Journal of Experimental Marine Biology and Ecology 393 (2010) 188–192

Contents lists available at ScienceDirect

Journal of Experimental Marine Biology and Ecology

j ourna l homepage: www.e lsev ie r.com/ locate / jembe

Non-lethal method to obtain stomach samples from a large marine predator and theuse of DNA analysis to improve dietary information

Adam Barnett a,b,⁎, Kevin S. Redd c, Stewart D. Frusher a, John D. Stevens b, Jayson M. Semmens a

a Tasmanian Aquaculture and Fisheries Institute, Marine Research Laboratories, University of Tasmania, Private Bag 49, Hobart, TAS 7001, Australiab CSIRO Marine and Atmospheric Research, GPO Box 1538, Hobart, TAS 7001, Australiac School of Zoology, University of Tasmania, Private Bag 5, Hobart, TAS 7001, Australia

⁎ Corresponding author. Tasmanian Aquaculture anResearch Laboratories, University of Tasmania, PrivatAustralia. Tel.: +61 3 6227 7275; fax: +61 6227 8035.

E-mail address: adam.barnett@utas.edu.au (A. Barne

0022-0981/$ – see front matter © 2010 Elsevier B.V. Adoi:10.1016/j.jembe.2010.07.022

a b s t r a c t

a r t i c l e i n f o

Article history:Received 16 February 2010Received in revised form 24 July 2010Accepted 27 July 2010

Keywords:DietGastric lavageNotorynchus cepedianusSharkStomach flushing

Dietary information of apex predators is crucial to understanding community dynamics and ecosystemprocesses. However, as dietary studies traditionally involve lethal sampling, obtaining this essentialinformation can have repercussions on predator populations and the structure and functioning of marineecosystems. With stronger emphasis being placed on conservation of species that are vulnerable tooverexploitation, the need for non-destructive methods of sampling is imperative, as is the requirement tomaximize the information obtained from each sample. Stomach flushing (gastric lavage) and DNA analysis ofstomach contents methods were tested on the broadnose sevengill shark Notorynchus cepedianus Peron1807. Acoustic tracking and recaptures of sharks implied high survivorship post-fishing and stomachflushing. From 85 prey items collected, 36 (43%) could be identified to species level using morphologicalanalysis. After DNA analysis, a further 35 items were identified to species level, doubling the informationobtained from these stomachs. The number of N. cepedianus that were confirmed to have eaten gummysharks Mustelus antarcticus Gunther 1870 also doubled after DNA analysis. Without DNA analysis (ofstomach contents) the importance of M. antarcticus in the diets of N. cepedianus would have beensubstantially underestimated. In addition, the non-lethal approach provides an opportunity to obtainmeaningful information from non-harvested, endangered or rare species or sampling of species withinprotected areas.

d Fisheries Institute, Marinee Bag 49, Hobart, TAS 7001,

tt).

ll rights reserved.

© 2010 Elsevier B.V. All rights reserved.

1. Introduction

Stomach content analysis is a key technique in animal ecology andfisheries research (Hyslop, 1980; Light et al., 1983; Hartleb andMoring, 1995). Dietary studies of fish traditionally involve lethalsampling and to ensure sampling integrity, large sample sizes areusually required to accurately measure a species' diet (Hartleb andMoring, 1995; Kamler and Pope, 2001). However, for many sharkspecies, there is growing concern about killing these animals, and theconservation of shark species is becoming a well publicized topic(Heupel and Simpfendorfer, in press). Nevertheless, the conservationstatus of most shark species is classified as being data deficient;highlighting the need for basic biological and ecological data to makeadequate conservation decisions (Heupel and Simpfendorfer, inpress). In addition, as many sharks are apex predators, dietaryinformation is also needed to help assess the role of these largepredators in marine communities, but accurate information is still

lacking for the majority of species (Braccini, 2008; Heithaus et al.,2008).

Therefore, while considering both conservation and the need forinformation, scientists are presently exploring non-lethal methodsof biological and ecological data collection (see Heupel andSimpfendorfer, in press). A non-destructive technique previouslyused in diet studies of fish is stomach flushing, also referred to asgastric lavage (Light et al., 1983). This technique involves pumpingwater via a tube down the throat of the animal into the stomach, andexpelling the stomach contents via the mouth.

