Habitat fragmentation can impede an animal’s ability to move through their habitat, affecting both local and long-distance movements. Each year, polar bears Ursus maritimus migrate to refuge habitats on land or to multiyear ice as annual sea ice breaks up. We used polar bear telemetry location data from 39 adult female polar bears tracked in Hudson Bay in 2013-2018 during break-up (2 May-23 July) to show variation in migratory movement and timing as break-up advances. We separated break-up into early and late periods and used standard deviation in temporal spatial autocorrelation (SASD) of sea ice concentration to quantify sea ice fragmentation. Higher spatial autocorrelation reflects dissimilarity in local habitat composition at a single point in time, while SASD reflects variation in local habitat composition over time. In late break-up, there was a significant positive correlation between polar bear path tortuosity and SASD. Individuals arrived on land significantly later when paths moved through sea ice with increasing SASD in late break-up. Reproductive status of adult female polar bears had no effect on the variability of the sea ice an individual travelled through. SASD provides a means of summarizing the complexity and dynamics of sea ice habitat and can be used to understand variation in movement and ecology of ice-associated organisms.

No abstract

Narwhals (Monodon monoceros) are an important subsistence harvest species for Inuit communities and their conservation is important for Inuit culture and ecosystem function. The northern Hudson Bay (NHB) narwhal population, which spends summer in northern Hudson Bay, Canada, has been assessed through periodic aerial surveys from 1981 to 2018. To estimate the population trajectory and predict future population trends under various harvest scenarios, a Bayesian population model was fit to four aerial survey estimates and harvest data from 1951 to 2018. The model resulted in a 2019 population estimate of ~14,400, 95% CI [10,300, 20,400] and an estimated starting population of 7,200, 95% CI [1,400, 19,000] in 1951. The model was extended 10 years into the future under three annual harvest scenarios (current harvest: 157, low harvest: 50, and high harvest: 300) and the probability of population decline was estimated. The model predicted a 6% chance of decline with an annual harvest quota of 50 narwhals, 78% for a harvest of 157, and 95% for a harvest of 300. This updated model provides the opportunity to shape conservation efforts by estimating past population trends and how those trends, combined with management action, can affect future population dynamics.

No abstract