lable at ScienceDirect

Animal Behaviour 81 (2011) 1039e1048

Contents lists avai

Animal Behaviour

journal homepage: www.elsevier .com/locate/anbehav

Site fidelity and sex-specific migration in a mobile apex predator:implications for conservation and ecosystem dynamics

Adam Barnett a,b,*, Kátya G. Abrantes a,c, John D. Stevens b,1, Jayson M. Semmens a

a Tasmanian Aquaculture and Fisheries Institute, Marine Research Laboratories, University of TasmaniabCSIRO Marine and Atmospheric ResearchcCoastal and Estuary Ecosystems, School of Marine and Tropical Biology, James Cook University, Australia

a r t i c l e i n f o

Article history:Received 11 October 2010Initial acceptance 4 December 2010Final acceptance 4 February 2011Available online 16 March 2011MS. number: 10-00702R

Keywords:acoustic telemetryestuaryhabitat partitioningmovement behaviourNotorynchus cepedianuspredatorepreyprotected area managementsatellite telemetryshark

* Correspondence: A. Barnett, Marine Research LHobart, TAS 7001, Australia.

E-mail address: adam.barnett@utas.edu.au (A. Bar1 J. D. Stevens is at CSIRO Marine and Atmosphe

Hobart, TAS 7001, Australia.

0003-3472/$38.00 � 2011 The Association for the Studoi:10.1016/j.anbehav.2011.02.011

Combining movement behaviour with other ecological information of predators and their prey isessential for an adequate understanding of ecosystem dynamics. The movement patterns of broadnosesevengill sharks, Notorynchus cepedianus , were monitored with acoustic and satellite technology incoastal areas of southeast Tasmania, Australia. Individuals were tagged in two habitats (Norfolk Bay andthe Derwent Estuary) for which we had ecological information such as diet, population structure andabundance. Notorynchus cepedianus showed seasonal site fidelity in the use of the coastal habitats. Thegeneral pattern was for sharks to exit coastal areas over winter and females to return the followingspring and males in summer. Their movement into these coastal areas coincided with high seasonalabundance of their known prey species during summer, suggesting feeding site fidelity. Individualstagged in two coastal areas showed low spatial and dietary overlap, suggesting localized site fidelity andfine spatial scale resource partitioning. This has rarely been reported for large mobile predators. Bothsatellite and acoustic methods showed that males make northerly migrations during winter to distancesof at least 1000 km. The combined use of tracking, diet and abundance information demonstrated thatN. cepedianus are likely to exert significant predation pressure on prey inhabiting these areas duringsummer. Overall, this study highlights the benefit of complementing movement data with otherecological information to understand the habitat use of large mobile predators and their potentialinfluences on ecosystem structure and function.� 2011 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.

Recently, studies have started to address the complex ecologicalroles of large predators with wide distributions (Ferretti et al.2010), but for the vast majority of these species, their roles areyet to be defined and quantified (Estes et al. 2001; Williams et al.2004; Ferretti et al. 2010). Because many large predators uselarge areas they can be important predators of multiple species ina number of systems and/or habitats (Ferretti et al. 2010). There-fore, characterization of spatiotemporal habitat use is crucial forunderstanding how large predators affect the structure and func-tion of the different systems they use (Williams et al. 2004; Austinet al. 2006; Smout & Lindstrom 2007) and for the conservation ofboth the predators and the systems of which they are integralcomponents (Jorgensen et al. 2009).

aboratories, Private Bag 49,

nett).ric Research, GPO Box 1538,

dy of Animal Behaviour. Published

Many large marine vertebrates that move over large spatialscales annually restrict their movement to smaller specific areas.Site fidelity, defined as the return to and reuse of an area over time,has mainly been attributed to mating, parturition and feeding(Switzer 1993; Speed et al. 2010). For example, numerous species ofturtles, seals and penguins migrate seasonally between specificbreeding and foraging areas (Lloyd et al. 1995; Stewart 1997;Charrassin & Bost 2001; Myers & Hays 2006; Lee et al. 2007).Killer whales, Orcinus orca, minke whales, Balaenoptera acutoros-trata, bluefin tuna, Thunnus thynnus, and a number of largemigratory sharks (e.g. white sharks, Carcharodon carcharias, tigersharks, Galeocerdo cuvier, and salmon sharks, Lamna ditropis)annually move to exploit seasonally abundant prey at aggregationsites (Lowe et al. 2006; Smout & Lindstrom 2007; Weng et al. 2008;Jorgensen et al. 2009; Walli et al. 2009; Foote et al. 2010; Meyeret al. 2010). Identification of these specific areas is often theinitial step in understanding habitat use of large mobile predators,and for formulating appropriate management and conservationstrategies (Martin et al. 2007).

One of the main conservation strategies is to protect habitatscritical to the survival of a species (Heupel et al. 2007). For shark

by Elsevier Ltd. All rights reserved.

A. Barnett et al. / Animal Behaviour 81 (2011) 1039e10481040

species, most research on critical habitats has focused on relativelysmall species or the use of nursery areas by juveniles. However,given that protecting nursery areas has only limited value inconserving shark populations (Kinney & Simpfendorfer 2009),there is a distinct need to understand when and how older lifestages are using coastal areas (Simpfendorfer & Heupel 2004;Speed et al. 2010). Similar scenarios are apparent for other largeslow-growing vertebrates such as sea turtles (Heppell et al. 1996;Mazaris & Matsinos 2006). For example, to sustain a healthy pop-ulation of sea turtles, it is more important to ensure the health ofadults than the survival of hatchlings.

The broadnose sevengill shark, Notorynchus cepedianus, isa large (up to 3 m) coastal-associated apex predator with a widetemperate distribution (Last & Stevens 2009). This species’ trophicposition rivals that of other species considered important uppertrophic predators, such as tiger sharks and white sharks (Cortés1999). Yet, in contrast to the latter two shark species, besides die-tary work there is very little information on the ecology or move-ment behaviour of N. cepedianus. This is a significant gap inunderstanding the behaviour of this shark and, in general, the roleapex predators such as this have in the coastal systems they inhabit.For other large mobile marine species such as pinnipeds, cetaceansand turtles, there is a better understanding of where and why theymove in coastal areas (Kuhn et al. 2009). Considering that many ofthese mesopredator species are either prey and/or competitors ofsharks in coastal systems, the lack of information on habitat andresource use for species such as N. cepedianusmeans an incompleteunderstanding of interspecies dynamics (e.g. competition andpredation) and the ecological role of co-occurring predators incoastal systems (Ferretti et al. 2010).

