u ‘Pi2- fS3 fISH U.S. Department of Commerce Volume 115 Number 1 January 2017 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 V' ★ / 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, Hawaii 96818 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. 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.fisherybulletin.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 ¥olyme 115 Number 1 January 2017 Fishery Bulletin Contents Artkies 1-12 Steinhorst, Kirk, Timothy Copeland, Michael W. Ackerman, William C. Schrader, and Eric C. Anderson Abundance estimates and confidence intervals for the run composition of returning salmonids 13-26 Lopez Quintero, Freddy Omar, Javier E. Contreras-Reyes, Rodrigo Wiff, and Reinaldo B. Arellano-Valle Flexible Bayesian analysis of the von Bertalanffy growth function with the use of a log-skew-f distribution 27-41 Dapp, Derek R., Charlie Huveneers, Terence I. Walker, John Mandeiman, David W. Kerstetter, and Richard D. Reina Using logbook data to determine the immediate mortality of blue sharks (Prionoce glauca) and tiger sharks (Galeocerdo cuvier) caught in the commercial U.S. pelagic longline fishery 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. Short contribution 42-49 Kilada, Raouf, Joel B. Webb, Kevin W. McNeel, Laura M. Slater, Quinn Smith, and Jayde Ferguson Preliminary assessment of a direct age-determination method for 3 commercially important crustaceans from Alaska Articles 50-59 Driggers III, William B., Matthew D. Campbell, Kristin M. Hannan, Eric R. Hoffmayer, Christian M. Jones, Lisa M. Jones, and Adam G. Pollack Influence of bait type on catch rates of predatory fish species on bottom longline gear in the northern Gulf of Mexico II Fishery Bulletin 1 14(4) 60-73 Colmenero, Ana I., Victor M. Tuset, and Pilar Sanchez Reproductive strategy of white anglerfish (Lophius piscatorius) in Mediterranean waters: implications for management 74-88 Mace III, Marvin M., and Lawrence P. Rozas Population dynamics and secondary production of juvenile white shrimp iLitopenaeus setiferus) along an estuarine salinity gradient 89-100 Dell'Apa, Andrea, Maria Grazia Pennine, and Christopher Bonzek Modeling the habitat distribution of spiny dogfish iSqualus acanthios), by sex, in coastal waters of the northeastern United States 101-116 Hueter, Robert E., John P. Tyminski, John J. Morris, Alexei Ruiz Abierno, and Jorge Angulo Valdes Horizontal and vertical movements of longfin makes Usurus paucus) tracked with satellite-linked tags in the northwestern Atlantic Ocean Short contribution 117-124 Ortega-Garcia, Sofia, Chugey Sepulveda, Scott Aalbers, Ulianov Jakes-Cota, and Ruben Rodriguez- Sanchez Age, growth, and length-weight relationship of roosterfish (Nemotistius pectoralis) in the eastern Pacific Ocean 125-128 Guidelines for authors 1 National Marine Fisheries Service NOAA Fishery Bulletin fy- established in 1881 Spencer F. Baird First U.S. Commissioner of Fisheries and founder of Fishery Bulletin Abundance estimates and confidence intervals for the run composition of returning salmonids Email address for contact author: tim.copeland@idfg.idaho.gov Abstract — In 2-stage fishery sam- pling, abundance is often estimated by using a primary sampling gear and total abundance is then parti- tioned into groups of interest by ap- plying data on composition derived from a secondary sampling gear. However, the literature is sparse on statistical properties of estimates of run composition. We examined the accuracy and precision of estima- tors of composition of wild steelhead (Oncorhynchus mykiss) in the Snake River, in the Pacific Northwest. We simulated estimators, using pooled and time-stratified data. We com- pared confidence intervals (CIs) de- termined on the basis of asymptoti- cal normality or a 2-stage bootstrap method. Stratified estimators were unbiased, except in a few cases. Joint CIs (all groups considered si- multaneously) had coverages near nominal. Conversely, pooled estima- tors performed poorly; the propor- tion of biased estimates increased as the number of groups estimated increased. Using empirical data, we show that CIs met precision goals for most groups. Half-widths of CIs decreased and stabilized as the number sampled and group abun- dance increased. In complex scenar- ios, estimates of small groups will yield poor precision and some may be biased, but a stratified estimate with a conservative joint Cl can be of practical use. The 2-step bootstrap approach is flexible and can incorpo- rate other sources of variability or sampling constraints. Manuscripts submitted 22 December 2015. Manuscripts accepted 25 August 2016. Fish. Bull. 115:1-12 (2017). Online publication date: 18 October 2016. doi: 10.7755/FB.115.1.1 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. Kirk Steinhorst' Timothy Copeland (contact author)^ Michael W. Ackerman^ William C. Schrader^ Eric C. Anderson^ ' Department of Statistical Science University of Idaho 415A Brink Hall 875 Perimeter Drive Moscow, Idaho 83844-1104 2 Idaho Department of Fish and Game 1414 East Locust Lane Nampa, Idaho 83686 In 2-stage sampling for fisheries monitoring and research, abundance is estimated with a primary sam- pling gear and then partitioned into groups of management interest by applying compositional data (e.g., species, stock, sex, age, and size) derived from a secondary sampling gear. For example, biological samples obtained from gillnetting or electro- fishing can be used to allocate abun- dance estimates from hydroacoustic counts to species (e.g., Tarbox and Thorne, 1996; Pritt et al., 2013; Rud- stam et ah, 2013; Hughes and High- tower, 2015). Alternatively, a more highly controlled sampling regime can be instituted by counting fish as they move past barriers (e.g., weirs or dams) and by collecting biological samples or data from some portion of the fish in order to partition counts (e.g., Wagner, 2007; Campbell et ah, 2012). However, the complexities of fishery sampling programs and the relevant groups into which the fish are parsed present difficulties for es- timating precision of the generated point estimates. Steelhead {Oncorhynchus mykiss) ^ Eagle Fish Genetics Laboratory Idaho Department of Fish and Game Pacific States Marine Fisheries Commission 1800 Trout Road Eagle, Idaho 83616 '' Fisheries Ecology Division National Marine Fisheries Service Southwest Fisheries Science Center, NOAA 1 10 Shaffer Road Santa Cruz, California 95060 are an important cultural, economic, and recreational resource in the Pa- cific Northwest of the United States. After the construction of hydro- electric dams on the Columbia and Snake rivers during the late 1960s and early 1970s, the abundance and survival of steelhead in the Snake River decreased (Raymond, 1988). In response, steelhead within the Snake River basin were listed as threatened under the Endangered Species Act in 1997. In recent years, abundances have increased slightly. However, the increase has been dominated by fish produced in hatcheries (intended to mitigate for reduced harvest opportu- nities and to supplement natural pop- ulations), while the returns of steel- head born in the natural environment remain critically low (Ford, 2011). Fishery biologists need to know how many wild versus hatchery-produced steelhead return in order to manage fisheries effectively, as well as to as- sess the conservation status of wild populations. Further, for wild fish, we need to know the numbers of fish returning by sex, age, and stock to inform viability analyses. 2 Fishery Bulletin 1 15(1) We collected data on the run composition of adult steelhead as they migrated past Lower Granite Dam (LGD) on the Snake River, 695 km from the ocean. Adults returning from the Pacific Ocean to spawn in tributaries of the Snake River must ascend fish ladders at 8 dams during their migration; Lower Granite Dam is the final dam they encounter before dispersing to spawn. An observation window on the LGD fish ladder (the primary sampling “gear”) allows the enumeration of fish by species as they migrate upstream. A trap- ping facility (a diversion gate in the fish ladder with chutes leading to a holding tank) located above the ob- servation window (the secondary sampling gear) allows the interception of fish and the collection of biological data (Harmon, 2003). Counting and sampling returning adult steelhead at the dam provide the data for calcu- lating run composition (Schrader et al.^). Surprisingly, the primary literature is sparse on the statistical prop- erties of estimates derived with current methods and applied to run composition. In our study, we examined the properties of estima- tors of fish composition and confidence intervals (CIs) derived from weightings of counts of fish at the obser- vation window, data on origin (wild versus hatchery) obtained from the samples taken at the trapping facili- ty, and compositional data (sex, age, and genetically de- fined stock) collected from wild fish subsampled at the trapping facility. We considered counts at the observa- tional window to provide a census of fish passing the dam. Initially, we assumed trapping rates (proportion of time the trap was open) could be precisely controlled to obtain a constant proportion of the fish throughout the run and, therefore, that data could be pooled across time to estimate abundance. However, logistical issues that affected trapping rates through time led us to in- vestigate temporally stratified estimators. Individual CIs (for each group considered independently) and joint CIs (for all groups within a variable of interest considered simultaneously) were derived 1) by using closed-form asymptotically normal equations or 2) a 2-step bootstrap sampling method, by origin (hatchery or wild) of the fish, by using compositional data collect- ed from wild fish. Using simulations, we compared the options for developing accurate estimates of abundance and CIs with good coverage; we then applied the pre- ferred method from the simulations to empirical data to develop guidance for sampling and interpreting fish- eries data on fish composition. Materials and methods We used data to describe the abundance and compo- sition of wild adult steelhead migrating past LGD to 1 Schrader, W. C., M. P. Corsi, P. Kennedy, M. W. Ackerman, M. R. Campbell, K. K. Wright, and T. Copeland. 2013. Wild adult steelhead and Chinook salmon abundance and com- position at Lower Granite Dam, spawn year 2011. 2011 annual report. Idaho Dep. Fish Game, IDFG Rep. 13-15, 89 p. [Available at website.] spawn in the Snake River during spring 2011. Spawn- ing year (SY) 2011 is defined as the year when adult steelhead migrate past LGD between 1 July 2010 and 30 June 2011. Although all steelhead in the Snake River basin spawn in the spring, the majority migrate past LGD during the previous fall, and a smaller por- tion migrates during the spring just before spawning. Later in this section, we describe the data set and es- timation procedures and the simulations developed to test the bias of the estimators and the coverage of the associated CIs. A complete description of the collec- tion methods and data used in this study is given by Schrader et al.^ Data collection Primary sampling stage (counts from the observation win- dow) Adult steelhead were counted as they passed a viewing window located in the LGD fish ladder, which they must ascend to migrate upriver. Counts of fish observed from the window were conducted during a majority of the year and occurred daily at 0400-2000 Pacific Time. Counts from videos were used in lieu of counts from the window in November, December, and March and occurred at 0600-1600. Most fish pass the window during the 10-16 h of daylight when counts are made. The ladder is drained and closed in January and February, and, as a result, adult steelhead can- not migrate upriver during those months. Count data were downloaded from the U.S. Army Corps of Engi- neers website (website). The steelhead count consists of all fish >30 cm in fork length identified as O. mykiss. Daily counts were aggregated on a weekly basis. We further combined weeks into longer time periods (up to 2 months) if few fish (<75 individuals) were passing the window during a week — a level observed typically early and late in the migration season. Secondary sampling stage (trapping rates) Trapping rates were determined by a committee of co-managers balancing sampling requirements for multiple projects with fish handling concerns. The trap is operational 24 h/day and the trapping rate determines how long a trap gate remains open 4 times/h, such that a daily systematic sample (by time) is taken from the fish as- cending the fish ladder. Thus, the trapping rate (pro- portion of an hour that the trap gate is open) approxi- mates the desired proportion of the population to be sampled. Trapping rates are typically 10-20%; for the majority of the SY2011 run, it was set at 10%. Trapped fish were anesthetized and examined to determine whether they were of hatchery or wild or- igin. In the Snake River basin, most hatchery-origin steelhead have a clipped adipose fin; however, some are released with an intact adipose fin to supplement natural populations. Therefore, undipped steelhead were examined for the presence of dorsal or ventral fin erosion, which often occurs in hatchery-reared fish (Latremouille, 2003). Undipped hatchery fish may also be identified by the presence of a coded wire tag, by Steinhorst et al.: Estimates of abundance for the run composition of salmonids 3 parentage determined genetically (Steele et al., 2013), or by a ventral-fin clip. Genotyping procedures for par- entage were conducted after the trapping season and gives accuracy rates approaching 100% (Steele et al., 2013). Fish not determined to be of hatchery origin were treated as wild fish. set containing sex, age, and stock for a subsample of wild fish that were trapped (for compositional data). These 3 data sets were used to produce estimates and CIs for the number of wild fish by sex or age or stock. Estimator and confidence intervals Subsampling of trapped wild steelhead Scale and tissue samples were then taken from a systematic subsample of trapped fish deemed wild. Percentages of the wild steelhead that were subsampled averaged around 50% during this study. Scale samples were used to deter- mine age on the basis of visual examination of scale annuli. Age data collected at LGD were used to assign returning adults back to a brood year (BY, the year in which their parents spawned). Tissue samples from the anal fin were used to deter- mine sex and the stock of origin. Stock composition was determined by using individual assignment, a method of genetic stock identification (Pella and Milner, 1987; Shaklee et al., 1999) based on single nucleotide poly- morphisms (SNPs). Adults were screened at 187 SNPs and with a sex-specific allelic discrimination assay (Campbell et al., 2012). Only individuals that were genotyped at >90% of SNPs were included. We used the maximum likelihood framework implemented in the program gsi_sim (Anderson et al., 2008; Anderson, 2010) to assign individuals to a stock. Each fish was assigned to the stock in which the probability of its genotype occurring was greatest by using the allocate- sum procedure (Wood et al., 1987). We did not attempt to identify out-of-basin strays. For this study, we as- sumed that the stock was determined without error (a future study will examine uncertainty in genetic as- signments). In essence, we treated the genetic data in the same way as we did for the age data. Ackerman et al.^ defined 10 genetically determined stocks used for assignments at LGD. The locations of these stocks included 1) the upper Salmon River (UPS); 2) Middle Fork Salmon River (including Chamberlain and Bargamin creeks) (MFS); 3) South Fork Salmon River (SFS); 4) lower Salmon River (LOS); 5) upper Clearwater River (Lochsa and Selway rivers) (UPC) ; 6) South Fork Clearwater River (including Clear Creek) (SFC); 7) lower Clearwater River (LOC); 8) Imnaha River (IMN); 9) Grande Ronde River (GRR); and 10) Tucannon River, Asotin Creek, and other tributaries to the Snake River downstream of the Clearwater River confluence (LSN). The sampling design produced 3 data sets: 1) a cen- sus of numbers of fish returning to and migrating past the dam (window counts), 2) a hatchery-versus-wild data set for all trapped fish (trap data), and 3) a data 2 Ackerman, M. W., N. V. Vu, J. McCane, C. A. Steele, M. R. Campbell, A. P. Matala, J. E. Hess, and S. R. Narum. 2014. Chinook and steelhead genotyping for genetic stock identification at Lower Granite Dam. Project progress report. 2013 annual report. Idaho Dep. Fish Game, IDFG Rep. 14- 01, 60 p. [Available at website] Abundance of wild steelhead The window-count data provided the abundance of adult steelhead migrating past LGD, but our focus was on wild fish; therefore, we first had to partition the overall abundance estimate into a wild-versus-hatchery abundance estimate. The proportions of wild and hatchery steelhead changed over the season, but within each weekly or monthly stratum, proportions were assumed to be relatively con- stant. Given the window counts by strata, C^, C2,...,Cs, the number of wild steelhead (W) was estimated with the following equation: (1) where Pi,P2,---,Ps = estimates of the proportion of wild steelhead by stratum from the trap ^ data; and Ps = NJt, (or denoted p, for a pooled estimate). These and all subsequent notations are defined in Ta- ble 1. Given the fixed numbers of adults counted at the dam for each stratum, we found a Cl for the number of wild fish by using either an asymptotically normal interval or by a parametric bootstrap. The asymptoti- cally normal interval is given as the conditional probabilities for wild females and males for strata 1, s. These pro- portions were estimated from data obtained from the subsample of the trapped fish; for example, for females ■^Fs = ^Fs /(^Fs + ^Ms)- Given estimates of these prob- abilities from wild fish examined in the trap, we used the following equation to estimate female abundance: '^ = l:li4F.Wi=E?.,«F,P.C„ (4) and to estimate wild male abundance (5) For the pooled estimators, we dropped the summa- tion and subscript. Similar estimates were made for A (ages, BY2004-BY2008 in this study), G (genetically identified stocks), and AxG age groups for each stock. That is, the number of wild steelhead in any group is a weighted sum of the stratified window counts, in which the weights are estimates of the probabilities of being wild and being a member of a particular group, includ- ing any combinations of the compositional variables. To find CIs for these estimates, we had to account for the variability of both the trap data and the com- positional data. For the asymptotically normal inter- val, we used Goodman’s (1960) estimated variance of a product: 2 „2 „2 „2 -S' S- Ps (6) yielding where s|g = (1 - |-)iiFs G - ^Fs )/(^s - D- For the pooled case, we used the following equation: 2 yu 2 ,->22 Sp — - S. 2 „2 (8) where s|^ = (l-:^)TYp(l-'Rp)/(r.-l); and (9) TYp = the proportion of wild females from the pooled sample. Similar formulae for M and estimates of age, stock, and ages by stock follow in the next paragraph. The bootstrap process described previously for ob- taining the Cl for the number of wild fish was extended by adding a conditional bootstrap loop based on the sex, age, and stock of wild fish in the trap. We defined pseudoreplicates parametrically, using (F’s*,M*) ~ binomial(rg,(7Yps,'rYMs)| Tg.wild) and (10) r * F' * Ml \ (■^Fs Bootstrap values for the total number of wild females, Fj ,...,Fg, were determined with the following equation: Steinhorst et al.: Estimates of abundance for the run composition of salmonids 5 Stratum Figure 1 Simulated abundance of wild steelhead trout {Oncorhynchus mykiss) by weekly time strata for (A) 5 age groups and (B) 10 stocks, which are given generic designations, such as Stock A. (11) The percentile bootstrap Cl for the true number of wild females, [Lp, l/p], is determined by finding the lOOa/2 and 100(1 — -1) percentiles. Similarly, we calculated a bootstrap Cl for the number of wild males. Changing the binomial described previously to a multinomial, we generated B sets of ('ir;^i,...,'KAi),...,('KiB,...,'K^) to ob- tain bootstrap CIs for the true number of wild fish of ages 1, ..., A. We followed the same procedure for stocks 1, ..., G and AxG ages by stock (ages within stocks). To ensure accuracy across all groups being evaluated at a particular time, joint CIs for numbers of wild fish by sex or age or stock were calculated by using the meth- ods of Mandel and Betensky (2008). Simulations Although our estimators and CIs are straightfor- ward, we did not know their statistical properties. We designed a simulation of the sampling process to examine the properties of sex, age, and stock estima- tors and CIs, using the methods defined previously to analyze each simulated sample. We set the total pas- sage of steelhead similar to the SY2011 observed count (200,000 fish). We set parameter values for all bino- mial and multinomial distributions similar to those of the stratified estimates obtained from the SY2011 data (Suppl. Tables 1 and 2). The percentage of wild steelhead ranged from 20% to 50%. The trapping rate varied from 3% to 14%, and the subsampling rate varied from 35% to 100%. Abundance and composition of the simulated population varied over 27 tem- poral strata loosely based on the character of the wild steelhead run in Snake River during SY2011 (Fig. 1). For simplicity, age and stock proportions in the population were generated by assuming age and stock are independent variables; therefore, we multiplied age and stock proportions to find age-by-stock propor- tions of the steelhead run. We generated 500 samples from the popula- tion in the following manner. First, we simu- lated number of trapped fish itg) by generating binomial samples for each time stratum with the number of binomial trials equal to the number of fish returning during that stratum and with probability equal to the proportion of fish trapped within that stratum. Second, we simulated the number of trapped fish that were wild for each stratum by generating bino- mial samples with the number of trials equal to tg and with probability equal to the true proportion of wild fish for that stratum. The remaining trapped fish were of hatchery origin. From these numbers, we generated a sample of trapped fish with 2 columns: time stratum and wild versus hatchery. The length of this data set was the sum of the numbers of wild fish trapped across the time strata Third, we calculated the number of wild fish whose sex, age, and stock had been determined in each stratum by multi- plying the simulated number of wild fish trapped by the proportion subsampled. These numbers by stratum were the number of binomial or multinomial trials for sex or age or genetic stock (rg). For example, we found the number of sampled fish that were wild females for a stratum by generating binomial trials of size Tg with probability equal to the true proportion of wild females during that stratum and with the remainder being males. Knowing the random number of wild females and males trapped in each stratum, we put together a simulated subsample of fish by sex by forming a data set with two columns (for stratum and sex). The size of this sample was equal to the sum across the time strata of the numbers of handled fish, J2s=i^s- Similar samples were simulated for age, stock, and stock by age (500 for each). The simulation generated 2 types of data from each sampling iteration: 1) a randomly generated trap sample with a random number of wild and hatchery fish, and 2) a randomly generated compo- sitional sample with random numbers of females and males or numbers of fish of various ages or numbers of fish of various stocks. 6 Fishery Bulletin 1 15(1) Table 2 Number of simulations in which criteria for coverage levels for individual confidence intervals coverage were not met for combinations of estimator type (scenario). Simula- tions were conducted for 4 variables of interest with varying numbers of categories ik). Scenario Criterion Variable of interest (number of categories) Sex {k=2) Brood year (/j=5) Stock 0fe=10) Agexstock (/2=50) Pooled asymptotically normal Coverage <0.85 0 1 5 18 Coverage <0.80 0 0 3 6 Pooled parametric bootstrap Coverage <0.85 0 0 5 13 Coverage <0.80 0 0 3 2 Stratified asymptotically normal Coverage <0.85 0 1 0 12 Coverage <0.80 0 0 0 3 Stratified parametric bootstrap Coverage <0.85 0 0 0 8 Coverage <0.80 0 0 0 0 We obtained estimates of abundance and boot- strapped CIs for the groups of interest for every simu- lation iteration, using the window counts and the pri- mary and secondary samples, as explained previously. After 500 simulation iterations, we had 500 estimates of the total number of wild steelhead, female and male wild, wild by age, and wild by stock. We also had 500 individual and joint CIs for each estimate. We saved the estimates and CIs for subsequent evaluation. The evaluation of estimator performance was based on bias and Cl coverage. We computed bias as the mean of the simulated estimates minus the true value. We computed the coverage of any individual Cl by tal- lying the number of times the true population number fell inside the CL For the pooled and stratified estima- tors, we tallied the number of cases for which the bias was >5%. Likewise, we tallied the number of cases for which the coverages for estimators were <0.85 (consid- ered poor) and <0.80 (considered very poor). Analysis of SY201 1 data We evaluated precision using the preferred estimator (determined from the simulations) that was applied to real data from SY2011; these data had more irregulari- ties than the simulated data. For this application, we determined the age-by-stock proportions from the data, not as the product of age and stock proportions. We measured precision as the half-width of a (l-a)100% Cl expressed as a percentage of the point estimate (Find for individual CIs or Pjoj values for joint CIs). Re- searchers often set a stringent goal of a Cl half-width <10% of the estimate. For management purposes, it is recommended that salmon stocks have unbiased abun- dance estimates with a coefficient of variation of 15% or less (Crawford and Rumsey^). For a 90% asymptotic Cl (which indicates a critical value of the t distribution at 1.645), it follows that |W - W| < 1.645se < 1.645(0. 15)W < 0.25W (12) or |W-W|/W<0.25 (13) (i.e., half of the width of the Cl interval should be <25% of the estimate). We compare the Pjnd and Pjoi values for all CIs with 0.10 and 0.25. To determine whether Pjnd was related to the number of fish sampled or estimated size of the target group, we fitted power curves, using the results from all cases. Results Simulations Performance of the pooled and stratified estimators was similar when the variable of interest had few categories, but the stratified estimators did better as complexity increased (Table 2). Detailed simulation results are provided in Suppl. Tables 3 and 4. All es- timators produced acceptable accuracy and Cl cover- age when numbers of wild fish were estimated by sex. Similarly, all estimators provided acceptable accuracy ^ Crawford, B. A., and S. M. Rumsey. 2011. Guidance for monitoring recovery of Pacific Northwest salmon and steel- head listed under the federal Endangered Species Act, 117 p. Northwest Region, National Marine Fisheries Service, NOAA. [Available at website.) Steinhorst et al.: Estimates of abundance for the run composition of salmonids 7 Table 3 Sample size {n, number of steelhead), abundance estimate, and individual and joint confidence interval half-width (%) for groups of wild steelhead trout {Oncorhynchus mykiss) that spawned in 2011. Values are given for groups defined by sex, brood year (BY), or stock (identified by the location where the stock spawns). Group n Abundance Individual Joint Total wild fish 4701 44,133 2.3 2.8 Females 1466 29,541 2.4 2.8 Males 732 14,592 2.8 3.2 BY2004 38 784 8.4 12.3 BY2005 520 11,239 3.6 4.8 BY2006 994 21,449 2.6 3.5 BY2007 473 10,103 3.7 5.0 BY2008 26 558 10.1 13.8 Grande Ronde 472 9442 7.1 11.9 Imnaha 168 3318 11.9 21.0 Lower Clearwater 173 3421 12.4 20.3 Lower Salmon 98 1941 16.3 26.5 Lower Snake 219 4374 10.3 17.5 Middle Fork Salmon 214 4312 10.8 15.9 South Fork Clearwater 233 4228 10.4 15.8 South Fork Salmon 135 2512 13.8 20.6 Upper Clearwater 215 3885 11.4 18.2 Upper Salmon 340 6699 8.1 12.8 when numbers by age were estimated, but the asymp- totically normal CIs had poor coverage in one case for each estimator type. Average Cl coverage among stocks was similar between the pooled estimators: 87.7% for the pooled asymptotically normal estimator and 88.2% for the pooled bootstrap estimator. Average Cl coverage was slightly higher for the stratified estimators: 88.1% for the stratified asymptotically normal estimator and 89.0% for the stratified bootstrap estimator. The pooled estimators had unacceptable bias and very poor Cl coverage for 3 of the 10 stocks, whereas the stratified estimators had acceptable accuracy for all stocks. Av- erage Cl coverage among stocks was similar for the pooled estimators: 81.5% for the pooled asymptotically normal estimator and 82.2% for the pooled bootstrap estimator. In contrast, average Cl coverage was higher for the stratified estimators, although it was similar between them: 88.4% for the stratified asymptotically normal estimator and 89.0% for the stratified bootstrap estimator. Problems with pooled estimators became even more prevalent when we addressed age by stock; however, the performance of the stratified estimators also began to suffer as the number of groups to be estimated in- creased to 50 (Table 2). The pooled estimators had un- acceptable levels of bias in 21 cases, whereas the strat- ified estimators had unacceptable bias in 3 cases. Poor performance was most common in groups composed of steelhead from the least abundant BYs. Instances of poor Cl coverage were usually, but not always, asso- ciated with unacceptably high bias. Overall, stratified estimators performed better than pooled estimators. Further, the bootstrap CIs had better coverage than the asymptotically normal CIs; in 3 instances, asymp- totically normal CIs had very poor coverage (<80%), but there were no such instances for the bootstrap CIs. For this reason, we applied the stratified bootstrap es- timator to the SY2011 data to develop guidelines for sampling and interpretation of such data. Application of the stratified bootstrap estimator to data from SY201 1 During SY2011, 208,296 steelhead were counted at LGD. Of these fish, 44,133 steelhead were estimated to be wild (21.2%, Table 3). The 90% Cl was 43,152-45,140 wild steelhead. There were approximately twice as many females as males. Sex ratio varies annually, but the ratios seen in 2011 were typical. The middle age groups had more returning fish than the youngest and oldest age groups. Stocks were not evenly represented (e.g., the GRR stock had almost 4 times the number as the LOS stock) (Table 4). There were 46 stock-by- age groups in SY2011; estimated abundance ranged from 4912 individuals in the GRR stock in BY2006 to 21 individuals in the LSN stock in BY2004 (Table 4). The composition of a real steelhead run was not as bal- anced as that in the simplification used to generate the simulated data in this study, and this uneven distribu- tion was most apparent in the age-by-stock groups. Effects of using individual versus joint CIs depended on the complexity (i.e., number of groups) in the vari- 8 Fishery Bulletin 1 15(1) Table 4 Sample size {n, number of steelhead), abundance estimate, and individual and joint confi- dence interval half-width (%) for age groups (brood years) within stocks of wild steelhead trout {Oncorhynchus mykiss) that spawned in 2011. Groups are identified by brood year (BY) and the location where the stock spawns. Group n Abundance Individual Joint Grande Ronde BY2004 6 116 19.0 34.1 Grande Ronde BY2005 67 1501 8.6 16.2 Grande Ronde BY2006 222 4912 7.4 12.8 Grande Ronde BY2007 133 2875 7.7 14.2 Grande Ronde BY2008 2 38 31.9 54.5 Imnaha BY2004 2 43 35.1 63.3 Imnaha BY2005 22 537 17.2 32.0 Imnaha BY2006 82 1781 12.9 23.7 Imnaha BY2007 43 957 13.8 26.0 Lower Clearwater BY2004 1 32 46.8 84.2 Lower Clearwater BY2005 24 522 14.7 25.7 Lower Clearwater BY2006 78 1693 12.7 23.2 Lower Clearwater BY2007 49 1097 13.5 24.2 Lower Clearwater BY2008 4 78 30.2 55.3 Lower Salmon BY2004 1 23 57.5 106.1 Lower Salmon BY2005 18 409 19.5 35.4 Lower Salmon BY2006 40 909 17.9 30.6 Lower Salmon BY2007 26 557 17.9 31.1 Lower Salmon BY2008 2 44 48.3 87.6 Lower Snake BY2004 1 21 50.8 91.9 Lower Snake BY2005 13 324 16.2 29.5 Lower Snake BY2006 99 2268 11.4 20.2 Lower Snake BY2007 72 1614 11.4 29.4 Lower Snake BY2008 6 147 23.1 40.5 Middle Fork Salmon BY2004 10 222 16.2 '31.4 Middle Fork Salmon BY2005 99 2245 11.7 21.7 Middle Fork Salmon BY2006 72 1543 11.8 23.2 Middle Fork Salmon BY2007 6 303 22.1 40.4 South Fork Clearwater BY2004 3 59 28.2 50.3 South Fork Clearwater BY2005 44 906 12.1 22.5 South Fork Clearwater BY2006 144 3010 10.5 19.6 South Fork Clearwater BY2007 12 254 22.6 40.9 South Fork Salmon BY2004 6 132 28 51.8 South Fork Salmon BY2005 75 1612 14.3 25.2 South Fork Salmon BY2006 31 695 18.3 33.5 South Fork Salmon BY2007 3 73 34.2 62.2 Upper Clearwater BY2004 5 99 25.1 47.8 Upper Clearwater BY2005 120 2425 11.0 20.1 Upper Clearwater BY2006 63 1283 12.2 22.7 Upper Clearwater BY2007 3 54 29.8 54.0 Upper Clearwater BY2008 1 24 53.0 97.6 Upper Salmon BY2004 2 39 29.7 54.2 Upper Salmon BY2005 35 758 10.6 21.0 Upper Salmon BY2006 156 3356 8.8 16.3 Upper Salmon BY2007 107 2319 9.1 17.9 Upper Salmon BY2008 10 227 15.4 28.1 able of interest (Table 3). Widths of the joint CIs for females and males were not markedly wider than those of the individual CIs. For age, the widths of the joint CIs were 1.2-3. 4 times the widths of the individual CIs. For stock, the joint CIs were 1.2-2. 5 times wider. For some age-by-stock groups, the joint CIs were consider- ably wider than those of the individual CIs (Table 4). Attainment of precision goals depended on the group and whether individual or joint CIs were considered (Tables 3 and 4). All sex and age groups met the 10% Steinhorst et a!.: Estimates of abundance for the run composition of salmonids 9 Figure 2 Relationship of confidence interval (Cl) half-width for groups of wild steel- head (Oncorhynchus my kiss) spawning in 2011 in Snake River, Pacific Northwest, to (A) number of fish sampled (Cl half-width=47. 48 [number sam- pled]“°'®^^, coefficient of multiple determination [i?^]=0.872), and (B) abun- dance estimate (Cl half-width=144.74[abundance]‘‘’-^®2^ R^=0.870). Precision criteria levels of 10% (dotted line) and 25% (dashed line) Cl half-widths are shown for reference. research goal for precision if Pi„d was used, except for the BY2008 group (Pi„d=10.1%). With the use of Pjoi, the 10% goal was met for all sex and age groups, except the BY2004 and BY2008 groups. Stock groups did not meet the 10% goal, except the GRR and UPS stocks, if Pj„d was used. All sex, age, and stock groups met the 25% management goal for individual and joint CIs, except the LOS stock if Pjoi was used. Of the stock by age groups, 11% met the 10% precision goal if Pi„d was used, but none met the goal if Pjo, was used. There was wider disparity in attainment of the 25% goal; 70% of the age-by-stock groups met the goal if Find was used, but only 35% of the age-by-stock groups met the goal if Pjoi was used. In general, half-widths of the CIs declined and stabilized as the num- ber of fish sampled and estimated abundance of the groups involved increased (Fig. 2). Precision scaled approximately with the cube root of sample size, indicating that re- ducing the Cl width by half would require approximately 8 times as many samples. Values of Pi„d with- in a percentage point of the 10% precision criterion were obtained when there were 26-233 fish from a given category in the subsample (mean=140) and when there were 558-4374 steelhead in the category (mean=2794). Values of Pi^d closest to the 25% criterion were obtained when there were 5 and 6 fish in a subsample and when there were 99-147 steelhead in a category. The power functions parameterized from the group estimates yielded values of 96 samples and 1982 steelhead at the 10% criterion and 7 samples and 147 steelhead at the 25% precision criterion. Discussion The stratified estimators had biases <5%, except for a few cases in the most complex analysis (age by stock). Individual CIs for most constituent groups had cov- erages very near the nominal 90%. For conservation assessments, the greatest need is data on abundance of each stock. Coverage of CIs was good even for the smallest stock. Age structure is important for compu- tation of productivity for each stock (i.e., for summa- rizing the adult progeny from a brood year returning across years and for computing progeny per parent). Accuracy of productivity estimates typically are large- ly controlled by the most abundant age groups, which had unbiased estimates in our study. When the strati- fied estimators were used, only a few of the smallest stock-by-age groups had bias >5.0% or Cl coverage <85%. Conversely, pooled estimators performed poorly ex- cept for the simplest variables of interest: sex and age. As variable complexity increased from age (5 cases) to stock (10 cases) to age-by-stock (50 cases), the propor- tion of estimates biased >5% increased from 0% to 30% to nearly half, respectively. Initially, we expected pooled estimators to be acceptable because of the highly con- 10 Fishery Bulletin 1 15(1) trolled sampling regime at LGD, which was very con- sistent for SY2011. The realized sampling rate in the simulation (the product of trap rate and subsample rate) averaged 4.7% (standard deviation 1.3%). How- ever, stock and age composition changed through the run, as did the number of steelhead crossing the dam. As the variable of interest became more complex, accu- racy and precision of pooled estimators decreased. Steinhorst, et al. (2010) estimated the run compo- sition of fall-run Chinook salmon (O. tshawytscha) at LGD on the basis of counts from observation windows from 18 August through 15 December (their meth- od 1). They used a stochastic model based on a fast Fourier transform to model the distribution of daily window counts, which were summed to obtain total abundance. Steinhorst, et al. (2010) used 2 bootstrap steps — a nonparametric bootstrap associated with the Fourier model and a parametric bootstrap applied to an estimate of composition pooled over the season. They did not report composition by stratum because composition was calculated with a complex accounting algorithm that could not be applied to individual stra- ta. In essence, they assumed that either the propor- tions of their sex-by-age-by-origin groups were fairly uniform over the season or that a constant proportion of the run was sampled for each stratum. However, if the groups of interest returned at different times, a pooled estimate of composition applied to total escape- ment would not be accurate, especially over longer temporal spans (e.g., the steelhead run). In our study, the simulation results from the pooled estimators in- dicated precisely that outcome. Precision may be computed for each group of inter- est, one at a time (i.e., Find); or more conservatively across all groups within a variable of interest (Pjoi), minimizing study-wide error. However, the conservative approach resulted in wider CIs; for example, joint CIs were 14-17% wider than the individual CIs for sex in the SY2011 run. For stocks, Pjoj values were about 47- 76% wider than Pjnd CIs. Given the number of stocks, we were not paying a large penalty for computing joint CIs. For age-by-stock groups, the joint CIs were on average 85% wider than the individual CIs (range: 71-158%). This difference likely was due to the uneven distribution of numbers by age and the large number of age-by-stock groups. Because we were trying to achieve joint coverage across so many groups simultaneously, a much greater expansion of the CIs was necessary. The results of our study show the cost to statistical power caused by the inclusion of many groups in an analysis. Investigators must consider whether the more conser- vative approach affects the usefulness of the resulting estimates and which groups are truly of management interest. For the latter consideration, investigators may combine some groups or decide that loss of precision is acceptable for their application. In our case, we com- bined strata to achieve greater sample sizes and used total age rather than the combinations of years spent in freshwater and years spent in saltwater that salmon biologists often use (Quinn, 2005). Precision is related to the amount of information available, and the quality of this information declines as group size becomes smaller or as the realized sam- ple rate is reduced. The problem in our case was that the steelhead run in Snake River is protracted over time, compounded by the complexity of the life history and stock structure of steelhead. Therefore, multinomi- al proportions must be estimated for many groups over many time strata unless the groups of interest can be simplified. Even so, we generally met the research goal of half the 90% Cl width within 10% of the estimate for sex and age groups present in SY2011. For stock groups, we met the management goal of half the 90% Cl width within 25% of the estimate for sex and age groups but our 10% precision goal was not attained, ex- cept with the 2 largest stocks when the less stringent Find measure was used. To develop guidance for interpretation of the esti- mates we obtained from the data, we relied on Pjnd because it was not affected by the number of other groups in the analysis. Precision of the estimates for individual groups declined rapidly when group abun- dance was <2500 individuals or when <100 individu- als from that group were collected in the subsample. However, if there were few fish in a group, analysts and managers probably would be content with a more lenient precision criterion. For example, if our esti- mate was 50 and the Cl was 20-80, the percent half- width would be 60% of the estimate but the fact that the true number is between 20 and 80 should be suf- ficiently precise for management purposes, especially if the numbers of fish in other groups are decidedly larger. With Pinj as a measure of precision, the 10% research precision goal could be reached if group abundance were to exceed 2000 individuals or if >100 samples from that group were collected. The 25% management precision goal was much more attainable and was achieved at group abundances >150 individu- als and when very few samples were collected (<10). These values can be used as thresholds for the lenient precision criterion in our application. Our results have implications for monitoring fish populations. If the interest is on the largest groups in a mixed population, most sampling programs will yield sufficient results. However, weak stocks are frequently a problem for conservation and fisheries management, and precision of abundance estimates of smaller groups becomes important. Obviously, there is a tradeoff be- tween sample size and number of subdivisions that can be maintained. Previous work by Gerritsen and McGrath (2007) has supported this notion, but their criteria for success focused on overall (average) preci- sion. Thompson (1987) found that a sample size of 510 fish should suffice under a worst case scenario for a = 0.05 (equal proportions among groups, but number of groups does not matter) as long as desired precision is expressed in absolute terms. If desired precision is ex- pressed in relative terms (as was done in our study), no sample size will be sufficient if group size approaches zero. However, our results provide useful guidance for Steinhorst et al.: Estimates of abundance for the run composition of saimonids 11 determining the groups that will have reliable esti- mates in complex scenarios. Even at lower fish abun- dances, we can define a lenient precision criterion with practical value. The estimation approach described in this article is very flexible and can be customized for many scenarios. In this study, we assumed that window counts were a census, and stock and age were determined without error. In most applications, abundance is determined from a sample rather than a census. Further, there is uncertainty in the determination of stock and age for each fish. Additional uncertainty (e.g., genetic variabil- ity or noncensus estimates of total abundance) can be incorporated into our framework by adding additional bootstrap steps to reflect the source of the additional variance (e.g., method 2 in Steinhorst et al., 2010). An- other practical issue is that samples for genetic anal- ysis can now be processed en masse, but ages must be read from scale samples individually; hence, more fish can be identified to stock than can be aged. The bootstrap routine can be altered such that all genetic information is used to estimate stock abundance, and age composition is applied within each stock estimate (i.e., age composition is conditional on stock). The stratified estimator in our study produced un- biased estimates, and the parametric bootstrap CIs had good coverage and acceptable precision. In com- plex scenarios, estimates of abundance of small groups will have poor precision and some may be biased, but a stratified estimate with a conservative joint Cl can be of practical use if the numbers of fish in other groups are much larger. The 2-step bootstrap approach is flex- ible and can be adapted to incorporate other sources of variability or sampling constraints. Acknowledgments Funding for this project was provided by Bonneville Power Administration under project numbers 1990- 055-00, 1991-073-00, and 2010-026-00. M. Campbell reviewed earlier drafts of this article. We gratefully acknowledge the thorough critiques given by 3 anony- mous reviewers. Literature cited Anderson, E. C. 2010. Assessing the power of informative subsets of loci for population assignment: standard methods are up- wardly biased. Mol. Ecol. Resour. 10:701-710. Anderson, E. C., R. S. Waples, and S. T. Kalinowski. 2008. An improved method for predicting the accuracy of genetic stock identification. Can. J. Fish. Aquat. Sci. 65:1475-1486. Campbell, M. R., C. C. Kozikay, T. Copeland, W. C. Schrader, M. W. Ackerman, and S. R. Narum. 2012. Estimating abundance and life history characteris- tics of threatened wild Snake River steelhead stocks by using genetic stock identification. Trans. Am. Fish. Soc. 141:1310-1327. Ford, M. J. (ed.). 2011. Status review update for Pacific salmon and steel- head listed under the Endangered Species Act: Pacific Northwest. NOAA Tech. Memo. NMFS-NWFSC-113, 281 p. Gerritsen, H. D., and D. McGrath. 2007. Precision estimates and suggested sample sizes for length-frequency data. Fish. Bull. 105:116-120. Goodman, L. A. 1960. On the exact variance of products. J. Am. Stat. As- soc. 55:708-713. Harmon, J. R. 2003. A trap for handling adult anadromous saimonids at Lower Granite Dam on the Snake River, Washing- ton. North Am. J. Fish. Manage. 23:989-992. Hughes, J. B., and J. E. Hightower. 2015. Combining split-beam and dual-frequency identi- fication sonars to estimate abundance of anadromous fishes in the Roanoke River, North Carolina. North Am. J. Fish. Manage. 35:229-240. Latremouille, D. N. 2003. Fin erosion in aquaculture and natural environ- ments. Rev. Fish. Sci. 11:315-335. Mandel, M., and R. A. Betensky. 2008. Simultaneous confidence intervals based on the per- centile bootstrap approach. Comput. Stat. Data Analys. 52:2158-2165. Pella, J. J., and G. B. Milner. 1987. Use of genetic marks in stock composition analy- sis. In Population genetics and fisheries management (N. Ryman and F. Utter, eds.), p. 274-276. Univ. Wash. Press, Seattle, WA. Pritt, J. J., M. R. DuFour, C. M. Mayer, P. M. Kocovsky, J. T. Tyson, E. J. Weimer, and C. S. Vandergroot. 2013. Including independent estimates and uncertainty to quantify total abundance of fish migrating in a large river system: walleye (Sander vitreus) in the Maumee River, Ohio. Can. J. Fish. Aquat. Sci. 70:803-814. Quinn, T. P. 2005. The behavior and ecology of Pacific salmon and trout, 388 p. Univ. Wash. Press. Seattle, WA. Raymond, H. L. 1988. Effects of hydroelectric development and fisheries enhancement on spring and summer Chinook salmon and steelhead in the Columba River basin. North Am. J. Fish. Manage. 8:1-24. Rudstam, L. G., J. M. Jech, S. L. Parker-Stetter, J. K. Horne, P. J. Sullivan, and D. M. Mason. 2013. Fisheries acoustics. In Fisheries techniques, S'"'* ed. (A. V. Zale, D. L. Parrish, and T. M. Sutton, eds.), p. 597-636. Am. Fish. Soc., Bethesda, MD. Shaklee, J. B., T. D. Beacham, L. Seeb, and B. A. White. 1999. Managing fisheries using genetic data: case stud- ies from four species of Pacific salmon. Fish. Res. 43:45-78. Steele, C. A., E. C. Anderson, M. W. Ackerman, M. A. Hess, N. R. Campbell, S. R. Narum, and M. R. Campbell. 2013. A validation of parentage-based tagging using hatchery steelhead in the Snake River basin. Can. J. Fish. Aquat. Sci. 70:1046-1054. 12 Fishery Bulletin 115(1) Steinhorst, K., D. Milks, G. P. Naughton, M. Schuck, and B. Arnsberg. 2010. Use of statistical bootstrapping to calculate confi- dence intervals for the fall Chinook salmon run recon- struction to Lower Granite Dam. Trans. Am. Fish. Soc. 139:1792-1801. Tarbox, K. E., and R. E. Thorne. 1996. Assessment of adult salmon in near-surface waters of Cook Inlet, Alaska. ICES J. Mar. Sci. 53:397-401. Thompson, S. K. 1987. Sample size for estimating multinomial propor- tions. Am. Stat. 41:42-46. Wagner, P. G. 2007. Fish counting at large hydroelectric projects. In Salmonid field protocols handbook: techniques for assess- ing status and trends in salmon and trout populations (D. H. Johnson, B. M. Shrier, J. S. O’Neal, J. A. Knutzen, X. Augerot, T. A. O’Neil, T. N. Pearsons, eds.), p. 173- 195. Am. Fish. Soc., Bethesda, MD. Wood, C. C., S. McKinnell, T. J. Mulligan, and D. A. Fournier. 1987. Stock identification with the maximum-likelihood mixture model: sensitivity analysis and application to complex problems. Can. J. Fish. Aquat. Sci. 44: 866-881. 