Monitoring the Effects of Tourism on Whale Shark Rhincodon Typus Behaviour in Mozambique

Monitoring the effects of tourism on whale shark Rhincodon typus behaviour in Mozambique

P ETER J. HASKELL,ANDREW M C G OWAN,ANNA W ESTLING,ADRIANA M ÉNDEZ- J IMÉNEZ,CHRISTOPH A. ROHNER,KYM C OLLINS,MARCELA R OSERO-CAICEDO J ODI S ALMOND,ARA M ONADJEM,ANDREA D. MARSHALL and S IMON J. PIERCE

Abstract The whale shark Rhincodon typus is a popular Introduction focal species in the marine tourism industry. We analysed  encounters with at least  individual sharks during iewing sharks in their natural setting is a popular tour-  – to assess their behaviour in the presence of Vism activity (Gallagher & Hammerschlag, ) and swimmers at Tofo Beach, Mozambique. Sharks varied in the income accrued creates an incentive to manage these size (estimated .–. m total length) and the majority charismatic species as a non-consumptive resource   (%) were males. The sharks displayed avoidance behav- (Brunnschweiler, ; Clua et al., ; Vianna et al.,  iours during .% of encounters. Encounter duration de- ). However, studies on elasmobranchs have documen- creased significantly, from  minutes  s with ted situations in which marine tourism has negative effects, undisturbed sharks to  minutes  s when sharks expressed including behavioural changes and increased energetic costs   avoidance behaviours, indicating that interactions with (Pierce et al., ; Fitzpatrick et al., ). Improved under- tourists affected the sharks’ short-term behaviour. standing of the actual or potential effects of tourism inter- However, during the .-year study period we found no actions is important for mitigating or avoiding longer-term trend in the mean encounter duration, the overall expres- effects and ultimately safeguarding employment and the sion of avoidance behaviour or the likelihood of an individ- tourism infrastructure. ual shark exhibiting avoidance behaviours. Potential effects The whale shark Rhincodon typus is the largest fish and of tourism may be mitigated by the non-breeding status and an iconic species for tourism (Gallagher & Hammerschlag,  transient behaviour of sharks at this aggregation site. ). Commercial whale shark interaction tours began in  in Western Australia after a seasonal aggregation of Keywords Behavioural observations, ecotourism, Mozambique, the species was discovered within the Ningaloo Marine Rhincodon typus, risk assessment, tourism management, whale Park (Davis et al., ). In  the whale shark tourism shark industry at this site was valued at AUD  million annually  This paper contains supplementary material that can be found (Catlin et al., ). Whale shark tourism industries have online at http://journals.cambridge.org now developed at several locations in all three tropical oceans and are mostly based on the presence of predictable seasonal feeding aggregations of sharks exploiting ephem- eral bursts in local productivity, such as mass fish or coral spawning events (Taylor, ; de la Parra Venegas et al., ). The global revenue from whale shark tourism was provisionally estimated to be USD  million in  PETER J. HASKELL* and ANDREW MCGOWAN School of Biosciences, University of (Graham, ). Exeter, Penryn, UK Whale sharks are categorized as Vulnerable on the IUCN ANNA WESTLING,ADRIANA MÉNDEZ-JIMÉNEZ,KYM COLLINS,MARCELA ROSERO- Red List (Norman, ). Although the economic value or CAICEDO and JODI SALMOND All Out Africa Research Unit, Lobamba, Swaziland potential of whale shark tourism has helped to justify legal CHRISTOPH A. ROHNER†,ANDREA D. MARSHALL‡ and SIMON J. PIERCE‡§ (Corresponding author) Marine Megafauna Foundation, Tofo Beach, protection for the species in some countries, concerns have Inhambane, Mozambique. E-mail [email protected] also been raised that specialist tourism industries could ARA MONADJEM All Out Africa Research Unit, Department of Biological Sciences, negatively affect the sharks. Sightings of whale sharks at University of Swaziland, Kwaluseni, Swaziland Gladden Spit in Belize declined during – *Also at: Marine Megafauna Foundation, Tofo Beach, Inhambane, Mozambique (Graham & Roberts, ) and, based on anecdotal reports †Also at: School of Geography, Planning and Environmental Management, The  University of Queensland, St Lucia, Australia, and Climate Adaptation Flagship, from guides, remained low at least until (Graham, CSIRO Marine and Atmospheric Research, EcoScience Precinct, Dutton Park, ). Graham () suggested that the rapid increase in Australia diver numbers at this site may have led to disturbance of ‡Also at: Wild Me, Tofo Beach, Inhambane, Mozambique §Also at: All Out Africa Research Unit, Lobamba, Swaziland snapper spawning behaviour (the main driver of whale Received  February . Revision requested  April . shark presence), and the whale sharks themselves, although Accepted  September . First published online  October . a dedicated study on disturbance by divers did not identify a