Although movement information is useful by itself, the incor-poration of movement behaviour with other ecological informationis necessary for a better understanding of ecosystem dynamics andspecies interactions. However, to date, not many spatial studies ofsharks have included information on other aspects of shark ecologysuch as diet or reproduction (Simpfendorfer & Heupel 2004; Speedet al. 2010). Recently, in an intensive study on the ecology ofN. cepedianus in two coastal habitats of southeast Tasmania,Australia, information on its diet, population structure, preyabundance and seasonality in abundance was collected (Barnett

N3 0 3 6 km

J

P

St

B

A

R

Derwent

Ralphs Bay

D5D4

D3

D1D2

D10D7

D9D8

D6

Figure 1. Map of study area. Filled circles represent receivers and each curtain or gate of rec

et al. 2010a, b).The addition of movement data makes this anideal case study to examine resource area use of apex predators intwo distinct coastal habitats in close proximity.

Our study had four aims. (1) The first was to test the hypothesisthat N. cepedianus display seasonality and site fidelity in the use ofcoastal areas. Catch rate information suggests a seasonal abun-dance of N. cepedianus in coastal areas, as catches decrease signif-icantly in winter (Barnett et al. 2010b). In addition, tag andrecapture data indicate that sharks show site fidelity to thesecoastal areas over multiple years (Barnett et al. 2010b). (2) Becauseprevious studies indicate relatively low dietary overlap betweensharks caught in two coastal locations (Barnett et al. 2010a), andsince individuals were mostly recaptured in the location they wereoriginally tagged (Barnett et al. 2010b), our second hypothesis wasthat N. cepedianus show fine-scale habitat partitioning and sitefidelity in these areas. (3) Since catch rates also indicate that femaleabundance increases in spring, whereas males are not caught inlarge numbers until summer (Barnett et al. 2010b), we predictedthat there is sexual segregation in the use of these areas. (4) Ourfinal aim was to evaluate the effectiveness of shark protected areasin southeast Tasmania.

METHODS

Study Area and Acoustic Receivers

An array of 74 acoustic receivers was deployed in coastal areas ofsoutheast Tasmania, Australia (VR2 receivers, VEMCO Ltd, Halifax,Canada; Fig. 1). Notorynchus cepedianus individuals were tagged intwo locations, Norfolk Bay and the Derwent Estuary (Fig. 1). NorfolkBay is a relatively shallow (average depth 15 m; maximum depth20 m), semi-enclosed bay with an area of 176 km2. The DerwentEstuary runs through the City of Hobart before opening into StormBay, and consistently reaches depths of 20e30 m, with a maximumdepth of 44 m. The acoustic array was set up as single curtains andgates, so that sharks could be detected moving across and betweenentrances and choke points of the coastal areas. With this design,an individual would need to be detected on two subsequentcurtains to confirm that it had entered a given area.

ittwater

PW

DC

FHBayE

F X

Norfolk Bay

H

GK

LEH

orm Bay

eivers has a designated code. Dashed line is the boundary of the shark protected areas.

A. Barnett et al. / Animal Behaviour 81 (2011) 1039e1048 1041

Range testing was conducted at a number of locations in NorfolkBay and Derwent Estuary to determine the detection radius of theacoustic receivers. From these tests, 100% of the tag transmissionswere detected within a radius of 400e500 m. Therefore, allreceivers were stationed at a maximum distance of 800 m apart sothat the detection radii slightly overlapped. In choke points, a singlereceiver was placed halfway between the margins, when thedistance between shorelines was less than 800 m (e.g. receiverD3eD10, Fig. 1). Since the study area encompasses shark protectedareas, receivers were deployed along the boundary of the protectedareas to determine the occurrence of N. cepedianus there (Fig. 1).The entire array was deployed from December 2007 to June 2009,with only the receivers at the entrance to Norfolk Bay (Line E; Fig. 1)and the Derwent Estuary (Line B) deployed until June 2010.

Animal Capture and Transmitter Deployment

Sharks were caught on bottom-set longlines (Barnett et al. 2010b)and fitted with acoustic-coded V16 transmitters (90 mmlength � 15mm diameter; <1% of the animal’s body weight) withbattery lives >2 years (N¼ 9) or >5 years (N¼ 34; VEMCO Ltd).Individualswere positioned into a cradle and then lifted on to the deckof the boat. Running water was pumped over the gills while aboard.The acoustic tagwas implanted into the peritoneal cavity via a 1e2 cmincision in the abdominal wall, and the incision closed by stitchinga surgeon’s suture using surgical needle and thread. Aseptic tech-niqueswereusedduring surgeryand theprocedurewas accomplishedin 3e5min, after which sharks were lowered back into the water.

Sharks were tagged during two periods. In the first year,33 sharks were tagged between December 2007 and March 2008(20 in Norfolk Bay and 13 in the Derwent Estuary; 24 females, ninemales). An additional 10 sharks, five for each location, were taggedbetween 17 October and 19 November 2008. In total, 43 sharkswere tagged (31 females and 12 males). Tagging in the DerwentEstuary was conducted between receivers D1 and D6 (see Fig. 1).The only exceptionwas the final three sharks, which were tagged inthe second year in Ralphs Bay, the lower section of the DerwentEstuary (Fig. 1). Tagging in Norfolk Bay was between Lines F and G(Fig. 1). Sharks ranged in size from 153 cm to 284 cm total length(TL; see Appendix Table A1 for sex and size ranges). Previousstudies have reported that females mature at 210e220 cm andmales at 150e170 cm TL (Ebert 1989, 1996; Lucifora et al. 2005).

Ethical Note

Owing to the large size of the sharks and the fast completion ofsurgery, the animalswere not sedated or anaesthetized. The option ofanaesthesia was discussed with the University of Tasmania AnimalWelfare Officer and the Animal Ethics Committee prior to this study.Their recommendation was that the negative effects of usinganaesthetic, including local anaesthetic, could be significant. It wouldbe impossible to maintain these large (up to 2.8 m) animals ina medium (water) in the boat while they recovered from full anaes-thesia. If sharks were released under the effect of the anaesthetic,theywould be too big to help swimuntil they recovered. Theywouldprobably just go to the bottomafter release, inwhich case theywouldhave difficulties in flushing water across their gills to breathe, andwould be vulnerable to attack from other N. cepedianus. Similarly,given that they needed to be released immediately, a local anaes-thetic could affect their swimming capabilities, leaving themvulnerable to attack from conspecifics. Surgically implanting acoustictags without anaesthetic is the common practice for large sharkspecies (e.g. Andrews et al. 2009; Meyer et al. 2010; Papastamatiouet al. 2010), given the concerns and logistical difficulties describedabove. To the best of our knowledge, N. cepedianus individuals

generally showedno ill effects from themethod of capture or surgery.We base this on the fact that all individuals swam off instantly andpowerfully after the procedures and, with the exception of oneindividual, theywere all detectedmoving in the array on subsequentdays (between 4 and 269 days; average 106) and animals returned tothe study areas over the 18-month course of the study (see AppendixTable A1). In addition, four recaptured sharks showed minimalevidence of even being tagged, that is, a very small well-healed scarwith no evidence of infection. All methods used were approved bythe University of Tasmania Animal Ethics Committee and theDepartment of Primary Industries and Water.

