MARINE ECOLOGY PROGRESS SERIESMar Ecol Prog Ser

Vol. 550: 163–174, 2016doi: 10.3354/meps11723

Published May 25

INTRODUCTION

Coastal ecosystems are characterized by high pri-mary productivity (Nixon et al. 1986) and high faunaldiversity (Ross 1986), which support broad ecologicalniches. Coastal bays within the northwestern Gulf ofMexico (GOM) fit this characterization, with highlevels of nutrient input from the Mississippi River,and other rivers, driving productive coastal ecosys-tems. The northwest GOM is also a system in flux,with increasingly high levels of disturbance due toglobal warming, including habitat destruction andhypoxia, leading to periodic fluctuations in biodiver-

sity (Rabalais et al. 2009). These varied modes of dis-turbance can affect the abundance of upper trophic-level predators such as sharks directly and throughmovement and density of their prey base (Heupel &Hueter 2002, Torres et al. 2006). One of the ways tounderstand the effects of environmental change onsharks is by analyzing regionally specific dietarytrends, allowing estimation of potential changes inecosystem structure.

Sharks within the Order Carcharhiniformes areamong the most common species in coastal waters ofthe GOM (Burgess et al. 2005, Drymon et al. 2010,Bethea et al. 2015). Among the most abundant of

© The authors 2016. Open Access under Creative Commons byAttribution Licence. Use, distribution and reproduction are un -restricted. Authors and original publication must be credited.

Publisher: Inter-Research · www.int-res.com

*Corresponding author: jplumlee@tamu.edu

Feeding ecology of three coastal shark speciesin the northwest Gulf of Mexico

Jeffrey D. Plumlee1,2,*, R. J. David Wells1,2

1Department of Marine Biology, Texas A&M University at Galveston, 1001 Texas Clipper Rd, Galveston, TX 77553, USA2Department of Wildlife and Fisheries Sciences, Texas A&M University, College Station, TX 77843, USA

ABSTRACT: The feeding ecology of 3 coastal shark species — Atlantic sharpnose Rhizoprionodonterraenovae, bonnethead Sphyrna tiburo, and blacktip Carcharhinus limbatus shark — wasexamined in the northwest Gulf of Mexico (GOM). A total of 601 sharks (305 Atlantic sharpnose,239 bonnethead, and 57 blacktip) were collected over 2 yr from recreational anglers in Galveston,TX. All individuals had stomach contents examined and a subset (50 Atlantic sharpnose, 50 bonnethead, and 36 blacktip sharks) was analyzed for stable isotopes (carbon, nitrogen, and sulfur) in muscle tissue, revealing short-term and long-term feeding strategies. Both blacktip andAtlantic sharpnose shark stomach contents consisted of teleost fishes with percent index of relative importance (%IRI) of 98.95 and 91.16, respectively, whereas bonnethead diets were dominated by crustaceans (%IRI = 87.20). Stable isotope analysis revealed bonnetheads hadhigher mean carbon (δ13C) and lower sulfur (δ34S) values, indicating inshore feeding and a preference for benthic invertebrates, respectively. Atlantic sharpnose and blacktip sharks wereshown to feed on similar prey using stomach content analysis, yet Atlantic sharpnose sharks hada broader diet, including cephalopods and crustaceans in addition to teleost fishes. Differenceswere further established using nitrogen (δ15N) values, which were significantly lower for Atlanticsharpnose than blacktip sharks. Collectively, stomach contents and stable isotope analyses sup-ported different feeding strategies of 3 common shark species. δ34S appeared to serve as a naturaltracer, distinguishing benthic versus pelagic feeding patterns in elasmobranchs. This study pro-vides important ecosystem-based feeding information of upper trophic-level predators in coastalwaters of the northwestern GOM.

KEY WORDS: Feeding ecology · Stable isotopes · Stomach contents · Gulf of Mexico

OPENPEN ACCESSCCESS

Mar Ecol Prog Ser 550: 163–174, 2016164

these are the Atlantic sharpnose Rhizoprionodon ter-raenovae, bonnethead Sphyrna tiburo, and blacktipCarcharhinus limbatus sharks (Bethea et al. 2015).All 3 of these species are abundant along the Texascoast, but vary seasonally in their distributions(Froeschke et al. 2010). While distribution and abun-dance data for Atlantic sharpnose, bonnethead, andblacktip sharks in the northwest GOM exist, littleinformation is available regarding their feedinghabits, which is essential to understanding how theirpresence affects ecosystem structure and function-ing. Elasmobranchs have been shown to have latitu-dinal and longitudinal dietary shifts that are impor-tant to understand when creating accurate ecosystemmodels (Bethea et al. 2007, Drymon et al. 2012).

While similar in range, these species differ slightlywith respect to assumed feeding patterns in otherregions (Castro 1996, Cortés et al. 1996, Bethea et al.2006). East of the Mississippi River, blacktip sharksare known piscivores, targeting teleost prey withinthe families Sciaenidae and Clupeidae, along withother baitfish and larger teleost fishes as they mature(Branstetter 1987, Castro 1996, Barry et al. 2008).Atlantic sharpnose sharks are more opportunistic,focusing on crustaceans (e.g. shrimp) when they arejuveniles and teleost fishes, cephalopods, and portu-nid crabs as they mature (Bethea et al. 2006, Drymonet al. 2012). Bonnetheads are primarily benthic feed-ers, targeting stomatopods, shrimp, portunid crabs,and cephalopods as juveniles, then becoming morespecialized on blue crabs Callinectes sapidus asadults (Cortés et al. 1996, Bethea et al. 2007).

Examination of diet and foraging patterns wherethere is known species mixing can aid in understand-ing resource partitioning and trophic structure(Papastamatiou et al. 2006, Kinney et al. 2011). Pair-ing dietary information with stable isotope analysisprovides a comprehensive assessment of feedingstrategies on a small spatial scale (Post 2002). Thiscombination of techniques offers useful informationabout short- and long-term feeding patterns, respec-tively (Wells et al. 2008, Kinney et al. 2011). Stomachcontents highlight feeding habits on a short temporalscale (hours to days) (Hynes 1950, Hyslop 1980,Cortés 1997), while stable isotope analysis using car-bon (δ13C), nitrogen (δ15N), and sulfur (δ34S) can pro-vide feeding information on temporal scales ofweeks, months, and years, depending upon species-specific and tissue-specific turnover rates (DeNiro &Epstein 1978, Post 2002, Hussey et al. 2012). δ13Cratios are widely used to identify the source(s) of pri-mary production and δ15N can determine trophic lev-els (DeNiro & Epstein 1978, Post 2002, Hussey et al.

