Polar bears (Ursus maritimus) are managed across Nunavut, Canada, under a harvest and monitoring system that seeks to ensure harvest is sustainable and identified management objectives are achieved. In recent decades, climatic changes across the Arctic have altered polar bear habitat at unprecedented rates. To retain viable polar bear subpopulations as part of the ecosystem and provide a subsistence resource for Inuit, scientific research and monitoring studies are conducted to evaluate subpopulation status and whether management objectives are being met. Here we report the results of a population study for polar bears inhabiting the M’Clintock Channel (MC) conducted 2014 – 2016. Current samples were collected using less-invasive genetic biopsy darting without immobilizing or physically handling bears. Our analyses included data from the 2014 – 2016 biopsy mark-recapture study, live-capture data collected under a mark-recapture study 1998 – 2000, and limited harvest recovery data over the entire period 1998 – 2016.
Results of a closed capture-recapture model, implemented in a Bayesian framework and fitted to data for independent animals (i.e., >2 years), suggest a mean abundance estimate of 716 (95% Credible Interval [CRI] = 545 – 955) for the period 2014 – 2016, indicating that the MC polar bear subpopulation increased from the mean abundance in 1998 – 2000 (325 [95% CRI = 220 – 484] in this study; 284 [SE: ± 59.3] in Taylor et al. ). Both the male and female segment of the subpopulation increased between study periods (1998 – 2000 and 2014 – 2016), likely because of a combination of reduced harvest pressure and improved habitat quality. We used a closed population model because data were too sparse for models with more parameters. Estimates of abundance should be interpreted with caution because they reflect the “superpopulation” (e.g., it includes all bears that use the MC management area, some of which spend time in other subpopulations as well) and likely include positive bias due to violation of model assumptions in addition to the negative bias caused by variation in the capture probability. The overall mean litter sizes for the period 2014 – 2016 were 1.70 (SE = 0.09) and 1.61 (SE = 0.11) for cubs-of-the-year and yearlings, respectively. The calculated mean number of yearlings per adult female declined from 0.39 (SE = 0.10) to 0.28 (SE = 0.06) between both study periods, but MC remains a productive polar bear subpopulation despite that decline and observed sea-ice changes. However, given the sparse reproductive data, we are not able to make any substantive inferences. Polar bear body condition (i.e., relative fatness), assessed in the spring, generally increased between the periods 1998 – 2000 and 2014 – 2016. Estimated apparent survival for bears aged 2 and older was 0.88 (SE = 0.02), although this is likely biased downward due to temporary or permanent movement of individual bears with respect to the study area and limited data availability concerning immigration and emigration. This is corroborated by the increase in abundance estimates across periods indicating the survival rate had to be greater than 0.88 to achieve such substantial growth. When we calculated adult survival using the change in abundance estimates between 1998 – 2000 and 2014 – 2016, our estimated rate of 0.93 suggests that the population growth is positive, with a growth rate of 2%. Overall, our findings align with local knowledge that the MC subpopulation recovered from over-harvest that occurred 1979 – 1999 (average harvest 34 bears/yr). Ecologically, we hypothesize that the observed improvements in body condition and strong population growth over time may be related to spatial and temporal reductions in sea-ice type and quantity providing transient benefits to the MC subpopulation due to lighter ice conditions (i.e., a reduction in thick, multiyear ice) and increased biological productivity. However, climate change is the primary long-term threat to polar bears and the threshold beyond which the MC subpopulation could be negatively affected by continued ice loss, like some other polar bear subpopulations, is currently unknown.
Estimating demographic parameters for the MC subpopulation proved to be challenging because small sample sizes, low probability of recapturing the same bear, and lack of movement information constrained analyses in this study such that the estimates of abundance and survival are almost certainly biased. Our estimates represent only the second time the MC subpopulation has been inventoried under a replicable, structured study design and thus offer many opportunities to learn from these experiences in analysis and data collection methodology. For other wildlife populations or ecosystems that share similarities with MC, we recommend collecting additional reproductive data and genetic samples at approximately the midpoint between the current study and the next comprehensive subpopulation assessment (in Nunavut’s case, that would be 5 – 7 years post-field work completion) or increasing study length (e.g. 4 – 5 years), to increase confidence in the survival rates, possible emigration, and reproduction. Further, movement data (satellite telemetry) are recommended. In the absence of satellite telemetry data on polar bear movements, we recommend conducting a meta-analysis to investigate exchange between MC and nearby subpopulations (i.e., Lancaster Sound, and Gulf of Boothia).