Geographical assignment of polar bears using multi-element isoscapes

Sci Rep. 2019 Jun 28;9(1):9390. doi: 10.1038/s41598-019-45874-w.

Abstract

Wide-ranging apex predators are among the most challenging of all fauna to conserve and manage. This is especially true of the polar bear (Ursus maritimus), an iconic predator that is hunted in Canada and threatened by global climate change. We used combinations of stable isotopes (13C,15N,2H,18O) in polar bear hair from > 1000 individuals, sampled from across much of the Canadian Arctic and sub-Arctic, to test the ability of stable isotopic profiles to 'assign' bears to (1) predefined managed subpopulations, (2) subpopulations defined by similarities in stable isotope values using quadratic discriminant analysis, and (3) spatially explicit, isotopically distinct clusters derived from interpolated (i.e. 'kriged') isotopic landscapes, or 'isoscapes', using the partitioning around medoids algorithm. A four-isotope solution provided the highest overall assignment accuracies (~80%) to pre-existing management subpopulations with accuracy rates ranging from ~30-99% (median = 64%). Assignment accuracies of bears to hierarchically clustered ecological groups based on isotopes ranged from ~64-99%. Multivariate assignment to isotopic clusters resulted in highest assignment accuracies of 68% (33-77%), 84% (47-96%) and 74% (53-85%) using two, three and four stable isotope groups, respectively. The resulting spatial structure inherent in the multiple stable isotopic compositions of polar bear tissues is a powerful forensic tool that will, in this case, contribute to the conservation and management of this species. Currently, it is unclear what is driving these robust isotopic patterns and future research is needed to evaluate the processes behind the pattern. Nonetheless, our isotopic approach can be further applied to other apex mammalian predators under threat, such as the large felids, providing that isotopic structure occurs throughout their range.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Animals
  • Canada
  • Climate Change
  • Geography
  • Isotope Labeling
  • Population Dynamics*
  • Spatial Analysis*
  • Ursidae*