13 National Marine Fisheries Service NOAA Fishery Bulletin fb- estabiished in 1881 Spencer F. Baird First U.S. Commissioner of Fisheries and founder of Fishery Bulletin Flexible Baf esiaii analysis of the won Bertalanffy growth function with the use of a log'Skew^t distribution Freddy Omar Lopez Quintero' Javier E= Contreras-Reyes (contact author)''* Rodrigo Wiff ^ Reinaldo 1. Arel!ano-¥alle^ Email address for contact author: javier.contrerasigifop.c! Abstract— The von Bertalanffy growth function is the model most widely applied to describe growth in fish populations. Parameters de- scribing this function usually are estimated from observed lengths at different ages by using maximum likelihood and by assuming Gauss- ian distributed errors. In harvested populations, observed length at age usually involves a high level of skewness and extreme values be- cause of the size-selective sampling process. Some approaches, based on the maximum-likelihood method for making inferences, have been devel- oped to resolve such issues. We pro- pose a Bayesian framework for esti- mating growth parameters for non- linear regression models — a frame- work that is based on the family of log-skew-^ distributions and which provides an approach that is flexible enough for modeling the presence of asymmetries and heavy tails. This framework based on a method in which 1) the error accounts for both skewness and heavy-tailed distribu- tions of a log-skev/-^ model, and 2) the observed length at each age has a heteroscedastic error distribution. The proposed method was applied and compared with the methods of previous models by using observed length-at-age data for the southern blue whiting (Micromesistius austra- lis), an important fish species har- vested in the southeast Pacific. Com- parisons indicated that the proposed model is the best for describing data on southern blue whiting. Manuscript submitted 6 January 2016. Manuscript accepted 30 September 2016. Fish. Bull. 115:13-26 (2017). Online publication date: 27 October 2016. doi: 10.7755/FB.115.1.2 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. ’ Departamento de Matematica Universidad Tecnica Federico Santa Maria Avenida Esparia 1680 Valparaiso 2390123, Chile 2 Division de Investigadon Pesquera Institute de Fomento Pesquero Avenida Almirante Blanco Encalada 839 Valparaiso 2361827, Chile Growth is one of the most impor- tant measurable life-history traits in individual organisms because it is fundamental in creating an un- derstanding of both population and ecosystem functions. Several models have been proposed to describe ani- mal growth. The most widely applied model, however, is the von Berta- lanffy growth function (VBGF; von Bertalanffy, 1938). This model has been used to describe length at age for a wide range of species across several taxa, such as mammals (Eng- lish et al., 2012), birds (Tj0rve and Tj0rve, 2010), and reptiles (Lehman and Woodward, 2008), although it is most extensively applied to fish spe- cies (Pardo et aL, 2013). The VBGF is based on principles underpinning the physiology of growth (von Ber- talanffy, 1938; Wiff and Roa-Ureta, 2008), gives an adequate description of growth with the use of only 3 pa- rameters, and states that the rate of growth of an individual is deter- mined by the difference between the buildup of body mass and loss due to energy expenditures for mainte- 3 Center of Applied Ecology and Sustainability Pontificia Universidad Catolica de Chile Avenida Libertador Bernardo O'Higgins 328 Santiago 8331150, Chile ^ Departamento de Estadfstica Pontificia Universidad Catolica de Chile Avenida Vicuna Mackenna 4860 Santiago 7820436, Chile nance. In harvested fish populations, the usual data available to estimate these parameters are cross-sectional, and a single length and age measure- ment is taken from each sampled individual. The VBGF describes the expected growth rate for the popu- lation, on the basis of length-at-age data composed of individuals with variable growth rates. Maximum-likelihood techniques derived from Gaussian and log- Gaussian errors normally are used to estimate VBGF parameters (Mil- lar, 2002; Siegfried and Sanso, 2006). Yet, in fish populations, this assump- tion often fails because these distri- butions are typically skewed, present heavy tails or have extreme values. Skewed distributions usually result from the size-selective sampling pro- cess (Montenegro and Branco, 2016). In addition, in harvested fish popu- lations, an accumulative effect of fishing exploitation exists for size at age. Growth rates vary among indi- viduals (Sainsbury, 1980), and fish- ing selectivity removes faster grow- ing individuals from each particular 14 Fishery Bulletin 115(1) age class. Hence, the bias in sampling for length at age that favors fast-growing individuals of each age class (Taylor et ah, 2005). Therefore, the assumption of Gaussianity to estimate parameters of the VBGF is not adequate (Contreras-Reyes and Arellano-Valle, 2013; Montenegro and Branco, 2016). Moreover, the as- sumption of Gaussianity implies that length may take negative values and, therefore, is nonsensical (Xiao, 1994; Millar, 2002). Different approaches have been proposed to over- come this drawback and fitting the VBGF. They can be separated roughly into 2 categories. In the first one, models, such as the one in Taylor et al. (2005), pro- vide a mechanistic approach to dealing with skewed length-at-age data, with a combined process of growth, selectivity, and mortality when fitting the VBGF. The second category is a more empirical approach in which skewed and heavy-tailed length-at-age data are mod- eled by using the maximum-likelihood method and as- suming a non-Gaussian distribution (Contreras-Reyes and Areliano-Valle, 2013) and by using Bayesian anal- ysis (Millar, 2002; Siegfried and Sanso, 2006). Millar (2002) proposed a Bayesian framework to estimate parameters of the VBGF, using a multiplicative error model with log-normal distribution. Contreras-Reyes and Arellano-Valle (2013) calculated the maximum- likelihood estimates for the VBGF with the family of skew- distributions (Azzalini and Capitanio, 2003), a flexible class that extends the known Student distri- bution (e.g., Geweke, 1993). Such models can incorpo- rate asymmetric and heavy-tailed errors, with presence of heteroscedasticity (Montenegro and Branco, 2016). Contreras-Reyes et al. (2014) reanalyzed the skew- ap- proach to incorporate a log-skew-^ distribution under multiplicative error distribution. In this study, we examined our proposed Bayesian method for estimating the VBGF parameters on the basis of a log-skew-^ distribution. This new framework merges the benefits provided by Bayesian analysis (Sieg- fried and Sanso, 2006) and the log-skew-i distribution (Contreras-Reyes et al., 2014) for estimating parameters of the VBGF for harvested fish populations. Addition- ally, our approach allows for heteroscedasticity in errors, modeled as power and exponential functions (Contreras- Reyes and Arellano-Valle, 2013). This Bayesian frame- work is applied to data of length-at-age composition of southern blue whiting {Micromesistius australis), an important species fished in the southeast Pacific. Materials and methods Log-skew-f von Bertalanffy growth model We let L(xi) be the expected value of the length related to an th individual at age Xj, L^>0, K > 0, tQ < min{xi, ... Xj,}, and n is the sample size. The VBGF defines growth in length as L(xi) = L<^(l-e-^<"‘-'»^). Equation 1 represents the simplest formulation of the VBGF, described by 3 parameters: where = the asymptotic length (in length units, e.g., centimeters); K - the growth rate coefficient expressed per unit of time; and ^o = the theoretical age (usually in years) when the length is zero. Parameters of the VBGF usually are estimated from observed length-at-age pairs, such as (x;, jj), i = 1, n, where is the ith observed length at age xj. Equation 1 was described in terms of multiplicative structure (Millar, 2002; Siegfried and Sanso, 2006; Contreras- Reyes et al., 2014) for random errors: = L(Xi)£i, (2) where £; = non-negative random errors, usually as- sumed to be independent, identically distributed errors with a mean of 1. Given this assumption, the VBGF in Equation 2 corresponds to the nonlinear regression with multiplicative random errors. We easily recovered the additive structure of the original model in Equa- tion 2 by applying log scale in the following way: y[ = L'(Xi ) + £,', with log y[, L'(x-^ ) = log L; = L[, and e( = log£i, i = 1, re, (3) in which el were assumed to be independent, identi- cally distributed, random errors with zero mean. Contreras-Reyes et al. (2014) assumed a log-skew-^ distribution (Azzalini et al., 2003) for the multiplica- tive and heteroscedastic random errors. Specifically, they assumed that the multiplicative errors Ei, i = 1, re, were independent random variables following a log-skew-^ distribution with parameters /fj G R (loca- tion), af > 0 (scale and dispersion), Aj G R (skewness and shape), and v > 0 (degrees of freedom), a distribu- tion that is denoted by el ~ LSTifii,af,X,v), i = l,...,n. (4) This approach is equivalent to considering that the transformed errors Sj, i = 1, ..., n, are independent and have a skew distribution (Branco and Dey, 2001; Az- zalini and Capitanio, 2003) denoted by el ~ ST(iii,af,X,v), i = 1, re. (5) In turn, this notation indicates that the transformed response variables (lengths) are derived from yl ~ STifi^ + Ll,af,X,v), i = l,...,n. (6) namely, that the density of yl is given by where — (yl ~Mi~ ) / (Tj is a standardized version of y ■ , (1) Lopez Quintero et al.: Bayesian analysis of the von Bertalanffy growth function 15 r v + 1 2 / r(v/2)(7rv)2 v+l V (8) z e M is the symmetric Student-^ density with v de- grees of freedom, and T{z;v) represents the respec- tive cumulative distribution function. In other words, we assumed that the original response yi followed a log-skew-^ distribution (Marchenko and Genton, 2010), which is denoted by ~LSr(/ii-LL;,af,A,v),f = l,..., n. (9) We assumed that v > 1 and considered the first mo- ment of the skew-^ distribution (Branco and Dey, 2001); therefore, the extra parameter had tobe chosen as = /7 n(v-i)/2] Agj Vtt r(v/2) ^1+^2 ’ (10) so that the transformed errors e- have a zero mean. This condition ensured that E{y[) = L[ and allowed us to identify the constant in the additive version of the regression model. Heteroscedasticity was introduced by means of the dispersion parameters erf and modeled by using a non- negative function m(p; Xj) depending on age x; and a heteroscedastic parameter p e M as af = (j^m(p; Xj), where > 0. When p = 0, homoscedasticity is recov- ered as (jf = cj^m(0; X;) = cr^. In our study, we considered 2 specific functions for modeling heteroscedasticity: the exponential function m(p; X;) = and the power func- tion m(p;Xi) = xf^. In both functions, if m(0;xi) = 1, it corresponds to the homoscedastic case. Asymmetry and heavy tails produced by extreme values of length-at-age data were controlled by the pa- rameters of shape (A) and degrees of freedom (v). Extension to a Bayesian framework We advanced a Bayesian analysis for the log-skew- t VBGF described in the previous section. Therefore, we first noted from the independence assumption and Equation 7 that the likelihood function of the unknown parameter vector 9 = {(3^ is fiy' I x,«) = !• + 1), (11) where P = {L^,K,Tq)^ are VBGF parameters, y' = {y'i, 1 x = (xi,...,x„)^, and Zj is as it was previously defined. To complete our Bayesian model specification, we needed to elicit a prior distribution for the unknown parameter vector, say niO). Therefore, the Bayesian in- ference on 0 (or function of 9) was based on the posteri- or distribution jd^d |x, y') oc/(y'|x, 9)7d9). This posterior distribution does not have a closed form (Cancho et al., 2011), but an estimation could still be calculated by using a Markov chain Monte Carlo (MCMC) algorithm (Chib and Greenberg, 1995; Cowles and Carlin, 1996; Robert and Casella, 2004). Given the available methods, we chose to implement a hand-tailored component-wise Metropolis-Hasting transition scheme; in other words, we selected an ap- proach in which 6 is divided into individual pieces that are easily updated sequentially with a random walk algorithm. Our selection was based on simplicity and the need to control all steps in the sampling. Other options included the use of variants of BUGS language (Lunn et al., 2012), the AD ModelBuilder (Fournier et al., 2012), or Stan software (Gelman et al., 2015). Note that our selected approach is different from that fol- lowed by Siegfried and Sanso (2006), who employed an algorithm that included both Gibbs, as well as Metrop- olis-Hasting steps. An advantage of using this MCMC scheme is that the procedure is subdivided into several univariate steps. All proposal distributions were tuned to achieve acceptance rates of 25-45% (Robert and Ca- sella, 2004). Specifically, we considered the different components of as independent (Siegfried and Sanso, 2006); in other words, idB) becomes the product of the marginal prior distributions of ,G^,p,\,v). It follows that only these marginal prior distributions must be elicited to complete our Bayesian model. Prior distributions for the VBGF parameters P were chosen as follows. Given that is strictly positive, we assumed a left truncated normal distribution with large variance (e.g., 100) as the prior distribution for this pa- rameter. Other possible and natural choices were log- normal, gamma, or even distributions with support in a reasonable and restricted interval. Because param- eters K and -to are both positive, we used gamma as prior distributions for them (Xiao, 1994; Siegfried and Sanso, 2006). Contreras-Reyes et al. (2014) reported estimates of around 0.16 for southern blue whiting, a value that incidentally conforms to the value obtained by Siegfried and Sanso (2006) for blue shark {Prionace glauca). We used this information to specify that the mean of the gamma prior distribution of was around 0. 15. For -^0, Contreras-Reyes et al. (2014) obtained -^0 = 2.5,, which indicates a prior distribution for this parameter. For the scale parameter cP, we considered the clas- sical inverse gamma prior distribution suitable for this type of parameter (Zhang et al., 2009). The het- eroscedastic parameter p usually takes positive or negative values. To give the full power of estimation to the data, we chose a noninformative prior, nip) 1. For the shape parameter A, we set a normal prior distribution with a zero mean and large variance. The parameter defining the degrees of freedom, v, should be strictly larger than 2 to ensure the existence of vari- ance in the log-skew-^ model; therefore, we considered an exponential prior distribution with mean equaling 2 (Geweke, 1993; Cancho et al., 2011) and truncated at the interval (2, 00). These prior specifications are sum- marized in Table 1. Comparisons and selection of models For sake of comparison, we considered 2 additional models with constant variance function derived from the log-normal distribution. The first one (hereafter, re- 16 Fishery Bulletin 115(1) Table 1 Elicited prior specifications for log-normal and log-skew-f models used to examine a Bayesian analysis of the von Bertalanffy growth function. ri¥(o,.=„)(0, 100) represents the MOjlOOl-density truncated at (0,=»), and T£(2^„){0.5) denotes the exponential density trun- cated at the (2,oo) interval. The parameters are the asymptotic length growth rate coefficient (K), theoretical age in years when the length is zero (-to), heteroscedastieity (p), inverted dispersion skewness (1), and degrees of freedom (v). Parameter Log-normal (type I) Log-normal (type II) Log-skew-^ L„(cm) niL„) 1 7W(o,„)(0,100) ri¥(o,„)(0,100) «y-i) Gamma( 15, 100) Gamma(15, 100) Gam?na(15,100) -^o(y) Gamma{10,4) Gamma{10,4) Gamma{10,4) P - - n(p) oc 1 GammaiO. 1,0.1) GammaiO. 1,0.1) Gamma(0. 1,0.1) 1 - - M15,100) V - - 7W(2,oo)(0.5) ferred to as type-I model) was similar to that developed by Siegfried and Sanso (2006), and the second one in- cluded a modification of the prior distribution of so that it was the same as that proposed in the log-skew-^ model. All these prior specifications are summarized in Table 1. The following models were considered: WA/C = Er=ilog i£log/(yn^i,^s) 8=1 ■PwAIC- (14) Also, the WAIC is related to the effective number of parameters: PWAIC — • Log-normal (type I) with constant variance function; • Log-normal (type II) with constant variance function; • Log-skew-^ with constant variance function; • Log-skew-^ with exponential variance function; • Log-skew-^ with power variance function. Selecting the “best” model is an important aspect in statistical analysis. In the rest of this section, we de- scribe how we implemented the deviance information criterion (DIG) and the widely applicable information criterion (WAIC) for model selection. 2Er.i log S=1 i E log /( ) - i E log /(y.'l ) S=1 . (15) Compared with DIG, WAIC has the property of averag- ing over the posterior density by using each iterated 0s, instead of being replaced by the mean 0. In addition, jowAic is more numerically stable than poic because it averages separately for each observation y[ (Gelman et al., 2014). Influential analysis Deviance information criterion The DIG proposed by Spiegelhalter et al. (2002) is based on the posterior mean of the deviance, and it can be approximated by the MCMC algorithm as follows: Die =2y:i^ log f{y[ i , 0 ) - A £ log f{y[ | X; , 0^ ) S=1 (12) where ^=-^Ef=i^s is the mean of a sample 0|,...,0b obtained from the posterior distribution 7r(0|(5). The DIG is related to the effective number of para- meters: Pdic — 2 log f{j' I X, 0 ) - i £ log f{j' I X, 03) S = 1 (13) The widely applicable information criterion The WAIC (e.g., Gelman et al., 2014) is based on the computed log-pointwise-posterior-predictive density, complement- ed by a correction for the effective number of para- meters to adjust for overfitting: The statistical stability of the proposed models exposed to perturbations of the data were analyzed by using in- fluential analysis. We considered the Kullback-Leibler (KL) divergence measure (Kullback and Leibler, 1951) to quantify the effect on the inferences produced by excluding one observation or a group of observations from the full data set. The KL-divergence had been considered previously in Bayesian influential analysis for elliptical and skew-elliptical models (Arellano-Valle et al., 2000; Vidal et al., 2006). We let P = 71(0 1 S) and iA; = 7r(0 1 S_j) be the posterior distribution of 9 obtained from the full data S = (x, y') and the data without the ith observation <§_; = (^j, y'_i), respectively. The KL-divergence between P and P_i was given by X(P,i>_i) = /i(e|S)iog{,^)d». (16) To identify influential observations, Peng and Dey (1995) showed that if pi » 1/2, where (17) Lopez Quintero et al.: Bayesian analysis of the von Bertalanffy growth function 17 then the rth observation is considered influential. Ad- ditionally, because the integral in Equation 16 cannot be written in closed form, it still can be approximated by sampling from the posterior distribution of 6 via the MCMC algorithm. In fact, if 6i,...,6b is a sample of size B from 7r(0 1 S), then the MCMC estimator of K(P,P_i) is computed as K{P,P_i) = logjiE S=1 + iIls=i^ogf{yl\x,0J, (18) with /'(y[|x,0g) given by Equation? and 9^ = (e.g., Cancho et al., 2011). It should be noted that we computed the KL-divergence be- tween P and P_i using the ith marginal sample density /’(y(|x,0s), but we did so without considering the pos- teriors 7r(0|S) and 7r(0|15 years) the observed length tends to converge to = 59.52 (Table 3). The log-skew-t model provides more precise 95% HPD in- tervals for older ages (>13 years; Fig. IB) and less pre- cise for young ages (0-5 years) in comparison with the log-normal model. Intervals of 95% HPD of log-skew-t model fit indicate that the observations were affected by the negative heteroscedastic parameter p (Fig. 1C). In addition, constant variance was assumed for the log-normal model and, therefore, the model underesti- mated the real variance in the age at length containing extreme values. The posterior densities of VBGF and variance parameters corresponding with the homosce- dastic log-normal and power-variance log-skew-^ mod- els are compared in Figure 2. The asymmetry and dis- persion of the posterior densities of VBGF were similar for the different error distributions. However, for the variance parameter, the posterior density was leptokur- tic when log-normal error distribution was used. Considering the boxplots of residuals by age from 20 Fishery Bulletin 115(1) Table 4 Summary of chain diagnostics for the fitted models: effective sample size (ESS), credibility R of Gelman index, Geweketest (G), Heidelberger-Welch test (HW), and Raftery-Lewis test (RL). In addition, the deviance information criterion (DIG) and widely applicable information criterion (WAIC) values for each model are reported, with their estimated number of parameters, Pd/c and pwA/c> respectively. The parameters are the asymptotic length (L^), growth rate coefficient (K), theoretical age in years when the length is zero (-?o), dispersion (a^), heteroscedasticity (p), skewness (A), and degrees of freedom (v). Model Parameter ESS R G HW RL Die Pdic WAIC PWAIC Log-normal (type I) Constant Loo 1134.673 1.002 1.543 0.817 39.4 -65275.66 4.338 -65275.66 7.079 K 760.156 1.003 -1.241 0.850 82.6 ~^0 934.834 1.003 1.518 0.799 63.0 52477.971 1.000 1.111 0.572 4.6 Log-normal (type II) Constant 749.129 1.005 -1.891 0.226 32.4 -65275.659 4.338 -66623.850 8.181 K 463.168 1.008 1.499 0.050 50.3 0 569.197 1.008 -1.817 0.456 35.3 29883.554 1.000 1.510 0.093 3.4 Log-skew-? Constant Loo 211.801 1.022 1.665 0.143 14.9 -66231.757 2.918 -66227.642 6.133 K 81.302 1.037 -1.118 0.500 30.0 0 91.705 1.041 1.230 0.393 60.5 (fi 656.456 1.002 1.662 0.654 6.7 A 596.318 1.004 -1.598 0.638 13.3 V 1927.670 1.005 -1.232 0.408 5.1 Log-skew-? Exponential Loo 204.714 1.054 -1.174 0.436 8.2 -66559.569 6.420 -66556.535 9.048 K 112.927 1.077 1.232 0.471 43.7 ~^0 162.885 1.078 -1.696 0.264 16.9 p 561.808 1.013 -1.335 0.504 8.5 179.131 1.029 1.015 0.673 116.0 A 1404.850 1.010 1.490 0.247 8.7 V 856.653 1.007 -0.771 0.159 10.9 Log-skew-? Power L„ 200.944 1.008 -1.293 0.087 7.7 -66625.510 6.482 -66623.850 8.181 K 99.489 1.052 0.979 0.086 16.0 “^0 128.952 1.053 -1.223 0.052 16.9 p 706.436 1.008 -0.580 0.096 8.7 (d- 189.195 1.029 0.631 0.615 23.6 A 1656.937 1.004 1.700 0.355 5.6 V 289.711 1.016 -0.481 0.495 24.5 le fitted log-skew-? model (Fig. 3), we can observe that (Fig. 4). As expected. we found that the log-norm. residuals indicate a flat pattern and that their mean is concentrated around zero. We noted also a decreasing variance in older fish, produced in part by the negative value of the estimated heteroscedasticity (p = -0.18). Furthermore, extreme values for younger and older fish (<6 and >15 years) were detected by the estimated de- gree of freedom (u =14.32; Table 3). Influential analysis Peng-Dey’s criterion (Eq. 17, p; = 0.5) is suitable for certain nonlinear regression models with normal er- rors and many observations are considered influential model has more influential observations than the pow- er log-skew-f model for each probability. Therefore, we selected, in Table 5, only the probabilities 0.70, 0.60, 0.55, and 0.51 for those influential observations in log- normal and power log-skew-i models. For the selected model, when p; = 0.51, the largest number of restricted observations was recorded and the RC of the error dis- tribution parameters was raised. When the number of influential observations increased (in terms of the pj) and were removed, the degree of freedom parameter also increased. Because several of these observations are extreme values (Contreras-Reyes et al., 2014), the error distribution shifts from log-skew-? to log-skew-? Lopez Quintero et al.: Bayesian analysis of the von Bertalanffy growth function 21 Age 16 18 20 22 24 Age 0.010- 0.008 0.006 c 46 m were reported in 50% of longline sets). We are unsure why gangion length was often recorded inaccurately by fishermen, but a greater awareness of proper recording techniques for this variable could result in its use in future studies of immediate mortality. We recommend that 2 additional variables that can influence mortal- ity or catch rates, namely air temperature during haul- ing and estimated set depth, should be added to future USPL reporting forms. In addition, several other vari- ables that were not examined in our study but could influence catch and mortality rates have been included in recent iterations of the reporting form for the USPL (2015 form available from website; 2003 form available from website). These variables — bait used, hook type, hook size, and hook offset — should be incorporated into future analyses in which the USPL data set is used. There were 19 categories (species or groupings) of sharks that were recorded by fishermen in the UPSL data set (Table 1). Only 4 of these categories were considered for analysis because of problems with iden- tification and uncertainty about the accuracy of the recordings. The International Plan of Action for the Conservation and Management of Sharks outlined im- proved shark identification as one of its primary objec- tives and as a critical step to acquiring data to assess shark stocks (FAO^). We recommend that fishermen use shark identification guides or undertake shark identi- fication training (or both) to improve the accuracy of entries in the USPL data set for cryptic species and species that are similar in appearance to other species. We are the first to use the USPL data set to as- sess immediate mortality rates of longline-caught ti- ger and blue sharks. Rates of immediate mortality for these species closely match results from previous stud- ies, and this similarity indicates accurate recording by commercial fishermen (Table 2). Because the USPL data set covers the entirety of the US. Atlantic com- mercial longline fishery, it can be used to determine long-term changes in mortality over a wide geographic scale that may not be achievable with other data sets. If additional variables are recorded by commercial fish- ermen, the USPL data set can be used to accurately discern causes of mortality during fishery capture in a wide range of species. Although our analysis was re- stricted to examining bycatch of sharks, the inclusion of additional variables could allow for mortality exami- nations of targeted teleosts and targeted sharks over a vast temporal and spatial scale. The results of our study indicate the value of log- book data for scientific studies. In addition to providing immediate mortality rates, logbook data can be used to analyze catch rates, spatiotemporal catch trends, and species distributions (Cheng et al., 2001; Mandelman et al., 2008). The accessibility of the USPL data set to the general public promotes its use and advances our knowledge of fisheries science. Many other long-term government data sets have been collected, but they are not publicly available for analysis. Although we rec- ognize the legal requirements for the protection of a fisherman’s personal information, greater availability of governmental logbook and observer data would al- low for the examination of fishery trends on a world- wide scale and will be necessary to implement effective global strategies for species management. Our results indicate that fisheries management regulations have had a positive effect on the survival of the 2 shark species examined. The establishment of fin-to-carcass ratios has reduced the number of ti- ger shark discarded dead in the US. commercial long- FAO (Food and Agriculture Organization of the United Na- tions). 1999. International plan of action for reducing in- cidental catch of seabirds in longline fisheries. International plan of action for conservation and management of sharks. International plan of action for the management of fishing capacity, 26 p. FAO, Rome. [Available from website.] Dapp et ai.: Immediate mortality of Prionoce glauca and Galeocerdo cuvier caught by pelagic longlines 39 line fishery. Additionally, fin-to-carcass ratios have not increased retention rates for tiger sharks (the mean percentages retained were 16% in 1992-1993 and 14% in 1993-2008) because these sharks are large-bodied animals, boats often have limited deck space, and their meat has a low commercial value (Mandelman et ah, 2008; Simpfendorfer, 2009). Similarly, regulations that require the use of circle hooks during commercial long- line operations have decreased the rate of immediate mortality for blue shark by 8.0%. Our study adds sup- port to the assumption that fisheries regulations have been beneficial to some species, and results from recent studies indicate that species of sharks caught in the U.S. longline fishery are stable or increasing in abun- dance (Carlson et ah, 2012). Acknowledgments Financial support to conduct this work was provided by the Australian Research Council Linkage Project LPl 10200572, the Holsworth Wildlife Research En- dowment, Fisheries Victoria, the Australian Fisher- ies Management Authority, Melbourne Aquarium, and Monash University. We thank 3 anonymous reviewers for providing feedback on previous iterations of our manuscript. Literature cited Afonso, A. S., and F. H. V. Hazin. 2014. Post-release survival and behavior and exposure to fisheries in juvenile tiger sharks, Galeocerdo cuvier, from the South Atlantic. J. Exp. Mar. Biol. Ecol. 454:55-62. Afonso, A. S., F. H. V. Hazin, F. Carvalho, J. C. Pacheco, H. Hazin, D. W. Kerstetter, D. Murie, and G. H. Burgess. 2011. Fishing gear modifications to reduce elasmobranch mortality in pelagic and bottom longline fisheries off Northeast Brazil. Fish. Res. 108:336-343. Afonso, A. S., R. Santiago, H. Hazin, and F. H. V. Hazin. 2012. Shark bycatch and mortality and hook bite-offs in pelagic longlines: interactions between hook types and leader materials. Fish. Res. 131:9-14. Baum, J. K., and R. A. Myers. 2004. Shifting baselines and the decline of pelagic sharks in the Gulf of Mexico. Ecol. Lett. 7:135-145. Baum, J. K., and W. Blanchard. 2010. Inferring shark population trends from generalized linear mixed models of pelagic longline catch and effort data. Fish. Res. 102:229-239. Baum, J. K., R. A. Myers, D. G. Kehler, B. Worm, S. J. Harley, and P. A. Doherty. 2003. Collapse and conservation of shark populations in the Northwest Atlantic. Science 299:389-392. Bonfil, R. 1994. Overview of world elasmobranch fisheries. FAO Fish. Tech. Pap. 341, 119 p. FAO, Rome. Bromhead, D., S. Clarke, S. Hoyle, B. Muller, P. Sharpies, and S. Harley. 2012. Identification of factors influencing shark catch and mortality in the Marshall Islands tuna longline fishery and management implications. J. Fish Biol. 80:1870-1894 Burgess, G. H., L. R. Beerkircher, G. M. Cailliet, J. K. Carlson, E. Cortes, K. J. Goldman, R. D. Grubbs, J. A. Musick, M. K. Musyl, and C. A. Simpfendorfer. 2005. Is the collapse of shark populations in the north- west Atlantic Ocean and Gulf of Mexico real? Fisheries 30:19-26. Campana, S. E., W. Joyce, and M. J. Manning. 2009. Bycatch and discard mortality in commercially caught blue sharks Prionace glauca assessed using archival satellite pop-up tags. Mar. Ecol. Prog. Ser. 387:241-253. Carlson, J. K., L. F. Hale, A. Morgan, and G. Burgess. 2012. Relative abundance and size of coastal sharks de- rived from commercial shark longline catch and effort data. J. Fish Biol. 80:1749-1764. Carruthers, E. H., D. C. Schneider, and J. D. Neilson. 2009. Estimating the odds of survival and identifying mitigation opportunities for common bycatch in pelagic longline fisheries. Biol. Conserv. 142:2620-2630. Cheng, C. H., T. Kawasaki, K. P. Chiang, and C. H. Ho. 2001. Estimated distribution and movement of hairtail Trichiurus lepturus in the Aru Sea, based on the logbook records of trawlers. Fish. Sci. 67:3-13. Cicia, A. M., L. S. Schlenker, J. A. Sulikowski, and J. W. Mandelman. 2012. Seasonal variations in the physiological stress re- sponse to discrete bouts of aerial exposure in the little skate, Leucoraja erinacea. Comp. Biochem. Physiol. A. 162:130-138. Clarke, S. C., J. E. Magnussen, D. L. Abercrombie, M. K. McAl- lister, and M. S. Shivji. 2006. Identification of shark species composition and pro- portion in the Hong Kong shark fin market based on molecular genetics and trade records. Conserv. Biol. 20:201-211. Clarke, S. C., M. P. Francis, and L. H. Griggs. 2013. Review of shark meat markets, discard mortal- ity and pelagic shark data availability, and a proposal for a shark indicator analysis. NZ Fish. Assess. Rep. 2013/65, 78 p. Minist. Prim. Ind., Wellington, New Zea- land. [Available at website, accessed January 2015.] Coelho, R., J. Fernandez-Carvalho, P. G. Lino, and M. N. Santos. 2012. An overview of the hooking mortality of elasmo- branchs caught in a swordfish pelagic longline fishery in the Atlantic Ocean. Aquat. Living Resour. 25:311- 319. Coelho, R., P. Infante, and M. N. Santos. 2013. Application of generalized linear models and gener- alized estimation equations to model at-haulback mortal- ity of blue sharks captured in a pelagic longline fishery in the Atlantic Ocean. Fish. Res. 145:66—75. Cramer, J., and H. Adams. 2000. Large pelagic logbook newsletter — 1998. NOAA Tech. Memo. NMFS-SEFSC-433, 25 p. Curran, D., and K. Bigelow. 2011. Effects of circle hooks on pelagic catches in the Ha- waii-based tuna longline fishery. Fish. Res. 109:265-275. Dapp, D. R., T. 1. Walker, C. Huveneers, and R. D. Reina. 2016a. Respiratory mode and gear type are important de- terminants of elasmobranch immediate and post-release mortality. Fish Fish. 17:507-524. 40 Fishery Bulletin 1 15(1) Dapp, D. R., C. Huveneers, T. I. Walker, M. Drew, and R. D. Reina. 2016b. Moving from measuring to predicting bycatch mortality: predicting the capture condition of a longline- caught pelagic shark. Front. Mar. Sci. 2:126. Dent, F., and S. Clarke. 2015. State of the global market for shark products. FAO Fish. Aquae. Tech. Pap. 590, 187 p. FAO, Rome. Dulvy, N. K., S. L. Fowler, J. A. Musick, R. D. Cavanagh, P. M. Kyne, L. R. Harrison, J. K. Carlson, L. N. K. Davidson, S. V. Fordham, M. P. Francis, et al. 2014. Extinction risk and conservation of the world’s sharks and rays. eLife 2014;3:e00590 Epperly, S. P., J. W. Watson, D. G. Foster, and A. K. Shah. 2012. Anatomical hooking location and condition of ani- mals captured with pelagic longlines: the Grand Banks experiments 2002-2003. Bull. Mar. Sci. 88:513-527. Francis, M. P., L. H. Griggs, and S. J. Baird. 2001. Pelagic shark bycatch in the New Zealand tuna longline fishery. Mar. Freshw. Res. 52:165-178. Frick, L. H., R. D. Reina, and T. I. Walker. 2010a. Stress related physiological changes and post-re- lease survival of Port Jackson sharks (Heterodontus por- tusjacksoni) and gummy sharks {Mustelus antarcticus) following gill-net and longline capture in captivity. J. Exp. Mar. Biol. Ecol. 385:29-37. Frick, L. H., T. I. Walker, and R. D. Reina. 2010b. Trawl capture of Port Jackson sharks, Heterodon- tus portusjacksoni, and gummy sharks, Mustelus antarct- icus, in a controlled setting: effects of tow duration, air exposure and crowding. Fish. Res. 106:344-350. Gallagher, A. J., E. S. Orbesen, N. Hammerschlag, and J. E. Serafy. 2014. Vulnerability of oceanic sharks as pelagic longline bycatch. Global Ecol. Conserv. 1:50-59. Gilman, E., S. Clarke, N. Brothers, J. Alfaro-Shigueto, J. Man- delman, J. Mangel, S. Petersen, S. Piovano, N. Thomson, P. Dalzell, et al. 2008. Shark interactions in pelagic longline fisher- ies. Mar. Policy 32:1-18. Gilman, E., M. Chaloupka, Y. Swimmer, and S. Piovano. 2016. A cross-taxa assessment of pelagic longline by-catch mitigation measures: conflicts and mutual benefits to elasmobranchs. Fish Fish. 17:748-784. Godin, A. C., J. K. Carlson, and V. Burgener. 2012. The effect of circle hooks on shark catchability and at-vessel mortality rates in longlines fisheries. Bull. Mar. Sci. 88:469-483. Griggs, L. H., and S. J. Baird. 2013. Fish bycatch in New Zealand tuna longline fisheries 2006-07 to 2009-10. NZ Fish. Assess. Rep. 2013/13, 73 p. Minist. Prim. Ind., Wellington, New Zealand. [Available from website, accessed May 2015.] Guida, L., T. I. Walker, and R. D. Reina. 2016. Temperature insensitivity and behavioural reduc- tion of the physiological stress response to longline cap- ture by the gummy shark, Mustelus antarcticus. PLoS One ll(2):e0148829. Hight, B. V., D. Holts, J. B. Graham, B. P. Kennedy, V. Taylor, C. A. Sepulveda, D. Bernal, D. Ramon, R. Rasmussen, and N. C. Lai. 2007. Plasma catecholamine levels as indicators of the post-release survivorship of juvenile pelagic sharks caught on experimental drift longlines in the Southern California Bight. Mar. Freshw. Res. 58:145-151. Hoenig, J. M., and S. H. Gruber. 1990. Life-history patterns in the elasmobranchs: impli- cations for fisheries management. In Elasmobranchs as living resources: advances in biology, ecology, systemat- ics, and the status of fisheries. NOAA Tech. Rep. NMFS 90 (H. L. Pratt Jr., S. H. Gruber, and T. Taniuchi, eds.), 1-16 p. Hoffmayer, E. R., J. M. Hendon, and G. R. Parsons. 2012. Seasonal modulation in the secondary stress re- sponse of a carcharhinid shark, Rhizoprionodon terrae- novae. Comp. Biochem. Physiol. A. 162:81-87. Kerstetter, D. W., and J. E. Graves. 2006. Effects of circle versus J-style hooks on target and non-target species in a pelagic longline fishery. Fish. Res. 80:239-250. Mandelman, J. W., P. W. Cooper, T. B. Werner, and K. M. Lagueux. 2008. Shark bycatch and depredation in the U.S. At- lantic pelagic longline fishery. Rev. Fish Biol. Fish. 18:427-442. Morgan, A., and J. K. Carlson. 2010. Capture time, size and hooking mortality of bottom longline-caught sharks. Fish. Res. 101:32-37. Morgan, A., P. W. Cooper, T. Curtis, and G. H. Burgess. 2009. Overview of the Lf.S. East Coast bottom longline shark fishery, 1994-2003. Mar. Fish. Rev.71(l):23-38. Moyes, C. D., N. Fragoso, M. K. Musyl, and R. W. Brill. 2006. Predicting postrelease survival in large pelagic fish. Trans. Am. Fish. Soc. 135:1389-1397. Musyl, M. K., R. W. Brill, D. S. Curran, N. M. Fragoso, L. M. McNaughton, A. Nielsen, B. S. Kikkawa, and C. D. Moyes. 2011. Postrelease survival, vertical and horizontal movements, and thermal habitats of five species of pe- lagic sharks in the central Pacific Ocean. Fish. Bull. 109:341-368. Petersen, S. L. 2008. Understanding and mitigating vulnerable bycatch in southern African trawl and longline fisheries. Ph.D. thesis, 285 p. Univ. Cape Town, Cape Town, South Africa. R Core Team. 2012. R: a language and environment for statistical com- puting. R Foundation for Statistical Computing, Vienna, Austria. [Available from website, accessed July 2013.] Reardon, M. B., L. Gerber, and R. D. Cavanagh. 2006. Isurus paucus. The lUCN Red List of Threatened Species 2006: e.T60225A12328101.. Scott-Denton, E., P. F. Cryer, J. P. Gocke, M. R. Harrelson, D. L. Kinseila, J. R. Pulver, R. C. Smith, and J. A. Williams. 2011. Descriptions of the U.S. Gulf of Mexico reef fish bot- tom longline and vertical line fisheries based on observer data. Mar. Fish. Rev. 73(2): 1-26. Serafy, J. E., E. S. Orbesen, D. J. G. Snodgrass, L. R. Beer- kircher, and J. F. Walter. 2012. Hooking survival of fishes captured by the Unit- edStates Atlantic pelagic longline fishery: impact of the 2004 circle hooks rule. Bull. Mar. Sci. 88:605-621. Simpfendorfer, C. 2009. Galeocerdo cuvier. The lUCN Red List of Threat- ened Species 2009: Smith, S. E., D. W. Au, and C. Show. 1998. Intrinsic rebound potentials of 26 species of Pacific sharks. Mar. Freshw. Res. 49:663-678. Dapp et al.: Immediate mortality of Prionace glauca and Galeocerdo cuvier caught by pelagic longiines 41 Sullivan, G. M., and R. Feinn. 2012. Using effect size — or why the P value is not enough. J. Grad. Med. Educ. 4:279-282. Vannuccini, S. 1999. Shark utilization, marketing and trade. FAO Fish. Tech. Pap. 389, 470 p. FAO, Rome. Watson, J. W., S. P. Epperly, A. K. Shah, and D. G. Foster. 2005. Fishing methods to reduce sea turtle mortality as- sociated with pelagic longiines. Can. J. Fish. Aquat. Sci. 62:965-981. Yokota, K., M. Kiyota, and H. Minami. 2006. Shark catch in a pelagic longline fishery: compari- son of circle and tuna hooks. Fish. Res. 81:337-341. 42 National Marine Fisheries Service NOAA FishBry Bulletin .%» established in 1881 -d=. Spencer F. Baird First U.S. Commissioner of Fisheries and founder of Fishery Bulletin Preliminary assessment of a direct age- determination method for 3 commercially important crustaceans from Jllaska Email address for contact author: rwkilada@gmail.com Abstract — Management of commer- cially important crab and shrimp species in Alaska has been hindered by the inability to directly determine the age of individual animals. We in- vestigated the applicability of a re- cently developed method of age de- termination to red king crab (Para- lithodes camtschaticus), southern Tanner crab (Chionoecetes bairdi), and spot shrimp (Pandalus platycer- os) from Alaska. The cuticle struc- tures of the mesocardiac ossicles of crabs and the eyestalks of spot shrimp were visualized with histo- logical staining to identify the en- docuticle, where growth bands have been observed in other crustaceans. For all species, paired light and dark bands were observed in longitudinal, thin sections of these structures in the majority of specimens exam- ined. The proximal portion of the mesocardiac ossicle, where growth bands were observed, was absent in the foregut exuviae of red king and southern Tanner crabs that molted in captivity. If validated, counts of growth bands hold promise as a reli- able measure for determining age of these species. Manuscript submitted 10 August 2015. Manuscript accepted 25 October 2016. Fish. Bull. 115:42-49 (2017). Online publication date: 15 November 2016. doi: 10.7755/FB.115.1.4 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. Raouf Kilada (contact author)''^ Joel B. Webb^ Kevin W. McNeel^ Laura M. Slater^ Quinn Smith^ Jayde Ferguson^ ' Department of Biology University of New Brunswick (Saint John) 100 Tucker Park Road Saint John, New Brunswick E2L 4L5, Canada ^ Department of Marine Science Suez Canal University Ismailia 41522, Egypt ^ Division of Commercial Fisheries Alaska Department of Fish and Game P.O. Box 115526 Juneau, Alaska 99811-5526 ^ Age Development Unit, Mark, Tag, and Age Laboratory Alaska Department of Fish and Game P.O. Box 115526 Juneau, Alaska 99811-5526 For many fish and invertebrate species, age can be determined di- rectly from growth bands recorded in calcified hard structures. These structures include bones, scales, and otoliths in fish species (Campana, 2001) and statoliths and shell sec- tions in a variety of invertebrates (Jensen, 1969; Kilada et ah, 2007; Abele et ah, 2009). Similar meth- ods have not been applied to de- capod crustaceans because of the presumed loss and replacement of calcified structures during ecdysis (Vogt, 2012). Instead, indirect meth- ods, including captive observations, tag-recapture experiments, accumu- lation of lipofuscin in neural tissue, and analysis of size-frequency dis- 5 Division of Commercial Fisheries Alaska Department of Fish and Game 351 Research Court Kodiak, Alaska 99615 * Division of Commercial Fisheries Alaska Department of Fish and Game P.O. Box 1 10024 Juneau, Alaska 99811-0024 ^ Pathology Laboratory Alaska Department of Fish and Game 333 Raspberry Road Anchorage, Alaska 99518-1599 tributions, have been applied to in- fer age (Hartnoll, 2001; Vogt, 2012; Pinchuk et ah, 2016). A lack of reliable age information impedes assessment and manage- ment of crustacean fisheries (Caddy, 1986). In Alaska, some major crab stocks are assessed and managed by using length-based population mod- els (e.g., Zheng et ah, 1995), in which data on abundance, harvest, growth, and mortality are integrated. Howev- er, the accuracy of these models may be compromised if the growth or mor- tality rates are not truly representa- tive of processes in situ. In contrast, age-structured models implicitly ac- count for variability in growth and mortality by incorporating compre- Kilada et al.: An age-deternnination nnethod for 3 commerciaiiy important crustaceans 43 Table 1 Collection location, date, sex, and number of individual red king crab {Paralithodes camtschaticus), southern Tanner crab {Chionoecetes bairdi), and spot shrimp {Pandalus platyceros) collected in Alaska during 2013-2014 for investigation of pres- ence of growth bands in thin sections of the mesocardiac ossicle of the gastric mill of red king and southern Tanner crabs and in eyestalks of spot shrimp. Species Collection location Collection date Sex n Red king crab Bristol Bay, Alaska June 2013 Female 30 Southern Tanner crab Marmot Bay, Kodiak, Alaska June 2014 Male 34 Spot shrimp Seymour Canal, southeast Alaska February and June 2014 Male and female 30 hensive size- and abundance-at-age data (Quinn and Deriso, 1999). A direct method for determining age based on count- ing bands in the endocuticle layer has been presented for decapod crustaceans (Leland et al., 2011; Kilada et al., 2012). Bands were initially described in the ossicles of the gastric mill for 6 species (Leland et al., 2011; Kilada et al., 2012) and in the eyestalks of 2 additional species (Kilada et al., 2012). These observations have since been extended to additional crab, lobster, and shrimp species and to euphausids (Kilada and Acuna, 2015; Kilada et al., 2015; Sheridan et al., 2015; Kilada and Ibrahim, 2016; Krafft et al., 2016) and have been further supported by technical development, corrobora- tion, and validation (Sheridan et al., 2015; Leland et al., 2015). Our objective was to evaluate the potential of this method for 3 commercially important species in Alaska: the red king crab (Paralithodes camtschaticus), south- ern Tanner crab (Chionoecetes bairdi), and spot shrimp (Pandalus platyceros). Feasibility was evaluated 1) by identifying the endocuticle layer by histological ex- amination, 2) by observing the presence or absence of growth bands in the endocuticle of the mesocardiac os- sicles of crabs and eyestalks of shrimp, and 3) by deter- mining whether the region of the mesocardiac ossicle, where growth bands have been observed, may be re- tained during ecdysis for these crab species. The me- socardiac ossicle was selected as the primary structure for investigation in the 2 crab species because of the presence of clear bands and evidence of possible reten- tion through ecdysis (Leland et al. 2011; Kilada et al., 2012). The eyestalks of shrimp were selected because of the dissimilarity between crab and shrimp gastric mill ossicles and because of the evidence of the presence of bands in the eyestalk of shrimp (Kilada et al., 2012). Materials and methods Histological examination Red king crab and spot shrimp were collected in south- east Alaska in 2014, by using pot gear, and south- ern Tanner crab were collected near Kodiak, Alaska, in 2014 by using trawl gear (n=3 for each species). Gastric mills of red king and southern Tanner crabs and paired eyestalks of spot shrimp were dissected and preserved in Bouin’s fixative for approximately 1 month and then transported to the Fish Pathology Lab- oratory of the Alaska Department of Fish and Game in Anchorage, Alaska. Before processing, mesocardiac ossicles of some red king crabs were trimmed to fit within histological cassettes (1 cm^), but most ossicles and eyestalks did not require trimming. Structures were transferred to 70% ethanol and decalcified with Evans and Krajian fluid (Evans and Krajian, 1930). Tissues were dehydrated, infiltrated, and embedded in paraffin with an automatic tissue processor. Histologi- cal cassettes were cut longitudinally into 6-pm sections with a rotary microtome, and sections were mounted onto glass slides. Sections of mesocardiac ossicles of crabs and eyestalks of spot shrimp were prepared and stained with Masson’s trichrome (Thermo Fisher Scientific^, Waltham, MA) and permanently mounted with Permount mounting medium (Thermo Fisher Sci- entific). Masson’s trichrome was expected to stain the endocuticle and exocuticle layers blue and the mem- branous layer (hypodermis) and epicuticle red. The cuticle layers (for detailed description, see Vatcher et al., 2015) were then examined with a Zeiss microscope (Carl Zeiss Microscopy, Jena, Germany) and photo- graphed with a Jenoptik digital camera and ProgRes CapturePro software (JENOPTIK Optical Systems Inc., Jupiter, FL). Presence of growth bands Red king crab, southern Tanner crab, and spot shrimp (n=30, 34, and 30, respectively) of a range of body sizes were collected across Alaska in 2013 and 2014 (Table 1). Carapace length (CL) and shell con- dition of red king crabs, carapace width (CW) and shell condition of southern Tanner crabs, CL of spot shrimp, and sex (identified from pleopod morphologi- cal features for shrimp) of all specimens were record- 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. 44 Fishery Bulletin 115(1) A Figure 1 Locations of the longitudinal axes used for prep- aration of thin sections indicated by red arrows for (A) the eyestalk of spot shrimp (Pandalus platyceros) and (B) the dorsal ossicle (urocar- diac and mesocardiac) of red king {Paralithodes camtschaticus) and southern Tanner {Chionoece- tes bairdi) crabs, as illustrated with the ossicle of a southern Tanner crab. Red king and southern Tanner crabs and spot shrimp were collected in Alaska during 2013-2014 for this study of a di- rect age-determination method. ed at the time of collection. Shell condition, a subjec- tive index of epibionts and wear on the exoskeleton, was noted because it has proven useful for evaluating differences in the recency of molting in crab species (Jadamec et al., 1999; Donaldson and Byersdorfer, 2002). Crab gastric mills and shrimp eyestalks were dissected, cleaned, and preserved in a mixture of eth- anol, glycerol, and distilled water (with a volume ra- tio of 70:4:26). Structures were embedded in cold cure epoxy resin and sectioned longitudinally (Fig. 1) with a diamond- bladed IsoMet Low Speed Saw (Buehler, Lake Bluff, IL) at the University of New Brunswick in Saint John, New Brunswick, Canada. Several serial sec- tions (with thickness of 160-180 pm) were prepared per structure and mounted with epoxy individual- ly on slides, polished by hand with dry, 0.3-pm-grit lapping film, covered with 90% ethyl alcohol, and viewed under transmitted light with a CX41 Olym- pus compound microscope (Olympus Corp., Tokyo) at 100-400X magnification. All cuticular layers were ex- amined throughout the thin sections, and bands were recognized as alternating (bipartite) translucent and opaque zones in the endocuticle of crab mesocardiac ossicles and shrimp eyestalks. Photographs were tak- en with a DP72 Olympus digital camera (Olympus Corp.) attached to the microscope. Examination of exuviaS gastric structures Two female red king crab were captured by divers in southeast Alaska and 3 male southern Tanner crab were captured by pot and trawl gear near Kodiak, Alaska, in 2014. Female red king crabs were immature and grasped by males in situ (precopulatory guarding of mate), indicating that the pubertal molt to matu- rity was imminent. All crabs were held in aquaria with flow-through seawater at ambient seawater tempera- ture and monitored daily until they molted. After ec- dysis, the exuvial gastric mill structures were removed and photographed. At 1 week after the molt, the whole stomach was dissected from the crab, and the gastric mill structures were photographed for comparison with the same structures from the exuviae. Results Histological examination of the cuticle The eedocuticular layer was visually differentiated from other cuticular layers in the mesocardiac ossicles of red king and southern Tanner crabs (Fig. 2) and in the eyestalks of spot shrimp (Fig. 3) after ossicles were stained with Masson’s trichrome. The epicuticle, exocu- ticle, and membranous layers were clearly visible in thin sections of crab mesocardiac, where red indicated the membranous layer and epicuticle (Fig. 2). Lamellar structure, of alternating light and dark stained bands, were more discernible in the mesocardiac of a southern Tanner crab (Fig. 2B) and the exo- and endocuticle of the eyestalk of a spot shrimp (Fig. 3B) than in the me- socardiac of a red king crab (Fig. 2C) at similar mag- nification. The cuticle layers of the eyestalks of spot shrimp tended to separate during histological section- ing (Fig. 3A). Kilada et al.i An age-determination method for 3 commercially important crustaceans 45 A Figure 2 Histological sections (6-pm thickness), stained with Masson’s trichrome, of the (A) dorsal ossicle of a red king crab (Paralithodes camtschaticus) that was col- lected in southeast Alaska in 2014 and used to represent both the red king crab and the southern Tanner crab (Chionoecetes bairdi). Included are the mesocar- diac (m), the urocardiac (u), and the medial tooth (mt). Higher magnification of the ossicle region denoted by the box in panel A revealed the cuticular layers of the arch of the mesocardiac, as evident in the high-magnification images of 2 other specimens collected in 2014: (B) the southern Tanner crab and (C) the red king crab. The epicuticle (Ep) and membranous layer (MI) stained red and the exocuticle (Ex) stained a darker shade of blue than the endocuticle (En). Presence of growth bands Thin sections were successfully obtained from the me- socardiac ossicles from 19 red king crabs and 32 south- ern Tanner crabs and single eyestalks from 18 spot shrimp. For mesocardiac sections of both crab species, bipartite band patterns (exclusive of the lamellae) were clearest in the location of maximum endocuticle thick- ness in relation to the exo- and epicuticle that was at the proximal end of the structure (Figs. 2A, 4, A-B). For spot shrimp, bands were clearest at the proximal end of the eyestalk with respect to the anterior tip of the cephalothorax (Figs. 3A, 4C). Ossicular retention throughout ecdysis Comparison of gastric mill structures in exuviae with those of postmolt stomachs of crabs indicated differ- ences in presence of regions of the gastric mill. The cusp of the medial tooth of the urocardiac and zygocar- diac ossicles of the gastric mill, along with the portion of the ossicle adjacent to each tooth, were the primary structures present in the exuviae of both southern Tan- ner and red king crabs (Fig. 5A). For both crab species, the anterior portion of mesocardiac, the pterocardiac, and the anterior portions of the zygocardiac ossicles were not visible in the exuviae, but they were present and robust in a postmolt crab, indicating that these portions were potentially retained or resorbed and sub- sequently replaced (Fig. 5, B and D). Discussion The presence of bands in the endocuticle layer of the mesocardiacs of southern Tanner and red king crabs and in the eyestalks of spot shrimp indicated that de- termining age on the basis of band counts may be fea- sible, as it is for other crustacean species (e.g., Leland et al., 2011; Kilada et al., 2012; Leland et al., 2015; Kilada et al., 2015; Sheridan et al., 2015; Kilada and Ibrahim, 2016; ICrafft et al., 2016). Histological charac- 46 Fishery Bulletin 115(1) :l Figure 3 Histological sections (B-pm thickness), stained with Masson’s trichrome, of the eyestalk of a spot shrimp (Pandalus platyceros) collected in Alaska in 2014, show- ing (A) the entire eyestalk and (B) the cuticular layer- ing, as seen in the inset in the smaller, lower small box in panel A, including the epicuticle (Ep), exocuticle (Ex), endocuticle (En), and membranous layer (Ml). The larger box indicates the location within the structure where growth bands were observed in Figure 4C. terization of the mesocardiac ossicles of the crab spe- cies and of the eyestalk of the spot shrimp defined the boundaries between cuticular layers — boundaries that are critical for developing band counting criteria. As with results from studies of other species (e.g., Kilada et ah, 2012; Leland et ah, 2015), there was prelimi- nary evidence that a portion of the mesocardiac ossicle, where band patterns are observed in the endocuticle, is either retained or replaced during molting. A structure that is retained through ecdysis would be useful for evaluating band information. The cuticle layers of the mesocardiac ossicles of red king and southern Tanner crabs and the eyestalks of spot shrimp were similar to those observed in other studies of decapod crustaceans (see Roer and Dillaman, 1984, and references therein; Vatcher et ah, 2015). Differences in staining results may be due to differ- ences in the specific calcium compound involved in the biomineralization within each region (Vatcher et al., 2015). As in results for blue crab (Callinectes sapidus) with the use of acridine orange (a different histological stain) (Vatcher et ah, 2015), the medial tooth of the urocardiac stained similarly and was continuously con- nected to the epicuticle in both crab species examined in this study (Fig. 2A). This result further supports the hypothesis that the hardened cusp of the tooth is of epicuticular rather than exocuticular origin (Vatcher et ah, 2015). Bipartite patterns were readily visible in the me- socardiac ossicles of red king and southern Tanner crabs and in the eyestalks of spot shrimp. Recurrence of this pattern in multiple individuals indicates that band counts may be promising as indicators of growth variability through the lifetime of these species. Before this experiment, bands were observed in eyestalks of snow crab (C. opilio), which is a congener of the south- ern Tanner crab (Kilada et ah, 2012). We evaluated the mesocardiac of the southern Tanner crab because of the possibility that this structure may be retained through ecdysis (Kilada et al., 2012; Leland et ah, 2015; but also see Vatcher et al., 2015). To our knowledge, de- scription of growth bands in red king crabs is a first for the family Lithodidae. The appearance of bands in spot shrimp was very similar to that observed in the eye- stalks of a congener, the northern shrimp (P. borealis) (Kilada et al., 2012). Finally, most notably for the red king crab, a high proportion of the structures evaluated for growth bands were damaged during embedding or preparing thin sections. Structures were, by necessity, shipped dry before they were embedded in resin, and this condition likely contributed to their fragility and high fracture rate during the embedding and sectioning processes. Embedding structures before shipping could effectively mitigate this problem. The absence of the basal region of the mesocardiac (where band patterns are observed in thin sections) in exuviae of southern Tanner and red king crabs indicat- ed that this portion of the ossicle was either retained or replaced within 1 week after molting (Fig. 5, A and C). Recently it has been hypothesized that the endocuticle region of the pterocardiac and mesocardiac ossicles is retained through molting (Kilada et al., 2012; Leland et al., 2015). Calcein marks in the endocuticle were visi- ble after several molts and portions of the ossicles were absent in the exuviae of lobster and crayfish species. However, for the blue crab, histological characteriza- tion of the cuticular layers of the dorsal ossicle (dor- somedial tooth) indicated that the dorsal cuticle (Roer and Dillaman, 1984), like the endocuticle, is resorbed during the premolt stage and resynthesized during the postmolt stage (Vatcher et al., 2015). As with brachy- uran crabs (Brosing, 2014), the gastric teeth differed Kilada et al.: An age-determination method for 3 commercially important crustaceans 47 Figure 4 Bipartite growth band patterns in the endocuticle of the arch of the mesocardiac ossicle of the gastric mill (for reference, see Fig. 2) of (A) a red king crab (Paralithodes camtschaticus) collected in Bristol Bay, Alaska, in June 2013; (B) a southern Tanner crab (Chionoecetes bairdi) collected in Marmot Bay, Alaska, in June 2014; and (C) in the proximal portion of the eyestalk of a spot shrimp {Pandalus platyceros) collected in Seymour Canal, Alaska, in May 2014, all under transmitted light. The endocuticle (En), exocuticle, (Ex), and membranous layer (Ml) are labeled for orientation. in biomineralization from other portions of the gastric mill (Vatcher et al., 2015) and remained in the exuviae of both crab species examined in this study. Our results are an initial step toward developing and further evaluating growth bands as a possible in- dicator of age for commercially valuable crustaceans in Alaska. Key areas for further research include evalua- tions of variability (precision) in band count and clar- ity among the primary ossicles (pterocardiac, zygocar- diac, and mesocardiac) of the gastric mill (Leland et al., 2011; Leland et al., 2015; Sheridan et al., 2015), innovation in preparation techniques for thin sections (Sheridan et ah, 2015), definitions of criteria for iden- tifying bands (e.g., Leland et al. 2015), determination of the fate (retention or replacement) of the endocuticle during molting (Vatcher et ah, 2015), and corroboration of band counts with current understanding of species- specific growth, life history, and longevity based on in- direct methods. Ultimately, hypotheses should also be developed regarding the mechanism by which growth bands are formed and retained in structures that are molted (shrimp eyestalk) or possibly retained (e.g., gas- tric mill ossicles) (Kilada et ah, 2012; Leland et ah, 2015; Vatcher et ah, 2015). Rigorous validation of bands as indices of age will also be necessary before their application in stock assessment and fisheries management (Beamish and McFarlane, 1983; Campana, 2001; Leland et ah, 2011). Validation techniques potentially applicable to the species investigated in this study include the use of autofluorescent stains and the use of specimens with known ages. Autofluorescent stains, such as cal- cein, can create discrete marks in calcified hard parts that can be used to examine band deposition with 48 Fishery Bulletin 1 15(1) Figure 5 (A) Image of the primary ossicles of the gastric mill seen in the exuvia of a red king crab {Paralithodes carntschaticus) collected in Bristol Bay, Alaska, in June 2013. Included are the lateral teeth of the zygocardiac ossicle (z) and the medial tooth (mt) and urocardiac (u) of the dorsal (uro- and mesocardiac) ossicle, showing that the portion of the ossicle containing the arch of the mesocardiac (m, indicated by the box and illustrated in Fig. 2) was absent. (B) Im- age of the dorsal ossicle of the same red king crab 1 week after ecdysis, showing that the entire mesocardiac was present (boxed). (C) The base of the dorsal ossicle for the exuvia of a southern Tanner crab {Chionoecetes bairdi) collected in Marmot Bay, Alaska, in June 2014, showing that the urocardiac was present but the mesocardiac (boxed) and pterocardiac ossicles were absent. (D) Image of the gastric mill ossicles, showing that the mesocardiac (boxed) and pterocardiac ossicles (p) were present in the ossicle of the southern Tanner crab 1 week after ecdysis. elapsed time after marking (Kilada et al., 2012; Le- land et al., 2015). Further, determining band counts for crustaceans for ^vhich ages are known (e.g., ani- mals reared in captivity) will be necessary to further understand dynamics of band formation (Leland et al., 2015). Acknowledgments This project was funded by the Alaska Department of Fish and Game. Thanks are extended to C. Siddon and D. Pengilly of the Alaska Department of Fish and Game for their support and to D. Oxman for an earlier review of this manuscript. Three anonymous reviewers contributed to significant improvement of the manu- script through review of a previous draft. This short contribution is professional publication number PP-280 of the Alaska Department of Fish and Game. Literature cited Abele, D., T. Brey, and E. Philipp. 