Although well known amongst shark researchers, only one studyhas reported the use of stomach flushing to extract stomach contentsfrom a shark species (Medved, 1985). Sandbar sharks Carcharinusplumbeus were flushed to investigate gastric evacuation rates. Themethod was verified by dissecting 18 individuals after flushing, todiscover that all stomachs contained no food and very little water.Despite the technique showing promise, to date, no study hasaddressed the possible long term effects from stomach flushing.Other non-lethal methods have also been used to extract stomachcontents from elasmobranchs. For instance, forceps have been used toevert the stomachs of small sharks (b136 cm TL) (Schurdak andGruber, 1989; Cortes and Gruber, 1990;Webber and Cech, 1998; Bush,

189A. Barnett et al. / Journal of Experimental Marine Biology and Ecology 393 (2010) 188–192

2003). However, since this method involves either restraining theshark upside down or anaesthetizing the animal before reaching intoits mouth with forceps and pulling the stomach out, it is notappropriate for larger species. Thornback skates Raja clavata andlesser spotted dogfish Scyliorhinus canicula in captivity have also beeninjected with a emetic to induce stomach eversion (Andrews et al.,1998; Sims et al., 2000). Vomiting occurred up to 10 min after theinjection and in some cases the animal everted its stomach multipletimes (e.g. nine times for one skate) (Sims et al., 2000). However, thismethod is not appropriate for wild animals that cannot be kept intanks until they vomit. Therefore gastric lavage may be the mostpromising method for extracting stomach contents from large sharksin the field.

Regardless of the method used to gather stomach contents,accurately quantifying stomach contents from morphological charac-teristics often fails to achieve species level identifications due to a lackof hard remains, soft bodied prey andwell digested remains (Haywood,1995; Reñones et al., 2002). In many cases, prey can only be identifiedto a broad taxonomic category. These broad categories may not besufficient to give accurate information on species interactions andunderestimate the importance of certain taxa as prey. This isparticularly relevant for generalist predators or when a predatorpreys upon a large number of species from similar taxonomic groups(Symondson, 2002). In these cases, DNA based methods can improvethe probability of species level identification (Jarman et al., 2004).

DNA analysis has been used to identify prey from faecal materialfor a number of seal species (Purcell et al., 2004; Orr et al., 2004;Kvitrud et al., 2005; Parsons et al., 2005; Casper et al., 2007), whales(Jarman et al., 2002; Jarman et al., 2004), whale sharks (Jarman andWilson, 2004), lobsters (Redd et al., 2008) and penguins (Jarmanet al., 2002; Jarman et al., 2004; Deagle et al., 2007). Two studies havealso usedmolecular techniques to investigate the stomach contents ofpredatory fish (Rosel and Kocher, 2002; Smith et al., 2005). All thestudies above with the exception of Smith et al. (2005) (predatoryfish) and Deagle et al. (2007) (penguins) developed group-specificPCR primers to amplify specific target species. In general, this usuallyrequires some prior knowledge of the likely diet (Valentini et al.,2008). For predators with diverse diets, a universal primer approach(primers designed to amplify a wide range of taxa) is moreappropriate (Valentini et al., 2008). With the universal approachDNA sequences are normally run through a barcode database (e.g.GenBank) to see if they match with sequences previously deposited inthe system. For example, universal primers identified prey items frompredatory fish fromwestern equatorial Pacific with 95–100% accuracyusing GenBank (Smith et al., 2005). Both group-specific and universalprimer sets have also been used to study penguin diets (Deagle et al.,2007), and, bothmethods produced similar results, demonstrating theeffectiveness of universal primers in dietary analysis (Valentini et al.,2008). However, to avoid biased conclusions caused from universalprimers failing to amplify all prey species, samples should be analyzedwith multiple universal primer sets to allow cross validation (Deagleet al., 2007).

Another limitation to using barcoding data bases is the misiden-tification or non-identification due to the reference database notcontaining a comprehensive list of the species in a group that is beingstudied (Deagle et al., 2007; Valentini et al., 2008). However, withthe ever increasing number of sequence data (barcoding markers)continually added to databases and the improved quality andrigorous design of new databases (e.g. Barcode of Life Data Systems,BOLD) this problem should be negligible for future DNA barcodedietary studies (see Valentini et al., 2008 for a review on the use ofbarcoding in ecology).