Data Analysis

Niche overlapTo investigate niche overlap, each line of receivers was classified

as a resource, and the number of days detected at each resourceused as an indication of how often an animal used that resourcearea. All lines were analysed as they are labelled in Fig. 1, exceptthe Derwent receivers (D1e10), which were grouped into D1e2 asone resource area, D3e7 as another resource and D8e10 as another.Niche overlap analysis was conducted in EcoSim 700 (Gotelli &Entsminger 2010) using Pianka’s index (O) permutated 1000times. The degree of overlap is presented in a 0e1 scale, where0 means no overlap and 1 means complete overlap. Niche overlapwas compared between locations (Norfolk Bay, Derwent Estuary)and between sexes.

Seasonal occurrenceData analysis was based on the number of days sharks were

detected at each line or group of receivers. For this analysis,receivers D1-2 were analysed as one area and classified as Der-went-mid, and receivers D3-7 were considered another area andclassified as the upper Derwent. Lines F, G, K, L and X were allpooled to represent Norfolk Bay. The two Ralphs Bay receivers (R1and R2) were added to Line A; Line J was incorporated with Line Band Lines H, E, PW were analysed as separate areas.

We used classification and regression tree (CART) analysis(De’ath & Fabricius 2000) to investigate spatial (line) and temporal(month of the year) differences in area used. Input data consisted ofthe number of individual sharks detected per 3-day blocks. The sizeof the tree was selected by cross-validation (1-SE rule). To inves-tigate further the seasonal use of this coastal area, the number ofsharks detected per day for each line/area for each month wasanalysed with circular statistics (Oriana 3 software, KovachComputing Services, Anglesey, Wales, http://www.kovcomp.co.uk/oriana). Rayleigh’s test was used to test for uniformity in detectionsover a year. For this purpose, we used a 12-month subsample of thedata, collected from February 2008 to January 2009.

Use of protected areasWe plotted the number of days that each animal was present in

the study area on a timeline to interpret visually the residencytimes of each individual in the shark protected areas. A CART wasalso constructed to identify any seasonality in use of the protectedarea, including differences between the sexes. Themodel was basedon the proportion of days each individual shark was detected insidethe protected area per season. Only animals that were tagged in thefirst season (December 2007eMarch 2008) were considered, anddata from January 2008 to June 2009 were analysed.

Long-distance Movements

Five male N.cepedianus were fitted with MK10 satellite archivaltags (PAT; Wildlife Computers, Redmond, WA, U.S.A.) that were

A. Barnett et al. / Animal Behaviour 81 (2011) 1039e10481042

programmed to release between 138 and 180 days after deploy-ment. The first two tags were deployed on the 30 April and 2 May2008 and another two were deployed on 5 May and 3 June 2009.The individual tagged in May 2009 was also fitted with an acoustictag, which was not used in the quantitative analysis (AppendixTable A2). A single tag was deployed on 30 May 2010. Sharkswere caught as described for acoustic tagging methods, but tagswere externally attached with a stainless steel tag head insertedinto the muscle region in front of the dorsal fin. The procedure tookless than a minute. Additionally, acoustic data were obtained fromother VR2 users on mainland Australia.

RESULTS

Of the 33 sharks tagged in the first year, one shark was notdetected again, giving a total of 32 sharks monitored in the firstyear. In the second year of the study, we also monitored 32 sharks,22 individuals from the first year and 10 tagged in the second year(see Appendix Table A1).

Niche Overlap

There was a low overlap in resource area use between sharkstagged in Norfolk Bay and in the Derwent Estuary (Pianka’s index:O ¼ 0.34). Resource overlap was very high between males andfemales for both locations (Pianka’s index: Norfolk Bay: O ¼ 0.89;Derwent: O ¼ 0.98). Fourteen sharks tagged in Norfolk Bay movedto the Derwent at some stage of the study (see Appendix

Month: May − Oct

Line: B,DU,DM,E,H,NFB,PW A

1.6 (425)

5.0 (61)

R2 = 52%

Figure 2. Five-leaf classification and regression tree explaining the number of acoustically tatext) and month of the year. Model was calculated based on number of sharks detected insharks/3 days are also presented below the leaves, and mean and sample sizes (in parenthes

Table A1). However, while the furthest up-river they moved wasD1-2, for most of the time they were not detected above Line A(Fig. 1). In contrast, no sharks tagged in the mid-upper region ofthe estuary (i.e. above D1-2) moved to Norfolk Bay (see AppendixTable A1), but these sharks ranged as far upstream as D7 (Fig. 1).The three sharks tagged in Ralphs Bay (lower Derwent) did movebetween the Derwent and Norfolk Bay (see Appendix Table A1,Fig. 1). However, as with the sharks tagged in Norfolk Bay, theynever moved up-river above D1-2. These three sharks showedhigh overlap in resource area use (Pianka’s index: O ¼ 0.85) withone another.

Seasonal Occurrence

The classification and regression tree, explaining 52% of thevariability, indicates that the shark’s abundance was mainlydependent on the time of year (Fig. 2). A strong nonuniform patternin area use (Rayleigh’s: P < 0001) was also evident for lines DU(Z ¼ 30.43), DM (Z ¼ 12.15), PW (Z ¼ 8.42), H (Z ¼ 23.73),E (Z ¼ 22.75)and NFB (Z ¼ 29.36), indicating distinct differences inmonthly occurrence of sharks over the year (Fig. 3). The generalpattern observed was of a high occurrence of sharks in the warmermonths from October/November to April, and low to no occurrencebetween May and September (Figs 2, 3).