Oryx, 2015, 49(3), 492–499 © 2014 Fauna & Flora International doi:10.1017/S0030605313001257 Downloaded from https://www.cambridge.org/core. IP address: 170.106.34.90, on 29 Sep 2021 at 10:41:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605313001257 Tourism and whale sharks 493

direct effect on the sharks (Heyman et al., ). Propeller injuries from small boats have also been observed on whale sharks at several aggregation sites (Rowat et al., ; Speed et al., ). Studies of the short-term behavioural responses of sharks to tourists and boats at Ningaloo Reef (Norman, ), Donsol in the Philippines (Quiros, ) and Tofo Beach in Mozambique (Pierce et al., ) have revealed that sharks routinely display avoidance behaviours, including banking, eye-rolling, fast swimming and diving, in response to close approaches by swimmers or boats. The ecological, social and economic sustainability of whale shark tourism at Ningaloo Reef was reviewed by Mau (). Although long-term empirical data on whale shark behaviour in the area were not available, the industry FIG. 1 Tofo Beach, Mozambique. The rectangle on the inset was judged to be ecologically sustainable. This was based on shows the location of the main map in south-east Africa. the lack of observed interruption of feeding behaviour, which may take place largely at night or at least outside tourist interaction times (Taylor, ), the regular re-sightings are provided by Pierce et al. (). Tourism operators of philopatric sharks (Holmberg et al., , ), and the offer daily -hour snorkelling trips that aim to locate lack of reproductive behaviour observed in this juvenile whale sharks and other marine megafauna. Vessels typically male-biased population (Meekan et al., ; Norman & survey a  km stretch of coastline south of Tofo, between the Stevens, ). However, concerns were raised about the surf line and c. , m from the shore, in waters c. – m potential for injury from boat strikes (Speed et al., ) deep. Sharks are located through visual inspection of the and an apparent decline in mean size over time water surface, where it is often possible to spot their dark (Bradshaw et al., ). Aside from Ningaloo Reef, where silhouettes or exposed fins. The boat is then positioned in the sharks have been relatively well studied, the longer-term relation to the shark’s direction of travel and clients enter sustainability of whale shark tourism has not been consid- the water to interact with the shark. ered explicitly. We collected data during January –June , Here we examine the tourism industry at Tofo Beach except during July and August  when no sampling (Praia do Tofo) in Mozambique, an international hotspot trips took place. Upon locating a whale shark, observers en- for whale shark encounters (Pierce et al., ). Whale tered the water alongside clients and recorded the total shark interactions are a key attraction for international number of swimmers and environmental characteristics, divers visiting the country but a lack of official management including weather conditions (categorized as sunny, slightly has raised questions about the sustainability of the industry overcast, overcast or raining), Beaufort sea state and under- (Pierce et al., ; Tibiriçá et al., ), particularly after a water visibility. The total length of the shark was estimated significant decline in sightings during – (Rohner visually (Rohner et al., ) and the sex identified by the et al., ). Previous work at Tofo Beach has examined presence or absence of claspers on the pelvic fins. The how interactions can be managed to minimize the potential presence and location of any injuries or scars were noted for short-term negative effects on the sharks (Pierce et al., and categorized post hoc as either major or minor (Speed ). Here we extend that dataset to evaluate whether lon- et al., ). A basic ethogram of each shark’s behaviour gitudinal encounter data reveal evidence of an increasing in the presence of swimmers was produced to record slow frequency of avoidance behaviours overall and whether swimming, equating to normal behaviour; fast swimming, individual-based analyses show evidence of changing avoid- where there was an obvious increase in the shark’s tail-beat ance behaviours over time. We consider the implications of frequency; diving, where the shark dived away from the sur- these data and evaluate the longer-term implications for the face; banking, where the shark rolled its back towards swim- ecology of whale sharks at this site. mers; changing direction, where the shark altered its direction of swimming in the presence of swimmers; and any other obvious avoidance behaviours, such as the violent  Methods shudder reported by Quiros ( ). Observations of feeding behaviour were also noted when they occurred. The total The village of Tofo Beach is situated in the Inhambane prov- encounter duration, defined as the time between the first ince of Mozambique, c.  km north-east of the capital swimmer entering the water and the last swimmer returning city, Maputo (Fig. ). A full site overview and description to the boat, was recorded (in minutes) following each of the commercial whale shark tourism industry in Tofo interaction. When possible, standardized identification

