\ = mean asymptotic TL; K = the growth coefficient; and /o = the age at which theoretical TL is zero. The age-based catch curve technique, estimated for all sexes combined, was used to determine the annual instantaneous rate of total mortality (Z). The natural logarithm of the number of fish between each age class was then plotted by age, and the annual instantaneous rate of Z was then taken as the slope of the line of best fit. The first 4 age classes (320-650 mm TL) were not included in this analysis because they represented fish not fully vulnerable to the fishing gear. Calculations of M were made with the equation of Pauly (1980), which incorporates water temperature and the VBGF param- eters of and K. A second method for estimating M, developed by Hoenig (1983), also was applied by using the general equation loggZ = 1.46 - 1.01 (loge^max)> (5) where Z is analogous to M in an unexploited population. The mean annual water temperature for Musandam is 26.8°C (Wilson^). The instantaneous mortality rate (F) was estimated by subtracting the estimate of M from Z estimated as the slope in the descending limb of the age-based catch curve (F=Z-M). An estimate of the rate of exploitation (F) was calculated as E=F/Z: ln(M) = 0.55 - 1.611n(L) + 1.441n(Linf) + ln(iD. (6) To characterize the state of the fishery, the following F-based biological reference points were calculated: 0.5 M for a precautionary target (Fop^) and 0.67 M for the mortality limit (Fiimit). Results Age determination and validation Annuli (opaque zones that appeared darker than the adjacent translucent ones) were counted on sectioned sagittal otoliths of orange-spotted grouper. Among the opaque zones, the first one was more difficult to define because it was usually much wider than the remainder. Of the 214 sections, 206 were read successfully, with ages of 1-15 years. The presence of an opaque zone at the margin of otolith (identified by edge analysis) indicated that annuli had a strong seasonality in depo- sition. More than 80% of all fish sampled from Septem- ber through December had opaque margins. A lag of 3 months was observed between the peak in seawater 3 Wilson, S. 2011. Unpubl. data. Five Oceans Environme- nal Services, Way 3021, Al Qurm, Muscat, Sultanate of Oman 494 Fishery Bulletin 1 14(4) Table 1 Number, size frequency (total length [TL] in millimeters), and mean gutted whole weight (GW), with standard errors (SEs), of samples of orange-spotted grouper (Epinephelus coioides) used in this study. Samples were collected in northern Oman between May 2004 and August 2005. n=number of fish sampled. Sex n TT mm-raax Mean TL (SE) GW mm-max Mean GW (SE) Female 175 281-900 589.58 (7.21) 250-9436 2680.37 (102.06) Transitional 11 449-748 646.00 (24.60) 1101-5156 3539.27 (1452.26) Male 13 443-1005 617.54 (44.88) 1008-18906 3866.15 (327.79) Size class (mm TL) Figure 2 Relationship between size class (total length in millimeters) and fre- quency of (A) immature female (n=74), (B) undetermined maturity- stage female (n = 17), (C) mature female (n=84), (D) male (n=13) , and (E) transitional phase (n=ll) orange-spotted grouper {Epinephelus coi- oides) within samples collected in northern Oman between May 2004 and August 2005. The dashed line indicates size at maturity. temperature in June and the formation of the opaque zone at the otolith edge in September. Size and age distribution The relationship between TL and GW for both sexes combined was best described by the equation GW = 5.27-6 X (TL3129)^ (7) where the sample size was 356 and the coefficient of determination (r^) was 0.97 (P<0.001). Females dominated the sample (n=178, 86.5 %) and ranged in size from 281 to 900 mm TL and in age from 1 to 11 years (average: 589.6 mm TL [standard error (SE) 7.21]) (Table 1; Fig. 2, A-C; Fig. 3, A-C). More than half the females (n=77) sampled, however, were classified as un- determined, inactive because it was not possible to establish their spawning his- tory. Males composed only 6.5% of the sampled population, but they were found throughout the size and age classes (Figs. 2D and 3D). The smallest male was 443 mm TL and 2 years old, and the largest male was 1005 mm TL and 15 years old (average: 5 96. 4 mm TL[SE 65.40]) (Table 1). The male-to-female ratio was 1:14. The size and age at sex change was de- scribed by a relatively narrow size range (449-748 mm TL) and age range (4-8 years) for transitional individuals (Figs. 2E and 3E). Growth models and mortality estimates Because of the small number of males that were sampled, the VBGF curve was fitted to all fish, regardless of sex. The plot of size at age revealed a trend in the growth trajectory that was almost linear and had no apparent asymptote (Fig. 4A). Estimates of the VBGF pa- rameters were as follows: L„,=927.97 mm, iir=0.135, to= -2.33. There was a curvilinear relationship between otolith weight and age, best described by the follow- ing equation: Age = 0.101 -I- -0.0023(OW) + 0.0017(OW)2, (8) where OW = otolith weight (to the nearest 0.001 g) (r^=0.555, n=204) (Fig. 4B). Mcliwain et al,: Demographic profile of Epinephelus coloides in northern Oman 495 30 20 10 0 30 20 10 0 30 ^ 20 0) g- 10 0) it 0 30 20 10 0 30 20 10 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Age (years) Figure 3 Relationship between age (years) and frequency of (A) immature fe- male (n=71), (B) undetermined female (n=16), (C) mature female (n=80), (D) male («=13) , and (E) transitional phase (rt=ll) orange- spotted grouper {Epinephelus coioides) within samples collected in northern Oman between May 2004 and August 2005. The instantaneous rate of mortality iZ), estimated with the age-based catch curve, was 0.722/year (Fig. 4C; Table 2). The estimate of M determined from Pauly’s (1980) equation was 0.14/year, which is comparable to the estimate derived from Hoenig’s equation (Hoenig, 1983). Although the estimates of F from the 2 meth- ods were similar (0.65/year and 0.59/year, respective- ly), both values were much greater than the and biological reference points. The estimate for E was 0.81/year. Description of sex and maturation stages Females As with most species of Epinephelus de- scribed to date, the gonads were fused posteriorly, bi- lobed, and unequal in size (Sadovy and Colin, 1995; Fennessy and Sadovy, 2002). Ovaries of immature females (n=71) were small, had a thin gonad wall, and lamellae were dominated by gonia and previtel- logenic oocytes in the primary growth stage (Fig 5B). Precursory dorsal sperm sinuses were present in 29% of these individuals. Undetermined, inactive females (n=17) were included as a secondary classification for females for which reproductive history could not be determined. For undetermined, inactive females, gonads were packed with gonia and previtellogenic, primary growth oo- cytes similar to those found in immature females, and no spermatogenic material was present. The tunica (gonad wall) for these individuals also resembled that of immature individuals. Precursory dor- sal sperm sinuses were present in 38% of undetermined, inactive females (Fen- nessy and Sadovy, 2002). The ovaries of mature inactive fe- males (n=27), were larger in diameter than the ovaries of immature females, and they were dominated by previtel- logenic oocytes at the cortical alveo- lar stage, the most advanced stage of oocytes (Fig. 5B). Evidence of recent spawning as a female was provided by the presence of a thick gonad wall and intralamellar muscle bundles in 26% of mature, inactive females caught during November and December. Mature, active females {n-57) contained oocytes in all stages of development (Fig. 50. Those females in active vitellogenesis (ripen- ing and ripe) were dominated by oocytes in the cortical alveolus and yolk globule stages, respectively. One spent female was observed (Fig. 5D). Transitional individuals In gonads of all transitional individuals (n=ll), evidence of prior female spawning was found. That evidence included a thick ovarian wall or the presence of intralamellar muscle bundles (scarring) and the occasional presence of brown bodies. Seven individuals were classified as early transitional fish because they contained crypts of primary and secondary spermatocytes or sperma- tids and oocytes in either the cortical alveoli stage or yolk globule stage (Fig. 6A). There was no evidence of degeneration of either male or female germ cells. In the 4 individuals classified as late-stage transitional fish, degenerating or atretic oocytes in the presence of maturing testicular tissue at differing stages of development were observed (Fig. 6B). The late-stage transitional fish were sampled within the perceived spawning season (March-May). Spermatozoa had not filled the dorsal sperm sinuses of any transitional individual. Males The gonads of immature males (n=ll) were small and compact and while contained the basic structure of an immature female, also had crypts of primary and secondary spermatocytes and spermatids (Fig. 6C). The gonia of 1 individual appeared to be de- generating. There was a relationship between fish size and total area of spermatogenic tissue, with larger fish containing more spermatogenic crypts scattered 496 Fishery Bulletin 1 14(4) 1100 1000 900 800 700 600 500 400 300 200 100 0 Lf=927.97(1 - exp(-0.135 (‘--2-33))) r2=z0.501 7 8 9 10 11 12 13 14 15 16 Age (years) Otolith weight (g) Figure 4 Plots (A) of size at age, (B) of otolith weight against age, and (C) of age against otolith weight for orange-spotted grouper {Epinephelus coioides) collected in northern Oman between May 2004 and August 2005. Seasonal reproduction The GSI for orange-spotted grouper indicat- ed a relatively protracted spawning season that peaks between March and May. A com- parison of values from the GSI with data on seawater temperature revealed that the onset of the spawning season in March oc- curred at the same time that temperatures rose above 26°C. The plotting of the fre- quency of mature, active and mature, inac- tive females by month revealed the timing of reproductive output. During the 3-month peak spawning period, more than 90% of the sample was composed of mature, active individuals (ripening and ripe), and only one mature, inactive RS female found. In addition, ripe females made up more than 30% of each of the monthly samples for 7 months of the year, indicating a protracted spawning season beyond the one described by the GSI. Size and age at maturation Plots of age and size frequency for females during the full spawning season (April- June) showed a large proportion of imma- ture or undetermined inactive females. By contrast, the numbers of mature inactive females within this period were low. The size and age at 50% maturity for female orange-spotted grouper sampled throughout the year were estimated to be 580 mm TL and 4 years, respectively. Discussion among the previtellogenic oocytes. Dorsal sperm si- nuses were present in 4 immature males. Two of these males were primary males with small gonads and no sign of prior female spawning. In mature active males {n-2), the testes were domi- nated by spermatogenic tissue and the the sperm si- nuses were filled with spermatozoa (Fig. 6D). Primary and secondary spermatocytes and spermatids were also present in both fish. There was no evidence of previtel- logenic oocytes in these testes. No spent males were sampled. The fishery for catch of orange-spotted