The exceptions to this pattern were Lines A and B, which werethe only lines to have a uniform distribution (Rayleigh’s: A:Z ¼ 0.68,P ¼ 0.504; B:Z ¼ 0.76, P ¼ 0.468), indicating that sharks occurred inthe lower part of the river all year (Fig. 3). In particular, sharks were

Nov − April

Line: B,DU,PW A,DM,E,H,NFB

2.0 (178)

6.7 (151)

9.9 (148)

Month: April,Nov,Dec Jan − Mar

gged sharks detected in the study area based on spatial resource (i.e. lines, as defined ineach line/group of receivers per 3-day blocks. Histograms of distribution of number ofes) for each group are also indicated. Ranges in histograms correspond to 0e24 sharks.

Jan

Jan

Jan

Jan Jan

Jan

Jan

Jan

Apr

Apr

Apr

Apr Apr

Apr

Apr

Apr

Jul

Jul

Jul

Jul Jul

Jul

Jul

Jul

Oct

Oct

Oct

Oct Oct

Oct

Oct

Oct

NFB

30

30 30

30

30 30

30

30

20

20

20

E

11−33−5>5

DU

A B

H PW

20 20

DM

Figure 3. Circular plots showing the number of N. cepedianus detected per day for eachmonth, at each line (A, B, E, H) or group of receivers (NFB ¼ Norfolk Bay, DU andDM ¼ upper and middle Derwent, PW ¼ Pithwater; see Fig. 1). Plots cover 1 year, fromFebruary 2008 to January 2009.

A. Barnett et al. / Animal Behaviour 81 (2011) 1039e1048 1043

detected on more days at Line A than at any other line duringwinter (Fig. 2). They appeared to congregate in the lower section ofthe estuary in late autumn to early winter (AprileJune) and this ishighlighted by the drop in number of sharks occurring in Lines NFB,E, H, DU coinciding with high numbers of occurrence on Lines A, Band DM at this time (Fig. 3). These peaks in occurrence occurredprior to individuals either departing the coastal area for winter orremaining in close proximity (i.e. individuals that were regularlydetectedmoving in and out of the study area) to the lower Derwentthroughout winter (Fig. 4). In fact, 13 of the 25 sharks tagged inNorfolk Baymoved to the lower Derwent (Lines A and B) in autumn,and this appears to be the time when sharks from both locationshad the most spatial overlap.

With the exception of the one male that left the coastal systemin mid-winter, individuals regularly detected over winter in thefirst year were females (Appendix Table A1, Fig. 4). Females thatwere absent from the study area over winter returned between4 September and 13 November 2008 (Fig. 4), whereas malesreturned considerably later, between 16 December 2008 and 19January 2009 (Fig. 4). For the second year, six females and one malewere regularly detected on Line B during winter. The male detectedduring winter was one of the three sharks tagged in Ralphs Bay(Appendix Table A1). The other two individuals, which werefemales, also remained in the coastal vicinity through winter. In thefollowing summer, eight N. cepedianus (four of each sex) returnedto the coastal region (Appendix Table A1). Three females remainedin the coastal area for two winters and one of these (tagged in theDerwent) never entered Norfolk Bay (Appendix Table A1). Fourindividuals were present for all three summers in coastal areas, andall of them only used the habitat theywere originally tagged in. Onefemale left the study area prior to winter 2008 and returned inspring 2009, so it was absent for an entire year, not just for winter.

Use of Protected Areas

There was a high degree of variation in the number of daysindividuals spent within the monitored coastal areas (Fig. 4). Thepercentage of days that sharks tagged in the first year were presentin the monitored areas ranged from 1 to 53% (Appendix Table A1).However, when we considered only the time between spring andautumn (i.e. omitting winter because most sharks leave the coastalarea), in the first year sharks spent 2e76% and in the second year2e93% of days inside the protected area.

The CART analysis, explaining 52% of the variability, shows thatN. cepedianus spent more time in the protected areas in summerand autumn than in spring and winter (Fig. 5). The proportion ofdays spent inside the protected areas also varied between the sexes(Fig. 5). For winter and spring, males spent a lower proportion oftheir time in these areas than females. Among females, theproportion of time spent in these areas was higher in spring than inwinter (Fig. 5). Males and females spent the same amount of time inprotected areas during summer (Fig. 5). However, in autumn, malesspent a lot more time inside the protected areas than females(Fig. 5).

Long-distance Movements

Only males were fitted with PAT tags because they showedadistinct patternof use of the coastal areas in autumn (Fig. 5), prior toleaving in winter and then not returning until summer (Fig. 4). Alltags released earlier than programmed (Appendix Table A2). All fivemales made northerly migrations. Three tags (one in each year)released in southern New South Wales (NSW; minimum distance970 km in 99 days) and the other two tags released in Bass Strait(minimum distance 460 km in 34 days; Fig. 6 and AppendixTable A2). The individual that was fitted with both an acoustic anda PAT tag moved to Bass Strait in 34 days (PAT tag released 8 June2009), and this animal was detected on the acoustic array back inTasmania on 28 December 2009.

Additionally, three of the acoustically tagged males weredetected in southern NSW between Batemans Bay and Jervis Bay byreceivers of other researchers. All these individuals departed thecoastal area of Tasmania in autumn, before being detected in NSWduring winterespring. One individual moved between Tasmaniaand NSW in 2 consecutive years and returned to Tasmania fora third summer. Another was detected moving back from NSW toTasmania in the following summer.

Summer Winter Summer Winter

Jan 2008 April 2008 July 2008 Oct 2008 Jan 2009 April 2009 July 2009Day

Shar

ks

Figure 4. Timeline showing the days each individual shark was detected in the protected area of southeast Tasmania, from January 2008 to July 2009. For each shark, the first pointmarked on the graph corresponds to the tagging day, with the exception of sharks that were tagged before January 2008. Females are in black and males in grey.

Season: Sp,Win Aut,Sum

Sex: M F

Season: Win Sp

Season: Aut Sum

Sex: F M

3.11% (23)

10.8% (33)

33.6% (15)

19.6% (34)

57.0% (16)

52.1% (49)

R2 = 52%

Figure 5. Six-leaf classification and regression tree, explaining the percentage of time each shark spent in the protected area, based on season and sex. Histograms of distribution ofthis percentage are presented below the leaves, and mean and sample sizes (in parentheses) are also indicated. Ranges in histograms correspond to 0e100%. Sp ¼ spring,Win ¼winter, Aut ¼ autumn, Sum ¼ summer. M ¼male, F ¼ female.