This study aims to test the use of stomach flushing to acquiredietary samples for sub-adult and adult (size range 150–290 cm TL)sevengill sharks Notorynchus cepedianus and evaluate post-flushingsurvival rates. The use of universal primers for DNA dietary analysis to

improve the quantity and quality of dietary information was alsoinvestigated for this species.

2. Methods

2.1. Field methods

Stomach samples were collected from the Derwent Estuary andNorfolk Bay south east Tasmania, Australia (43.00°S; 147.76°E).N. cepedianus were caught using bottom-set longlines that were setfor 4–6 h. Soak times of 4 h were used as it normally takes 2–4 h toattract a number of sharks to the lines (Pers. obs.). Once landed, theirstomachs were flushed, total length measured, and they were taggedin the dorsal fin with plastic Jumbo tags (Daltons, Henley-on-Thames,England) (20 individuals were also acoustically tagged, see below fordetail) and returned to the water.

Stomach flushing was undertaken by restraining the shark, while aplastic hose (~3 cm diameter) attached to a submersible electric pumpwas inserted through the mouth into the stomach. Seawater waspumped into the shark's stomach. Once the stomach was filled withwater, which was evident by the expansion of the stomach region, thehosewas removed and gentle pressure applied to the abdominal regioncausing the water and any food items to be regurgitated (~3 min perflush). Any material regurgitated was collected in a sieve, bagged,labeled and placed on ice for subsequent morphological and molecularanalysis. Ten sharks were dissected after flushing to validate theeffectiveness of this flushingmethod. If prey species that are vulnerableto capture from longlines showed no signs of digestion, they werejudged to be recently ingested and were excluded from analysis due tothe likelihood that they were eaten from the longline.

2.2. Survival after capture and stomach flushing procedures

To assess the survivorship of sharks after fishing and stomachflushing, we recorded the number of sharks recaptured. In addition tothe use of conventional tags (recapture data), 10 individuals weretagged with acoustic transmitters after being stomach flushed, andtheir post-release survivorship compared to 10 other individuals thatwere also tagged but not stomach flushed. Acoustically tagged sharkswere released within an acoustic monitoring array consisting of 72receivers deployed throughout the study area. Coded acoustictransmitters (VEMCO Ltd., Halifax, Canada) were inserted though a~2 cm incision in the abdominal wall into the body cavity. The incisionwas closed with a surgeon's suture. The entire procedure wasnormally accomplished in 3 to 5 min. Running water was pumpedover the shark's gills throughout the procedure to ensure the gillsremained wet.

2.3. Dietary analysis

In the laboratory, stomach contents were identified to their lowesttaxonomic level using morphological characteristics. For stomachsamples containing prey that was sectioned in multiple pieces(n=38), we first had to determine the number of prey. As withmost dietary studies, there was the potential for some inaccuracy indetermining the number of prey at this stage, however, we used amethodology aimed at minimising the risk of identifying multiplepieces of the same prey as separate prey items. This was done bypiecing body parts together and looking for changes in the size orrepetition of body parts (e.g. two sets of jaws would indicate two ofthat particular prey species present or disproportionate body partswould indicatemore than one individual). Stomach contents that onlycontained a small amount of a single tissue type e.g. b50 g of musclewere assumed to be from a single animal. For prey that could not bepositively identified to species level, a small piece of tissue(approximately 2 mm3) was removed for molecular analysis. This

190 A. Barnett et al. / Journal of Experimental Marine Biology and Ecology 393 (2010) 188–192

was cut from the centre of the prey remains (all tissue types) with asterile blade to reduce the chance of cross contamination from otherstomach contents.

DNA was extracted from tissue samples using a Qiagen DNeasy Kit(Qiagen, Inc.). The sample DNA was amplified by PCR using threeuniversal primer sets from different mitochondrial regions (Table 1);CO1 barcoding primers (Fish F1/R1 and Fish F2/R2), and universal 16sprimers (LRN-13398/LRJ-12887) (Kasper et al., 2004; Ratnasinghamand Hebert, 2007). These primer sets yielded PCR products between556 bp and 776 bp in length. The components of these 25 μL PCRswere; 5 μL 1× KAPA HiFi Fidelity Buffer, 0.75 mM MgCl2, 0.75 mMdNTPs, 0.75 mM forward primer, 0.75 mM reverse primer, 0.5 μL KAPAHiFi DNA polymerase (2.5 Units) and 3 μL template DNA (~1–5 ng).PCRwas conducted using aMJ Research Tetrad thermocycler under thefollowing conditions: initial denaturation and polymerase activation at98 º C for 20 s followed by 35 cycles of 94 °C for 45 s, annealingtemperature for 45 s, 72 °C for 30 s; and a final extension at 72 °C for1 min.