photographs of both the left and right flanks of each shark were taken and uploaded to the global Wildbook for Whale Sharks photo-identification library (Arzoumanian et al., ; Marshall & Pierce, ). Each of these encounters was then assigned to a new or previously identified shark in the library. It was difficult to establish whether behaviours exhibited by whale sharks in the presence of swimmers would also occur in a natural, undisturbed setting. Within-effect comparisons (i.e. behaviour in the presence of swimmers) were FIG. 2 Mean (± SD) number of whale sharks Rhincodon typus therefore employed instead of attempting to determine sighted per daily trip during the -month sampling period. cause-and-effect relationships. This approach is useful in Numbers above the bars denote the number of sampling days in cases where there are no baseline or control data, and that month. focuses on assessing behavioural responses under gradations of the effect (Bejder & Samuels, ). Fast A generalized linear model was fitted, using the number swimming, diving, banking and changing direction are typ- of encounters as the response variable and search effort ically classified as avoidance behaviours in whale sharks (in hours) and year as explanatory variables, to establish (Norman, ; Quiros, ; Pierce et al., ). When whether encounter rate changed over the course of the these or other obvious avoidance behaviours were observed study. As the dataset ended in June , a second model during an interaction they were each assigned a score of one. was fitted to -month periods, using the number of encoun- The total score at the end of each encounter was thus equal ters as the response variable and search effort and -month to the total number of avoidance behaviours observed. All block as explanatory variables. A quasi-poisson error struc- statistical analyses were carried out using Rv... (R ture was used for both analyses because of overdispersion of Development Core Team, ), with MASS and nlme. residual deviance. A third model was used to test for a The relationship between avoidance score and encounter change in encounter duration during the study period, duration was tested using a generalized linear mixed model using log-transformed encounter duration as the response approach. Encounter duration, log transformed to ensure variable and month and year as explanatory variables. We that error structure was normally distributed, was used as  used χ to test for a population-level increase in avoidance the response variable. Avoidance score was used as the response over the five -month periods. A generalized linear explanatory variable and shark identification was added as mixed effects model was used to test whether the likelihood a random effect to avoid pseudo replication. Given the sig- of avoidance was related to the number of previous encoun- nificant relationship between avoidance score and encounters in which the whale shark had been successfully identi- ter duration (see results; Pierce et al., ), encounter fied and its behaviours recorded over the study period. duration was used in all further analyses as a proxy for We used the Wildbook library to establish how many previ- avoidance. A further model was fitted to determine which ous encounters had been recorded for each individual shark. variables significantly affected encounter duration. A binomial score for avoidance was used as the response Encounter duration was entered as the response variable variable, encounter number as the explanatory variable, and tested against nine main effect explanatory variables: and shark identity as a random factor. Significance was whale shark variables (size, sex, presence/absence of scar- accepted at the % confidence interval in all analyses. ring or feeding behaviour), environmental variables (weather, sea state (–) and underwater visibility), number of swimmers present during an encounter, and the Julian Results day (Day  =  January ). We tested second-order interactions between all combinations of whale shark variables; We recorded a total of  whale shark encounters from  all combinations of environmental variables; size, sex and trips during January –June , with a total search scarring and all environmental variables; and between the effort of . hours. The mean encounter rate was . ± SD number of swimmers and underwater visibility. Year and . per trip, with – encounters recorded on each trip, month of encounter were entered as nested random effects. equating to . ± SD . sharks per hour of search effort. Only encounters in which shark identity had been con- At least one shark was sighted on .% of trips. There firmed were used in the analysis; a single interaction was was no significant trend in sightings over the course of the randomly selected from those individuals for which mul- study (linear regression, df = ,P=.; Fig. ). A total of tiple encounters had been recorded. We used a generalized  sharks were positively identified using the Wildbook linear model to test for a significant relationship between the photo-identification library. Of  sexed sharks,  were amount of scarring on an individual and its total length. female (.%) and  were male (.%), which was

