FISH andFISHERIES Social behaviour in sharks and rays: analysis, patterns and implications for conservation David M P Jacoby1,2, Darren P Croft2 & David W Sims1,3 1Marine Biological Association of the United Kingdom, The Laboratory, Citadel Hill, Plymouth PL1 2PB, UK; 2Centre for Research in Animal Behaviour, College of Life and Environmental Sciences, University of Exeter, Exeter EX4 4QG, UK; 3Marine Biology and Ecology Research Centre, School of Marine Sciences and Engineering, University of Plymouth, Drake Circus, Plymouth PL4 8AA, UK Abstract Correspondence: There are widespread records of grouping behaviour in both adult and juvenile sharks David MP Jacoby, Marine Biological and rays (Class Chondrichthyes, Subclass Elasmobranchii). Yet despite burgeoning Association of the descriptions of these events, many of the proximate and ultimate causes of group United Kingdom, living in these top predators remain elusive. Given the documented negative The Laboratory, anthropogenic effects on many shark populations globally, there is an increasing Citadel Hill, Plymouth need to understand how behaviourally mediated grouping influences population PL1 2PB, UK Tel.: distributions and abundance, and the role this plays in exacerbating vulnerability to +01752 633277 fishing mortality. Here, we analyse group living in elasmobranchs: we describe our Fax: +01752 633102 current understanding of the patterns, mechanisms and functions of both aggrega- E-mail: david. tion (where grouping is not driven by social mechanisms) and social grouping (where [email protected] grouping is influenced by social interaction) and discuss some of the current methods Received 25 Feb 2011 used to study social behaviour in this taxa. In particular, social preferences in Accepted 15 Jul 2011 elasmobranchs have received relatively little attention. We propose that the study of shark aggregations may benefit from a more fine-scale analytical approach offered by detailed exploration of social interactions using social network analysis. Better understanding of the frequency and longevity of social relations, in conjunction with current long-term data on habitat use and site philopatry, will likely serve for a more informed approach to coastal and pelagic elasmobranch conservation initiatives. Keywords Aggregation, fisheries impact, sharks, social behaviour, social networks, social organization Introduction 2 Patterns of grouping behaviour in elasmobranch fishes 3 Aggregation 3 Social grouping 7 Mechanisms and functions of grouping 8 Aggregation 8 Social grouping 8 Methods for studying shark social behaviour 11 Tracking and telemetry 11 Social network analysis 12 Implications and future directions 14 Ó 2011 Blackwell Publishing Ltd DOI: 10.1111/j.1467-2979.2011.00436.x 1 Shark social behaviour D M P Jacoby et al. Conclusions 15 Acknowledgements 15 References 15 (Cutts and Speakman 1994; Herskin and Steffensen Introduction 1998). Conversely, there are costs associated with Group living and social behaviour have been grouping behaviour, typically a reduction in forag- documented in animals from a wide range of ing efficiency or an increased risk of parasite or terrestrial, freshwater and marine taxa (Krause disease transmission (Johnson et al. 2002; Hoare and Ruxton 2002). The formation of social groups et al. 2004) to name a few. As a result, the fitness of may involve both active and passive processes. For an individual in a group is likely to vary as a example, individuals may actively prefer to associ- function of both group size and composition and the ate with conspecifics and orientate to their direc- context under which grouping has occurred. Unsur- tion of locomotion (Couzin et al. 2005; Guttal and prisingly, group living has been the subject of Couzin 2010). Some fish species, for example, intense research by behavioural ecologists with show both polarized schooling behaviour, defined by particular focus on optimum group size and the highly synchronous swimming when moving from decision to join or leave a group (Caraco 1979; Coˆte´ one place to another or evading a predator, and and Poulin 1995; Krause and Ruxton 2002), the less organized, uncoordinated shoaling behaviour genetic consequences of interacting with kin when aggregating for social purposes (Pitcher (Hamilton 1964; Hain and Neff 2007), the mech- 1983). Such patterns of grouping can be main- anisms underlying patterns of social organization tained by each individual obeying a few simple, (Krause et al. 2000; Croft et al. 2005) and those localized rules of attraction orientation and repul- required to support repeated individual interaction sion (Couzin et al. 2002; Sumpter 2006). In such as social recognition and familiarity (Barber contrast, many animal aggregations do not involve and Wright 2001; Tibbetts