Prey sequences were analyzed using the NCBI (GenBank)Nucleotide Database by BLAST searching as well as the Barcode ofLife (BOLD) Database using the Identify Specimen utility. Thesedatabases represent the most comprehensive list of DNA sequencescurrently available and the fish fauna of southern Australia areparticularly well represented (Ward et al. 2005). We considered 95%certainty as a strong match based on a previous study on predatoryfish diets (Smith et al., 2005) and the relatively low diversity ofexpected species within taxonomic groups (e.g. ~5 species of sharks).If a database matched more than one species with a high certainty,species distribution was used to select the correct match. For exampleone sample recognized Mustelus lenticulatus and Mustelus antarcticusas equally strong matches (95%), however, M. lenticulatus is endemicto New Zealand and M. antarcticus is the only Mustelus species foundin the Tasmanian region (Compagno et al., 2005).

The percent index of relative importance (% IRI) was calculated asper Cortes (1997) to examine diet prior and post DNA analysis.

3. Results

Stomach flushing was conducted on 100 N. cepedianus individuals,of these 50 contained prey. Of the 50 that did not contain prey, 10 hadeverted stomachs. An additional ten stomach samples were obtainedfrom fisherman. From these 60 stomachs a total of 85 individual preyitems were collected.

Eight of the 10 sharks flushed and subsequently dissected tovalidate the flushing method had consumed prey. Flushing success-fully removed the entire stomach contents from seven of the eightsharks. For one individual, a piece ofM. antarcticus (~8 cm diameter)was still present, although five other items including threecephalopod beaks were successfully evacuated. The remaining twosharks were dissected to see if flushing had failed to remove theprey. Dissection confirmed that these sharks had empty stomachswhen flushed.

Table 1PCR primers used (Kasper et al., 2004; Ratnasingham and Hebert, 2007) includingprimer name, sequence and annealing temperature used in DNA amplification.

Primer Sequence Annealingtemp.

Targetregion

Fragmentlength

Fish F2 (5′-TCGACTAATCATAAAGATATCGGCAC-3′) 54 °C CO1 655 bpFish R2 (5′-ACTTCAGGGTGACCGAAGAATCAGAA-3′) 54 °C CO1 655 bpFish F1 (5′-TCAACCAACCACAAAGACATTGGCAC-3′) 54 °C CO1 655 bpFish R1 (5′-TAGACTTCTGGGTGGCCAAAGAATCA-3′) 54 °C CO1 655 bpLRN-13398

(5′-CGCCTGTTTATCAAAAACAT-3′) 52 °C 16s 550–750 bp

LRJ-12887

(5′-CCGGTCTGAACTCAGATCACG-3′) 52 °C 16s 550–750 bp

To test the accuracy of the molecular analysis, 15 prey (from 7species) collected from stomach samples that could be positivelyidentified by morphological analysis were subjected to the molecularprey identification protocol described above. DNA analysis correctlyidentified all 15 samples to the species level. No disagreements inspecies identification between databases (GenBank and BOLD) werefound during the course of this experiment.

Morphological analysis identified 36 of the 85 (42%) prey items tospecies level. Due to the advanced stage of digestion or a minimalamount of tissue present, the majority of the remaining prey couldonly be classified to broader groups such as shark or teleost. After DNAanalysis, 85% of the total items were identified to species level and 13prey samples remained unidentified. Of these, 11 samples consisted ofhard tissue parts. No DNA could be extracted from fish lenses (n=2)or cephalopod beaks (n=4), and extraction was only successful forfour of the eight teleost skeletal elements. Only one of the sevenelasmobranch cartilage samples did not amplify.