FIG. 4 The total number of sightings (unique encounter days) for each photo-identified shark. FIG. 3 Frequency distribution of the estimated lengths of photo-identified sharks. All potential explanatory variables were present in  encounters and, after controlling for multiple encounters  significantly different from a  :  sex ratio (χ = ., with individual sharks,  encounters were available to P # .). The mean total length was . ± SD . m test for other significant variables affecting encounter dur- (range –. m) for females and . ± SD . m (range ation. A linear mixed effects model, with year and month –. m) for males (Fig. ), with no significant difference retained as nested random effects, demonstrated that scar-  between the sexes (t-test, t = −.,df=,P=.). ring was highly significant (linear mixed effects, χ = ., The total number of encounters with identified sharks P=.). Encounters with individuals that had no appar- (counted as the number of days on which interactions ent scarring were significantly shorter than with individuals with a particular shark were recorded) was – over the that had some degree of scarring (linear mixed effects,  course of the study, although the inclusion of additional χ = .,P=.); the severity of the scarring did not sig- data from the Wildbook library showed that % of the indi- nificantly affect encounter duration (linear mixed effects,  vidual sharks considered in this study had been exposed to χ = .,P=.). There was no significant relationship swimmers prior to the start of this work in . Overall, between the total length of an individual and the degree of     sharks were encountered on a mean of . ± SD . occa- scarring observed (ANOVA, F(,) = . ,P=. ). sions, with up to  encounters per individual (Fig. ). The Encounters were significantly longer when sharks were  reliance on suitable photographs having been submitted to feeding (linear mixed effects, χ = .,P=.) and the library means that these figures represent the minimum when more swimmers were present (linear mixed effects,  number of previous encounters. Scarring was observed on χ = .,P=.). .% of identified individuals and whale sharks were ob- The number of whale sharks encountered was not served feeding during .% of all encounters. correlated with search effort between years (ANOVA,     The mean duration of encounter was  minutes  s ± F(,) = . ,P= . ). There was a significant difference   SD  minutes  s(n=). One or more avoidance re- in encounter rates between years (ANOVA, F(,) = . , sponses were recorded during .% of encounters. P=.) but analysis of -month blocks indicates that the Evaluation of  encounters with  identified sharks in- source of this difference was the second half of , for dicated that encounter duration was significantly related to which July and August were not included (ANOVA,   #   expressed avoidance response (generalized linear mixed F(,) = . ,P . ). Encounter rate was generally  model, χ = .,P=.). The mean duration of low during the austral winter, so excluding these  months encounter when a shark showed no avoidance was  min- may have affected this result. No other variables were utes  s ± SD  minutes s(n=), which is significantly significant. There was no significant difference in the longer than encounters with sharks that exhibited avoid- frequency of avoidance behaviours expressed between the   ance (generalized linear mixed model, χ = ., five -month periods (χ = .,P=.). Encounter P , .). There was no significant difference in the duration was significantly shorter in  ( minutes  s ± duration of encounters where one or more avoidance SD  minutes  s) than in  ( minutes  s ± SD  behaviours were expressed (generalized linear mixed minutes  s) and  ( minutes  s ± SD  minutes     #    model, χ = .,P=.); mean duration of such en- s; ANOVA, F(,) = . ,P . ). A total of counters was  minutes  s ± SD  minutes  s(n=). encounters with  whale sharks were analysed to establish A binomial response (avoidance or no avoidance) was that the number of previous encounters with swimmers did therefore used in subsequent analyses. not affect the likelihood that a whale shark would display