2012). In addition, δ34S has been shown to contrastbenthic versus pelagic foraging strategies in teleostfishes (Peterson & Fry 1987, Thomas & Cahoon 1993,Wells et al. 2008). δ34S values tend to be lower in ben-thic zones due to the increased percentage of sulfidesin the sediment, but higher in the water columnwhere an increase in sulfates occurs (Peterson 1999,Fry et al. 2008, Kiyashkoa et al. 2011). Collectively,these 3 natural tracers, combined with stomach con-tent analysis, have not been used in elasmobranchfeeding studies to describe overall trophic and eco-system structure within the marine food web.

The objective of this study was to examine the feed-ing strategies of 3 coastal shark species in the north-west GOM. Factors of sex, year, month, location, andontogeny were used when identifying intraspecifictrends in diet. We used stomach content analysis toquantify dietary breadth and to examine the impor-tance of individual prey items in the diet of each sharkspecies. In addition, we used bulk stable isotope ratiosof δ13C, δ15N, and δ34S to examine food web structure,to elucidate feeding ecology of the 3 species, andcompare individual dietary preferences. Ultimately,we aim to better understand the dietary niche of eachspecies and use these data for ecological models forecosystem-based fisheries management.

MATERIALS AND METHODS

Sample collection

This study draws from a dataset of opportunisticsamples collected from April through October of2013 and 2014 from Galveston, Texas. The majorityof samples (n = 593) were collected dockside fromrecreational anglers via hook and line, along withspecimens collected from the Texas Parks andWildlife Department bottom longline survey (n = 8).Sample location and information regarding bait typeused was assessed through personal communicationwith anglers. All sharks used for the study weretaken during day trips throughout waters surround-ing Galveston, which minimized likelihood for varia-tions in the base of the food web. Each shark wassexed and measured along a straight line to the near-est centimeter to obtain total length (TL), fork length(FL), and precaudal length (PCL). Stomachs wereremoved from individuals at the dock and sealed viazip tie at the esophageal end of the stomach and theanterior end of the scroll valve so that no contentswere lost. Epaxial muscle tissue was removed fromeach specimen anterior to the primary dorsal fin for

Plumlee & Wells: Feeding ecology of coastal sharks 165

stable isotope analysis. If the dorsal fin location couldnot be assessed after processing, muscle tissue wasremoved from the dorsal portion of the vertebral column.

Stomach content and stable isotope procedures

Stomachs were preserved via a 48 h fixing processin 10% formalin and then moved to a solution of 70%ethyl alcohol for longer-term storage. Each stomachwas measured for full wet weight (g), opened, andseparated with a series of 3 metal mesh sieves sized1.27 cm, 1400 μm, and 500 μm. All contents includingotolith, bone and carapace found within the stomachswere identified to the lowest possible taxon. Onceidentified, contents were sorted and weighed to thenearest 0.1 g. If bait was found that was previouslyidentified through angler interviews, it was weighedand removed from subsequent analysis.

Muscle tissue samples were immediately cata-logued and frozen at −20°C upon return to the lab.During processing, tissue samples were dried at 60°Cfor 48 h in a Heratherm OGS180 drying oven (Ther-moScientific) then lipid was extracted via the DionexAccelerated Solvent Extractor 35. The extraction pro-cess used 34 ml cells packed with layered tissue sam-ples separated by 30 mm Whatman filter papers, andran in cycles of 5 min saturations with petroleumether at 100°C and 1500 psi (105.5 kg cm–2) in orderto reach thermal equilibrium, followed by a flushwith fresh solvent. This procedure was repeated 3times per cell to ensure the removal of lipids. Follow-ing lipid extraction, tissue was homogenized via aWig-L-Bug® grinding mill and further dried at 60°Cfor 24 h to remove any additional solvent. Once dry,the tissue was encapsulated using 5 × 9 mm tin cap-sules, placed in a 96 plate well, and shipped foranalysis. Analysis of the stable isotopes δ13C and δ15Nwas performed using a PDZ Europa ANCA-GSL ele-mental analyzer interfaced to a PDZ Europa 20-20isotope ratio mass spectrometer (IRMS) (Sercon), andδ34S analysis was done using an Elementar vario ISO-TOPE cube interfaced to a 20-22 IRMS (Sercon).Heavy isotopes were compared to laboratory stan-dards; carbon was compared via Vienna PeeDeeBelemnite, nitrogen was compared via atmosphericN2, and sulfur was compared via Vienna Canon Dia-blo Trilobite. All analysis was done through the Sta-ble Isotope Facility at the University of California atDavis. Stable isotope data were presented in deltanotation, δX = [(Rsample/ Rstandard) − 1] × 1000, where Xis the heavy isotope, Rsample is the ratio of heavy to

light isotope in the sample, and Rstandard is the ratio ofheavy to light isotope in the reference standard. Theneed for lipid extraction was confirmed using repli-cate samples of extracted and non-extracted tissue.For each species, 10 samples of muscle tissue wereselected for comparison. Half of each sample selectedunderwent lipid extraction and the remaining half ofthe sample was left unaltered. The samples ofextracted and non-extracted tissue were comparedvia paired Student’s t-test. Significant differences(α ≤ 0.05) were detected between mean δ13C andδ15N stable isotope ratios in both blacktip and At lan -tic sharpnose sharks, while δ34S, along with all stableisotope ratios in bonnetheads, had no significant dif-ferences between extracted and non-extrac ted tis-sue. Post-extraction tissue was raised to the thresholdC:N ratio described in Hussey et al. (2011) of 3.0,which is closest to pure protein and ideal when con-ducting stable isotope analysis. To ensure homo-geneity among all tissues, all samples were lipid-extracted to remove effects of high lipid concen tra -tion, as well as soluble urea, on isotopic ratios.

Data analysis

Feeding patterns were investigated according tointerspecific differences. In addition to differencesamong species, intraspecific differences were identi-fied using sex, year, ontogeny (length and maturity),and month as factors. Shark species were separatedby maturity for intraspecific analysis, based on lengthat 50% maturity (LF50) measurements for each spe-cies (Baremore & Passerotti 2013, Hoffmayer et al.2013, Frazier et al. 2014); however, owing to a lowcomparable sample size between mature and imma-ture for all species, samples were placed into 10 cmsize bins using FL for interspecific analysis. Analysisregarding stomach contents was done by organizingthe taxonomic groups found within the stomachs intohigher categories; highest level taxa were achievedat the subphylum and infraclass levels (Teleostei,Crustacea, and Cephalopoda), while less commontaxa were grouped into Other (Echinodermata,Bivalvia, Gastropoda, and various algae). For furtherclassification detail among groups, Teleostei andCrustacea were broken down using prey groupingssimilar to the ones described in Bethea et al. (2004)and Bethea et al. (2007). Groupings included epi -benthic teleost, pelagic teleost, penaeid shrimp,brachyurans, other crustaceans, and cephalopods.Unidentified material from Teleostei and Crustaceawere removed from this analysis. A percent index of

