Determining Kill Rates of Ungulate Calves by Brown Bears Using Neck-Mounted Cameras

Predation of moose (Alces alces) and caribou (Rangifer tarandas) calves by brown bears (Ursus arctos) has been extensively studied, because predation has population implications for both predator and prey species. Existing methods have provided estimates of population-level predation rates, but they have not been able to estimate kill rates by individual bears until recently and the accuracy of these estimates was unknown. Our objectives were to evaluate whether collars equipped with video cameras would be able to record predation events by brown bears and provide useful estimates of ungulate kill rates. We fitted 17 brown bears in the Nelchina Basin of Alaska, USA, with Global Positioning System (GPS) collars equipped with cameras in the spring of 2011, 2012, and 2013. We retrieved the collars in late June of each year to download the video data and associated GPS locations. To compensate for incomplete sampling, we constructed a calf risk model from previous calf mortality studies and used it to predict total calf kills for each bear through the end of June. The camera collars documented kill rates considerably greater than previous estimates. Median handling times by bears were 40 min for caribou calves and 60 min for moose calves. These short handling times should be considered by future researchers when trying to evaluate kill detectability and when factoring calf mortality into population models. We demonstrate that this technology can be successfully applied in the field and provide recommendations on the video sampling intervals necessary to detect predation rates. (C) 2017 The Wildlife Society.