avoidance behaviour (generalized linear mixed model, Whale shark encounters were spatially aggregated along  χ = .,P=.).  km of coast immediately south of Tofo Beach. Although the sharks are also sighted more widely along the Mozambican coast (Cliff et al., ), the small size of the Discussion primary aggregation area and its almost complete daily search coverage by commercial operators suggest the possi- We found no significant longitudinal effects on encounter bility of displacement if whale sharks have negative inter- duration or the raw proportion of avoidance behaviours ex- actions with swimmers or boats. Over half (%) of pressed over the -month study period. Furthermore, there identified sharks had some form of scarring, although not was no significant relationship between the number of all of these were from anthropogenic sources such as propel- previous tourist encounters and the likelihood of avoidance ler strikes. This percentage is lower than that recorded in behaviours being displayed by individual sharks. At this Djibouti, where propeller or boat strike scars were observed stage there is no evidence of short-term avoidance beha- on % of identified sharks (Rowat et al., ), but empha- viours translating into medium- to longer-term behavioural sizes the importance of working with skippers to ensure that change among the sharks present in Mozambican waters. a safe minimum distance from the sharks is maintained The potential negative effects of tourism may be ameli- (Pierce et al., ). Although tourism operators have not orated by the ecology and population structure of whale perceived a shift in whale shark distribution, the lack of sharks in Mozambique. The majority of sharks encountered standardized spatial sampling coverage precludes a defini- were male (%). Maturity in male sharks occurs when they tive assessment. reach a total length of c.  m, based on records from Our results support the use of encounter duration as a Ningaloo Reef (Norman & Stevens, ), and in female means of measuring the effect of swimmer presence on sharks at a total length of . . m, based on South whale shark behaviour, as sharks that are disturbed will rou- African and other records (Beckley et al., ; Norman tinely end an encounter by diving or increasing their swim- & Stevens, ). Given that no sharks of total length ming speed. However, it is important to note that encounter . . mor,  m were encountered, most sharks were duration, although a useful proxy of avoidance, may only be immature and were probably several years old at the time able to provide resolution at the binomial scale (avoidance of first sighting (Wintner, ). Therefore, tourism at or no avoidance). Encounter duration has previously been this site is unlikely to interfere with reproduction, if it used to quantify avoidance by Hector’s dolphins does in fact occur in these waters. The majority (%) of Cephalorhynchus hectori of swimmers in New Zealand sharks had only one or two encounters recorded on the (Bejder et al., ) and to assess long-term trends in en- Wildbook photo-identification library by  June  counters with dwarf minke whales Balaenoptera acutoros- (with the earliest record from ), including from this trata on the Great Barrier Reef, Australia (Birtles et al., study, suggesting that many of the sharks are transient to ). However, these studies used different definitions of the area, as noted in other aggregations (Holmberg et al., encounter duration: Bejder et al. () defined it as the ; Rowat et al., ; Fox et al., ). There was no evi- length of time that a swimmer spent within  mof dence of a learnt avoidance behaviour amongst individual the nearest dolphin and Birtles et al. () defined it as sharks and, based on data from the Philippines (Quiros, the time between the first sighting of a whale and the end ), a degree of habituation may be more likely to occur. of the vessel’s contact with that whale. Here we define The percentage of whale sharks observed feeding during encounter duration as the time between the first swimmer daylight hours at Tofo (.%) is comparable to the rates re- entering the water in the presence of a whale shark and ported at Donsol in the Philippines (%in and %in the last swimmer returning to the boat. This definition ; Quiros, ) but substantially lower than the % re- could be less useful in areas with significant currents or ported from Bahía de Los Angeles, Mexico, where research- where the boat would take longer to retrieve swimmers. ers targeted feeding whale sharks (Nelson & Eckert, ). We found a positive relationship between the number of Contrary to results from the Philippines, where feeding swimmers and encounter duration. This result may be an sharks were . times more likely to exhibit a dive response artefact of the definition in use, as larger groups of swim- in the presence of swimmers (Quiros, ), encounter mers are likely to contain members with a wider range of durations in this study were significantly longer when feed- swimming abilities. The presence of more able swimmers ing behaviour was observed. Whale sharks are flexible in is likely to lead to an increase in encounter duration, as their foraging strategies, depending on prey type and verti- they are able to maintain visual contact with a shark for cal distribution (Graham et al., ; Nelson & Eckert, ; longer than slower or less fit individuals. Another important Taylor, ; Brunnschweiler et al., ; Motta et al., ), point to consider in future work is the number of swimmers so the effects of tourism on feeding are likely to be that are actually in contact with the shark from a behav- context-specific. ioural perspective, investigating how swimmer proximity