Mar Ecol Prog Ser 550: 163–174, 2016

relative importance (%IRI) was computed for preyitems using percent weight (%W), percent numericalquantity (%N), and percent frequency of occurrence(%O) (Pinkas et al. 1971, Cortés 1997):

IRI = (%N + %W) × %O (1)

(2)

For analysis using stomach contents, %W was usedto calculate differences among species, as it is a met-ric that is frequently used to quantify nutritional con-tribution (Rooker 1995). Stomach contents by %Wwere analyzed using ANOSIM with a Bray-Curtissimilarity matrix, and additional information to iden-tify the most important prey items was providedusing SIMPER analysis. Both ANOSIM and SIMPERnon-para metric statistical techniques were carriedout using PRIMER v.7 (Clarke & Gorley 2015). Stom-achs containing contents with negligible weight, but identifiable contents, were excluded from %W calcu-lations, yet included on %IRI analysis. Further analy-sis was conducted using the Shannon-Wiener diver-sity index (H ’), incorporating both species evennessindex (J ’) and species richness (S), to quantify thediversity of the diet of each species:

(3)

where pi is the proportional abundance of species i:

(4)

H max = ln S and J ’ is therefore H' divided by H_maxor (ln S):

(5)

Along with prey diversity, trophic level was esti-mated using stomach contents and standardized preyitem trophic positions were estimated similar toCortés (1999). Taxonomic richness was also assessedvia cumulative prey curve (CPC) to assess if thebreadth of each species diet had been fully described(Ferry & Cailliet 1996).

MANOVA models were applied to incorporateδ13C, δ15N, and δ34S to assess differences among spe-cies. ANOVA models were then used to compare dif-ferences among species using individual stable iso-tope values. Simple linear regressions were used todetermine any length effects correlated with δ13C,δ15N, and δ34S. When length was determined to havea significant effect, it was selected as a covariate and

incorporated into ANCOVA models, which wereused for intraspecific analyses using ANOVA models. Statistical significance was assessed at α ≤0.05 and all parametric tests were analyzed usingSYSTAT (Cranes Software Inter national).

Quadratic discriminant function analysis (QDFA)was used to test the ability of %W of major taxonomicgroups and stable isotope ratios of shark tissue to distinguish uniqueness of their feeding strategiesbased upon individual species. Jackknife cross- validated classifications were used to quantify classification success to respective species basedupon dietary contribution. QDFA models were basedon residuals of %W of major taxonomic groups(Teleostei, Crustacea, Cephalopoda, and Other) andδ13C, δ15N, and δ34S. Classification success has aninverse relationship to levels of dietary overlap: highlevels of overlap are reflected in low percent classification success, while low levels of overlap arereflected in high percent classification successamong the 3 species.

RESULTS

A total of 601 (n = 305 Atlantic sharpnose, n = 239bonnethead, and n = 57 blacktip sharks) stomachswere analyzed in this study. Of those, 83.3% (n =254) Atlantic sharpnose, 92.5% (n = 221) bonnet-head, and 61.4% (n = 35) blacktip shark stomachscontained identifiable contents and were used forstatistical analysis. Along with stomachs analyzed,136 (n = 50 Atlantic sharpnose, n = 50 bonnethead,and n = 36 blacktip shark) tissue samples were usedfor stable isotope analysis. Size ranges and sex ratiosvaried for collected samples of each species: Atlanticsharpnose shark 51.3 to 89.5 cm FL (48 females,[mean ± SE] FL = 80.7 ± 1.0 cm; 257 males, FL =77.4 ± 0.4), bonnethead 49 to 102 cm FL (166 females,FL = 82.4 ± 0.7 cm; 73 males, FL = 77.2 ± 0.7 cm), andblacktip shark 52.7 to 143.7 FL (31 females, FL =108.1 ± 3.0 cm; 26 males, FL 94.5 ± 3.5 cm) (Fig. 1).

Stomach contents

Among the 3 species, 54 individual taxonomicgroups were identified through stomach contentanalysis, with 23 taxa being identified to the specieslevel. Samples from Atlantic sharpnose sharks con-tained 37 taxonomic groups, bonnetheads contained23 taxonomic groups, and blacktip sharks contained14 taxonomic groups. Both Atlantic sharpnose sharks

%IRIIRI

IRIprey

total

=⎛⎝⎜

⎞⎠⎟× 100

H p pi

S

i i’ ==∑

1

ln

pii=

n

N

JH

H’

’

max

=

166

Plumlee & Wells: Feeding ecology of coastal sharks

and bonnetheads had CPCs reaching an asymptote,while blacktip sharks did not, indicating further sam-pling is needed to fully describe the diet of this spe-cies in this region. Trophic position, estimated using

stomach contents, for blacktip sharks (4.22) was sim-ilar to what was found by Cortés (1999) off of thewestern Atlantic Ocean. However, Atlantic sharp-nose sharks (4.20) and bonnetheads (3.40) werefound to be closer than estimated previously in theGOM, and values were higher than those estimatedfrom the western Atlantic (Delorenzo et al. 2015).Dietary taxonomic diversity was determined usingH ’, supporting that Atlantic sharpnose sharks hadthe highest S (Atlantic sharpnose shark = 37, bonnet-head = 23, blacktip shark = 14) and bonnetheads hadthe highest J ’ (Atlantic sharpnose shark = 0.52, bon-nethead = 0.62, blacktip shark = 0.32) and overall H ’(Atlantic sharpnose shark = 1.99, bonnethead = 2.39,blacktip shark = 1.22).

Stomach contents from both blacktip and Atlanticsharpnose sharks (Tables S1 & S2, respectively, inthe Supplement at www.int-res.com/articles/suppl/m550p163_supp.pdf) consisted primarily of un -identified teleosts, (%IRI = 91.16 and 98.95, re -spectively). Primary prey categories for Atlanticsharpnose sharks included unidentified Teleostei(87.40%IRI), Penaeidae (3.56%IRI), and Teuthoidea(2.92%IRI). Within identified Teleostei, the largestcontributing fish taxon to the diet of Atlantic sharp-nose sharks was family Sciaenidae (1.84%IRI).Blacktip sharks had primary prey categories ofunidentified Teleostei (88.52%IRI) and Atlanticcroaker Micro pogonias undulatus (7.46%IRI), whileadditional overall non-teleost prey categories (Crus-tacea, Cephalopoda, and Other) summed to1.15%IRI. Diet of bonnetheads (Table S3) consistedprimarily of crustaceans (87.20%IRI), with the top 3contributors as unidentified Brachyuran (48.91%IRI),Callinectes sapidus (18.06%IRI), and C. similis(3.32%IRI) (Fig. 2).