A. Barnett et al. / Animal Behaviour 81 (2011) 1039e10481044

NSW

Sydney

Jervis Bay

Batemans Bay

Eden Victoria

Tasmania

Bass Strait

PAT tagging location

N

0 100 km

Figure 6. Release locations of PAT tags for the five male Notorynchus cepedianus, represented by stars. The three acoustic-tagged males were detected in New South Wales (NSW) atreceivers stationed in the coastal region between Jervis Bay and Batemans Bay.

A. Barnett et al. / Animal Behaviour 81 (2011) 1039e1048 1045

DISCUSSION

Seasonality, Site Fidelity and Habitat Partitioning

Notorynchus cepedianus showed seasonality and site fidelity inthe use of the coastal habitats of southeast Tasmania. The generalpattern was for sharks to leave the coastal areas over winter andreturn the following spring or summer. However, intraspecificdifferences in this pattern were evident. For instance, both move-ment analysis and previous catch rate information show that mostfemales appear in spring, whereasmales arrive later in summer andtheir use of coastal areas peaks in autumn (Barnett et al. 2010b).More specifically, in Norfolk Bay there was a distinct seasonalpattern with sharks abundant in summer and absent in winter.However, in the Derwent Estuary, some females still occurredduring winter (Barnett et al. 2010b).

At a regional scale, N. cepedianus showed site fidelity tocoastal areas in Tasmania. However, the low spatial overlapbetween N. cepedianus tagged in the Derwent Estuary and Nor-folk Bay suggests that once in these coastal areas, individualswere site-specific to a particular habitat (Norfolk Bay or DerwentEstuary). Also, 55 of a total of 56 recaptures occurred in theiroriginal tagging location (Barnett et al. 2010b) and sharks fromboth locations showed relatively low dietary overlap (Pianka’sindex: O ¼ 0.45; Barnett et al. 2010a). Low spatial and dietary

overlap suggests resource partitioning and/or localized sitefidelity over relatively fine spatial scales (about 30 km). Mostknowledge on habitat partitioning in shark species is related tothe segregation of juveniles from adults (Speed et al. 2010), andthere is very little information about intraspecific habitat parti-tioning between groups of adult sharks. However, there area couple of examples: white sharks show similar site fidelitypatterns to N. cepedianus, where individuals annually travel2000e5000 km to return to the central California region, andthen exhibit clear preferences for particular coastal sites sepa-rated by only kilometres (Jorgensen et al. 2009). Blacktip reefsharks, Carcharhinus melanopterus, from two adjacent lagoons atPalmyra Atoll, also show fine-scale site fidelity and niche sepa-ration, rarely moving into the lagoon that they were not taggedin (Papastamatiou et al. 2010).

Site fidelity has been attributed to mating, pupping, feeding andreuse of natal sites (reviewed in Speed et al. 2010). The lack ofevidence relating the use of these coastal areas by N. cepedianus toreproduction and the absence of smaller size classes (<150 cm)suggest that the main purpose for site fidelity is not mating,pupping or use of natal sites (Barnett et al. 2010b). The highseasonal abundance of known prey species during summersuggests feeding site fidelity (Stevens & West 1997; Barnett et al.2010a). Recurring site fidelity is likely to increase foragingsuccess, as spatial familiarity and increased knowledge of the local

A. Barnett et al. / Animal Behaviour 81 (2011) 1039e10481046

prey can improve foraging efficiency (Staniland et al. 2004;Jorgensen et al. 2009; Van Moorter et al. 2009). For example,intraspecific differences in foraging locations of Antarctic fur seals,Arctocephalus gazella, from the same breeding colony were attrib-uted to individual knowledge of successful foraging sites (Stanilandet al. 2004). If this is the case, then the site-specific behaviour andlow dietary overlap (Barnett et al. 2010a) shown by N. cepedianus inTasmania may be influenced by the individuals’ foraging success ineither habitat (bay or estuary).

Although resource partitioning by N. cepedianus was evident,there was some spatial overlap in the lower Derwent, particularlyin late autumn, when sharks from the upper Derwent and NorfolkBay moved to the lower section of the estuary prior to mostindividuals exiting the coastal system for winter. Spatial overlap inthis area is also evident by the three sharks tagged in Ralphs Bay(lower Derwent) not moving to the mid-upper region of theestuary, but moving to Norfolk Bay. Also, the single individual thatwas recaptured moving between the two locations was tagged inNorfolk Bay and subsequently recaptured in Ralphs Bay (Barnettet al. 2010b). Similarly, three social groups of killer whales thatshowed spatial separation in their use of inshore waters in theBritish Columbia region had some spatial overlap in specific areas(Hauser et al. 2007). The significance of the lower Derwent inautumn is unclear. It is possible that individuals form temporaryaggregations for social reasons, or they may be following prey. Ifthe lower Derwent area is used for mating in autumn prior tomost animals exiting the coastal systems for winter, then ourprevious surveys would have failed to detect females with freshmating scars (Barnett et al. 2010b). This may explain why malesarrive later to coastal areas and why they peak in occurrence(both catch rates and movement data) during autumn (Barnettet al. 2010b).

Long-distance Movement

The movement of three acoustically tagged males (oneindividual in 2 consecutive years) to a similar vicinity in southernNSW as three of the PAT-tagged males, along with two other PATtags releasing in Bass Strait, suggests a northern seasonal migrationby males. The return of two of these acoustically tagged animals(one individual for three consecutive summers) to the Tasmaniancoastal area, as well as the return of the male that was doubletagged with a PAT and an acoustic tag, further supports long-rangesite fidelity. Northerly migrations by males during winter arefurther supported by 80% of N. cepedianus caught in the NSW sharkcontrol (beach meshing) programme being male and primarilyoccurring during AugusteOctober (winterespring; Krogh 1994).The near absence of N. cepedianus in NSW catches in Decem-bereJune (summereautumn) also coincides with tracked malesreturning to coastal areas of Tasmania and their appearance incatches at this time (Barnett et al. 2010b).

If males go north, where do females go, and why do somefemales remain in close proximity to coastal areas over winterwhile others depart? Sex-specific migrations are evident in a rangeof taxonomic groups including mammals, fish, birds and amphib-ians and have been attributed to reproductive requirements(reviewed in Handley & Perrin 2007). Since N.cepedianus havea biennial reproductive cycle (Ebert 1989, 1996), and the coastalareas of this study do not appear to be nursery areas or puppinggrounds (Barnett et al. 2010b), one hypothesis worth furtherinvestigation is that the separation of females is related to pregnantindividuals moving somewhere specifically to pup. Intraspecificvariations in large-scale movement behaviour have also beenassociated with variations in resource availability. For example, inSweden, half of a population of moose, Alces alces, migrates while

the other half remains (Ball et al. 2001). Similar behavioural vari-ations were exhibited by grey seals, Halichoerus grypus, at SableIsland, Canada (Austin et al. 2004). Migrant and nonmigrant indi-viduals may persist in a population if they receive approximatelyequal payoffs (Swingland & Lessells 1979). The segregation ofN. cepedianus individuals over winter could thus be related toresource partitioning. Here, males move north after summer whenresources in coastal regions are less abundant, while some femalesremain close to coastal areas and other females may move furtherafield.