grou- per in northern Oman comprised primar- ily young, mostly immature females with a maximum age of 11 years. Males were rare and made up only 6.5% of the sample. We found that this species is slow growing (had a low K), and produces a VBGF curve best described as linear (with no apparent asymptote). This latter result, coupled with the maximum age (15 years) that is lower than the age of orange-spotted grouper in other locales, indicates that the observed age structure is heavily truncated as a result of overfishing. The histological evidence, based on representative fish from all size classes and samples throughout the year, indicates that the population of orange-spotted grouper from northern Oman is diandric protogynous — a finding that is similar to that from reports on this species within the Arabian Gulf (Grandcourt et ah, 2005, 2009). Furthermore, we found through the presence of primary males (smaller males than those at first reproduction) that the devel- Mcliwain et al.: Demographic profile of Epinephelus coloides in northern Oman 497 Table 2 Estimates of mortality for orange-spotted grouper {Epinephelus coioides) calculated for this study and by Grandcourt (2005) by using methods of Pauly (1980) and Hoenig (1983). Also included are the precautionary biological reference points, precautionary target (i^opt) arid limit (i^iimit)> which are based on Hoenig’s estimate of natural mortality (M) and exploitation rate. Parameter (value/year) This study Grandcourt (2005) Total mortality — (from the catch curve) 0.72 0.97 Natural mortality (Hoenig) 0.08 0.19 Natural mortality (Pauly) 0.14 - Fishing mortality (Hoenig) 0.65 0.78 Fishing mortality (Pauly) 0.59 - Exploitation rate 0.81 0.80 Fopt = 0-5 M (Hoenig) 0.04 0.10 ^limit = 0.67 M (Hoenig) 0.05 0.13 Figure S Photomicrographs of histological sections taken from gonads of orange-spotted grouper (Epineph- elus coioides) collected in northern Oman between May 2004 and August 2005: (A) immature fe- male gonad, with previtellogenic oocyte (PVO), gonad wall (GW), lumen (LU), and sperm sinus (SS); (B) mature, inactive female gonad, with CA and GW; (C) mature active (ripe) female gonad, with CA and yolk globule stage oocyte (YG); and (D) mature, active (spent) female gonad, with YG and postovulatory follicle (POF). 498 Fishery Bulletin 1 14(4) Figure 6 Photomicrographs of histological sections taken from gonads of orange-spotted grouper (Epineph- elus coioides) collected in northern Oman between May 2004 and August 2005: (A) early-stage transitional gonad, with previtellogenic oocyte (PVO), yolk globule-stage oocyte (YCJ), and sper- matogenic crypt (SPC); (B) late-stage transitional gonad, with atretic oocyte (AO), lumen (LU), gonad wall (GW), and SPC; (C) immature male gonad, with PVO, SPC, spermatids or spermatozoa (SP); and (D) mature (ripe) male gonad, with dorsal sperm sinus (DSS), spermatocytes (SC), SP, gonad wall (GW), and blood vessel (BV). opmental pathway of male orange-spotted grouper from Oman is diandric and occurs in one of 2 ways, directly from juveniles or by sex change from fully functioning mature females (Rhodes and Sadovy, 2002). Demography and fishing impacts The demographic characteristics of orange-spotted grouper from Oman were very similar to the character- istics observed in 2 other growth studies of this species conducted in the region (Kuwait: Mathews and Samuel, 1987; Abu Dhabi: Grandcourt, 2005). All the growth parameters, L„ (93.07 cm TL; 97.9 cm TL), K (0.1655; 0.14), and to (-0.39; -1.5), for fish in Kuwait and Abu Dhabi were almost identical to the parameters in our results. However, the most significant and perhaps most worrying differences within the Oman population of or- ange-spotted grouper are the values for the two demo- graphic parameters that are often cited as indicators of overfishing: a reduced maximum age and a truncated age structure. The maximum ages for orange-spotted grouper from the Abu Dhabi study and from our Oman study were 12 and 15 years, significantly lower than the maximum age of 22 years reported for fish from Kuwait during the 1980s (Mathews and Samuel, 1987). Similarly, nearly 23% of the samples from Kuwait (av- eraged over 4 years of sampling) represented fish older than 10 years, compared with 0.5% of the samples from Abu Dhabi (Grandcourt et ah, 2005, Grandcourt, 2012) and 2% of the samples from Oman. Although it is still not clear how sex change is con- trolled in orange-spotted grouper, the large size range over which sex change occurs indicates behavioral con- trol, which acts independently of the social group size (Shapiro, 1987). It has been shown that if sex change is controlled by behavior, the removal of the larger, older males through fishing results in rapid sex change by a large female in a group (Huntsman and Schaaf, 1994). Such a compensatory effect means that sex ratios are not necessarily affected by fishing; however, the size Mcliwain et al.: Demographic profile of Epinephelus coloides in northern Oman 499 and age at which sex change occurs is reduced signifi- cantly (Adams et al., 2000). For socially mediated sex change to occur, female assessment of either the size (size-ratio assessment) or sex (sex-ratio assessment) of the group is made. In the latter scenario, when males are rare, the numbers of fish undergoing transition in- creases (Coleman et al., 1996). Although male orange-spotted grouper were uncom- mon in the population in northern Oman, there was no evidence that the numbers of transitional fish had increased, and therefore the compensatory mechanism has largely been overridden (Levin and Grimes, 2001). Dramatic reductions in male numbers have previously been documented for gag (Mycteroperca microlepis) and scamp {Mycteroperca phenax) in the Gulf of Mexi- co, with dire consequences for those populations (Cole- man et al., 1996). We argue that the smaller numbers of older fish and secondary males in our sample are not attributed to fishing selectivity (e.g., gear type or depth). Sampling was rigorous and representative of the fishing fleet, which included large wood dhows (>10 m length overall) and small fibreglass boats (4-10 m length overall). Unpublished creel surveys of the Musandam region reveal that dhows are landing the largest individuals because they fish for longer periods in waters up to 300 m (senior author, unpubl. data). Wire-mesh traps, the most common gear type for this fishery, have an average opening size of 100 cm (Al Masoori et al., 2004), big enough to allow the largest orange-spotted grouper to enter and to ensure that they are vulnerable to being caught by the gear type. Grouper populations, whose male biomass is eroded through selective fishing, face a significant loss of re- productive output. Although there are no baseline data on the sex ratio of orange-spotted grouper, the males composed 6.5% of the sampled population, similar to results of study on the Atlantic coast of the United states of male gag, whose numbers declined from 19.6% to 1.9% over 12 years (McGovern et al., 1998). A loss of males can seriously affect female reproductive suc- cess to the point where some groups of females with limited access to males remain reproductively inactive (Coleman et al., 1996). Limited contact between the sexes also can destabilize social hierarchies (Levin and Grimes, 2001) and, therefore, the overall functioning of the aggregation (Coleman et al., 1996). Historical aggregations of individuals that return each year to spawn in a particular place can be reduced substantial- ly in density if newly recruited fish have no older, ex- perienced fish from which to learn the migration route (Coleman et al., 1996). Future biological monitoring of the population in Oman should include an increased sample size during the spawning season that would al- low for detection of nonspawning females. Likewise, a reduction in the male-to-female sex ratio could be a useful indicator of stock recovery after the introduc- tion of regulations, and male biomass can be used as a proxy for reproductive potential (Levin and Grimes, 2001). Spawning season The peak of the spawning season for the Oman popula- tion of orange-spotted grouper occurred over a 3-month period from March through May, and there was evi- dence of protracted spawning across the majority of the year. This spawning period contrasts not only with the spawning periods of populations of orange-spotted grouper in the Arabian Gulf but also with those of a range of other species in the gulf in which spawn- ing periods are much more constrained when compared with the spawning periods of conspecific populations outside the gulf Populations of the pink ear emperor {Lethrinus lentjan), spangled emperor (L. nebulosus), and black spot snapper {Lutjanus fuluiflamma) in the Arabian Gulf all spawn for 2-4 months between April and July (Grandcourt et al., 2006a, 2006b), whereas conspecific populations (pink ear emperor and span- gled emperor) within the Indo-Pacific, Indian Ocean, and Red Sea will spawn for the whole year or between August and March (Carpenter and Allen, 1989; Kailola et al., 1993). Such constrained spawning would be ex- pected to be associated with seasonal extremes in wa- ter temperature; all spawning of gulf species occurs as water temperatures increase into summer, and spawn- ing ceases during the hottest summer months (August and September). Implications for the regional and global management of groupers There is no evidence to indicate that the structure of stocks of orange-spotted grouper within Oman waters at the time of this writing (2015) is any different from the structure we have presented here (data collected in 2005). The high rate of F and high rate of E, considered unsustainable for slow growing grouper, are similar to the rates reported for species of grouper worldwide (Sa- dovy de Mitcheson et al., 2013). We argue that immediate management of popula- tions of orange-spotted grouper and of species of grou- per throughout this region is warranted. However, the management options are limited for the populations of this species (and the majority of species of Epineph- elidae) within Oman and the wider Indian Ocean sim- ply because these species predominantly inhabit deep water (Heemstra and Randall, 1993; senior author, unpubl. data). For this group of species, implementing minimum size limits and quotas for catch would be im- practical, not only because barotrauma experienced by fish upon release causes a significant rate of mortality but also because this fishery includes a number of dif- ferent serranid species (senior author, unpubl. data). One management option that has been shown to be useful in a range of regions around the globe is area closure, and we argue that area closures may be an ef- fective management option for this species and species of Epinephelidae because this approach has the poten- tial not only to reduce overall fishing effort but also to protect key benthic habitats (Anderson et al., 2014). 