Prior to DNA analysis, 24 dietary samples were categorized asunidentifiable elasmobranch. However, DNA analysis identified 23ofthe samples to the species level. The use of DNA analysis significantlyincreased species identification and showed which elasmobranchspecies are prominent in N. cepedianus diets (Table 2). For example,morphological analysis showed that M. antarcticus occurred in 17% ofthe 60 stomachs sampled. DNA analysis increased the resolution,showing thatM. antarcticuswas present in 35% of these stomachs andincreased the %IRI from 25 to 72% (Table 2). Morphological analysisidentified marine mammals in eight of the stomach samples. Five ofthese samples were suspected of being cetaceans and three pinnipeds.Molecular results agreed with this broad estimation, but furtheridentified the cetaceans as bottlenose dolphins, Tursiops truncatus,and the pinnipeds as Australian fur seals, Arctocephalus pusillus(Table 2). This resulted in T. truncatus %IRI changing from 0 to 8%(Table 2). Overall, seven of the 22 species consumed in the study wereonly identified by DNA analysis and six of these were teleosts(Table 2).

The survivorship of sharks post-fishing and stomach flushingappeared to be high. Of the 100 sharks that were stomach flushed, 13were subsequently recaptured by research longlining and five werecaught by fishers. The time at liberty for recaptures ranged from 11 to715 days. All 20 acoustic tagged sharks (including the 10 that werestomach flushed)were detectedmoving in and out of the array area inthe subsequent months (receivers deployed 18 months) after tagging,suggesting high post-flushing survivorship.

4. Discussion

Stomach flushing is a simple and effective method for obtainingdietary samples from N. cepedianus without sacrificing animals. Thesubsequent recapture of 18% of flushed sharks, which is a relativelyhigh percent in shark studies (common for b5%, see Kohler andTurner, 2001 for review), and the detection of all acoustic taggedindividuals implied high post-fishing and stomach flushing survivor-ship. The successful application of this method by Medved (1985)confirms that it is appropriate to obtain stomach samples for at leasttwo species of sharks. However, similar results are foreseeable forother elasmobranchs, as many species are able to evert their stomachswith no apparent ill effects, suggesting that the gut morphology iswell suited for easy and fast extraction of food in this manner(Schurdak and Gruber, 1989; Sims et al., 2000; Simpfendorfer et al.,2001; Brunnschweiler et al., 2005). Further studies on other speciesare required to test this hypothesis. The method is easily applied andthe main concern for future work on other species is to adjust thewater pressure from the hose to suit the size of the animals beingflushed to avoid possible damage.

A disadvantage of the stomach flushing method is the possibilitythat not all dietary items in the stomach are evacuated during the

Table 2Number of samples positively identified by their morphology (Mor) and the number ofsamples that were positively identified by DNA analysis that were not identified bymorphology, (% DNA is the increased identification of each species from using DNAanalysis). The index of relative abundance prior (% IRI) and post (% IRI DNA)DNA analysis.

Species Common name Total Mor DNA %DNA

%IRI

% IRIDNA

ChondrichthyansMustelus antarcticus Gummy shark 21 10 11 52 25 72Galeorhinus galeus School shark 3 1 2 66 0.2 1Squalus acanthias Whitespotted

spurdog7 4 3 43 2 5

Myliobatis australis Southern eagleray

4 3 1 25 2 3

Spiniraja whitleyi Melbourne skate 4 1 3 75 0.2 2Urolophus cruciatus Banded stingaree 3 3 0 – 1 1Notorynchuscepedianus

Broadnosesevengill shark

3 2 1 33 2 3

Pristiophorusnudipinnis

Southernsawshark

1 1 0 – 0.2 0.3

Callorhinchus milii Elephant fish 4 2 2 50 0.4 2Cephaloscylliumlaticeps

Draughtboardshark

2 1 1 50 0.1 0.4

MammalsArctocephalus pusillus Australian fur

seal3 2 1 33 1 1

Tursiops truncatus Bottlenosedolphin

5 0 5 100 0 8

CephalopodNotodarus gouldi Goulds squid 2 2 0 – 0.3 0.3

TeleostsArripis trutta Australian

salmon1 1 0 – 0.1 0.1

Leptatherinapresbyteroides

Silver fish 1 0 1 100 0 0.1

Thyrsites atun Barracouta 1 1 0 – 0.1 0.1Anguilla australis Shortfinned eels 1 0 1 100 0 0.1Conger verreauxi Conger eel 2 0 2 100 0 0.1Latridopsis forsteri Bastard

