Nicholas W. Pilfold, Alysa McCall, Andrew E. Derocher, Nicholas J. Lunn and Evan Richardson


Published date April 14, 2016

Migratory response of polar bears to sea ice loss: to swim or not to swim

  • States migratory responses to climate change may vary across and within populations, particularly for species with large geographic ranges
  • Predicts an increase in the frequency of long-distance swims (> 50 km) as a consequence of climate change for polar bears Ursus maritimus
  • Examines GPS satellite-linked telemetry records of 58 adult females and 18 subadults from the Beaufort Sea (BS), and 59 adult females from Hudson Bay (HB), for evidence of long-distance swimming during seasonal migrations in 2007–2012
  • Identifies 115 swims across both populations
  • Finds median swim duration was 3.4 d (range 1.3–9.3 d) and median swim distance was 92 km (range 51–404 km)
  • Finds swims were significantly more frequent in the BS (n = 100) than HB (n = 15). In the BS, subadults swam as frequently as lone adult females, but more frequently than adult females with offspring
  • ​Models the likelihood of a polar bear engaging in swims using collar data from the BS
  • Finds swims were more likely for polar bears without offspring, with the distance of the pack ice edge from land, the rate at which the pack ice edge retreated, and the mean daily rate of open water gain between June–August
  • Finds the yearly proportions of BS adult females swimming in 2004–2012 were positively associated with the rate of open water gain
  • Results corroborate the hypothesis that long-distance swimming by polar bears is likely to occur more frequently as sea ice conditions change due to climate warming. However, results also suggest that the magnitude of the effect likely varies within and between populations