500 Fishery Bulletin 114(4) Closure of the fishery within Oman at certain times of the year, specifically those times associated with the peak spawning period (March-May) may be useful, and the introduction and increased support of no-take zones or marine protected area that encompass critical habitat and aggregation or spawning sites of grouper will be vital for the persistence of viable populations of orange-spotted grouper and other species Epinepheli- dae within Oman. Acknowledgments This research was made possible by financial aid from the Ministry of Agriculture and Fisheries, Oman (Agri- cultural Development Fund). We also thank E. Grand- court and K. Rhodes for earlier discussions at the man- uscript stage. Literature cited Adams, S., B. D. Mapstone, G. R. Russ, and C. R. Davies. 2000. Geographic variation in the sex ratio, sex specific size, and age structure of Plectropomus leopardus (Ser- ranidae) between reefs open and closed to fishing on the Great Barrier Reef. Can. J. Fish. Aquat. Sci. 57: 1448-1458. Anderson, A. B., R. M. Bonaldo, D. R. Barneche, C. W. Hack- radt, F. C. Felix-Hackradt, J. A. Garcia-Charton, and S. R. Fleeter. 2014. Recovery of grouper assemblages indicates effec- tiveness of a marine protected area in southern Bra- zil. Mar. Ecol. Prog. Ser. 514:207—215. Beverton, R. and S. J. Holt. 1957. On the dynamics of exploited fish populations. Fish. Invest. Series II, 19, Chapman & Hall, London. Brule, T., T. Colas-Marrufo, A. Tuz-Sulub, and C. Deniel. 2000. Evidence for protogynous hermaphroditism in the serranid fish Epiiiephelus drummondhayi (Perciformes: Serranidae) from the Campeche Bank in the southern Gulf of Mexico. Bull. Mar. Sci. 66:513-521. Bullock, L. H., M. D. Murphy, M. F. Goodcharles, and M. E. Mitchell. 1992. Age, growth, and reproduction of jewfish Epineph- eliis itajara in the eastern Gulf of Mexico. Fish. Bull. 90:243-249. Carpenter, K. E., and G. R. Allen. 1989. FAO species catalogue, vol. 9. Emperor fishes and large-eye breams of the world (family Lethrinidae). An annotated and illustrated catalogue of lethrinid species known to date. FAO Fish. Synop. 125, 118 p. FAO, Rome. Cheat, J. H., and D. R. Robertson. 2002. Age-based studies. In Coral reef fishes: dynamics and diversity in a complex ecosystem (P. F. Sale, ed.), p. 57-80. Academic Press, San Diego, CA. Choat, J. H., D. R. Robertson, J. L. Ackerman, and J. M. Posada. 2003. An age-based demographic analysis of the Carib- bean stoplight parrotfish Sparisoma viride. Mar. Ecol. Prog. Ser. 246:265-277. Coleman, F. C., C. C. Koenig, and L. A. Collins. 1996. Reproductive styles of shallow-water groupers (Pi- sces: Serranidae) in the eastern Gulf of Mexico and the consequences of fishing spawning aggregations. Envi- ron. Biol. Fish. 47:129-141. Coleman, F. C., C. C. Koenig, G. R. Huntsman, J. A. Musick, A. M. Ecklund, A.M., J. C. McGovern, G. R. Sedberry, R. W. Chapman, and C. B. Grimes. 2000. Long-lived reef fishes: the grouper-snapper com- plex. Fisheries 25:14-20. Colin, P. L. 1992. Reproduction of the Nassau grouper, Epinephelus striatus (Pisces: Serranidae) and its relationship to envi- ronmental conditions. Environ. Biol. Fish. 34:357-377. Cornish, A., and M. Harmelin-Vivien. 2004. Epinephelus coioides. The lUCN Red List of Threatened Species 2004: e.T44674A10934751. [Avail- able at website, accessed November 2015.] DeMartini, E. E., A. R. Everson, and R. S. Nichols. 2011. Estimates of body sizes at maturation and at sex change, and the spawning seasonality and sex ratio of the endemic Hawaiian grouper (Hyporthodus quernus, F. Epinephelidae). Fish. Bull. 109:123-134. Fennessy, S. T, and Y. Sadovy. 2002. Reproductive biology of a diandric protogynous her- maphrodite, the serranid Epinephelus andersoni. Mar. Freshw. Res. 53:147-158. Ferreira, B. R, and G. R. Russ. 1992. Age, growth and mortality of the inshore coral trout Plectropomus rnaculatus (Pisces: Serranidae) from the central Great Barrier Reef, Australia. Aust. J. Mar. Freshw. Res. 43:1301-1312. Grandcourt, E. M. 2002. Demographic characteristics of a selection of ex- ploited reef fish from the Seychelles: a preliminary study. Mar. Freshw. Res. 53123 — 130. 2005. Demographic characteristics of selected epineph- eline groupers (Family: Serranidae; Subfamily; Epineph- elinae) from Aldabra Atoll, Seychelles. Atoll Res. Bull. 539:201-216. 2012. Reef fish and fisheries in the Gulf In Coral reefs of the Gulf: adaptation to climatic extremes (B. M. Riegl and S. Purkis, eds.), p. 127-161. Springer, Dordrecht, Netherlands. Grandcourt, E.M., T. Z. A1 Abdessalaam, F. Francis, and A. T. A1 Shamsi. 2005. Population biology and assessment of the orange- spotted grouper, Epinephelus coioides (Hamilton, 1822), in the southern Arabian Gulf Fish. Res. 74:55-68. Grandcourt, E. M., T. Z. A1 Abdessalaam, and F. Francis. 2006. Age, growth, mortality and reproduction of the blackspot snapper, Lutjanus fulviflamma (Forsskal, 1775), in the southern Arabian Gulf Fish. Res. 78:203- 210. Grandcourt, E. M., T. Z. A1 Abdessalaam, A. T. A1 Shamsi, and F. Francis. 2006. Biology and assessment of the painted sweetlips (Diagramma pictum (Thunberg, 1792)) and the spangled emperor (Lethrinus nebulosus (Forsskal, 1775)) in the southern Arabian Gulf Fish. Bull. 104:75-88. Grandcourt, E. M., T. Z. A1 Abdessalaam, F. Francis, A. T. A1 Shamsi, S. A. Hartmann. 2009. Reproductive biology and implications for manage- ment of the orange-spotted grouper Epinephelus coioides in the southern Arabian Gulf J. Fish Biol. 74:820-841. Mcliwain et al.: Demographic profile of Epinephelus coloides in northern Oman 501 2011. Demographic parameters and status assessments of Lutjanus ehrenbergii, Lethrinus lentjan, Plectorhinchus sordidus and Rhabdosargus sarba in the southern Ara- bian Gulf. J. Appl. Ichthyol. 27:1203-1211. Heemstra, P. C., and J. E. Randall. 1993. FAO species catalogue, vol. 16. Groupers of the world (family Serranidae, subfamily Epinephelinae): an annotated and illustrated catalogue of the grouper, rock- cod, hind, coral grouper and lyretail species known to date. FAO Fish. Synop. 125, 382 p. FAO, Rome. Hoenig, J. M. 1983. Empirical use of longevity data to estimate mortal- ity rates. Fish. Bull. 81:898-902. Huntsman, G. R., and W. E. Schaaf 1994. Simulation of the impact of fishing on reproduction of a protogynous grouper, the graysby. North Am. J. Fish. Manage. 14:41-52. Kailola, P. J., M. J. Williams, P. C. Stewart, R. E. Reichelt, A. McNee, and C. Grieve. 1993. Australian fisheries resources, 422 p. Bureau of Resource Sciences and Fisheries Research Development Corporation, Canberra, Australia. Levin, P. S., and C. B. Grimes. 2001. Reef fish ecology and grouper conservation and management. In Coral reef fishes: dynamics and diver- sity in a complex ecosystem (P. F. Sale, ed.), p. 377-389. Academic Press, San Diego, CA. Liu, M., and Y. Sadovy de Mitcheson. 2009. Gonad development during sexual differentiation in hatchery-produced orange-spotted grouper {Epineph- elus coioides) and humpback grouper (Cromileptes altive- lis) (Pisces: Serranidae, Epinephelinae). Aquaculture 287:191-202. Luckhurst, B. E. 2008. Site fidelity and return migration of tagged red hinds {Epinephelus guttatus) to a spawning aggrega- tion site in Bermuda. Proc. Gulf Caribb. Fish. Inst. 50:750-763. Manooch, C. S., III. 1987. Age and growth of the snappers and groupers. In Tropical snappers and groupers: biology and fisheries management (J. J. Polovina and S. Ralston, eds), p. 329- 374. Westview Press, London. Mathews, C. R, and M. Samuel. 1987. Growth, mortality and assessments for groupers from Kuwait. Kuwait Bull. Mar. Sci. 9:173-191. McGovern, J. C., D. M. Wyanski, O. Pashuk, C. S. Manooch II, and G. R. Sedberry. 1998. Changes in the sex ratio and size at maturity of gag, Mycteroperca microlepis, from the Atlantic coast of the southeastern United States during 1976-1995. Fish. Bull. 96:797-807. Newman, S. J., D. M. Williams, and G. R. Russ. 1996. Age validation, growth and mortality rates of the tropical snappers (Pisces: Lutjanidae) Lutjanus adetii (Castelnau, 1873) and L. quinquelineatus (Bloch, 1790) from the central Great Barrier Reef, Australia. Mar. Freshw. Res. 47:575-584. Pauly, D. 1980. On the interrelationships between natural mortal- ity, growth parameters, and mean environmental tem- perature in 175 fish stocks. J. Cons. Int. Explor. Mer 39:175-192. Pears, R. J., J. H. Choat, B. D. Mapstone, and G. A. Begg. 2006. Demography of a large grouper, Epinephelus fus- coguttatus, from Australia’s Great Barrier Reef: impli- cations for fishery management. Mar. Ecol. Prog. Ser. 307:259-272. 2007. Reproductive biology of a large, aggregation-spawn- ing serranid, Epinephelus fuscoguttatus (Forsskal): man- agement implications. J. Fish Biol. 71:795-817. Polunin, N. V. C., C. M. Roberts, and D. Pauly. 1996. Developments in tropical reef fisheries science and management. In Reef fisheries (N. V. C. Polunin and C. M. Roberts, eds.), p. 361-377. Chapman and Hall, London. Quinitio, G. F., N. B. Caberoy, D. M. Reyes Jr. 1997. Induction of sex change in female Epinephelus coioi- des by social control. Isr. J. Aquae. Bamidgeh 49:77-83. Radebe, P. V., B. Q. Mann, L. E. Beckley, and A. Govender. 2002. Age and growth of Rhabdosargus sarba (Pisces: Sparidae) from KwaZulu-Natal, South Africa. Fish. Res. 58:193-201. Randall, J., 1995. Coastal fishes of Oman, 439 p. Univ. Hawaii Press, Honolulu, Hawaii Reinboth, R. 1967. Biandric teleost species. Gen. Comp. Endocrinol. 9:146A. Rhodes, K. L., and Y. Sadovy. 2002. Reproduction in the camouflage grouper (Pisces: Serranidae) in Pohnpei, Federated States of Microne- sia. Bull. Mar. Sci. 70:851-869. Robinson, J., N. A. J. Graham, J. E. Cinner, G. R. Almany, and P. Waldie. 2015. Fish and fisher behaviour influence the vulnerabil- ity of groupers (Epinephelidae) to fishing at a multispe- cies spawning aggregation site. Coral Reefs 34:371-382. Russ, G. R., D. C. Lou, and B. P. Ferreira. 1996. Temporal tracking of a strong cohort in the popula- tion of a coral reef fish, the coral trout, (Plectropomus leopardus) Serranidae: Epinephaline, in the central Great Barrier Reef, Australia. Can. J. Fish. Aquat. Sci. 53:2745-2741. Russ, G. R., D. C. Lou, J. B. Higgs, and B. P. Ferreira. 1998. Mortality rate of a cohort of the coral trout, Plec- tropomus leopardus, in zones of the Great Barrier Reef Marine Park closed to fishing. Mar. Freshw. Res. 49:507-511. Sadovy, Y. J. 1996. Reproduction of reef fishery species. In Reef fish- eries (N. V. C. Polunin and C. M. Roberts, eds.), p.l5- 59. Chapman and Hall, London. Sadovy, Y, and P. L. Colin. 1995. Sexual development and sexuality in the Nassau grouper. J. Fish Biol. 46:961-976. Sadovy, Y., A. Rosario, and A. Roman. 1994. Reproduction in an aggregating grouper, the red hind, Epinephelus guttatus. Environ. Biol. Fish. 41:269-286. Sadovy de Mitcheson, Y., and M. Liu. 2008. Functional hermaphroditism in teleosts. Fish Fish. 9:1-43. Sadovy de Mitcheson, Y., M. T. Craig, A. A. Bertoncini, K. E. Carpenter, W. W. L. Cheung, J. H. Choat, A. S. Cornish, S. T. Fennessy, B. P. Ferreira, P. C. Heemstra, et al. 2013. Fishing groupers towards extinction: a global as- 502 Fishery Bulletin 114(4) sessment of threats and extinction risks in a billion dol- lar fishery. Fish Fish. 14:119-136. Samoilys, M. A., and L. C. Squire. 1994. Preliminary observations on the spawning behav- ior of coral trout, Plectropomus leopardus (Pisces: Ser- ranidae), on the Great Barrier Reef. Bull. Mar. Sci. 54:332-342. Shapiro, D. Y. 1987. Reproduction in groupers. In Tropical snappers and groupers: biology and fisheries management (J. J. Polovina and S. Ralston, eds.), p. 295-327. Westview Press, London. Siddeek, M. S. M., M. M. Fouda, and G. V. Hermosa Jr. 1999. Demersal fisheries of the Arabian Sea, the Gulf of Oman and the Arabian Gulf. Est. Coast. Shelf Sci. 49:87-97. Sluka, R. D., M. Chiappone, and K. M. Sullivan Sealey. 