167

Fig. 1. Size distributions of Atlantic sharpnose (n = 305), bon-nethead (n = 239), and blacktip (n = 57) sharks collected in2013 and 2014 from Galveston, Texas. Lines indicatingmedian ages of maturity (dashed lines: male; dotted lines:female) are noted on the histograms based on length at

50% maturity (LF50) measurements for each species

Fig. 2. %IRI contributions of each of the dietary categories,using major overall taxonomic groups for Atlantic sharpnose(n = 254), bonnethead (n = 221), and blacktip sharks (n = 35)

Mar Ecol Prog Ser 550: 163–174, 2016

ANOSIM was used to compare stomach contents(%W) for 3 species of shark across the following cat-egories: Teleostei, Crustacea, Cephalopoda, andOther. ANOSIM for among species analysis was sig-nificantly different (global R = 0.501, p ≤ 0.05). Pair-wise analysis within ANOSIM revealed significantdifferences between Atlantic sharpnose sharks andbonnetheads (global R = 0.534, p ≤ 0.05) as well asblacktip sharks and bonnetheads (global R = 0.892,p ≤ 0.05), while no difference was denoted betweenAtlantic sharpnose and blacktip sharks (global R =−0.186, p > 0.05). SIMPER analysis showed %W ofCrustacea was the most important contributor driv-ing diet differences between bonnetheads andblacktip sharks (93.27 average dissimilarity), andbonnetheads and Atlantic sharpnose sharks (75.52average dissimilarity), with highest values associatedwith bonnetheads. High %W of Teleostei combinedwith low %W of Crustacea were noted for blacktipsharks, while higher %W of Cephalopoda was alsoshown to differ in Atlantic sharpnose shark diet rela-tive to blacktip sharks (32.81 average dissimilarity).

To infer further relationships, and due to low ap -plicable samples sizes of various taxonomic groups,ANOSIM was run among species using prey group-ings epibenthic teleost, pelagic teleost, brachyuran,penaeid shrimp, other crustaceans, and cephalopods.Using these groupings, significant differences werefound among species (global R = 0.576, p ≤ 0.05).Pairwise analysis within ANOSIM revealed signifi-cant differences between Atlantic sharpnose sharksand bonnetheads (global R = 0.573, p ≤ 0.05), andblacktip sharks and bonnetheads (global R = 0.903,p ≤ 0.05), while no difference was denoted betweenAtlantic sharpnose and blacktip sharks (global R =−0.011, p > 0.05). SIMPER analysis showed %W ofbrachyurans in bonnetheads and %W of epibenthicand pelagic teleosts in blacktip sharks (99.07 averagedissimilarity) along with Atlantic sharpnose sharks

(75.52 average dissimilarity) being the largest differ-ence among the 3 species. High %W of epibenthicteleosts combined with low %W of penaeid shrimpwere noted differences between blacktip and At -lantic sharpnose shark stomach contents. ANOSIMwas also run using sex, month, year, length, andmaturity level as intraspecific factors for each of the 3species. No intraspecific factor was found to be sig-nificant using stomach content data.

Stable isotope analysis

Analysis with MANOVA for among species com-parisons using all stable isotope ratios (δ13C, δ15N,and δ34S) indicated significant differences amongspecies (F = 29.697, p ≤ 0.05). Individual ANOVAswere then conducted using individual stable isotoperatios among species. δ13C was significantly differentamong species (F = 3.119, p ≤ 0.05), with highest val-ues in bonnetheads (−16.89 ± 0.05‰, mean ± SE), fol-lowed by blacktip sharks (−16.94 ± 0.04‰), andAtlantic sharpnose sharks (−17.03 ± 0.03‰). δ15N wasalso significantly different across species (F = 9.453,p ≤ 0.05) with highest values for blacktip sharks(16.43 ± 0.09‰), followed by Atlantic sharpnosesharks (16.04 ± 0.08‰), and lastly bonnetheads(15.91 ± 0.08‰). δ34S was significantly differentamong species (F = 8.840, p ≤ 0.05), with highest val-ues in blacktip sharks (16.79 ± 0.14‰), followed byAtlantic sharpnose sharks (16.70 ± 0.16‰), and bon-netheads (15.94 ± 0.16‰) (Table 1, Fig. 3).

Intraspecific analysis was performed usingANOVA with month, sex, maturity, and year as factors for each of the stable isotope ratios. δ13C washigher for mature bonnetheads (F = 13.347, p ≤ 0.05)and blacktip sharks (F = 5.430, p ≤ 0.05), as well asblacktip sharks collected in 2013 relative to 2014 (F =5.271, p ≤ 0.05). δ15N was higher for female, relative

168

Test R2 F p Pairwise Tukey’s HSD p

δ13C 0.047 3.119 0.047 Atlantic sharpnose shark vs. bonnethead shark 0.036Atlantic sharpnose shark vs. blacktip shark 0.325Blacktip shark vs. bonnethead shark 0.682

δ15N 0.124 9.453 <0.001 Atlantic sharpnose shark vs. bonnethead shark 0.440Atlantic sharpnose shark vs. blacktip shark 0.005Blacktip shark vs. bonnethead shark <0.001

δ34S 0.117 8.840 <0.001 Atlantic sharpnose shark vs. bonnethead shark 0.001Atlantic sharpnose shark vs. blacktip shark 0.913Blacktip shark vs. bonnethead shark 0.001

Table 1. ANOVA results for stable isotope ratios among species. Bold: significantly different at p ≤ 0.05

Plumlee & Wells: Feeding ecology of coastal sharks

to male Atlantic sharpnose sharks (F = 39.214, p ≤0.05); however, no sex-specific differences wereobserved for the other 2 species. δ34S was signifi-cantly higher for all species in 2014 relative to 2013:Atlantic sharpnose sharks (F = 14.155, p ≤ 0.05), bonnetheads (F = 11.927, p ≤ 0.05), and blacktipsharks (F = 5.906, p ≤ 0.05). δ34S was also higher inimmature bonnetheads relative to mature for bothyears (F = 9.340, p ≤ 0.05) (Table 2).

Regression analysis was run for each stable isotopeon each species and compared to FL to examineontogenetic patterns. Blacktip shark length had asignificant positive linear relationship with δ13C (R2 =0.114, p = < 0.05) and δ15N (R2 = 0.128, p = < 0.05),and bonnethead length had a significant negativelinear relationship with δ34S (R2 = 0.107, p = < 0.05),while Atlantic sharpnose sharks showed no signifi-cant linear relationships (Fig. 4).

Jackknife reclassification success using QDFA wascalculated for each species to estimate dietary over-lap. Reclassification success was highest with %Wcombined with all 3 stable isotope ratios for bonnet-heads (93%) as well as blacktip sharks (76%), andlowest for Atlantic sharpnose sharks (24%) due tohigher overlap. Reclassification using %W alone wasmost useful in identifying specialization of feedingwithin the dataset (92% bonnethead, 96% blacktipshark, and 17% Atlantic sharpnose shark), whilereclassification using only stable isotope ratiosyielded less accurate yet more evenly distributedclassification success (58% bonnethead, 61% black-tip shark, and 38% Atlantic sharpnose shark).