trumpeter1 0 1 100 0 0.1

Emmelichthys nitidus Redbait 1 0 1 100 0 0.1Platycephalusbassensis

Sand flathead 1 1 0 – 0.1 0.1

Engraulis australis Australiananchovy

2 0 2 100 0 0.1

191A. Barnett et al. / Journal of Experimental Marine Biology and Ecology 393 (2010) 188–192

sampling protocol. However, in this study, only one dietary item wasnot removed from the 10 sharks that were flushed and then dissected,showing that this disadvantage is likely to be minor. The conservationbenefits of non-lethal shark research outweigh any disadvantagesfrom obtaining some incomplete dietary samples and can becompensated for with large sample sizes.

Many of the shark prey collected in the present study were lackingkey diagnostic features, making identification difficult. This is notuncommon in sharks such as N. cepedianus, which tend to slice uptheir prey into pieces prior to swallowing (Ebert, 1991; Braccini,2008). In addition, several individuals have also been observedfeeding on the same prey item (Ebert, 1991), in which case only apiece of flesh may be evident in the stomach sample. Without the useof molecular techniques, over 50% of N. cepedianus dietary sampleswould have only been identified to broad scale groupings, such asfamily or class. Similar results were reported when using DNAmethods to identify prey from Pacific Harbour seal Phoca vitulinarichardsi and Australian fur seal scats (Orr et al., 2004; Casper et al.,2007).

In addition to identifying the diversity of species in N. cepedianusdiet, the increased taxonomic resolution also provided improvedinformation on the frequency of occurrence of species in the diet andtherefore the importance of species-specific trophic linkages.Morphological analysis had identified M. antarcticus as a relatively

important prey species of N. cepedianus, however DNA analysisshowed that N. cepedianus had consumed twice the number of M.antarcticus, further emphasising the importance of predator–preylink between these species. Other studies have used DNA analysis toinvestigate specific predator–prey relationships. For example,several studies used DNA analysis to delineate abundant and raresalmon species within harbour seal diets to gauge the likely impactof pinnipeds on commercial and endangered salmon species (Orret al., 2004; Purcell et al., 2004; Kvitrud et al., 2005). Molecularmethods have also demonstrated the possible impact of predatoryfish on the recovery of commercial cod species from over fishing(Rosel and Kocher, 2002).

Extraction and amplification techniques in this study were used tocover a broad array of prey species and tissue types. Future studieswhere predators contain predominately hard parts in their stomachcontents may benefit frommolecular analyses using methods tailoredfor specific tissues and prey group (Deagle et al., 2007). For example asignificant proportion of unidentifiable teleosts present in shortfinmako shark Isurus oxyrinchus stomachs were eye lenses (Maia et al.,2006). Here, the use of specific DNA methods designed for hard bonetype tissues may have substantially increased prey identification.

The present study demonstrates that stomach flushing had asimilar efficiency in obtaining dietary samples as destructive samplingfor N. cepedianus (Ebert, 2002; Lucifora et al., 2005; Braccini, 2008).This technique could potentially be used with other elasmobranchspecies and future studies should test its suitability prior to sacrificinganimals for dietary information. In particular, the method may beapplicable where destructive sampling is not allowed (marineprotected areas) or where it is potentially unethical (threatened andendangered species). By undertaking molecular analysis of unidenti-fiable prey, the number of species with precise identification wasdoubled. This results in a better understanding of both the diversity ofprey items consumed and also the importance of specific prey speciesto N. cepedianus diet. The use of both non-destructive and fine-scaletaxonomic resolution dietary methods that can help identify theecological roles of these top level predators is important forconservation and management of marine ecosystems.

Acknowledgements

J. Yick and A. Pender are acknowledged for field assistance and M.Bell for additional stomach samples. This study was supported bygrants from the Save Our Seas Foundation, Winifred Violet ScottFoundation and the Holsworth Wildlife Research Endowment. Allresearch was conducted with approval from the University ofTasmania Animal Ethics Committee (#A0009120) and the Depart-ment of Primary Industries and Water (Permit # 8028). [RH]

References

Andrews, P.L.R., Sims, D.W., Young, J.Z., 1998. Induction of emesis by the sodiumchannel activator veratrine in the lesser spotted dogfish, Scyliorhinus canicula(Chondrichthyes: Elasmobranchii). J. Mar. Biol. Assoc. UK 78, 1269–1279.