2001. Influence of habitat on grouper abundance in the Florida Keys, U.S.A. J. Fish Biol. 58:682-700. Smith, C. L. 1965. The patterns of sexuality and the classification of serranid fishes. Am. Mus. Novit. 2207:1-20. Starr, R. M., E. Sala, E. Ballesteros, and M. Zabala. 2007. Spatial dynamics of the Nassau grouper Epineph- elus striatus in a Caribbean atoll. Mar. Ecol. Prog. Ser. 343:239-249. Tuz-Sulub, A., and T. Brule. 2015. Spawning aggregations of three protogynous grou- pers in the southern Gulf of Mexico. J. Fish Biol. 86:162-185. Wakefield, C. B., S. J. Newman, R. J. Marriott, D. K. Bodding- ton, and D. V. Fairclough. 2013. Contrasting life history characteristics of the eight- bar grouper Hyporthodus octofasciatus (Pisces: Epineph- elidae) over a large latitudinal range reveals spawn- ing omission at higher latitudes. ICES J. Mar. Sci. 70:485-497. Wakefield, C. B., A. J. Williams, S. J. Newman, M. Bunel, D. K. Boddington, E. Vourey, and D. V. Fairclough. 2015. Variations in growth, longevity and natural mortal- ity for the protogynous hermaphroditic eightbar grouper Hyporthodus octofasciatus between the Indian and Pa- cific Oceans. Fish. Res. 172:26-33. Walters, C., and P. H. Pearse. 1996. Stock information requirements for quota manage- ment systems in commercial fisheries. Rev. Fish Biol. Fish. 6:21-42. West, G. 1990. Methods of assessing ovarian development in fishes: a review. Aust. J. Mar. Freshw. Res. 41:199- 222. Williams, A. J., C. R. Davies, B. D. Mapstone, and G. R. Russ. 2003. Scales of spatial variation in demography of a large coral-reef fish — an exception to the typical model? Fish. Bull. 101:673-683. Zeller, D. C., and G. R. Russ. 1998. Marine reserves: patterns of adult movement of the coral trout (Plectropomus leopardus (Serranidae)). Can. J. Fish. Aquat. Sci. 55917-924. 503 NOAA National Marine Fisheries Service Fishety Bulletin fy- established 1881 -ds Spencer F. Baird First U.S. Commissioner of Fisheries and founder of Fishery Bulletin Envirorimeiital conditions of 2 river drainages into the northern Golf of Mexico during successful hatching of Alabama shad iAIosa alabamae} Email address for contact author: iake.schaefer@usm.edu Abstract — -In recent years, the Ala- bama shad (Alosa alabamae) has experienced dramatic declines and extirpations from portions of its na- tive range. Habitat degradation and barriers to migration are considered contributing factors to contraction in the distributional range this spe- cies. To identify conditions during successful spawning, river tempera- tures and discharges in 2 drainages of the northern Gulf of Mexico (the Apalachicola and Pascagoula rivers) were characterized during successful hatching “windows.” Sampling dur- ing 2005-2009 yielded 400 juvenile Alabama shad of which 261 were aged from counts of rings on sagit- tal otoliths. Results from logistic regression revealed that successful spawning coincided with increases in temperature within a specific range (9.4-21.5°C) and with an average drainage-dependent discharge vol- ume (625.6 m^/s in the Apalachicola River and >400.7 m^/s in the Pasca- goula River). Timing of successful hatching windows differed between drainages but not between years within each drainage. Documenting and identifying the river conditions during successful reproduction pro- vide important information on how to manage rivers to aid in the recov- ery of this species of conservation concern. Manuscript submitted 9 February 2016. Manuscript accepted 26 August 2016. Fish. Bull. 114:503-512 (2016). Online publication date: 15 September 2016. doi: 10.7755/FB.114.4.11 The views and opinions expressed or implied in this article are those of the author (or authors) and do not necessarily reflect the position of the National Marine Fisheries Service, NOAA. Paul F. Mickle’ Jacob F. Schaefer (contact author)’ Susan B. Adams^ Brian R. Krelser’ William T. Slack^ ' Department of Biological Sciences The University of Southern Mississippi 118 College Drive, #5018 Hattiesburg, Mississippi 39406-5018 ^ Southern Research Station U.S. Forest Sen/ice U.S. Department of Agriculture 1000 Front Street Oxford, Mississippi 38655 3 Waterways Experiment Station EE-A Engineer Research and Development Center U.S. Army Corp of Engineers 3909 Halls Ferry Road Vicksburg, Mississippi 39180-6199 Migratory species invest significant resources in moving through a va- riety of disparate habitats to reach spawning sites (Gross, 1987; Roff, 1988). For fish species, these costs can be substantial (i.e., osmoregula- tion in diadromous fishes, or move- ment through suboptimal habitat that increases exposure to preda- tors, disease, or anthropogenic dis- turbances), and spawning migrations require energetic demands associated with gonad maturation (Leggett and Whitney, 1972; Hodgson and Quinn, 2002). These costs also are linked inextricably to a number of species life history traits (e.g., body size, fe- cundity, and age of maturation) that have co-evolved with migration to maximize individual fitness (Gross, 1987; Kinnison et ah, 2001). Despite the evolution of complex behaviors, physiology, and morphology, many migratory species show substantial flexibility in the timing of migration and the routes taken (Berthold, 2001; Alerstam et al., 2003). River conditions that are pos- sible cues for migration in fish spe- cies include discharge, flow velocity, temperature, suspended sediment, pH, conductivity, and dissolved oxy- gen (McLean et al., 1982; Quinn and Adams, 1996; Hewitt, 2003). These parameters are susceptible to rapid changes during spring rains when river discharge increases dramati- cally, possibly ultimately influencing the potential for successful recruit- ment (Maurice et al., 1987). Flow and temperature are correlated with oxygen levels and egg development time and ultimately with hatching success (Mann, 1996). After hatching, temperatures, nutrient levels, and turbidity are expected also to influ- ence growth and survivorship. There- fore, individuals spawned at different 504 Fishery Bulletin 114(4) 33°N 32°N 3rN 30°N 29°N Figure 1 Map of the 2 river drainages sampled for Alabama shad (Alosa alabamae) during 2005-2009. Five sites were sampled in the Pascagoula River basin in Mississippi: 1) Eastabuchie; 2) Shubuta; 3) McLain; 4) Merrill; and 5) Wade. Three sites were sampled in the Apalachicola River basin in Flori- da: 6) Woodruff; 7) Wewahitchka; and 8) Blountstown. times (early versus late) in a season may have different growth and mortality rates (Limburg, 1996). Fish otoliths provide a detailed history of the daily and annual growth of an individual and can provide a useful tool for retrospective assessment of the growth rates of early juveniles (Parsons and Peters, 1989; Gef- fen, 1992). Daily growth rings on otoliths consist of alternating calcium- and protein-rich layers (Geffen, 1992; Armstrong et al., 2004). Diel feeding cycles lead to variable growth, resulting in alternating opaque and translucent rings in each 24-h period. Unlike oth- er skeletal elements, otoliths do not undergo bone re- modeling that would potentially resorb layers (Simkiss, 1974). Therefore, the daily deposition of otolith rings provides a method for determining age (in days) in spe- cies of bony fish. Alabama shad {Alosa alabamae) is an anadromous fish that ascends rivers in the Mississippi River basin and northern Gulf of Mexico to spawn during spring months (February-May) of each year (Mettee and O’Neil, 2003). A special conservation status has been conferred on Alabama shad by several states, including Alabama, Arkansas, Florida, Georgia, Kentucky, Louisi- ana, Mississippi, and Missouri (Meadows et al., 2006). Population-level genetic data indicate a high degree of site fidelity within some populations of Alabama shad (Bowen, 2005). As a result, differences in spawning tim- ing and recruitment due to genetic drift or local adap- tation are possible. Until recently, much of the basic biology for Alabama shad had been inferred from research on the Ameri- can shad (A. sapidissima) in northern Atlantic basins. The American shad is fairly well studied; published works address fecundity, spawning, feeding behavior, and even restoration (Olney and McBride, 2003; Walter and Olney, 2003). Much less is known about the ecol- ogy of Alabama shad (Mickle et al., 2010). The goals of this study were to improve our understanding of river conditions during successful hatching of Alabama shad and compare these conditions among river drain- ages. Our objectives were 1) to assess differences in successful hatching timing of Alabama shad between 2 basins, 2) to evaluate and compare river conditions during successful hatching between the basins, and 3) to assess interannual variability in both hatch timing and river conditions during successful hatching. For this study, we defined successful hatching as hatching that resulted in larval survival to the juvenile stage. Once the timing of successful hatching was determined (from daily age data), we were able to compare river condition and hatch timing for the 2 rivers and sam- pling years — findings that would help to identify the necessary conditions for recruitment of Alabama shad. Materials and methods Hatch timing and river conditions during hatching were compared between 2 rivers in the central Gulf of Mexico (Fig. 1). The Apalachicola River is a large river Mickle et al.: Hatch window for Alosa alabamae 505 in the Florida panhandle and contains the largest ex- tant population of Alabama shad (Mettee and O’Neil, 2003). This river has a dam on the mainstem (Jim Woodruff Lock and Dam, at river kilometer 171), and Alabama shad spawn below the dam, although some pass through a lock system that is opened twice a day (Laurence and Yerger, 1967; Ely et al., 2008). The Pas- cagoula River, located in Mississippi, is the last large, free-flowing river in the contiguous United States (Dynesius and Nilsson, 1994). No spawning grounds of Alabama shad have been documented for the Pasca- goula River, although juveniles are consistently present within this river system (Mickle et al., 2010). To investigate drainage-level differences in success- ful hatch timing, juveniles were collected from the Pas- cagoula River basin in Mississippi during 2005-2009 and from the Apalachicola River basin in Florida dur- ing 2007-2008 and then aged by counting daily otolith rings. Samples were taken during June and October in the Apalachicola River basin and in June through October in the Pascagoula River basin, excluding Sep- tember and October 2005 after Hurricane Katrina. The year 2007 was the only year during the sampling period that was categorized as a severe drought year by the Palmer Drought Severity Index for the region (U.S. Drought Monitor, website). The other years (2005, 2006, 2008, and 2009) were characterized as low water, but not drought, years. Fish were collected with a SR-14EB^ electrofishing boat (Smith-Root Inc., Vancouver, WA) operated at 5000 W and 16 A with pulses-per-second ranging from 7.5 to 120. Electrofishing effort typically occurred for 1200 s at each site and was focused on the habitat types of sand bar, open channel, and bank. Alabama shad were tagged individually and placed in 95% ethanol. During low-water periods, some sites (typically shal- low sand bars) were not accessible by the electrofishing boat. Because Alabama shad undergo ontogenetic shifts in habitat use (Mickle et al., 2010), these sites were seined and occasionally a cast net was used to ensure individuals of all ages were sampled throughout the sampling period. Cast nets had diameters of 1.52-2.43 m and a bar mesh of 1.59 cm. Seines were 3. 