DISCUSSION

Both stomach content and stable isotope analysessupport different feeding patterns for 3 commonshark species in the northwest GOM. Several studieshave shown an overlap of range for these 3 species inthe GOM (Drymon et al. 2010, Bethea et al. 2015) andalong the coast of Texas (Froeschke et al. 2010).However, until now there have been few observa-tions regarding their compared feeding strategies inthe northwest GOM. Bonnetheads were found to bemost different, with the vast majority of the diet con-sisting of benthic invertebrates. Blacktip and Atlanticsharpnose sharks both had diets consisting mostly ofteleost fishes, primarily from the family Sciaenidae.Fishes that feed throughout the water column con-tain a higher amount of sulfates (SO4) and higherrespective δ34S values than organisms that specializeon benthic prey such as crustaceans and other ben-thic invertebrates (Peterson & Fry 1987), and thesepatterns were clearly demonstrated by bonnetheadswith lower δ34S values and a diet primarily of crus-taceans (87%IRI). In contrast, higher δ34S valueswere reflective in both blacktip and Atlantic sharp-nose sharks, which both had teleost-dominated diets(98 and 91%IRI, respectively). Blacktip sharks weremore specialized feeders, when compared to Atlanticsharpnose sharks, consuming almost exclusively

169

Fig. 3. Biplots of mean (±SE) stable isotope ratios for Atlanticsharpnose, bonnethead, and blacktip shark species using

δ13C, δ15N, and δ34S

Mar Ecol Prog Ser 550: 163–174, 2016

tele ost fishes with corresponding higher δ15N valuesreflective of a diet that consisted of higher trophic-level prey items. Atlantic sharpnose sharks werefound to feed more generally, with stomachs con -taining a mixture of teleost fishes along with crusta ceans and cephalopods. Atlantic sharpnosesharks also had corresponding lower δ15N values,consistent with a diet of lower trophic-level teleostprey items and increased contributions from otherlower trophic-level taxa. Bonnethead samples con-tained the lowest δ15N values, which reflected feed-ing on the lowest trophic level of the 3 species. All 3species had overlapping δ13C values; however, bon-netheads had significantly higher mean values, sug-gesting a more inshore feeding strategy versus anoffshore feeding strategy shown by the lower meanδ13C values of the other 2 species (Hussey et al. 2012).

The piscivorous feeding strategy of blacktip sharksfound in the northwest GOM is similar to other stud-ies throughout the GOM. A full breadth of diet analy-sis was challenging due to the large proportion ofempty stomachs in blacktip sharks (61% of stom-achs), which has also been seen in previous studies(Hoffmayer & Parsons 2003, Barry et al. 2008). Thisproportion of empty stomachs is possibly due to thecaveat of hook and line sampling. Hook and linesampling has the potential to attract active animals

on the search for food versus othermore active gear types, which are lessdiscriminatory towards animals withhigher levels of satiation (Cortés1997). The diet of blacktip sharks pri-marily consisted of teleost fishes, amajority of which were in the family Sciaenidae. The dominant fish spe-cies identified in the stomachs ofblacktip sharks was Atlantic croakerMicropogonias un dulatus. Atlanticcroaker was found to be far moreabundant in blacktip shark diets rela-tive to gulf menhaden Brevoortiapatronus, which dominated in otherstudies (Hoffmayer & Parsons 2003,Bethea et al. 2004, Barry et al. 2008).Blacktip sharks have also been shownto have high levels of sympatry withmenhaden, being a high percentageof bycatch in the menhaden commer-cial fishery (de Silva et al. 2001). Bothspecies are abundant in GalvestonBay (Rozas & Zimmerman 2000) andalong the nearshore waters of theTexas coast (Lewis et al. 2007). Gulf

menhaden and the Atlantic menhaden B. tyrannushave been shown to be the primary prey for blacktipsharks from studies in Louisiana, Mississippi, andFlorida (Hoffmayer & Parsons 2003, Bethea et al.2004, Barry et al. 2008), yet were found in lowerabundance when compared to Atlantic croaker inblacktip shark diets. The GOM experienced adecline of over 50% in total catch of menhaden from2009 to 2013, while for those same years, total catchof Atlantic croaker in the GOM stayed relatively con-stant (4% increase) (NOAA Fisheries 2015). Annualfluctuations in abundance may contribute to the lownumber of gulf menhaden found in the diets of black-tip sharks in our study along with the aforemen-tioned lack of samples.

Ontogenetic diet shifts were detected using stableisotope ratios for blacktip shark. Both δ13C and δ15Nwere found to have positive relationships with length,confirming shifts in the diets of blacktip sharks withincreasing age. Ontogenetic diet shifts have beenidentified in teleost fishes and chondrichthyans (Fryet al. 1999, Albo-Puigserver et al. 2015) using bothδ13C and δ15N. Increasing δ15N is indicative of preda-tion on larger prey as well as increased trophic levelwith age, which corresponds to known dietary infor-mation for blacktip sharks (Castro 1996, Hoffmayer &Parsons 2003, Barry et al. 2008, Bornatowski et al.

170

Dependent variable δ13C (‰) δ15N (‰) δ34S (‰)

Atlantic sharpnose sharkMaturity (mature/immature) −17.02 ± 0.03/ 16.05 ± 0.09/ 16.68 ± 0.19/

−17.04 ± 0.06 16.01 ± 0.16 16.79 ± 0.36Year (2013/2014) −17.07 ± 0.03/ 15.94 ± 0.12/ 16.15 ± 0.26/

−16.98 ± 0.04 16.15 ± 0.11 17.25 ± 0.12Sex (male/female) −17.04 ± 0.03/ 16.25 ± 0.07/ 16.77 ±0.16/

−16.99 ± 0.06 15.33 ± 0.10 16.46 ± 0.48

Bonnethead Maturity (mature/immature) −16.83 ± 0.04/ 15.94 ± 0.07/ 15.78 ± 0.16/

−17.33 ± 0.19 15.69 ± 0.47 17.15 ± 0.27Year (2013/2014) −16.84 ± 0.07/ 15.90 ± 0.11/ 15.45 ± 0.20/

−16.95 ± 0.07 15.91 ± 0.12 16.43 ± 0.20Sex (male/female) −16.81 ± 0.04/ 16.06 ± 0.08/ 15.74 ± 0.24/

−16.93 ± 0.07 15.83 ± 0.11 16.13 ± 0.20

Blacktip sharkMaturity (mature/immature) −16.83 ± 0.07/ 16.67 ± 0.21/ 16.71 ± 0.21/

−17.00 ± 0.04 16.31 ± 0.08 16.84 ± 0.19Year (2013/2014) −17.06 ± 0.05/ 16.64 ± 0.20/ 16.30 ± 0.18/

−16.89 ± 0.04 16.34 ± 0.09 17.01 ± 0.18Sex (male/female) −17.00 ± 0.05/ 16.31 ± 0.08/ 16.68 ± 0.24/