Braccini, J.M., 2008. Feeding ecology of two high-order predators from south-easternAustralia: the coastal broadnose and the deepwater sharpnose sevengill sharks.Mar. Ecol. Prog. Ser. 371, 273–284.

Brunnschweiler, J.M., Andrews, P.L.R., Southall, E.J., Pickering, M., Sims, D.W., 2005.Rapid voluntary stomach eversion in a free-living shark. J. Mar. Biol. Assoc. UK 85,1141–1144.

Bush, A., 2003. Diet and diel feeding periodicity of juvenile scalloped hammerheadsharks, Sphyrna lewini, in Käne'ohe Bay, Oahu, Hawaii. Environ. Biol. Fishes 67,1–11.

Casper, R.M., Jarman, S.N., Gales, N.J., Hindell, M.A., 2007. Combining DNA andmorphological analyses of faecal samples improves insight into trophic interactions:a case study using a generalist predator. Mar. Biol. 152, 815–825.

Compagno, L.J.V., Dando, M., Fowler, S., 2005. Sharks of the World. HarperCollins,London.

Cortes, E., 1997. A critical review of methods of studying fish feeding based on analysisof stomach contents: application to elasmobranch fishes. Can. J. Fish. Aquat. Sci. 54,726–738.

192 A. Barnett et al. / Journal of Experimental Marine Biology and Ecology 393 (2010) 188–192

Cortes, E., Gruber, S.H., 1990. Diet, feeding-habits and estimates of daily ration of younglemon Sharks, Negaprion brevirostris (Poey). Copeia 1, 204–218.

Deagle, B.E., Gales, N.J., Evans, K., Jarman, S.N., Robinson, S., Trebilco, R., Hindell, M.A.,2007. Studying seabird diet through genetic analysis of faeces: a case study onmacaroni penguins (Eudyptes chrysolophus). PLoS ONE 2 (9), e831.

Ebert, D.A., 1991. Observations on the predatory behaviour of the sevengill sharkNotorynchus cepedianus. S. Afr. J. Mar. Sci. 11, 455–465.

Ebert, D.A., 2002. Ontogenetic changes in the diet of the sevengill shark (Notorynchuscepedianus). Mar. Freshwater Res. 53, 517–523.

Hartleb, C.F., Moring, J.R., 1995. An improved gastric lavage device for removingstomach contents from live fish. Fish. Res. 24, 261–265.

Haywood, M.D.E., 1995. Rates at which post-larval prawns are digested by predatoryfish and the implications for predation studies. J. Fish Biol. 47, 337–340.

Heithaus, M.R., Frid, A., Wirsing, A.J., Worm, B., 2008. Predicting ecological consequencesof marine top predator declines. Trends Ecol. Evol. 23, 202–210.

Heupel, M.R., Simpfendorfer, C.A., in press. Science or slaughter: need for lethalsampling of sharks. Conserv. Biol.

Hyslop, E.J., 1980. Stomach contents analysis — a review of methods and theirapplication. J. Fish Biol. 17, 411–429.

Jarman, S.N., Wilson, S.G., 2004. DNA-based species identification of krill consumed bywhale sharks. J. Fish Biol. 65, 586–591.

Jarman, S.N., Gales, N.J., Tierney, M., Gill, P.C., Elliott, N.G., 2002. A DNA-based methodfor identification of krill species and its application to analysing the diet of marinevertebrate predators. Mol. Ecol. 11, 2679–2690.

Jarman, S.N., Deagle, B.E., Gales, N.J., 2004. Group-specific polymerase chain reaction forDNA-based analysis of species diversity and identity in dietary samples. Mol. Ecol.13, 1313–1322.

Kamler, J., Pope, K.L., 2001. Nonlethal methods of examining fish stomach contents. Rev.Fish. Sci. 9, 1–11.

Kasper, M.L., Reeson, A.F., Cooper, S.J.B., Perry, K.D., Austin, A.D., 2004. Assessment ofprey overlap between a native (Polisties humilis) and an introduced (Vespulagermanica) social wasp using morphology and phylogenetic analysis of 16s rDNA.Mol. Ecol. 13, 2037–2048.