and Dale 2007; Ward social attraction and form as a result of animals et al. 2007). being drawn to aggregate because of a limited The evolution of both shoaling and schooling resource such as food or specific habitat require- behaviour has been highly selected for in extremely ments (Johnson et al. 2002) or because of syn- variable three-dimensional (3D) aquatic environ- chronized patterns of daily or seasonal activity ments. Some small freshwater teleost fish, however, (Guttal and Couzin 2010). Thus, an important also shoal under laboratory conditions, and there- distinction needs to be made between aggregations fore, much of what we know today about social that do not involve social attraction (referred to behaviour in fish can be attributed to research on hereafter simply as aggregation) and those that do model teleost species such as the guppy (Poecilia (hereafter, social groups). For the purposes of this reticulata, Poeciliidae; Magurran et al. 1994; Croft review, ‘aggregation’ will also be referred to when et al. 2004) or the three-spined stickleback (Gaster- there is no clear indication or sufficient research to osteus aculeatus, Gasterosteidae; Ward et al. 2002, support that grouping is socially derived, although 2008; Frommen et al. 2007). future research will surely address these current Sharks and rays (Class Chondrichthyes, Subclass grey areas. Elasmobranchii; known collectively as elasmo- Animal groups arise from a complex trade-off of branchs) are also frequently observed grouping in costs and benefits associated with both conspecific large numbers; however, little is known about the and heterospecific interaction. Freshwater teleost mechanisms driving this behaviour. Indeed, quan- fishes, for example, gain antipredator benefits such tifying aggregation or social interactions in marine as the dilution of risk or the confusion effect when fishes presents a significant challenge in comparison shoaling with group mates (Krause and Ruxton with smaller, freshwater teleost species. Laboratory 2002; Hoare et al. 2004). Schooling behaviour in experiments have repeatedly demonstrated that larger fish and, equally, formation flight in some predator avoidance behaviour constitutes a com- migratory birds also appear to facilitate a reduction mon driver of shoaling among many teleost fishes in the energetic costs associated with movement (Lachlan et al. 1998; Krause et al. 2000; Hoare 2 Ó 2011 Blackwell Publishing Ltd, F I S H and F I S H E R I E S Shark social behaviour D M P Jacoby et al. et al. 2004). This idea, although never empirically interactions in gregarious animals whilst discussing tested, is often alluded to in the studies of juvenile the benefits of applying such analyses to a elasmobranch behaviour (Morrissey and Gruber K-selected species of marine predator. With impor- 1993; Heupel and Simpfendorfer 2005), whom tant recent advances in telemetry technology for themselves are likely to be vulnerable to a range large marine predators (Sims 2010) taken together of larger predators. With the exception of human with appropriate analytical approaches, the review fishing behaviour, however, many highly predatory concludes by proposing a more holistic approach to species of shark occupy apex positions within their the understanding of shark social behaviour, and respective food webs, suggesting that there are with it the potential to influence how elasmobranch arguably other significant factors dictating elasmo- populations are managed under ever more intensive branch grouping behaviour, in adults at least. fishing pressure (Baum et al. 2003). Shark aggregations and the physical or environ- ment variables that underpin these events are Patterns of grouping behaviour in reasonably well documented in the scientific liter- elasmobranch fishes ature (Economakis and Lobel 1998; Heupel and Simpfendorfer 2005; Dewar et al. 2008). In con- Elasmobranchs are highly diverse, marine verte- trast, there is considerably less known about the brate taxa that have adapted to fill apex predatory occurrence of social groups in wild sharks, although roles within the estuarine, coastal and oceanic some species have been hypothesized to engage in environment. In general contrast to bony fish, diel periods of social refuging behaviour (Sims elasmobranchs are much slower to gain maturity, 2003). It is well known, for instance, that scalloped produce fewer, more well-developed offspring and hammerhead sharks (Sphyrna lewini, Sphyrnidae), regularly live for periods of decades, rather than which are largely solitary foragers, exhibit regular, years. These K-selected life-history traits are consis- polarized schooling behaviour associated with spe- tent across all species of elasmobranch despite cific locations such as underwater seamounts substantial