0-3. 7 m wide by 1.8-2. 4 m deep and had a 0.3-cm-wide mesh. Age was estimated in days by counting daily rings on the sagittal otoliths (Secor et al., 1992). One oto- lith per individual fish was removed and mounted on a slide with Crystalbond mounting adhesive (Ted Pella. Inc., Redding, CA). Otoliths were mounted with the pri- mordia facing down and sanded by hand with sequen- tially finer grit-size paper, as necessary, to expose the rings. Daily rings were counted with a Wild Heerbrugg compound microscope (Leica Microsystems Inc., Buffalo Grove, IL). Magnification ranged from 290x to 1080x depending on the diameter (0.25-0.50 mm) of the oto- lith. Otolith images were taken with a SPOT Insight 1 Mention of trade names or commercial companies is for iden- tification purposes only and does not imply endorsement by the National Marine Fisheries Service, NOAA. Color digital camera (Diagnostic Instruments Inc., Sterling Heights, MI) and by using SPOT Advanced software, vers. 3.3 (Diagnostic Instruments Inc.), and were enhanced by using Image-Pro Express, vers. 4.0.1 (Media Cybernetics Inc., Rockville, MD). Age rings on each otolith were counted 3 times by the same person during separate sessions, and counts were averaged over the 3 observations. Of the otoliths read, 5% were randomly selected for independent vali- dation by another reader and compared with the range of original readings. The otoliths of older fish (>250 days, all from individuals >100 mm in total length [TL] captured later in the year) were thick and brittle, making sanding and accurate reading of daily rings dif- ficult. These older fish were removed from our analysis. As suggested by Geffen (1992), fish age was determined by adding 10 days to each daily ring count to compen- sate for the posthatching yolk stage that precedes daily ring formation in certain species. The hatch date for each individual was determined by counting back the age of the fish from its collection date. The successful hatching period (average time from the start of hatch- ing to the end of hatching [hereafter “hatch window”] ) for each drainage and year was defined as the period between the earliest and latest hatching date for all Alabama shad younger than 250 days captured within a given drainage in a given year. To evaluate river conditions during the successful hatching time, mean daily river discharge and temper- atures were analyzed for the period January-July of each year. For the Pascagoula River basin, river data were collected from the U.S. Geological Survey flow- gauging station (02479000) at Merrill (at river kilo- meter 137; Fig. 1). For the Apalachicola River basin, the data were collected from a US. Geological Survey gauging station (02358000) and a Florida Department of Environmental Protection station at Jim Woodruff Dam (at river kilometer 171; Fig. 1). All missing data (usually 1 or 2 daily mean values per data lapse) were compensated with mean datum from before and after the lapse period (<2% of all data). Overall difference in the timing of successful hatch- ing between rivers was assessed as the modal differ- ence between rivers and with years pooled. To test for significance of this value, we used a randomization procedure to build a null distribution of expected mod- al differences given the observed hatching times. For 10,000 permutations, 50 observations were randomly drawn (without replacement) for 2 groups (represent- ing the 2 rivers) and a modal difference calculated (by using the sample function in R statistical software, vers. 3.2.3 [R Core Team, 2015]). The significance of the observed modal difference between rivers was as- sessed by comparison with the distribution of permuted values. We used logistic regression to assess our ability to predict hatching periods from river conditions. In each of the 7 river-year combinations, we divided the first 182 d into 26 7-d periods. For each period, we calcu- lated standardized (z-score) mean temperature and 506 Fishery Bulletin 1 14(4) Table 1 Summary table of successful hatch windows for Alabama shad (Alosa alabamae) and river conditions during those windows in 2005-2009 in the Pascagoula River basin, Mississippi, and in 2007-2008 in the Apalachicola River basin, Florida (only data from within hatch windows are summarized). Minimums and maximums are for daily means during hatch windows. Year Number of Alabama shad collected Julian days of hatching Hatch window duration (d) Mean temp- erature (°C) Mean discharge rate (mYs) Minimum temp- erature (“O Maximum temp- erature PC) Minimum discharge rate (mYs) Maximum disharge rate (mYs) Pascagoula River 2005 85 32-58 27 13.1 747.8 12.8 15.8 255.1 1322.4 2006 79 38-73 36 14.9 436.5 10.2 16.5 187.7 914.6 2007 49 38-65 28 14.2 187.9 10.8 21.5 130.5 294.5 2008 57 32-79 48 15.2 422.2 10.7 21.0 166.2 852.3 2009 34 27-54 28 13.9 209.2 11.0 16.1 115.2 535.2 Grand Mean 60.8 33.4-65.8 33.4 14.3 400.7 11.1 18.18 170.9 783.8 Apalachicola River 2007 52 6-64 58 13.3 592.3 9.8 18.4 385.1 974.1 2008 44 10-67 58 14.5 658.4 9.4 18.8 285.9 1602.7 Grand Mean 48.0 8.0-65.5 58.0 13.9 625.6 9.6 18.60 335.5 1288.4 discharge (m^/s) and the change in temperature and discharge (difference in standardized temperature and flow from the previous period). Each period was scored as if it were part of the hatch window if it overlapped by >1 d with the established hatch window for that river and year. The logistic regression model (devel- oped by using the glm function in R) predicted hatch- ing from the mean and change in temperature and flow conditions and with year nested within river as factors. Results Over the 5 years of this study, 400 juvenile Alabama shad were collected; 304 from the Pascagoula River ba- sin (85, 79, 49, 57, 34 individuals in each year during the period 2005-2009, respectively) and 96 individuals from the Apalachicola River basin (52 and 44 in 2007 and 2008, respectively). Although it was sometimes dif- ficult to find otoliths clear enough to reveal a continu- ous sequence of rings, daily ring counts were completed successfully on otoliths from 208 Alabama shad from the Pascagoula River basin and from 53 Alabama shad from the Apalachicola River basin. Variability between the repeated counts of rings for individuals was low. Repeated counts typically differed by less than 5 rings; maximum disparity of 18 occurred in 1 otolith and the individual with maximum disparity was removed from analyses. Validation of the ring count was confirmed by the second reader whose counts fell within the range of the original 3 readings for each otolith. In all readable otoliths, the interior rings (around the primordium) remained clear; however, some fish collected later in the year had otoliths that were too thick to be aged accurately. Daily age validation was conducted by comparing natural, date-specific markers with specific river con- ditions (MacLellan and Saunders, 1995). We compared the number of rings during low growth periods (high- water events) with the number of days of high wa- ter events. This identification of events also allowed date-specific markers to be used for a comparison of ages by year and river. A 1:1 ratio of incremental count versus age was observed — a relationship that is consistent with results from other clupeid species for which daily otolith rings have been validated (Geffen, 1982; Campana et al., 1987; Moksness and Wespestad, 1989). In the Pascagoula River basin, successful hatch win- dows began from late January to early February (Ju- lian days 32, 38, 35, 32, and 27 in each year in the pe- riod 2005-2009, respectively) and ended from late Feb- ruary to late March (Julian days 58, 73, 65, 79, and 55) (Table 1, Figs. 2-5). For the Apalachicola River basin, the successful hatch windows began in early January (Julian days 6 and 9 in 2007 and 2008, respectively) and ended in early February (Julian days 64 and 67) (Table 1). Median hatching days (midway within each window) in the Pascagoula River basin occurred from early to late February (Julian days 45, 55.5, 50, 55.5, and 41 for the years in the period 2005-2009, respec- tively) compared with early to mid-February in the Apalachicola River basin (Julian days 35 and 38 for 2007 and 2008). Length of time of a successful hatch window was shorter in the Pascagoula River basin; an average window length was 30.8 d (SD 10.8), compared with 58 d (SD 0.0) in the Apalachicola River basin (Ta- ble 1, Figs. 2 and 3). All successful hatches occurred in the first 12 weeks of each year in both river basins. When all years were pooled, the modal hatching time Mickle et al.: Hatch window for Ahsa alabamae 507 Predicted HW Observed HW w B3 E 't' o Q. O N r, 3 ji t A il I s ? ^ Day of year Figure 2 Standardized (z-score) discharge (black line) and temperature (gray line) values for the Pascagoula River basin in Mississippi during (A) 2005; (B) 2006; (C) 2007; (D) 2008; and (E) 2009. Bars represent lo- gistic regression results for the probability of a time period being in the hatch window (HW) (Table 1). Dark gray bars represent periods of the observed hatch windows. Numbers above bars represent the total number of individuals aged back to that 7-d period. was 33 days later in the Pascagoula River basin (day 52) than in the Apalachicola River basin (day 19). This difference was greater than 99.3% of the permuted val- ues (P<0.003). The logistic regression allowed us to correctly clas- sify 86% of 7-d periods (with the use of a 50% prob- ability cutoff) as within or outside hatch windows. Mean temperature and discharge were significant contributors to the model and hatch windows occurred during periods of lower temperatures and higher dis- charges (Table 2). Neither change in temperature, nor change in discharge, were significant factors, but riv- er as a factor was marginally significant. Averaging across years, grand mean temperature during hatch- ing was 14.3°C in the Pascagoula River and 13.9°C in the Apalachicola River, with average rates of dis- charge during hatching of 400.7 m^/s and 625.4 m^/s, respectively (Table 1). In general, predicted and ob- served hatch windows were similar in the Apalachico- la River, whereas observed hatching in the Pascagoula tended to be 2-3 weeks later than that predicted by the logistic regression. 