may influence these results. Future studies that rely on in- characteristics in terms of population structure. The results water observers could trial re-defining encounter duration of this study demonstrate that encounter duration is a useful as the time between the observer’s first and last visual con- metric for assessing whale shark behaviour, and provide a tact of the shark underwater, to reduce the influence of template for future studies in Mozambique and other swimming ability and group size on results. locations. Consistent with results from Quiros () but contrary to Norman () encounter duration was significantly longer with scarred whale sharks. This could be because Acknowledgements scarred individuals have slower reaction times and a re- We thank the All Out Africa volunteers and staff for their duced level of agility as a consequence of their previous efforts in collecting the majority of the data used in this injuries. However, the severity of scarring did not signifi- study, and Marine Megafauna Foundation staff, volunteers cantly alter encounter duration. Another hypothesis is that and students who also contributed. We thank the developers older individuals, who would presumably be more likely to of, and contributors to, the Wildbook whale shark database. have accumulated scars, show less avoidance than younger The support of Casa Barry Lodge and Tofo Scuba through- individuals. However, we found no significant relationship out the field portion of the project is greatly appreciated. between shark size and the severity of scarring. A third This study was made possible by support from the plausible explanation is that some sharks have a slower nat- National Marine Aquarium (UK), Shark Foundation, GLC ural reaction time to potential threats and are therefore Charitable Trust, Project AWARE Foundation, the Rufford more likely to accumulate injuries than sharks with faster re- Small Grant Foundation, PADI Foundation, Ocean actions. Sharks may also vary in their inquisitiveness. Revolution, Fondation Ensemble and private donors. We Further research is required to confirm or refute these thank Philip Dearden, Adrian Gutteridge, Conrad Speed hypotheses. and the anonymous reviewers for their comments on this A mean of . swimmers per encounter was recorded. article. This exceeds the maximum recommended by most management plans, which limit swimmer numbers to three in   Mexico (Remolina Suárez et al., ) and in Australia References (Catlin & Jones, ), although these limits may be par- tially influenced by operational constraints in those loca- ARZOUMANIAN, Z., HOLMBERG,J.&NORMAN,B.()An tions. From a management perspective it is also important astronomical pattern-matching algorithm for computer-aided to consider customer satisfaction with the experience, which identification of whale sharks Rhincodon typus. Journal of Applied Ecology, , –. can be significantly reduced in crowded scenarios (Davis BECKLEY, L.E., CLIFF, G., SMALE, M.J. & COMPAGNO, L.J.V. () et al., ). Recent strandings and sightings of whale sharks in South Africa. The inclusion of well-defined and easily observed para- Environmental Biology of Fishes, , –. meters of known significance that can be recorded by either BEJDER, L., DAWSON, S.M. & HARRAWAY, J.A. () Responses by ’ trained specialists (e.g. researchers) or non-specialist obser- Hector s dolphins to boats and swimmers in Porpoise Bay, New Zealand. Marine Mammal Science, , –. vers (e.g. tour operators or volunteers) would increase the BEJDER,L.&SAMUELS,A.() Evaluating the effects of amount of information available, enabling routine sustain- nature-based tourism on cetaceans. In Marine Mammals: Fisheries, ability assessments and facilitating inter-site comparisons Tourism and Management Issues (eds N. Gales, M. Hindell & (Graham, ). We have created a template data collection R. Kirkwood), pp. –. CSIRO Publishing, Melbourne, sheet (Supplementary Material ), which is currently in use Australia. BIRTLES, R.A., ARNOLD,P.W.&DUNSTAN,A.() Commercial in Mozambique. We also encourage the submission of swim programs with dwarf minke whales on the northern Great standardized photographs of all whale sharks encountered Barrier Reef, Australia: some characteristics of the encounters with to the global whale shark photo-identification library management implications. Australian Mammalogy, , –. (Wild Me, ). This facilitates the incorporation of BRADSHAW, C.J.A., F ITZPATRICK, B.M., S TEINBERG, C.C., B ROOK,B.W. individual-based analyses into future studies, concurrently &MEEKAN, M.G. () Decline in whale shark size and abundance at Ningaloo Reef over the past decade: the world’s largest fish is getting enabling more accurate regional and ocean-wide studies of  –  smaller. Biological Conservation, , . population size and interconnectivity (Graham, ; BROOKS, K., ROWAT, D., PIERCE, S.J., JOUANNET,D.