2009. Bivalve models of aging and the determination of molluscan lifespans. Exp. Gerontol. 44:307-315. Beamish, R. J., and G. A. McFarlane. 1983. The forgotten requirement for age validation in fisheries biology. Trans. Am. Fish. Soc. 112:735-743. Brosing, A. 2014. Foregut structures of freshly moutled exuviae from Maja crispate, Cancer pagurus and Pseudosesarma moeschi (Decapoda: Brachyura). J. Nat. Hist. 48:543- 555. Kilada et al.: An age-determination method for 3 commercially important crustaceans 49 Caddy, J. F. 1986. Modelling stock-recruitment processes in Crusta- cea: some practical and theoretical perspectives. Can. J. Fish. Aquat. Sci. 43:2330—2344. Campana, S. E. 2001. Accuracy, precision, and quality control in age de- termination, including a review of the use and abuse of age validation methods. J. Fish Biol. 59:197-242. Donaldson, W. E., and S. C. Byersdorfer. 2005. Biological field techniques for lithodid crabs. Alas- ka Sea Grant College Program, AK-SG-05-03, 82 p. Univ. Alaska, Fairbanks. AK. Evans, N. and A. Krajian. 1930. New method of decalcification. Arch. Path. 10:447. Hartnoll, R. G. 2001. Growth in Crustacea — twenty years on. In Ad- vances in decapod crustacean research: proceedings in the 7^^ colloquium Crustacea Decapoda Mediterranea; Lisbon, Portugal, 6-9 September 1999 (J. P. M. Paula, A. A. V. Flores, and C. H. J. M. Fransen, eds.), p. 111- 122. Springer Science-i-Business Media, Dordrecht, Netherlands. Jadamec, L. S., W. E. Donaldson, and P. Cullenberg. 1999. Biological field techniques for Chionoecetes crabs. Alaska Sea Grant College Program AK-SG-99-02, 80 p. Univ. Alaska, Fairbanks, AK. Jensen, M. 1969. Age determination of echinoids. Sarsia 37:41-44. Article Kilada, R., and E. Acuna. 2015. Direct age determination by growth band counts of three commercially important crustacean species in Chile. Fish. Res. 170:134-143. Kilada, R., and N. K. Ibrahim. 2016. Preliminary investigation of direct age determina- tion using band counts in the gastric mill of the blue swimmer crab (Portunus pelagicus Linneus, 1758) in two salt-water lakes in the eastern Mediterranean. J. Crust. Biol. 36:119-128. Kilada, R. W., D. Roddick, and K. Mombourquette. 2007. Age determination, validation, growth and mini- mum size of sexual maturity of the greenland smooth- cockle (Serripes groenlandicus , Bruguiere, 1789) in east- ern Canada. J. Shellfish Res. 26:443-450. Kilada, R., B. Sainte-Marie, R. Rochette, N. Davis, C. Vanier, and S. Campana. 2012. Direct determination of age in shrimps, crabs, and lobsters. Can. J. Fish. Aquat. Sci. 69:1728 — 1733. Kilada, R., A.-L. Agnalt,, N. H. Arboe, S. Bjarnason, A. Bur- meister, E. Farestveit, 0. S. Gfslason, A. Gudlaugsdottir, D. GuSmundsdottir, J. P. Jonasson, et al. 2015. Feasibility of using growth band counts in age de- termination of four crustacean species in the Northern Atlantic. J. Crust. Biol. 35:499-503. Krafft, B. A., M. Kvalsund, G. Spvik, E. Farestveit, and A.-L. Agnalt. 2016. Detection of growth zones in the eyestalk of the Antarctic krill Euphausia superba (Dana, 1852) (Eu- phausiacea). J. Crust. Biol. 36:267-273. Leland, J. C., J. Coughran, and D. J. Bucher. 2011. A preliminary investigation into the potential value of gastric mills for ageing crustaceans. In New frontiers in crustacean biology: proceedings of the TCS summer meeting, Tokyo, 20-24 September 2009. Crustacean Monographs 15 (A. Asakura, ed.), p. 57-68. Leland, J. C., D. J. Bucher, and J. Coughran 2015. Direct age determination of a subtropical freshwa- ter crayfish (redclaw, cherax quadricarinatus) using os- sicular growth marks. PLoS ONE 10(8):e0134966. Pinchuk, A. L, H. R. Harvey, and G. L. Eckert. 2016. Development of biochemical measures of age in the Alaskan red king crab Paralithodes camtschaticus (Anomura): validation, refinement and initial assess- ment. Fish. Res. 183:92-98. Quinn, T. J., II, and R. B. Deriso. 1999. Quantitative fish dynamics, 560 p. Oxford Univ. Press Inc., New York. Roer, R., and R. Dillaman. 1984. The structure and calcification of the crustacean cu- ticle. Am. Zool. 24:893-909. Sheridan, M., R. Officer, 1. O’Connor, and C. Lordan. 2015. Investigating the feasibility of using growth incre- ments for age determination of Norway lobster (Nephrops norvegicus) and brown crab {Cancer pagurus). J. Crust. Biol. 35:495-498. Vatcher, H. E., R. D. Roer, and R. M. Dillaman. 2015. Structure, molting, and mineralization of the dorsal ossicle complex in the gastric mill of the blue crab, Cal- linectes sapidus. J. Morphol. 276:1358-1367. Vogt, G. 2012. Ageing and longevity in the Decapoda (Crustacea): a review. Zool. Anz. 251: 1-25. Zheng, J., M. C. Murphy, and G. H. Kruse. 1995. A length-based population model and stock-recruit- ment relationships for red king crab, Paralithodes camts- chaticus, in Bristol Bay, Alaska. Can. J. Fish. Aquat. Sci. 52:1129-1246. 50 National Marine Fisheries Service NOAA Abstract— Identifying effective methods of reducing shark bycatch in hook-based fisheries has received little attention despite reports of declines in some shark populations. Previously proposed shark bycatch mitigation measures include gear modifications, time and area clo- sures, avoidance of areas with high shark abundance, use of repellents, and use of specific bait types. Re- gardless of the method of shark by- catch reduction, knowledge of the effects of the chosen method on the catch rates of targeted fish species should be understood. To examine the effects of bait type on catch rates of sharks and teleosts on bot- tom longline gear, standardized gear was deployed with bait alternating between Atlantic mackerel {Scomb- er scombrus) and northern shortfln squid (Illex illecebrosus). For all shark species examined, except the scalloped hammerhead (Sphyrna lewini), a preference for hooks baited with Atlantic mackerel was observed. Commercially and recreationally im- portant teleosts had no significant preference for a specific bait, with the exception of the red drum {Sci- aenops ocellatus), which had a sig- nificant preference for hooks baited with northern shortfin squid. Bait preference decreased as total catch rate increased on individual longline sets. Our results point to the use of specific baits as a viable method to reduce shark catch rates without de- creasing catches of targeted teleosts. Manuscript submitted 10 November 2015. Manuscript accepted 27 October 2016, Fish. Bull. 115:50-59 (2017). Online publication date: 10 November 2016. doi: 10.7755/FB.115.1.5 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. Fishery BuHetin ^ established in 1881 Spencer F. Baird First U.S. Commissioner of Fisheries and founder of Fishery Bulletin Inflyence of bait type on catch rates of predatoi^ fish species on bottom longline gear in the northern Gulf of Mexico William B. Driggers ill (contact author)' Matthew D. Campbell' Kristin M. Hannan^ Eric R. Hoffmayer' Christian M. Jones' Lisa M. Jones' Adam G. Pollack^ Email address for contact author: william.driggers@noaa.gov ' Mississippi Laboratories Southeast Fisheries Science Center National Marine Fisheries Service, NOAA P.O. Drawer 1207 Pascagoula, Mississippi 39567 2 Riverside Technology Inc. Mississippi Laboratories Southeast Fisheries Science Center National Marine Fisheries Service, NOAA P.O. Drawer 1207 Pascagoula, Mississippi 39567 Because of reported declines in some shark populations, there has been increasing interest in mitigat- ing bycatch rates of shark species in longline fisheries that target teleosts (Francis et al., 2001; Beerkircher et al., 2002; Gilman et al., 2008; Ward et al., 2008). Within the northern Gulf of Mexico, commercial fishing for sharks has occurred since the 1920s (Rogers, 1920) and continues to present day. Although directed commercial fishing effort for sharks in the region has waxed and waned over the years, shark bycatch con- tinues to be an important source of mortality, particularly in hook-based fisheries, such as the snapper and grouper bottom longline fishery (e.g., Scott-Denton et a!., 2011). Proposed or enacted efforts to cur- tail shark bycatch in longline fisher- ies have included gear modifications (Ward et al., 2008), reductions in gear soak time (Carruthers et al., 2011), adjustments in fishing depth (Rey and Mueoz-Chapuli, 1991), time and area closures (Watson et al., 2009), avoidance of areas of known high shark abundance (Walsh et al., 2009), use of repellents (Robbins et al., 2011) and use of specific bait types (Gilman et aL, 2007). Ultimate- ly, for any bycatch reduction method to be fully embraced within a fishery, it will be necessary that catch rates of targeted species not be negatively affected when a specific approach is applied. Ideally, a bycatch reduction approach would minimize bycatch rates and not affect catch rates of target species and cause the least amount of economic hardship and changes to proven fishing practices. Among measures proposed to miti- gate shark bycatch, use of bait that Driggers et al.: Influence of bait type on catch rates of predatory fish on longline gear 51 does not decrease catch rates of target species yet re- duces rates of shark capture could be the most easily implemented and likely to be readily accepted within a fishery. Past studies have shown that catch rates of certain shark species are affected by the use of specific bait types. For example, Gilman et al. (2007) analyzed data collected in the pelagic longline fishery in Hawaii and determined that the catch rate of blue sharks (Priona- ce glauca) was reduced by 36% when fish, rather than squid, were used as bait. Similarly, Watson et al. (2005) conducted experimental longline sets in the western North Atlantic Ocean and found catch rates of blue sharks were 31-40% (depending on hook type) lower on hooks baited with fish than on hooks baited with squid. Although results of these studies are potentially biased by use of multiple hook types and sizes, their results strongly indicate that use of a specific bait type could be a means to reduce bycatch of sharks. Hook-based fisheries in the northern Gulf of Mexico that target grouper (Serranidae), snapper (Lutjani- dae), and tilefish (Malacanthidae) species frequently capture sharks (Gulak et al., 2013). For example, the Atlantic sharpnose shark (Rhizoprionodon terraenouae) was the sixth-most captured fish species among the ap- proximately 180 fish species reported in observer data collected from the bottom longline fishery for reef fish in the Gulf of Mexico (Scott-Denton et al., 2011). Fur- thermore, smoothhound (Mustelus spp.) and blacknose (Carcharhinus acronotus) sharks were among the 20 most frequently captured fish taxa within the same fishery (Scott-Denton et al., 2011). The size and type of hooks vary in the bottom longline fishery for reef fish in the Gulf of Mexico (Gulak et al., 2013), and bait type is inconsistent, depending on personal preference, availability, and price, among other factors (Prytherch, 1983; Scott-Denton et al., 2011). Because there is high variability in the gear and bait used within this fish- ery, use of observer data to examine potential effects of bait type on the catch rates of specific species is prob- lematic. Our goal was to conduct a controlled experi- ment to test the effects of 2 readily available bait types commonly used in the bottom longline fishery for reef fish on catch rates of sharks and economically impor- tant teleosts in this region. Additionally, we examined density-dependent effects on preferences for the 2 bait types for both groups of fish species. Materials and methods Bottom longline gear was deployed from the NOAA Ship Oregon II at sampling sites in the northern Gulf of Mexico from 11 March through 13 April 2015. Sam- pling sites were selected on the basis of 18.5-km grids within predefined geographic areas (from Pascagoula, Mississippi, to Cape San Bias, Florida) and depth (9-1000 m) constraints. However, obstructions (e.g., other vessels, reefs, petroleum platforms, and safety fairways) caused the locations of some sampling sites to be different from the positions that were originally planned. The bounds of the sampling universe were se- lected to maximize sampling opportunities within tem- poral limitations. Bottom longline gear consisted of 1842 m of 4.0-mm monofilament mainline and 100 gangions. Each gan- gion was 3.7 m in length and constructed of an AK snap (size 150), 3.2 m of 3.0-mm-diameter monofilament, 0.5 m of 2.4-mm-diameter fishing wire, and a 15/0 circle hook (Mustad #39960Di, O. Mustad & Son A.S, Gjorvik, Norway). Each gangion was baited with Atlantic mack- erel {Scomber scombrus) or northern shortfin squid {II- lex illecebrosus). Both bait types were cut so that they were of the same approximate dimensions and appro- priately sized for the hook. Gangions were deployed so that each bait type alternated along the length of the mainline (i.e., northern shortfin squid, Atlantic mack- erel, northern shortfin squid, Atlantic mackerel), and the starting bait type was selected randomly for each longline set. Gear soak time, defined as time elapsed between de- ployment of the terminal high flyer during gear deploy- ment and retrieval of the first high flyer during haul- back, was approximately 1 h at sampling sites with depths less than 400 m and 2 h at sampling sites with depths greater than 400 m. However, the actual time each hook spends in the water can vary due to a num- ber of factors, such as hook position along the mainline, differences in gear setting and retrieving speeds, and delays related to handling times that were associated with the number of organisms captured. Therefore, the time each hook entered the water at deployment and exited the water during retrieval was electronically monitored. The elapsed time between deployment and retrieval of each hook was considered hook soak time. The status of each retrieved hook was monitored and recorded as whole bait present, partial bait pres- ent, no bait present, missing hook, or organism cap- tured. To determine whether baits of Atlantic mackerel and northern shortfin squid were retained equally on hooks, data from gangions classified with a status of whole bait present were compared by using chi-square tests with Yates correction for continuity. Using the same test, we investigated potential differences in bait retention between Atlantic mackerel and northern shortfin squid when there were hook interactions with feeding organisms other than those retained on a hook. For this investigation, the categories partial bait pres- ent and no bait present were combined. We combined them because of the subjective nature of the partial bait present category (i.e., the status of a hook was classified the same whether, for example, a small piece of bait tissue had been removed or most of the bait had been removed). The category missing hook was not included in analyses because, in the limited number of cases for which this status was recorded, the hook 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. 52 Fishery Bulletin 1 15(1) Table 1 Catch composition and length ranges, by bait type, of fish species caught on bottom longline gear in the northern Gulf of Mexico in March and April 2015. The bait types are Atlantic mackerel {Scomber scombrus) and northern shortfin squid {Illex illecebrosus). All measurements are reported in fork length, with the exceptions of values for tilefish {Lopholatilus chamae- leonticeps) and red drum (Sciaenops ocellatus) that are reported in total length, as well as values for clearnose skates {Raja eglanteria) that are reported in disc width. Results of Kolmogorov-Smirnov (K-S), Mann-¥/hitney (M-W), and chi-square (with Yates correction) tests, used to test for differences in species-specific length distributions, median lengths, and catch rates between bait types, respectively, are provided with associated P-values. Species Atlantic mackerel n size range (mm) Northern shortfin squid n size range (mm) K-S {P) M-W (P) Chi- square {P) Carcharhinus acronotus 61 563-1053 12 587-1061 1.19 (0.12) 238.0 (0.06) 31.56 (<0.01) Carcharhinus plumbeus 45 1025-1800 7 1183-1546 0.95 (0.33) 98.0 (0.13) 26.33 (<0.01) Centrophorous uyato 43 610-951 21 478-948 0.77 (0.59) 419.5 (0.65) 6.89 (<0.01) Galeocerdo cuvier 20 910-2620 10 785-1812 1.12 (0.16) 52.5 (0.05) 2.70 (0.10) Hyporthodus flavolimbatus 14 593-865 6 560-795 0.62 (0.83) 31.5 (0.83) 2.45 (0.12) Lopholatilus chamaeleonticeps 12 497-895 14 465-905 0.70 (0.72) 97.0 (0.52) 0.04 (0.85) Lutjanus campechanus 56 370-876 45 536-860 0.55 (0.92) 1208.0 (0.89) 0.99 (0.32) Mustelus canis 18 754-1187 6 810-1081 0.86 (0.45) 39.0 (0.42) 5.04 (0.02) Mustelus sinusmexicanus 40 686-1167 24 713-1120 0.76 (0.62) 379.0 (0.34) 3.52 (0.06) Raja eglanteria 7 384-410 16 349-446 0.78 (0.57) 47.0 (0.97) 2.78 (0.09) Rhizoprionodon terraenovae 344 574-895 139 486-860 0.69 (0.73) 21,969.0 (0.80) 86.16 (<0.01) Sciaenops ocellatus 5 825-910 21 838-1005 1.32 (0.06) 82.0 (0.06) 8.65 (<0.01) Sphyrna lewini 12 860-1890 9 750-1860 1.13 (0.14) 78.0 (0.09) 0.19 (0.66) Squalus complex 21 410-680 10 445-655 0.64 (0.80) 89.5 (0.66) 3.22 (0.07) could have been hung on the bottom and, as a result, gear failure could have occurred during retrieval. Ad- ditionally, to determine whether retention of the 2 bait types differed with increasing soak time, the distribu- tion of hook soak times for hooks retrieved with whole bait present of Atlantic mackerel was compared to the distribution of soak times for hooks retrieved with whole bait present of northern shortfin squid by using a Kolmogorov-Smirnov test. All captured individuals were identified to the low- est possible taxonomic level and measured to the near- est millimeter. Fork length (FL) and total length (TL) were measured from the tip of the rostrum to the cau- dal notch and to the tip of the upper lobe of the cau- dal fin while the fin was fully extended, respectively. For batoids, disc width was measured between apices of the pectoral fins. Species-specific identifications were not possible in a limited number of instances as a re- sult of a fish escaping a hook before being landed (i.e., catch was confirmed but identification was not possi- ble). In those cases, all individuals could be identified to at least the family level. A Kolmogorov-Smirnov test was used to test for differences in the species-specific length distributions of individuals caught on each bait type. When length distribution data of a species were normally distributed (as indicated by values of kurtosis and skewness being between -2 and 2) and homosce- dastic (as assessed with an F-test), a ^-test was used to determine whether differences existed in species-spe- cific mean length at capture for each bait type. When length distribution data failed to meet the assumptions of parametric statistics, a Mann-Whitney W test was used to compare median length of each species caught by bait type. j Chi-square tests with Yates correction for continuity i v/ere used to determine whether there was a bait-relat- ed effect on species-specific catch rates. Species includ- ed in analyses were limited to those that had a mini- mum of 20 individuals captured regardless of bait type (Table 1). Additionally, because of low species-specific capture rates and the close morphological similarities i of the Cuban dogfish (Squalus cubensis) and the short- spine spurdog {S. mitsukurii) (both species with fewer than 20 individuals captured), these squalid shark spe- cies were treated as a complex. For those species that showed a significant bait preference and were captured on a minimum of 10 longline sets, we examined if there was a change in the degree of bait preference with in- creasing total catch. For this examination, an index of bait preference (IBP) was calculated with the following equation: IBP = {number of individuals captured on mackerel - number of individuals captured on squid) / total number of individuals captured. The IBP values ranged from 1 (all individuals cap- tured with Atlantic mackerel bait) to -1 (all individuals captured with northern shortfin squid bait). The value of 0 indicated that an equal number of individuals were caught on both bait types. To obtain a nonparametric estimation of the relationship between species-specif- ic IBP and total catch, a locally weighted scatterplot Driggers et al.; Influence of bait type on catch rates of predatory fish on longline gear 53 Figure 1 Map of sites where sampling was conducted with bottom longline gear in the northern Gulf of Mexico in March and April 2015. Black circles represent each sampling site. The grey lines indicate the isobaths at 10, 50, 100, 200, 500, and 1000 m. smoothing (LOWESS) with a 75% smoothing factor was applied to scatterplots to determine whether a trend was present in the relationship. All statistical tests were considered significant at an a level of 0.05. Results During March and April 2015, 131 longline sets were completed (Fig. 1) for this study, resulting in deploy- ments of 13,100 hooks and the capture of 1196 indi- vidual fish. Hook soak times ranged from 77 to 257 min (mean of 118.6 min, standard deviation [SD] 39.4), and hook-specific bait status was recorded for 12,888 gan- gions. Bait status of 212 hooks was not recorded due to technical difficulties; however, these hooks represented less than 2% of the gangions deployed. Of the moni- tored hooks, 8 were retrieved with the hook missing as a result of crimp failure (e.g. splitting or slipping). Of the 1944 gangions that were retrieved with whole bait present, there was no significant difference in the num- ber of Atlantic mackerel (n=956) and northern shortfin squid (/z=988) retained on hooks (x2=:0.49, P=0.48). This result indicates that both bait types were equally re- tained on hooks during deployment, soak, and retrieval in the absence of interactions with feeding organisms. Interactions between bait and feeding organisms were evident on 9742 gangions that were retrieved with partial or no bait remaining on hooks. In these cases, bait damage or loss was more common with northern shortfin squid (/i=5033) than with Atlantic mackerel («=4709) (x^=10.71, P<0.01). There was no significant difference in the distributions of hook soak times when hooks were retrieved with whole bait present (Kol- mogorov-Smirnov statistic=0.97, P=0.31) (Fig. 2). All captured organisms were fish species with the exception of 1 loggerhead sea turtle {Caretta caretta), captured with bait of northern shortfin squid, and 1 giant isopod, Bathynomus giganteus, captured with At- lantic mackerel. The total catch and size range of each species analyzed are presented in Table 1. There were no significant differences in species-specific length dis- tributions for individuals caught by bait type (P-values for all species were >0.05) (Table 1), and length distri- bution data were non-normal or heteroscedastic, with the exception of data for the red drum. The mean size of red drum captured on hooks baited with Atlantic mackerel (mean: 859.0 mm TL [SD 32.1]) was signifi- cantly smaller than the mean size of individuals cap- tured with northern shortfin squid (mean: 903.7 mm TL [SD 40.0]) it= -2.31, P=0.03). There was a significant difference in the median length of tiger sharks {Galeocerdo cuvier) captured on each bait type, with the median length of tiger sharks being smaller on hooks baited with northern shortfin squid (1337 mm FL) than on hooks baited with Atlantic mackerel (1922 mm FL) (W=52.5, P=0.05). There were no statistically significant species-specific differences in the median length of any other species between the 2 bait types at a=0.05 (Table 1). However, at an a level of 0.10, there were significant differences in the median length of blacknose sharks (1^=238.0, P=0.06) and scal- loped hammerheads {Sphyrna lewini) (W=78.0, P=0.09) captured on the 2 bait types. For blacknose sharks, me- dian length at capture was smaller when these sharks were caught with northern shortfin squid (1073 mm FL versus 1120 mm FL), whereas for scalloped hammer- heads the median length at capture was smaller when these sharks were caught with Atlantic mackerel (1443 mm FL versus 1812 mm FL). The results of chi-square tests indicate significant differences in the expected and observed catches for 5 shark and 1 teleost species (Table 1). All of the shark species for which a statistically significant preference 54 Fishery Bulletin 1 15(1) Figure 2 Comparison by bait type of the number of hooks retrieved with whole bait present over the range of hook soak times used in this study for which predatory fish species were caught on bot- tom longline gear in the northern Gulf of Mexico during March and April 2015. Gray bars represent the bait that was Atlantic mackerel (Scomber scombrus) and white bars represent the bait that was northern shortfin squid (Illex illecebrosus). for a single bait type was observed were cap- tured more frequently with Atlantic mackerel than with northern shortfin squid. For exam- ple, sandbar sharks (Carcharhinus plumbeus) were caught 6.4 times more frequently with Atlantic mackerel than with northern shortfin squid. Furthermore, for those shark species that showed no statistically significant prefer- ence for a specific bait type, all were captured more frequently on hooks baited with Atlantic mackerel, and chi-square test P-values were <0.10, with the exception of the value for the scalloped hammerhead (F-0.66). Unlike other elasmobranchs, clearnose skates (Baja eglanteria) were caught more fre- quently on hooks baited with northern short- fin squid; however, there was no statistically significant difference in the observed and ex- pected catch of this species between the 2 bait types (x^=2.78, P=0.09). Among teleosts, red drum was the only species for which a signifi- cant bait preference (x^=8.65, P<0.01) was ob- served, and this species was caught 4.2 times more frequently with northern shortfin squid than with Atlantic mackerel. Other commer- cially and recreationally important teleosts that were captured include red snapper (Lut- janus campechanus) (x^=0.99, P=0.32) and tile- fish (Lopholatilus chamaeleonticeps) (x^ =0.04, P=0.85), neither of which had an observed significant bait preference. Although a significant bait preference was not found for yellowedge grouper (Hyporthodus flavolimbatus) (x^=2.45, P=0.12), 70% of all individu- als of this species were captured on hooks baited with Atlantic mackerel. The relationship between IBP and total catch was examined for Atlantic sharpnose shark, blacknose shark, sandbar shark, and red snapper. For all 3 shark species, a preference for hooks baited with Atlantic mackerel was observed at low rates of total catch; however, as catch rates increased, IBP values trended toward 0 (Fig. 3). This trend was most evident for sand- bar sharks, which were captured exclusively on hooks baited with Atlantic mackerel when the total catch con- sisted of 12 of fewer fish (Fig. 3). A preference for hooks baited with Atlantic mackerel also was observed for red snapper at low rates of total catch; however, unlike what was observed for sharks, red snapper were cap- tured regularly on hooks baited with northern shortfin squid regardless of total catch size (Fig. 3). Further- more, when more than 20 individuals were captured on a longline set, red snapper were captured exclusively on hooks baited with northern shortfin squid (Fig. 3). Discussion The results of our study indicate that when given the choice between hooks baited with northern shortfin squid or with Atlantic mackerel, the degree of bait pref- erence varies among fish species and that, when pres- ent, bait preference generally declines with increasing rates of total catch on longline gear. Specifically, shark species commonly captured as bycatch on longline gear in the northern Gulf of Mexico preferentially selected Atlantic mackerel over northern shortfin squid, par- ticularly when total catch rates were low. Although a statistically significant preference for hooks baited with Atlantic mackerel was documented only for the Atlantic sharpnose shark, blacknose shark, sandbar shark, smooth dogfish (Mustelus canis), and little gulp- er shark (Centrophorous uyato), the same trend was evident for all elasmobranch species captured, with the exceptions of the scalloped hammerhead and clearnose skate. There was no difference between the 2 bait types in catch rates, and therefore no difference in bait pref- erence, for economically important teleosts, with the exception of the red drum. Together, these results pro- vide support for an easily applied measure to reduce rates of shark bycatch and cause no effect on catch rates of target species or need for gear modifications. Several previous studies reported declines in catch rates of blue sharks that were associated with specific bait types (Watson et ah, 2005; Gilman et ah, 2007; Foster et ah, 2012) and found that bycatch of blue sharks was reduced when fish rather than squid were used as bait. Although our findings superficially seem in opposition to those of Watson et al. (2005) and Gil- man et al. (2007), differences among the studies can be explained by the diets of the species examined and, therefore, support the use of specific bait types to re- Driggers et al.: Influence of bait type on catch rates of predatory fish on longline gear 55 A B Total number of individuals captures (all species) Figure 3 Comparison of relationships between index of bait preference and total capture number of all species on individual long- line sets deployed in the northern Gulf of Mexico during March and April 2015 for (A) Atlantic sharpnose sharks {Rhi- zoprionodon terraenovae), (B) blacknose sharks {Carcharhinus acronotus), (C) sandbar sharks (C. plumbeus), and (D) red snapper (Lutjanus campechanus). The black lines represent locally weighted regression lines (determined through application of locally weighted scatterplot smoothing). duce bycatch rates of sharks. Watson et al. (2005) and Gilman et al. (2007) examined bycatch of blue sharks in pelagic longline fisheries and found that catch rates of this species declined when mackerel were used as bait. Similarly, Foster et al. (2012) found the use of Atlantic mackerel as bait to decrease bycatch of blue sharks; however, they also indicated that hooks baited with Atlantic mackerel were more efficient at capturing porbeagles (Lamna nasus) and shortfin makos (Isurus oxyrinchus). Cortes (1999) presented an exhaustive literature re- view of the diets of 149 shark species and found that the diet of blue sharks was dominated by cephalopods (49.4%). By comparison, teleost fish species composed 55-98.2% of the diets of Atlantic sharpnose sharks, blacknose sharks, porbeagles, sandbar sharks, scalloped hammerheads, and shortfin makos (Cortes, 1999). The 2 shark species that we examined and that do not have a primarily piscivorous diet, according to Cortes (1999), were the tiger shark and the smooth dogfish. The diet of tiger sharks was primarily composed of teleost fish species (35.4%), and sea turtles (23.8%), whereas crus- taceans (64.3%) and teleost fish species (16.6%) domi- nated the diet of smooth dogfish. Cephalopods were re- ported by Cortes (1999) to constitute 0.0-15.5% of the diets of the shark species we captured. The results of our study, combined with those of Watson et al. (2005), Gilman et al. (2007), and Foster et al. (2012), indicate that, despite being frequently characterized as opportu- nistic (e.g., Strasburg, 1958; Lowe et al., 1996), sharks, when presented with a choice between a bait that is a common prey item and one that is not a significant dietary component, will actively select the former. Although not statistically significant, the catch rate of clearnose skates was 2.7 times higher on hooks with northern shortfin squid than on hooks with At- lantic mackerel, indicating a strong preference for the former. This result was similar to the findings of Ariz et al. (2006) and Coelho et al. (2012), who found that batoids, more specifically ray species, captured on long- 56 Fishery Bulletin 115(1) line gear are caught at a higher rate on hooks baited with northern shortfin squid than on hooks baited with Atlantic mackerel. Ebert and Bizzarre (2007) showed that the clearnose skate has a diverse diet, with ap- proximately 21% fish and less than 1% squid species as prey items reported in stomach contents. However, Schwartz (1996) found that the fish component of the diet of clearnose skates was composed of small-bodied fish species, such as the striped anchovy (Anchoa hep- setus), Atlantic croaker (Micropogonias undulatus), spot {Leiostomus xanthurus), and blackcheek tongue- fish {Symphurus plagiusa). This finding indicates that a limited gape size resulted in more clearnose skates being captured on hooks with northern shortfin squid because that bait type is more malleable and, there- fore, more easily manipulated than Atlantic mackerel. Among teleosts, no consistent trend in preference for one bait type over another was found. For example, red snapper and tilefish showed no significant preference for either bait type; however, red drum had a signifi- cant preference for northern shortfin squid. Conversely, although not statistically significant, there was an ob- vious trend with yellowedge grouper toward a prefer- ence for Atlantic mackerel. The lack of bait preference shown by red snapper and tilefish was expected because both species are widely reported to be omnivorous and opportunistic (e.g., Steimle et al., 1999; Callaway et al., 2009; Moser et al.^). Yellowedge grouper have been reported to feed primarily on brachyuran crab and te- leost fish species (Heemstra and Randall, 1993); there- fore, the trend toward a preference for hooks baited with Atlantic mackerel was not unexpected. In contrast, the preference for bait of northern shortfin squid exhibited by red drum was not clearly related to the known diet of this species. The results of a number of studies indicate that red drum forage on a diverse group of prey, including invertebrate and fish species and that their prey varies depending on life stage, habitat, and season (e.g., Overstreet and Heard, 1978; Scharf and Schlicht, 2000). For example, Booth- bly and Avualt (1971) examined the stomach contents of red drum sampled within a coastal marsh system along the coast of Louisiana and determined that crus- taceans dominated the diet from late spring through fall but fish species became more important during colder months. As pointed out by Overstreet and Heard (1971) and Scharf and Schlicht (2000), dietary shifts likely were related to seasonally mediated changes in abundance of prey species. However, Matlock (1987) re- ported that, in general, adult red drum consume more individuals that are fish species than individuals that are invertebrate species. Therefore, the preference that red drum showed for the northern shortfin squid in 2 Moser, J. G., Jr., A. G. Pollack, G. W. Ingram Jr., C. T. Gled- hill, T. A. Henwood, and W. B. Ilriggers III. 2012. Develop- ing a survey methodology for sampling red snapper, Lutja- nus campechanus, at oil and gas platforms in the northern Gulf of Mexico. Southeast Data, Assessment, and Review, SEDAR31-DW26, 23 p. [Available from website.] our study cannot be explained on the basis of dietary composition reported in the literature because Atlan- tic mackerel would have been the expected preferred bait type. Although the most likely explanation is gape limitation, another possible explanation is that red drum were outcompeted for hooks baited with Atlantic mackerel and opportunistically fed on northern short- fin squid. However, it does not appear that red drum were outcompeted because all sets where red drum were captured had relatively low catch rates of other fish species. There was a clear decrease in bait preference with increasing total catch rates for Atlantic sharpnose, blacknose, and sandbar sharks. This trend was most obvious for sandbar sharks: on sets with less than 13 other captured individuals, regardless of species, 100% of sandbar sharks were caught on hooks baited with Atlantic mackerel. At catch rates of 13 or more indi- viduals, regardless of species, on a set, the likelihood that sandbar sharks would be captured on hooks baited with northern shortfin squid increased. This decrease in bait preference indicates that Atlantic sharpnose, blacknose, and sandbar sharks become more opportu- nistic when in the presence of competitors or that few- er hooks baited with Atlantic mackerel were available later in the set (as a result of depredation of preferred bait types or a captured fish occupying a hook) and late arriving individuals did not have an equal number of both bait types from which to choose. As total catch rates increased, the trend of Atlan- tic sharpnose, blacknose, and sandbar sharks moved toward no preference for a particular bait type (i.e., IBP approached 0) and the trend of red snapper moved toward a preference for northern shortfin squid (i.e., IBP<0). This result indicates that sharks were still actively selecting hooks baited with Atlantic mackerel despite more hooks baited with northern shortfin squid being available. Conversely, the trend in bait prefer- ence of red snapper at increasing total catch rates in- dicates that they were feeding opportunistically on the most abundant bait type available. This latter point can be more clearly demonstrated by comparing the IBP values of the Atlantic sharpnose shark with those of the red snapper. On longline sets with greater than 25 individuals captured, IBP values were exclusively greater than 0 for the Atlantic sharpnose shark and less than 0 for red snapper. Another possible explana- tion for the shift toward a reduced preference for At- lantic mackerel at high catch rates could be differences in retention rates of the 2 bait types as hook soak time increased. However, there was no significant difference in the distribution of soak times for hooks retrieved with whole bait present for each bait type. Beyond species-specific dietary preferences, it is pos- sible that differences in foraging strategies among the species we examined, at least in part, led to differences in catch rates of fish species on the 2 bait types. All shark species that we examined are active, roaming predators. Speed et al. (2010) estimated that the home range (excluding seasonal migrations) of small-bodied Driggers et al.: Influence of bait type on catch rates of predatory fish on longiine gear 57 coastal shark species that are not reef obligate or con- strained within a bounded area (e.g., bay or estuary) can be up to 100 km^. For example, through acoustic monitoring, Heupel et al. (2006) determined the mean home range of the bonnethead {Sphyrna tiburo) was 8.31 km^; however, some individuals used areas of up to approximately 74 km^. Larger-bodied species, such as the tiger shark, have been documented to have vast home ranges, on the order of 1000s of square kilome- ters (Heithaus et al., 2007). Conversely, most of the teleosts that we captured, with the exception of the red drum, are relatively sed- entary and have a high degree of site fidelity in off- shore waters to discrete structures, such as lumps and depressions (e.g., Able et al., 1993; Gallaway et al., 2009). For example, Jones et al. (1989) documented tilefish and yellowedge grouper in the western Gulf of Mexico occupying discrete burrows with openings rang- ing from 0.25 to 8 m wide. Able et al. (1982) hypothe- sized that these structures serve as refuge from preda- tors, and Jones et al. (1989) suggested that individuals have long-term fidelity to specific burrows. Although it is unknown how far these individuals move from their burrows to forage, the aforementioned studies all indi- cate that, in contrast with coastal shark species, most teleost species that we examined remain in relatively close proximity to a specific location. Therefore, it is possible that the teleosts that exhibited this behavior in our study were attracted to bait on the basis of prox- imity rather than preference. In contrast, because shark species rely, in part, on chemotaxis to locate prey from a distance (e.g., Shel- don, 1911; Lpkkeborg et al., 2014), it is possible that the area of bait influence was greater for Atlantic mackerel than for northern shortfin squid and led to sharks homing in more frequently on hooks baited with Atlantic mackerel. Chemotaxis, however, is unlikely to have affected bait preference given the relatively close gangion spacing (~ 18 m apart), diffusion and mixing of scent plumes from individual baits with increasing distance from the gear, and the setting of gear parallel to the axis of a current that resulted in a single plume of odorants from both Atlantic mackerel and northern shortfin squid. Therefore, given that these 2 species have significantly different relative concentrations of low-molecular-weight metabolites that are known at- tractants or stimulants of feeding behavior in fish spe- cies (Carr et al., 1996), it is more likely that higher catch rates of most shark species on hooks baited with Atlantic mackerel was a result of preference and not the area of bait influence. A potential criticism of our study is that we used alternating bait types on each longiine set rather than making comparisons on the basis of single-bait sets. Although numerous bait preference studies have used an alternating bait design similar to the one in our study (e.g., Broadhurst and Hazin, 2001; Woll et al., 2001; Yokota et al., 2009), it was suggested by Watson et al. (2005) and Foster et al. (2012) that the use of alternating bait types can bias results because of a po- tential interaction effect of bait types. Both Watson et al. (2005) and Foster et al. (2012) examined effects of bait and hook type on catch rates of epipelagic organ- isms caught on pelagic longiine gear and therefore they focused on highly mobile species, such as sea turtles, sharks, swordfish, and tunas, that occupy a single habi- tat (epipelagic zone). Conversely, we used bottom longiine gear in a highly dynamic area in terms of prey density, foraging behav- iors of target species, currents, depth, dissolved oxy- gen, salinity, temperature, turbidity, substrate types, and patchy habitats. Therefore, because many of these variables affect feeding behavior and the ability of fish species to locate baited hooks (Lpkkeborg et al., 2014), we believe the use of alternating bait types is justi- fied and best suited to answer the questions we were addressing. Had we used a single-bait approach, we would have needed to account for each of the biotic and abiotic variables for individual longiine sets. However, although we acknowledge that the use of alternating bait types could have introduced a potential bait inter- action effect, we did expose an equal number of both bait types to all conditions encountered, thereby limit- ing the number of potentially biasing factors to one. Additionally, several of the species we encountered are infrequently captured and occur in large aggrega- tions. For example, of the 79 little gulper sharks col- lected, 62% were caught on 2 sets. Further, although little gulper sharks are infrequently captured in shelf waters of the northern Gulf of Mexico, when present, they are found in large schools (senior author, personal observ.). Had single-bait sets been used and had a set occurred in proximity to a school of little gulper sharks, the resulting data would be indicative of the high den- sity of little gulper sharks in the area and not of a preference for a specific bait type. Ward et al. (2004) examined the theoretical effect of hook soak time on pelagic longiine catches in the Pacific Ocean and determined that catch rates can be affected by, among other issues, baits falling off during deployment of longiine sets, by deterioration that re- sults in baits falling off hooks, or by a time-related re- duction in the degree of attraction. Godin et al. (2012), in a metadata analysis, stated that the catch of sharks, in general, was reduced when Atlantic mackerel were used as bait. They went on to reason that squid, when compared to mackerel, was a more effective bait be- cause it remains on hooks longer, does not deteriorate as rapidly, and does not lose its attraction properties over time. Although we were not able to quantify the attrac- tant properties of the 2 bait types, our results are in direct opposition to those of Godin et al. (2012) in that we found both bait types were retained on hooks equal- ly and that bait damage or loss was greater for hooks baited with northern shortfin squid. The disparity in the 2 studies is likely attributable to the data sources of Godin et al. (2012) primarily reporting catch of blue sharks and to differences in hook soak times among studies. For example, Godin et al. (2012) included fish- 58 Fishery Bulletin 115(1) eries-dependent data sources, and commercial longline gear is generally set for much longer periods (e.g., >8 h) than the 1-4 h we allowed gear to soak. Future work should examine the “endurance” of the 2 bait types be- yond the maximum hook soak time of 4 h in our study as well as the effect of hook soak time on catch rates and bait performance. Our results, as well as those of the aforementioned studies, indicate that species-specific catch rates can be impacted significantly by the type of bait used. There- fore, an understanding of species-specific bait preferenc- es of target and bycatch species is imperative, a point also highlighted by Coelho et al. (2012). Furthermore, because ontogenetic (e.g., Wells et ah, 2008) and sea- sonal shifts (e.g., Boothby and Avault, 1971) in dietary preferences are well established among fish species, it will be necessary to identify preferences across all life stages and seasons so that no one stage is adversely affected or seasonally vulnerable. Future research will be needed to determine whether the use of a single bait type (i.e., northern shortfin squid) effectively will reduce the catch rates of shark species in the Gulf of Mexico snapper and grouper bottom longline fishery or whether sharks will opportunistically feed on a single bait type at the same rate as a preferred bait type in the absence of choice. Acknowledgments We thank the crew of the NOAA Ship Oregon II, S. Garner, M. Hendon, N. Hopkins, J. Lewis, J. McKinney, J. Moser, B. Noble, B. Prohaska, K. Rademacher, and T. Wallace for assistance in the collection of samples. Literature cited Able, K. W., C. B. Grimes, R. A. Cooper, and J. R. Uzmann. 1982. Burrow construction and behavior of the tilefish, Lopholatilus chamaeleonticeps, in Hudson Submarine Canyon. Environ. Biol. Fish. 7:199-205. Able, K. W., C. B. Grimes, R. S. Jones, and D. C. Twichell. 1993. Temporal and spatial variation in habitat charac- teristics of tilefish {Lopholatilus chamaeleonticeps) off the east coast of Florida. Bull. Mar. Sci. 53:1013-1026. Ariz, J., A. Delgado de Molina, M. L. Ramos, and J. C. Santana. 2006. Check list and catch rate data by hook type and bait for bycatch species caught by Spanish experimen- tal longline cruises in the south-western Indian Ocean during 2005. Indian Ocean Tuna Commission, lOTC- 2006-WPBy-04, 10 p. Beerkircher, L. R., E. Cortes, and M. Shivji. 2002. Characteristics of shark bycatch observed on pe- lagic longline off the southeastern United States, 1992- 2000. Mar. Fish. Rev. 64(4):40-49. Boothby, R. N, and J. W. Avault Jr. 1971. Food habits, length-weight relationship, and condi- tion factor of the red drum (Sciaenops ocellata) in south- eastern Louisiana. Trans. Am. Fish. Soc. 100:290-295. Broadhurst, M. K., and F. H. V. Hazin. 2001. Influence of type and orientation of bait on catches of swordfish {Xiphias gladius) and other species in an artisanal sub-surface longline fishery off northeastern 1 Brazil. Fish. Res. 53:169-179. j Carr, W. E. S., J. C. Netherton III, R. A. Gleeson, and C. D. Derby. ■ 1996. Stimulants of feeding behavior in fish: analyses of tissues of diverse marine organsims. Biol. Bull. I 190:149-160. Carruthers, E. H., J. D. Neilson, and S. C. Smith. 2011. Overlooked bycatch mitigation opportunities in pe- lagic longline fisheries: soak time and temperature ef- fects on swordfish {Xiphias gladius) and blue shark {Pri- onace glauca) catch.. Fish. Res. 108:112-120. Coelho, R., M. N. Santos, and S. Amorim. 2012. Effects of hook and bait on targeted and bycatch ■ fishes in an equatorial Atlantic pelagic longline fish- ; ery. Bull. Mar. Sci. 88:449-467. i Cortes, E. 1999. Standardized diet compositions and trophic levels of sharks. ICES J. Mar. Sci. 56:707-717. Ebert, D. A., and J. J. Bizzarre. ' 2007. Standardized diet compositions and trophic levels of i skates (Chondrichthyes: Rajiformes: Rajoidei). Environ. : Biol. Fish. 80:221-237. Foster, D. G., S. R Epperly, A. K. Shah, and J. W. Watson. ! 2012. Evaluation of hook and bait type on the catch rates in the western North Atlantic Ocean pelagic longline i fishery. Bull. Mar. Sci. 88:529-545. ' Francis, M. R, L. H. Griggs, and S. J. Baird. j, 2001. Pelagic shark bycatch in the New Zealand tuna longline fishery. Mar. Freshw. Res. 52:165-178. Gallaway, B. J., S. T. Szedlmayer, and W. J. Gazey. | 2009. A life history review for red snapper in the Gulf of [ Mexico with an evaluation of the importance of offshore I petroleum platforms and other artificial reefs. Rev. |. Fish. Sci. 17:48-67. Gilman, E., D. Kobayashi, T. Swenarton, N. Brothers, P. Dal- zell, and I. Kinan-Kelly. 2007. Reducing sea turtle interactions in the Hawaii- based longline swordfish fishery. Biol. Conserv. 139:19- 28. !■ Gilman, E., S. Clarke, N. Brothers, J. Alfaro-Shigueto, J. Man- delman, J. Mangel, S. Petersen, S. Piovano, N. Thomson, P. Dalzell, et al. : 2008. Shark interactions in pelagic longline fisher- ies. Mar. Policy 32:1-18. !■ Godin, A. C., J. K. Carlson, and V. Burgener. 2012. The effect of circle hooks on shark catchability and at-vessel mortality rates in longline fisheries. Bull. Mar. Sci. 88:469-483. Gulak, S. J. B., M. P Enzenauer, and J. K. Carlson. 2013. Characterization of shark and reef fish bottom long- iine fisheries: 2012. NOAA Tech. Memo. NMFS-SEFSC- 652, 42 p. Heemstra, R C., and J. E. Randall. 1993. FAO species catalogue, vol. 16. Grouper of the world (family Serranidae, subfamily Epinephelinae): an an- notated and illustrated catalogue of the grouper, rock- cod, hind, coral grouper and lyretail species known to date. FAO Fisheries Synopsis 125, 382 p. FAO, Rome. Heithaus, M. R., A. J. Wirsing, L. M. Dill, and L. I. Heithaus. 2007. Long-term movements of tiger sharks satellite- Driggers et al.: Influence of bait type on catch rates of predatory fish on longline gear 59 tagged in Shark Bay, Western Australia. Mar. Biol. 151:1455-1461. Heupel, M. R., C. A. Simpfendorfer, A. B. Collins, and J. P Tyminski. 2006. Residency and movement patterns of bonnethead sharks, Sphyrna tiburo, in a large Florida estuary. En- viron. Biol. Fish. 76:47—67. Jones, R. S., E. J. Gutherz, W. R. Nelson, and G. C. Matlock. 1989. Burrow utilization by yellowedge grouper, Epineph- elus flavolimbatus, in the northwestern Gulf of Mexi- co. Environ. Biol. Fish. 26:277-284. Lpkkeborg, S., S. I. Siikavuopio, O. B. Humborstad, A. C. Utne- Palm, and K. Ferter. 2014. Towards more efficient longline fisheries: fish feed- ing behaviour, bait characteristics and development of alternative baits. Rev. Fish Biol. Fish. 24:985-1003. Lowe, C. G., B. M. Wetherbee, G. L. Crow, and A. L. Tester. 