508 Fishery Bulletin 1 14(4) o O' fli cr ■< O B) 3 Q. O Day of year Figure 3 Standardized (z-score) discharge (black line) and temperature (gray line) val- ues for the Apalachicola River basin in Florida in (A) 2007 and (B) 2008. Bars represent logistic regression results for the probability of a time period being in the hatch window (HW) (Table 1). Dark gray bars represent periods of the observed hatch windows. Numbers above bars represent the total number of individuals aged back to that 7-d period. Discussion Hatch windows for Alabama shad were consistently longer and earlier in the Apalachicola River basin than in the Pascagoula River basin. There are 3 pos- sible explanations for these observed differences: 1) the cues, or timing of cues, that trigger spawning runs or acts of spawning differ between these river basins; 2) spawning runs and hatching take place at the same time, but in the years sampled, differences in recruit survival gave the appearance of significantly different hatch windows; or 3) the fish populations are differ- ent in the river basins, and they respond differently to cues. Previous work on population genetics of Ala- bama shad has indicated some drainage-level fidelity that would allow for interpopulation variability in the timing of spawning runs (Bowen, 2005; Bowen et ah, 2008). Other, unmeasured variables may also play key roles as migration or spawning cues. It should be noted that a lower sample size in the Apalachicola River ba- sin (2 versus 5 years sampled) could have contributed to differences. However, annual variation in timing of hatch windows within the drainages was clearly less than the variation in timing of hatch windows across drainages. The distribution of successful hatching dates from each drainage and year indicates that prolonged or multiple spawning events occurred (Figs. 2 and 3). The overall lengths of the windows for successful hatching from the Apalachicola River basin were comparable with lengths of hatch window documented for Ameri- can shad on the St. Lawrence River, between Canada and the United States (Maltais et ah, 2010). In gen- eral, in our study, hatch windows of Alabama shad co- incided with spring floods that had above-average flows and low, but increasing water temperatures. Spawning migrations of American shad are also linked to river temperature (Leggett and Whitney, 1972; Quinn and Adams, 1996). The Apalachicola River is impounded, and the effects of that control have the potential to ar- tificially influence temperature and flow. The buffering of these parameters by the impoundment would be ex- pected to prolong spawning windows — an outcome that would be consistent with our logistic regression results, which indicated that longer periods of lower tempera- tures were associated with successful spawning. In contrast to our findings, previous studies by Lau- rence and Yerger (1967) and Mills (1972) have reported that spawning temperatures fall in a range of 19-22°C. However, Ely et al. (2008) found adult Alabama shad present on the spawning grounds when temperatures were similar (13.1-15.2°C) to temperatures found in other studies (Laurence and Yerger, 1967; Mills, 1972) during the hatch windows that we detected. We found that mean daily water temperatures of 19-22 C did not occur until late March and April, after 96% of all docu- mented successful hatching was complete. Mean temperature and discharge were the most im- portant variables for predicting hatch windows. With these variables, hatch windows were found to be simi- lar in the 2 basins, but, in general, observed hatch win- dows were later than predicted in the Pascagoula River Mickle et al.: Hatch window for /A/oso alabamae 509 Day of year Figure 4 Raw discharge (black line) and temperature (gray line) values for the Pascagoula River basin, Mississippi, in (A) 2005; (B) 2006; (C) 2007; (D) 2008; and (E) 2009. Shaded areas represent the observed hatch windows (HWs) basin. This result indicates that temperature and dis- charge conditions in the Pascagoula River basin may be suitable when there is either limited spawning or recruitment earlier in the year. It should be noted that we cannot rule out the possibility that spawning in one or both rivers occurs over longer periods and that the differences in successful hatch timing among drainages reflect differences only in the time of successful recruit- ment. It is also possible that other cohorts of Alabama shad were present but not collected, or represented in the analysis; we deem this possibility unlikely given the extensive sampling in the drainages and the ab- sence of outliers to the observed hatch windows. Size data (Mickle et al., 2010) also are consistent with size of fish when hatching occurs in early spring (no small- er individuals [<55 mm TL] were sampled later in the year than July) and growth occurs through the summer and early fall. Therefore, drainage differences were more likely due to other unmeasured variables that might have contributed to the reduced hatch window or successful recruitment in the Pascagoula River basin. The data regarding differences in successful hatch windows provide valuable information but do not al- low for definitive conclusions on the mechanisms at work. Additional spatial and temporal data are needed to elucidate these mechanisms. Unfortunately, the Ala- bama shad is a species in decline and sample sizes are likely to be too small, or populations judged too imper- 510 Fishery Bulletin 114(4) Table 2 Results from the logistic regression in predicting hatch windows from mean and change (A) in standardized temperature and discharge. Significant predictors (chi- squared statistic) are shown in bold font. df=degrees of freedom. df Estimate Standard error P Discharge 1 0.576 0.260 4.91 0.027 A Discharge 1 0.121 0.261 0.64 0.645 Temperature 1 -2.490 0.463 51.89 <0.001 A Temperature 1 -1.207 0.734 2.76 0.097 River 1 -1.109 0.815 0.972 0.061 River*Year 5 1.53 0.909 iled, to allow for intensive adult sampling for migration analyses. The initial experimental design of this study included a comparison of samples from inland drain- ages in Arkansas and Missouri, where spring increases in water temperature would lag behind those in Gulf of Mexico drainages. However, 3 years of sampling during August within these systems, at sites where fish had been collected in earlier years, yielded no juvenile fish. Within impounded water systems, flow regimes may be managed through water releases to mimic natural cues and allow populations to complete their spawn- ing runs. Castro-Santos and Letcher (2010) found that flow-regulated systems altered spawning success and migration timing of American shad both upstream and downstream. Other riverine species have also been af- fected adversely by altered flow regimes. Manipulated flows physically altered habitats, resulting in decreased diversity of fish species (Freeman et ah, 2001). Alter- ing systems in which Alabama shad reproduce may add stressors to migrating juveniles and adults. Unfortu- nately, the drainages along the coast of the northern Gulf of Mexico are also areas that have been rapidly developed over the past 75 years and that development may ultimately lead to further extirpations of this spe- cies (Turner, 1990). Acknowledgments We thank K. Limburg and 2 anonymous reviewers for feedback provided on an earlier draft. We also thank M. Bland, A. Commens-Carson, G. McWhirter, and C. Harwell of the U.S. Forest Service (USFS) and B. Bow- en, J. Spaeth, B. Zuber, J. Bishop, and S. Jackson of Mickle et al.: Hatch window for Alosa alabamae 511 the University of Southern Mississippi (USM) for field assistance. We thank the USM Department of Biologi- cal Sciences and the USFS for providing vehicles and boats. Funding was provided by the National Marine Fisheries Service, NOAA (0-2003-SERl), the USFS Southern Research Station (SRS 03-CA-11330127-262), the Mississippi Department of Wildlife, Fisheries, and Parks (SWG04), the American Sportfishing Association, Fish America Foundation (FAF-4078R), and USM. Lo- gistical support was provided by R. Long, A. Knapp, C. Purtlebaugh, and J. Jackson of the Florida Fish and Wildlife Conservation Commission. All methods and procedures were approved by the USM Institutional Animal Care and Use Committee. Literature cited Alerstam, T., A. Hedenstrom, and S. Akesson. 2003. Long-distance migration: evolution and determi- nants. Oikos 103:247-260. Armstrong, J. D., P. S. Fallon-Cousins, and P. J. Wright. 2004. The relationship between specific dynamic action and otolith growth in pike. J. Fish Biol. 64:739-749. Berthold, P. 2001. Bird migration: a general survey, 2"^ ed. (H.-G. Bau- er and V. Westhead, transl.), 272 p. Oxford Univ. Press, New York. Bowen, B. R. 2005. Alabama shad phylogeography. M.S. thesis, 95 p. Dep. Biol., Univ. Southern Mississippi, Hattiesburg, MS. Bowen, B. R., B. R. Kreiser, P. F. Mickle, J. F. Schaefer, and S. B. Adams. 2008. Phylogenetic relationships among North American Alosa species (Clupeidae). J. Fish Biol. 72:1188-1201. Campana, S. E., A. J. Gagne, and J. Munro. 1987. Otolith microstructure of larval herring (Clupea harengus): image or reality? Can. J. Fish. Aquat. Sci. 44: 1922-1929. Castro-Santos, T., and B. H. Letcher. 2010. Modeling migratory energetics of Connecticut River American shad (Alosa sapidissima): implications for the conservation of an iteroparous anadromous fish. Can. J. Fish. Aquat. Sci. 67:806-830. Dynesius, M., and C. Nilsson. 1994. Fragmentation and flow regulation of river systems in the northern third of the world. Science 266:753-762. Ely, P. C., S. P. Young, and J. J. Isley. 2008. Population size and relative abundance of adult Alabama shad reaching Jim Woodruff Lock and Dam, Apalachicola River, Florida. North Am. J. Fish. Man- age. 28:827-831. Freeman, M. C., Z. H. Bowen, K. D. Bovee, and E. R. Irwin. 2001. Flow and habitat effects on juvenile fish abun- dance in natural and altered flow regimes. Ecol. Appl. 11:179-190. Geffen, A. J. 1982. Otolith ring deposition in relation to growth rate in herring (Clupea harengus) and turbot (Scophthalmus maximus) larvae. Mar. Biol. 71:317-326. 1992. Validation of otolith increment deposition rate. In Otolith microstructure examination and analysis (D. K. Stevenson and S. E. Campana, eds.), p. 101-113. Can. Spec. Publ. Fish. Aquat. Sci. 117. Gross, M. R. 1987. Evolution of diadromy in fishes. Am. Fish. Soc. Symp. 1:14-25. Hewitt, D. A. 2003. Abundance and migratory patterns of anadromous fish spawning runs in the Roanoke River, North Caro- lina. M.S. thesis, 113 p. North Carolina State Univ., Raleigh, NC. [Available at website.] Hodgson, S., and T. P. Quinn. 2002. The timing of adult sockeye salmon migration into fresh water: adaptations by populations to prevailing thermal regimes. Can. J. Zool. 80:542—555. Kinnison, M. T., M. J. Unwin, A. P. Hendry, and T. P. Quinn. 2001. Migratory costs and the evolution of egg size and number in introduced and indigenous salmon popula- tions. Evolution 55:1656-1667. Laurence, G. C., and R. W. Yerger. 1967. Life history studies of the Alabama shad, A/osa ala- bamae, in the Apalachicola River, Florida. Proc. Annu. Conf SEAFWA 20:260-273. [Available at website.] Leggett, W. C., and R. R. Whitney. 1972. Water temperature and the migrations of American shad. Fish. Bull. 70:659-670. Limburg, K. E. 1996. Growth and migration of 0-year American shad (Alosa sapidissima) in the Hudson River estuary: oto- lith microstructural analysis. Can. J. Fish. Aquat. Sci. 53:220-238. MacLellan, S. E., and M. W. Saunders. 1995. A natural tag on the otoliths of Pacific hake (Mer- luccius productus) with implications for age validation and migration. In Recent developments in fish otolith research (D. H. Secor, J. M. Dean, and S. E Campana, eds.), p. 567-580. Univ. South Carolina Press, Colum- bia, SC. Maltais, E., G. Daigle, G. Colbeck, and J. J. Dodson. 