−16.89 ± 0.06 16.54 ± 0.15 16.90 ± 0.17

Table 2. Mean (±SE) differences of stable isotope ratios among maturity(mature/immature), year (2013/2014), and sex (male/female). Bold: significantly

different using ANOVA at p ≤ 0.05

Plumlee & Wells: Feeding ecology of coastal sharks

2014). The overall change in δ13C indicates an in-crease in movement and expansion of home range asontogenetic dietary shifts occur. Dietary shifts result-ing from movement offshore can account for up to a4‰ in δ13C, which has been demonstrated throughteleost fishes and elasmobranchs (Leakey et al. 2008,Hussey et al. 2011). Blacktip sharks have been shownto be highly migratory, with regular seasonal migra-

tions away from the coast as they age, which aredriven primarily by seasonal changes in temperaturein the GOM (Heupel & Simpfendorfer 2002, Hueter etal. 2005). These seasonal movements to maximizetheir ectothermic meta bolisms (Papastamatiou &Lowe 2012) may explain the difference in sourcepathway, δ13C, as the sharks begin to migrate as theygrow larger and mature.

Bonnetheads were found to have a very narrow di-etary niche, with a diet dominated by the blue crabCallinectes sapidus. Blue crab has been shown inseveral studies throughout the GOM and the north-west Atlantic Ocean to be a primary item in the diet ofbonnetheads (Cortés et al. 1996, Bethea et al. 2007).Our findings confirm a similar diet preference forblue crab with a focus on lesser blue crab C. simi lisand stomatopods. A significant amount of plant mate-rial, primarily benthic macroalgae, was found in 36%of the stomachs of bonnetheads, which was likely aresult of incidental ingestion when pursuing benthicinvertebrates. Cortés et al. (1996) found contributionsfrom 3 species of sea grass (Tha lassia testudinum, Syringodium filiforme, and Halo dule wrightii) in 56%of bonnethead stomachs in southwest Florida. Betheaet al. (2007) also found significant contributions fromplant matter in the diets of bonnetheads, especiallyfor young-of-the-year sharks. The primary vegetativematter found for our study area was strictly green(Chlorophyta) and brown algae (Phaeophyta), withno contributions from seagrasses or other angio -sperms. This confirms the regionally related trendsre garding vegetative matter ingestion, with primaryvegetative matter consumed being algae, rather thanseagrass, due to Galveston Bay’s low amount of sea-grass beds and high amount of benthic macroalgae(Pulich & White 1991).

Stable isotope analysis for bonnetheads mirroredthe unique feeding strategies revealed through stom-ach content analysis. Analysis revealed significantlylower δ34S and δ15N, indicating benthic invertebratepredation and lower trophic-level feeding. Ontoge-netic diet shifts were detected for bonnetheads be -tween δ34S values and length. Previous studies ofbonnetheads observed ontogenetic diet shifts withincreasing specificity towards larger blue crab, andmoving away from other smaller prey items. Betheaet al. (2007) and Cortés et al. (1996) found strong positive correlations between carapace length ofingested blue crabs and size of bonnetheads, con-firming changes in prey preference with size. Ourfindings did not show any correlation for diet prefer-ence changing with size for bonnetheads; however,δ34S showed a negative relationship with shark FL,

171

Fig. 4. Simple linear regression lines for Atlantic sharpnose,bonnethead, and blacktip sharks, comparing fork length

(FL) to stable isotope values δ13C, δ15N, and δ34S

Mar Ecol Prog Ser 550: 163–174, 2016

indicating an increase of specificity in benthic inver-tebrate consumption (Fry et al. 2008, Wells et al.2008). Ontogenetic changes were also detected inδ13C between immature and mature bonnetheads,which may also be linked to blue crabs. Blue crabsspawn over several months from April to November,primarily in June through August, with largerfemales having several batches of eggs per season(Dickinson et al. 2006, Graham et al. 2012). Whenspawning, females move into offshore waters ofhigher salinity when a corresponding pattern of olderindividuals are observed offshore (Dickinson et al.2006). This movement of prey can potentially drivelarger bonnetheads further from the coast, which isreflected in the decrease in δ13C and confirmed usingobserved offshore seasonal movement patterns(Heupel et al. 2006, Ubeda et al. 2009, Driggers et al.2014).

Atlantic sharpnose sharks were shown to be oppor-tunistic predators, consuming a wide assortment oftaxonomic groups including teleost fishes, crusta -ceans, and cephalopods. Atlantic sharpnose sharkshad the highest taxonomic richness found among the3 species, with 37 individual taxonomic groups foundin their stomachs. The majority of the Atlantic sharp-nose shark diet consisted primarily of teleost fishes, agroup that was shared with blacktip sharks, with themost abundant fish taxon found being the family Sci-aenidae. A study from the eastern and central GOMby Drymon et al. (2012) showed regional variance inAtlantic sharpnose shark diet using both stomachcontents and stable isotope ratios, indicating a wideprey base. This trend in feeding corresponds todietary information found in this study. While Atlan -tic sharpnose sharks showed very large taxonomicrichness in their diet, they had a low species even-ness and diversity. A consumer that exhibits trophicplasticity may have a large prey base but feedsopportunistically, taking advantage of less numeri-cally available prey items to supplement their diet,whereas a feeder that has a less taxonomically di -verse diet may feed evenly throughout its prey base,such as bonnetheads. Further evidence of the widetrophic niche of the Atlantic sharpnose shark wasindicated by its low classification success usingQDFA. Low classification success of Atlantic sharp-nose sharks (24%) indicates a large amount of over-lap with the other 2 species within the model, whichwas contrasted by high levels of classification in bonnetheads (93%), which had very little dietaryoverlap. Atlantic sharpnose sharks also had the lowest mean δ13C values, indicating they fed the furthest offshore of the 3 species. They have been

shown to have a relatively small home range, havinga 95% estimated home range of <9 km (Carlson et al.2008) as juveniles, little range separation via onto -geny (Bethea et al. 2015), and are thought to pup offshore (Drymon et al. 2010). These trends correlateto the low mean δ13C values suggestive of offshorefeeding where Atlantic sharpnose shark abundanceis highest.

Establishing feeding patterns among migratorypredators is crucial to understanding ecosystemdynamics and predator interactions. Assumptionsthat group species together as predators, ignoringdifferences in prey base and dietary trends, do notallow for accurate estimates of species influences onthe ecosystem. This study suggests that all 3 spe-cies — Atlantic sharpnose, bonnethead, and blacktipshark — have different dietary preferences. Thisimportant distinction between species with signifi-cant range overlap can be used to provide estimatesof their combined ecosystem influence throughouttheir range in the northwest GOM. Accurate esti-mates of feeding patterns can add crucial data to eco-system models and provide useful information forecosystem-based fishery management.