1996. Ontogenetic dietary shifts and feeding behavior of the tiger shark, Galeocerdo cuvier, in Hawaiian wa- ters. Environ. Biol. Fish. 47:203-211. Matlock, G. C. 1987. The life history of the red drum. In Manual of red drum aquaculture (G. W. Chamberlain, R. J. Miget, and M. G. Haby, eds.). p. 1-47. Texas Agricultural Exten- sion Service and Sea Grant College Program, Texas A&M University, College Station, TX. Overstreet, R. M., and R. W. Heard. 1978. Food of the red drum, Sciaenops ocellatus, from the Mississippi Sound. Gulf Res. Rep. 6:131-135. Prytherch, H. F. 1983. A descriptive survey of the bottom longline fishery in the Gulf of Mexico. NOAA Tech. Memo. NMFS-SE- FC-122, 33 p. Rey, J. C., and R. Munoz-Chapuli. 1991. Relation between hook depth and fishing efficiency in surface longline gear. Fish. Bull. 89:729-732. Robbins, W. D., V. M. Peddemors, and S. J. Kennelly. 2011. Assessment of permanent magnets and electroposi- tive metals to reduce the line-based capture of Gala- pagos sharks, Carcharhinus galapagensis. Fish. Res. 109:100-106. Rogers, A. 1920. Industrial uses for the shark and porpoise. J. Am. Leather Chem. Assoc. 15:241-244. Scharf, F. S., and K. K. Schlicht. 2000. Feeding habits of the red drum (Sciaenops ocella- tus) in Galveston Bay, Texas: seasonal diet variation and predator-prey size relationships. Estuaries 23:128-139. Schwartz, F. J. 1996. Biology of the clearnose skate, Raja eglanteria, from North Carolina. Fla. Sci. 59:82-95. Scott-Denton, E., P. F. Cryer, J. P. Gocke, M. R. Harrelson, D. L. Kinsella, J. R. Pulver, R. C. Smith, and J. Williams. 2011. Descriptions of the U.S. Gulf of Mexico reef fish bot- tom longline and vertical line fisheries based on observer data. Mar. Fish. Rev. 73(2): 1-26. Sheldon, R. E. 1911. The sense of smell in selachians. J. Exp. Zool. 10:51-62. Speed, C. W., I. C. Field, M. G. Meekan, and C. J. A. Bradshaw. 2010. Complexities of coastal shark movements and their implications for management. Mar. Ecol. Prog. Ser. 408:275-293. Strasburg, D. W. 1958. Distribution, abundance, and habits of pelag- ic sharks in the central Pacific Ocean. Fish. Bull. 58:335-361. Steimle, F. W., C. A. Zetlin, P. L. Berrien, D. L. Johnson, and S. Chang. 1999. Tilefish, Lopholatilus chamaeleonticeps, life history and habitat characteristics. NOAA Tech. Memo. NMFS- NE-152, 30 p. Walsh, W. A., K. A. Bigelow, and K. L. Sender. 2009. Decreases in shark catches and mortality in the Hawaii-based longline fishery as documented by fishery observers. Mar. Coast. Fish. 1:270-282. Ward, R, R. A. Myers, and W. Blanchard. 2004. Fish lost at sea: the effect of soak time on pelagic longline catches. Fish. Bull. 102:179-195. Ward, P, E. Lawrence, R. Darbyshire, and S. Hindmarsh. 2008. Large-scale experiment shows that nylon leaders reduce shark bycatch and benefit pelagic longline fish- ers. Fish. Res. 90:100-108. Watson, J. W., S. P. Epperly, A. K. Shah, and D. G. Foster. 2005. Fishing methods to reduce sea turtle mortality as- sociated with pelagic longlines. Can. J. Fish. Aquat. Sci. 62:965-981. Watson, J. T., T. E. Essington, C. E. Lennert-Cody, and M. A. Hall. 2009. Trade-offs in the design of fishery closures: manage- ment of silky shark bycatch in the eastern Pacific Ocean tuna fishery. Conserv. Biol. 23:626-635. Woll, A. K., J. Boje, R. Holst, and A. C. Gundersen. 2001. Catch rates and bait selectivity in longline fishery for Greenland halibut (Reinhardtius hippoglossoides, Walbaum) at East Greenland. Fish. Res. 51:237-246. Wells, R. J. D., J. H. Cowan Jr., and B. Fry. 2008. Feeding ecology of red snapper Lutjanus campecha- nus in the northern Gulf of Mexico. Mar. Ecol. Prog. Ser. 361:213-225. Yokota, K., M. Kiyota, and H. Okamura. 2009. Effect of bait species and color on sea turtle bycatch and fish catch in a pelagic longline fishery. Fish. Res. 97:53-58. 60 National Marine Fisheries Service NOAA Fishery Bulletin ^ established in 1881 •<#. Spencer F. Baird First U.S. Commissioner of Fisheries and founder of Fishery Bulletin Reproductiwe strategy of white anglerfish iLophim piscatorim} in Mediterranean waters; implications for management Email address for contact author: colmenero@icm.csic.es Institut de Ciencies del Mar (ICM) Consejo Superior de Investigaciones Cienfficas (CSIC) Passeig Maritim de la Barceloneta 37-49 08003 Barcelona, Spain Abstract— Reproductive parameters of the white anglerfish (Lophius pis- catorius) in the northwestern Medi- terranean Sea were studied in 556 specimens collected monthly aboard commercial fishing vessels that were trawling at depths of 12-836 m. The main spawning season occurred from February through June. The size at maturity was estimated to be 48.8 cm in total length (TL) for males, 59.9 cm TL for females, and 51.3 cm TL for both sexes combined. The white anglerfish has group-synchro- nous oocyte development and deter- minate fecundity. It is a total spawn- er (spawns all its eggs once during a spawning season) and has a batch fecundity ranging from 661,647 to 885,214 oocytes, a relative batch fe- cundity of 66-128 oocytes per gram of female gutted weight, and a po- tential fecundity with values from 54,717 to 104,506 oocytes per kilo- gram of female total weight. This study is the first to provide the re- productive biology of white angler- fish in the northwestern Mediterra- nean Sea and provide valuable infor- mation that can be used to improve the stock assessment and ensure proper management of this species. Manuscript submitted 2 November 2015. Manuscript accepted 21 October 2016. Fish. Bull. 115:60-73 (2017). Online publication date; 15 November 2016. doi: 10.7755/FB.115.1.6. 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. Ana I. Colmenero fcontact author) Victor M. Tuset Pilar Sanches The genus Lophius, commonly known as anglerfish, monkfish, or goosefish, belong to a family of bathydemersal fishes, which live and feed on the bottom of the seafloor generally be- low 200 m (Caruso, 1986). It includes 7 species distributed around the world. The white anglerfish (Lophius piscatorius) is found in the northeast Atlantic Ocean and the Mediterra- nean Sea, and the black anglerfish (Lophius budegassa) coexists with white anglerfish over most of its range, although the black anglerfish has a more southerly distribution in the Atlantic (Caruso, 1986). The shortspine African angler (Lophius vaillanti) is found in the eastern At- lantic (Maartens and Booth, 2005). The devil anglerfish (Lophius vome- rinus) occupies the southeast Atlan- tic and the northern and western In- dian Ocean (Walmsiey et ah, 2005). The blackfin goosefish (Lophius gastrophysus) inhabits the western Atlantic, and the goosefish (Lophius americanus) occurs in the northwest Atlantic (Caruso, 1983). Finally, the yellow goosefish (Lophius litulon) is distributed in the northwest Pacific, in the Gulf of Po-Hai, in the Yel- low Sea, and in the East China Sea (Yoneda et al., 1997). We focused on white anglerfish, which can be found on the conti- nental shelf and slope, inhabiting depths from the shoreline to >1000 m (Afonso-Dias and Hislop, 1996). In the Mediterranean Sea, this species cohabits with black anglerfish, and even though the distributions of both species overlap, no ecological compe- tition exists between them because of a temporal segregation in their daily biorhythms (Colmenero et al., 2010). Both of these species of Lophius play an important role in the trophic structure of benthodemersal ecosys- tems because they represent major predators, along with the European hake (Merluccius merluccius) (Diaz et al., 2008). In the community struc- ture of the northwestern Mediterra- nean Sea, species of anglerfish are considered top predators (Coll et al., 2006; Vails et al., 2014). They are also reported to be important in the deep- sea community (depths from 200 m to the bottom of the ocean) because they are the most abundant species (Labropoulou and Papaconstantinou, 2000; Maiorano et al., 2010). Despite the fact that the deep sea is the largest ecosystem on the planet, is highly diverse, and has a wealth of resources, it is still mostly unknown and poorly understood in comparison with shallov/-water ar- Colmenero et al.; Reproductive strategy of Lophius piscatorius in Mediterranean waters 61 eas: therefore environmental management in deep wa- ters is difficult (Ramirez-Llodra et al., 2010). In the last few decades, the decline of traditional fisheries on the continental shelf, the increasing demand for food sourc- es, and rapid technological developments have resulted in an increasing exploitation of deep-sea resources (Ko- slow et al., 2000; Ramirez-Llodra et al., 2011) and in an incremental increase in the global mean depth of fishing (Watson and Morato, 2013). This rise in deep-sea fishing has affected catches of Lophius species, given the growing demand for hu- man consumption of this group of fish that is leading to an increase in worldwide commercial exploitation and targeting of anglerfishes (Farina et al., 2008). To- tal catch reported globally for white anglerfish reached more than 26,500 metric tons (t) in 2014 (FAO Global Capture Production database, website) and total catch of anglerfishes in the northwestern Mediterranean Sea for the same year added up to 660 t (Tudo Vila^). Landings in our study area were composed primarily of black anglerfish (86%) and generally only a small percentage of white anglerfish (14%) (Tudo Vila^), but, for landings in Atlantic waters, the opposite is true; white anglerfish (94%) dominate the catch (Dobby et ah, 2008). Although the European Commission previ- ously has conducted stock assessments of black ang- lerfish in the western Mediterranean Sea, there is no corresponding assessment for white anglerfish. The lack of information about the structure of the popula- tion of white anglerfish in this region and the lack of knowledge of the basic biology of this species are the main reasons for the absence of any assessment. The actual management regulations applied for black ang- lerfish generally are those applied to bottom trawling (European Union Council Regulation 1967/2006), with recommendations aimed at reducing the fishing effort of the fleet in order to avoid loss in stock productivity and decreases in landings (Cardinale et al.^). The small quantity of white anglerfish available from landings in Mediterranean waters makes studies of this species challenging. Studies conducted in the Mediterranean Sea have been scarce, and they have been focused on temporal and spatial distribution of this species (Ungaro et al., 2002; Colmenero et ah, 2010), age and growth (Tsimenidis and Ondrias, 1980; Tsimenidis, 1984), feeding ecology (Lopez et al., 2016), morphometries (Negzaoui-Garali and Ben Salem, 2008), parasites (Colmenero et al., 2015a), and ova character- istics (Colmenero et al., 2015b). Among these studies, only Ungaro et al. (2002) analyzed some of the biologi- cal features of this species by using data available from trawl surveys, including data on distribution, abun- ^ Tudo Vila, P. 2015. Unpubl. data. Directorate of Fishing and Maritime Affairs, Government of Catalonia, Avinguda Diagonal 523-525, 08029 Barcelona, Spain. ^ Cardinale, M., D. Damalas, and C. G. Oslo (eds.). 2015. Sci- entific, Technical and Economic Committee for Fisheries (STECF) — Mediterranean Assessments, part 2 (STECF-15- 06), 396 p. Publications Office of the European Union, Lux- embourg. [Available from website.] dance, stock demography, and size at maturity. The lat- ter work is valuable but is limited because sampling occurred only in the spring and summer; a whole year of sampling is recommended to obtain more accurate biological information. A study of reproductive ecology is important for an understanding of population dynamics, and it is critical for assessing the effects of harvesting on fish popula- tions when attempting to develop appropriate manage- ment strategies. Recruitment is recognized as a key process for maintaining sustainable populations, and the relationship between the reproductive output of the population and the resulting recruitment is central to understanding how a fish population will respond to constant stressors such as fishing (Chambers and Trip- pel, 1997). Although knowing more about the relation- ships between life history strategies and productivity with depth could help managers understand the poten- tial response of a deep-sea species to fishing (Brazen and Haedrich, 2012), it is first necessary to conduct biological studies of fish to gain knowledge of the re- productive system of a species (Koslow et al., 1995). Such studies include gonad morphology (external and cellular description of the ovary and testis), reproduc- tive pattern (hermaphroditism or gonochorism), repro- ductive behavior, reproductive cycle, spawning season duration, size at maturity, sex ratio, size at sexual transition, and fecundity. All of this information can be applied at the popu- lation level to evaluate reproductive potential and to serve as a basis for limits on fishing that aim in or- der to keep recruitment at sustainable levels (Garcfa- Diaz et ah, 2006). Because reproductive strategy varies within species, depending on the area of distribution of each species and the depth distribution of each species in each area (Rotllant et al., 2002), there is a need for knowledge about reproduction of deep-sea fish species. Such information is needed particularly in the Mediter- ranean Sea because the data available for this region are limited (Morales-Nin et al., 1996; D’Onghia et al., 2008; Munoz et ah, 2010; Bustos-Salvador et al., 2015), and, furthermore, target species of fisheries have been the focus of only a few studies (Rotllant et ah, 2002; Recasens et al., 2008). The goal of this study was to describe the reproduc- tive parameters — gonadal morphology, spawning sea- son, size at sexual maturity, oocyte development, and fecundity — of white anglerfish in the northwestern Mediterranean Sea in order to provide valuable infor- mation and scientific background to improve stock as- sessments and effective management for Lophius spe- cies in Mediterranean waters. Materials and methods Sampling and data collection Between June 2007 and December 2010, 556 white anglerfish, with total lengths (TLs) of 9-120 cm, were 62 Fishery Bulletin 115(1) Figure 1 Map of the northwestern Mediterranean Sea showing the study area where white anglerfish (Lophius piscatorius) were collected from fishing grounds off the Catalan coast between June 2007 and December 2010. collected monthly aboard commercial fishing ves- sels that were trawling at depths of 12-836 m. Fish were sampled from 467 stations located in the fish- ing grounds off the Catalan coast in the northwestern Mediterranean Sea from 40°5.980'N to 43°39.310'N and from 0°32.922'E to 3°35.718'E (Fig. 1). For each individual, TL was measured to the nearest centimeter, total weight (TW) and gutted weight (GW) were mea- sured to the nearest gram, and gonad weight (GNW) and liver weight (LW) were measured to the nearest 0.01 g. The sex of all fish was determined and assigned macroscopically to a gonadal stage on the basis of a scale of 5 maturity phases proposed by Colmenero et al. (2013): immature (phase I), developing or regener- ating (phase II), spawning capable (phase III), actively spawning (phase IV), and regressing (phase V). Fish that were too small (<20 cm TL) for their sex to be determined or for assignment to a gonadal phase were classified as indeterminate. Macroscopic gonadal stage was validated histologically, according to the most advanced cell within the gonad (West, 1990). Gonads were fixed in 10% buffered formalin solution, dehydrated in ascending solutions of alcohols and em- bedded in a methacrylate polymer resin, sectioned at a thickness of 4 pm with a manual microtome Leica Reichert-Jung 2040^ (Leica Microsystems, Wetzlar, Ger- many), stained with Lee’s stain (methylene blue and basic fuchsin), and mounted in a synthetic resin of di- butyl phthalate xylene on microscope slides. Gonads were classified according to their size and color and the presence or absence of specific inclusions (oil droplets, ^ 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. yolk, postovulatory follicles, or sperm), as well as the type of oocytes (Wallace and Selman, 1981). Reproductive biology The spawning season was estimated by analyzing the monthly variation in the percentage of maturity phas- es and the changes in gonadosomatic index (GSI) and hepatosomatic index (HSI) for mature fish of each sex (Afonso-Dias and Hislop, 1996; Colmenero et al., 2013). Because indeterminate individuals (n=27) were not considered, 251 males and 278 females were used to determine both indices, which were calculated accord- ing to Yoneda et al. (2001) as and GSI = (GNW / GW) X 100 HSI = (LW / GW) X 100. (1) (2) The lengths at which 25%, 50%, and 75% of sampled fish reached sexual maturity were estimated by fitting the proportion of sexually mature males and females (phase III, phase IV, or phase V) and for both sexes combined to the logistic equation (Colmenero et al., 2013): P = 100 / (1 + exp [a -I- bTL]), (3) where P = the percentage of mature individuals as a function of size class (measured in TL); and a and b are specific parameters that can change during the life cycle. A logarithmic transformation was applied to this equa- tion to calculate the parameters a and b by means of linear regression. Colmenero et al.: Reproductive strategy of Lophius piscatorius in Mediterranean waters 63 Oocyte development and fecundity Oocyte diameters obtained from 21 randomly selected specimens at all phases of maturity were measured to the nearest 0.01 pm with an image analysis program (Image-Pro Plus, vers. 5.0, Media Cybernetics, Inc., Rockville, MD) in combination with an Axioskop 2 Plus microscope (Carl Zeiss Microscopy, LLC, Thornwood, NY), and a ProgRes C14 digital microscope camera (Jenoptik AG, Jena, Germany). Only oocytes sectioned through the nucleus were taken into account. The de- velopmental stages of the oocytes were categorized ac- cording to the descriptions in Colmenero et al. (2013) that were adapted from Wallace and Selman (1981). The mean oocyte diameter by developmental stage was determined by calculating the diameter of all oocytes encountered in each subsample, and the range was set with the smallest and largest oocytes found at each de- velopmental oocyte stage. Fecundity was determined by using the gravimetric method described by Hunter and Goldberg (1980). Be- cause homogeneity in oocyte distribution within ovaries of white anglerfish has already been established (Afon- so-Dias and Hislop, 1996), ovarian tissue subsamples of approximately 500 mg were taken randomly from 2 specimens with ovaries in phase III that had nei- ther postovulatory follicles nor atretic oocytes present. Whole tissue subsamples were placed on several slides and covered with cover slips, then photographed with a Canon Powershot SD870 IS digital camera (Canon USA, Melville, NY). Oocytes were counted manually with Image-Pro Plus. Batch fecundity (BF), the total number of hydrat- ed oocytes produced in a single spawning event by an individual female, of each female was determined by means of this equation: BF = {oocyte number / sampled GNW) x total GNW, (4) where BF is the product of the number of secondary vitellogenic oocytes per unit of weight multiplied by the total ovarian weight (Yoneda et al., 2001). Relative batch fecundity (RBF), the total number of mature eggs released by a female during the spawning batch per gram of female GW, was calculated with the following equation (Pavlov et al., 2009): RBF = BF / GW. (5) Potential fecundity was calculated as the number of vitellogenic oocytes divided by TW in kilograms for each mature female and then averaged (Murua et al., 2003). Results Gonad morphology j The gonad of female white anglerfish has 2 ribbon-like ovarian lobes connected to each other at their posterior end. One side of the “ribbon” consists of an ovigerous membrane from which a single layer of oocyte clus- ters, which contain oocytes at different developmental stages, projects into the lumen. The other side is no- novigerous and secretes a gelatinous material during maturation that fills the ovarian lumen, where mature oocytes develop (Fig. 2). During maturation, the gonad increases in size until it fills the abdominal cavity (Fig. 3). Testes are a pair of elongated organs with a bean shape in transverse section. Spermatogenesis takes place in a capsule-like sac called a cyst, but it is com- pleted in the lumina of the lobules. The cysts appear to be arranged with a gradient of germ cells of increasing maturation from the cortex to the sperm duct (Fig. 4). Spawning season The monthly distribution of maturity phases (Fig. 5) revealed a peak in reproduction during spring, when a major portion of the spawning females and the high- est value of GSI (0.77) were found. Spawning capable females (phase III) were caught primarily between April and June, and females in the actively spawning phase (IV) were observed in November, December, and March — the latter month having the maximum occur- rence (11%). Females in immature, regressing, and de- veloping or regenerating phases (I, V, and II, respec- tively) were found year-round, although the highest percentage of immature individuals (49%) was observed in January. The GSI values followed the same pattern shown in these maturity phases: highest during spring, decreasing during summer and autumn, and increas- ing again during winter. Males in all maturity phases were observed throughout the year, but with a maxi- mum percentage of mature males (66%) in February and March. Immature males were found primarily in July (69%). The mean GSI for females increased as their ovaries developed and peaked in phase IV. For males, the mean GSI increased with testicular develop- ment and reached a maximum in phase IV (Table 1). The mean HSI for females and males increased during the summer and autumn months and decreased during winter and spring. On the basis of these observations, a main spawning season was found from February through June and a secondary one occurred in Novem- ber and December. Size at sexual maturity The maturity ogive for males indicates that the length at which 50% of them reached sexual maturity (L50) was 48.4 cm TL (Fig. 6A). Maturity in males occurred at about 37% of their maximum observed TL. The smallest mature male found was 32.5 cm TL, and the largest immature male was 50 cm TL. The maturity ogive for females indicates that L50 was 59.9 cm TL (Fig. 6B). Female maturity occurs at about 30% of their maximum observed TL. Like the smallest male, the smallest mature female was 32.5 cm TL. The largest immature female measured 56 cm TL. The maturity ogive for the sexes combined indicates an L50 of 51.3 64 Fishery Bulletin 1 15(1) Figure 2 Histological sections from ovaries of female white ang- lerfish (Lophius piscatorius) in 3 phases of maturity: (A) immature, (B) spawning capable, and (C) actively spawning. Ov=ovigerous membrane, Nov=nonovigerous membrane, n=nucleus, Pg=primary growth stage, Vt=vitellogenesis stage, Od=oil droplet, Yv=yolk vesi- cle, Mm^mucus matrix. Scale bars=100 pm. cm TL. The lengths at which 25% and 75% of fish at- tained maturity were 43.5 and 53.4 cm TL for males, 48.6 and 71.1 cm TL for females, and 44.7 and 58 cm TL for the sexes combined. Oocyte development and fecundity Oocytes in different developmental stages were found in each maturity phase. They were organized in clus- ters where a gradient in the size of the oocyte was ob- served. A group of oocytes differentiated from others as the ovaries developed, indicating that white anglerfish has group-synchronous oocyte development and can be considered to have determinate fecundity (Fig. 7). Ova- ries at each maturity phase contained primary oogonia- and perinucleolar-stage oocytes. Chromatin nucleolar were difficult to find and were present only in imma- ture phase. Females at the cortical alveolar stage were not found in our samples. Vitellogenic and hydrated oocytes were located in females capable of spawning. Oocyte diameters at each stage of oocyte development are shown in Table 2. Batch fecundity ranged from 661,647 to 885,214 oocytes from 2 females that measured 76 and 105 cm TL, 6331 and 16,178 g TW, and 5182 and 13,330 g GW, respectively. Relative batch fecundity ranged from 66 to 128 oocytes/g GW (average of 97 oocytes/g GW [standard deviation, SD 43] ). Potential fecundity values moved from 54,717 to 104,506 oocytes/kg TW with a mean of 79,612 oocytes/kg TW (SD 35,206). Discussion Relevance of reproductive traits for sustainable management Fishing activity during spawning seasons may affect population parameters, specifically composition of the size distribution, mortality rate, sexual structure of the population, size at maturity, and changes in the spawn- ing season. These parameters, in turn, can increase the risk of over-exploitation of a stock. Fishing during spawning periods may result in tar- geting a specific size class of the population and thus increasing the chance of catching the older (and larger) age classes and making the stock vulnerable to repro- ductive collapse (van Overzee and Rijnsdorp, 2015). Because spawning is generally limited to specific ar- eas and times (Cushing, 1990), the conservation of re- sources can be enhanced by limiting fishing activity in a spatiotemporal frame. Furthermore, fishing pressure has been documented to have reduced initial size at maturity — an issue that is a concern particularly for late-maturing species (Stewart et al., 2010). If size of capture is below the size at first maturity, there is a genuine risk of recruitment overfishing. Therefore, knowledge of the spawning season and the size at ma- turity can help managers establish closed seasons and prevent fishing at this vulnerable time in the life cycle Colmenero et al.: Reproductive strategy of Lophius piscatorius in Mediterranean waters 65 oOO fjm ■ , 5 cm Figure 3 Images of a female of the species Lophius showing (A) an ovary that occupies the entire abdominal cavity; (B) a close-up of the female gonad; and (C) a hexagonal chamber of the mucoid veil, which contains an unfertilized egg of white anglerfish (Lophius piscatorius) and (D) a scanning electron micrograph of yolk vesicles from the inside of the ovum of a white anglerfish. of fish species by preserving breeding individuals and establishing a legal minimum size. The results of our study of white anglerfish in the northwestern Mediterranean Sea indicate that a long spawning period occurs during mid-winter and late i spring, from February through June, although a sec- I ondary breeding period has been observed in November I and December. These results agree with those obtained I in studies that were focused on the northeastern Atlan- I tic Ocean, where this species spawns from November I through June (Fulton, 1898; Afonso-Dias and Hislop, 1 1996; Hislop et al., 2001). However, a previous study in the northwestern Mediterranean Sea identified a spawning season during spring-summer (Ungaro et al., 2002). Discrepancies between the latter study and our work may be explained by the differences in sampling periods. Nevertheless, spawning seasonality, which is associ- ated with environmental conditions and local oceano- graphic features, varies between species as well as by geographical area. An example of this variability in spawning seasonality can be observed in 2 locations along the Atlantic-Iberian coast: on the Portuguese and western Spanish coasts, spawning of the white anglerfish takes place during winter-spring (Duarte et al., 2001), whereas on the northern Spanish coast (Bay 66 Fishery Bulletin 115(1) Figure 4 Histological sections of testes from actively spawning male white anglerfish (Lophius pis- catorius) showing (A) lobular organization, (B) spermatozoa in the lumen of the seminal lobules and in the sperm duct, (C) seminal lobules during spermatogenesis, and (D) a close-up of the seminal lobules. Ta=tunica albuginea, L=seminal lobule, Bv=blood vessel, Sz=spermatozoa, Sg=spermatogonia, Sc^spermatocyte, St=spermatid, Scale bars=25 pm. of Biscay), spawning occurs during summer (Quinco- ces et al.^). In fact, spawning activity for one of its congenerics, the black anglerfish, in the northwestern Mediterranean seems to occur from November through March and a secondary spawning occurs in August and September (Colmenero et ah, 2013). Although a little overlap exists between spawning seasons of both of these Lophius species in Mediterranean waters, the main period is markedly different, and that difference lessens competition among these species. Usually, species of Lophius have long spawning peri- ods ranging between 4 and 6 months. Black anglerfish off the Spanish-Atlantic coasts spawn from November through February (Duarte et al., 2001), and in the Bay of Biscay the peak spawning period is from May through July (Quincoces et al.®). The goosefish off the ^ Quincoces, I., M. Santurtiin, and P. Lucio. 1998. Biological aspects of white anglerfish (Lophius piscatorius) in the Bay of Biscay (ICES Division Villa, b, d), in 1996-1997. ICES Council Meeting (C.M.) Documents 1998/0:48, 29 p. ^ Quincoces, I., P. Lucio, and M. Santurtiin. 1998. Biology of black anglerfish Lophius budegassa in the Bay of Biscay East Coast of the United States has its reproductive period from May through June (Armstrong et al., 1992), spawning for the blackfin goosefish off the Brazilian coasts takes place during spring and summer (Valen- tim et al., 2007), and the yellow goosefish spawns be- tween February and May in the East China Sea and the Yellow Sea (Yoneda et al., 2001). The devil ang- lerfish off the coast of South Africa has a well-defined summer breeding season (Griffiths and Hecht, 1986), and individuals of this species off the coast of Namibia spawn throughout the year with a slight increase be- tween autumn and spring (Maartens and Booth, 2005). Most deep-sea fish species reach sexual maturity at sizes larger than those of species that inhabit the con- tinental shelf reach maturity, and, in some cases, males mature at smaller sizes than females (Rotllant et al., 2002; Pajuelo et al., 2008). A similar pattern was ob- served for white anglerfish — one in which females ma- ture sexually at larger sizes (59.9 cm TL) than those recorded for males (48.4 cm TL). This pattern has also waters, during 1996-1997. ICES Council Meeting (C.M.) Documents 1998/0:47, 28 p. Colmenero et al.: Reproductive strategy of Lophius piscatorius in Mediterranean waters 67 i! Dll I3III bIV dV D •GSI -o-HSI Figure S Monthly distribution of maturity phases of gonads for (A) males (n=251) and (B) females (n=278) and monthly changes in the mean gonadosomatic (GSI) and hepatosomatic (HSI) indices for (C) males (n=135) and (D) females (n=202) of white anglerfish {Lophius piscatorius) collected from the northwestern Mediterranean Sea between June 2007 and December 2010. On the basis of macroscopic examination, specimens were assigned to the following 5 phases: immature (phase I), developing or regenerating (phase II), spawning capable (phase III), actively spawning (phase IV), and regressing (phase V). Error bars indicate standard error of the mean. been found for white anglerfish in other areas (Ofstad and Laurenson®) and for other species of anglerfish. Fe- male black anglerfish, for example, mature at 48.2 cm TL, whereas males attain first maturity at 33.4 cm TL (Colmenero et ah, 2013), and female devil anglerfish reach sexual maturity at 58.2 cm TL, whereas males of this species mature at 39.9 cm TL (Maartens and Booth, 2005). For the goosefish, L50 for females and males was estimated at 48.5 and 36.9 cm TL, respectively (Arm- strong et ah, 1992), and female yellow goosefish mature at 56.7 cm TL and males of this species mature at 36.2 cm (Yoneda et ah, 2001). This dissimilarity in size at maturity is usually associated with a trade-off between life history traits, where early maturity involves a larg- I er size but a slower growth (Stearns and Koella, 1986; I Charnov, 2008). Reproductive strategy ' The reproductive strategy of white anglerfish is one of ' discontinuous oogenesis with synchronous development ® Ofstad, L. H., and C. Laurenson. 2007. Biology of angler- fish Lophius piscatorius in Faroese waters. ICES Council Meeting (C.M.) Documents 2007/K:07, 16 p. of vitellogenic oocytes and is, therefore, this species is considered a total spawner (Afonso-Dias and Hislop, 1996). The oocytes ovulate at once, and the eggs are released in either a unique event or over a short period of time, as part of a single episode during the spawning season (Murua and Saborido-Rey, 2003; Pavlov et ah, 2009). This pattern of oocyte development and spawn- ing patterns is also found in other species of Lophius (Leslie and Grant, 1990; Armstrong et ah, 1992; Col- menero et ah, 2013). Yoneda et ah (2001) suggested that yellow goosefish may have the potential to spawn more than once a year, on the basis of the observation of a captive specimen that released several infertile egg masses. However, this spawning behavior cannot be considered normal. Female anglerfish spawn their eggs in a mucoid veil that floats near the surface. The veil consists of individual chambers that contain 1-3 eggs and has an opening that provides water circulation. In our study, we recognized in some chambers the presence of 2 eggs sharing the same chamber. Although this way of re- leasing eggs is not common among fish species, some Scorpaeniformes, such as the shortfin turkeyfish (Den- drochirus brachypterus) (Fishelson, 1978) or the short- Maturity (%) 68 Fishery Bulletin 115(1) Table 1 Gonadosomatic (GSI) and hepatosomatic (HSI) indices at each maturity phase for male and female white anglerfish {Lophius piscatorius) collected from the northwestern Mediterranean Sea between June 2007 and December 2010. SE=standard error. Sex Maturity phase GSI range Mean GSI (SE) HSI range Mean HSI (SE) n Male I 0.01-0.41 0.10 (0.01) 1.04-4.65 2.37 (0.07) 106 II 0.06-1.07 0.25 (0.03) 0.27-5.11 2.67 (0.14) 54 III 0.21-1.30 0.61 (0.05) 1.92-6.72 3.20 (0.18) 35 IV 0.30-1.70 0.70 (0.09) 0.40-5.39 3.34 (0.28) 17 V 0.19-1.11 0.50 (0.06) 2.04-5.10 3.35 (0.22) 17 Female I 0.01-0.86 0.23 (0.02) 0.92-5.33 2.37 (0.11) 66 II 0.04-1.22 0.40 (0.02) 0.42-7.79 2.87 (0.10) 133 III 0.61-1.65 1.13 (0.52) 2.15-5.83 3.99 (1.85) 2 IV 1.59-3.86 2.81 (0.52) 2.68-8.50 5.80 (1.20) 4 V 0.18-2.44 0.66 (0.06) 0.36-8.59 3.03 (0.17) 60 spine thornyhead (Sebastolobus alascanus) , (Erickson and Pikitch, 1993), also spawn buoyant gelatinous egg masses. It has been proposed that the advantages of releasing eggs in these veils facilitate their dispersal; the egg veil floats near the surface and is | subject to the actions of wind, currents, and waves. The veil also serves as protection for eggs against predation because of the pres- ence of obnoxious or toxic substances in the veils (Armstrong et al., 1992). Moreover, the | veil may help with the fertilization of eggs. When males are present and the egg ribbon is laid, the ribbon keeps the eggs together and prevents their dispersion through the water. The males then eject milt near the veil to guarantee fertilization of all the eggs (Dahlgren, 1928). Armstrong et al. (1992) ; suggested that sperm reach oocyte cham- bers through the pores that connect the chambers when the ribbon is extruded from I; the female and starts to absorb water. I Another feature of the reproduction of the white anglerfish to highlight is its type of spermatogenesis, which is known to be semicystic. Spermatogenesis starts inside the cysts that contain germinal cells in dif- ferent stages of development from spermato- i gonia to spermatids, but it is not completed within the cyst. During spermatogenesis, the cyst breaks and spermatids are released . from the cyst into the lumen of the lobule, '! where they become spermatozoa. This kind of spermatogenesis has been described pre- viously in the family Lophiidae only in the ^ blackmouth angler (Lophiomus setigerus) (Yoneda et al., 1998) and in black anglerfish Figure 6 Maturity ogives for (A) male and (B) female white anglerfish (Lophius piscatorius) collected from the northwestern Mediterra- nean Sea between June 2007 and December 2010. Colmenero et al.: Reproductive strategy of Lophius piscatorius in Mediterranean waters 69 Table I Oocyte diameters, ranges and means with standard errors (SEs), and histological characteristics of ovarian follicles in white anglerfish {Lophius piscatorius), collected from the northwestern Mediterranean Sea between June 2007 and December 2010. SE=standard error, at each stage of oocyte development. Stages of Mean oocyte Oocyte diameter oocyte development diameter (pm) (SE) (pm) range Histological characteristics Primary growth stage 82.79 (2.34) 12-203 Nucleus contains a large nucleolus and some peripheral nucleoli. Yolk granules are not pres- ent in the cytoplasm. Cortical alveolar stage 256.10 (3.08) 207-316 Cortical alveolar vesicles and oil droplets appear in the c3doplasm. Yolk granules are not yet pres- ent in the cytoplasm. Nucleus is central within the yolk Vitellogenesis 729.31 (17.58) 324-876 Yolk granules appear between cortical alveolar vesicles. As vitellogenesis advances, yolk gran- ules fill the c3d;oplasm until they are in contact with the nucleus, which remains in a central position. Migratory nucleus 939.31 (7.77) 902-1008 Yolk granules and oil droplets start to fuse. Nucleus migrates to one pole of the oocyte. Hydration 1672.50 (4.77) 1523-1750 Yolk granules form a single mass. Nucleus is not present in the cytoplasm. (Colmenero et al., 2013). This specialized spermato- genesis also has been found in other deep-sea species of Neoceratiidae (Jespersen, 1984) and of Macruridae (Fernandez-Arcaya et al., 2013), in the shore clingfish {Lepadogaster lepadogaster) (Mattei and Mattei, 1978), in species of Blennidae (Lahnsteiner and Patzner, 1990), in a species of Ophidion (Mattei et al., 1993), in the dusky jawfish (Opistognathus whitehursti) (Manni and Rasotto, 1997), and in species of Scorpaena (Munoz et al., 2002; Sabat et al., 2009), which also release their eggs in gelatinous substances. Fecundity Because of their particular reproduction behavior, which includes a high parental investment in the off- spring, white anglerfish are likely to spawn once a year, and the population dynamics of this species are expected to be highly sensitive to external biological and ecosystem factors (ICES"^). Spawning occurs in deep waters because mature white anglerfish have been described by Hislop et al. (2001) as migrating to deeper water before spawning. The same behavior is seen in yellow goosefish: adult fish migrate to deeper waters in response to seasonal changes in water tem- perature and gonadal maturation (Yoneda et al., 2002). ICES (International Council for the Exploration of the Sea). 2012. Report of the working group on the assess- ment of southern shelf stocks of hake, monk and megrim (WGHMM), 10-16 May 2012, ICES Headquaters, Copenha- gen, Denmark. ICES CM 2012/ACOM:11, 617 p. These vertical migrations into deeper waters where commercial fishing and scientific surveys cannot reach could be the reason that very few mature females were captured during our study — a trend that is common in other studies of Lophius species (Ofstad and Lauren- son®). Clearly, this low number of mature females will affect the estimation of such reproductive parameters as fecundity. Generally, deep-sea species have low fecundity and large egg sizes (Gage and Tyler, 1991; Herring, 2002). The white anglerfish has determinate fecundity with values between 661,647 to 885,214 oocytes — levels that are high in comparison with other deep-sea species that inhabit the same depth strata but that are similar to the mean potential fecundity of its Mediterranean con- generic, the black anglerfish (Colmenero et al., 2013). Fecundity values vary among populations as a result of adaptations to local environmental conditions, and they are related to abiotic factors, such as temperature and salinity (Nissling and Dahlman, 2010; Thorsen et al., 2010), and to biotic factors, such as food supply, population density, allocation of energy to reproduction, and fish size (Treasurer, 1981; Merrett, 1994; Nash et al., 2000). In this study, we were not able to determine cor- relations between fecundity and these factors because only 2 actively spawning females were collected. Eggs of white anglerfish have been reported to have a mean diameter of 2.72 mm (SD 0.08) (Colmenero et al., 2015b), a size that is considered large for pelagic eggs, which typically range from 0.5 to 5.5 mm in diameter (Ahlstrom and Moser, 1980). Larger eggs have more 70 Fishery Bulletin 115(1) Figure 7 Distribution of oocyte diameters in (A) spawning capable and (B) ac- tively spawning female white anglerfish {Lophius piscatorius) collected from the northwestern Mediterranean Sea between June 2007 and De- cember 2010. yolk, which increases the potential for larval survival (Duarte and Alcaraz, 1989). The only information avail- able about egg diameters for other species of Lophius is for yellow goosefish, which occupy a bathymetric range that is similar to that occupied by white anglerfish and have a similar egg size (Yoneda et ah, 2001). In con- trast, the black anglerfish has an egg diameter of 1.88 mm (SD 0.12), a size that is nearly 1.5 times smaller than the diameters reported for the white anglerfish and yellow goosefish, and inhabits shallower waters than those inhabited by the other 2 species (Colmenero et ah, 2015b). A comparative study of egg sizes in deep- sea species found that egg size increased significantly with depth (Fernandez-Arcaya®). Egg size is important to offspring survival in many organisms, and large eggs survive better than small ones in environments ® Fernandez-Arcaya, U. 2015. Personal common. Departa- mento de Recursos Marinos Renovables, Institut de Ciencies del Mar, Passeig Man'tim de la Barceloneta, 37-49, 08003 Barcelona, Spain. where dissolved oxygen is low (Hendry I and Day, 2003). r General remarks In this study, we estimated the spawning i season, size at sexual maturity, and fe- I cundity of white anglerfish. Considering the parameter values that we obtained, j we can conclude that this species is one ' that employs a K reproductive strategy. In general, this strategy is defined by a large body size, longevity, late matura- tion, and low fecundity (Pianka, 1970, 1974). A wide range of deep-sea demersal fish species generally display life history characteristics consistent with if-selec- tion (Adams, 1980; Gage and Tyler, 1991). These traits make deep-sea fish stocks highly vulnerable to fishing and capable of little resilience to over-exploitation, increasing the urgency for the conserva- tion and management of this group of animals (Koslow et ah, 2000; Morato et ~ ah, 2006; Norse et ah, 2012). Theoretically, the /^-strategy for deep- sea fish species should imply a low fecun- dity; however, some species, such as the North Pacific armorhead iPseudopentace- ros wheeleri), wreckfish {Polyprion amer- icanus), and splendid alfonsino (Beryx splendens), have high fecundities (Sed- berry et ah, 1996; Lehodey et ah, 1997; Humphreys, 2000). White angelfish and species of Lophius in general also should be included in this group because of their high fecundity (Afonso-Dias and Hislop, 1996; Colmenero et ah, 2013). This variability in re- productive strategy is the result of adaptation to envi- ronmental changes, such as temperature, bathymetric pressure, light, and food availability (Herring, 2002; Brown-Peterson et ah, 2011). Likely, the high fecundity and the low economic value of the white anglerfish, at least until the last decades of the 20th century, has allowed the stock to be sustainable within acceptable limits. With the recent expansion of anglerfish fisher- ies, sustainability is in question, and our study is the first step toward an informed assessment of this deep- j sea resource and its management with an ecosystem j perspective. Acknowledgments The authors would like to thank the crew of the fishing vessels Avi Pau, Estel-lada, Germans Felix, San Benito, and Port de Roses for allowing us to conduct sampling aboard their vessels. We also thank M. Baeta and L. Colmenero et al.: Reproductive strategy of Lophius piscatorius in Mediterranean waters 71 Martinez for helping with data collection, A. Ospina for assisting with the map of the study area, and C. Barria, N. Amat, and R. Alarcon for their valuable comments. We offer special thanks to K. Denning for revising the English of the early draft. This study was part of the project Monitoratge del recursos pesquers i marisquers al litoral catala of the Directorate of Fishing and Mari- time Affairs, Government of Catalonia. Literature cited Adams, P. B. 1980. Life history patterns in marine fishes and their conse- quences for fisheries management. Fish. Bull. 78:1—12. Afonso-Dias, I. R, and J. R. G. Hislop. 1996. The reproduction of anglerfish Lophius piscatorius Linnaeus from the north-west coast of Scotland. J. Fish Biol. 49:18-39. Ahlstrom, E. H., and H. G. Moser. 1980. Characters useful in identification of pelagic marine fish eggs. CalCOFI Rep. 21:121-131. Armstrong, M. P., J. A. Musick, and J. A. Colvocoresses. 1992. Age, growth, and reproduction of the goosefish Lophius americanus (Pisces: Lophiiformes). Fish. Bull. 90:217-230. 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. Bustos-Salvador, R, U. Fernandez-Arcaya, E. Ramirez-Llodra, J. Aguzzi, L. Recasens, and G. Rotllant. 2015. Reproductive biology of NW Mediterranean tongue- fish Symphurus nigrescens and Symphurus ligulatus. J. Mar. Biol. Assoc. U.K. 95:1041-1049. Caruso, J. H. 1983. The systematics and distribution of the lophiid ang- lerfishes: 11. Revisions of the genera Lophiomus and Lophius. Copeia 1983:11-30. 1986. Lophiidae. In Fishes of the North-eastern Atlantic and the Mediterranean (P. J. P. Whitehead, M.-L. Bau- chot, J.-C. Hureau, J. Nielsen, and E. Tortonese, eds.), p. 1362-1363. UNESCO, Paris. Chambers, R. C., and E. A. Trippel (eds.). 1997. Early life history and recruitment in fish populations, 596 p. Chapman and Hall, London. Charnov, E. L. 2008. Fish growth: Bertalanffy k is proportional to repro- ductive effort. Environ. Biol. Fish. 83:185-187. Colmenero, A. I., J. Aguzzi, A. Lombarte, and A. Bozzano. 2010. Sensory constraints in temporal segregation in two species of anglerfish, Lophius budegassa and L. piscato- rius. Mar. Ecol. Prog. Ser. 416:255-265. Colmenero, A. L, V. M. Tuset, L. Recasens, and P. Sanchez. 2013. Reproductive biology of Black Anglerfish {Lophius bu- degassa) in the northwestern Mediterranean Sea. Fish. Bull. 111:390-401. Colmenero, A. I., C. Barria, S. W. Feist, and V. M. Tuset. 2015a. Observations on the occurrence of Spraguea lophii in Mediterranean lophiids. Parasitol. Res. 114: 1977-1983. Colmenero, A. L, V. M. Tuset, J.-M. Fortune, and P. Sanchez. 2015b. The chorion ultrastructure of ova of Lophius spp. J. Fish Biol. 86:1881-1886. Coll, M., 1. Palomera, S. Tudela, and F. Sarda. 2006. Trophic flows, ecosystem structure and fishing im- pacts in the South Catalan Sea, Northwestern Mediter- ranean. J. Mar. Syst. 59:63-96. Cushing, D. H. 1990. Plankton production and year-class strength in fish populations: an update of the match/mismatch hypoth- esis. Adv. Mar. Biol. 26:249-293. D’Onghia, G., P. Maiorano, and L. Sion. 2008. A review on the reproduction of grenadiers in the Mediterranean with new data on the gonad maturity and fecundity. Am. Fish. Soc. Symp. 63:169-184. Dahlgren, U. 1928. The habits and life history of Lophius, the angler- fish. Nat. Hist. 28:18-32. Diaz, R, J. Santos, F. Velasco, A. Serrano, and N. Perez. 2008. Anglerfish discard estimates and patterns in Span- ish Northeast Atlantic trawl fisheries. ICES J. Mar. Sci. 65:1350-1361. Dobby, H., L. Allan, M. Harding, C. H. Laurenson, and H. A. McLay. 2008. Improving the quality of information on Scottish ang- lerfish fisheries: making use of fishers’ data. ICES J. Mar. Sci. 65:1334-1345. Drazen, J. C., and R. L. Haedrich. 2012. A continuum of life histories in deep-sea demersal fishes. Deep-Sea Res. Oceanogr., I 61:34-42. Duarte, C. M., and M. Alcaraz. 1989. To produce many small or few large eggs: a size- independent reproductive tactic of fish. Oecologia 80:401-404. Duarte, R., M. Azevedo, J. Landa, and P. Pereda. 2001. Reproduction of anglerfish {Lophius budegassa Spi- nola and Lophius piscatorius Linnaeus) from the Atlan- tic Iberian coast. Fish. Res. 51:349-361. Erickson, D. L., and E. K. Pikitch. 1993. A histological description of shortspine thornyhead, Sebastolobus alascanus, ovaries: structures associated with the production of gelatinous egg masses. Environ. Biol. Fish. 36:273-282. Farina, A. C., M. Azevedo, J. Landa, R. Duarte, P. Sampedro, G. Costas, M. A. Torres, and L. Canas. 2008. Lophius in the world: a synthesis on the com- mon features and life strategies. ICES J. Mar. Sci. 65:1272-1280. Fernandez-Arcaya, U., E. Ramirez-Llodra, G. Rotllant, L. Re- casens, H. Murua, 1. Quaggio-Grassiotto, and J. B. Company. 2013. Reproductive biology of two macrourid fish, Nezumia aequalis and Coelorinchus mediterraneus, inhabiting the NW Mediterranean continental margin (400-2000 m). Deep-Sea Res. Oceanogr., II 92:63-72. Fishelson, L. 1978. Oogenesis and spawn-formation in the pigmy lion fish Dendrochirus brachypterus (Pteroidae). Mar. Biol. 46:341-348. Fulton, T. W. 1898. The ovaries and ovarian eggs of the angler or frog- fish {Lophius piscatorius), and of the John Dory {Zeus faber). Annu. Rep. Fish. Board Scotl. 16:125-134. Gage, J. D., and P. A. Tyler. 1991. Deep-sea biology: a natural history of organisms at the deep-sea floor, 524 p. Cambridge Univ. Press, Cam- bridge, UK. 72 Fishery Bulletin 1 15(1) Garcia-Diaz, M., J. A. Gonzalez, M. J. Lorente, and V. M. Tuset. 2006. Spawning season, maturity sizes, and fecundity in blacktail comber (Serranus atricauda) (Serranidae) from the eastern-central Atlantic. Fish. Bull. 104:159-166. Griffiths, M. H., and J. Hecht. 1986. A preliminary study of age and growth of the monk- fish Lophius upsicephalus (Pisces: Lophiidae) on the Agulhas Bank, South Africa. S. Afr. J. Mar. Sci. 4:51-60. Hendry, A. P., and T. Day. 2003. Revisiting the positive correlation between female size and egg size. Evol. Ecol. Res. 5:421-429. Herring, R 2002. The biology of the deep ocean, 314 p. Oxford Univ. Press, Oxford. Hislop, J. R. G., A. Gallego, M. R. Heath, F. M. Kennedy, S. A. Reeves, and P. J. Wright. 2001. A synthesis of the early life history of the anglerfish, Lophius piscatorius (Linnaeus, 1758) in northern British waters. ICES J. Mar. Sci. 58:70-86. Humphreys, R. L, Jr. 2000. Otolith-based assessment of recruitment varia- tion in a North Pacific seamount population of armor- head Pseudopentaceros wheeleri. Mar. Ecol. Prog. Sen 204:213-223. Hunter, J. R., and S. R. Goldberg. 1980. Spawning incidence and batch fecundity in northern anchovy, Engraulis mordax. Fish. Bull. 77:641-652. Jespersen, . 1984. Spermatozoans from a parasitic dwarf male of Neoc- eratias spinifer Pappenheim, 1914. Vidensk. Medd. Dan. Nathist. Foren. 145:37-42. Koslow, J. A., J. Bell, P. Virtue, and D. C. Smith. 1995. Fecundity and its variability in orange roughy: ef- fects of population density, condition, egg size, and senes- cence. J. Fish Biol. 47:1063-1080. Koslow, J. A., G. W. Boehlert, J. D. M. Gordon, R. L. Haedrich, P. Lorance, and N. Parin. (northeastern Mediterranean). Hydrobiologia 440:281—296. 2000. Continental slope and deep-sea fisheries: implications for a fragile ecosystem. ICES J. Mar. Sci. 57:548-557. Labropoulou, M., and C. Papaconstantinou. 2000. Community structure of deep-sea demersal fish in the North Aegean Sea. Hydrobiologia 440:281-296. Lahnsteiner, F., and R. A. Patzner. 1990. Spermiogenesis and structure of mature spermatozoa in blenniid fishes (Pisces, Blenniidae). J. Submicros. Cy- tol. Pathol. 22:565-576. Lehodey, P., R. Grandperrin, and P. Marchal. 1997. Reproductive biology and ecology of a deep-demersal fish, alfonsino Beryx splendens, over the seamounts off New Caledonia. Mar. Biol. 128:17-27. Leslie, R. W., and W. S. Grant. 1990. Lack of congruence between genetic and morphologi- cal stock structure of the southern African anglerfish Lophius vomerinus. S. Afr. J. Mar. Sci. 9:379-389. Lopez, N., J. Navarro, C. Barria, M. Albo-Puigserver, M. Coll, and 1. Palomera. 2016. Feeding ecology of two demersal opportunistic pred- ators coexisting in the northwestern Mediterranean Sea. Est. Coast. Shelf Sci. 175:15-23. Maartens, L., and A. J. Booth. 2005. Aspects of the reproductive biology of monkfish Lophi- us vomerinus off Namibia. Afr. J. Mar. Sci. 27:325-330. Maiorano, P., L. Sion, R. Carlucci, F. Capezzuto, A. Giove, G. Costantino, M. Panza, G. D’Onghia, and A. Tursi. 2010. The demersal faunal assemblage of the north-west- ern Ionian Sea (central Mediterranean): current knowl- edge and perspectives. Chem. Ecol. 26:219-240. Manni, L., and M. B. Rasotto. 1997. Ultrastructure and histochemistry of the testicular efferent duct system and spermiogenesis in Opistogna- thus whitehurstii (Teleostei, Trachinoidei). Zoomorphol- ogy 117:93-102. Mattel, C., and X. Mattel. 1978. La spermiogenese d’un poisson teleosteen {Lepado- gaster lepadogaster). 11. Le spermatozoi'de. Biol. Cell. 32:267-274. Mattel, X., Y. Siau, O. T. Thiaw, and D. Thiam. 1993. Peculiarities in the organization of testis of Ophidian sp. (Pisces Teleostei). Evidence for two types of spermato- genesis in teleost fish. J. Fish Biol. 43:931-937. Merrett, N. R. 1994. Reproduction in the North Atlantic oceanic ichthyo- fauna and the relationship between fecundity and spe- cies’ sizes. Environ. Biol. Fish. 41:207-245. Morales-Nin, B., E. Massuti, and C. Stefanescu. 1996. Distribution and biology of Alepocephalus rostratus from the Mediterranean Sea. J. Fish Biol. 48:1097-1112. Morato, T., R. Watson, T. J. Pitcher, and D. Pauly. 2006. Fishing down the deep. Fish Fish. 7:24—34. Munoz, M., M. Casadevall, and S. Bonet. 2002. Testicular structure and semicystic spermatogenesis in a specialized ovuliparous species: Scorpaena notata (Pisces, Scorpaenidae). Acta Zool. 83:213-219. Munoz, M., C. Dimitriadis, M. Casadevall, S. Vila, E. Delgado, J. Lloret, and F. Saborido-Rey. 2010. Female reproductive biology of the bluemouth Heli- colenus dactylopterus dactylopterus: spawning and fecun- dity. J. Fish Biol. 77:2423-2442. Murua, H., and F. Saborido-Rey. 2003. Female reproductive strategies of marine fish spe- cies of the North Atlantic. J. Northwest Atl. Fish. Sci. 33:23-31. Murua, H., G. Kraus, F. Saborido-Rey, P. R. Witthames, A. Thorsen, and S. Junquera. 2003. Procedures to estimate fecundity of marine fish spe- cies in relation to their reproductive strategy. J. North- west Atl. Fish. Sci. 33:33-54. Nash, R. D. M., P R. Witthames, M. Pawson, and E. Alesworth. 2000. Regional variability in the dynamics of reproduction and growth of Irish Sea plaice, Pleuronectes platessa L. J. Sea Res. 44:55-64. Negzaoui-Garali, N., and M. Ben Salem. 2008. Morphological characters and meristic counts in two anglerfishes, Lophius budegassa and L. piscatorius (Osteichthyes; Lophiidae) from Tunisian coastal waters (central Mediterranean). Ann. Sen Hist. Nat. 18:17-26. Nissling, A., and G. Dahlman. 