An individual-based movement model for contacts between mule deer (Odocoileus hemionus)

Abstract

Contacts between individuals are key for the spread of infectious disease. Although essential to understanding disease spread, contact rates are difficult to predict, based simply on population demographics in wildlife populations, because contact rates depend upon environmental features as well as the nature of social interactions within and between groups of individuals. We developed a detailed, behaviorally structured, individual-based model (IBM) in Netlogo to simulate contacts between- and within-groups of individual mule deer (Odocoileus hemionus), a species particularly susceptible to chronic wasting disease. The model tracks contacts (defined as two individuals coming within five meters of one another), recorded as between- or within-group depending on the social group membership of the two individuals (dyad). We parameterized the model with data from mule deer with global positioning systems (GPS) collars in east-central Alberta, Canada. Individuals move according to habitat preferences, home range attraction, and grouping behaviours. Animals were tracked at two-hour time steps and were modelled as selecting locations relative to preferred resources based on sex-specific integrated step-selection functions (iSSFs) with steps biased toward a home range centroid. Total within-group contacts increased with group size and were sensitive to changes in movement cohesion of the group and movement persistence, particularly movement cohesion. Total between-group contacts were sensitive only to the number of groups. We compared model predictions for where the locations of deer contacts occurred against an existing statistical model for the relative contact probabilities (RCP) on the same landscape (Dobbin et al. 2023). Predicted locations of deer contacts generally were consistent with higher predicted RCP values. When disease transmission is a function of contact rate, the model can be used to assess the interaction between model components (e.g., movement rates, grouping rules, home ranges, animal densities) and the spatial distribution of key natural and artificial resources that may attract deer and potentially increase disease spread.