2010. Spawning dynamics of American shad (Alosa sapi- dissima) in the St. Lawrence River, Canada-USA. Ecol. Freshw. Fish. 19:586-594. Mann, R. H. K. 1996. Environmental requirements of European non-sal- monid fish in rivers. Hydrobiologia 323:223-235. Maurice, K. R., R. W. Blye, P. L. Harmon, and D. Lake. 1987. Increased spawning by American shad coincident with improved dissolved oxygen in the tidal Delaware River. In Common strategies of anadromous and catad- romous fishes (M. J. Dadswell, R. J. Klauda, C. M. Mof- fitt, and R. L. Saunders, eds.), p. 79-88. Am. Fish. Soc., Bethesda, MD. McLean, R. B., P. T. Singley, D. M. Lodge, and R. A. Wallace. 1982. Synchronous spawning of threadfin shad. Copeia 4:952-955. Meadows, D. W, S. B. Adams, and J. F. Schaefer. 2006. Threatened fishes of the world: Alosa alabamae (Jor- dan and Evermann, 1896) (Clupeidae). Environ. Biol. Fish. 82:73-174. Mettee, M. R, and P. E. O’Neil. 2003. Status of Alabama shad and skipjack herring in Gulf of Mexico drainages. Am. Fish. Soc. Symp. 35:157-170. 512 Fishery Bulletin 114(4) Mickle, P. F., J. F. Schaefer, S. B. Adams, and B. R. Kreiser. 2010. Habitat use of age 0 Alabama shad in the Pascagou- la River drainage, USA. Ecol. Freshw. Fish 19:107—115. Mills, J. F. 1972. Biology of the Alabama shad in Northwest Florida. Florida Dep. Nat. Resour. Mar. Res. Lab., Tech. Ser. 68, 24 p. [Available from Florida Fish and Wildlife Conser- vation Commission, Farris Bryant Bldg., 620 S. Meridian St., Tallahassee, FL 32399-1600.] Moksness, E., and V. Wespestad. 1989. Ageing and back-calculating growth rates of Pacific herring, Clupea pallasii, larvae by reading daily otolith increments. Fish. Bull. 87:509-513. Olney, J. E., and R. S. McBride. 2003. Intraspecific variation in batch fecundity of Ameri- can shad: revisiting the paradigm of reciprocal latitudi- nal trends in reproductive traits. Am. Fish. Soc. Sym. 35:185-192,. Quinn, T. R, and D. J. Adams. 1996. Environmental changes affecting the migratory timing of American shad and sockeye salmon. Ecology 77:1151-1162. Parsons, K. M., and K. M. Peters. 1989. Age determination in larval and juvenile sheeps- head, Arc/iosaTjgzis prohatocephalus. Fish. Bull. 87:985- 988. R Core Team. 2015. R: a language and environment for statistical com- puting. R Foundation for Statistical Computing, Vien- na, Austria. [Available at website, accessed December 2015.] Roff, D. A. 1988. The evolution of migration and some life his- tory parameters in marine fishes. Environ. Biol. Fish. 22:133-146. Secor, D. H., J. M. Dean, and E. H. Laban. 1992. Otolith removal and preparation for microstructur- al examination. In Otolith microstructure examination and analysis (D. K. Stevenson and S. E. Campana, eds.), p. 19-57. Can. Spec. Publ. Fish. Aquat. Sci. 117. Simkiss, K. 1974. Calcium metabolism of fish in relation to ageing. In The ageing of fish (T. B. Bagenal, ed.), p. 1-12. Unwin Brothers Ltd., London. Turner, R. E. 1990. Landscape development and coastal wetland losses in the northern Gulf of Mexico. Am. Zool. 30:89-105. Walter, J. F., and J. E. Olney. 2003. Feeding behavior of American shad during spawn- ing migration in the York River, Virginia. Am. Fish. Soc. Symp. 35:201-209. 513 Acknowledgment of reviewers The editorial staff of the Fishery Bulletin would like to acknowledge the scientists who reviewed manuscripts in 2015-16. Their contributions have helped ensure the publication of quality science. Kenneth Able Andre Afonso Valerie Allain Joel Anderson Dominic Andradi-Brown Peter Auster Ronnie Baker Joseph Ballenger Earl Becker Daniel Beckman Igor Belkin Hugues Benoit Keith Bigelow Joseph Bizzarre Aaron Blake Bodil Bluhm Katherine Bockrath James Bohnsack John Boreman Jr. Richard Brill Jon Brodziak Richard Brodeur Elizabeth Brooks Thierry Brule Demarest Jeffrey Buckel Michael Burton Carlos Bustamante Diaz Steven Cadrin Matthew Campbell John Carlson Christopher Chizinski John Claydon Rui Coelho Jeremy Collie Richard Condrey David Conover Jason Cope Kevin Craig Daniel Curran Tanya Darden David Demer Sandra Diamond-Tissue Edward Dick Jennifer Doerr Marina Dolbeth Jynessa Dutka-Gianelli Katy Echave Nelson Ehrhardt Brad Erisman Kristen Ferry Dariusz Fey Daniel Foster G. Mark Fowler Chris Francis James Franks Gregory Fulling Yongwen Gao James Gartland Nicholas Gidmark Jenny Giles John Gold John Graves Wayne Hajas Norman Hall Scott Hamilton Dana Hanselman Patrick Harris Alastair Harry Melissa Head Daniel Hennen Ronald Hill Thomas Hoff Kim Holland Lisa Hollensead Hsiang-Wen Huang Peter Hulson Robert Humphreys Jr. J. Jeffery Isely Larry Jacobson Josef Jech Christian Jorgensen Stephen Karl Craig Kastelle Aimee Keller David Kerstetter Holly Kindsvater Nikolai Klibansky Jeff Kneebone Donald Kobayashi Stanislaw Kotwicki Matthew Kupchik Thomas Lankford Cristiano Lazoski Christopher Legault Jesse Leland Rosangela Lessa Karin Limburg Mary Livingston Milton Love Carlos Lozano Gavin Macaulay Gustavo Macchi Anna Malek Mercer John Manderson Jeffrey Marliave Rich McBride Edward McGinley John McGovern Dianne McLean Dwayne Meadows Rick Methot Jonathan Mitchell Bradley Moore Alexia Morgan David Morgan Kentaro Morita Shokoh Morita Bruce Mundy Thomas Munroe Debra Murie Brenda Norcross Cornelia Oedekoven James Orr Ferran Palero Andrew Piercy Tyler Pigler Andre Punt Nuno Queiroz Chet Rakocinski R. Anne Richards Martin Robards Eric Robillard Cara Rodgveller Paul Rogers Christopher Rooper 514 Fishery Bulletin 1 14(4) Skyler Sagarese Ian Stewart Ingrid van Putten David Sampson Allan Stoner James Vasslides William Satterthwaite James Sulikowski Douglas Vaughan Amy Schueller Yonat Swimmer Chugey Sepulveda Ian Taylor William Walsh Seifu Seyoum Benjamin Walther Nancy Shackell Verena Trenkel Robert Ward Gary Shepherd Edward Trippel William Weber Colin Simpfendorfer Nathan Truelove Chantel Wetzel Michael Simpkins Victor Tuset Andreas Winter John Skalski Wann-Nian Tzeng Robert Sluka Steven Smith Clara Ulrich Mary Yoklavich Susan Sogard Audrey Vails Yuying Zhang Dave Somerton Kray Van Kirk Joseph Zydlewski 515 Fishery Bulletin Guidelines for authors Contributions published in Fishery Bulletin describe original research in marine fishery science, fishery en- gineering and economics, as well as the areas of ma- rine environmental and ecological sciences (including modeling). Preference will be given to manuscripts that examine processes and underlying patterns. Descriptive reports, surveys, and observational papers may occa- sionally be published but should appeal to an audience outside the locale in which the study was conducted. Although all contributions are subject to peer reviev/, responsibility for the contents of papers rests upon the authors and not on the editor or publisher. Submission of an article implies that the article is original and is not being considered for publication elsewhere. Plagiarism and double publication are considered serious breaches of publication ethics. To verify the originality of the research in papers and to identify possible previous publication, manuscripts may be screened with plagiarism-detection software. Manuscripts must be written in English; authors whose native language is not English are strongly advised to have their manuscripts checked by Eng- lish-speaking colleagues before submission. Once a paper has been accepted for publication, on- line publication takes approximately 3 weeks. There is no cost for publication in Fishery Bulletin. Types of manyscripts accepted by the joyrnal Articles generally range from 20 to 30 double-spaced typed pages (12-point font) and describe an original contribution to fisheries science, engineering, or eco- nomics. Tables and figures are not included in this page count, but the number of figures should not ex- ceed one fi.gure for every four pages of text. Articles contain the following divisions: abstract, introduc- tion, methods, results, and discussion. Short contributions are generally less than 20 double spaced typed pages (12-poiiit font) and, like articles, describe an original contribution to fisheries science. They follow the same format as that for articles: ab- stract, introduction, results and discussion, but the results and discussion sections may be com- bined. They are distinguished from full articles in that they report a noteworthy new observation or discov- ery— such as the first report of a new species, a unique finding, condition, or event that expands our knowledge of fisheries science, engineering or economics — and do not require a lengthy discussion. Companion articles are presented together and pub- lished together as a scientific contribution. Both arti- cles address a closely related topic and may be articles that result from a workshop or conference. They must be submitted to the journal at the same time. Rewiew articles generally range from 40 to 60 dou- ble-spaced typed pages (12-point font) and address a timely topic that is relevant to all aspects of fisheries science. They should be forward thinking and address novel views or interpretations of information that en- courage new avenues of research. They can be reviews based on the outcome from thematic workshops, or con- tributions by groups of authors who want to focus on a particular topic, or a contribution by an individual who chooses to review a research theme of broad interest to the fisheries science community. A review article will include an abstract, but the format of the article per se will be up to the authors. Please contact the Scientific Editor to discuss your ideas regarding a re- view article before embarking on such a project. Preparation of manuscript Title page should include authors’ full names, mailing addresses, and the senior author’s e-mail address. Abstract should be limited to 200 words (one-half typed page), state the main scope of the research, and empha- size the authors conclusions and relevant findings. Do not review the methods of the study or list the contents of the paper. Because abstracts are circulated by ab- stracting agencies, it is important that they represent the research clearly and concisely. General tent must be typed in 12-point Times New Ro- man font throughout. A brief introduction should con- vey the broad significance of the paper; the remainder of the paper should be divided into the following sec- tions: Materials and methods, Results, Discussion, and Acknowledgments. Headings within each section must be short, reflect a logical sequence, and follow the rules of subdivision (i.e., there can be no subdivision with- out at least two subheadings). The entire text should be intelligible to interdisciplinary readers; therefore, all acronyms, abbreviations, and technical terms should be written out in full the first time they are mentioned. Abbreviations should be used sparingly because they are not carried over to indexing databases and slow readability for those readers outside a discipline. They should never be used for the main subject (species, method) of a paper. For general style, follow the U.S. Government Print- ing Office Style Manual (2008) [available at website] and Scientific Style and Format: the CSE Manual for Authors, Editors, and Publishers (2014, 8*^^ ed.) pub- lished by the Council of Science Editors. For scientific nomenclature, use the current edition of the American Fisheries Society’s Common and Scientific Names of 516 Fishery Bulletin 114(4) Fishes from the United States, Canada, and Mexico and its companion volumes {Decapod Crustaceans, Mollusks, Cnidaria and Ctenophora, and