2010. Fecundity of flounder, Pleuronectes flesus, in the Bal- tic Sea — reproductive strategies in two sympatric popu- lations. J. Sea Res. 64:190-198. Norse, E. A., S. Brooke, W. W. L. Cheung, M. R. Clark, I. Eke- land, R. Froese, K. M. Gjerde, R. L. Haedrich, S. S. Heppell, T. Morato, et al. 2012. Sustainability of deep-sea fisheries. Mar. Policy 36:307-320. Pajuelo, J. G., J. M. Lorenzo, and R. Dominguez-Seoane. 2008. Gonadal development and spawning cycle in the digynic hermaphrodite sharpsnout seabream Diplodus Colmenero et al.: Reproductive strategy of Lophius piscatorius in Mediterranean waters 73 puntazzo (Sparidae) off the Canary Islands, northwest of Africa. J. Appl. Ichthyol. 24:68-76. Pavlov, D. A., N. G. Emel’yanova, and G. G. Novikov. 2009. Reproductive dynamics. In Fish reproductive biol- ogy: implications for assessment and management (T. Jakobsen, M. J. Fogarty, B. A. Megrey, and E. Moksness, eds.), p. 48-90. Wiley-Blackwell, Oxford, UK. Pianka, E. R. 1970. On r- and K-selection. Am. Nat. 104:592-597. 1974. Niche overlap and diffuse competition. Proc. Natl. Acad. Sci. U.S.A. 71:2141-2145. Ramirez-Llodra, E., A. Brandt, R. Danovaro, B. De Mol, E. Es- cobar, C. R. German, L. A. Levin, P. M. Arbizu, L. Menot, P. Buhl-Mortensen, B. E. Narayanaswamy, C. R. Smith, D. P. Tittensor, P. A. Tyler, A. Vanreusel, and M. Vecchione. 2010. Deep, diverse and definitely different: unique attri- butes of the world’s largest ecosystem. Biogeosciences 7:2851-2899. Ramirez-Llodra, E., P. A. Tyler, M. C. Baker, O. A. Bergstad, M. R. Clark, E. Escobar, L. A. Levin, L. Menot, A. A. Rowden, C. R. Smith, et al. 2011. Man and the last great wilderness: human impact on the deep sea. PLoS ONE 6(8):e22588. Recasens, L., V. Chiericoni, and P. Belcari. 2008. Spawning pattern and batch fecundity of the Euro- pean hake {Merluccius merluccius (Linnaeus, 1758)) in the western Mediterranean. Sci. Mar. 72:721-732. Rotllant, G., J. Moranta, E. Massuti, F. Sarda, and B. Morales-Nin. 2002. Reproductive biology of three gadiform fish species through the Mediterranean deep-sea range (147-1850 m). Sci. Mar. 66:157-166. Sabat, M., F. Lo Nostro, M. Casadevall, and M. Munoz. 2009. A light and electron microscopic study on the organi- zation of the testis and the semicystic spermatogenesis of the genus Scorpaena (Teleostei, Scorpaenidae). J. Morphol. 270:662-672. Sedberry, G. R., J. L. Carlin, R. W. Chapman, and B. Eleby. 1996. Population structure in the pan-oceanic wreckfish. Polyprion americanus (Teleostei: Polyprionidae), as in- dicated by mtDNA variation. J. Fish Biol. 49:318-329. Stearns, S. C., and J. C. Koella. 1986. The evolution of phenotypic plasticity in life-history traits: predictions of reaction norms for age and size at maturity. Evolution 40:893-913. Stewart, J., K. Rowling, A.-M. Hegarty, and A. Nuttall. 2010. Size and age at sexual maturity of snapper Pagrus auratus in New South Wales 2008/09, 38 p. Indus- try & Investment NSW, Cronulla, North South Wales, Australia. Thorsen, A., P. R. Witthames, G. Marteinsdottir, R. D. M. Nash, and O. S. Kjesbu. 2010. Fecundity and growth of Atlantic cod (Gadus morhua L.) along a latitudinal gradient. Fish. Res. 104:45-55. Treasurer, J. W. 1981. Some aspects of the reproductive biology of perch Perea fluviatilis L. Fecundity, maturation and spawning behaviour. J. Fish Biol. 18:729-740. Tsimenidis, N. 1984. The growth pattern of otoliths of Lophius piscatorius L., 1758 and Lophius budegassa Spinola, 1807 in the Ae- gean Sea. Cybium 8:35-42. Tsimenidis, N. C., and J. C. Ondrias. 1980. Growth studies on the angler-fishes Lophius pisca- torius L., 1758 and Lophius budegassa Spinola, 1807 in Greek waters. Thalassografika 3:63-94. Ungaro, N., G. Marano, R. Auteri, A. Voliani, E. Massuti, M. Garcfa-Rodrfguez, and K. Osmani. 2002. Distribution, abundance and biological features of anglerfish {Lophius piscatorius and Lophius budegassa) (Osteichthyes: Lophiiformes) in the Mediterranean Sea. Sci. Mar. 66:55-63. Valentim, M. F. M., E. P. Caramaschi, and M. Vianna. 2007. Biology and ecology of anglerfishes of the genus Lophius (Lophiidae, Lophiiformes) with emphasis in Lophius gastrophysus Miranda-Ribeiro, 1915: current status. Oecol. Bras. 11:503-520. Vails, M., C. J. Sweeting, M. P. Olivar, M. L. Fernandez de Pu- elles, C. Pasqual, N. V. C. Polunin, and A. Quetglas. 2014. Structure and dynamics of food webs in the water column on shelf and slope grounds of the western Medi- terranean. J. Marine Syst. 138:171-181. van Overzee, H. M. J., and A. D. Rijnsdorp. 2015. Effects of fishing during the spawning period: im- plications for sustainable management. Rev. Fish Biol. Fish. 25:65-83. Walmsley, S. A., R. W. Leslie, and W. H. H. Sauer. 2005. The biology and distribution of the monkfish Lophius vomerinus off South Africa. Afr. J. Mar. Sci. 27:157-168. Wallace, R. A., and K. Selman. 1981. Cellular and dynamic aspects of oocyte growth in te- leosts. Am. Zool. 21:325-343. Watson, R. A., and T. Morato. 2013. Fishing down the deep: accounting for within-species changes in depth of fishing. Fish. Res. 140:63-65. West, G. 1990. Methods of assessing ovarian development in fishes: a review. Aust. J. Mar. Freshw. Res. 41:199-222. Yoneda, M., M. Tokimura, H. Fujita, N. Takeshita, K. Takeshi- ta, M. Matsuyama, and S. Matsuura. 1997. Age and growth of anglerfish Lophius litulon in the East China Sea and the Yellow Sea. Fish. Sci. 63:887-892. 1998. Reproductive cycle and sexual maturity of the ang- lerfish Lophiomus setigerus in the East China Sea with a note on specialized spermatogenesis. J. Fish Biol. 53:164-178. 2001. Reproductive cycle, fecundity, and seasonal distribu- tion of the anglerfish Lophius litulon in the East China and Yellow seas. Fish. Bull. 99:356-370. Yoneda, M., M. Tokimura, H. Horikawa, K. Yamamoto, M. Mat- suyama, and S. Matsuura. 2002. Spawning migration of the anglerfish Lophius li- tulon in the East China and Yellow Seas. Fish. Sci. 68:310-313. 74 National Marine Spencer F. Baird Fisheries Service Fishery BuHetin First U.S. Commissioner w NOAA ftf established in 1881 of Fisheries and founder of Fishery Bulletin Population dynamics and secondary production of juvenile white shrimp iLitopmnaeus setiferus} along an estuarine salinity gradient Marvin M. Mace IIP Lawrence P. Rozas^ Email address for contact author: marvin.mace.iii@gmail.com Abstract — We used estimates of shrimp density, growth, mortality, and secondary production during an 84-d sampling period to compare the value of nursery habitat for juvenile white shrimp {Litopenaeus setiferus) among 3 salinity zones (interme- diate, brackish, and saline zones) within Sabine Lake, an estuary of the northern Gulf of Mexico. Densi- ty, growth, mortality, and secondary production were generally higher in the saline or brackish zones and lowest in the intermediate zone. The saline and brackish zones appeared to provide the most important nurs- ery habitat on a per-area basis, but the intermediate zone also may con- tribute substantially to total shrimp production; although production in the intermediate zone was modest, this zone encompasses a relatively large portion (26%) of coastal wet- lands in Louisiana. The relative val- ue of nursery areas can be dynamic; variation occurs both spatially (e.g., within an estuary and among es- tuaries) and temporally (e.g., from year to year). We documented with- in-estuary differences (i.e., differ- ences among salinity zones within the estuary) in the value of nursery habitat for white shrimp in Sabine Lake and expect this value, espe- cially in the intermediate zone, to vary interannually. The dynamic na- ture of habitats should be considered when assessing the value of estua- rine nursery areas. Manuscript submitted 16 October 2015. Manuscript accepted 26 September 2016. Fish. Bull. 115:74-88 (2017). Online publication date: 22 November 2016. doi: 10.7755/FB.115.1.7 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. ' Department of Biology University of Louisiana at Lafayette P.O. Box 42451 Lafayette, Louisiana 70504 ^ Estuarine Habitats and Coastal Fisheries Center Southeast Fisheries Science Center National Marine Fisheries Service, NOAA 646 Cajundome Boulevard Lafayette, Louisiana 70506 Penaeid shrimps support several valuable commercial fisheries in the United States. Three species, brown shrimp {Farfantepenaeus aztecus), pink shrimp (F. duorarum), and white shrimp {Litopenaeus setiferus), within the family make up most of the landings in terms of weight and economic value (NMFS^). Most white shrimp are landed in the western Gulf of Mexico, specifically in Louisi- ana and Texas (NMFS^). The white shrimp is an estuarine- dependent species that occurs in- shore during the early phase of its life history. Spawning takes place off- shore from late spring to early fall, postlarvae migrate into estuarine habitat, and juvenile shrimp use this nursery habitat for -7-16 weeks be- fore returning offshore to complete their life cycle (Muncy, 1984; Baker et ah, 2014). The importance of es- tuarine nursery areas in support- ing populations of juvenile penaeid shrimps, such as white shrimp, is 1 NMFS (National Marine Fisheries Ser- vice). 2015. Commercial fishery land- ings. [Available at website, accessed October 2015.] widely accepted, but the suitability of this habitat for penaeid shrimps often varies among and within es- tuaries (Vance et ah, 1998; Rozas and Minello, 2010, 2011; Palmer and Montagna, 2015). Vegetated aquatic habitats (e.g., tidal marshes, mangroves, and sea- grasses) within estuaries are thought to provide essential habitat for many species of fish and crustaceans (Kneib, 1997; Kathiresan and Bing- ham, 2001; Heck et al., 2003; Minello et al., 2003). Tidal marshes, in par- ticular, are believed to be important, especially for the young of commer- cially important species, including white shrimp (Boesch and Turner, 1984; Kneib, 1997). In coastal Louisi- ana, tidal marshes are disappearing at a rapid rate because of a variety of natural and anthropogenic causes (Turner, 1990), and identifying areas that are important nurseries for fish- ery species should be a priority for conserving coastal habitats. The density (abundance) and vi- tal rates (growth and mortality) of juvenile white shrimp in estuaries may be useful indicators of habitat value and their contribution to adult Mace and Rozas: Population dynamics of juvenile Litopenaeus setiferus 75 stocks (Diop et aL, 2007; NMFS, 2010; Baker et al., 2014). Abundance of early juvenile white shrimp is, for example, a good predictor of late juvenile shrimp abundance, which is itself a good predictor of the num- ber of shrimp landed by the fishery (Diop et al., 2007). The white shrimp is harvested as an annual fishery crop (Nance et al., 2010), and survival during the ju- venile life stage may have a larger effect on the adult population than adult mortality or fecundity (Baker et al., 2014). Mortality may decrease as the size of juve- nile white shrimp increases (Baker and Minello, 2010); therefore, growth also could be an important factor that influences the recruitment success of white shrimp to the fishery. Salinity is a major environmental factor thought to influence the use of estuarine nursery areas by white shrimp and other nekton at the landscape scale (Wein- stein et al., 1980; Rakocinski et al., 1992; Wagner and Austin, 1999; Rozas and Minello, 2010). For example, juvenile white shrimp are reported to be most abun- dant in mesohaline and polyhaline environments of es- tuaries (Gunter et al., 1964; Howe et al., 1999; Minello, 1999; Rozas and Minello, 2010), although they range much more widely in salinities <1 and >35 (Gunter et al., 1964). These observations are based mainly on comparisons of shrimp density, abundance, or catch per unit of effort from samples collected within estuarine habitats of differing salinity. Density alone, however, may be an insufficient indicator of habitat quality (Van Horne, 1983); additional measures, such as growth, mortality, or secondary production, can provide a more comprehensive picture of habitat quality (Van Horne, 1983; Beck et al., 2001; NMFS, 2010; Dolbeth et al., 2012). We are unaware of any other study in which this full suite of metrics (density, growth, mortality, and secondary production) has been incorporated to com- pare nursery areas of white shrimp along an estuarine salinity gradient. Few studies have examined the ef- fect of salinity on growth and mortality of young white shrimp (Zein-Eldin and Griffith, 1969; Rozas and Mi- nello, 2011), and available estimates indicate that sec- ondary production of white shrimp is higher in estu- aries of the northern Gulf of Mexico than in those of the U.S. Atlantic coast (Zimmerman et al., 2000; Kneib, 2003; Minello et al., 2008). Habitat-specific vital rates (growth and mortality) and estimates of secondary pro- duction are needed to fully assess and compare nursery areas among different salinity regimes. This informa- tion can be used 1) to identify important nursery areas and essential habitats (Beck et al., 2001), 2) to develop detailed population models and improve stock assess- I ment models for white shrimp (Baker et al., 2014), and I 3) to calibrate ecosystem models used to predict effects I of human activities on this and other estuarine species I (Rose et al., 2014). j The purpose of our study was to measure and com- I pare the value of nursery habitat for white shrimp in 3 salinity zones in the Sabine Lake estuary along an estuarine salinity gradient. We collected quantitative density data from each salinity zone in summer and fall, when white shrimp were most abundant in the estuary. Length-frequency data from these samples were used to examine size distributions and to esti- mate growth and mortality rates among the 3 salin- ity zones. We used these data and the size-frequency method (Garman and Waters, 1983) to estimate and compare secondary production of white shrimp among salinity zones. Materials and methods Study site Our study was conducted in the Sabine Lake estuary (hereafter referred to as Sabine Lake) which is located in southwest Louisiana within the area of the coast known as the Chenier Plain (Fig. 1). Sabine Lake en- compasses an area of 375,979 ha; approximately half (49%) of that area is composed of marshes (Gosselink et al-^). Marshes in coastal Louisiana are classified into salinity zones based on dominant vegetation (Chab- reck, 1970; Visser et al., 1998, 2000). We selected 3 sa- linity zones (intermediate, brackish, and saline) using this vegetation classification rather than water salin- ity measured at sampling sites because conditions in estuaries are in constant flux. Vegetation (e.g., marsh type) represents average environmental conditions (e.g. salinity regime) integrated over time better than a sin- gle salinity measurement. The salinity ranges for these vegetation-based zones tend to be 0. 5-5.0 for interme- diate, 5.0-18.0 for brackish, and 18.0-36.0 for saline. Salinity in any zone, however, occasionally may extend outside the typical range. These 3 salinity zones are comparable to the oligohaline, mesohaline, and polyha- line zones, respectively, of the Venice system (Anony- mous, 1958; Visser et al., 1998). Sampling procedure White shrimp were sampled during 6 sampling trips in 2011 (Table 1) that commenced on July 12 (trip 1) and 26 (trip 2), August 9 (trip 3), September 7 (trip 4) and 20 (trip 5), and October 4 (trip 6). A l-m^ drop sampler (Zimmerman et al., 1984) was used to collect 45-60 samples during each of these sampling trips, which each required 3 days to complete (with one salinity zone completed each day). Details of our sampling de- sign are given in Mace and Rozas (2015). Briefly, we se- lected an area of 1 kmxl km within each salinity zone, divided it into 16 squares of equal size (0.25 kmxO.25 km), and, before each sampling trip, we randomly se- lected 5 of these squares for sampling. To obtain a rep- ^ Gosselink, J. G., C. L. Cordes, and J. W. Parsons. 1979. An ecological characterization study of the Chenier Plain coastal ecosystem of Louisiana and Texas. 3 vols., FWS/OBS-78/9, 78/10, and 78/11. Natl. Coast. Ecosyst. Team, Off. Biol. Serv., U.S. Fish Wildl. Serv., Slidell, LA. 76 Fishery Bulletin 115(1) « Intermediate. Sabine Lake Louisiana Brackish Gulf of Mexico Texas Saline 0 10 Km 1 I Figure 1 Map of the study area in Sabine Lake (at the border of Texas and Loui- siana). Samples of juvenile white shrimp {Litopenaeus setiferus) were collected in 2011 to estimate density, natural mortality, growth, and secondary production at 3 locations along the estuarine salinity gradi- ent (zones=intermediate, brackish, and saline) that were each 100 ha in size. Black dots represent locations where individual drop samples were collected. resentative sample of the shrimp population from each selected square, a drop sample was taken from each of 4 habitat types where shrimp are known to occur: 1) marsh vegetation <1 m from the interface of ro.arsh and open water (marsh edge), 2) shallow water <1 m from the marsh edge (SWl), 3) shallow water 1-5 m from the marsh edge (SWl-5), and 4) shallow water >5 m from the marsh edge (SW>5). Locations of SW>5 sampling sites were determined by randomly selecting distances between 6 m from the nearest shore and the middle of the water body. Only 3 of the 4 habitat types were sampled on the first 3 sample dates in most of the salinity zones when low water levels precluded sampling at the marsh edge Size (Table 1). Before collecting each drop sample, salinity, temperature, depth, and turbidity were measured according to the protocol described in Rozas et al. (2012). Each nekton sample was placed on ice, preserved in 10% formalin at the end of the day, and transported to the laboratory for processing. In the labora- tory, all organisms were separated from detritus, and juvenile penaeid shrimps were identified to species by using the characters from Perez Farfante^, Ditty (2011), and references therein. Cara- pace length was measured for all juve- nile white shrimp as described in Mace and Rozas (2015). The duration of flooding for each habitat was estimated by using water levels measured in the field and equa- tions fitted for each salinity zone (Mi- nello et al., 2012). We collected data on water levels at 20 randomly located transects within each salinity zone. Along each transect, water levels were measured at the marsh edge and at 0.5, 1, 2, 3, 4, and 5 m on each side of the marsh edge. To derive equations for es- timating long-term water levels in each salinity zone, we regressed water-level data from temporary tide gauges placed in each salinity zone against data from nearby tide gauges of the NOAA Center for Operational Oceanographic Products and Services (CO-OPS: website) and the U.S. Geological Survey’s Coastwide Reference Monitoring System (CRMS; website): NOAA station 8770570 for the saline zone, NOAA station 8770475 for the brackish zone, and CRMS site 0662 for the intermediate zone. We es- timated the flooding durations (percent- age of time water depth was >5 cm) at habitats in each salinity zone from July through October 2011 by relating the water depth measured at each transect sampling site to concurrent tide data calculated from these fitted equations. We examined size, measured as total length in mil- limeters, of white shrimp collected in our samples by comparing box plots of size for each salinity zone and sample date. Descriptive statistics, such as minimum, 3 Perez Farfante, I. 1970. Diagnostic characters of juve- niles of the shrimps Penaeus aztecus aztecus, P. duorarum duorarum, and P. brasiliensis (Crustacea, Decapoda, Penaei- dae). U.S. Fish Wiidl. Serv., Sped. Sci. Rep. Fish. 599, 26 p. I Mace and Rozas; Population dynamics of juvenile Litopenaeus setiferus 77 Table 1 Distribution of samples used to estimate density, natural mortality, growth, and secondary production of juvenile white shrimp (Litopenaeus setiferus) in Sabine Lake. Samples were collected on 6 sampling trips in 2011 in 3 salinity zones (intermediate, brackish, and saline) and 4 habitat types (1 — marsh edge, or marsh vegetation <1 m from the interface of marsh and open water [ME]; 2 — shallow water <1 m from the marsh edge [SWlj; 3 — shallow water 1-5 m from the marsh edge [SWl-5]; and 4 — shallow water >5 m from the marsh edge [SW>5]). Zone Trip Date ME SWl Habitat SWl-5 SW>5 Total Intermediate 1 07/12/11 0 5 5 5 15 2 07/26/11 0 5 5 5 15 3 08/9/11 0 10 5 5 20 4 09/9/11 5 5 5 5 20 5 09/20/11 5 5 5 5 20 6 10/4/11 5 5 5 5 20 Brackish 1 07/13/11 0 5 5 5 15 2 07/27/11 0 5 5 5 15 3 08/10/11 0 10 5 5 20 4 09/7/11 5 5 5 5 20 5 09/21/11 5 5 5 5 20 6 10/5/11 5 5 5 5 20 Saline 1 07/14/11 0 5 5 5 15 2 07/28/11 5 5 5 5 20 3 08/11/11 0 10 5 5 20 4 09/8/11 5 5 5 5 20 5 09/22/11 5 5 5 5 20 6 10/6/11 5 5 5 5 20 maximum, and mean size, were also computed for sa- linity zones and sample dates. Density We restricted our analysis of shrimp density to the last 3 sampling trips when sample sizes were equal both among habitats types and salinity zones (Table 1). Before analysis, these density data were ln(x+l) transformed to remove the relationship between the mean and variance present in untransformed data (Milliken and Johnson, 1992). We used an analysis of variance (ANOVA) with 3 factors: 1) salinity zone (in- termediate, brackish, saline), 2) habitat type (marsh edge, SWl, SWl-5, SW>5), and 3) sampling trip (4 to 6). The ANOVA model included a 3-way interaction, all 2-way interactions, and all main effects. After a signifi- cant main effect of habitat type, 3 a priori contrasts were made: 1) marsh edge versus SWl, 2) SWl versus i SWl-5+SW>5, and 3) marsh edge versus SWl+SWl- ! 5+SW>5. The first and third contrasts allowed us to I compare marsh edge with adjacent shallow water and i all shallow-water sites combined, respectively. We com- pared shallow-water sites near (<1 m) and those sites farther away from the marsh edge with the second con- trast. Results were considered significant at F<0.05. We also estimated omega squared (co^) for each factor in the ANOVA, using the formulas for measured factors given in Olejnik and Algina (2003). Omega squared is an estimate of effect size that can be interpreted as the proportion of variance explained by a given factor in an ANOVA. All density data presented in the text, tables, and figures are untransformed means. Mortality We converted length-frequency distributions to age-fre- quency distributions to estimate mortality. Before con- verting to age-frequency distributions, length-frequency distributions were weighted to account for differences in the relative amount of habitat types within each sa- linity zone (Suppl. Table 1, avail, online). Weighting was accomplished by first constructing length-frequency distributions for each habitat type within a salinity zone and then multiplying the number of individuals in each size class by the total area of a given habitat type in that zone, resulting in 4 different length-fre- quency distributions for each zone on each sample date. Length-frequency data from samples collected in marsh edge habitat were applied to the entire marsh within 5 m of the shoreline, although densities of juvenile white shrimp may decline as distance from the shoreline in- creases from 1 to 5 m into marsh vegetation (Minello et al., 2008). Because we were interested in the entire 78 Fishery Bulletin 1 15(1) shrimp population in each salinity zone on each sample date, we combined length-frequency distributions from all 4 habitat types in each zone from each sample date. Finally, we converted the combined length-frequency distribution in each salinity zone on each sample date into a relative frequency distribution and apportioned the total number of shrimp collected in each zone on each sample date to fit this distribution. These weight- ed length-frequency distributions were used in subse- quent analyses of mortality and growth. Daily instantaneous mortality rates (Z) were es- timated in each salinity zone by using a horizontal catch-curve analysis (Vetter, 1988). We assumed Z was equal to the daily instantaneous natural mortal- ity rate (M) because there was no fishery for juvenile white shrimp in our study area, and we removed only a very small fraction of the total population during our sampling. We combined and converted length-frequency data from multiple sampling trips within each salin- ity zone to age-frequency data by assuming a growth rate of 1 mm TL/d, a reasonable assumption for juve- nile penaeid shrimps in general (Dali et ah, 1990) and white shrimp specifically (Rozas and Minello, 2011). Using the age-frequency data, we calculated mortal- ity with the following 2 methods: 1) a linear regres- sion of the ln(density+0.1) against age, where the slope is an estimate of mortality (Ricker, 1975) and 2) the Chapman-Robson estimator (Chapman and Robson, 1960) with the standard error corrected for overdisper- sion (Smith et ah, 2012). We started the catch curve with the regression method at the age of highest abun- dance and included all ages up to but not including the first age at which time there were <1 individual (Smith et ah, 2012). For the Chapman-Robson estimator, we included all ages greater than the age of peak abun- dance. We compared mortality rates between catch- curve methods and among salinity zones by first taking the difference between 2 estimates and then construct- ing a 95% confidence interval (CD for the difference. We considered results significantly different when the 95% Cl of the difference between 2 estimates did not include zero (Schenker and Gentleman, 2001). Growth We calculated growth rates by following individual co- horts and using the mean size of a cohort on 2 consecu- tive sample dates to estimate a mean growth rate be- tween sample dates. Individual cohorts of shrimp were identified from each of the length-frequency distribu- tions on different sample dates in each salinity zone by using the mixdist package, vers. 0.5-4 (Macdonald and Du, 2012) in R, vers. 3.1.0 (R Core Team, 2014), and the mean growth rate for a cohort between sample dates was calculated as p _ /^t+i ~ Mt '-^absolute , . > n+1 ~ h where /J.^ = the mean carapace length at time t; and /^t + 1 = the mean carapace length at time ^ + 1. Before calculating growth, carapace length was con- verted to total length by using the formula TL = CL X 4.944 (Baker and Minello, 2010). This conversion makes it easier to compare our growth rates with pub- lished values. Mean growth rates among salinity zones were compared as described above for mortality rates. For each sample date in each salinity zone, we at- tempted to model cohorts by using normal, lognormal, or gamma distributions based on previous observations of length-frequency distributions of penaeid shrimp as they immigrated into estuaries (Baxter and Renfro, 1967). We used Akaike’s information criterion adjusted for small sample size (AICc), AAICc values (the differ- ence in AICc values between a given model and the model with the lowest AICc value), and values (AIC weights, which can be interpreted as an estimate of the probability that a given model is the best among all models considered, given the data) to compare and se- lect the best model or models that described the shape and the mean size and standard deviation of the mean of shrimp cohorts (Burnham and Anderson, 2002). Secondary production Secondary production that occurred in each salinity zone over the 84-d sampling period was estimated in kilograms per hectare with the size-frequency meth- od (Garman and Waters, 1983) because we were not able to track cohorts over all sample dates. The size- frequency method uses the mean number and weight of individuals in each size class over time to estimate the biomass lost as individuals move through the size- frequency distribution. We used shrimp up to 60 mm TL to estimate secondary production because shrimp larger than this size begin to emigrate from estuaries into the Gulf of Mexico (Pullen and Trent, 1969). We used the length-to-weight conversion given in Minello et al. (2008) to estimate mean shrimp weights and as- sumed this relationship was similar for all 3 salinity zones. We also estimated total biomass, measured in kilograms per hectare, within each salinity zone for the 84-d sampling period as the sum of the mean biomass of all size classes over all sample dates. One of the most influential parameters affecting production estimates with the use of the size-frequency method is the cohort production interval (CPI) (Benke, 1979; Garman and Waters, 1983). The size-frequency method was originally developed to estimate produc- tion for insects whose larvae develop in aquatic habi- tats and produce only one generation per year. This original method was modified by including the CPI into the calculation to account for species that have multiple generations per year. Benke (1979) originally defined the CPI in terms of the amount of time taken to complete larval development (i.e., the aquatic stages when growth and production occur). Garman and Wa- ters (1983) defined the CPI for fish as the average max- imum age obtained by individuals in the population. In our study, the age of a shrimp depended on its length because we estimated age from length; there- Mace and Rozas: Population dynamics of juvenile Litopenaeus setiferus 79 Figure 2 Mean water (A) salinity, (B) temperature, (C) dissolved oxygen, (D) turbidity (in nephelometric turbidity units, NTUs), and (E) depth within intermediate, brack- ish, and saline zones measured on 6 trips during which samples of juvenile white shrimp (Litopenaeus setiferus) were collected in 2011. Means and standard errors (1 SE) were computed from 15 to 20 replicate measurements. fore, growth rates could have affected our age esti- mates and the resultant CPI. We used a conservative approach for initial estimates of secondary production by assuming that shrimp in each salinity zone had the same CPI of 53 d (i.e., the maximum age of a shrimp since recruitment to the population as a 7 mm TL post- larva until migration out of the population at 60 mm TL, assuming a growth rate of 1 mm TL/d). We also calculated secondary production using CPIs for shrimp in each salinity zone based on the growth rates we es- 1 timated as described above. For example, if the mean growth rate that we estimated for a salinity zone was 2 mm TL/d, a 60-mm-TL shrimp that arrived in the population at 7 mm TL would have aged 27 d since it arrived in the population. Therefore, the CPI based on this 2 mm TL/d growth rate would be 27. Results Environmental variables Environmental variables, other than salinity, differed little among salinity zones (Fig. 2). Southwest Loui- 80 Fishery Bulletin 115(1) Figure 3 Comparison of mean flooding duration (number of hours water depth was > 5 cm/total number of hoursxlOO) for marsh edge and shallow-water sites among 3 differ- ent salinity zones (intermediate, brackish, and saline) in Sabine Lake from July 2011 through October 2011. Negative values indicate distances (from the marsh edge) within marsh vegetation, and positive values are distances from the marsh edge over shallow water. Means and standard errors (1 SE) were calculated from 20 replicate measurements. siana experienced drought conditions starting in late 2010 until the end of 2011 (NOAA National Centers for Environmental Information, Climate Data Online, web- site, accessed October 2015). As a result, mean salini- ties during our study in Sabine Lake were slightly out- side the range typical for the brackish zone and well above the normal range for the intermediate zone (Fig. 2). Mean water temperature varied little among salin- ity zones during each sampling trip and was slightly lower during the last 3 trips than during the first 3 trips. Mean dissolved oxygen was less variable during the first 3 sampling trips than during the last 3 trips, but no clear trend was detected in dissolved oxygen among salinity zones. We observed no consistent trend in either mean water depth or turbidity among salin- ity zones. Flooding duration varied among salinity zones and distances from the marsh edge (Fig. 3). The marsh edge habitat was flooded for the highest percentage of the time in the intermediate zone (70%), followed by the brackish (50%) and saline (42%) zones. The mean dura- tion of flooding of the vegetated marsh surface >0.5 m from the marsh edge (i.e., the negative numbers on the a:-axis in Fig. 3) was relatively short (<20% of the time) in all 3 salinity zones. In the saline zone, shallow-water sites within 5 m of the marsh edge were flooded for relatively short durations as well (Fig. 3). Size White shrimp ranged in size from 4 to 126 mm TL, on the basis of data from all 6 sampling trips to the 3 salinity zones. Most shrimp (98%), however, were ju- veniles <60 mm TL, and, in general, size distributions and mean sizes were similar among salinity zones and sampling dates (Fig. 4). Mean size on most sampling dates ranged from 12 to 35 mm TL, except in the in- termediate zone on trip 5 when mean size was 41 mm TL. Density Mean densities varied among habitat types, salinity zones, and sample dates (Table 2). The overall ANO- VA model was significant (P<0.001) and accounted for 60% of the variation in the shrimp density data. All 3 main effects were significant, as were the interactions between habitat type and salinity zone and between salinity zone and sampling trip; the 3-way interaction was not significant (Table 3). Habitat type was the most important factor in the model, and it accounted for an order of magnitude more variation (45%) in the density data than that of any other factor (<6%, Table 3). The interaction of habi- tat type with salinity zone was weak, accounting for only 4% of the variation in the density data. Results from a priori contrasts indicated that shrimp densities were higher in the marsh edge habitat than in adja- cent shallow water (P=0.0379), higher in shallow water near the marsh edge than in other shallow-water sites (P<0.0001), and higher at the marsh edge sites than at all shallow-water sites combined (P<0.0001). Mean shrimp density in each habitat type was gen- erally highest in the saline zone. For example, mean shrimp densities at the marsh edge sites were estimat- ed as 41.5 individuals/m^ (standard error [SE] 10.2), 56.7 individuals/m^ (SE 11.2), and 85.1 individuals/ m^ (SE 22.6) for the intermediate, brackish, and sa- line zones, respectively (Table 2). The SWl habitat type was an exception; mean density at this type of habitat (based on the last 3 sampling trips) was highest in the brackish zone and lowest in the intermediate zone (Fig. 5). Although the main effect of salinity zone was signif- icant, this effect varied by sample date; mean density was usually highest in the saline zone, except for on trip 5 when mean density was higher in the brackish zone than in the saline zone (Fig. 6). Mortality We estimated mortality by combining shrimp length- frequency data from the last 3 sampling trips in each salinity zone and conducting 2 catch-curve analyses per zone. Within a salinity zone, no difference in mor- tality rates could be detected between the results of the 2 catch-curve analyses (Table 4). Among salinity zones, no statistically significant difference in mortal- ity rates was detected between the saline (0.09 [SE Mace and Rozas: Population dynamics of juvenile Litopenaeus setiferus 81 the intermediate zone (0.05 [SE 0.006] and 0.07 [SE 0.008]). Only the linear- regression estimates indicated significantly greater mortality in the saline zone than in the intermediate zone. Growth We were unable to consistently track individual co- horts of shrimp over sampling trips in each salinity zone. Identifying individual cohorts on some sample dates was difficult because there were no clear modes that indicated separate cohorts, and the mixdist pack- age could not adequately fit any of the 3 distributions to the data. On most sampling dates when we were able to track individual cohorts, more than 1 model had some support from AIC values; for these dates, we used model-averaged estimates for mean carapace length from models that had a combined AIC weight >0.95 (Suppl. Tables 2 and 3, avail, online). Mean growth rate estimates among all salin- ity zones and cohorts combined ranged from 0.72 mm TL/d (SE 0.28) to 1.83 mm TL/d (SE 0.23) (Table 5). Among salinity zones, mean growth rates were 1.22 mm TL/d (SE 0.13), 0.99 mm TL/d (SE 0.19), and 1.62 mm TL/d (SE 0.12) in the intermediate, brackish, and saline zones, respectively (Table 5). Mean growth rates were significantly higher in the saline zone than in the brackish and intermediate zones. No significant differ- ence in mean growth rates was detected between the brackish and intermediate zones. Secondary production Secondary production of juvenile white shrimp esti- mated with the conservative approach (i.e., the one based on the same growth rate of 1 mm TL/d in each salinity zone) varied by salinity zone and was significantly higher in the saline than in the inter- mediate zone (Table 6). Secondary production in the saline zone during the 84-d sampling period was es- timated as 382 kg/ha (95% Cl, 187 to 577), which was approximately 3 times the value of 116 kg/ha (95% Cl, 27 to 205) for the intermediate zone. In the brackish zone, production was estimated as 232 kg/ ha (95% Cl, 102 to 361), approximately twice that of the intermediate zone. Production-to-biomass ratios over the 84-d sampling period were estimated as 5.0, 6.5, and 7.4 in the intermediate, brackish, and sa- line zones, respectively (Table 6). When we used the mean growth rates estimated for each salinity zone (Table 5) to compute shrimp age and CPIs, the over- all pattern of secondary production during the 84-d sampling period was the same among salinity zones, although some estimates of production were higher (intermediate=142 kg/ha, brackish=227 kg/ha, and sa- line=614 kg/ha) when compared with results from the 0.007] and 0.08 [SE 0.006]) and brackish (0.08 [SE conservative approach (Table 6). Differences in pro- 0.005] and 0.09 [SE 0.006]) zones, but estimates that duction estimates appear to have been driven mainly resulted from the use of both methods showed signifi- by differences in shrimp density among salinity zones cantly higher mortality in the brackish zone than in (Table 2). 140 - C 120 - 100 - 80 60 40 20 •[ 0 1210 344 75 702 X 1 2 3 4 5 6 Trip Figure 4 Box plots of the size, measured as total length in millimeters, of juvenile white shrimp {Litopenaeus setiferus) collected on 6 sampling trips in 3 salinity zones, (A) intermediate, (B) brackish, and (C) sa- line zones, in Sabine Lake in 2011. The black line inside boxes represents the mean size, the boxes extend to the 25^^ and 75*^ percentiles, and the whiskers extend to the minimum and maximum size. Numbers above each box-and-whisker plot are the sample size for each sampling trip. 82 Fishery Bulletin 1 15(1) Table 2 Mean densities, measured in individuals per square meter, of juvenile white shrimp {Litopenaeus setiferus) in Sabine Lake during 6 sampling trips in 2011 in 3 salinity zones (intermediate, brackish, and saline) and 4 habitat types: 1 — marsh edge, or marsh vegetation <1 m from the interface of marsh and open water (ME); 2 — shallow water <1 m from the marsh edge (SWl); 3 — shallow water 1-5 m from the marsh edge (SWl-5); and 4 — shallow water >5 m from the marsh edge (SW>5). A dash indicates that no sample was taken in that habitat on that date. Sample sizes for each mean range from 5 to 10 and are provided in Table 1. Standard errors (SEs) of the means are given in parentheses. Zone Intermediate Brackish Saline Trip ME SWl SWl-5 SW>5 ME SWl SWl-5 SW>5 ME SWl SWl-5 SW>5 1 _ 1.2 (0.8) 0.0 (0.0) 0.0 (0.0) _ 11.2 (5.9) 0.4 (0.4) 2.4 (1.9) _ 6.0 (5.3) 7.6 (4.4) 1.6 (0.7) 2 - 7.8 (4.7) 0.4 (0.4) 0.0 (0.0) - 13.6(2.7) 4.6 (4.1) 3.0 (2.0) 63.0 (28.2) 29.4 (26.2) 3.8 (3.1) 4.0 (1.5) 3 - 12.0 (5.3) 1.4 (0.5) 0.0 (0.0) - 27.7(10.4) 1.0 (0.8) 2.2 (1.6) - 46.4(11.7) 29.4(12.2) 10.0(3.2) 4 61.8(25.4) 25.2 (9.4) 0.4 (0.4) 0.2 (0.2) 74 (22.1) 58.6(23.1) 2.8 (1.5) 1.8 (0.6) 60.4 (24.4) 65.0 (36.2) 6.2 (3.9) 10.4 (3.4) 5 16.0(4.7) 19.4 (4.8) 0.4 (0.4) 0.6 (0.6) 64.8 (20.0) 51.0(15.8) 8.4 (5.1) 1.6 (0.9) 43.0(9.2) 9.0 (3.9) 10.8 (3.7) 5.6 (2.6) 6 46.8(12.6) 47.8 (20.7) 4.0 (1.3) 0.6 (0.2) 31.4(13.4) 28.8 (6.6) 5.8 (2.1) 9.4 (8.4) 174.0 (74.6) 33.0(16.1) 17.0 (3.8) 22.8 (3.5) Overall mean 41.5(10.2) 17.9 (4.2) 11(0.3) 0.2 (0.1) 56.7(11.2) 31.2 (5.5) 3.8 (1.2) 3.4 (1.5) 85.1(22.6) 33.6 (7.8) 12.4 (2.8) 9.1 (1.6) Table 3 Results of 3-way analysis of variance (ANOVA) to examine differences in mean density of ju- venile white shrimp (Litopenaeus setiferus) among salinity zones, habitat types, and sampling trips in Sabine Lake in 2011. A measure of effect size, omega squared (co^), is given for each factor and can be interpreted as an estimate of the proportion of variance accounted for by each source of variation. Data used in this analysis were obtained from trips 4-6. Factor df Sum of squares F ratio P 0)^ Zone 2 27.90 13.86 <0.0001 0.06 Trip 2 8.80 4.37 0.0143 0.02 Habitat type 3 208.57 69.10 <0.0001 0.45 Zone*Trip 4 13.35 3.32 0.0124 0.02 Zone*Habitat type 6 25.30 4.19 0.0006 0.04 Trip*Habitat type 6 8.92 1.48 0.1899 0.01 Zone*Trip*Habitat type 12 13.70 1.34 0.3373 0.00 Error 144 144.90 p ii Discussion The suitability of habitat for juvenile white shrimp varies among salinity zones within Sabine Lake. The saline and brackish zones provide more important nursery habitat for white shrimp than does the inter- mediate zone, on the basis of our estimates of shrimp density, growth, and secondary production. The inter- mediate zone contains the least important nursery area, although shrimp mortality rates were lower there than in the 2 other salinity zones. Although growth, mortality, and secondary production, together with den- sity, are considered necessary for a comprehensive as- sessment of nursery habitat (Beck et ah, 2001), and density alone may sometimes be a misleading indica- tor of habitat quality (Van Horne, 1983), in our study. density appeared to be an accurate indicator of habitat quality for juvenile white shrimp. Habitat type accounted for most of the variation in density of white shrimp. Density of white shrimp was highest for habitat <1 m from the marsh edge and de- clined with increasing distance into open water in all 3 salinity zones. This general pattern for white shrimp, namely high density near the marsh edge and a decline in density in open water, has been observed in another area of Sabine Lake (Nevins et al., 2014), in Galveston Bay, Texas (Minello et al., 2008), and in Barataria Bay, Louisiana (Rozas and Minello, 2015). Densities of white shrimp at the marsh edge sites in our study area were among the highest densities reported for this species in the northern Gulf of Mexico. This concentration of shrimp may be a response to the short flooding dura- Mace and Rozas: Population dynamics of juvenile Litopenaeus setiferus 83 Figure § Comparison of mean densities of juvenile white shrimp {Litopenaeus setiferus) among 4 habitat types and 3 sa- linity zones (intermediate, brackish, and saline) within Sabine Lake, based on data from the last 3 sampling trips (trips 4-6) in 2011. Habitat types are 1 — marsh edge (ME), or marsh vegetation <1 m from the interface of marsh and open water; 2 — shallow water <1 m from the marsh edge (SWl); 3 — shallow water 1-5 m from the marsh edge (SWl-5); and 4 — shallow water >5 m from the marsh edge (SW>5). Means and standard er- rors were computed from 15 replicate samples. Figure represents the significant interaction of habitat type and salinity zone. tion of marsh >1 m from the marsh edge because this short duration limited access to the vegetated marsh surface and concentrated individuals at the marsh edge sites. Nekton density patterns in shallow water also can be influenced by the distribution of submerged aquatic vegetation (SAV) or other structure (Rozas and Minello, 2010). In our study area in 2011, SAV was absent be- cause of an ongoing drought; however, SAV can be pres- ent, especially in the intermediate and brackish zones of estuaries during periods of normal rainfall (Chab- reck, 1971). Because densities of juvenile white shrimp are similar in marsh vegetation and SAV (Howe and Wallace, 2000; Rozas and Minello, 2006), we expect that the distribution pattern of shrimp in our inter- mediate and brackish zones could be quite different in years of normal rainfall, when SAV beds in these loca- tions may be extensive. The pattern we observed in growth rates of juve- nile white shrimp among salinity zones was similar to that reported for the Cape Fear River estuary in North Carolina (Laney and Copeland^), but the growth rates we observed in our study area were overall higher than ^ Laney, R. W., and B. J. Copeland. 1981. Population dynam- ics of penaeid shrimp in two North Carolina tidal creeks. Rep. 81-1, 161 p. Carolina Power & Light Co., Raleigh, NC. Trip Figure 6 Comparison of mean densities of juvenile white shrimp {Litopenaeus setiferus) among 3 sampling trips (trips 4-6) made in 2011 and among 3 salinity zones (in- termediate, brackish, and saline) within Sabine Lake. Means and standard errors were computed from 20 replicate samples. Figure represents the significant in- teraction between sample date and salinity zone. those reported in Barataria Bay (Rozas and Minello, 2011). In laboratory experiments, extremes in salinity have been shown to reduce growth rates of juvenile white shrimp (Zein-Eldin and Griffith, 1969). Salinity may directly influence growth rates through increased metabolic costs for shrimp in low-salinity (e.g., oligo- haline) areas (Rozas and Minello, 2011), although this explanation did not likely pertain to our study because salinities in the intermediate zone were elevated by the ongoing drought and more typical of the brackish zone. Although comparisons of natural mortality rates among salinity zones are rare, results of available stud- ies indicate that mortality may increase with salinity for a variety of species. Mortality of juvenile white shrimp within marsh tidal creeks of the Cape Fear River estuary was lower in low-salinity areas than in high-salinity areas (Laney and Copeland^). The mortal- ity rate for juvenile spot {Leiostomus xanthurus) in the Cape Fear estuary was lower at a low-salinity site than at a high-salinity site during the first year, although it was similar between the 2 sites in the second year (Weinstein and Walters, 1981). Mortality rates for this species and Atlantic croaker {Micropogonias undula- tus) were also lower at low-salinity, up-estuary sites than at high-salinity sites near tidal inlets in both the Cape Fear River and Pamlico Sound estuaries of North Carolina (Ross, 2003). Possible explanations given by Weinstein and Walters (1981) and Ross (2003) for this pattern include the following: 1) higher density of pred- ators at high-salinity sites, 2) higher stress on juvenile fish induced by high salinity indirectly increasing mor- tality, and 3) higher migration rates from saline sites biasing mortality rates upwards in those areas. 84 Fishery Bulletin 115(1) Table 4 Daily instantaneous natural mortality estimates (and standard errors) for juvenile white shrimp (Litopenaeus setiferus) within intermediate, brackish, and saline zones of Sabine Lake. Mortality rates were es- timated with 2 different catch-curve analyses, linear regression and the Chapman-Robson estimator, by using data from trips 4-6 in 2011. Analysis Intermediate Brackish Saline Linear regression 0.05 (0.006) 0.08 (0.005) 0.09 (0.007) Chapman-Robson 0.07 (0.008) 0.09 (0.006) 0.08 (0.006) Table 5 Absolute growth rate estimates (Gabsoiute) standard errors calculated by following individual cohorts of Ju- venile white shrimp {Litopenaeus setiferus) collected during 3 sampling trips (trips 4-6) in 2011 within in- termediate, brackish, and saline zones of Sabine Lake. Growth rates per day. are given in total length in millimeters Zone Start End ^absolute Intermediate Trip 4 Trip 5 1.22 (0.13) Brackish Trip 4 Trip 5 0.77 (0.45) Trip 4 Trip 5 0.74 (0.28) Trip 5 Trip 6 1.72 (0.50) Trip 5 Trip 6 0.72 (0.28) mean 0.99 (0.19) Saline Trip 2 Trip 3 1.42 (0.07) Trip 2 Trip 3 1.83 (0.23) Mean 1.62 (0.12) Although we did not directly address these possible explanations for our results, we know that juvenile white shrimp can survive in a wide range of salini- ties (Zein-Eldin and Griffith, 1969), and we believe it is unlikely that environmental conditions would be more stressful for juvenile shrimp in a saline zone than in an intermediate zone (Rozas and Minello, 2011). We also tried to minimize any bias due to migration by considering only shrimp <60 mm TL in our mortality estimates. White shrimp do not begin to migrate from estuaries into the northern Gulf of Mexico until they reach 60 mm TL (Pullen and Trent, 1969). We, there- fore, conclude that increased predation may be respon- sible for the relatively higher mortality rates we esti- mated in the saline and brackish zones, compared with the rates in the intermediate zone. We did not quantify the density of all potential predators of juvenile white shrimp in each salinity zone to test this possibility, but previous studies indicate that juvenile penaeid shrimps are more abundant in the diets of fish predators from high-salinity areas of estuaries than in the diets of predators in areas with low salinity (Minello et al., 1989) and in the diets of predators at estuarine sites near the sea than at sites farther up an estu- ary (Salini et ah, 1990). Alternatively, this higher shrimp mortality may have been the result of less access to marsh edge habitat in the brackish and saline zones than in the intermediate zone. In the in- termediate zone, marsh edge habitat was flooded for longer periods and, therefore, may have provided more protection from predators than that provided by this type of habitat in the other salinity zones. Our secondary production estimates for juvenile white shrimp in shallow marsh habitats of Sabine Lake are similar to, and within the range of, production esti- mates for this species reported elsewhere. Comparisons among published values should be made cautiously, however, because of differences in definitions of the term secondary production and the methods used to estimate it. We know of no other comparisons of white shrimp production among salinity zones, but produc- tion (technically yield) of white shrimp in aquaculture ponds of different salinities was approximately 3.5 times greater in 4 ponds with high salinity (2 ponds at 15 and 2 ponds at 21) than in 2 low-salinity (7) ponds (Hysmith and Colura, 1976). Our estimate of 382 kg/ha for production during the 84-d sampling period in the saline zone is higher than the estimate of annual pro- duction for juvenile white shrimp (109 kg/ha) reported for saline marshes in Galveston Bay (Minello et ah, 2008; Table 4). The difference in these 2 estimates of production of juvenile white shrimp is most likely due to the higher densities we documented in our study area than those reported for Galveston Bay. Using shrimp landings data and area of emergent wetlands in the U.S. states bordering the Gulf of Mexico, Engle (2011) estimated mean annual production (technically yield) of penaeid shrimps (all species combined) as 241 kg/ha (range: 57-1660 kg/ha). The intermediate zone contained less valuable habi- tat than the saline zone on a per-hectare basis during the 84-d sampling period, but the estimated produc- tion from the intermediate zone, which occupies a large proportion of coastal marshes within Louisiana, was not trivial (116 kg/ha). Therefore, the production from these low-salinity areas would contribute substantially to the total production of white shrimp in Louisiana estuaries. Multiplying our production estimates in the saline and intermediate zones by the total area of these 2 zones determined for coastal Louisiana in 2013 (Sasser et al.®), the total production over the 84-d sam- pling period from the intermediate zone would equal ® Sasser, C. E., J. M. Visser, E. Mouton, J. Linscombe, and S. B. Hartley. 2014. Vegetation types in coastal Louisiana in 2013: U.S. Geological Survey Scientific Investigations Map 3290, 1 sheet, scale 1:550,000. [Available at website.] Mace and Rozas: Population dynamics of juvenile Litopenaeus setiferus 85 Table 6 Estimates of secondary production (with 95% confi- dence interval and measured as kilograms per hectare), mean biomass (measured as kilograms per hectare), and production-to-biomass ratios for juvenile white shrimp {Litopenaeus setiferus) within intermediate, brackish, and saline zones of Sabine Lake. Production (P) and biomass (B) estimates are based on data collected over an 84-d period from July through October 2011 (trips 1-6) and production was estimated by using the size- frequency method (Garman and Waters, 1983). Salinity zone Production Biomass P:B Intermediate 115.7 (26.8 to 204.7) 23.3 5.0 Brackish 231.6(102.4 to 360.8) 35.6 6.5 Saline 382.0 (186.8 to 577.3) 51.7 7.4 44 million kg in comparison with 113 million kg from the saline zone. Our study was conducted during drought conditions in southwest Louisiana, and the relative habitat value of the 3 salinity zones in our study area may differ during nondrought periods. During years of normal rainfall, the intermediate zone may provide less valu- able habitat for white shrimp. For example, the mean density of juvenile white shrimp within the marsh edge habitat of the intermediate zone during September in 2012, a year of normal rainfall, was 80% less than the density we observed there in the same month in 2011 and report here; whereas, shrimp densities in the same habitat type in the brackish and saline zones differed between these 2 years by <20% (senior author, unpubl. data). Juvenile penaeid shrimp abundance and com- mercial landings may be higher during warm, dry pe- riods with low freshwater inflows to estuaries (Moller et al., 2009; Piazza et al., 2010), although Gunter and Hildebrand (1954), Browder (1985), and Palmer and Montagna (2015) reported examples of a positive rela- tionship between rainfall or freshwater inflow and the abundance or commercial landings of penaeid shrimp. More comparisons of secondary production from other estuaries conducted over several years are needed be- fore definitive conclusions can be drawn about patterns of juvenile white shrimp production among salinity zones. Inferences from our results are limited to the lo- cations and time period we sampled. Our study was confined to a single year, and sampling sites were rep- licated in space but drawn from an area limited to 100 ha in each salinity zone. Support for extending the inferences from our study more generally, however, comes from studies of other estuaries that corroborate our results. For example, the spatial distribution and growth rates of juvenile white shrimp among salinity zones that we observed are consistent with the pat- tern reported from studies of Barataria Bay, Louisiana (Rozas and Minello, 2010, 2011). In a meta-analysis of 5149 samples collected from multiple locations in the northern Gulf of Mexico, mean densities of juvenile white shrimp were highest in mesohaline and polyha- line areas (equivalent to our brackish and saline salin- ity zones, respectively) (Minello, 1999) — densities that were consistent with our results. Additional compari- sons of demographic rates and secondary production from other estuaries of the Gulf of Mexico would allow broader inferences to be drawn in the future. Our estimates of density, biomass, growth, natural mortality, and secondary production of juvenile white shrimp were generally higher in the saline or brack- ish zones and lowest in the intermediate zone. To our knowledge, this study is the first attempt to simultane- ously estimate and compare population dynamics and secondary production of juvenile penaeid shrimps along an estuarine salinity gradient. Although inference from our work is limited in scope to the 3 locations and the time period we sampled, the saline and brackish zones provided more important nursery habitat for juvenile white shrimp than such habitat in the intermediate zone. The total amount of production from the inter- mediate zone, however, was not trivial when the area that this zone covers in coastal Louisiana is consid- ered. Nursery habitats that provide a small contribu- tion on a per-area basis, such as the intermediate zone in our study area, may still have a large effect at the population level because the habitat covers a relative- ly large total area (Dahlgren et al., 2006). Moreover, the relative value of nursery habitats can be dynamic, with variation occurring both spatially (e.g., within as well as among estuaries) and temporally, from year to year (Kraus and Secor, 2005). We documented within- estuary (i.e., among salinity zones) differences in habi- tat value for white shrimp in Sabine Lake and expect this value, especially in the intermediate zone, to vary interannually. For example, the habitat value of the intermediate zone in our study area likely would be less in a year in which rainfall was greater than or equal to average levels (i.e., in a year of lower salinity) than in the year we documented, 2011, a year of se- vere drought and relatively high salinity. This dynamic nature of habitat value should be considered when as- sessing estuarine nursery areas. Acknowledgments We thank S. Hillen and J. Salas of the NOAA South- east Fisheries Science Center’s (SEFSC) Galveston Laboratory; L. Broussard, S. Beck, and D. O’Malley of Louisiana State University; A. Cummings of the SEF- SC Estuarine Habitats and Coastal Fisheries Center in Lafayette, Louisiana; and J. C. Robichaux and J. Thompson of the University of Louisiana at Lafayette for help collecting and processing samples. We would also like to thank D. Richard at Stream Wetland Ser- vices, Lake Charles, Louisiana, for helping to locate sites and for providing lodging, the staff at Murphree 86 Fishery Bulletin 115(1) Wildlife Management Area for providing lodging, and the staff of the Texas Point National Wildlife Refuge for access to one of our field sites. We thank P. Caldwell for conducting the GIS analysis needed to weight the length-frequency distributions and for creating Fig- ure 1. A teaching assistaetship from the University of Louisiana at Lafayette, the Rockefeller Wildlife Schol- arship, the Coastal Conservation Association/Ted Beau- lieu Sr. Scholarship, a research grant from the Wetland Foundation, and the SEFSC provided funding for this project. We also thank T. Minello and 3 anonymous re- viewers whose comments improved this article. Literature cited Anonymous. 