Grizzly bear response to linear features and human recreational activity
A major side-effect of industrial activity is the associated linear-feature footprint and increase in recreational access. Alberta's threatened grizzly bear (Ursus arctos) populations overlap with a multitude of different potential forms of human disturbance, including human recreational activity. Identifying the effects of recreation on grizzly bear behaviour and population recovery is challenging; both in terms of quantifying spatiotemporal variation in multiple types of human recreation at
... man recreation at scales relevant to grizzly bears, and in documenting population-level responses by such a behaviourally variable species as grizzly bears. My first objective was to develop a method to quantify motorised and non-motorised recreational activity across a linear-feature network using trail cameras. Using a generalized linear mixed-effects model to estimate temporal variation in sampling, and Ordinary Kriging to interpolate spatial variation across a linear-feature network, I was able to create spatio-temporally varying maps of recreation that can be incorporated into habitat selection studies. I incorporated grizzly bear radiocollar data within an integrated step selection analysis (iSSA) to predict the importance, significance and directional effect of motorised and non-motorised recreation on grizzly bear habitat selection and movement. I concluded that grizzly bears select for trails when recreation is absent, however they display no response when recreational activity is high. Male grizzly bears also altered their movement behaviour in proximity of trails with high recreational activity; reducing movement speed when activity was absent and increasing speed when recreational activity was high. In general, males showed greater responses to recreation than females, and both male and female bears showed a stronger response to motorised versus non-motorised recreation. Using trail camera data on grizzly bears and black bears (U. americanus), I investigated the influence of recreational activity on bear habitat use iii within a multi-species framework. Grizzly bears and black bears displayed spatial segregation, rarely co-occurring on the landscape. Species' occurrence was not influenced significantly by the presence or absence of recreational activity, however, both species used sites where motorised activity was present less intensively, and this response was strongest in grizzly bears. As with the results from the iSSA, negative responses to motorised recreation were greater relative to non-motorised. Finally, I used the opportunity of having concurrent DNA capture-recapture, radiotelemetry and trail camera data to directly compare two methods of grizzly bear density estimation; 1) spatial capture-recapture and 2) spatial mark-resight. Results for both methods were similar, in terms of both accuracy and precision, highlighting options for density estimation of large mammals without relying on complete individual identification for the entire sampled population. Overall, my results demonstrate the need to incorporate intensity and type of recreational activity within habitat selection studies of grizzly bears and interactions with sympatric species such as black bears. Access management is a challenge that needs to be addressing in large mammal conservation, and will require methods for monitoring human activity and changes in wildlife behaviour and population demography. iv Preface This thesis is an original work by Andrew Ladle. Field methods were approved by the University of Alberta Animal Care and Use Committee (Protocol # AUP00000436) and University of Saskatchewan Animal Care Committee (#20010016).