Gray Seal Cannibalism at the Largest Colony in the World, Sable Island

Intraspecific predation, or cannibalism, is found across taxonomic groups and has been associated with high population densities and limited availability of alternative prey resources (Fox 1975; Polis 1981). In otariids, that are highly sexually dimorphic and polygynous, infanticide at breeding colonies has been attributed to sexual aggression by subordinate males (Campagna et al. 1988). However, adult males of some species such as Steller sea lions (Eumetopias jubatus) and New Zealand sea lions (Phocarctos hookeri) have been reported killing and subsequently feeding on pups, which presumably provides some nutritional benefit during the breeding season (Ryazanov et al. 2018; Wilson et al. 2010). Furthermore, the closely related behavior of interspecific (intraguild) predation of marine mammals has been reported in a number of pinniped populations (Baylis et al. 2024; Boveng et al. 1998; Bradshaw et al. 1998; Byrnes and Hood 1994; Gentry and Johnson 1981; Haelters et al. 2012; Harcourt 1991; Hiruki et al. 1999; Penry et al. 2013; Robinson et al. 1999; Siniff and Bengtson 1977; van Neer et al. 2015). Steller sea lion predation of harbor seals (Phoca vitulina richardii) in Alaska has been attributed to the rapid expansion of sea lion populations, which increased the frequency of adverse interspecific interactions (Mathews and Adkison 2010), and the apparent increase in gray seal (Halichoerus grypus) predation of harbor seals (P. v. vitulina) in Scotland (Langley et al. 2025) correlated to an exponential increase in gray seal pup production in the North Sea (Thomas et al. 2019). However, it is difficult to discern whether these correlations are driven by the foraging strategy emerging in response to increased pressure on prey resources, or if the events are more readily observed as densities increase. Cannibalism was first documented in Atlantic gray seals on Amet Island, Nova Scotia, in 1992, when an adult male was observed feeding on a weaned pup during the breeding season (Bédard et al. 1993). Cannibalism events were recorded in their entirety in subsequent breeding seasons (Kovacs et al. 1996) and years later across the Atlantic at a breeding colony in southeast Scotland (Bishop et al. 2016). Cannibalism and intraguild predation of sympatric harbor seals (van Neer et al. 2015) and harbor porpoises, Phocoena phocoena (Bouveroux et al. 2014; Haelters et al. 2012; van Bleijswijk et al. 2014) have since been found to be widespread throughout gray seal subpopulations and exhibited by multiple, seemingly specialist individual adult males (Langley et al. 2025). Unlike other pinniped examples of intraguild predation, gray seal predation of harbor seals is not restricted to the gray seal breeding season but was found in Scotland to peak during the harbor seal breeding season (Langley et al. 2025). Phocid carcasses with significant traumatic injuries have been reported throughout the northwest Atlantic since at least 1980, including on Sable Island (Lucas and Stobo 2000). Some of these cases were attributed to predation by white sharks (Carcharodon carcharias) based on tooth fragments and rake marks, but most carcasses (including gray seal pups and harbor seal adults and pups) had what have been described as spiral or corkscrew lacerations attributed to a different marine predator, proposed to be the Greenland shark (Somniosus microcephalus; Lucas and Natanson 2010). In the northeast Atlantic, an alternative hypothesis was that the injuries were consistent with seals being drawn through the ducted propellers of ships (Thompson et al. 2010) until the behavior was directly observed on a breeding colony in southeast Scotland (Bishop et al. 2016). Direct observations were used to ground-truth strandings data, and these spiral lacerations, along with other pathological attributes, including seal claw rake marks in the blubber and canine punctate lesions on the muzzle, were attributed to predation by gray seals (Brownlow et al. 2016). Evaluating the cause of death of carcasses with these types of injuries is therefore critical as incorrectly attributing mortalities to an anthropogenic origin, and/or to alternative predators, has management implications. The individual gray seal recorded by drone in 2023 (observation 1) could not be identified through photo ID, and photo ID data from the at-sea events in 2024 (observations 2 and 3) were below the quality required for conservative mark-recapture analysis. However, unique pelage and scarring patterns visible from individuals involved in these at-sea events, along with good quality photo ID data from events recorded on the beach (observations 4 and 5), indicate that observations reported here involved multiple adult male gray seals. There are a number of pathological attributes of marine mammal carcasses associated with predation by gray seals (Brownlow et al. 2016; van Neer et al. 2020, 2021). The following attributes listed by Brownlow et al. (2016) were present in gray seal pup carcasses on Sable Island 2023–2025 and photographic examples were taken (Figure 2): significant areas of the skin or tissue missing, undermining of blubber, rake marks in the blubber, smooth wound margin, single linear lesion with one or more rotations (i.e., corkscrew laceration), ragged wound margin, evidence of skeletal trauma, avulsion of one or both scapula, punctate lesions everywhere, lesion beginning at the mouth, and punctate lesions on the muzzle. The only pathological attribute listed in Brownlow et al. (2016) that was not documented on Sable Island was skeletal trauma to the head. Notably, fractures to the face and jaw were recorded in corkscrew carcasses from Sable Island in a previous study attributing these injuries to Greenland sharks (Lucas and Natanson 2010). Moreover, that study described further attributes that we also reported here: wounds running between the chin and pectoral area (Figure 2j), no (apparent) tissue missing (Figure 2e), partial or complete removal of the fore-flippers and scapulas, that is, the “jacket” (Figure 