1958. Symposium on the classification of brackish waters. Oikos 9:311-312. Baker, R., and T. J. Minello. 2010. Growth and mortality of juvenile white shrimp Lito- penaeus setiferus in a marsh pond. Mar. Ecol. Prog. Sen 413:95-104. Baker, R., M. Fujiwara, and T. J. Minello. 2014. Juvenile growth and mortality effects on white shrimp Litopenaeus setiferus population dynamics in the northern Gulf of Mexico. Fish. Res. 155:74-82. Baxter, K. N., and W. C. Renfro. 1967. Seasonal occurrence and size distribution of postlar- val brown and white shrimp near Galveston, Texas, with notes on species identification. Fish. Bull. 66:149-158. 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. Benke, A. C. 1979. A modification of the Hynes method for estimat- ing secondary production with particular significance for multivoltine populations. Limnol. Oceanogr. 24:168- 171. Boesch, D. F., and R. E. Turner. 1984. Dependence of fishery species on salt marshes: the role of food and refuge. Estuaries 7:460-468. Browder, J. A. 1985. Relationship between pink shrimp production on the Tortugas Grounds and water Sow patterns in the Florida Everglades. Bull. Mar. Sci. 37:839-856. Burnham, K. P., and D. R. Anderson. 2002. Model selection and multimodel inference: a practical information-theoretic approach, 2"*^ ed., 488 p. Springer, New York. Chabreck, R. H. 1970. Marsh zones and vegetation types in the Louisiana coastal marshes. Ph.D. diss., 113 p. Louisiana State Univ., Baton Rouge, LA. 1971. Ponds and lakes of the Louisiana coastal marshes and their value to fish and wildlife. Proc. Annu. Conf. SEAFWA 25:206-215. Chapman, D. G., and D. S. Robson. 1960. The analysis of a catch curve. Biometrics 16:354- 368. Dahlgren, C. P., G. T. Kellison, A. J. Adams, B. M. Gillanders, M. S. Kendall, C. A. Layman, J. A. Ley, I. Nagelkerken, and J. E. Serafy. 2006. Marine nurseries and effective Juvenile habi- tats: concepts and applications. Mar. Ecol. Prog. Ser. 312:291-295. Dali, W., B. J. Hill, P. C. Rothlisberg, and D. J. Sharpies. 1990. The biology of the Penaeidae. Adv. Mar. Biol. 27:1-484. Diop, H., W. R. Keithly Jr., R. F. Kazmierczak Jr., and R. F. Shaw. 2007. Predicting the abundance of white shrimp {Litope- naeus setiferus) from environmental parameters and pre- vious life stages. Fish. Res. 86:31-41. Ditty, J. G. 2011. Young of Litopenaeus setiferus, Farfantepenaeus aztecus and F. duorarum (Decapoda: Penaeidae): A re- assessment of characters for species discrimination and their variability. J. Crustac. Biol. 31:458-467. Dolbeth, M., M. Cusson, R. Sousa, and M. A. Pardal. 2012. Secondary production as a tool for better under- standing of aquatic ecosystems. Can. J. Fish. Aquat. Sci. 69:1230-1253. Engle, V. D. 2011. Estimating the provision of ecosystem services by Gulf of Mexico coastal wetlands. Wetlands. 31:179-193. Garman, G. C., and T. F. Waters. 1983. Use of the size-frequency (Hynes) method to es- timate annual production of a stream fish popula- tion. Can. J. Fish. Aquat. Sci. 40:2030-2034. Gunter, G., and H. H. Hildebrand. 1954. The relation of total rainfall of the state and catch of the marine shrimp (Penaeus setiferus) in Texas wa- ters. Bull. Mar. Sci. 4:95-103. Gunter, G., J. Y. Christmas, and R. Killebrew. 1964. Some relations of salinity to population distribu- tions of motile estuarine organisms, with special refer- ence to penaeid shrimp. Ecology (USA) 45:181-185. Heck, K. L., Jr., G. Hays, and R. J. Orth. 2003. Critical evaluation of the nursery role hypothesis for seagrass meadows. Mar. Ecol. Prog. Ser. 253:123-136. Howe, J. C., and R. K. Wallace. 2000. Relative abundance of postlarval and juvenile penaeid shrimps in submerged aquatic vegetation and emergent marsh habitats. Gulf Mex. Sci. 18:130-137. Howe, J. C., R. K. Wallace, and F. Scott Rikard. 1999. Habitat utilization by postlarval and juvenile penaeid shrimps in Mobile Bay, Alabama. Estuaries 22:971- 979. Hysmith, B. T., and R. L. Colura. 1976. Effect of salinity on growth and survival of penaeid shrimp in ponds. J. World Aquat. Soc. 7:289-303. Kathiresan, K., and B. L. Bingham. 2001. Biology of mangroves and mangrove ecosystems. Adv. Mar. Biol. 40:81-251. Kneib, R. T. 1997. The role of tidal marshes in the ecology of estuarine nekton. Oceanogr. Mar. Biol. Annu. Rev. 35:163—220. 2003. Bioenergetic and landscape considerations for scal- ing expectations of nekton production from intertidal marshes. Mar. Ecol. Prog. Ser. 264:279-296. Kraus, R. T., and D. H. Secor. 2005. Application of the nursery-role hypothesis to an es- tuarine fish. Mar. Ecol. Prog. Ser. 291:301-305. Mace and Rozas: Population dynamics of juvenile Litopenaeus setiferus 87 Macdonald, R, and J. Du. 2012. Mixdist: finite mixture distribution models. R package vers. 0.5-4. [Available at website, accessed Oc- tober 2015.] Mace, M., Ill, and L. P. Rozas. 2015. Estimating natural mortality rates of juvenile white shrimp Litopenaeus setiferus. Estuar. Coasts 38:1580- 1592. Milliken, G. A., and D. E. Johnson. 1992. Analysis of messy data. Vol. 1: designed experi- ments, 473 p. Chapman and Hall, London. Minello, T. J. 1999. Nekton densities in shallow estuarine habitats of Texas and Louisiana and the identification of essential fish habitat. Am. Fish. Soc. Symp. 22:43-75. Minello, T. J., R. J. Zimmerman, and T. E. Czapla. 1989. Habitat-related differences in diets of small fishes in Lavaca Bay, Texas, 1985-1986. NOAA Tech. Memo. SEFC-NMFS-236, 16 p. Minello, T. J., K. W. Able, M. P. Weinstein, and C. G. Hays. 2003. Salt marshes as nurseries for nekton: testing hy- potheses on density, growth and survival through meta- analysis. Mar. Ecol. Prog. Ser. 246:39-59. Minello, T. J., G. A. Matthews, P. A. Caldwell, and L. P. Rozas. 2008. Population and production estimates for decapod crustaceans in wetlands of Galveston Bay, Texas. Trans. Am. Fish. Soc. 137:129-146. Minello, T. J., L. P. Rozas, and R. Baker. 2012. Geographic variability in salt marsh flooding pat- terns may affect nursery value for fishery species. Es- tuar. Coasts 35:501-514. Moller, O. O., Jr., J. P. Castello, and A. C. Vaz. 2009. The effect of river discharge and winds on the in- terannual variability of the pink shrimp Farfantepenaeus paulensis production in Patos Lagoon. Estuar. Coasts 32:787-796. Muncy, R. J. 1984. Species profiles: life histories and environmental requirements of coastal fishes and invertebrates (Gulf of Mexico) — white shrimp. U.S. Fish Wildl. Serv. FWS/ OBS-82/11.20. U.S. Army Corps Eng., TR EL-82-4, 19 p. Nance, J. M., C. W. Caillouet Jr., and R. A. Hart. 2010. Size-composition of annual landings in the white shrimp, Litopenaeus setiferus, fishery of the northern Gulf of Mexico, 1960-2006: its trend and relationships with other fishery-dependent variables. Mar. Fish. Rev. 72:1-13. Nevios, J. A., J. B. Pollack, and G. W. Stunz. 2014. Characterizing nekton use of the largest unfished oyster reef in the United States compared with adjacent estuarine habitats. J. Shellfish Res. 33:227-238. NMFS (National Marine Fisheries Service). 2010. Marine fisheries habitat assessment improvement plan. Report of the National Marine Fisheries Service Habitat Assessment Improvement Plan Team. NOAA Tech. Memo. NMFS-F/SPO-108, 115 p. Olejnik, S., and J. Algina. 2003. Generalized eta and omega squared statistics: measures of effect size for some common research de- signs. Psychol. Methods 8:434-447. Palmer, T. A., and P. A. Montagna. 2015. Impacts of droughts and low flows on estuarine wa- ter quality and benthic fauna. Hydrobiologia 753:111- 129. Piazza, B. P., M. K. La Peyre, and B. D. Keim. 2010. Relating large-scale climate variability to local spe- cies abundance: ENSO forcing and shrimp in Breton Sound, Louisiana, USA. Glim. Res. 42:195-207. Pullen, E. J., and W. L. Trent. 1969. White shrimp emigration in relation to size, sex, temperature and salinity. FAO Fish. Rep. 57:1001-1014. R Core Team. 2014. R: a language and environment for statistical com- puting. R Foundation for Statistical Computing, Vienna, Austria. [Available from website, accessed April 2014] Rakocinski, C. F., D. M. Baltz, and J. W. Fleeger. 1992. Correspondence between environmental gradients and the community structure of marsh-edge fishes in a Louisiana estuary. Mar. Ecol. Prog. Ser. 80:135-148. Ricker, W. E. 1975. Computation and interpretation of biological sta- tistics of fish populations. Bull. Fish. Res. Board Can. 191, 382 p. Rose, K. A., H. Huang, D. Justic, and K. de Mutsert. 2014. Simulating fish movement responses to and poten- tial salinity stress from large-scale river diversions. Mar. Coast. Fish. 6:43-61. Ross, S. W. 2003. The relative value of different estuarine nursery areas in North Carolina for transient juvenile marine fishes. Fish. Bull. 101:384-404. Rozas, L. P., and T. J. Minello. 2006. Nekton use of Vallisneria americana Michx. (Wild celery) beds and adjacent habitats in coastal Louisi- ana. Estuar. Coasts 29:297-310. 2010. Nekton density patterns in tidal ponds and adja- cent wetlands related to pond size and salinity. Estuar. Coasts 33:652-667. 2011. Variation in penaeid shrimp growth rates along an estuarine salinity gradient: implications for managing river diversions. J. Exp. Mar. Biol. Ecol. 397:196-207. 2015. Small-scale nekton density and growth patterns across a saltmarsh landscape in Barataria Bay, Louisi- ana. Estuar. Coasts 38:2000-2018. Rozas, L. P., T. J. Minello, and D. D. Dantin. 2012. Use of shallow lagoon habitats by nekton of the northeastern Gulf of Mexico. Estuar. Coasts 35:572- 586. Salini, J. P., S. J. M. Blaber, and D. T. Brewer. 1990. Diets of piscivorous fishes in a tropical Australian estuary, with special reference to predation on penaeid prawns. Mar. Biol. 105:363-374. Schenker, N., and J. F. Gentleman. 2001. On judging the significance of differences by ex- amining the overlap between confidence intervals. Am. Stat. 55:182-186. Smith, M. W., A. Y. Then, C. Wor, G. Ralph, K. H. Pollock, and J. M. Hoenig. 2012. Recommendations for catch-curve analysis. North Am. J. Fish. Manage. 32:956—967. Turner, R. E. 1990. Landscape development and coastal wetland losses in the northern Gulf of Mexico. Am. Zool. 30:89-105. Van Horne, B. 1983. Density as a misleading indicator of habitat quality. J. Wildl. Manage. 47:893-901. Vance, D. J., M. D. E. Ha3nvood, D. S. Heales, R. A. Kenyon, and N. R. Loneragan. 1998. Seasonal and annual variation in abundance of 88 Fishery Bulletin 115(1) postlarval and juvenile banana prawns Penaeus mer- guiensis and environmental variation in two estuaries in tropical northeastern Australia: a six year study. Mar. Ecol. Prog. Ser. 163:21-36. Vetter, E. F. 1988. Estimation of natural mortality in fish stocks: a re- view. Fish. Bull. 86:25-43. Visser, J. M., C. E. Sasser, R. H. Chabreck, and R. G. Linscombe. 1998. Marsh vegetation types of the Mississippi River Deltaic Plain. Estuaries 21:818-828. Visser, J. M., C. E. Sasser, R. H. Chabreck, and R. G. Linscombe. 2000. Marsh vegetation types of the Chenier Plain, Loui- siana, USA. Estuaries. 23:318-327. Wagner, C. M., and H. M. Austin. 1999. Correspondence between environmental gradients and summer littoral fish assemblages in low salinity reaches of the Chesapeake Bay, USA. Mar. Ecol. Prog. Ser. 177:197-212. Weinstein, M. R, and M. P. Walters. 1981. Growth, survival and production in young-of-year populations of Leiostomus xanthurus Lacepede residing in tidal creeks. Estuaries 4:185-197. Weinstein, M. R, S. L. Weiss, and M. F. Walters. 1980. Multiple determinants of community structure in shallow marsh habitats. Cape Fear River estuary. North Carolina, USA. Mar. Biol. 58:227-243. Zein-Eldin, Z. R, and G. W. Griffith. 1969. An appraisal of the effects of salinity and tempera- ture on growth and survival of postlarval penaeids. FAO Fish. Rep. 57:1015-1026. Zimmerman, R. J., T. J. Minello, and G. Zamora Jr. 1984. Selection of vegetated habitat by brown shrimp, Penaeus aztecus, in a Galveston Bay salt marsh. Fish. Bull. 82:325-336. Zimmerman, R. J., T. J. Minello, and L. P. Rozas. 2000. Salt marsh linkages to productivity of penaeid shrimps and blue crabs in the northern Gulf of Mex- ico. In Concepts and controversies in tidal marsh ecology (M. P. Weinstein and D. A. Kreeger, eds.), p. 293-314. Kluwer Academic Publishers, Dordrecht, Netherlands. 89 National Marine Fisheries Service NOAA Fishery Bulletin fb- established in 1881 •100 m), and highest occurrence of adult females was found inshore, in waters (depths <78 m) shallower than the waters in which adult males oc- curred, and farther inside the Cape Cod peninsula than the other locations surveyed (DelFApa et al., 2014). DelFApa et al. (2014) suggested that adult females may actively seek shallow waters as a strategy to elude adult males and energy-demanding copulation — a tactic that results in spatial segregation. Alternatively, in deeper coastal areas, where males are usually more abundant, adult females may synchronize their different habitat- selection behavior with that of adult males, resulting in temporal segregation (DelFApa et al., 2014). Overall, the results of that study indicate that sex- ual segregation in spiny dogfish, in the Cape Cod area, occurs both spatially and temporally and is strongly in- fluenced by depth. Different behavioral choices by each of the sexes are reflected in their preferred habitat as- sociations and indicate that depth and time, but likely not temperature, are potential key drivers for sexual segregation in spiny dogfish (DelFApa et al., 2014). However, that study was based on limited data collect- ed through fishery-dependent surveys in a regionally restricted inshore coastal area, and the findings may not be representative of the entire U.S. Atlantic stock of spiny dogfish. Therefore, for fishery management purposes it would be useful to analyze existing data sets from a larger coastal area, one that encompasses the entire range of the current commercial fishery be- tv/een New England and North Carolina. In addition, it would be beneficial to analyze com- parable regional data from fishery-independent inshore surveys. In this regard, the nearshore trawl surveys conducted by the Northeast Area Monitoring and As- sessment Program (NEAMAP) in the mid-Atlantic rep- resent the most suitable data source that is available for completing a more extensive study on the influence of environmental habitat characteristics, season, and time of day on the distribution of spiny dogfish in in- shore coastal waters of the northeastern United States. The aim of this study was to use the NEAMAP sur- vey data to model the abundance of spiny dogfish, by sex, in inshore coastal waters of the U.S. Atlantic, by considering oceanographic (i.e., sea-surface tempera- ture [SST], salinity, and chlorophyll-a [chl-a] concen- tration), topographic (i.e., depth, slope, and distance to land), and temporal (i.e., season and time of day) char- acteristics as predictive variables for catch per unit of effort (CPUE). This study involved the use of hierarchi- cal Bayesian spatiotemporal models and is the first one to use a large fishery-independent database as a source of data for analysis and prediction of the habitat dis- tribution of spiny dogfish along the northwest Atlantic inshore coastal area, by sex and by time of day. The results of this study provide information on the spatial and temporal distribution of adult spiny dogfish that will be valuable for fishery managers because it will enable the adoption of enhanced management strate- gies for the fishery for spiny dogfish in the U.S. Atlantic that are based on the sexual segregation exhibited by this species. Materials and methods Mid-Atlantic surveys NEAMAP surveys are conducted in the coastal waters bounded by the western edge of Cape Cod, Massachu- setts, and by Cape Hatteras, North Carolina. From Montauk, New York, and southward, sampling is con- fined within the 18.3-m depth contour. In the deeper nearshore waters off southern New England, the deep- est stations extend to about 36.6 m (Fig. 1). NEAMAP conducts 2 cruises per year, one each in spring (April- May) and fall (September-October), timed to rough- ly coincide with offshore surveys carried out by the NOAA Northeast Fisheries Science Center (NEFSC). Each NEAMAP cruise samples 150 stations distributed among 17 regions and 2 depth strata. To allow compari- sons with current NEFSC surveys, NEAMAP adopted the bottom trawl design developed for the NEFSC by the Northeast Trawl Survey Advisory Panel, joint board of the Mid- Atlantic and New England Fishery Manage- ment Councils. The NEAMAP conducts surveys by fish- ing with a 3-bridle, 4-seam bottom trawl with a net of 400 meshes (of 12 cm width) for a total length of 48 Dell'Apa et al.: Modeiing the distribution of Squalus acanthias, by sex 91 Map of the study area. Black circles represent the locations of 1820 trawl hauls conducted in inshore coastal waters of the mid-Atlantic during surveys by the Northeast Area Monitoring and Assessment Program. Data from these surveys were used to determine the dis- tribution of spiny dogfish (Squalus acanthias) between New England and North Carolina in 2007-2013. m, a 7.6-cm cookie sweep, and a 2.5-cm knotless liner in the eodend. The doors are 1.7-m Thyboron Type IV. Data collection, sources, and analysis Fishery-independent data on spiny dogfish were col- lected during the NEAMAP surveys in spring and fall of 2007-2013 on the FA^ Darana R. As described by Bonzek et al.^, for each haul, up to approximately 18 spiny dogfish were examined to determine individual length (precaudal length [PCL] in centimeters), weight, sex, and maturity (males only, by external examination of the claspers). All other specimens were weighed in total and measured individually. In rare cases of very 1 Bonzek, C. F., J. Gartland, R. A. Johnson, and J. D. Lange Jr. 2008. NEAMAP Near Shore Trawl Survey: peer review documentation. A report to the Atlantic States Marine Fish- eries Commission by the Virginia Institute of Marine Science (VIMS), Gloucester Point, VA. [Available from VIMS, P.O. Box 1346, Gloucester Point, VA 23062-1346]. large catches, those individuals not processed as de- scribed were counted (or weighed) in full without fur- ther processing, before being released. When subsam- pling occurred, the attributes of the subsampled por- tion (e.g., sex ratios and lengths) were expanded to the total catch (Bonzek et al.^). Additionally, for each haul, the date, starting and ending times, and location (lati- tude and longitude) were recorded. For the purpose of our analysis, the CPUE for spiny dogfish, by sex, for each trawl haul was calculated as the total number of individuals of each sex for each 20 min of trawling. The length of individuals was con- verted from PCL to total length (TL) in centimeters by using the conversion factor iTL=PCL/0. 807) available for this species in FishBase (website). The analysis for this study included only adult females (>80 cm TL) and adult males (>60 cm TL), on the basis of the size at maturity reported for this species in the northwest At- lantic (Nammack et al., 1985). It is worth noting that at the time of the study by Nammack et al.’s (1985), 92 Fishery Bulletin 115(1) the stock of spiny dogfish was not subjected to the in- tense exploitation by the domestic commercial fishery that began in the early 1990s. This exploitation caused this stock to be declared overfished in 1998 and result- ed in the development of a fishery management plan in 1999 for this species in federal waters (5-322 km offshore) by the New England and Mid-Atlantic Fish- ery Management Councils (MAFMC^). The federal plan was further reinforced by an interstate fishery manage- ment plan for state waters (0-5 km offshore) developed by the Atlantic States Marine Fisheries Commission in 2002 (ASMFC3; Dell’Apa et ah, 2015). Additionally, the commercial fishery has preferen- tially targeted adult females because of their larger size. This fishing strategy has resulted in a recent in- crease in the adult male:female sex ratio in the catch of this species and in a decrease in the average size at maturity for adult females in the U.S. Atlantic stock from the sizes reported by Nammack et al. (1985) (Sos- ebee, 2005; Rago and Sosebee”*) — a drop from 80 cm TL to about 74.5 cm TL, according to Bubley et al. (2013). Overall, this decline indicates that the actual size at maturity of adult females from the NEAMAP surveys during 2007-2013 was likely to have been smaller than the 80 cm TL used in our analysis. However, because no interannual variability in the predicted CPUE of adult females was found by using the 80-cm-TL size at maturity criterion and because of the inherent difficul- ties in choosing an alternative size criterion as a result of consistent fiuctuation in the annual average size at maturity reported for adult females (Marques da Silva and Ross®), we opted to adopt a more conservative ap- proach by using the most commonly accepted size at maturity reported by Nammack et al. (1985). For statistical purposes, the time of each set was classified into 3 categories: morning (between 6:00 AM and 12:59 PM), afternoon (between 1:00 PM and 6:59 PM), and night (between 7:00 PM and 5:59 AM), ac- cording to the time partitioning used by DelFApa et al. (2014). Six environmental variables were included in the analysis: bathymetry (mean depth of each haul in feet and converted to meters for analysis), distance to shore (measured in meters), slope of the seabed (percent grade), monthly mean SST (measured in degrees Cel- ^ MAFMC (Mid-Atlantic Fishery Management Council). 1999. Spiny dogfish fishery management plan, 292 p. Mid- Atlantic Fishery Management Council, Dover, DE. [Avail- able from website.] ^ ASMFC (Atlantic States Marine Fisheries Commis- sion). 2002. Interstate fishery management plan for spiny dogfish. ASMFC, Fish. Manage. Rep. 40, 98 p. [Available from website.] Rago, R, and K. Sosebee. 2012. Update on the status of spiny dogfish in 2012 and initial evaluation of harvest at the Fmsy proxy, 43 p. Science and Statistical Committee, Mid-Atlantic Fishery Management Council, Dover, DE. ® Marques da Silva, H., and M. R. Ross. 1993. Reproduc- tive strategies of spiny dogfish, Squalus acanthias, in the NW Atlantic. ICES Council Meeting (C.M.) Documents 1993/ G:51,18 p. [Available from website.] sius), monthly mean chl-a concentration (measured in milligrams per cubic meter), and monthly mean values of practical salinity. Data for 3 variables — SST, chl-a concentration, and salinity — were extracted from the NASA Earth Obser- vations website (website) as long-term monthly mean climate data. Bathymetry was derived from the same NASA Earth Observations website by using the Gen- eral Bathymetric Chart of the Oceans (GEBCO) grid (website). In addition, data for bathymetry were col- lected at each haul location. These data were used to correct and check the information on final mean depth. When a discrepancy occurred between the GEBCO and survey data, a mean of the values in the 2 data sets was used. Distances to the coast and slope gradients were derived from the bathymetry map created with the GEBCO grid, by using the Near (World Equidistant Cy- lindrical coordinate system) and Slope Spatial Analyst tools in ArcGis 9.2® (Esri, Redlands, CA). All the covariates were aggregated at a resolution of 0.25°x0.25° and were transformed into raster lay- ers with the raster package (Hijmans, 2013) in R, vers. 3.1.2 (R Core Team, 2014). To check collinearity be- tween explanatory environmental variables, a drafts- man’s plot and the Pearson’s correlation index were used. Because variables were not correlated highly with coefficients of correlation (r) <0.5, they were con- sidered in further analyses. Modeling species abundance For the purpose of our analysis and modeling, we used the spatial distribution approach, which combines ob- servations of species occurrence or abundance with en- vironmental estimates to predict spiny dogfish distribu- tion at locations that were not sampled (Austin, 2007; Elith and Leathwick, 2009). Different approaches and methods can be used to model the spatial distribution of a species. However, most of the common applications do not always provide accurate results when run with traditional prediction methods (i.e., frequentist infer- ence), often because of a large amount of spatiotempo- ral variability in the data that characterizes dynamic marine ecosystems (Roos et al., 2015). To account for this variability, we used hierarchical Bayesian spatio- temporal models in our study. Bayesian approaches have several advantages over traditional methods and have been applied successfully to fisheries studies (Colloca et al., 2009; Munoz et al., 2013; Pennine et al., 2014). Bayesian methods allow the inclusion of both the observed data and model pa- rameters as random variables (Banerjee et al., 2004) and provide more realistic and accurate estimations of uncertainty (Pennine et al., 2014). Additionally, they al- low the use of spatial and temporal components as 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. Deli'Apa et al.; Modeling the distribution of Squalus acanthias, by sex 93 Table 1 Summary of variables as potential fixed effects that influence the distribution of spiny dogfish (Squalus acanthias) in the mid-Atlantic and included in Bayesian models. SST=sea-surface temperature; chl- a=chiorophyll-a concentration; practical salinity is a ratio and does not have physical units. Variable Description Units Bathymetry Mean fishing depth of haul m Slope Seabed slope at the sampling station % grade Distance to shore Distance from the coast at the sampling station m SST SST monthly value of haul °c Chl-a Chl-a monthly value of haul mg/m® Salinity Salinity of the water - Season Season when haul was sampled Spring, fall Time Time when haul was sampled Morning, afternoon, night random-effect term, reducing their influence on the es- timation of the habitat variables (Gelfand et al., 2006). Hierarchical Bayesian spatiotemporai models were used to predict abundance of spiny dogfish, by sex, with respect to explanatory variables, as well as to describe the main spatial distribution changes over time, by sex. These models are extremely applicable to studies characterized by data observed at continuous locations within a defined spatial area, as was the case for the data set used in this study. Values of CPUE from the NEAMAP surveys were considered appropriate proxies for levels of abundance of spiny dogfish. The spatial variation of the CPUE values for spiny dogfish, by sex, was modeled by using a hierarchical Bayesian spatiotemporai approach, specifically a Pois- son point process model with log-linear intensity. It was assumed that the number of spiny dogfish at each station sampled, has a Poisson distribution with rate tj where is the observation time at site i and A.i is proportional to relative species abundance at sta- tion i and measures the survey expectation for a unit observation time, according to this general formulation: logikOi - a + ZijP + Fj + Wi, where a P Wi the intercept; the matrix of covariates at the year j and the station I; the vector of the regression coefficients; the component of the temporal unstructured random effect at the year j; and the spatially structured random effect at the station i. In this model, independence between the sampling locations is assumed. However, some spatial autocor- relation may be present in the data set because the abundance of a species at nearby stations is influenced by similar environmental parameters. Consequently, adjacent, or nearby, stations would be expected to be similar in terms of abundance of spiny dogfish. The W, accounts for this influence. For each of the models for both sexes, 8 potential fixed-effects were considered: 6 environmental variables and 2 temporal variables (Table 1). For all the parameters considered in the fixed-effects model, a vague zero-mean Gaussian prior distribution with a variance of 100 was assigned, and a zero-mean Gaussian prior distribution with a Matern covariance structure was assumed for the spatial effect (for more details about the spatial component, see Munoz et al., 2013). Finally, for the temporal effect, a LogGamma prior distribution with the parameters of shape and scale equal to 1 and 5x10®, respectively, was assumed for the log-precision parameter Aj, and j represented the year. For each parameter, a posterior distribution was obtained. Unlike the mean and confidence interval produced by classical analyses, this type of distribu- tion enables explicit probability statements about the parameter. Therefore, the region bounded by the 0.025 and 0.975 quantiles of the posterior distribution results in an intuitive interpretation: for a specific model, the unknown parameter is 95% likely to fail within this range of values (95% credibility interval [CrI]). All models obtained by combining environmental, spatial, and temporal variables and the possible inter- actions were fitted and compared by using the mea- sures of the deviance information criterion (DIG) (Spie- gelhalter et al., 2002) and the cross-validated logarith- mic score (LCPO) (Roos and Held, 2011). Specifically, smaller DIG and LCPO values indicate better fit and predictive quality. All the analyses were performed with the integrated nested lapiace approximation (INLA) method that is implemented in Rue et a!., 2009; Martins et al., 2013 and with the R-INLA package (website) in R software. Model validation Two approaches were used to assess the predictive ac- curacy of the selected model. First, the predicted and observed values from the full data set were compared. 94 Fishery Bulletin 115(1) Table 2 Summary of the posterior distributions of the fixed effects for the best model of distribution of spiny dogfish {Squalus acanthias), by sex. This summary contains the mean, standard deviation (SD), median (Qo s). and 95% credible interval (Q0.025 to Q0.975). the latter of which is a central interval contain- ing the 95% of the probability under the posterior distribution. SST=sea-surface temperature; chl- a=chlorophyll-a concentration. Sex Predictors of distribution Mean SD Qo.025 Qo.5 Qo.975 Females Intercept 0.55 0.94 0.03 0.62 1.22 Bathymetry -1.14 0.05 -2.52 -1.05 -0.22 SST 0.35 0.03 0.08 0.30 1.98 Chl-a 0.25 0.02 0.04 0.19 1.05 Salinity -0.94 0.04 -1.65 -0.92 -0.11 Season (spring) 1.45 0.05 0.34 1.35 2.06 Time (night) -0.65 0.03 -1.24 -0.57 -0.12 Time (morning) 0.88 0.03 0.32 0.67 1.43 Males Intercept 0.88 0.12 0.24 0.86 1.87 Bathymetry 1.85 0.05 1.44 1.83 2.36 SST -1.02 0.08 -2.33 -0.98 -1.51 Chl-a -0.15 0.01 -1.56 -0.18 -0.02 Salinity 0.65 0.03 0.23 0.59 1.43 Season (spring) -0.37 0.02 -2.24 -0.32 -0.02 Time (morning) -0.43 0.06 -1.52 -0.35 -0.04 Time (night) -0.25 0.01 -1.09 -0.23 -0.02 Second, a 50-fold cross-validation based on a random half of the data set was performed to build the model, and the remaining data were used to test the predic- tion (Fielding and Bell, 1997). For both of these approaches, 3 statistics were cal- culated: Pearson’s r, root mean square error (RMSE), and the average error (avg. error). Pearson’s r measures the linear dependence between predicted and observed values. It can vary from -1 to 1, with 1 representing a perfect positive correlation between the 2 data sets. The RMSE represents the standard error of the differ- ences between predicted values and observed values, and the avg. error represents the mean error between observed and predicted values. The closer these 2 sta- tistics are to zero, the better the prediction (Potts and Elith, 2006). Results Initial results During 1820 trawl hauls, 2372 adult spiny dogfish were caught, of which 2252 were females and 120 males. Seasonally, 2085 females were caught in spring and 167 females were captured in fall, and 64 and 56 males were caught in spring and in fall, respectively. The final model with the best fit (based on the low- est DIG and LCPO) for CPUE for each sex includes as relevant covariates bathymetry, SST, salinity, chl-a con- centration, season and time of the survey, and the ran- dom spatial effect for both sexes (Table 2). Slope of the seabed and distance from the coast were not relevant variables for the model of CPUE of spiny dogfish for both sexes. No relevant interannual differences were found in CPUE variability for both the females and males in the sampling area because all models with the yearly temporal effect resulted in higher DIC than those DIC from models without it. Females Results for adult females indicated that CPUE had a negative relationship with bathymetry (posterior mean: -1.14; CrI: -2.52 to -0.22), and salinity (poste- rior mean: -0.94; CrI: -1.65 to -0.11); therefore, higher values for CPUE of females were predicted to be found in shallow, less saline waters compared with CPUE lev- els predicted for other depths and more saline waters. Conversely, values of SST (posterior mean: 0.35; CrI: 0.08 to 1.98) and chl-a concentration (posterior mean: 0.25; CrI: 0.04 to 1.05) indicated a positive relationship with the expected CPUE, indicating that higher CPUE should be expected in warmer waters with higher pri- mary productivity (i.e., higher concentrations of chl-a). Additionally, there was a seasonal effect for the esti- mated probability of CPUE of adult females; the high- est estimated value occurred in spring (posterior mean: 1.45; CrI: 0.34 to 2.06) compared with the reference level (fall season). Dell'Apa et al.; Modeling the distribution of Squalus acanthias, by sex 95 B 40° N 35° N 80° W 70' W Figure 2 Map of the median of the posterior estimates of the probability of the catch per unit of effort of the mature (A) female and (B) male spiny dogfish (Squalus acanthias) caught during fishery-independent sur- veys conducted between New England and North Carolina in 2007- 2013. Values of CPUE, which ranged from 16 to 0 for females and from 6 to 0 for males, were used as a proxy for relative abundance. Morning was the period of the day during a sur- vey with the highest estimated probability of CPUE of adult females (posterior mean: 0.88; CrI: 0.32 to 1.43) with respect to the reference level (afternoon). Conversely, hauls conducted at night had the lowest estimated CPUE (posterior mean: -0.65; CrI: -1.24 to -0.12). Finally, higher estimated values of CPUE for adult females were found for waters off the southern area of the Cape Cod peninsula and off Georges Bank than for CPUE estimates for other areas sampled (Fig. 2A). Males For adult males, a positive relationship between ba- thymetry and CPUE (posterior mean: 1.85; CrI: 1.44 to 2.36) was observed, as well as for salinity and CPUE (posterior mean: 0.65; CrI: 0.23 to 1.43). Predicted val- ues from the model indicated a negative relationship for adult males between expected CPUE and SST (pos- terior mean: -1.02; CrI: -2.33 to -1.51) and between ex- pected CPUE and chl-a concentration (posterior mean: -0.15; CrI: -1.56 to -0.02). We interpret this result as indicating a higher probability of catching adult males as depth and salinity increase and water temperature and chl-a concentration decrease. The estimated probability of CPUE for adult males indicated a seasonal effect, with lower values in spring (posterior mean: -0.37; CrI: -2.24 to -0.02) than in fall, the season of the reference level. Surveys conducted in the morning had the lowest estimated CPUE of adult males (posterior mean: -0.43; 96 Fishery Bulletin 115(1) CrI: -1.52 to -0.04) compared with surveys conducted in the afternoon (the reference level), which was the period of time with the highest CPUE. Night-time sur- veys also had lower values of CPUE (posterior mean = -0.25; CrI: -1.09 to -0.02) compared with the values from afternoon surveys. Higher estimated values of CPUE for adult males occurred at higher latitudes, above 40°N, and off Georges Bank, than estimated values of CPUE from other sampled areas (Fig. 2B). Model performance For model validation, reasonably high values for Pear- son’s r were obtained for both sexes. In particular, from the model for adult females, an r value of -0.65 was obtained in the cross-validation with the original data set, and an r value of -0.71 was obtained in the cross- validation with half of the data set. For adult males, the r value was -0.68 in the cross-validation with the original data set and was -0.74 in the cross-validation with half of the data set. Low values of RMSE and avg. error were achieved for adult females, with an RMSE of 0.98 and an avg. error of 0.045 in the cross-validation with the original data set and with an RMSE of 1.14 and an avg. er- ror of -0.023 in the cross-validation with half of the data set (Table 3). For adult males, low values were also observed, with an RMSE of 1.15 and an avg. error of 0.032 in cross-validation with the original data set and an RMSE of 1.11 and an avg. error of -0.018 in the cross-validation with half of the data set (Table 3). These validation results indicate a good performance of the 2 models. Discussion This study provides predictive information on the habi- tat distribution of spiny dogfish in US. Atlantic coastal waters, by modeling the CPUE in the NEAMAP survey as a proxy for the abundance of this species. The re- sults of this study offer insight into the key environ- mental and temporal variables that influence the habi- tat selection for each of the sexes of this species. For our modeling approach, we assumed CPUE is a proxy for species abundance and that CPUE should be the same for both female and male spiny dogfish, although CPUE is not always a viable proxy for species abun- dance and is not always the same for each sex because of inherent variability in the catchability coefficient (see Hilborn and Walters, 1992; Walters, 2003; Maun- der and Punt, 2004; Maunder et al., 2006). However, the lack of linearity between CPUE and fish abundance is largely reduced when data from standardized fish- ery-independent surveys are used (Maunder and Punt, 2004), as was done in our study. Another confounding issue in our analysis was that in the NEAMAP survey, female spiny dogfish mark- edly outnumbered males, by a ratio of 20:1. Ideally, the Table 3 Statistics used in the hierarchical Bayesian spatiotem- poral model to estimate abundance of spiny dogfish iSqualus acanthias), by sex. The statistics include root mean square error (RMSE) with the original data set, average error (avg. error) with the original data set, root mean square error (RMSE-cross) with half of the data set, and average error (avg. error-cross) with half of the data set. RMSE- avg. error Sex RMSE avg. error cross -cross Eemales 0.98 0.045 1.14 -0.023 Males 1.15 0.032 1.11 -0.018 analysis of data from a simultaneous offshore survey would have been more suitable for determining wheth- er patterns inshore and offshore are the same, but such data were not available. This lack of information points to a need for concurrent offshore sampling to support the extrapolation of our conclusions to the offshore component of and for the generalization of our predic- tions for the entire US. Atlantic stock of spiny dogfish. Our results related to sex-based segregation and distribution are in line with those reported in other studies. As expected, a higher number of adult females were caught in shallower, inshore, and warmer waters in comparison with adult males, which are more com- monly reported in deeper, offshore, and colder waters (Shepherd et al., 2002; Methratta and Link, 2007; Sagarese et al., 2014a). On the basis of the results of the hierarchical Bayes- ian spatiotemporal model, abundance of adult female spiny dogfish is predicted to decrease with depth, but abundance of adult males is predicted to increase with depth. In addition, CPUE of adult females is predicted to be higher in warmer, less saline waters and in wa- ters with higher concentrations of chl-a, but CPUE of adult males is predicted to be higher in colder, more sa- line waters, and in waters with lower concentrations of chl-a. Nevertheless, adult females occasionally move to deeper waters, mainly in the spring. More studies are needed to understand the drivers for these movements into deeper waters. However, considering that partu- rition in spiny dogfish is likely to occur in offshore, deeper areas off the edge of the continental shelf and in deep basins (Jensen, 1966; Nammack et al., 1985; Hanchet, 1988; Campana et al., 2009) and that the time of parturition is commonly reported to be between November and January in the northwest Atlantic popu- | lation (Nammack et al., 1985), despite the observation in a recent study of neonate in more inshore waters off Rhode Island in February (Sulikowski et al., 2013), it is likely that these offshore movements of adult females in the spring are not related to parturition events but might be the result of a strategy to avoid males and Dell'Apa et al.: Modeling the distribution of Squalus acanthias, by sex 97 energy-demanding copulation (Pratt and Carrier, 2001; Sims et al., 2001; Dell’Apa et al., 2014, 2015). More studies are needed to test this hypothesis. That the CPUE predicted in our analysis for adult females was higher in warmer waters than in less warm waters is in accordance with results of Sagarese et al. (2014a), a study in which adult females in the northeastern U.S. continental shelf large marine eco- system were observed to occupy significantly warmer waters than those occupied by adult males. In regard to adult male abundance, the results of the Bayesian model obtained from our study are also in agreement with results from Sagarese et al. (2014a): results from both studies indicate a higher occurrence of adult males than adult females in deeper, colder waters. In spiny dogfish, gestation lasts for almost 2 years, and it has been hypothesized that this species may have an evolutionary advantage in that adult, pregnant fe- males actively seek inshore, warmer waters because such habitats can maximize the growth rates of em- bryos (Sagarese et al., 2014a). Salinity was found to be a relevant environmental variable for predicting the habitat distribution of spiny dogfish for both sexes — a finding that is in agreement with Shepard et al. (2002) and Sagarese et al. (2014a), but we interpret this result and the results of other studies as a possible artifact of the association of salin- ity with other key environmental habitat characteris- tics, such as depth and temperature, that better define habitat selection of spiny dogfish. Generally, inshore waters are less saline than offshore environments be- cause of the increased contribution of freshwater runoff. On the basis of the Bayesian model, we predict that abundance of adult females should be higher in the spring than in the fall and higher in the morning than in the afternoon and night. Conversely, abundance of adult males is predicted to be higher in the fall than in the spring, as well as higher in the afternoon than at other times of day. For adult females, this predicted seasonal pattern of abundance is in agreement with re- sults from the NEFSC survey, which is conducted far- ther offshore than the NEAMAP survey (Sagarese et al., 2014b). This pattern also indicates that the habitat distribution of adult females in the spring is mainly driven by environmental factors (i.e., temperature and depth) and that habitat distribution in the fall is main- ly influenced by ecological factors (i.e., prey abundance) (Sagarese et al., 2014a, 2014b). For adult males, the results of this study contrast with those reported by Sagarese et al. (2014b), in that we predicted a greater abundance of males in the fall than in the spring. We acknowledge that our analysis is based on fewer males than females, but we are confi- dent that the level of uncertainty that may result from this limited number is included within the parameters analyzed by a Bayesian approach. In combination, how- ever, our results and those of Sagarese et al. (2014b), for inshore and offshore waters, respectively, indicate that abundance of adult males should be higher in the fall than in the spring in inshore coastal waters, al- though a greater chance for catching larger aggrega- tions of adult males may occur in offshore, deeper wa- ters in the spring. More studies are needed to support this hypothesis and to analyze specific seasonal differ- ences in habitat distribution and abundance of spiny dogfish, by sex, throughout its range. Unfortunately, it was not possible to run the hierarchical Bayesian spa- tiotemporal model and develop probability distribution maps for each sex, by season, as with the maps ob- tained for CPUE for each sex, because of the consistent presence in the survey of locations that were sampled multiple times. This study is the first one to provide a predictive model for the spatial distribution of each sex of spiny dogfish on the basis of time of day, which previously has been suggested to be an important aspect as- sociated with the sexual segregation of this species (Dell’Apa et al., 2014). Such information could be used to enhance fishery sustainability by developing fishery strategies based on the distribution and habitat associ- ation of each of the sexes. For example, targeting adult females inshore, in shallower waters, would result in greater pressure on the entire U.S. Atlantic stock of spiny dogfish. Fishery sustainability is measured in terms of fishing mortality and spawning stock biomass, and adult female biomass is used as a biological refer- ence point. Although the spiny dogfish is not currently considered overfished and overfishing is not occurring, low numbers of adult females are expected to be re- cruiting to the stock over the next few years (Rago and Sosebee^; Dell’Apa et al., 2015). Considering the concentration of chl-a as a valid proxy for primary production (de Leiva Moreno et al., 2000), adult females were predicted to associate with habitats characterized by higher productivity, but adult males were predicted to occur in less productive waters. As with our analysis of results for salinity, we interpreted this result for chl-a concentration as the direct association of primary productivity with the preferred habitat for females rather than as a driver per se. Because adult females are found most com- monly inshore, in shallower and warmer waters where primary production is usually highest, it is likely that chl-a concentration and primary production should not be considered as an important predictive variable for abundance of spiny dogfish. However, for coastal elasmobranch species, such as the Atlantic sharpnose shark {Rhizoprionodon terraenouae), chl-a concentra- tion has been found to be an important environmen- tal variable that affects the abundance and habitat association of that species (Drymon et al., 2013). Pa- pastamatiou et al. (2013) reported that for both sexes of tiger shark (Galeocerdo cuvier), immigration across the Hawaiian Archipelago was correlated with chl-a concentration — a result that they interpreted to be in- dicative of a foraging activity. It is, therefore, likely that for large pelagic shark species, such as the blue shark (Prionace glauca), primary production is an im- portant variable that influences species habitat distri- bution (Mitchell et al., 2014). 98 Fishery Bulletin 115(1) Our results indicate that adult female spiny dogfish may select warmer inshore habitats not for foraging but to reduce gestation time of embryos, as has been reported to occur in the round stingray (Urobatis hal- leri) and may be common in other elasmobranchs (Jirik and Lowe, 2012). Further research should be focused on understanding the influence of primary productivity on habitat selection in spiny dogflsh, by sex, to evalu- ate the link with gestation and the possible link with foraging success (Dell’Apa et ah, 2015). From a management perspective, the results of this study indicate a higher probability of catching spiny dogfish at higher latitudes, mainly on the Georges Bank shelf area, where the stock of this species is com- monly reported to aggregate between spring and fall (Rulifson, 2010). Results also indicate that there is a higher probability of adult male aggregations inshore in coastal waters during the fall. Additionally, a lower abundance of adult females was predicted to be found in the fall in these same waters. Combined, these re- sults indicate that the Georges Bank shelf area may be explored for the development of a male-only directed fishery in the fall. However, further studies are needed to confirm this hypothesis, studies that possibly could combine data analyses of both fishery-independent (e.g., NEFSC surveys) and fishery-dependent surveys in the New England region. Additionally, the results from this study indicate that a potential male-only directed fishery, mainly in the Georges Bank shelf area, could be limited to fishing in the afternoon, when fewer aggregations of adult females are predicted to be associated with coastal waters, and in the fall, when higher numbers of adult females are predicted to select deeper waters. Sagarese et al. (2014a) noted that during the fall, adult females may move into offshore waters as a strategy to avoid encountering adult males — a suggestion also made by other authors (Veris- simo et al., 2011; Dell’Apa et al., 2014). Future research should be undertaken to compare and integrate the spa- tiotemporal habitat distribution of spiny dogfish, by sex, found in this study with that of fishery-independent surveys, such as the surveys that the NEFSC has con- ducted in offshore waters and for a longer period of time, and with that of fishery-dependent surveys, which provide less standardized but more direct information on the influence of fishery exploitation of this species be- tween New England and North Carolina. Because tem- perature and depth are key environmental variables for prediction of the abundance in both sexes, more studies are also needed to understand the influence of climate change in the spatial distribution of spiny dogfish along the northwest Atlantic coast and continental shelf (Nye et al., 2009; Sagarese et al., 2014a). Acknowledgments We thank all the people who have worked on the FA^ Darana R at the Virginia Institute of Marine Sci- ence in Gloucester Point, Virginia, from crew members to those people who volunteered to help collect data and samples during the surveys. We are grateful to S. Sagarese and to 3 anonymous reviewers for their valu- * able inputs and comments that helped to improve the quality of this manuscript. Literature cited Austin, M. 2007. Species distribution models and ecological theory: a critical assessment and some possible new approaches. Ecol. Model. 200:1-19. Banerjee, S., B. P. Carlin, and A. E. Gelfand. 2004. Hierarchical modeling and analysis for spatial data, 474 p. Chapman & Hall/CRC, Boca Raton, FL. Bubley, W. J., J. A. Sulikowski, D. M. Koester, and P. C. W. Tsang. 2013. Using a multi-parameter approach to reassess maturity of spiny dogfish, Squalus acanthias, following increased fishing pressure in the Western North Atlan- tic. Fish. Res. 147:202-212. Campana, S. E., D. W. Kulka, and S. Joyce. 2009. Growth and reproduction of spiny dogfish off the eastern coast of Canada, including inferences on stock structure. In Biology and management of dogfish sharks (V. F. Gallucci, G. A. McFarlane, and G. G. Barg- mann, eds.), p. 195-208. Am. Fish. Soc., Bethesda, MD. Colloca, F., V. Bartolino, G. J. Lasinio, L. Maiorano, P. Sartor, and G. Ardizzone. 2009. Identifying fish nurseries using density and persis- tence measures. Mar. Ecol. Prog. Ser. 381:287-296. Compagno, L., M. Dando, and S. Fowler. 2005. Princeton field guides: sharks of the world, 496 p. | Princeton Univ. Press, Princeton, NJ. Conradt, L. 2005. Definitions, hypotheses, models, and measures in the study of animal segregation. In Sexual segregation in vertebrates: ecology of the two sexes (K. E. Ruckstuhl and P. Neuhaus, eds.), p. 11-32. Cambridge Univ. Press, Cambridge, UK. de Leiva Moreno, J. I., V. N. Agostini, J. F. Caddy, and F. Carocci. 2000. Is the pelagic-demersal ratio from fishery landings a useful proxy for nutrient availability? A preliminary data exploration for the semi-enclosed seas around Eu- rope. ICES J. Mar. Sci. 57:1091-1102. DelFApa, A., J. C. Johnson, D. G. Kimmel, R. A. Rulifson. 