World Fishes Impor- tant to North Americans). For species not found in the above mentioned AFS publications and for more recent changes in nomenclature, use the Integrated Taxonom- ic Information System (ITIS) (available at website), or, secondarily, the California Academy of Sciences Cata- log of Fishes (available at website) for species names not included in ITIS. Common (vernacular) names of species should be lowercase. Citations must be given of taxonomic references used for the identification of specimens. For example, “Fishes were identified accord- ing to Collette and Klein-MacPhee (2002); sponges were identified according to Stone et al. (2011).” Dates should be written as follows: 11 November 2000. Measurements should be expressed in metric units, e.g., 58 metric tons (t); if other units of measure- ment are used, please make this fact explicit to the reader. Use numerals, not words, to express whole and decimal numbers in the general text, tables, and fig- ure captions (except at the beginning of a sentence). For example: We considered 3 hypotheses. We collected 7 samples in this location. Use American spelling. Re- frain from using the shorthand slash ( / ), an ambiguous symbol, in the general text. Word usage and grammar that may be useful are the following: • Aging For our journal the word aging is used to mean both age determination and the aging pro- cess (senescence). Authors should make clear which meaning is intended where ambiguity may arise. • Fish and fishes For papers on taxonomy and biodi- versity, the plural of fish is fishes, by convention. In all other instances, the plural is fish. Examples: The fishes of Puget Sound [biodiversity is indicated] ; The number of fish caught that season [no emphasis on biodiversity]; The fish were caught in trawl nets [no emphasis on biodiversity]. The same logic applies to the use of the words crab and crabs, squid and squids, etc. • Sex For the meaning of male and female, use the word sex, not gender. • Participles As adjectives, participles must modify a specific noun or pronoun and make sense with that noun or pronoun. Incorrect: Using the recruitment model, estimates of age-1 re- cruitment were determined. [Estimates were not using the recruitment model.] Correct: Using the recruitment model, we determined age- 1 estimates of recruitment. [The participle now modifies the word we, i.e., those who were using the model.] Incorrect: Based on the collected data, we concluded that the mortality rate for these fish had increased. [We were not based on the collected data.] Correct: We concluded on the basis of the collected data that the mortality rate for these fish had increased. [Eliminate the participle and replace it with the adverbial phrase on the basis of.] Equations and mathematical symbols should be set from a standard mathematical program (MathType) and tool (Equation Editor in MS Word). LaTex is accept- able for more advanced computations. For mathemati- cal symbols in the general text (a, y}, Ji, ±, etc.), use the symbols provided by the MS Word program and itali- cize all variables, except those variables represented by Greek letters. Do not use photo mode when creating these symbols in the general text. Number equations (if there are more than 1) for fu- ture reference by scientists; place the number within parentheses at the end of the first line of the equation. Literature cited section comprises published works and those accepted for publication in peer-reviewed journals (in press). Follow the name and year system for cita- tion format in the “Literature cited” section (that is to say, citations should be listed alphabetically by the au- thors’ last names, and then by year if there is more than one citation with the same authorship. A list of abbreviations for citing journal names can be found at website. Authors are responsible for the accuracy and com- pleteness of all citations. Literature citation format: Author (last name, followed by first-name initials). Year. Title of article. Abbreviated title of the journal in which it was published. Always include number of pages. For a sequence of citations in the general text, list chrono- logically: (Smith, 1932: Green. 1947; Smith and Jones, 1985). Digital object identifier (doi) code ensures that a publication has a permanent location online. Doi code should be included at the end of citations of published literature. Authors are responsible for submitting ac- curate doi codes. Faulty codes will be deleted at the page-proof stage. Cite all software, special equipment, and chemical solutions used in the study within parentheses in the general text: e.g., SAS, vers. 6.03 (SAS Inst., Inc., Cary, NO. Guidelines for authors 517 Footnotes are used for all documents that have not been formally peer reviewed and for observations and personal communications. These types of references should be cited sparingly in manuscripts submitted to the journal. All reference documents, administrative reports, internal reports, progress reports, project reports, contract reports, personal observations, personal communications, unpublished data, manuscripts in re- view, and council meeting notes are footnoted in 9 pt font and placed at the bottom of the page on which they are first cited. Footnote format is the same as that for formal literature citations. A link to the online source (e.g., [http://www/ , accessed July 2007.]), or the mailing address of the agency or department holding the document, should be provided so that readers may obtain a copy of the document. Tables are often overused in scientific papers; it is sel- dom necessary to present all the data associated with a study. Tables should not be excessive in size and must be cited in numerical order in the text. Headings should be short but ample enough to allow the table to be in- telligible on its own. All abbreviations and unusual symbols must be ex- plained in the table legend. Other incidental comments may be footnoted with italic numeral footnote markers. Use asterisks only to indicate significance in statistical data. Do not type table legends on a separate page; place them above the table data. Do not submit tables in photo mode. • Notate probability with a capital, italic P. • Provide a zero before all decimal points for values less than one (e.g., 0.07). • Round all values to 2 decimal points. • Use a comma in numbers of five digits or more (e.g., 13,000 but 3000). Figures must be cited in numerical order in the text. Graphics should aid in the comprehension of the text, but they should be limited to presenting patterns rather than raw data. Figures should not exceed one figure for every four pages of text and must be labeled with the number of the figure. Place labels A, B, C, etc. within the upper left area of graphs and photos. Avoid placing labels vertically (except for the y axis). Figure legends should explain all symbols and abbre- viations seen in the figure and should be double-spaced on a separate page at the end of the manuscript. Line art and halftone figures should be submitted as pdf files with >800 dpi and >300 dpi, respectively. Color is allowed in figures to show morphological differences among species (for species identification), to show stain reactions, and to show gradations in temperature con- tours within maps. Color is discouraged in graphs, and for the few instances where color may be allowed, the use of color will be determined by the Managing Editor. Approved color figures should be submitted as TIFF or JPG files in CMYK format. • Capitalize the first letter of the first word in all la- bels within figures. • Do not use overly large font sizes in maps and for axis labels in graphs. • Do not use bold fonts or bold lines in figures. • Do not place outline rules around graphs. • Place a North arrow and label degrees latitude and longitude (e.g., 170°E) in all maps. • Use symbols, shadings, or patterns (not clip art) in maps and graphs. Supplementary materials that are considered essential, but are too large or impractical for inclusion in a paper (e.g., metadata, figures, tables, videos, websites), may be provided at the end of an article. These materials are subject to the editorial standards of the journal. A URL to the supplementary material and a brief ex- planation for including such material should be sent at the time of initial submission of the paper to the journal. • Metadata, figures, and tables should be submitted in standard digital format (Word docx) and should be cited in the general text as (Suppl. Table, Suppl. Fig., etc.). • Websites should be cited as (Suppl. website) in the general text and be made available with doi code (if possible) at the end of the article. • Videos must not be larger than 30 MB to allow a swift technical response for viewing the video. Au- thors should consider whether a short video uniquely captures what text alone cannot capture for the un- derstanding of a process or behavior under exami- nation in the article. Supply an online link to the location of the video. Copyright law does not apply to Fishery Bulletin, which falls within the public domain. However, if an author reproduces any part of an article from Fishery Bulle- tin, reference to source is considered correct form (e.g.. Source: Fish. Bull. 97:105). Failure to follow these guidelines and failure to correspond with editors in a timely manner will delay publication of a manuscript. Submission of manuscript Submit manuscript online at the ScholarOne website. Commerce Department authors should submit papers under a completed NOAA Form 25-700. For further de- 518 Fishery Bulletin 1 14(4) tails on electronic submission, please contact the As- sociate Editor, Kathryn Dennis, at kathryn.dennis@noaa.gov When requested, the text and tables should be submit- ted in Word format. Figures should be sent as separate PDF files (preferred), TIFF files, or JPG files. Send a copy of figures in the original software if conversion to any of these formats yields a degraded version of the figure. Questions? If you have questions regarding these guidelines, please contact the Managing Editor, Sharyn Matriotti, at sharyn.matriotti@noaa.gov Questions regarding manuscripts under review should be addressed to Kathryn Dennis, Associate Editor. Fishery Bulletin Subscription form Superintendent of Documents Publications Order Form *5178 I I YESj please send me the following publications: Subscriptions to Fishery Bulletin for $32.00 per year ($44.80 foreign) The total cost of my order is $ . Prices include regular domestic postage and handling and are subject to change. (Company or Personal Name) (Please type or print) (Additional address/attention line) (Street address) (City, State, ZIP Code) (Daytime phone including area code) (Purchase Order No.) Charge your order. IT’S EASY! Please Choose Method of Payments I I Check payable to the Superintendent of Documents I I GPO Deposit Account your orders (202) 512-2104 (Credit card expiration date) I I VISA or MasterCard Account (Authorizing Signature) Mail to: Superintendent of Documents Thank you for P.O. Box 979050, St. Louis, Missouri 63197 your order! Tel. no. (U.S. callers): 866-512-1800 Tel. no. (International callers): 1-202-512-1800 Also available online at http://bookstore. gpo.gov/actions/GetPublication. do?stocknumber=703-023-00000-2 email: ContactCenter@gpo.gov SMtTHSONIAN LIBRARIES