Polar bear predatory behaviour reveals seascape distribution of ringed seal lairs

Ringed seal (Pusa hispida) breeding distribution has been extensively studied across near-shore habitats, but has received limited attention at a seascape scale due to the difficulty in accessing offshore sea ice environments. Employing highly visible predation attempts by polar bears (Ursus maritimus) on ringed seals in subnivean lairs observed by helicopter, the spatial relationship between predatory behaviour and ringed seal breeding habitat was examined. Resource selection functions were used to determine the relative probability of predation attempts on ringed seals in lairs as a function of habitat during a period of low ringed seal natality (2004–2006). Ringed seal pup kill locations were compared between years of low (2003–2006) and high (2007–2011) natality to assess the effect of reproductive output on habitat use. During years of low natality, polar bear hunting attempts were more likely in near-shore fast ice, and pup kills were observed predominately in fast ice (fast = 65 %, pack = 29 %, P = 0.002) at a median distance of 36 km from shore. In years of high natality, pup kills were observed farther from shore (median = 46 km, P = 0.03), and there was no difference in the proportion of observations in fast ice and pack ice (fast = 43 %, pack = 52 %, P = 0.29). These results suggest that the facultative use of adjacent offshore pack ice by breeding ringed seals may be influenced by natality. This study illustrates how documenting the behaviour of a predator can facilitate insight into the distribution of a cryptic prey.