2013. The international trade and fishery management of spiny dogfish: a social network approach. Ocean Coast. Manage. 80:65-72. DelFApa, A., J. Cudney-Burch, D. G. Kimmel, and R. A. Rulifson. 2014. Sexual segregation of spiny dogfish in fishery- dependent surveys in Cape Cod, Massachusetts: po- tential management benefits. Trans. Am. Fish. Soc. 143:833-844. DelFApa, A., C. W. Bangley, and R. A. Rulifson. 2015. Who let the dogfish out? A review of management and socio-economic aspects of spiny dogfish fisher- ies. Rev. Fish Biol. Fish. 25:273-295. Drymon, J. M., L. Carassou, S. P. Powers, M. Grace, J. Dindo, and B. Dzwonkowski. 2013. Multiscale analysis of factors that affect the distri- Dell'Apa et al.: Modeling the distribution of Squalus acanthias, by sex 99 bution of sharks throughout the northern Gulf of Mexi- co. Fish. Bull. 111:370-380. Elith, J., and J. R. Leathwick. 2009. Species distribution models: ecological explanation and prediction across space and time. Annu. Rev. Ecol. Evol. Syst. 40:677-697. Espinoza, M., M. Cappo, M. R. Heupel, A. J. Tobin, and C. A. Simpfendorfer. 2014. Quantifying shark distribution patterns and spe- cies-habitat associations: implications of marine park zoning. PLoS One 9(9):el06885. Fielding, A. H., and J. F. Bell. 1997. A review of methods for the assessment of predic- tion errors in conservation presence/absence models. Environ. Conserv. 24:38-49. Ford, E. 1921. A contribution to our knowledge of the life-histories of the dogfishes landed at Plymouth. J. Mar. Biol. Assoc. U. K. 12:468-505. Gelfand, A. E., J. A. Silander Jr., S. Wu, A. Latimer, P. O. Lew- is, A. G. Rebelo, and M. Holder. 2006. Explaining species distribution patterns through hi- erarchical modeling. Bayesian Anal. 1:41-92. Hanchet, S. 1988. Reproductive biology of Squalus acanthias from the east coast, South Island, New Zealand. N.Z. J. Mar. Freshw. Res. 22:537-549. Hijmans, R. J. 2013. Introduction to the “raster” package (version 2.1- 25). R Foundation for Statistical Computing, Vienna, Austria. Hilborn, R., and C. J. Walters (eds). 1992. Quantitative fisheries stock assessment: choice, dy- namics and uncertainty, 570 p. Chapman & Hall, New York. Jensen, A. C. 1966. Life history of the spiny dogfish. Fish. Bull. 65:527-554. Jirik, K. E., and C. G. Lowe. 2012. An elasmobranch maternity ward: female round stingrays Urobatis halleri use warm, restored estuarine habitat during gestation. J. Fish Biol. 80:1227-1245. Lack, M. 2006. Conservation of spiny dogfish Squalus acanthias: a role for CITES?, 40 p. TRAFFIC International, Cam- bridge, UK. [Available from website.] Martins, T. G., D. Simpson, F. Lindgren, and H. Rue. 2013. Bayesian computing with INLA: new features. Comput. Stat. Data Anal. 67:68-83. Maunder, M. N., and A. E. Punt. 2004. Standardizing catch and effort data: a review of re- cent approaches. Fish. Res. 70:141-159. Maunder, M. N., J. R. Sibert, A. Fonteneau, J. Hampton, P. Kleiber, and S. J. Harley. 2006. Interpreting catch per unit effort data to assess the status of individual stocks and communities. ICES J. Mar. Sci. 63:1373-1385. Methratta, E. T., and J. S. Link. 2007. Ontogenetic variation in habitat association for four groundfish species in the Gulf of Maine-Georges Bank region. Mar. Ecol. Prog. Ser. 338:169-181. Mitchell, J. D., K. J. Collins, P. 1. Miller, and L. A. Suberg. 2014. Quantifying the impact of environmental variables upon catch per unit effort of the blue shark Prionace glauca in the western English Channel. J. Fish Biol. 85:657-670. Munoz, F., M. G. Pennino, D. Conesa, A. Lopez-Quilez, and J. M. Bellido. 2013. Estimation and prediction of the spatial occurrence of fish species using Bayesian latent Gaussian mod- els. Stoch. Environ. Res. Risk Assess. 27:1171-1180. Nammack, M. F., J. A. Musick, and J. A. Colvocoresses. 1985. Life history of spiny dogfish off the northeastern United States. Trans. Am. Fish. Soc. 114:367-376. Nye, J. A., J. S. Link, J. A. Hare, and W. J. Overholtz. 2009. Changing spatial distribution of fish stocks in rela- tion to climate and population size on the Northeast United States continental shelf Mar. Ecol. Prog. Ser. 393:111- 129. Overholtz, W. J., and A. V. Tyler. 1985. Long-term responses of the demersal fish assem- blages of Georges Bank. Fish. Bull. 83:507-520. Papastamatiou, Y. P., C. G. Meyer, F. Carvalho, J. J. Dale, M. R. Hutchinson, and K. N. Holland. 2013. Telemetry and random-walk models reveal complex patterns of partial migration in a large marine predator. Ecology 94:2595-2606. Pennino, M. G., F. Munoz, D. Conesa, A. Lopez-Quilez, and J. M. Bellido. 2014. Bayesian spatio-temporal discard model in a demer- sal trawl fishery. J. Sea Res. 90:44-53. Perry, R. L, and S. J. Smith. 1994. Identifying habitat associations of marine fishes us- ing survey data: an application to the northwest Atlan- tic. Can. J. Fish. Aquat. Sci. 51:589-602. Pittman, S. J., and K. A. Brown. 2011. Multi-scale approach for predicting fish species dis- tributions across coral reef seascapes. PLoS One 6(5): 620583. Potts, J. M., and J. Elith. 2006. Comparing species abundance models. Ecol. Mod- el. 199:153-163. Pratt, H. L., Jr., and J. C. Carrier. 2001. A review of elasmobranch reproductive behavior with a case study on the nurse shark, Ginglymostoma cirratum. Environ. Biol. Fish. 60(1):157-188. R Core Team. 2014. R: a language and environment for statistical com- puting. R Foundation for Statistical Computing, Vien- na, Austria. [Available from website, accessed August 2014.] Rods, M., and L. Held. 2011. Sensitivity analysis in Bayesian generalized linear mixed models for binary data. Bayesian Anal. 6:259- 278. [Available from website.] Roos, N. C., A. R. Carvalho, P. F. Lopes, and M. G. Pennino. 2015. Modeling sensitive parrotfish (Labridae: Scarini) habitats along the Brazilian coast. Mar. Environ. Res. 110:92-100. Ruckstuhl, K. E., and T. H. Clutton-Brock. 2005. Sexual segregation and the ecology of the two sexes. In Sexual segregation in vertebrates: ecology of the two sexes (K. E. Ruckstuhl and P. Neuhaus, eds.), p. 3-10. Cambridge Univ. Press, Cambridge, UK. Rue, H., S. Martino, and N. Chopin. 2009. Approximate Bayesian inference for latent Gauss- ian models by using integrated nested Laplace approxi- mations. J. R. Stat. Soc., B 71:319-392. 100 Fishery Bulletin 115(1) Rulifson, R. A. 2010. Tagging spiny dogfish overwintering in North Caro- lina, and summering in Bay of Fundy, Canada. In Pro- ceedings of the Transboundary Resources Assessment Committee (TRAC) spiny dogfish review; Woods Hole, MA, 30 March-2 April 2009, p. 15-19. [Available from website.] Sagarese, S. R., M. G. Frisk, T. J. Miller, K. A. Sosebee, J. A. Musick, and P. J. Rago. 2014a. Influence of environmental, spatial, and ontoge- netic variables on habitat selection and management of spiny dogfish in the Northeast (US) shelf large marine ecosystem. Can. J. Fish. Aquat. Sci. 71:567-580. Sagarese, S. R., M. G. Frisk, R. M. Cerrato, K. A. Sosebee, J. A. Musick, and P. J. Rago. 2014b. Application of generalized additive models to ex- amine ontogenetic and seasonal distributions of spiny dogfish {Squalus acanthias) in the Northeast (US) shelf large marine ecosystem. Can. J. Fish. Aquat. Sci. 71:847-877. Shepherd, T, F. Page, and B. Macdonald. 2002. Length and sex-specific associations between spiny dogfish {Squalus acanthias) and hydrographic variables in the Bay of Fundy and Scotian Shelf. Fish. Oceanogr. 11:78-89. Sims, D. W. 2005. Differences in habitat selection and reproductive strategies of male and female sharks. In Sexual seg- regation in vertebrates: ecology of the two sexes (K. E. Ruckstuhl and P. Neuhaus, eds.), p. 127-147. Cam- bridge Univ. Press, Cambridge, UK. Sims, D. W., J. P. Nash, and D. Morritt. 2001 Movements and activity of male and female dogfish in a tidal sea lough: alternative behavioural strategies and apparent sexual segregation. Mar. Biol. 139:1165- 1175. Sosebee, K. A. 2005. Are density-dependent effects on elasmobranch ma- turity possible? J. Northwest Atl. Fish. Sci. 35:115-124. Spiegelhalter, D. J., N. G. Best, B. P. Carlin, and A. van der Linde. 2002. Bayesian measures of model complexity and fit. J. R. Stat. Soc., B. 64:583-639. Springer, S. 1967. Social organization of shark populations. In Sharks, skates, and rays (P. W. Gilbert, R. F. Mathewson, and D. P. Rail, eds.), p. 149-174. The Johns Hopkins Press, Baltimore, MD. Sulikowski, J. A., B. K. Prohaska, A. E. Carlson, A. M. Cicia, C. T. Brown, and A. C. Morgan. 2013. Observations of neonate spiny dogfish, Squalus acanthias, in Southern New England: a first account of a potential pupping ground in the Northwestern Atlan- tic. Fish. Res. 137: 59-62. Verfssimo, A., D. Grubbs, J. McDowell, J. Musick, and D. Portnoy. 2011. Frequency of multiple paternity in the spiny dogfish Squalus acanthias in the Western North Atlantic. J. Hered. 102:88-93. Walters, C. 2003. Folly and fantasy in the analysis of spatial catch rate data. Can. J. Fish. Aquat. Sci. 60:1433-1436. Wearmouth, V. J., and D. W. Sims. 2008. Sexual segregation in marine fish, reptiles, birds and mammals: behaviour patterns, mechanisms and con- servation implications. Adv. Mar. Biol. 54:107-170. 101 National Marine Spencer F. Baird V Fisheries Service Fishery Bulletin First U.S. Commissioner S NOAA fy- established in 1881 of Fisheries and founder mSjK of Fishery Bulletin iMy Horizontal and ¥ertical movements of longfin makos itsurus paucm} tracked with satellite- linked tags in the northwestern Atlantic Ocean Robert E. Hueter (contact author)’ John P. Tyminski’ John J. Morris’ Alexei RuIe Abierno^ Jorge Angulo Valdes^'® Email address for contact author: rhueter@mote.org ' Center for Shark Research Mote Marine Laboratory 1600 Ken Thompson Parkway Sarasota, Florida 34236 2 Centro de Investigaciones Marinas Universidad de la Habana Calle 16, No. 114 e/ Ira y 3ra Miramar, Playa, La Habana CP 11300, Cuba 3 School of Natural Resources and Environment University of Florida P.O. Box 116455 Gainesville, Florida 32611 Abstract— The longfin mako (Isurus paucus) is a poorly studied oceanic shark taken in fisheries throughout its worldwide range in temperate and tropical waters. Satellite-linked tags were deployed to investigate the movements of 2 mature males, one tagged in the northeastern Gulf of Mexico (GOM) and the other off northern Cuba. Horizontal tracks estimated by using likelihood meth- ods were similar for these sharks; comparable movements were docu- mented from the GOM, through the Straits of Florida and the Ba- hamas, and into the open Atlantic Ocean where they converged on the Mid-Atlantic Bight. Depth and tem- perature ranges were 0-1767 m and 4.0-28.8°C. A diel pattern of vertical movement was evident for both in- dividuals, along with regular forays from cold daytime depths to warmer near-surface waters, possibly as an adaptation for thermoregulation. The vertical movements of longfin makos allow them to exploit verti- cally migrating prey but these move- ments increase their vulnerability to pelagic longlining. The horizontal movements of these sharks reveal the limited benefit of areas current- ly closed to pelagic longlining off the southeastern United States and also indicate the connectivity of U.S., Cuban, Mexican, and Bahamian wa- ters for this species. Because of the vulnerability of longfin makos to overexploitation, improved biological information is needed for accurate stock assessments and appropri- ate management and conservation measures. Manuscript submitted 15 April 2016. Manuscript accepted 29 November 2016. Fish. Bull.: 115:101-116 (2017). Online publication date: 20 December 2016. doi: 10.7755/FB.115.1.9 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. The shark genus Isurus comprises 2 species with similar morphologi- cal features, the shortfin mako (I. oxyrinchus) and the longfin mako (I. paucus) (Compagno, 2001). The longfin mako is a global oceanic spe- cies that inhabits both tropical and subtropical waters (Compagno, 2001) but also ranges into temperate seas (Queiroz et ah, 2006; Bustamante et ah, 2009; Mucientes et ah, 2013). In the western Atlantic Ocean, the long- fin mako has been documented in the Gulf Stream off the eastern United States, off Cuba’s northern coast, as far south as southern Brazil (Com- pagno, 2001), and as far north as Georges Bank (Kohler et ah, 1998). In the Gulf of Mexico (GOM), long- fin makos have been observed in- frequently but have been reported in the northern GOM off the Mis- sissippi River and south of Panama City, Florida (Killam and Parsons, 1986) and in the southern GOM off Tabasco, Mexico (Wakida-Kusunoki and de Anda-Fuente, 2012). Despite its large size (to at least 417 cm in total length [TL]; Gilmore, 1983), the longfin mako is an understudied species, partly because of its relative rarity, its pelagic nature, and observ- ers’ confusion with its congener, the shortfin mako (Castro, 2011). The longfin mako is not directly targeted in any fishery, but is taken as bycatch throughout its range in tropical and temperate waters by pelagic longline fisheries that tar- get swordfish {Xiphias gladius), tuna species {Thunnus spp.), and other shark species (Reardon et ah, 2006; Mucientes et ah, 2013; Fredou et ah, 2015). In a study of shark bycatch in the small-scale, pelagic longline fish- ery of northwest Cuba, Guitart Man- day (1975) reported that the longfin mako was the sixth most common shark by weight of the 11 shark spe- cies reported; a more recent study of this fishery ranked the longfin mako higher in landings (Aguilar et ah, 102 Fishery Bulletin 115(1) 2014). Fins of longfin makos are of desirable quality and have been reported in the Hong Kong (Clarke et al., 2006), Chilean (Sebastian et ah, 2008), and Indone- sian fin trades (Sembiring et ah, 2015). Consequently, in some fisheries but not in the Cuban fishery, this spe- cies may be finned and discarded at sea; hence, land- ings of longfin makos may be underreported (Reardon et ah, 2006). Given the apparent declines in some populations of the shortfin mako (Baum et ah, 2003; Dulvy et ah, 2008), it is likely that populations of the longfin mako have been affected by intensive pelagic longline fisheries (Reardon et ah, 2006). Because of its rarity, low reproductive potential, and bycatch-induced mortality, the longfin mako is listed as vulnerable in the lUCN Red List of Threatened Species (Reardon et ah, 2006) and in 2008 was added to Appendix II of the Convention on the Conservation of Migratory Species of Wild Animals (Kyne et ah, 2012). In U.S. waters, reten- tion of longfin makos has been prohibited for both com- mercial and recreational fishermen since 2000 under the National Marine Fisheries Service (NMFS) fishery management plan for sharks that inhabit the Atlantic Ocean and adjacent waters (NMFS^). In an ecological risk assessment of shark species caught in Atlantic pelagic longline fisheries, Cortes et ah (2015) ranked the vulnerability of the longfin mako among species at highest risk and highlighted the need for better basic biological information for this shark. Satellite-linked tagging technologies have provided researchers with effective tools for revealing home ranges, movement and migration routes, and habitat- use patterns of marine predators (Hammerschlag et ah, 2011). Most lamnid species have been studied by using satellite tracking. These species include the shortfin mako (Loefer et ah, 2005; Stevens et ah, 2010; Rog- ers et ah, 2015), white shark {Carcharodon carcharias) (Bruce et ah, 2006; Weng et ah, 2007; Nasby-Lucas et ah, 2009), porbeagle (Lamna nasus) (Fade et ah, 2009; Saunders et ah, 2011; Francis et ah, 2015), and salmon shark (L. ditropis) (Weng et ah, 2005, 2008). However, there are no detailed reports of satellite-tracked long- fin makos from any part of the global range of this species. Conventional tagging results in U.S. waters, although sparse, indicate movement of longfin makos from the eastern GOM into the western North Atlantic Ocean, likely through the Straits of Florida (Kohler et ah, 1998). We are the first to use satellite tracking as a means of assessing the behavior, ecology, and vulner- ability to fisheries of this species. Materials and methods Two specimens of the longfin mako, one from the northeastern GOM and one from the southeastern ^ NMFS (National Marine Fisheries Service). 1999. Fi- nal fishery management plan for Atlantic tuna, swordfish, and sharks, 97 p. Nath Mar. Fish. Serv., Silver Spring, MD. [Available from website.] GOM, were captured and tagged with pop-up satel- lite archival tags to track their horizontal and vertical movements. In 2012, a male longfin mako (LFMl) was . captured during an overnight pelagic longline set de- ployed on 27 April in the northeastern GOM (28.40°N, 85.84°W) from the RV Weatherbird II of the Florida Institute of Oceanography. In 2015, a second male long- fin mako (LFM2) was captured during an overnight pelagic longline set deployed from an artisanal Cuban fishing vessel on 13 February off Cojimar in northwest Cuba (23.26°N, 81.98°W). When LFMl was captured in 2012, the sea-surface temperature (SST) was 25.3°C and depth to the bot- , tom was approximately 334 m. The gear targeted pe- lagic fish species and consisted of 26 km of mainline with 30-m gangions composed of 136-kg monofilament connected through a 9/0 nickel-plated swivel to 1 m of 0.8-mm stainless steel cable. The 202 hooks deployed were 18/0 circle hooks with zero offset and were baited with Spanish mackerel (Scomberomorus maculatus) or little tunny {Euthynnus alletteratus) and were sus- pended at depths 30-60 m below the surface. Chemical glow sticks (Chemilures,^ World Plastics, San Carlos, CA) were attached to the gangions approximately 2 m above each baited hook as a fish attractant. Upon haulback of the gear the following morning on 28 April, ; one longfin mako was among the catch. The captured shark was lifted out of the water and brought on deck with a specially designed cradle to support the shark’s body weight (Grace et al., 2007). The animal remained within this cradle for measuring and tagging proce- dures, during which time its gills were irrigated with | seawater from a hose inserted into its mouth. The shark was tagged with a pop-up satellite ar- chival tag (MklO; Wildlife Computers, Redmond, WA). - The tag archived measurements of ambient tempera- ture, pressure, and light level at 3-s intervals and summarized these data into 8-h periods to facilitate data transmission. The tag was programmed to detach after 90 d on the shark, float to the sea surface, and transmit a summary of its archived data by way of the Argos satellite system with time-at-depth and time-at- temperature histograms in 14 user-defined bins. Black antifouling paint (EP-2000; ePaint Company, East Fal- mouth, MA) before deployment had been applied to the tag, excluding its sensors and label. At deployment, the tag was inserted into the shark’s dorsal muscula- ture just below the first dorsal fin by using a stainless steel dart (Type SSD; 34.0 x 8.5 mm; Hallprint Pty. Ltd., Hindmarsh Valley, Australia) attached to a 15-cm tether composed of 55-kg coated, braided wire (Berkley, ' Spirit Lake, lA). To avoid tag destruction from extreme | depth, a mechanical release device (RD1800; Wildlife Computers), designed to release at a depth of 1800 m, was threaded inline at the midpoint of the tether. The tether, excluding the portion with the RD1800 device, 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. Hueter et al.: Horizontal and vertical movements of Isurus paucus in the northwestern Atlantic Ocean 103 was protected with heat shrink tubing (3M, Two Har- bors, MN). A conventional Rototag (Dalton ID Systems Ltd, Henley-on-Thames, UK) and dart tag (Hallprint Pty. Ltd.) also were applied to the shark’s first dorsal fin and dorsal musculature, respectively. The condition of the tagged shark at release was scored as condition 4 (>30 s of revival time, limited swimming on release) according to the release condition categories of Hueter et al. (2006). In 2015, when LFM2 was captured, the SST was 25.7°C and depth to the bottom was approximately 750 m. The gear targeted pelagic fish species, in particu- lar istiophorids and swordfish, and consisted of 8 km of mainline with gangions of 20-30 m, both composed of 2-mm twisted monofilament. The 66 hooks deployed were either 15/0 or 16/0 J hooks baited with salted clupeid and cyprinid fish species (e.g., silver carp [Hy- pophthalmichthys molitrix\), and suspended approxi- mately 40 m below the surface. The fishermen attached battery-powered light-emitting devices (known locally in Cuba as lampos) to the gangions approximately 3 m above each baited hook. During gear retrieval in the early morning of 14 February, the hooked shark was transferred from the fishing vessel to a research boat, secured by tail rope at the stern, and maintained in the water during the process of taking measurements and tagging. This shark (LFM2) was tagged with a MiniPAT (Wildlife Computers) that archived measurements of temperature, pressure, and light level at 5-s intervals and summarized these data into 6-h periods. The tag was programmed to detach from the shark after 150 d, and the time-at-depth and time-at-temperature his- tograms of the data were distributed among 12 user- defined bins. Unlike the MklO tag, which was deployed on LFMl, the MiniPAT sends time-series data of depth and temperature by way of the Argos satellite system in addition to archiving that data. A clear antifouling coating (Propspeed, Propspeed USA, Miami, FL) had been applied to the tag, excluding its sensors and re- lease pin. At deployment, the tag was inserted into the dorsal musculature of the shark just below the first dorsal fin by using a plastic anchor (Domeier; 20x14 mm; Wildlife Computers) attached to a 15-cm tether composed of stainless steel wire rope with a 23-kg load capacity (type 18-8; McMaster-Carr, Santa Fe Springs, CA). A newer version of the RD1800 device was em- ployed with an internal pin that breaks under pres- sure at a depth of 1800 m, releasing the tag from the tether. The tether, excluding the portion of the RD1800 device, was protected with heat shrink tubing (3M). No conventional tag was applied to this shark. The condi- tion of the shark at release was classified as condition 2 (no revival time required, slow but strong swimming on release; Hueter et al., 2006). For both sharks, species identification was deter- mined by the presence of taxonomic characteristics of the genus Isurus (Campagno, 2001), together with a first dorsal fin well behind the free rear tip of the pec- toral fin, dark coloration on the ventral surface of the snout, and long pectoral fins, all of which collectively distinguish the longfin mako from the shortfin mako (Guitart Manday, 1966; Garrick, 1967; Bustamante et al., 2009). Maturity of the 2 males was assessed by stage of clasper development and its condition. Compiled data collected through the Argos system were uploaded to the Wildlife Computers Data Portal (website) for processing with GPE3 software (Wildlife Computers). This statistical processing tool runs ex- clusively on the tag manufacturer’s Internet servers. The GPE3 software uses tag data and corresponding SST (NOAA Optimum Interpolation (01) SST V2 High Resolution) and bathymetry (NOAA ETOPOl global re- lief model. Bedrock version) reference data as inputs into its gridded hidden Markov model to generate the most likely animal location for a given time, as well as a distribution of likelihoods as an indicator of loca- tion quality. This model provides an overall score as an indicator of how well the model fits the observed data. We ran the model with varying inputs for the parameter of animal speed to generate a fit with an optimal score and realistic maximum likelihood track (MLT). Optimal MLTs for LFMl and LFM2 were gener- ated by using animal speed inputs of 4.5 and 2.5 m/s, respectively. The total distance of the MLT was calcu- lated with GE-Path software (vers. 1.4.5). Likelihood surfaces were generated by using the raster and ncdf packages for statistical software R, vers. 3.2.3 (R Core Team, 2015) and by using the script made available by the tag manufacturer. For comparing the MLT with SST, we produced imagery in R from the Group for High Resolution Sea Surface Temperature (GHRSST) global 1-km SST data set (website) using functions in the fields, maps and raster packages for R. We assigned a diel period to each record (day, night, dawn, dusk) in the time-series data sent by the Mini- PAT. To approximate the times of sunrise and sunset for a given date and location (from the MLT), we con- sulted an online calculator (website). On the basis of these estimates, dawn was defined as the 30-min period before and after sunrise, and dusk was defined as the 30-min period before and after sunset. To evaluate dif- ferences in percent time at depth and percent time at temperature between the 2 sharks and to test differ- ences between day and night, we performed 2-sample Kolmogorov-Smirnov (K-S) tests. Mean depths between diel periods were compared with Welch’s unequal vari- ances ^-test. These statistical analyses were performed by using the stats package for R (R Core Team, 2015). To further investigate the environmental drivers of behavior, we used the time-series depth data from the MiniPAT to calculate vertical speed (as a proxy for activity level) for comparison with the corresponding temperature at depth. The difference between sequen- tial depth data points was used to determine vertical velocity. The absolute value of the vertical velocity was considered the vertical speed. We then examined the relationship between daily mean vertical speed and minimum (daytime) and maximum (nighttime) temper- atures during that segment of the day, using a linear 104 Fishery Bulletin 115(1) 90°W 85°W SO-W 75“W 70°\N 90°W SS'-W 80°W 75°W 70°W I Bottom 1 — |— depth (m) 10,000 8000 6000 4000 2000 500 Figure 1 The model-derived maximum likelihood tracks for 2 longfin makos (Isurus paucus) and the locations where they were tagged and where their pop-up tags were released: (A) the first longfin mako, LFMl, was tagged 28 April 2012 off northwest Florida, and (B) the second longfin mako, LFM2, was tagged 14 February 2015 off Cojimar, Cuba. Shaded areas around the tracks represent 99% likelihood surfaces of track locations. The bathymetric scale represents water depth (in meters) to the bottom. least-squares regression analysis. This analysis was also used to examine the relationship between daytime interforay duration and mean temperature, as well as to test for a possible correlation between mean night- time depth and moon phase. For the latter analysis, moon illumination data were acquired from the U.S. Naval Meteorology and Oceanography Command (web- site) and were arcsine-transformed before statistical analysis (Abascal et ah, 2010). All regression analyses were performed in SigmaPlot 10.0 (Systat Software Inc., San Jose, CA). Results On 28 April 2012, LFMl, a mature male (precaudal length [PCL]=201 cm), was tagged with a MklO pop- up satellite archival tag in the northeastern GOM. The 90-d tag was released on schedule and began trans- mitting on 27 July 2012. The first location recorded through the Argos system (37.46°N, 72.35°W) was 330 km east-northeast of Virginia Beach, Virginia, a mini- mum at-sea distance of 2233 km from the tagging site (Fig. lA). Transmitted data were received over a period of 8 d, providing approximately 70% of the summary data. Therefore, although the condition of this shark was judged to be relatively poor (condition 4) when tagged and released, the transmitted data revealed that this animal survived the capture-and-release event. On 14 February 2015, LFM2, a mature male mea- suring 190 cm PCL, was tagged with a MiniPAT off the northern coast of Cuba. The 150-d tag was released on schedule and began transmitting on 15 July 2015. The first location recorded through the Argos system (37.21°N, 73.51°W) was 223 km east-northeast of Vir- ginia Beach, Virginia, a minimum at-sea distance of 1762 km from the tagging site (Fig. IB). The tag trans- mitted data for 21 days, providing approximately 82% of the summary data. Horizontal movements During the first 3 weeks after tagging, LFMl remained off the continental shelf in the eastern GOM, moving in a southeasterly direction (Fig. lA). By the third week of May, the shark entered the Straits of Florida and then continued on an easterly path in the waters be- tween Cuba and the Florida Keys. Then LFMl moved in a northeasterly direction and entered waters of the Bahamas by the end of May, passed to the west of Grand Bahama Island, and moved into the open At- lantic Ocean by early June. During the first 3 weeks of June, the shark maintained a northeasterly direction until reaching continental shelf edge waters offshore of the mouth of Chesapeake Bay. For the rest of its track. Hueter et al.: Horizontal and vertical movements of Isurus paucus in the northwestern Atlantic Ocean 105 38°N 42°N 40'’N 68°W 36°N — I — 70°W 34''N itic Ocean 32”N 78°W 76°W 74°W 15 10 km — I — 72'W 78°W 76°W 74“W 72°W 70°W 68°W 78°W 76°W 74‘>W 72‘>W 70‘=W 68°W 78°W 76°W 74°W 72°W 70'W 68°W Figure 2 Locations of warm water from the Guif Stream during late June and July and estimated positions of longfin makos {Isurus paucus) tracked with satellite-linked tags in the northwestern Atlantic Ocean. (A) Track of the first shark, LFBJl, from 29 June to 5 July 2012 in relation to sea-surface temperature (SST) for 2 July 2012. (B) Track of LFMl for the period 21-27 July 2012 in relation to SST for 24 July 2012. (C) Track of the second shark, LFM2, from 29 June to 5 July 2015 in relation to SST for 2 July 2015. (D) Track of LFM2 for the period 9-15 July 2015 in relation to SST for 12 July 2015. Blue and white circles represent positions in June and July, respectively. Red triangles mark the locations where the pop-up tag was released. The gray lines represent the 1000-m and 4000-m depth contours. LFMl remained within warm Gulf Stream waters and largely off the continental shelf 100-400 km from the coasts of Delaware, Maryland, and Virginia (Fig. lA; Fig. 2, A and B). The northernmost position of the track (39.25°N, 72.83°W) was reached on 11 July. The total distance of the MLT was 6809 km, representing a mean movement rate of 75.7 km/d. The MLT for LFM2 displayed movement from the northern coast of Cuba into the Straits of Florida in the latter half of February, followed by movements into waters of the Bahamas by early March (Fig. IB). Then LFM2 remained in Gulf Stream waters between Grand Bahama Island and southeast Florida for a pe- riod of about 3 weeks. By late March and early April, 106 Fishery Bulletin 115(1) 40°N as^N 30“N 25°N 90°W 85°W 80°W 70°W Figure 3 | Maximum likelihood tracks for 2 iongfin makes (Isurus paucus) * tracked with satellite-linked tags in different years (2012 and I 2015) in the northwestern Atlantic Ocean. Only the common cal- endar days of tracking (28 April-15 July) are shown, which show a high degree of seasonal synchronicity. Shaded areas around tracks represent 99% likelihood surfaces. Both tracks indicate movement largely outside area closures and gear-restricted areas as mandated by the National Marine Fisheries Service (yellow and green poly- gons). Areas 1-4 with seasonal fishing restrictions are in effect for 1) 1 April-31 May, 2J 1 Feb-30 April, 3) 1 Dec-30 April, and 4) the month of June. the shark moved back through the Straits of Florida in a westerly direction. After reaching the edge of the Yucatan shelf by mid-April, LFM2 moved northward into the GOM. By the beginning of May, the shark initiated southerly movements and re-entered the Straits of Florida, and it was located south of the Florida Keys by the first week of May. Next, LFM2 contin- ued on an easterly to northeasterly path through the Bahamas (north of Andros Is- land), into the open Atlantic Ocean, and off the continental shelf. This shark continued in a northeasterly direction during the lat- ter half of May. Then LFM2 remained in pelagic waters and shifted to a northerly direction during the month of June. On 24 June, the shark reached its northernmost position (39.38°N, 70.65°W) before moving in a southwesterly direction during July to the point where its tag was detached (on 15 July). During late June and the first half of July, LFM2 remained within the Gulf Stream and off the continental shelf, approximately 140-330 km from the coasts of New Jersey, Delaware, Maryland, and Virginia (Fig. 2, C and D). The total MLT covered 8826 km, a mean movement rate of 58.8 km/day. By comparing the 2 tracks for the calen- dar days they had in common, one in 2012 and the other in 2015, one can observe a high degree of synchronicity in the move- ments of these sharks in the GOM and Straits of Florida, in their parallel tracks northward in the Atlantic Ocean, and in their convergence in the Mid-Atlantic Bight (MAB) (Fig. 3). In U.S. territorial waters, a broad system of areas in the GOM and off the Atlantic coast are closed year-round or seasonally to pelagic longline fishing (NMFS^). Both LFMl and LFM2 stayed largely outside these protected areas (Fig. 3). Vertical movements Both sharks undertook daily vertical movements and portions of most days were spent near the surface and at depths in excess of 200 m. For LFMl, the depth and temperature ranges experienced during its recorded track were 6-952 m and 4.6-28.8°C. The mean daily vertical range (i.e., difference between minimum and maximum depths) was 494.7 m (standard deviation [SD] 173.8). For LFM2, the ranges in depth and tem- perature were 0-1767 m and 4.0-28.4°C. The mean ^ NMFS (National Marine Fisheries Service). 2016. Pelagic longline restrictions. In HMS commercial compliance guid- ance: guide for complying with the Atlantic tunas, swordfish, shark, and billfish regulations, p. 17-23. Natl. Mar. Fish. Serv., Silver Spring, MD. [Available from website.] daily vertical range of this shark was 435.4 m (SD 147.0). The depth profile for LFM2 indicates a pat- tern of diel vertical migration (DVM), with the shark spending nighttime toward the surface and daytime at greater depths and with periods of dawn and dusk spent largely at intermediate depths (Fig. 4). We noted evidence of seasonal variation in vertical habitat use; LFM2 remained at shallower depths at night during June and July than during other months (Fig. 4). The mean depth at daytime (321.7 m [SD 107.6]) was sig- nificantly deeper for LFM2 than the mean depth at nighttime (94.2 m [SD 90.2]; P<0.0001) (Fig. 5). The mean depth at dawn (245.1 m [SD 104.5]) was shallow- er than the mean depth at dusk (258.7 m [SD 111.9]; P=0.023) (Fig. 5). The results from a comparison of the histogram data for the 2 sharks indicated similar time-at-depth distributions (K-S test: F=0.575; Fig. 6A), although some differences were noted. For example, LFM2 spent 26.1% of its time in the depth range of 300-400 m and LFMl spent 10.0% of its time within that range. The first Iongfin mako spent more time at depths >500 m (10.3%) than did LFM2 (1.6%; Fig. 6A). Time-at-tem- Hueter et al.; Horizontal and vertical movements of Isurus paucus in the northwestern Atlantic Ocean 107 2/14/2015 3/14/2015 4/11/2015 5/9/2015 6/6/2015 7/4/2015 Date Figure 4 Depth profile for the 150-d track of a longfin mako {Isurus paucus-, LFM2) tagged with a pop-up satellite archival tag off the northwestern coast of Cuba in 2015. The time- series depth data points are color-coded by diel period. The daily maximum depth is based on all received summary data sources for this shark. The line for weekly mean depth at nighttime corresponds with values on the right axis. Note that the left axis is split in 2 places. perature distributions were not significantly different for these sharks (K-S test: F=0.989), and both sharks spent the largest proportion of their time in waters with a temperature range of 24-27°C (Fig. 6B). Large proportions of time in cold temperatures were observed; LFMl and LFM2 spent 19.6% and 14.5% of their time in temperatures <12°C, respectively. We separated the binned histogram data for LFM2 into 12-h blocks of time that roughly corresponded to day (0800-2000) and night (2000-0800). The differences between day and night for time at depth (K-S test: P=0.536; Fig. 60 and time at temperature (K-S test: P=0.9895; Fig. 6D) were not significant but further highlighted the DVM pattern. Plotting time-series depth and temperature data for LFM2 was useful for visualizing dynamic patterns in vertical movement, although the relatively low resolu- tion of these data (10-min intervals) precluded accurate calculations of ascent and descent rates. This shark spent daytime periods largely at depth (mean daily depth: 321.7 m); however, we noted upward vertical for- ays during these periods that appeared to be regularly timed and came closer to the surface when the tem- perature at depth was coldest (Fig. 7). The depth profile in Figure 7D shows that LFM2 experienced a change in temperature as high as 18.8°C during its upward movements but spent a relatively brief period of time in the warmer, near-surface waters (approximately 10-15 min) before initiating descent. When we examined this daytime pattern over LFM2’s entire track, we found an inverse linear relationship between temperature and vertical speed (coefficient of multiple determination [i?2]=0.63; Fig. 8A). In contrast, vertical speed during the night for this shark did not appear to be correlated with tempera- ture (i?^=0.001). For LFM2, mean vertical speed in the daytime (2.3 m/min) was higher than in the nighttime (1.91 m/min; P=0.0133). Mean vertical speed at dawn (3.46 m/min) did not differ from that at dusk (3.48 m/min; P=0.921}, although both crepuscular periods showed that this shark traveled at significantly higher speeds than during day and night (F<0.0001). Further, we examined the durations between the upward day- time forays from depths of 250-531 m for the entire track of LFM2 («=36) and found that they ranged from 1.2 to 3.8 h. The results of a linear regression indicate that the duration between these forays was correlated with the mean temperature experienced by this shark at depth {R^=0.60; Fig. 8B). Time-series data further revealed that LFM2 did not always remain in near-sur- face waters during the night because forays to depths >250 m were not uncommon, particularly during the pe- riod February-May (Fig. 9). Unlike the regularly timed daytime movements from depth (>250 m) toward the surface, the nighttime dive profiles for this shark were more variable and indicated a period of bottom time 108 Fishery Bulletin 115(1) before ascent (20-40 min). Additionally, an examina- tion of lunar phase with mean nighttime depth did not detect any significant correlation {R^=0.0005). Discussion This article is the first published report on the move- ments and vertical habitat use of the longfin make, a poorly studied shark species for which our knowledge has largely been based on incidental catches from pe- lagic fisheries. This study is based on 2 satellite tracks, and the results revealed similar movements of the 2 longfin makos from the GOM to the western North At- lantic Ocean, as well as a diel pattern of vertical move- ment, a tolerance for extended periods in deep cold wa- ter, and dynamic vertical forays. Horizontal movements and distribution The MLTs for the 2 longfin makos revealed long-dis- tance movements from the eastern GOM, through the Straits of Florida, and into the open Atlantic Ocean off the northeastern coast of the United States. Although direct comparisons between movement rates are made difficult by differences in tracking or analytical meth- ods, the rates of movement for LFMl (75.7 km/d) and LFM2 (58.8 km/d) are comparable with those reported for other lamnids. Short-term acoustic tracks of shortfin makos have indicated mean speeds of 53 km/d (2.2 km/h; Sepulveda et ah, 2004). Migratory tracks from satellite tagging of white sharks have shown mean movement rates of 74.4 km/d (Bruce et ah, 2006) and that individual white sharks travel as fast as 119 km/d (Weng et ah, 2007). Data from conventional tagging of longfin makos, al- though scant, have indicated movements consistent with those observed with our satellite-tracked longfin makos: 2 sharks conventionally tagged in the eastern GOM were recaptured off the coasts of northern Cuba and eastern Florida (Kohler et al., 1998). Data from a more recent return of a conventional tag from a longfin mako indicated movement from the eastern Caribbean to the continental slope off of Delaware Bay (Kohler‘S). Off the U.S. north Atlantic coast, this species has been tagged in waters along the shelf edge (at depths >200 m), and the few recaptures from this area indicate a pattern of move- ment eastward into deeper pelagic waters (Kohler et al., 1998). The presence of LFMl and LFM2 in the Straits of Florida is consistent with fisheries data from this region. In an overview of the landings of large pelagic species along the northern coast of Cuba, Guitart Manday (1975) reported that the longfin mako was captured in nearly ev- ery month but noted peaks in abundance in surveys conducted during August-November in 1971 and dur- ing April-May and August-November in 1972. Obser- vations from recent studies of the artisanal pelagic longline fishery along the northern coast of Cuba have indicated that the longfin mako continues to be caught year-round and peaks in relative abundance during January-March (J. Angulo Valdes, unpubl. data). Do- drill and Gilmore (1979) documented a beached speci- men of longfin mako at Melbourne Beach, along the east coast of Florida, during the month of December (in 1975) and made note in their addendum of 2 additional specimens captured with drift longlines at depths of 200-400 m between Jupiter and Sebastian Inlets dur- ing the months of April and May (in 1978). In a 2-year study of the shark hycatch in the swordfish fishery off the east coast of Florida, Berkeley and Campos (1988) recorded 2 longfin makos captured during the months of October and December (in 1982). In our study, both male longfin makos moved into the MAB during the months of June and July and used outer continental shelf, slope, and oceanic habitats dur- ing this period. Conventional tagging data from this area indicated a similar distribution of captures, along with a male:female ratio of 1. 9:1.0, for the longfin mako (Kohler et al., 1998). Data provided by the NMFS Pelagic Observer Program (POP) for the period 1992- 2014 indicated that 18% of longfin makos documented Kohler, N. 2015. Personal commun. Northeast Fish. Sci. Cent., Natl. Mar. Fish. Serv., 28 Tarzwell Dr., Narragansett, RI 02882. Hueter et al.: Horizontal and vertical movements of Isurus paucus in the northwestern Atlantic Ocean 109 B 0-50 50-100 g, 100-200 O) m 200-300 Q. 300-400 <13 Q 400-500 >500 D 27-30 - O 24-27 O CD 21-24 a> § 18-21 - £ 15-18 - :□ 2 12-15 - 0 ?■ 9-12 - h- 6-9 - 3-6 - 40 40 Percent time Figure 6 Depth and temperature histograms for 2 longfin makos (Isurus paucus) tracked with sat- ellite-linked tags in the northwestern Atlantic Ocean in 2012 and 2015; (A) mean time-at- depth comparison for the 2 sharks LFMl and LFM2, (B) time-at-temperature comparison for LFMl and LFM2, (C) day (0800-2000) versus night (2000-0800) comparison of time at depth for LFM2, and (D) day versus night comparison of time at temperature for LFM2. were from the MAB area (delineated as 35-43°N; 71- 78°W) and that the greatest proportion were landed in July and August (combined 45%) (Cortes®). Of the 11 geographic areas covered by the POP, only the GOM had a larger proportion of longfin mako captures (34%) than the MAB (Cortes®). Diel vertical movement The depth and temperature data from the tags of LFMl and LFM2 indicate a daily pattern of vertical movement ® Cortes, E. 2015. Personal commun. Southeast Fish. Sci. Cent., Natl. Mar. Fish. Serv., 3500 Delwood Beach Rd., Pan- ama City, FL 32408-7403. ® Cortes, E. 2015. Personal commun. Southeast Fish. Sci. Cent., Natl. Mar. Fish. Serv., 3500 Delwood Beach Rd., Pan- ama City, FL 32408-7403. between near-surface waters and the mesopelagic zone, with overall depth and temperature ranges of 0-1767 m and 4.0-28.8° C. A pattern of DVM was evident with greater time at depth during the day and more time in the mixed layer at night. These observations are con- sistent with longline catches of longfin makos in that most captures occur during overnight sets with baits set at depths of 0-220 m below the surface (Dodrill and Gilmore, 1979; Queiroz et al., 2006; Hemida and Capape, 2008; Bustamante et al., 2009). Satellite track- ing results for its congener, the shortfin mako, indicate a similar DVM pattern, but shortfin makos do not ap- pear to have as great a tolerance for depth and cold temperatures. In the western North Atlantic Ocean, for example, a tagged shortfin mako had depth and temper- ature ranges of 0-556 m and 10.4-28.6° C (Loefer et al., 2005), and results from tracking a shortfin mako in the 110 Fishery Bulletin 115(1) Time (GMT -4 hrs) Figure 7 Variability in daytime vertical movements of a longfin mako ilsurus paucus; LFM2) tagged off the northwestern coast of Cuba in 2015. (A) Temperature-depth profile dur- ing a week-long period in June showing limited temperature variability. (B) Tempera- ture-depth profile in early July showing movement through a temperature-stratified water column. (C) 39-h period beginning 8 June when the shark remained largely at depths of 300-400 m during the day within this relatively unstratified water column. (D) 31-h period beginning 3 July showing regular daytime forays from cold depths into warmer surface waters. GMT-4 h=4 hours behind Greenwich Mean Time. southwest Pacific indicated a comparable vertical move- ment pattern of shallower depths at night than those during the day, descents to 620 m, and a temperature range of 8.8-23.4°C (Stevens et ah, 2010). Similar DVM behavior has been observed in a num- ber of other pelagic shark species, including the bigeye thresher (Alopias superciliosus; Stevens et ah, 2010), white shark (Nasby-Lucas et ah, 2009), and porbeagle (Saunders et ah, 2011), as well as in pelagic teleosts such as the swordfish (Carey and Robison, 1981; Abas- cal et ah, 2010; Sepulveda et ah, 2010) and bigeye tuna {T. obesus; Musyl et ah, 2003), and it often has been suggested that this behavior is a response to the movement of their prey (Musyl et ah, 2003; Stevens et ah, 2010; Saunders et ah, 2011). Although the diet of the longfin mako has not been described fully, the presumed prey of this species includes schooling fish and pelagic squid species (Compagno, 2001). The lat- ter may be the more important dietary component because squid and squid beaks have been identified in the contents of stomachs of longfin makos (Dodril and Gilmore, 1979; Castro, 2011) and are similarly a known component of the shortfin mako diet (Stillwell and Kohler, 1982; Maia et ah, 2006). The stomach of a 310-cm-TL female longfin mako from the southeast- ern Pacific Ocean contained squid remains, most like- ly from jumbo squid (Dosidicus gigas; Bustamante et ah, 2009). Pelagic cephalopods also have been shown to be diel vertical migrators (Roper and Young, 1975). A distinct DVM pattern was observed for jumbo squid tagged with satellite tags: they spent most daylight hours at depths >250 m, rose toward the surface at dusk, and spent the majority of time at night at depths <150 m (Gilly et ah, 2006). This pattern is similar to Hueter et al.: Horizontal and vertical movements of Isurus paucus in the northwestern Atlantic Ocean 111 the pattern of movements of longfin makos observed in our study and indicate that the diel pattern of the longfin mako may be related to a habit of following cephalopod prey. The importance of the convergence of the tracks of LFMl and LFM2 in the MAB in summer is unclear, but their movements likely are influenced directly or indi- rectly by water temperature, given their observed ten- dency to remain near or within the Gulf Stream (Fig. 2). The water column in this region becomes strongly stratified by mid-summer and is characterized by the formation of a warm, mixed surface layer of about 30 m in depth (Castelao et al., 2008). We noted a shallow- er mean depth at nighttime (-30 m) for LFM2 during June and July that indicates a nighttime preference for this mixed layer. Similarly, satellite tagging of bluefin tunas (T. thynnus) in the Atlantic Ocean has indicat- ed a shallower summertime depth distribution when bluefin tunas occupy well-stratified water in this region (Galuardi and Lutcavage, 2012). The 60-d track of a shortfin mako tagged in Gulf Stream waters off South Carolina also had a decrease in nightly depth range after the summer solstice, indicating that the shark or its prey was tracking a thermal regime at night (Loefer et al., 2005). In a study of the diet of large, pelagic, predatory fish species in this same region of the North Atlantic Ocean, the most common component of stom- ach contents was cephalopods, and the highest biomass was represented by ommastrephid squid species (Logan et al., 2013). Taken together, the data from these studies indicate that longfin makos use this area off the MAB as a sum- mer feeding area and adjust their vertical movement patterns in response to seasonal changes in distribu- tions of cephalopod or teleost prey to maximize foraging efficiency. Because both LFMl and LFM2 were sexually mature males, we cannot discount the possibility that this area may also be a mating ground for this species. Using NMFS POP data from the MAB (Keene'^), we es- timated that 58% of female and 46% of male longfin makos caught in this area are of a sexually mature size. There is evidence that other lamnids, such as the white shark, use the MAB as both feeding and mating grounds (Gilmore, 1993). However, additional informa- tion, such as the presence of female longfin makos with fresh mating scars, will be needed to verify that the MAB is a mating ground for this species. Use of deep, cold habitat and endothermy Profiles of time at depth and time at temperature for the 2 longfin makos (Fig. 6) indicated that this species is capable of using cold, deepwater habitat for extended periods of time (42-54% of time at depths >200 m; 14- 20% of time at temperatures <12° C), particularly dur- ing daytime hours. This pattern contrasts with that of its congener, the shortfin mako, which appears to make only brief excursions below the thermocline (Abascal et al., 2011) and reportedly spends only 4% of its time at depths greater than 300 m in the southwest Pacific (Stevens et al., 2010) and only about 6-10% of its time at depths exceeding 200 m in the northwestern Atlan- tic Ocean (Vaudo®). Keene, K., Jr. 2016. Personal commun. Southeast Fish. Sci. Cent., Natl. Mar. Fish. Serv., 75 Virginia Beach Dr., Bldg. 2, Miami, FL 33149-1003. ® Vaudo, J. 2013. Personal commun. The Guy Harvey Res. Inst., Nova Southeastern Univ., 8000 N Ocean Dr., Dania Beach, FL 33004. 