Kohler, N.E., Turner, P.A., 2001. Shark tagging: a review of conventional methods andstudies. Environ. Biol. Fishes 60, 191–224.

Kvitrud, M.A., Riemer, S.D., Brown, R.F., Bellinger, M.R., Banks, M.A., 2005. Pacific harborseals (Phoca vitulina) and salmon: genetics presents hard numbers for elucidatingpredator–prey dynamics. Mar. Biol. 147, 1459–1466.

Light, R., Adler, P., Arnold, D., 1983. Evaluation of gastric lavage for stomach analyses.N. Am. J. Fish. Manage. 3, 81–85.

Lucifora, L.O., Menni, R.C., Escalante, A.H., 2005. Reproduction, abundance and feedinghabits of the broadnose sevengill shark Notorynchus cepedianus in north Patagonia,Argentina. Mar. Ecol. Prog. Ser. 289, 237–244.

Maia, A., Queiroz, N., Correia, J.P., Cabral, H., 2006. Food habits of the shortfin mako,Isurus oxyrinchus, off the southwest coast of Portugal. Environ. Biol. Fishes 77,157–167.

Medved, R.J., 1985. Gastric evacuation in the sandbar shark, Carcharhinus plumbeus.J. Fish Biol. 26, 239–253.

Orr, A.J., Banks, A.S., Mellman, S., Huber, H.R., DeLong, R.L., Brown, R.F., 2004.Examination of the foraging habits of Pacific harbor seal (Phoca vitulina richardsi)to describe their use of the Umpqua River, Oregon, and their predation onsalmonids. Fish. Bull. 102, 108–117.

Parsons, K.M., Piertney, S.B., Middlemas, S.J., Hammond, P.S., Armstrong, J.D., 2005.DNA-based identification of salmonid prey species in seal faeces. J. Zool. 266,275–281.

Purcell, M., Mackey, G., LaHood, E., Huber, H., Park, L., 2004. Molecular methods for thegenetic identification of salmonid prey from Pacific harbor seal (Phoca vitulinarichardsi) scat. Fish. Bull. 102, 213–220.

Ratnasingham, S., Hebert, P.D.N., 2007. BOLD: the barcode of life data system. Mol. Ecol.Notes 7, 355–364.

Redd, K.S., Jarman, S.N., Frusher, S.D., Johnson, C.R., 2008. A molecular approach toidentify prey of the southern rock lobster. Bull. Entomol. Res. 98, 233–238.

Reñones, O., Polunin, N.V.C., Goni, R., 2002. Size related dietary shifts of Epinephelusmarginatus in a western Mediterranean littoral ecosystem: an isotope and stomachcontent analysis. J. Fish Biol. 61, 122–137.

Rosel, P.E., Kocher, T.D., 2002. DNA-based identification of larval cod in stomachcontents of predatory fishes. J. Exp. Mar. Biol. Ecol. 267, 75–88.

Schurdak, M., Gruber, S.H., 1989. Gastric evacuation of the lemon shark Negaprionbrevirostris (Poey) under controlled conditions. Exp. Biol. 48, 77–82.

Simpfendorfer, C.A., Goodreid, A.B., McAuley, R.B., 2001. Size, sex and geographicvariation in the diet of the tiger shark, Galeocerdo cuvier, from Western Australianwaters. Environ. Biol. Fishes 61, 37–46.

Sims, D.W., Andrews, P.L.R., Young, J.Z., 2000. Stomach rinsing in rays. Nature 405, 38.Smith, P.J., McVeagh, S.M., Allain, V., Sanchez, C., 2005. DNA identification of gut

contents of large pelagic fishes. J. Fish Biol. 67, 1178–1183.Symondson, W.O.C., 2002. Molecular identification of prey in predator diets. Mol. Ecol.

11, 627–641.Valentini, A., Pompanon, F., Taberlet, P., 2008. DNA barcoding for ecologists. Trends

Ecol. Evol. 24, 110–117.Ward, R.D., Zemlak, T.S., Innes, B.H., Last, P.R., Hebert, P.D.N., 2005. DNA barcoding

Australia's fish species. Philos. Trans. R. Soc. London, Ser. B. 360, 1847–1857.Webber, J.D., Cech, J.J., 1998. Nondestructive diet analysis of the leopard shark from two

sites in Tomales Bay, California. Calif. Fish Game 84, 18–24.