Protected Areas

Protected area management for shark conservation has mainlyfocused on life stages using nursery areas, and more information isrequired to determine the extent to which older, more mobileindividuals are vulnerable to threats (Speed et al. 2010). Thepresent study is an example of the degree of protection that may beafforded to large coastal predators with current managementstrategies. Despite the variation in the number of days thatN. cepedianus individuals were detected within protected areas, therelatively low average (24% of days) indicates that N. cepedianus arevulnerable to exploitation for a significant amount of time. Even injust the peak season (spring to autumn), many individuals are onlyin protected areas for about a third of the time. In fact, twoacoustically tagged N. cepedianus were killed by fishing soon afterleaving the protected areas (tags returned to A. Barnett byfishermen).

The challenge of using marine protected areas to conservepredators is to design an area that is large enough to afford suffi-cient protection to species that are highly mobile and hard toobserve, while appeasing human activities (Ashe et al. 2010;Pichegru et al. 2010). Studies on killer whales and Africanpenguins, Spheniscus demersus, used behavioural information toprioritize habitats primarily used for the activity in which thespecies were most affected by anthropogenic disturbance (Asheet al. 2010; Pichegru et al. 2010). In both cases, the protection ofessential foraging grounds was of greater benefit than protectinghabitats generically. In this regard, the current study has shownthat the protection of coastal areas in Tasmania has included animportant foraging ground for N. cepedianus.

Conclusion

Knowledge of the movement patterns of large mobile predatorshas been limited by the large spatial scale of their ranges. Conse-quently, the factors that influence movement patterns and distri-butions of predators in relation to resources remain poorlyunderstood (Austin et al. 2004). In particular, the lack of synthesisof movement data with other ecological information means ourunderstanding of the ecology of upper trophic level predators ispoor, impeding their effective conservation and management(Speed et al. 2010). The incorporation of movement behaviourcollected in this study with previously gathered ecological infor-mation gives a much clearer picture of the habitat use patterns ofN. cepedianus in coastal areas of Tasmania. The fine-scale resourcepartitioning evident in N. cepedianus has rarely been reported forlarge mobile predators. Population ecology has traditionally over-looked individual differences, but understanding the causes andconsequences of intrapopulation differences in movement patternscan provide important insight into what factors influence thespatial dynamics of apex predator populations (Jorgensen et al.2009). Our results suggest that N. cepedianus exert significantpredation pressure on prey inhabiting these areas and play a crucialrole as the apex predator in coastal systems of Tasmania. Owing to

A. Barnett et al. / Animal Behaviour 81 (2011) 1039e1048 1047

their high trophic level throughout their global distribution (seeBarnett et al. 2010a for a summary of regional diet studies), it islikely that they play a similar role in other coastal systems. Overall,this study highlights the need to complement movement data withother ecological information (multidata approaches) to understandthe habitat use of large mobile predators and their potentialinfluences on ecosystem structure and function.

Acknowledgments

We thank E. Forbes, J. Yick and D. Jones for field assistance andRuss Bradford for advice with setting up PAT tags. We also thank DrM. Heupel and Dr C. Simpfendorfer for advice on array design.Australian Animal Tagging and Monitoring System (AATAMS) forthe loan of 12 receivers. Nick Otway and Megan Ellis from South-East Australian Coastal Acoustic Monitoring System (SEACAMS) andthe NSW Department of Industry and Investment for providingdata from the detections of our sharks in their receivers. This studywas supported by grants to A.B. from the Save Our Seas Foundation,Winifred Violet Scott Foundation and the Holsworth WildlifeResearch Endowment.

References

Andrews, K. S., Williams, G. D., Farrer, D., Tolimieri, N., Harvey, C. J., Bargmann, G.& Levin, P. S. 2009. Diel activity patterns of sixgill sharks, Hexanchusgriseus: theups and downs of an apex predator. Animal Behaviour, 78, 525e536.

Ashe, E., Noren, D. P. & Williams, R. 2010. Animal behaviour and marine protectedareas: incorporating behavioural data into the selection of marine protectedareas for an endangered killer whale population. Animal Conservation, 13,196e203.

Austin, D., Bowen, W. D. & McMillan, J. I. 2004. Intraspecific variation in move-ment patterns: modelling individual behaviour in a large marine predator.Oikos, 105, 15e30.

Austin, D., Bowen, W. D., McMillan, J. I. & Iverson, S. J. 2006. Linking movement,diving and habitat to foraging success in a large marine predator. Ecology, 87,3095e3108.

Ball, J. P., Nordengren, C. & Wallin, K. 2001. Partial migration by large ungulates:characteristics of seasonal moose Alcesalces ranges in northern Sweden.WildlifeBiology, 7, 39e47.

Barnett, A., Abrantes, K., Stevens, J. D., Yick, J., Frusher, S. D. & Semmens, J.M. 2010a.Predatoreprey relationships and foraging ecology of a marine apex predator witha wide temperate distribution. Marine Ecology Progress Series, 416, 189e200.

Barnett, A., Stevens, J. D., Frusher, S. D. & Semmens, J. M. 2010b. Seasonaloccurrence and population structure of the broadnose sevengill shark Notor-ynchuscepedianus in coastal habitats of southeast Tasmania. Journal of FishBiology, 77, 1688e1701.

Charrassin, J. B. & Bost, C. A. 2001. Utilisation of the oceanic habitat by kingpenguins over the annual cycle. Marine Ecology Progress Series, 221, 285e297.

Cortés, E. 1999. Standardized diet compositions and trophic levels of sharks. ICESJournal of Marine Science, 56, 707e717.

De’ath, G. & Fabricius, K. E. 2000. Classification and regression trees: a powerfulyet simple technique for ecological data analysis. Ecology, 81, 3178e3192.

Ebert, D. A. 1989. Life history of the sevengill shark, Notorynchuscepedianus (Peron1807), in two northern California Bays. California Fish and Game, 75, 102e112.