2e,h), and a clean-edged chevron-shaped wound at the base of the dorsal side indicative of “jacket” removal (Figure 2e). In addition to the independent observations of gray seal killing, feeding and cannibalism events (Figure 1), and the evidence of gray seal cannibalism in strandings (Figure 2), the number of dead pups with spiral lacerations was recorded along a 20 km stretch of both North and South beach in the winter breeding seasons of 2023 and 2024; the first systematic surveys of carcasses with these injuries since 2003 (Lucas and Natanson 2010). In 2023, 11 surveys were completed twice a week between peak pupping in early January to the end of the research program in early February, and carcasses were tagged with cattle ear tags so that counts were not duplicated across surveys. A total of 765 pup carcasses with spiral lacerations were recorded on North beach (n = 506) and South beach (n = 259). Of these, 68% on North beach (n = 344) and 92% on South beach (n = 237) were resighted at least once over this period. In 2024, carcasses were not tagged, but the maximum recorded in any one day was 359 on the final survey of the season on 7 February: 211 on North beach and 148 on South beach. These counts and resightings suggest that during the 2023 and 2024 breeding seasons, gray seal predation was more prevalent on North beach, and carcasses on North beach were also more likely to be missed from surveys, possibly buried in the sand or washed out to sea at a greater rate than on South beach. Observations of cannibalism have been reported from gray seal populations across the North Atlantic for a number of decades (Bédard et al. 1993; Bishop et al. 2016; Kovacs et al. 1996; Langley et al. 2025) despite it never being directly observed on Sable Island. Indeed, the number of pup carcasses with spiral lacerations counted here, in a narrow temporal window during the breeding season, demonstrated that most cannibalism events go unobserved, particularly as strandings data underrepresent gray seal predation (i.e., carcasses are scavenged by birds and can be buried in the sand, washed out to sea, or were never on land and therefore unavailable to be counted). Indeed, the colony is surveyed weekly throughout the breeding season (Badger et al. 2023; den Heyer et al. 2014), which would suggest that gray seal intraspecific predation events mostly occur at times or locations where direct observations are unlikely; for example, at night, at sea, during high tide, and/or in stormy conditions. Observations reported here were biased toward land-based surveys in fair to good weather conditions, during daylight hours, and at low to mid-tidal states when beaches were more accessible. Direct observations are opportunistic and therefore often missed during dedicated surveys, which is why this type of data collection elsewhere lends itself to citizen-science (e.g., Langley et al. 2025). On Sable Island, where there is minimal observation effort year-round, sampling is limited to secondary evidence of mortality through strandings data, which is why detailed understanding into the pathology of different causes of death is critical. There are an estimated 80,000 gray seal pups born annually on Sable Island, which represents around 80% of the breeding population in the northwest Atlantic (den Heyer et al. 2024). The Sable Island gray seal population increased exponentially from the 1960s to the 2000s and in recent years has seen reduced growth associated with a documented decline in the survival of pups from weaning to recruitment (Bowen et al. 2003; den Heyer et al. 2014; Hammill et al. 2023; Rossi et al. 2021). Although there is currently no evidence that the predation of juvenile gray seals in the northwest Atlantic is impacting the gray seal population trend, a focused effort to estimate causes of death and mortality rates of weaned pups, young of the year, and adults would be needed to assess the contribution of cannibalism. Conversely, during the period of exponential increase in the gray seal population, the harbor seal population on Sable Island collapsed (Bowen et al. 2003). This decline was attributed to predation by sharks (Boulva and McLaren 1979; Bowen et al. 2003; Brodie and Beck 1983; Lucas and Stobo 2000) and nutritional stress driven by prey competition with gray seals, although no evidence for decreased provisioning of offspring was observed (Bowen et al. 2003). The current study suggests that many, if not all, of the apparent Greenland shark predation events previously documented could have been caused by adult male gray seals. Thus, the collapse of the harbor seal population on Sable Island should be investigated in the context of gray seal predation and the potential impact this could have on marine mammal population dynamics. Izzy Langley: conceptualization, methodology, formal analysis, data curation, writing – original draft, writing – review and editing. Damian Lidgard: conceptualization, methodology, data curation, writing – review and editing. Priyanka Varkey: conceptualization, methodology, data curation, writing – review and editing. Milagros Sanchez: conceptualization, methodology, data curation, writing – review and editing. Michelle Rivard: conceptualization, methodology, data curation, writing – review and editing. Cornelia E. den Heyer: conceptualization, methodology, data curation, writing – review and editing, funding acquisition. The authors would like to acknowledge all staff and volunteers who assisted in data collection, including those from Parks Canada, DFO and collaborating institutes, without whom this study would not have been possible. This work was supported by Fisheries and Oceans Canada (Research Permits SINP-2020-37578, SINPR-20023-45671). I.L. was also supported by the Ocean Frontier Institute Visiting Fellowship and was awarded ethical approval by the University of St Andrews Animal Welfare and Ethics Committee (BL18476). This work was supported by Fisheries and Oceans Canada and Ocean Frontier Institute. The authors declare no conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request.