ºìÐÓÊÓƵ

Topic

Applying predation risk theory to wildlife behaviour during non-consumptive ecotourism in Tweedsmuir Provincial Park, Nuxalk Territory: insights into risk, foraging, and conflict dynamics

Department of Geography

Date & location

• Tuesday, January 14, 2025

• 10:30 A.M.

• Clearihue Building

• Room B007 and Virtual

Reviewers

Supervisory Committee

• Dr. Chris Darimont, Department of Geography, University of Victoria (Supervisor)

• Dr. Paul C. Paquet, Department of Geography, Uvic (Member)

• Dr. Douglas Clark, School of Environment and Sustainability, University of Saskatchewan (Outside Member)

• Dr. Donald Kramer, Department of Biology, UVic (Outside Member) 

External Examiner

• Dr. Daniel Blumstein, Department of Ecology and Evolutionary Biology, University of California

Chair of Oral Examination

• Dr. Stan Dosso, School of Earth and Ocean Sciences, UVic

   

Abstract

Managing to minimize human disturbance of wildlife populations and the individuals that comprise them can draw on theory and data related to non-consumptive effects of risk on wildlife. Predation risk theory is a behavioural-ecological framework that describes how the risk of predation influences the behavior, physiology, and spatial distribution of prey. Interconnected with predation risk theory is a linkage to foraging behaviour. The asset protection principle (APP) predicts that variable food supply and its associated risks will affect antipredator behavior; responses to predation risk should dominate when food reserves are high, whereas nutritional risk becomes more important when food reserves are limited. Additionally, the human shield hypothesis (HSH) describes how some individuals might seek human presence if it repels other and presumably more dangerous potential sources of risk. Owing to these contexts and more, tolerance of human-caused disturbance stimuli can vary among animals, often mediated by how individuals and demographic sub-groups perceive risk. My research applied these frameworks to community-driven research questions related to the potential effects of ecotourism on several dimensions of grizzly bear (Ursus arctos horribilis) behaviour in a protected area in Nuxalk Territory, British Columbia, Canada. 

In my first data chapter (Chapter 2), we tested components of the APP and HSH where ecotourism co-occurs with grizzly bear foraging during hyperphagia. We used generalized linear mixed effects and multinomial regression models to understand how weekly detection rates (measured by camera traps) varied with food abundance (salmon and berries), ecotourist visitors, time of day, and age-sex class. When salmon abundance was high, bear activity (weekly detections) decreased by 13% with every 100 visitors/week. Under low salmon conditions, we observed the opposite pattern; bear activity increased with visitor numbers, creating ‘high bear high visitor’ conditions. Consistent with HSH, detection data revealed an increased likelihood of detecting subordinate age-sex classes compared with adult males when visitor numbers were high. Our findings suggest that when salmon are low, managers might consider limiting visitors to mitigate disturbance. More broadly, understanding how wildlife allocate anti-predator behavior as a function of risk and food can inform conservation science and practice.

In my second data chapter (Chapter 3), we approached the potential influence of ecotourism using different approaches, tools, and spatial scales. Specifically, we assessed whether and how a grizzly bear population might show among- and across-individual variation in tolerance to ecotourism activity using both spatial and behavioural lenses. Despite the high mobility of grizzlies and access to regionally-abundant salmon resources at ecotour sites, our genetic tagging approach revealed that only a small subset of individuals detected at least twice (n = 11 of 80) showed spatial detection histories that favoured areas of pronounced human activity. Most of these individuals (n = 9 of 11) were female, providing additional support for the human shield hypothesis. Our companion behavioural observations at ecotourism sites considered how ecotourism intensity (number of visitors, number of boats) and environmental variables (presence of other bears, salmon biomass) might affect fishing and alertness behaviour. Our findings revealed no evidence for an effect of measured ecotour metrics on grizzly behaviour. Instead, we found a negative association between salmon biomass and alertness, a pattern opposite to predictions based on predation risk theory. We speculate that reduced alertness at higher salmon densities related to (mostly female) bears, shielded from dangerous male bears, focusing their attention on hyperphagic behaviour during high salmon periods to maximize fitness gains related to reproductive potential. Collectively, our multi-lens approach illustrates how comprehensive insight into wildlife tolerance of human activity can emerge.

Finally, in my last data chapter (Chapter 4) we asked whether ecotour bears, which show tolerance to human activity, can lead to tolerating other presumably risky scenarios that lead to human-wildlife conflict in communities downstream of ecotourism sites. We screened for genetic matches between individuals that encountered conflict (n = 30) and 118 individuals detected upstream via hair snags (including 34 at ecotour sites). Of these 34, one encountered conflict. In analysis scenario 1, we considered all detected and undetected bears in the region as freely mixing, and used Bayes’ theorem to account for imperfect detection of ecotour bears among conflict samples, deriving an estimate of 1.47 (rounded to 2). Accounting for this uncertainty, we used a probability approach to ask how large the unknown non-ecotour bear population would have to be to observe this frequency of conflict among ecotour bears (2/34) by chance. The resulting population level exceeded available estimates, suggesting ecotour bears are less likely to encounter conflict. In scenario 2, we assumed that downstream bears are not necessarily from the same population as those sampled upstream, and compared the proportions of known ecotour and non-ecotour bears among conflict samples. We found no evidence of a significant difference. Collectively, these analyses suggest that tolerance to ecotourism did not predispose ecotourism bears to engage in risky scenarios, suggesting other human-caused drivers of conflict. In general, my dissertation research contributes to our understanding of the contexts in which non-lethal human disturbance stimuli can affect wildlife behaviour. Specifically, I highlight that neglecting context-dependent decision-making, such as resource availability, limits the practical utility of measuring non-consumptive risk effects. I also highlight how this work can support local, system-specific management decisions related to ecotour management.