Ebert, D. A. 1996. Biology of the sevengill shark Notorynchus cepedianus (Peron,1807) in the temperate coastal waters of southern Africa. South African Journalof Marine Science, 17, 93e103.

Estes, J. A., Crooks, K. & Holt, R. 2001. Predators, ecological role of. In: Encyclopediaof Biodiversity (Ed. by S. A. Levin & J. Lubchenco), pp. 857e878. New York:Academic Press.

Ferretti, F., Worm, B., Britten, G. L., Heithaus, M. R. & Lotze, H. K. 2010. Patternsand ecosystem consequences of shark declines in the ocean. Ecology Letters, 13,1055e1071.

Foote, A. D., Simila, T., Vikingsson, G. A. & Stevick, P. T. 2010. Movement, sitefidelity and connectivity in a top marine predator, the killer whale. EvolutionaryEcology, 24, 803e814.

Gotelli, N. J. & Entsminger, G. L. 2010. EcoSim: Null Models Software for Ecology.Version 7. Jericho, Vermont: Acquired Intelligence Inc. & Kesey-Bear. www.garyentsminger.com/ecosim/index.htm.

Handley, L. J. L. & Perrin, N. 2007. Advances in our understanding of mammaliansex-biased dispersal. Molecular Ecology, 16, 1559e1578.

Hauser, D. D. W., Logsdon, M. G., Holmes, E. E., VanBlaricom, G. R. &Osborne, R. W. 2007. Summer distribution patterns of southern resident killerwhales Orcinus orca: core areas and spatial segregation of social groups. MarineEcology Progress Series, 351, 301e310.

Heppell, S. S., Limpus, C. J., Crouse, D. T., Frazer, N. B. & Crowder, L. B. 1996.Population model analysis for the loggerhead sea turtle, Carettacaretta, inQueensland. Wildlife Research, 23, 143e161.

Heupel, M. R., Carlson, J. K. & Simpfendorfer, C. A. 2007. Shark nursery areas:concepts, definition, characterization and assumptions. Marine Ecology ProgressSeries, 337, 287e297.

Jorgensen, S. J., Reeb, C. A., Chapple, T. K., Anderson, S., Perle, C.,Van Sommeran, S. R., Fritz-Cope, C., Brown, A. C., Klimley, A. P. & Block, B. A.2009. Philopatry and migration of Pacific white sharks. Proceedings of the RoyalSociety B, 277, 679e688.

Kinney, M. J. & Simpfendorfer, C. A. 2009. Reassessing the value of nursery areas toshark conservation and management. Conservation Letters, 2, 53e60.

Krogh, M. 1994. Spatial, seasonal and biological analysis of sharks caught in theNew-South-Wales protective beach meshing program. Australian Journal ofMarine and Freshwater Research, 45, 1087e1106.

Kuhn, C. E., Crocker, D. E., Tremblay, Y. & Costa, D. P. 2009. Time to eat:measurements of feeding behaviour in a large marine predator, the northernelephant seal Mirounga angustirostris. Journal of Animal Ecology, 78, 513e523.

Last, P. R. & Stevens, J. D. 2009. Sharks and Rays of Australia. 2nd edn. Melbourne:CSIRO Publishing.

Lee, P. L. M., Luschi, P. & Hays, G. C. 2007. Detecting female precise natal philopatryin green turtles using assignment methods. Molecular Ecology, 16, 61e74.

Lloyd, S. D., Dee Boersma, P. & Gordon, S.1995. Court satellite telemetry of thewintermigration of Adélie penguins (Pygoscelis adeliae). Polar Biology, 16, 221e225.

Lowe, C. G., Wetherbee, B. M. & Meyer, C. G. 2006. Using acoustic telemetrymonitoring techniques to quantify movement patterns and site fidelity ofsharks and giant trevally around French Frigate Shoals and Midway Atoll. AtollResearch Bulletin, 543, 281e303.

Lucifora, L. O., Menni, R. C. & Escalante, A. H. 2005. Reproduction, abundanceand feeding habits of the broadnose sevengill shark Notorynchus cepedia-nus in north Patagonia, Argentina. Marine Ecology Progress Series, 289,237e244.

Martin, T. G., Chades, I., Arcese, P., Marra, P. P., Possingham, H. P. & Norris, D. R.2007. Optimal conservation of migratory species. PloS One, 2, e751.

Mazaris, A. D. & Matsinos, Y. G. 2006. An individual based model of sea turtles:investigating the effect of temporal variability on population dynamics.Ecological Modelling, 194, 114e124.

Meyer, C. G., Papastamatiou, Y. P. & Holland, K. N. 2010. A multiple instrumentapproach to quantifying the movement patterns and habitat use of tiger(Galeocerdo cuvier) and Galapagos sharks (Carcharhinus galapagensis) at FrenchFrigate Shoals, Hawaii. Marine Biology, 157, 1857e1868.

Myers, A. E. & Hays, G. C. 2006. Do leatherback turtles Dermochelyscoriacea forageduring the breeding season? A combination of data-logging devices providenew insights. Marine Ecology Progress Series, 322, 259e267.

Papastamatiou, Y. P., Friedlander, A. M., Caselle, J. E. & Lowe, C. G. 2010. Long-term movement patterns and trophic ecology of blacktip reef sharks (Carch-arhinus melanopterus) at Palmyra Atoll. Journal of Experimental Marine Biologyand Ecology, 386, 94e102.

Pichegru, L., Gremillet, D., Crawford, R. J. M. & Ryan, P. G. 2010. Marine no-takezone rapidly benefits endangered penguin. Biology Letters, 6, 498e501.

Simpfendorfer, C. A. & Heupel, M. R. 2004. Assessing habitat use and movement.In: Biology of the Sharks and their Relatives (Ed. by J. C. Carrier, J. A. Musick &M. R. Heithaus), pp. 553e572. London: CRC Press.

Smout, S. & Lindstrom, U. 2007. Multispecies functional response of the minkewhale Balaenoptera acutorostrata based on small-scale foraging studies. MarineEcology Progress Series, 341, 277e291.

Speed, C. W., Field, I. C., Meekan, M. G. & Bradshaw, C. J. A. 2010. Complexities ofcoastal shark movements and their implications for management. MarineEcology Progress Series, 408, 275e305.

Staniland, I. J., Reid, K. & Boyd, I. L. 2004. Comparing individual and spatialinfluences on foraging behaviour in Antarctic fur seals Arctocephalus gazella.Marine Ecology Progress Series, 275, 263e274.