&VELY,M. Brooks et al., ). () Seeing spots: photo-identification as a regional tool for whale With increasing global interest in human interactions shark identification. Western Indian Ocean Journal of Marine with whale sharks there is a clear need for the implementa- Science, , –.  tion of monitoring initiatives at tourism sites to understand BRUNNSCHWEILER, J.M. ( ) The Shark Reef Marine Reserve: a marine tourism project in Fiji involving local communities. Journal and avoid long-term anthropogenic effects. Other import- of Sustainable Tourism, , –. ant whale shark tourism sites, such as those in Mexico, BRUNNSCHWEILER, J.M., BAENSCH, H., PIERCE, S.J. & SIMS, D.W. the Seychelles and Western Australia, have similar () Deep-diving behaviour of a whale shark, Rhincodon typus,

during long-distance movement in the western Indian Ocean. MEEKAN, M.G., BRADSHAW, C.J.A., PRESS, M., MCLEAN, C., Journal of Fish Biology, , –. RICHARDS, A., QUASNICHKA,S.&TAYLOR, J.G. () Population CATLIN,J.&JONES,R.() Whale shark tourism at Ningaloo size and structure of whale sharks Rhincodon typus at Ningaloo Reef, Marine Park: a longitudinal study of wildlife tourism. Tourism Western Australia. Marine Ecology Progress Series, , –. Management, , –. MOTTA, P.J., MASLANKA, M., HUETER, R.E., DAVIS, R.L., DE LA CATLIN, J., JONES, T., NORMAN,B.&WOOD,D.() Consolidation PARRA, R., MULVANY, S.L. et al. () Feeding anatomy, in a wildlife tourism industry: the changing impact of whale shark filter-feeding rate, and diet of whale sharks Rhincodon typus during tourism expenditure in the Ningaloo coast region. International surface ram filter feeding off the Yucatan Peninsula, Mexico. Journal of Tourism Research, , –. Zoology, , –. CLIFF, G., ANDERSON-READE, M.D., AITKEN, A.P., CHARTER, G.E. & NELSON, J.D. & ECKERT, S.A. () Foraging ecology of whale sharks PEDDEMORS, V.M. () Aerial census of whale sharks (Rhincodon (Rhincodon typus) within Bahía de Los Angeles, Baja California typus) on the northern KwaZulu-Natal coast, South Africa. Fisheries Norte, Mexico. Fisheries Research, , –. Research, , –. NORMAN,B.() Rhincodon typus.InThe IUCN Red List of CLUA, E., BURAY, N., LEGENDRE, P., MOURIER,J.&PLANES,S.() Threatened Species v. .. Http://www.iucnredlist.org [accessed  Business partner or simple catch? The economic value of the March ]. sicklefin lemon shark in French Polynesia. Marine and Freshwater NORMAN, B.M. () Aspects of the biology and ecotourism industry of Research, , –. the whale shark Rhincodon typus in north-western Australia. MPhil DAVIS, D., BANKS, S., BIRTLES, A., VALENTINE,P.&CUTHILL,M. thesis. Murdoch University, Perth, Australia. () Whale sharks in Ningaloo Marine Park: managing tourism in NORMAN, B.M. & STEVENS, J.D. () Size and maturity status of the an Australian marine protected area. Tourism Management, , – whale shark (Rhincodon typus) at Ningaloo Reef in Western . Australia. Fisheries Research, , –. DE LA PARRA VENAGAS, R., HUETER, R.E., GONZALES CANO, J., PIERCE, S.J., MENDEZ-JIMENEZ, A., COLLINS, K., ROSERO-CAICEDO, TYMINSKI, J.P., GREGORIO REMOLINA, J., MASLANKA, M. et al. M. & MONADJEM,A.() Developing a Code of Conduct for () An unprecedented aggregation of whale sharks, Rhincodon whale shark interactions in Mozambique. Aquatic Conservation: typus, in Mexican coastal waters of the Caribbean Sea. PLoS ONE, Marine and Freshwater Ecosystems, , –. (), e. QUIROS, A.L. () Tourist compliance to a Code of Conduct and the FITZPATRICK, R., ABRANTES, K.G., SEYMOUR,J.&BARNETT,A.() resulting effects on whale shark (Rhincodon typus) behavior in Variation in depth of whitetip reef sharks: does provisioning Donsol, Philippines. Fisheries Research, , –. ecotourism change their behaviour? Coral Reefs, , –. RDEVELOPMENT CORE TEAM () R: A Language and Environment FOX, S., FOISY, I., DE LA PARRA VENEGAS, R., GALVÁN PASTORIZA,B. for Statistical Computing. R Foundation for Statistical Computing, E.GRAHAM, R.T., HOFFMAYER, E.R. et al. () Population Vienna, Austria. structure and residency of whale sharks Rhincodon typus at Utila, REMOLINA SUÁREZ, J.F., PÉREZ RAMÍREZ, J.J., GONZÁLEZ CANO, J.M., Bay Islands, Honduras. Journal of Fish Biology, , –. DE LA PARRA VENAGAS, R., BETANCOURT SABATINI, N., TRIGO GALLAGHER, A.J. & HAMMERSCHLAG,N.