Acknowledgements. We thank the students at Texas A&MUniversity at Galveston (E. Williams, M. Fort, J. Carter, J.Cullen, F. Shopnitz, M. Benitez, S. Hoskinson, P. Faulkner,K. Clark, V. Quesnell, and T. Harrington) for their assistancein the field and the laboratory. In addition, a special thanksto Williams Party Boats, Galveston Party Boats, Get HookedCharters, 3G Charters, and the Galveston Yacht Basin forallowing us to collect samples for the project. This work wassupported in part by the Texas Sea Grant and the TexasInstitute of Oceanography. We also thank T. TinHan, J.Mohan, and T. Richards along with the 3 anonymousreviewers for their comments that improved the manuscript.

LITERATURE CITED

Albo-Puigserver M, Navarro J, Coll M, Aguzzi J, Cardona L,Sáez-Liante R (2015) Feeding ecology and trophic posi-tion of three sympatric demersal chondrichthyans in thenorthwestern Mediterranean. Mar Ecol Prog Ser 524: 255−268

Baremore IE, Passerotti MS (2013) Reproduction of theblacktip shark in the Gulf of Mexico. Mar Coast Fish 5: 127−138

Barry KP, Condrey RE, Driggers WB III, Jones CM (2008)Feeding ecology and growth of neonate and juvenileblacktip sharks Carcharhinus limbatus in the Tim-balier−Terrebone Bay complex, LA, USA. J Fish Biol 73: 650−662

Bethea DM, Buckel JA, Carlson JK (2004) Foraging ecologyof the early life stages of four sympatric shark species.Mar Ecol Prog Ser 268: 245−264

Bethea DM, Carlson JK, Buckel JA, Satterwhite M (2006)Ontogenetic and site-related trends in the diet of the

172

Plumlee & Wells: Feeding ecology of coastal sharks 173

Atlantic sharpnose shark Rhizoprionodon terranovaefrom the northeast Gulf of Mexico. Bull Mar Sci 78: 287−307

Bethea DM, Hale L, Carlson JK, Cortes E, Manire CA, Gel-sleichter J (2007) Geographic and ontogenetic variationin the diet and daily ration of the bonnethead shark,Sphyrna tiburo, from the eastern Gulf of Mexico. MarBiol 152: 1009−1020

Bethea DM, Ajemian MJ, Carlson JK, Hoffmayer ER andothers (2015) Distribution and community structure ofcoastal sharks in the northeastern Gulf of Mexico. Envi-ron Biol Fishes 98: 1233−1254

Bornatowski H, Braga RR, Abilhoa V, Corrêa MFM (2014)Feeding ecology and trophic comparisons of six sharkspecies in a coastal ecosystem off southern Brazil. J FishBiol 85: 246−263

Branstetter S (1987) Age and growth estimates for blacktip,Carcharhinus limbatus, and spinner, C. brevipinna,sharks from the northwestern Gulf of Mexico. Copeia1987:964−974

Burgess GH, Beerkircher LR, Cailliet GM, Carlson JK andothers (2005) Is the collapse of shark populations in theNorthwest Atlantic Ocean and Gulf of Mexico real? Fisheries 30: 19−26

Carlson JK, Heupel MR, Bethea DM, Hollensead LD (2008)Coastal habitat use and residency of juvenile Atlanticsharpnose sharks (Rhizoprionodon terraenovae). EstuarCoast 31: 931−940

Castro JI (1996) Biology of the blacktip shark, Carcharhinuslimbatus, off the southeastern United States. Bull Mar Sci59: 508−522

Clarke KR, Gorley RN (2015) Primer v. 7: user manual andtutorial. PRIMER-E, Plymouth

Cortés E (1997) A critical review of methods of studying fishfeeding based on analysis of stomach contents: applica-tion to elasmobranch fishes. Can J Fish Aquat Sci 54: 726−738

Cortés E (1999) Standardized diet compositions and trophiclevels of sharks. ICES J Mar Sci 56:707–717

Cortés E, Manire CA, Hueter RE (1996) Diet, feeding habits,and diel feeding chronology of the bonnethead shark,Sphyrna tiburo, in southwest Florida. Bull Mar Sci 58: 353−367

de Silva JA, Condrey RE, Thompson BA (2001) Profile ofshark bycatch in the U.S. Gulf of Mexico MenhadenFishery. N Am J Fish Manage 21: 111−124

Delorenzo DM, Bethea DM, Carlson JK (2015) An assess-ment of the diet and trophic level of Atlantic sharpnoseshark Rhizoprionodon terraenovae. J Fish Biol 86: 385−391

DeNiro MJ, Epstein S (1978) Influence of diet on the distribution of carbon isotopes in animals. Geochim Cosmochim Acta 42: 495−506

Dickinson GH, Rittschof D, Latanich C (2006) Spawningbiology of the blue crab, Callinectes sapidus, in NorthCarolina. Bull Mar Sci 79: 273−285

Driggers WB, Frazier BS, Adams DH, Ulrich GF, Jones CM,Hoffmayer ER, Campbell MD (2014) Site fidelity ofmigratory bonnethead sharks Sphyrna tiburo (L. 1758) tospecific estuaries in South Carolina, USA. J Exp Mar BiolEcol 459: 61−69

Drymon JM, Powers SP, Dindo J, Dzwonkowski B, HenwoodTA (2010) Distributions of sharks across a continentalshelf in the northern Gulf of Mexico. Mar Coast Fish 2: 440−450

Drymon JM, Powers SP, Carmichael RH (2012) Trophic plasticity in the Atlantic sharpnose shark (Rhizoprion-odon terraenovae) from the north central Gulf of Mexico.Environ Biol Fishes 95: 21−35

Ferry L, Cailliet G (1996) Sample size and data analysis: arewe characterizing and comparing diet properly? In: MacKinley D, Shearer K (eds) Feeding ecology and nutri-tion in fish. Proc Symp American Fisheries Society, July14–18, 1996, San Francisco State University, CA, p 71–80

Frazier BS, Driggers WB III, Adams DH, Jones CM, LoeferJK (2014) Validated age, growth and maturity of the bonnethead Sphyrna tiburo in the western NorthAtlantic Ocean. J Fish Biol 85: 688−712

Froeschke J, Stunz GW, Wildhaber ML (2010) Environ -mental influences on the occurrence of coastal sharks inestuarine waters. Mar Ecol Prog Ser 407: 279−292

Fry B, Mumford PL, Tam F, Fox DD, Warren GL, Havens KE,Steinman AD (1999) Trophic position and individualfeeding histories of fish from Lake Okeechobee, Florida.Can J Fish Aquat Sci 56: 590−600

Fry B, Cieri M, Hughes J, Tobias C, Deegan LA, Peterson B(2008) Stable isotope monitoring of benthic−planktoniccoupling using salt marsh fish. Mar Ecol Prog Ser 369: 193−204