112 Fishery Bulletin 115(1) I I I I I M I I I I I I I I I I I I I I I I I M I I I I I I I I I I I I I I"! I ri t I I I I I I I I I I I I 0000 0600 1200 1800 0000 0600 1200 1800 0000 3/17/2015 3/18/2015 3/19/2015 Time Figure 9 Time and depth profile for a 2-d period in March 2015 for a longfin mako (Isurus paucus] LFM2) tracked with a satellite-linked tag in the northwest- ern Atlantic Ocean, indicating nighttime forays by this shark from warm, near-surface waters to cold waters at depths around 300 m. The shaded areas approximate nighttime. Depth data points are color-coded by tempera- ture where available. Lamnid shark species have the ability to conserve metabolic heat by means of vascular countercurrent heat exchangers (a complex called the retia mirahilia) and can maintain their tissues at temperatures signifi- cantly above ambient temperatures, likely as a means for broadening their habitat (Block and Carey, 1985; Goldman, 1997; Goldman et al., 2004). In a study of body temperature measurements and anatomical fea- tures related to heat production and conservation for lamnid species, the longfin mako was found to have a relatively small amount of red muscle and poorly de- veloped retia mirahilia and consequently was rated as the least endothermic of the lamnid shark species (Carey et al., 1985). However, these authors suggest that the retia mirahilia of the longfin mako could in- crease its thermal inertia and prolong cooling time so that this shark species can maintain a body tempera- ture higher than the ambient temperature in cold, deep waters after spending periods of time in warm surface waters. Our findings that daytime vertical activity and interforay duration both are related to temperature at depth (Figs. 7 and 8) are consistent with this hypoth- esis. Tracked shortfin makos have shown a similar pat- tern in the southeastern Pacific Ocean in that reduced thermal structure of the water column coincided with a decrease in the vertical activity of these fish (Abascal et al., 2011). We hypothesize that LFM2 conducted regular for- ays during daytime from cold depths to surface waters to gain heat lost at depth. A similar daytime move- ment pattern from depth to surface waters has been described for the blue shark (Prionace glauca-, Carey et al., 1990), and this movement pattern is a thermoregu- latory strategy that is aided by the property of muscle to warm up more quickly than the time it takes to cool down (Carey and Gibson, 1987). In contrast to the blue shark, the endothermic longfin mako may be able to remain warmer than the ambient water temperature for longer periods of time at depth and then re-warm rapidly during relatively brief forays to the surface. This form of behavioral thermoregulation may enable longfin makos to remain active predators in cold water and to exploit agile, vertically migrating prey, such as pelagic squid species that would largely become un- available (during the day) to competing ectothermic predators. We also noted nighttime vertical oscillations of LFM2 from near surface waters to depths >300 m (Fig. 9). The subsequent ascents brought this shark within 3 m of the surface, perhaps to regain heat after a period (20-40 min) at cold depths. For the shortfin mako, directed descents into deeper water, followed by rapid ascents (also known as bounce dives), have been associated primarily with daytime hours (Abascal et al., 2011) and have been correlated with success- ful feeding events at depth (Sepulveda et al., 2004). Although a DVM pattern has been reported for sat- ellite-tracked jumbo squid, a highly variable amount of nighttime diving to depths in excess of 300 m has been observed for this species (Gilly et al., 2006). It is likely that the nighttime forays we observed with LFM2 are related to foraging and may be a response to variable nighttime movement patterns of its pe- lagic prey. The large size of the longfin mako (second largest of all lamnid species after the white shark, by length) may also be thermally advantageous by fur- ther minimizing conductive heat loss to surrounding sea water. Larger shortfin makos have been found to Hueter et a!.: Horizontal and vertical movements of isurus paucus in the northwestern Atlantic Ocean 113 dive to greater maximum depths than those attained by smaller specimens (Sepulveda et aL, 2004). As with all other lamnid shark species, the longfin mako has the capacity to maintain its visceral temperature sub- stantially above ambient sea temperature through a suprahepatic rete (Carey et al., 1985; Bernal et aL, 2012), which may be a mechanism for enhancing rates of digestion and food assimilation (Goldman 1997; Newton et al., 2015). Implications for consereatisn strategies Pelagic longline fisheries pose a principal threat to longfin makes (Reardon et ai., 2006). Although bycatch data for this species are insufficient, there are pub- lished accounts of longfin makos having been captured in gear that targeted swordfish (Dodrill and Gilmore, 1979; Buencuerpo et ah, 1998; Queiroz et al., 2006; Mucientes et al., 2013). In a report on the swordfish fishery along the east coast of Florida, Berkeley and Campos (1988) noted that more shark species than tar- get species were captured and that the overall mortal- ity rate was 66% for 13 species of hooked sharks. Fur- thermore, in a recent 1-year survey (conducted from October 2010 through November 2011) of the pelagic longline fishery along the northwestern coast of Cuba, the longfin mako was the second-most abundant shark species landed (by number), exceeded only by the short- fin mako, out of the 15 different shark species observed (Aguilar et al., 2014). A nearly 5-year (from October 2010 through April 2015) pilot monitoring program of the same fishery documented the longfin mako as the most abundant pelagic shark species landed (MINAL, 2015). Currently, there are no restrictions on catches of longfin makos in the waters of Cuba. In U.S. waters, the year-round or seasonal closing of areas to pelagic longline fisheries in the GOM and Atlantic Ocean ap- pear to offer minimal protection for this species. Al- though based on only 2 tracks from 2 individuals, our data do not indicate that these sharks spent much time inside these closed areas, where they could have been protected from pelagic longlining. The effect of this fishing pressure may be profound for the longfin mako, given its lower fecundity (2-8 pups per litter; Guitart Manday, 1975; Compagno, 2001) than that of the shortfin mako (4-25 pups per litter; Mollet et al., 2000; Compagno, 2001). The combi- nation of this low productivity and high susceptibility of longfin makos to longline gear has led to this species being ranked as highly vulnerable in ecological risk as- sessments of shark species caught in pelagic longline fisheries (Cortes et al., 2010, 2015). With this vulner- ability and generally lower abundance in comparison with its conspecific, for which it is often misidentified, the longfin mako is a shark species of special conserva- tion concern in today’s oceans. The depths at which sharks distribute themselves profoundly affect their susceptibility to commercial fisheries (Speed et al., 2010). The diel vertical move- ment pattern of the longfin mako contributes to its vulnerability to pelagic longline fisheries, especially to fisheries that target swordfish. Results from a recent study, in v/hich the satellite tracks of sharks were compared with fishing vessel movements in the North Atlantic Ocean, indicate how effectively pelagic long- line fleets are exploiting key oceanic shark habitats, almost entirely overlapping fishing effort with pre- ferred habitat of shark species for much of the year (Queiroz et al, 2016). With only 2 satellite tracks of longfin makos that covered periods of 3-5 months, the results of our study reveal connectivity in the range of this species among the territorial waters of 4 coun- tries— the United States, Cuba, Mexico, and the Ba- hamas— and hence the vulnerability of this species to fishing fleets from multiple nations. More than 75% of species of pelagic sharks and rays have an elevat- ed risk of extinction due to overfishing (Dulvy et al., 2008), and large pelagic shark species, such as the longfin mako, have an approximately 80% probabil- ity of being threatened (Dulvy et al., 2014). Effective management and assessment efforts, however, are of- ten impeded by a lack of species-specific fishery and biological data. Although our study contributes new information to the limited knowledge of the longfin mako, continued efforts to improve data collection and the monitoring of pelagic shark catches, particularly in areas of concentrated abundance of longfin makos, such as those off the northern coast of Cuba, are war- ranted. The recent implementation of the first Nation- al Plan of Action for conservation and management of species of sharks and rays by Cuba (MINAL, 2015) is a positive step toward this goal. Acknowledgments We are indebted to the captains and crews of Florida Institute of Oceanography’s RV Weatherbird II and of the MV Marinabella and FV Poco a Poco in Cuba. We sincerely thank our many Cuban colleagues, including L. Garcia Lopez, A. Alvarez Aleman, and C. Aguilar Betancourt of the Centro de Investigaciones Marinas, Universidad de la Habana, and F. Pina of the Centro de Investigaciones de Ecosistemas Costeros. Special grati- tude is extended to Tandem Stills + Motion, Herzog & Co., and Discovery Communications for documenting the work in Cuba. Permission to deploy satellite tags in Cuba was made possible through the dedicated ef- forts of D. Whittle of the Environmental Defense Fund and of Mundo Latino TV of Cuba. We thank J. Stewart for the R script used to generate likelihood surfaces and B. Galuardi for his helpful advice in constructing figures of sea-surface temperatures. We are grateful to R Rogers and 2 anonymous reviewers for their input toward improving the manuscript. This work was fund- ed by the Gulf of Mexico Research Initiative through the Florida Institute of Oceanography and by the Guy Harvey Ocean Foundation, Christopher Reynolds Foun- dation, Environmental Defense Fund, and Discovery Communications. 114 Fishery Bulletin 115(1) Literature cited Abascal, F. J., J. Mejuto, M. Quintans, and A. Ramos-Cartelle. 2010. Horizontal and vertical movements of swordfish in the Southeast Pacific. ICES J. Mar. Sci. 67:466-474. Abascal, F. J., M. Quintans, A. Ramos-Cartelle, and J. Mejuto. 2011. Movements and environmental preferences of the shortfin mako, Isurus oxyrinchus, in the southeastern Pacific Ocean. Mar. Biol. 158:1175-1184. Aguilar, C., G. Gonzalez-Sanson, R. Hueter, E. Rojas, Y. Ca- brera, A. Briones, R. Borroto, A. Hernandez, and P. Baker. 2014. Captura de tiburones en la region noroccidental de Cuba. Lat. Am. J. Aquat. Res. 42:477-487. Baum, J. K., R. A. Myers, D. G. Kehler, B. Worm, S. J. Harley, and P. A. Doherty. 2003. Collapse and conservation of shark populations in the northwest Atlantic. Science 299:389-392. Berkeley, S. A., and W. L. Campos. 1988. Relative abundance and fishery potential of pe- lagic sharks along Florida’s east coast. Mar. Fish. Rev. 50(1):9-16. Bernal, D., J. K. Carlson, K. J. Goldman, and C. G. Lowe. 2012. Energetics, metabolism, and endothermy in sharks and rays. In Biology of sharks and their relatives, 2nd ed. (J. C. Carrier, J. A. Musick, and M. R. Heithaus, eds.), p. 211-237. CRC Press, Boca Raton, FL. Block, B. A., and F. G. Carey. 1985. Warm brain and eye temperatures in sharks. J. Comput. Physiol. B 156:229-236. Bruce, B. D., J. D. Stevens, and H. Malcom. 2006. Movements and swimming behaviour of white sharks (Carcharodon carcharias) in Australian waters. Mar. Biol. 150:161-172. Buencuerpo, V., S. Rios, and J. Moron. 1998. Pelagic sharks associated with the swordfish, Xiphi- as gladius, fishery in the Eastern North Atlantic Ocean and the Strait of Gibraltar. Fish. Bull. 96:667-685. Bustamante, C., F. Concha, F. Balbontin, and J. Lamilla. 2009. Southernmost record of Isurus paucus Gitart Man- day, 1966 (Elasmobranchii: Lamnidae) in the southeast Pacific Ocean. Rev. Biol. Mar. Oceanogr. 44:523-526. Carey, F. G., and B. H. Robison. 1981. Daily patterns in the activities of swordfish, Xiphi- as gladius, observed by acoustic telemetry. Fish. Bull. 79:277-292. Carey, F. G., and Q. H. Gibson. 1987. Blood flow in the muscle of free-swimming fish. Physiol. Zool. 60:138-148. Carey, F.G., J. G. Casey, H. L. Pratt, D. Urquhart, and J. E. McCosker. 1985. Temperature, heat production and heat exchange in lamnid sharks. Mem. Calif Acad. Sci. 9:92-108. Carey, F. G., J. V. Scharold, and A. J. Kalmijn. 1990. Movements of blue sharks (Prionace glauca) in depth and course. Mar. Biol. 106:329-342. Castelao, R., S. Glenn, O. Schofield, R. Chant, J. Wilkin, and J. Kohut. 2008. Seasonal evolution of hydrographic fields in the central Middle Atlantic Bight from glider observa- tions. Geophys. Res. Lett. 35:L03617. Castro, J. 1. 2011. The sharks of North America, 613 p. Oxford Univ. Press, New York. Clarke, S. C., J. E. Magnussen, D. L. Abercrombie, M. K. McAl- lister, and M. S. Shivji. 2006. Identification of shark species composition and pro- portion in the Hong Kong shark fin market based on molecular genetics and trade records. Conserv. Biol. 20:201-211. Compagno, L. J. V. 2001. Sharks of the world: an annotated and illustrated catalogue of shark species known to date. Vol. 2: Bull- head, mackerel and carpet sharks (Heterodontiformes, Lamniformes and Orectolobiformes). FAO Species Cata- logue for Fishery Purposes 1, vol. 2, 269 p. FAO, Rome. Cortes, E., F. Arocha, L. Beerkircher, F. Carvalho, A. Domingo, M. Heupel, H. Holtzhausen, M. N. Santos, M. Ribera, and C. Simpfendorfer. 2010. Ecological risk assessment of pelagic sharks caught in Atlantic pelagic longline fisheries. Aquat. Living Re- sour. 23:25-34. Cortes, E., A. Domingo, P. Miller, R. Forselledo, F. Mas, F Aro- cha, S. Campana, R. Coelho, C. Da Silva, F. H. V. Hazin, et al. 2015. Expanded ecological risk assessment of pelagic sharks caught in Atlantic pelagic longline fisheries. Col- lect. Vol. Sci. Pap. ICCAT 71:2637-2688. Dodrill, J. W, and R. G. Gilmore. 1979. First North American continental record of the longfin mako (Isurus paucus Guitart Manday). Fla. Sci. 42:52-58. Dulvy, N. K., J. K. Baum, S. Clarke, L. J. V. Compagno, E. Cortes, A. Domingo, S. Fordham, S. Fowler, M. P. Francis, C. Gibson, et al. 2008. You can swim but you can’t hide: the global sta- tus and conservation of oceanic pelagic sharks and rays. Aquat. Conserv: 18:459-482. Dulvy, N. K., S. L. Fowler, J. A. Musick, R. D. Cavanagh, P. M. Kyne, L. R. Harrison, J. K. Carlson, L. N. K. Davidson, S. V. Fordham, M. P. Francis, et al. 2014. Extinction risk and conservation of the world’s sharks and rays. eLife 2014;3:e00590. Francis, M. R, J. C. Holdsworth, and B. A. Block. 2015. Life in the open ocean: seasonal migration and diel diving behaviour of Southern Hemisphere porbeagle sharks (Lamna nasus). Mar. Biol. 162:2305—2323. Fredou, F. L., M. T. Tolotti, T. Fredou, F. Carvalho, H. Hazin, G. Burgess, R. Coelho, J. D. Waters, P. Travassos, and F. H. V. Hazin. 2015. Sharks caught by the Brazilian tuna longline fleet: an overview. Rev. Fish Biol. Fish. 25:365-377. Galuardi, B, and M. Lutcavage. 2012. Dispersal routes and habitat utilization of juvenile Atlantic bluefin tuna, Thunnus thynnus, tracked with mini PSAT and archival tags. PLoS ONE 7(5):e37829. Garrick, J. A. F. 1967. Revision of sharks of genus Isurus with description of a new species (Galeoidea, Lamnidae). Proc. U.S. Natl. Mus. 118:663-690. Gilly, W. F, U. Markaida, C. H. Baxter, B. A. Block, A. Bousta- ny, L. Zeidberg, K. Reisenbichler, B. Robison, G. Bazzino, and C. Salinas. 2006. Vertical and horizontal migrations by the jum- bo squid Dosidicus gigas revealed by electronic tag- ging. Mar. Ecol. Prog. Ser. 324:1-17. Hueter et al.: Horizontal and vertical movements of Isurus paucus in the northwestern Atlantic Ocean 115 Gilmore, R. G. 1983. Observations on the embryos of the longfin make, Isurus paucus, and the bigeye thresher, Alopias supercil- iosus. Copeia 1983:375—382. 1993. Reproductive biology of lamnoid sharks. Environ. Biol. Fish. 38:95-114. Goldman, K. J. 1997. Regulation of body temperature in the white shark, Carcharodon carcharias. J. Comput. Physiol. B 167:423-429. Goldman, K. J., S. D. Anderson, R. J. Latour, and J. A. Mustek. 2004. Homeothermy in adult salmon sharks, Lamna dit- ropis. Environ. Biol. Fish. 71:403-411. Grace, M. A., W. B. Driggers III, J. Forrester, and N. Hopkins. 2007. Landing sling use increases data accuracy and sampling opportunities for large marine fishes captured during fishery-independent longline surveys. Fish. Res. 88:146-149. Guitart Manday, D. 1966. Nuevo nombre para una especie de tiburdn del ge- nereo Isurus (Elasmobranchii: Isuridae) de aguas Cuba- nos. Poeyana (Inst. Biol.), A 15:1—9. 1975. Las pesquerias pelagico-oceanicas de corto radio de accion en la region noroccidental de Cuba. Sen Oceanol. Acad. Cienc. Cuba 31:1-26. Hammerschlag, N., A. J. Gallagher, and D. M. Lazarre. 2011. A review of shark satellite tagging studies. J. Exp. Mar. Biol. Ecol. 398:1-8. Hemida, F., and C. Capape. 2008. On the occurrence of the longfin mako, Isurus paucus (Chondrichthyes: Isuridae) off the Algerian coast (south- western Mediterranean). Acta Adriat. 49(2): 185— 189. Hueter, R. E., C. A. Manire, J. P. Tyminski, J. M. Hoenig, and D. A. Hepworth. 2006. Assessing mortality of released or discarded fish us- ing a logistic model of relative survival derived from tag- ging data. Trans. Am. Fish. Soc. 135:500-508. Killam, K., and G. Parsons. 1986. First record of the longfin mako, Isurus paucus, in the Gulf of Mexico. Fish. Bull. 84:748-749. Kohler, N. E., J. G. Casey, and P. A. Turner. 1998. NMFS Cooperative Shark Tagging Program, 1962- 93: an atlas of shark tag and recapture data. Mar. Fish. Rev. 60(2): 1-87. Kyne, P. M., J. K. Carlson, D. A. Ebert, S. V. Fordham, J. J. Bizzarre, R. T. Graham, D. W. Kulka, E. E. Tewes, L. R. Har- rison, and N. K. Dulvy. 2012. The conservation status of North American, Cen- tral American, and Caribbean chondrichthyans, 148 p. lUCN Species Survival Commission Shark Specialist Group, Vancouver, Canada. Loefer, J. K., G. R. Sedberry, and J. C. McGovern. 2005. Vertical movements of a shortfin mako in the west- ern north Atlantic as determined by pop-up satellite tag- ging. Southeast. Nat. 4:237-246. Logan, J. M., R. Toppin, S. Smith, B. Galuardi, J. Porter, and M. Lutcavage. 2013. Contribution of cephalopod prey to the diet of large pelagic fish predators in the central North Atlantic Ocean. Deep Sea Res. (Part H: Top. Stud. Oceanogr.) 95:74-82. Maia, A., N. Queiroz, J. P. Correia, and H. Cabral. 2006. Food habits of the shortfin mako, Isurus oxyrinchus, off the southwest coast of Portugal. Environ. Biol. Fish. 77:157-167. MINAL (Ministry for the Food Industry). 2015. National plan of action for the conservation and management of Chondrichthyes in the Republic of Cuba, 48 p. Ministry for the Food Industry, Havana, Cuba. Mollet, H. F., G. Cliff, H. L. Pratt Jr., and J. D. Stevens. 2000. Reproductive biology of the female shortfin mako, Isurus oxyrinchus Rafinesque, 1810, with comments on the embryonic development of lamnoids. Fish. Bull. 98:299-318. Mucientes, G., R. Banon, and N. Queiroz. 2013. Updated distribution range of longfin mako Isurus paucus (Lamniformes: Lamnidae) in the North Atlan- tic. J. Appl. Ichthyol. 29:1163-1165. Musyl, M. K., R. W. Brill, C. H. Boggs, D. S. Curran, T. K. Ka- zama, and M. P. Seki. 2003. Vertical movements of bigeye tuna {Thunnus obe- sus) associated with islands, buoys, and seamounts near the main Hawaiian Islands from archival tagging data. Fish. Oceanogr. 12:152-169. Nasby-Lucas, N., H. Dewar, C. H. Lam, K. J. Goldman, and M. L. Domeier. 2009. White shark offshore habitat: a behavioral and environmental characterization of the eastern Pacific shared offshore foraging area. PLoS ONE 4(12):e8163. Newton, K. C., J. Wraith, and K. A. Dickson. 2015. Digestive enzyme activities are higher in the short- fin mako shark, Isurus oxyrinchus, than in ectothermic sharks as a result of visceral endothermy. Fish Physiol. Biochem. 41:887-898. Pade, N. G., N. Queiroz, N. E. Humphries, M. J. Witt, C. S. Jones, L. R. Noble, and D. W. Sims. 2009. First results from satellite-linked archival tagging of porbeagle sharks, Lamna nasus: area fidelity, wider- scale movements and plasticity in diel depth changes. J. Exp. Mar. Biol. Ecol. 370:64-74. Queiroz, N., S. Araujo, P. A. Ribeiro, P. Tarroso, R. Xavier, and A. M. Santos. 2006. A first record of longfin mako, Isurus paucus, in the mid-North Atlantic. Mar. Biodivers. Rec. I:e34. Queiroz, N., N. E. Humphries, G. Mucientes, N. Hammer- schlag, F. P. Lima, K. L. Scales, P. 1. Miller, L. L. Sousa, R. Seabra, and D. W. Sims. 2016. Ocean-wide tracking of pelagic sharks reveals ex- tent of overlap with longline fishing hotspots. Proc. Natl. Acad. Sci. U.S.A. 113:1582-1587. R Core Team. 2015. R: a language and environment for statistical com- puting. R Foundation for Statistical Computing, Vien- na, Austria. [Available from website, accessed January 2016.] Reardon, M. B., L. Gerber, and R. D. Cavanagh. 2006. Isurus paucus. The lUCN Red List of Threatened Species 2006: e.T60225A12328101. Rogers, P. J., C. Huveneers, B. Page, S. D. Goldsworthy, M. Coyne, A. D. Lowther, J. G. Mitchell, and L. Seuront. 2015. Living on the continental shelf edge: habitat use of juvenile shortfin makos Isurus oxyrinchus in the Great Australian Bight, southern Australia. Fish. Oceanogr. 24:205-218. Roper, C. F. E., and R. E. Young. 1975. Vertical distribution of pelagic cephalopods. Smith- son. Contrib. Zool. 209, 51 p. 116 Fishery Bulletin 115(1) Saunders, R. A., F. Royer, and M. W. Clarke. 2011. Winter migration and diving behaviour of porbeagle shark, Lamna nasus, in the Northeast Atlantic. ICES J. Mar. Sci. 68:166-174. Sebastian, H., R A. Haye, and M. S. Shivji. 2008. Characterization of the pelagic shark-fin trade in north-central Chile by genetic identification and trader surveys. J. Fish Biol. 73:2293-2304. Sembiring, A., N. P. D. Pertiwi, A. Mahardini, R. Wulandari, E. M. Kurniasih, A. W. Kuncoro, N. K. D. Cahyani, A. W. Ang- goro, M. Ulfa, H. Madduppa, et al. 2015. DNA barcoding reveals targeted fisheries for endan- gered sharks in Indonesia. Fish. Res. 164:130-134. Sepulveda, C. A., S. Kohin, C. Chan, R. Vetter, and J. B. Graham. 2004. Movement patterns, depth preferences, and stomach temperatures of free-swimming juvenile mako sharks, Isurus oxyrinchus, in the Southern California Bight. Mar. Biol. 145:191-199. Sepulveda, C. A., A. Knight, N. Nasby-Lucas, and M. L. Domeier. 2010. Fine-scale movements of the swordfish Xiphias gla- dius in the Southern California Bight. Fish. Oceanogr. 19:279-289. Speed, C. W., I. C. Field, M. G. Meekan, and C. J. A. Bradshaw. 2010. Complexities of coastal shark movements and their implications for management. Mar. Ecol. Prog. Ser. 408:275-293. Stevens, J. D., R. W. Bradford, and G. J. West. 2010. Satellite tagging of blue sharks (Prionace glauca) and other pelagic sharks off eastern Australia: depth be- haviour, temperature experience and movements. Mar. Biol. 157:575-591. Stillwell, C. E., and N. E. Kohler. 1982. Food, feeding habits, and estimates of daily ration of the shortfin mako (Isurus oxyrinchus) in the northwest Atlantic. Can. J. Fish. Aquat. Sci. 39:407-414. Wakida-Kusunoki, A. T., and D. de Anda-Fuente. 2012. Presence of longfin mako shark Isurus paucus (Chondrichthyes: Lamnidae) in the south-eastern Gulf of Mexico, Tabasco, Mexico. Mar. Biodivers. Rec. 5:e92. Weng, K. C., P. C. Castilho, J. M. Morrissette, A. M. Landeira- Fernandez, D. B. Holts, R. J. Schallert, K. J. Goldman, and B. A. Block. 2005. Satellite tagging and cardiac physiology reveal niche expansion in salmon sharks. Science 310:104-106. Weng, K. C., A. M. Boustany, P. Pyle, S. D. Anderson, A. Brown, and B. A. Block. 2007. Migration and habitat of white sharks (Carcha- rodon carcharias) in the eastern Pacific Ocean. Mar. Biol. 152:877-894. Weng, K. C., D. G. Foley, J. E. Ganong, C. Perle, G. L. Shil- linger, and B. A. Block. 2008. Migration of an upper trophic level predator, the salmon shark Lamna ditropis, between distant ecore- gions. Mar. Ecol. Prog. Ser. 372:253-264. 117 National Marine Fisheries Service NOAA Fishery Bulletin established in 1881 Spencer F. Baird First U.S. Commissioner of Fisheries and founder of Fishery Bulletin Age^ gromtli^ and length-imeiglit relationship of roosterfish pectoraiis} in the eastern Pacific Ocean Abstract — -Growth of roosterfish {Nematistius pectoraiis) was estimat- ed by analyzing daily growth incre- ments of sagittal otoliths collected from individuals captured in Ei Goifo Duke, Costa Rica during 2013-2014 and in southern Baja California Sur, Mexico during 2010-2015. Isometric growth was observed for all individ- uals and no significant differences were observed in the length-weight relationships between sexes or loca- tions. Age estimates ranged from 18 d (0.05 years) to 545 d (1.5 years), although 26% of otoliths (57-133 cm) were not legible because narrow daily grov#th increments were diffi- cult to differentiate in older fish. In- dividual growth parameters indicate that roosterfish grow at a rapid rate during the first year of life, reach- ing sizes of around 60-70 cm in fork length. Although future field valida- tion is necessary, the results of our study provide insight into the life history of this valuable resource of the eastern Pacific Ocean. Manuscript submitted 11 March 2016. Manuscript accepted 30 November 2016. Fish. Bull. 115:117-124 (2017). Online publication date: 27 December 2016. doi: 10.7755/FB.115.1.10 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. Sefia Ortega-Garda (contact author)' Chugey Sepulveda^ Scott Aalbers^ Uliano¥ Jakes-Cota' Ruben Rodriguez-Sanchei' Email address for contact author: sortega@ipn.mx ' Instituto Politecnico Nadonai Centro Interdisciplinario de Ciencias Marinas Avenida Instituto Politecnico Nacional s/n Colonia Playa Paio de Santa Rita C.P. 23096 La Paz, Baja California Sur, Mexico ^ Pfieger Institute of Environmental Research 21 10 South Coast Highway Oceanside, California 92054 A game fish that inhabits the neritic waters of the subtropical and tropical eastern Pacific Ocean (Eschmeyer et al., 1983), the roosterfish {Nematis- tius pectoraiis) is the only species in the genus Nematistius (family Nematistiidae). The most notable ex- ternal feature of this species is the presence of 7 elongated dorsal spines, which give rise to its common name (Rosenblatt and Bell, 1976) and al- lows it to be easily distinguished from other species (Mem, 1995). The roosterfish is a coastal species that has been found to occur in the Pacific Ocean from San Clemente, Southern California to San Lorenzo Island, Peru, including the Gulf of California and the Galapagos Islands (Love et al., 2005). Results from work to date indicate that roosterfish re- side primarily in warm (23-3 1°C), shallow waters where they may spend up to 90% of their time be- tween the sea surface and a depth of 12 m (Sepulveda et al., 2015). All age classes are considered neritic; juveniles are often found along the shoreline and larger individuals are commonly associated with both the surf line and complex near shore habitats (i.e., reefs and sand bars) (Niem, 1995; Sepulveda et al., 2015). Roosterfish can attain sizes up to 191 cm in total length and more than 51.7 kg (Robertson and Allen, 2015). They are active predators that have been reported to feed in shallow wa- ters (i.e., depths of 3-4 m) on school- ing fish species, such as species of mojarra (i.e., Pacific spotfin mojarra [Eucinostomus dowii] and graceful mojarra [E. gracilis}), and species of anchovy (i.e., sharpnose anchovy [Anchoa ischana} and Anchoa spp.) (Hobson, 1968; Rodriguez-Romero et al., 2009). The roosterfish is a very impor- tant resource for sportfishing indus- tries, especially for those that oper- ate at destinations favored by eeo- tourists, such as Baja California Sur, mainland Mexico, and the coast of Central America. Roosterfish contrib- ute significantly to the local economy because it is a common target of in- shore anglers; in fact, this species is sought out among anglers around the 118 world (Eschmeyer et al., 1983). To a lesser degree, the roosterfish also supports small-scale artisanal fisher- ies along the coast of Baja California Sur and Central America; however, because of its limited market value, it is not typically the primary target of such fishing operations (Lindner, 1947; Niem, 1995). Despite the ecological and economic importance of roosterfish, very few scientific studies have focused on this species and basic biological parameters remain largely unknown. In particular, questions pertaining to age and growth, natural mortality, longevity, and reproductive biology remain unanswered. These para- meters are essential for fisheries management and form the basis for most stock assessment models (Gold- man, 2005). This information is also essential for un- derstanding other biological and population traits such as productivity, yield per recruit, predator and prey dy- namics, and habitat requirements (DeVries and Erie, 1996; Campana, 2001; Robinson and Motta, 2002). Given the importance of this species to regional fish- eries off the coast of Baja California Sur and Central America, we estimated age and growth in our study by counting presumed daily growth increments (DGIs) within the sagittal otolith. The results of our study pro- vide the first estimates of growth for this species and allow intraspecific comparisons of growth between 2 geographically distinct locations along the range of the roosterfish in the eastern Pacific Ocean. Materials and methods Collection details From 2010 through 2015, roosterfish specimens were obtained from the sportfishing fieet in Cabo San Lu- cas and La Paz, Baja California Sur, Mexico. Similarly, from 2013 through 2014, specimens were collected from commercial and recreational fleets that operated out of the Golfo Duke, Costa Rica. Each specimen was mea- sured to determine fork length (EL) to the nearest 0.1 cm and weighed to the nearest 0.01 kg. Sex was deter- mined by visual inspection of the gonads, and, when possible, sagittal otolith pairs were removed, cleaned, and stored dry. Length-weight relationships Length-weight relationships (LWRs) were estimated by using the allometric equation: W = aFLb, where W = the weight; and a and b = the intercept and the slope of the regression line, respectively (Ricker, 1975). As with previous works, we assumed that, when b was equal to 3, the relationship was considered to be isometric (Sangun et al., 2007). Similarly, b values ?*3 were associated with allometric growth (Froese, 2006). Estimates of LWRs were calculated independently for Fishery Bulletin 115(1) both sex (males and females) and location (Baja Cali- fornia Sur and Golfo Duke) and compared by using a i! Student’s t-test (Zar, 2010). A Student’s t-test was also | used to evaluate whether b values were significantly ■ different from the null hypothesis for isometric growth ] {Hq. 6=3) (Sangun et al., 2007; Zar, 2010). ! Preparation and analysis of otoliths i 1 Preparation of otoliths followed closely the protocol de- 5 scribed by Secor et al., 1992. Briefly, the right otolith of j each specimen was embedded in crystal polyester resin ! and allowed to harden and dry for a 24-h period. We created transverse sections (0.6 mm) through each oto- ' lith that included the core (Fig. 1) by using an IsoMet i Low Speed Saw^ (model 11-1280-160; Buehler, Lake j Bluff, IL) equipped with a diamond watering blade (se- ries 15HC, Buehler). Because otolith increments (e.g., daily or annual) of many perciform fish species are not deposited in the sagittal plane, the transverse or the frontal planes typically are used for assessing DGIs (Secor et al., 1992). Most otoliths need some form of preparation before their microstructure can be accurately determined; therefore, a polishing procedure was used to remove material, expose the core region, and reveal the pre- sumed DGIs (Secor et al., 1992). To clearly define DGIs and facilitate readings, transversal sections were mounted on histological slides with Cytoseal mounting medium (Thermo Fisher Scientific, Waltham, MA) and hand polished with a series of micrometric sandpaper of decreasing grit size (15-3 pm, Diamond Lapping Film disc; Buehler). For finishing, sections were pol- ished with 0.3-pm MicroPolish alumina (Buehler) and 0.3-pm MicroCloth micrometric sandpaper (Buehler). Two readers independently counted presumed DGIs of prepared otolith sections without prior knowledge of fish length and weight. Readers counted DGIs on transverse sections by using a microscope with trans- mitted light (40-100x). Daily growth increments were counted from the core toward the dorsal edge of the otolith along the same transect (Fig. 1). The consistency or concordance of counts between readers was estimated by using a coefficient of varia- tion (CV) (Chang, 1982): where CVj R CVi = 100%x^^^=^ , (1) the age precision estimate for the jth fish; the ith age determination of the jth fish; the mean age estimate of the jth fish; and the number of times each fish was aged. Lengths at age derived from otolith readings were used to estimate the 3 parameters of the standard von Bertalanffy growth model: ^ 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. Ortega-Garcia et al.: Growth of Nematistius pectoralis in the eastern Pacific 119 = (2) where = length (FL in centimeters) at age t; = average maximum length; k =the individual growth coefficient; ^o=the hypothetical age when the length is equal to 0; and ^ =age. A power function was also tested; however, because our results were similar to those of the von Bertalanffy growth model, only results from the von Bertalanffy growth model were included for consistency and com- parison with other growth studies. Growth parameters were estimated independently During the sampling period, 290 roosterfish (190 specimens from Baja California Sur and 100 speci- mens from Costa Rica) were mea- sured and weighed. Specimens from Baja California Sur (66 males, 74 females, and 50 fish of unknown sex) ranged from 5.2 to 133.0 cm FL and from 0.01 to 25.87 kg in weight. Specimens from Costa Rica (47 males, 18 females, and 35 fish of unknown sex) ranged from 8.3 to 116.0 cm FL and 0.01 to 17.80 kg (Fig. 2). The estimated parameters of the LWR by area and sex are provided in Table 1. Comparisons of growth parameters (values of the exponent b) revealed that all LWRs were not significantly different from a value of 3 (Student’s ^-test: P>0.05) and indicated that roost- erfish growth is isometric (Table 1). Given the find- ing of isometric growth for both location and sex, the length and weight data were pooled into one represen- tative LWR for this study (Fig. 3, Table 1). For age estimation, 182 pairs of sagittal otoliths were collected and processed, 103 pairs from fish col- lected in Baja California Sur (5.2-120.0 cm FL) and 79 pairs from fish collected in Costa Rica (13.5-116.0 cm FL). Of these pairs, 130 pairs or 71.5% (Baja California Sur=68 and Costa Rica=62) were readable and used in the age estimation analyses. The remaining 52 otolith pairs (Baja California Sur=35 and Costa Rica=17) were discarded for one or more of the following reasons: the otoliths were broken, crosscuts did not include the core. Figure 1 (A) Right sagitta of a 63.5-cm-FL roosterfish (Nematistius pectoralis) cap- tured in Baja California Sur, Mexico, in 2014. Dotted white lines represent a transverse section through the core, which is indicated by the black circle. (B) Sagitta showing daily growth increments (that consists of two rings: gray and white) from a 5.2-cm-FL roosterfish (18 d in age) captured in Baja Cali- fornia Sur in 2014. for both study areas (Baja California Sur and Costa Rica), and an analysis of the residual sum of squares was used to evaluate the possible differ- ences in parameters between the 2 locations (Ratkowsky, 1983). Parameters for both the LWR and von Bertalanffy growth model were estimated by using nonlinear least squares by means of an itera- tive process with the Gauss-Newton algorithm (Bates and Watts, 1988), which allows estimation of nonlin- ear model parameters without need- ing to transform the data into a linear function. Statistical analyses were performed with R statistical software, vers. 3.2.3 (R Core Team, 2015). Significance was determined by using a ^-test with an alpha level of 0.05. Results 120 Fishery Bulletin 115(1) 35 n 10 20 30 40 50 60 ■ Mates (BCS) im Males (CR) M Females (BCS) □ Females (CR) □ Unknown sex (BCS) □ Unknown sex (CR) 70 80 90 100 110 120 130 140 Fork length (cm) Figure 2 Size distribution (measured in fork length) of roosterfish {Nematistius pectoralis) captured in Baja California Sur (BCS), Mexico, during 2010-2015 and in Golfo Duke, Costa Rica (CR), 2013-2014. Table 1 Parameters (a and b) of the length-weight relationship of roosterfish (Nematistius pectoralis) captured from 2 locations, Baja California Sur (BCS), Mexico, during 2010-2015 and Costa Rica (CR) during 2013-2014, by area and sex, and Student’s Gtest results for the testing of isometric growth. Area Sex a b t P BCS Males 2.4x10-5 2.84 -0.18 0.85 BCS Females 2.8x10-5 2.81 -0.25 0.80 BCS Sexes combined 2.7x10-5 2.82 -0.25 0.79 CR Males 2.7x10-5 2.81 -0.48 0.62 CR Females 2.8x10-5 2.81 -0.50 0.62 CR Sexes combined 2.7x10-5 2.81 -0.60 0.54 Combined Sexes combined 2.6x10-5 2.82 -0.29 0.76 sections were overpolished, and otolith sections were illegible because of increasingly narrow widths of DGIs in older fish. The estimated ages of the fish assessed from Baja California Sur ranged from 18 d (0.05 years, 5.2 cm FL) to 548 d (1.49 years, 86 cm FL). The range of age estimates from Costa Rica was 32 d (0.08 years, 13.5 cm FL) to 448 d (1.22 years, 78.7 cm FL). In most cases, it was possible to count DGIs from sagittal otoliths of fish <57 cm FL (the size reached at an age of approximately 1 year). The agreement be- tween the estimated ages assigned independently by the 2 readers revealed a CV of 9.02%, indicating a high consistency. The average CV for different size classes increased from 5.9 for fish with 18-100 DGIs to 6.8 for fish with 101-199 DGIs and 9.5 for fish with 200- Ortega-Garda et ai.: Growth of Nematistius pectoralis in the eastern Pacific 121 30 n 25 -- £ O) ■(D 5 20 15 10 - 5 - 0 - o Males • Females ♦ Unknown sex Fitted model /• □ • / /'•D •/ 20 40 60 80 I 100 120 140 Fork length (cm) Figure 3 Collective length-weight relationship of all roosterfish (Nematistius pectoralis) examined in this study. Specimens were captured in Baja California Sur, Mexico, in 2010-2015 and in Golfo Duke, Costa Rica, 2013-2014. 549 DGIs. Reader confidence was reduced for samples from fish that ranged from 57 to 86 cm FL, primarily because of the narrowing and softening of DGI edges. Similarly, for roosterfish >86 cm FL, DGIs were not easily differentiated and, therefore, these fish were not included in the age-estimation portion of our study. An- nual growth marks were not observed in sagittal oto- liths or in cross sections. The estimated parameters of the von Bertalanffy growth model are given by area in Table 2. The analy- sis of residual sums of squares did not reveal signifi- cant differences in growth parameters between Baja California Sur and Costa Rica (P>0.05). Therefore, the data for both locations was combined into a single growth model (Fig. 4). Estimates of growth parameters indicate that roosterfish grow rapidly during the first year of life, attaining a size of approximately 60-70 cm FL within the first year (40% of L„). Discussion Our study provides the first estimates of age and growth for the roosterfish, a species that supports a substantial recreational fishery throughout the east- ern Pacific Ocean. The growth estimates presented here indicate that roosterfish grow at a rapid rate and that they may reach up to 60-70 cm FL (40% of L^) during the first year of life. Accelerated growth in the first year of life has been reported for other warm- water pelagic species, such as the cobia (Rachycentron canadum', Franks et ah, 1999), dolphinfish (Coryphaena hippu- rus; Schwenke and Buckel, 2008), blue marlin (Makaira nigricans', Prince et ah, 1991), and sailfish (Istiophorus platyp- terus', Alvarado-Castillo and Felix-Uraga, 1996). As hypothesized previously, rapid early growth is likely advantageous for survival because it is linked to swimming speed, foraging success, and predation (Prince et ah, 1991). Although field vali- dation is necessary, our work provides an initial hypothesis regarding the growth of this valuable eastern Pacific Ocean species and increases scientific knowl- edge of this poorly known species. Samples of roosterfish were obtained primarily from recreational fisheries that operate seasonally throughout Baja Cali- fornia Sur and Central America. Because roosterfish are typically released in most of these recreational fisheries, acquisition of samples was challenging; specimens from smaller size classes (<50 cm FL) proved to be especially difficult to obtain. In addition, because larger individuals are particularly prized in the catch-and- release fisheries, guides and captains rarely harvest specimens >100 cm FL, a tendency that further inhibited the collection of larger individuals for our study. Length-weight relationships The estimated b values (Table 1) were within the nor- mal, expected range for fish (6=2. 5-3. 5) (Froese, 2006). Bagenal and Tesch (1978) suggested that when the Table 2 Parameter values from the von Bertalanffy growth model for roosterfish (Nematistius pectoralis) captured in Baja California Sur (BCS), Mexico, and Golfo Duke, Costa Rica (CR), during 2010-2015: average maximum length (L„), measured as fork length in centimeters, in- dividual growth coefficient (k), and the hypothetical age when the length is equal to 0 (Iq), measured in years. Area k ^0 BCS 166.90 0.48 -0.02 CR 171.72 0.47 -0.08 Areas combined 163.77 0.47 -0.08 122 Fishery Bulletin 115(1) 100 ra c o 80 - 60 - 40 20 - 0 - Males Females Unknown sex Fitted VBGM ^ • m 7. •• • • I® • ~ □ o ^ no* «P □ D 5 »• o*' r * / ♦ 0 60 120 180 240 300 360 420 480 540 Number of presumed daily growth increments Figure 4 The combined (sex and location) von Bertalanffy growth model (VBGM) of roosterfish (Nematistius pectoralis) captured in Baja California Sur, Mexico, in 2010-2015 and in Golfo Dulce, Costa Rica, 2013-2014. value of b is between 2.8 and 3.2, growth is considered to be isometric (i.e., growth of all body parts is consistent and pro- portional throughout development). Based on t-test results, estimated b values were not significantly different from 3 (Table 1), indicating that isometric growth for roost- erfish is independent of location and sex. Because we did not identify differ- ences in the LWR, samples were pooled and collectively presented (see Froese, 2006). A similar b value was reported by Gonzalez-Sanson et al. (2014) for rooster- fish caught in a coastal lagoon off Barra de Navidad, Jalisco, Mexico. However, the Gonzalez-Sanson et al. (2014) study was based on a limited sample size and size range of fish (6=2.92, n=8; 7.5-29.3 cm in total length). Values of the exponent 6 can be influenced by statistical proce- dure and sample size (Bolger and Con- nolly, 1989), as well as by variations as- sociated with size range, maturation, sex, and time of year (Cone, 1989). Because the spawning dynamics of this species (i.e., season and time of year) remain unknown and we were unable to obtain samples from exceptionally large indi- viduals, it is possible that differences in LWR exist for larger, mature roosterfish. Age and growth The sagittal otolith is the preferred hard structure for estimating age in fish because deposition occurs con- tinuously throughout life, a scenario that enhances age estimation when compared with the use of other hard structures (Campana and Thorrold, 2001). Otoliths are also preferred for age estimation because the struc- tures are not lost or shed (as they are with scales) or reabsorbed (as with bones and spines) (Ramirez-Perez et al., 2011). Sectioned and polished otoliths from in- dividuals <57 cm FL provided visible DGIs that were readily distinguishable. However, in otoliths from larger individuals (>57 cm FL), it was difficult to dif- ferentiate between recent outer edge DGIs, thus caus- ing increased uncertainty in age estimates. Campana (1999) reported that calcium, oxygen, and carbon domi- nate the elemental composition of the otolith and that these elements form the calcium carbonate matrix of the otolith. An excess of calcium carbonate in sagittal otoliths may hinder ridge quantification to varying de- grees (Hill et al., 1989). In cross sections, DGIs were counted from the core to the dorsal edge of the otolith. The path of visual counts were not always in a straight line from the core to the outer edge of a sagitta. The optimal reading path that provided the best visual clarity of DGIs typically fol- lowed a somewhat circuitous route that shifted from one area of the sagitta to another (Uchiyama et al., 1986). The precision of age estimates for the 2 readers of this study was high (CV=9.02%) and aligns with CV values from other studies in which counts of annual marks were used (Prince et ah, 1991; DeMartini et ah, 2007). Campana (2001) proposed that there is no a pri- ori value of CV that can be assigned as a target level for studies of age because it is highly influenced by the species and the nature of the hard structure itself In addition, the consistency between readings often de- creases as fish age increases because growth marks are closer together in older fish (Steward et ah, 2009). We found that the precision of DGI counts decreased with increasing age (18-548 d) and FL (5.2-86.0 cm). Prince et al. (1991) suggested that estimating the age of large or old fish by using DGI counts may result in an underestimation of age and an overestimation of the growth rate. Because it was not possible to count DGIs precisely in roosterfish >86 cm FL (1.5 years old), it is recommended that otolith DGIs be used only for indi- viduals of ages less than 1 year (the age reached at a size of approximately 66 cm FL). Although it was not possible to detect annual marks on the otoliths of roosterfish in our study, a previous study had focused on assessing the potential use of dorsal spines for aging this species (Chavez-Arellano, 2016). Chavez-Arellano (2016) analyzed roosterfish be- tween 14 and 133 cm FL and found a similar average length (70.61 cm FL) for year-1 individuals. However, Ortega-Garcia et al.: Growth of Nematistius pectoralis in the eastern Pacific 123 robust comparisons between these studies are not pos- sible because the work of Chavez-Arellano (2016) was not focused on age determination; he assessed only dor- sal spine suitability for aging purposes. Despite differ- ences in our study and that of Chavez-Arellano (2016), both works support the hypothesis of rapid growth in this species, especially in the first year of life. As has been reported for others species like the dolphinfish, additional structures (e.g., scales) may prove useful in future age assessments of the roosterfish (Schwenke and Buckel, 2008). Although field validation is necessary to confirm the proposed growth hypothesis for roosterfish, the inclu- sion of small size classes (5 cm FL) and a large size range provides support for the use of DGIs for age esti- mation. Oxenford and Hunte (1983) assumed DGIs for dolphinfish, and Uchiyama et al. (1986) subsequently validated those values with both hatchery-reared and fish caught in the wild. For some species, DGIs have been validated by experiments with fish reared in cap- tivity, otolith marking, radiochemical dating, or various other techniques (Prince et al., 1991; Campana, 2001). Ideally, validation experiments should include the pe- riod during which the initial growth ring is formed and should evaluate the regularity of growth-ring forma- tion during all life stages (e.g., spawning, migration, and periods of starvation) because all of these factors may influence the regularity of deposition rates (Uchi- yama et al., 1986). Values of L„ estimated from the von Bertalanffy growth equation were greater than the length of the largest roosterfish sampled in our study (133 cm FL) and slightly lower than the maximum length recorded by Robertson and Allen (2015). This difference likely occurred because the estimated parameters were based only on the initial phases of growth for roosterfish (up to 1.49 years). The relatively rapid growth rate that we report here is similar for both regions (Baja California Sur and Costa Rica), despite differences in oceanographic condi- tions (e.g., annual sea-surface temperature) or potential prey sources and availability. These similarities may be due to the lack of larger individuals in this study (e.g., readability of fish >86 cm FL), a scenario that may have masked potential ontogenetic differences. It may also be that roosterfish from the 2 regions have deposi- tion rates that differ from each other. If ring formation occurs at a rate that is less than one ring per day, it may be that this study underestimates the actual age of the fish surveyed. Regardless, this work presents the first data supporting a growth hypothesis for this spe- cies, and future investigations should focus on the use of field validation techniques. Acknowledgments We wish to thank the Pfleger Institute of Environmen- tal Research and the Institute Politecnico Nacional for the research grant SIP20150861. We also are grateful to C. McCue and M. Gutierrez for their help during sam- pling and to the sportfishing fleet at Cabo San Lucas, Baja California Sur, Mexico. The senior author and R. Rodriguez-Sanchez received a Comision de Operacion y Fomento de Actividades Academicas fellowship. We thank the anonymous reviewers for their constructive comments, which helped us to improve the manuscript. Literature cited Alvarado-Castillo, R. M., and R. Felix-Uraga. 1996. Determinacion de la edad de Istiophorus platypter- us (Pisces: Istiophoridae) al sur del Golfo de California, Mexico. Rev. Biol. Trop. 44(1):233— 239. Bagenal, T. B., and F. W. Tesch. 1978. Age and growth. In Methods for assessment of fish production in freshwater (T. B. Bagenal, ed.), p. 101—136. Blackwell Scientific Publications, Oxford, UK. Bates, D. M., and D. G. Watts. 1988. Nonlinear regression analysis and its applications, 365 p. John Wiley and Sons Inc., New York. Bolger, T., and P. L Connolly. 1989. The selection of suitable indices for the measure- ment and analysis of fish condition. J. Fish Biol. Campana, S. E. 1999. Chemistry and composition of fish otoliths: path- ways, mechanisms and applications. Mar. Ecol. Prog. Ser. 188: 263-297. 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. Campana, S. E., and S. R. Thorrold. 2001. Otoliths, increments, and elements: keys to a com- prehensive understanding of fish population. Can. J. Fish. Aquat. Sci. 58:30-38. Chang, W. Y. B. 1982. A statistical method for evaluating the reproduc- ibility of age determination. Can. J. Fish. Aquat. Sci. 39:1208-1210. Chavez-Arellano, R. 2016. Uso potencial de las espinas de la aleta dorsal del pez gallo, Nematistius pectoralis Gill, 1862, para la de- terminacion de edad. Bachelor’s thesis, 48 p. Universi- dad Autonoma de Baja California Sur, La Paz, Baja Cali- fornia Sur, Mexico. Cone, R. S. 1989. The need to reconsider the use of condition indices in fisheries science. Trans. Am. Fish. Soc. 118:510-514. DeMartini, E. E., J. H. Uchiyama, R. L. Humphreys Jr., J. D. Sampaga, and H. A. Williams. 2007. Age and growth of swordfish (Xiphias gladius) caught by the Hawaii-based pelagic longline fish- ery. Fish. Bull. 105:356-367. DeVries, D. R., and R. V. Frie. 1996. Determination of age and growth. In Fisheries techniques, 2"^ ed. (B. R. Murphy and D. W. Willis, eds.), p. 483-512. Am. Fish. Soc., Bethesda, MD. Eschmeyer, W. N., E. S. Herald, and H. Hammann. 1983. A field guide to Pacific coast fishes of North Ameri- ca, 336 p. Houghton Mifflin Co., Boston, MA. Franks, J. S, J. R. Warren, and M. V. Buchanan. 1999. Age and growth of cobia, Rachycentron canadum. 124 Fishery Bulletin 115(1) from the northeastern Gulf of Mexico. Fish. Bull. 97:459-471. Froese, R. 2006. Cube law, condition factor and weight-length re- lationships: history, meta-analysis and recommenda- tions. J. Appl. Ichthyol. 22:241-253. Goldman, K. J. 2005. Age and growth of elasmobranch fishes. In Man- agement techniques for elasmobranch fisheries (J. A. Musick and R. Bonfil, eds.), p. 76-102. FAO Fish. Tech. Pap. 474, Rome. Gonzalez-Sanson, G., C. Aguilar-Betancourt, D. Kosonoy-Acev- es, G. Lucano-Ramirez, S. Ruiz-Ramirez, J. R. Flores-Ortega, and F. Silva-Batiz. 2014. Weight-length relationships for 38 fish species of Barra de Navidad coastal lagoon, Jalisco, Mexico. J. Appl. Ichthyol. 30:428-430. Hill, K. T., G. M. Cailliet and R. L. Radtke. 1989. A comparative analysis of growth zones in four cal- cified structures of Pacific blue marlin, Makaira nigri- cans. Fish. Bull. 87: 829-843. Hobson, E. S. 1968. Predatory behavior of some shore fishes in the Gulf of California. U.S. Fish Wildl. Serv., Res. Rep. 73, 92 p. Lindner, M. J. 1947. The commercial marine fishes, crustaceans and mol- luscs of the west coast of Mexico. Trans. Am. Fish. Soc. 74:71-80. Love, M. S., C. W. Mecklenburg, T. A. Mecklenburg, and L. K. Thorsteinson. 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. Dep. Interior, U.S. Geol. Surv., Biol. Res. Div., Seattle, WA. Niem, V. H. 1995. Nematistiidae. In Pacifico centro-oriental. Volu- men HI. Vertebrados — Parte 2 (W. Fischer, F. Krupp, W. Schneider, C. Sommer, K. E. Carpenter, and V. H. Niem, eds.), p. 1322. Guia FAO para la identificacion de espe- cies para los fines de la pesca. FAO, Rome. Oxenford, H. A., and W. Hunte. 1983. Age and growth of dolphin, Coryphaena hippurus, as determined by growth rings in otoliths. Fish. Bull. 81:906-909. Prince, E. D., D. W. Lee, and J. R. Zwelfel. 1991. Estimating age and growth of young Atlantic blue marlin, Makaira nigricans from otolith microstructure. Fish. Bull. 89:441-459. R Core Team. 2015. R: a language and environment for statistical com- puting. R Foundation for Statistical Computing, Vien- na, Austria. [Available from website, accessed Decem- ber 2015.] Ramirez-Perez, J. S., F. N. Melo-Barrera, and L. E. Ayala-Bobadilla. 2011. Age and growth of Pacific golden-eye tilefish (Cau- lolatilus affinis) in the central region of the Gulf of Cali- fornia. Cienc. Mar. 37:71-85. Ratkowsky, D. A. 1983. Nonlinear regression modeling: a unified practical approach, 288 p. Marcel Dekker Inc., New York. Ricker, W. E. 1975. Computation and interpretation of biological sta- tistics of fish populations. Bull. Fish. Res. Board Can. 191, 382 p. Robertson, D. R., and G. R. Alien. 2015. Shorefishes of the tropical eastern Pacific: online information system: online information system. Version 2.0. Smithsonian Tropical Research Institute, Balboa, Panama. Robinson, M. P., and P. J Motta. 2002. Patterns of growth and the effects of scale on the feeding kinematics of the nurse shark (Ginglymostoma cirratum). J. Zool. 256:449—462. Rodriguez-Romero, J., X. G. Moreno-Sanchez, L. A. Abitia- Cardenas, and D. S. Palacios-Salgado. 2009. Trophic spectrum of the juvenile roosterfish, Nematistius pectoralis Gill, 1862 (Perciformes: Nematisti- idae), in Almejas Bay, Baja California Sur, Mexico. Rev. Biol. Mar. Oceanogr. 44:803-807. Rosenblatt, R. H. and M. A. Bell. 1976. Osteology and relationships of the roosterfish, Nematistius pectoralis Gill. Nat. Hist. Mus. Los Ang. Cty., Contrib. Sci. 279, 23 p. Sangun, L., E. Akamca, and M. Akar. 2007. Weight-length relationships for 39 fish species from the north-eastern Mediterranean coast of Turkey. Turk- | ish J. Fish. Aquat. Sci. 7:37-40. Schwenke, K. L., and J. A. Buckel. 2008. Age, growth, and reproduction of dolphinfish (Co- { ryphaena hippurus) caught off the coast of North Caro- j lina. Fish. Bull. 106:82-92. 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. Sepulveda, C. A., S. A. Aalbers, and D. Bernal. 2015. Post-release survival and movements patterns of roosterfish (Nematistius pectoralis) off the Central Amer- ican coastline. Lat. Am. J. Aquat. Res. 43:162-175. Steward, C. A., K. D. DeMaria, and J. M. Shenker. 2009. Using otolith morphometries to quickly and inex- pensively predict age in the gray angelfish (Pomacanthus arcuatus). Fish. Res. 99:123-129. Uchiyama, J. H., R. K. Burch, and S. A. Kraul Jr. 1986. Growth of dolphins, Coryphaena hippurus and C. equiselis, in Hawaiian waters as determined by daily in- crements on otoliths. Fish. Bull. 84:186-191. Zar, J. H. 2010. Biostatistical analysis, ed., 944 p. Prentice Hall, Upper Saddle River, NJ. 125 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 review, 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 manuscripts accepted by the fournal 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 figure 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 I spaced typed pages (12-point font) and, like articles, I describe an original contribution to fisheries science, j They follow the same format as that for articles: ab- i straet, introduction, results and discussion, but the results and discussion sections may be com- j bined. They are distinguished from full articles in that j 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. Companioti 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. Review 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 text 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 126 Fishery Bulletin 115(1) 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] iquations 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, 'xf, n, ±, 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 and do not cut and paste equations and letters or symbols of variables from a different software program. 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). Acknowledgments should be no more than 6 lines of text. Only those who have contributed in an outstand- ing way should be acknowledged by name. For recogni- tion of other persons or groups, use a general term, such as “crew,” “observers,” “research coordinators,” and do not include names with these terms. 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- Guidelines for authors 127 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. Fo©tn©tes 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., Dittp://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 ail 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. Capy right lawi 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. 128 Fishery Bulletin 115(1) 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- 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 YES, 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) I (Purchase Order No.) Charge your order. IT’S EASY! Please Choose Method of Payment: I I Check payable to the Superintendent of Documents I I GPO Deposit Account I I VISA or MasterCard Account your orders I I I I I I I I I I p-p |— |— |-^ -I (202) 512-2104 (Credit card expiration date) (Authorizing Signature) Mail to: Superintendent of Documents Thank you for P.O. Box 979050, St. Louis, Missouri 63197 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(s>gpo.gov SMrTHSONIAN LIBRARIES 3 9088 01921 4964