() Global shark MENDOZA, M. et al. () Whale shark management strategies, currency: the distribution, frequency, and economic value of shark with the participation of local stakeholders, in Yum Balam, Mexico. tourism. Current Issues in Tourism, , –. In The First International Whale Shark Conference: Promoting GRAHAM, R.T. () Whale sharks of the western Caribbean: an International Collaboration in Whale Shark Conservation, Science overview of current research and conservation efforts and future and Management (eds T.R. Irvine & J.K. Keesing), pp. –. CSIRO needs for effective management of the species. Gulf and Caribbean Marine and Atmospheric Research, Perth, Australia. Research, , –. ROHNER, C.A., PIERCE, S.J., MARSHALL, A.D., WEEKS, S.J., BENNETT, GRAHAM, R.T. & ROBERTS, C.M. () Assessing the size, growth rate M.B. & RICHARDSON, A.J. () Trends in sightings and and structure of a seasonal population of whale sharks (Rhincodon environmental influences on a coastal aggregation of manta rays and typus Smith ) using conventional tagging and photo whale sharks. Marine Ecology Progress Series, , –. identification. Fisheries Research, , –. ROHNER, C.A., RICHARDSON, A.J., MARSHALL, A.D., WEEKS, S.J. & GRAHAM, R.T., ROBERTS, C.M. & SMART, J.C.R. () Diving PIERCE, S.J. () How large is the world’s largest fish? Measuring behaviour of whale sharks in relation to a predictable food pulse. whale sharks Rhincodon typus with laser photogrammetry. Journal Journal of the Royal Society Interface, , –. of Fish Biology, , –. HEYMAN, W.D., CARR, L.M. & LOBEL, P.S. () Diver ecotourism ROWAT,D.,MEEKAN,M.G.,ENGELHARDT,U.,PARDIGON,B.&VELY, and disturbance to reef fish spawning aggregations: it is better to be M. () Aggregations of juvenile whale sharks (Rhincodon typus)inthe disturbed than to be dead. Marine Ecology Progress Series, , – Gulf of Tadjoura, Djibouti. Environmental Biology of Fishes, , –. . ROWAT, D., SPEED, C.W., MEEKAN, M.G., GORE, M.A. & BRADSHAW, HOLMBERG, J., NORMAN,B.&ARZOUMANIAN,Z.() Robust, C.J.A. () Population abundance and apparent survival of the comparable population metrics through collaborative Vulnerable whale shark Rhincodon typus in the Seychelles photo-monitoring of whale sharks Rhincodon typus. Ecological aggregation. Oryx, , –. Applications, , –. SPEED, C.W., MEEKAN,M.G.,ROWAT,D.,PIERCE,S.J.,MARSHALL, HOLMBERG, J., NORMAN,B.&ARZOUMANIAN,Z.() Estimating A.D. & BRADSHAW,C.J.A.() Scarring patterns and relative population size, structure, and residency time for whale sharks mortality rates of Indian Ocean whale sharks. Journal of Fish Rhincodon typus through collaborative photo-identification. Biology, , –. Endangered Species Research, , –. TAYLOR, J.G. () Seasonal occurrence, distribution and movements MARSHALL, A.D. & PIERCE, S.J. () The use and abuse of of the whale shark, Rhincodon typus, at Ningaloo Reef, Western photographic identification in sharks and rays. Journal of Fish Australia. Marine and Freshwater Research, , –. Biology, , –. TAYLOR, J.G. () Ram filter-feeding and nocturnal feeding of whale MAU,R.() Managing for conservation and recreation: the sharks (Rhincodon typus) at Ningaloo Reef, Western Australia. Ningaloo whale shark experience. Journal of Ecotourism, , –. Fisheries Research, , –.

TIBIRIÇÁ, Y., BIRTLES, A., VALENTINE,P.&MILLER, D.K. () ANDREW MC GOWAN’ S research interests include avian cooperative Diving tourism in Mozambique: an opportunity at risk? Tourism in breeding, conservation of sea turtles, and tropical seabird populations. Marine Environments, , –. ANNA WESTLING is a coral reef ecologist interested in citizen science. VIANNA, G.M.S., MEEKAN, M.G., PANNELL, D.J., MARSH, S.P. & ADRIANA MÉNDEZ-JIMÉNEZ is interested in the human dimension of MEEUWIG, J.J. () Socio-economic value and community the environment, and in the way people perceive environmental is- benefits from shark-diving tourism in Palau: a sustainable use of reef sues. CHRISTOPH ROHNER is interested in the movement and feeding shark populations. Biological Conservation, , –. ecology of large pelagic animals, and linking their behaviour to bio- WILD ME () Wildbook for Whale Sharks. Http://www.whaleshark. logical and oceanographic drivers. KYM COLLINS specializes in the org [accessed  March ]. behaviour and acoustics of marine mammals. MARCELA ROSERO- WINTNER,S.P.() Preliminary study of vertebral growth CAICEDO is interested in marine conservation and specializes in sci- rings in the whale shark, Rhincodon typus, from the east entific and environmental communication in volunteer programmes coast of South Africa. Environmental Biology of Fishes, , and local communities. JODI SALMOND studies Australian reef eco- –. systems in a global context, specializing in citizen science and marine education. ARA MONADJEM is interested in the application of quan- titative ecological techniques to conservation-related problems in Biographical sketches small mammals and birds. ANDREA MARSHALL studies the biology, ecology, monitoring and management of manta rays globally. SIMON PETER HASKELL specializes in behavioural ecology and is interested PIERCE is a marine conservation ecologist specializing in whale in the role that tourism can play in elasmobranch conservation. sharks.