Predictive modeling of Alaskan brown bears (Ursus arctos): assessing future climate impacts with open access online software

Abstract

As vital representative indicators of the state of the ecosystem, Alaskan brown bear (Ursus arctos) populations have been studied extensively. However, an updated statewide density estimate is still absent, as are models predicting future occurrence and abundance. This kind of information is crucial to ensure population viability by adapting conservation planning to future needs. In this study, a predictive model for brown bear densities in Alaska was developed based on brown bear estimates derived on the best publicly available data (Miller et al. 1997). Salford’s TreeNet data mining software was applied to determine the impact of different environmental variables on bear density and for the first state-wide GIS prediction map for Alaska. The results emphasize the importance of ecoregions, climatic factors in December, human influence and food availability such as salmon. In order to assess the influence of changing climate conditions on brown bear populations, two different IPCC scenarios (A1B and A2) were applied to establish different predictive climate models. The results of these projections indicate a large expansion of brown bear densities within the next 100 years. High density habitat would thus expand from southern coastal areas towards central Alaska. Based on the modeling results, optimum potential protected areas were determined by means of the program Marxan. According to the outcome, the protection of brown bear populations and bear habitat should accordingly focus on areas along the southern coast of Alaska. The study provides a first digital GIS modeling infrastructure for bear densities in Alaska. Through the pro-active temporal and spatial identification of important brown bear habitats and connectivity zones ahead of time, measures ranging from conservation to the planning of transport facilities could be more effectively focused on minimizing and mitigating impacts to these critical areas before real-world problems occur, as well as in an Adaptive Sustainability Management framework.