Graham DJ, Perry H, Biesiot P, Fulford R (2012) Fecundityand egg diameter of primiparous and multiparous bluecrab Callinectes sapidus (Brachyura: Portunidae) in Mississippi waters. J Crustac Biol 32: 49−56

Heupel MR, Hueter RE (2002) Importance of prey density inrelation to the movement patterns of juvenile blacktipsharks (Carcharhinus limbatus) within a coastal nurseryarea. Mar Freshw Res 53: 543−550

Heupel MR, Simpfendorfer CA (2002) Estimation of mortal-ity of juvenile blacktip sharks, Carcharhinus limbatus,within a nursery area using telemetry data. Can J FishAquat Sci 59: 624−632

Heupel MR, Simpfendorfer CA, Collins AB, Tyminski JP(2006) Residency and movement patterns of bonnetheadsharks, Sphyrna tiburo, in a large Florida estuary.Environ Biol Fishes 76: 47−67

Hoffmayer ER, Parsons GR (2003) Food habits of three sharkspecies from the Mississippi Sound in the northern Gulfof Mexico. Southeast Nat 2: 271−280

Hoffmayer ER, Driggers WB, Jones LM, Hendon JM,Sulikowski JA (2013) Variability in the reproductive biology of the Atlantic sharpnose shark in the Gulf ofMexico. Mar Coast Fish 5: 139−151

Hueter RE, Heupel MR, Heist EJ, Keeney DB (2005) Evidence of philopatry in sharks and implications for themanagement of shark fisheries. J Northwest Atl Fish Sci35: 239−247

Hussey NE, Dudley SFJ, McCarthy ID, Cliff G, Fisk AT(2011) Stable isotope profiles of large marine predators: viable indicators of trophic position, diet, and movementin sharks? Can J Fish Aquat Sci 68: 2029−2045

Hussey NE, MacNeil MA, Olin JA, McMeans BC, KinneyMJ, Chapman DD, Fisk AT (2012) Stable isotopes andelasmobranchs: tissue types, methods, applications andassumptions. J Fish Biol 80: 1449−1484

Hynes HBN (1950) The food of fresh-water sticklebacks(Gasterosteus aculeatus and Pygosteus pungitius), with areview of methods used in studies of the food of fishes.J Anim Ecol 19: 36−58

Hyslop EJ (1980) Stomach contents analysis — a review ofmethods and their application. J Fish Biol 17: 411−429

Mar Ecol Prog Ser 550: 163–174, 2016174

Kinney MJ, Hussey NE, Fisk AT, Tobin AJ, SimpfendorferCA (2011) Communal or competitive? Stable isotopeanalysis provides evidence of resource partitioningwithin a communal shark nursery. Mar Ecol Prog Ser439: 263−276

Kiyashkoa SI, Velivetskayab TA, Ignatievb AV (2011) Sulfur,carbon, and nitrogen stable isotope ratios in soft tissuesand trophic relationships of fish from the near shorewaters of the Peter the Great Bay in the Sea of Japan.Ecology 37: 297−302

Leakey CDB, Attrill MJ, Jennings S, Fitzsimons MF (2008)Stable isotopes in juvenile marine fishes and their inver-tebrate prey from the Thames Estuary, UK, and adjacentcoastal regions. Estuar Coast Shelf Sci 77: 513−522

Lewis M, Jordan S, Chancy C, Harwell L, Goodman L, Quarles R (2007) Summer fish community of the coastalnorthern Gulf of Mexico: characterization of a large-scale trawl survey. Trans Am Fish Soc 136: 829−845

Nixon SW, Oviatt CA, Frithsen J, Sullivan B (1986) Nutrientsand the productivity of estuarine and coastal marine ecosystems. J Limnol Soc South Afr 12: 43−71

NOAA Fisheries (2015) Annual commercial landings statistics. http: //www.st.nmfs.noaa.gov/commercial-fisheries/ commercial-landings/annual-landings/index(accessed on 11 October 2015)

Papastamatiou YP, Lowe CG (2012) An analytical andhypothesis-driven approach to elasmobranch movementstudies. J Fish Biol 80: 1342−1360

Papastamatiou YP, Wetherbee BM, Lowe CG, Crow GL(2006) Distribution and diet of four species of car-charhinid shark in the Hawaiian Islands: evidence forresource partitioning and competitive exclusion. MarEcol Prog Ser 320: 239−251

Peterson BJ (1999) Stable isotopes as tracers of organic matter input and transfer in benthic food webs: a review.Acta Oecol 20: 479−487

Peterson BJ, Fry B (1987) Stable isotopes in ecosystem studies. Annu Rev Ecol Syst 18: 293−320

Pinkas L, Oliphant MS, Iverson ILK (1971) Food habits ofalbacore, bluefin tuna, and bonito in California waters.Calif Dept Fish Game Fish Bull 152

Post DM (2002) Using stable isotopes to estimate trophicposition: models, methods, and assumptions. Ecology 83: 703−718

Pulich WM, White WA (1991) Decline of submerged vegeta-tion in the Galveston Bay system — chronology and rela-tionships to physical processes. J Coast Res 7: 1125−1138

Rabalais NN, Turner RE, Diaz RJ, Justic D (2009) Globalchange and eutrophication of coastal waters. ICES J MarSci 66: 1528−1537

Rooker JR (1995) Feeding ecology of the schoolmaster snapper, Lutjanus apodus (Walbaum), from south -western Puerto-Rico. Bull Mar Sci 56: 881−894

Ross ST (1986) Resource partitioning in fish assemblages—a review of field studies. Copeia 1986: 352−388

Rozas LP, Zimmerman RJ (2000) Small-scale patterns ofnekton use among marsh and adjacent shallow nonveg-etated areas of the Galveston Bay Estuary, Texas (USA).Mar Ecol Prog Ser 193: 217−239

Thomas CJ, Cahoon LB (1993) Stable isotope analyses dif-ferentiate between different trophic pathways support-ing rocky-reef fishes. Mar Ecol Prog Ser 95: 19−24

Torres LG, Heithaus MR, Delius B (2006) Influence of teleostabundance on the distribution and abundance of sharksin Florida Bay, USA. Hydrobiologia 569: 449−455

Ubeda AJ, Simpfendorfer CA, Heupel MR (2009) Move-ments of bonnetheads, Sphyrna tiburo, as a response tosalinity change in a Florida estuary. Environ Biol Fishes84: 293−303

Wells RJD, Cowan JH, Fry B (2008) Feeding ecology of redsnapper Lutjanus campechanus in the northern Gulf ofMexico. Mar Ecol Prog Ser 361: 213−225

Editorial responsibility: Yves Cherel, Villiers-en-Bois, France

Submitted: November 25, 2015; Accepted: April 6, 2016Proofs received from author(s): May 4, 2016

➤

➤

➤

➤

➤

➤

➤

➤

➤

➤

➤

➤

➤

➤

➤

➤