Integrating telemetry data into spatial capture–recapture modifies inferences on multi‐scale resource selection

Abstract Estimating population size and resource selection functions ( RSF s) are common approaches in applied ecology for addressing wildlife conservation and management objectives. Traditionally such approaches have been undertaken separately with different sources of data. Spatial capture–recapture ( SCR ) provides a hierarchical framework for jointly estimating density and multi‐scale resource selection, and data integration techniques provide opportunities for improving inferences from SCR models. Despite the added benefits, there have been few applications of SCR ‐ RSF integration, potentially due to complexities of specifying and fitting such models. Here, we extend a previous integrated SCR ‐ RSF model to accommodate overlap of individuals between data sources and add the functionality to an open‐source R package ( oSCR ). We illustrate an application of the integrated model to a population of American marten ( Martes americana ) in alpine forests of northern New England. Spatial encounter data from camera traps were combined with telemetry locations from radio‐collared individuals to examine how density and space use varied with spatial environmental features. We compared multi‐model inferences between the integrated SCR ‐ RSF model with telemetry and a standard SCR model with no telemetry. The integrated SCR ‐ RSF model supported more complex relationships with spatial variation in third‐order resource selection (i.e., individual space use), including selection for areas with shorter distances to mixed conifer forest and rugged terrain. Both models indicated increased second‐order selection (i.e., density) for areas close to mixed conifer forest, while the integrated SCR ‐ RSF model had a lower effect size due to modulation from spatial variability in space use. Our application of the integrated SCR ‐ RSF model illustrates the improved inferences from spatial encounter data that can be achieved from integrating auxiliary telemetry data. Integrated modeling allows ecologists to join empirical data to ecological theory using a robust quantitative framework to better address conservation and management objectives.