Stevens, J. D. & West, G. J. 1997. Investigation of School and Gummy Shark NurseryAreas in South Eastern Australia. Fisheries Research and Development CorporationProject 93/061 Final Report. Hobart: CSIRO Marine Research.

Stewart, B. S. 1997. Ontogeny of differential migration and sexual segregation innorthern elephant seals. Journal of Mammalogy, 78, 1101e1116.

Swingland, I. R. & Lessells, C. M. 1979. Natural regulation of giant tortoise pop-ulations on Aldabra Atoll: movement polymorphism, reproductive success andmortality. Journal of Animal Ecology, 48, 639e654.

Switzer, P. V. 1993. Site fidelity in predictable and unpredictable habitats. Evolu-tionary Ecology, 7, 533e555.

Van Moorter, B., Visscher, D., Benhamou, S., Borger, L., Boyce, M. S. &Gaillard, J. M. 2009. Memory keeps you at home: a mechanistic model forhome range emergence. Oikos, 118, 641e652.

Walli, A., Teo, S. L. H., Boustany, A., Farwell, C. J., Williams, T., Dewar, H., Prince, E.& Block, B. A. 2009. Seasonal movements, aggregations and diving behavior ofAtlantic bluefin tuna (Thunnus thynnus) revealed with archival tags. PloS One, 4,e6151.

Weng, K. C., Foley, D. G., Ganong, J. E., Perle, C., Shillinger, G. L. & Block, B. A. 2008.Migration of an upper trophic level predator, the salmon shark Lamnaditropis,between distant ecoregions. Marine Ecology Progress Series, 372, 253e264.

Williams, T. M., Estes, J. A., Doak, D. F. & Springer, A. M. 2004. Killer appetites:assessing the role of predators in ecological communities. Ecology, 85,3373e3384.

A. Barnett et al. / Animal Behaviour 81 (2011) 1039e10481048

APPENDIX

Table A1. Details of N. cepedianus fitted with acoustic transmitters

Location Date Sex TL Days % Days Ret/Rem y1 Ret/Rem y2 Mixed

First year 2007e2008NFB 4 Dec F 234 100 19 Remained Remained YesNFB 4 Dec F 226 92 18 No return No return YesNFB 2 Jan F 157 164 32 Ret-NFB Ret-NFB NoNFB 2 Jan F 153 50 10 No return No return YesNFB 2 Jan F 208 38 7 No return No return NoNFB 2 Jan F 200 129 25 Remained No return YesNFB 2 Jan F 196 269 52 Remained Remained YesNFB 2 Jan M 204 99 19 No return No return YesNFB 2 Jan F 198 44 9 No return No return NoDerwent 22 Jan M 212 196 40 Ret-Der No return NoDerwent 22 Jan F 224 95 19 Ret-Der No return NoDerwent 26 Jan F 243 67 14 No return* e NoDerwent 26 Jan F 247 4 1 Ret-Der No return NoDerwent 26 Jan F 217 86 18 No return No return NoDerwent 26 Jan F 257 29 6 No return No return NoNFB 30 Jan F 200 260 53 Ret-NFB Ret-NFB NoNFB 1 Feb F 188 38 8 Ret-NFB No return YesDerwent 2 Feb F 256 75 16 No return Ret-Der NoDerwent 2 Feb F 262 137 28 Remained No return NoDerwent 2 Feb F 260 0 0 No detections e e

Derwent 2 Feb F 277 87 18 Remained No return NoDerwent 2 Feb F 246 116 24 Remained Remained NoNFB 9 Feb F 255 121 25 Remained Ret-NFB YesNFB 9 Feb F 262 102 21 Ret-NFB No return YesNFB 9 Feb F 249 130 27 Remained No return YesNFB 9 Feb F 284 110 23 Ret-NFB No return NoNFB 17 Feb M 215 194 41 Ret-NFB No return YesNFB 21 Feb M 204 67 14 No returny No return NoNFB 22 Feb M 219 145 31 Ret-NFB No return* NoDerwent 26 Feb M 232 187 41 Ret-Der Remained NoDerwent 6 Mar M 246 183 40 Ret-Der Ret-Der NoNFB 7 Mar M 222 158 35 Ret-NFBy Ret-NFBy NoNFB 7 Mar M 224 96 21 Ret-Dery No return Yes

Second year 2008e2009Derwent 17 Oct F 208 26 12 e No return NoDerwent 17 Oct F 155 29 13 e No return NoNFB 23 Oct M 221 64 29 e No return NoDerwentR 6 Nov F 228 101 49 e Remained YesDerwentR 6 Nov F 210 144 70 e Remained YesDerwentR 6 Nov M 199 191 93 e Remained YesNFB 18 Nov F 267 11 6 e No return NoNFB 18 Nov M 164 165 85 e Ret- NFB YesNFB 18 Nov F 240 4 2 e No return NoNFB 19 Nov F 274 68 35 e Remained Yes

NFB ¼ Norfolk Bay; Derwent ¼ Derwent Estuary; DerwentR ¼ Ralphs Bay (lower region of the Derwent Estuary). M ¼male, F ¼ female. Date is the date of tagging and TL istotal length (cm). For acoustic tags, Days ¼ the number of days detected, %Days ¼ percentage of days detected from date of tagging until receivers removed. Ret/Rem indicateswhether sharks returned after winter or remained in coastal region throughwinter for the first (y1) and second (y2) years. For animals that returned, the area where they weredetected is indicated (NFB/Der); no return indicates they left the area and did not return. The Mixed column gives information on whether the animal moved between the twohabitats (Derwent and Norfolk Bay) or not.

* Animals killed by fishermen after leaving protected areas.y Individuals that were detected on receivers in NSW.

Table A2. Details of N. cepedianus fitted with PAT transmitters

Location Date Sex TL Days P days R date/location Return Distance travelled

Derwent 30 Apr 2008 Male 208 53 180 BS e 21 Jun 2008 e 633Derwent 1 May 2008 Male 211 57 180 NSW e 26 Jun 2008 e 830Norfolk Bay* 5 May 2009 Male 214 34 138 BS e 8 Jun 2009 NFB 461Derwent 3 Jun 2009 Male 216 99 150 NSW e 10 Sept 2009 e 970Norfolk Bay 30 Apr 2010 Male 234 156 180 NSW e 4 Oct 2010 930

P days ¼ programmed number of days after tagging for tags to be released. R date/location ¼ date and place of tag release: NSW ¼ New SouthWales, BS ¼ Bass Strait. Distancetravelled is the shortest distance travelled (km).

* Shark also fitted with an acoustic tag.