Abstract
One hundred and 12 burrow systems were defined in terms of wood-chuck (Marmota monax rufescens) use rather than burrow structure. The spatial distribution of these burrows was fitted by a negative binomial. Part of the clustering in this distribution resulted from preferences for steep slopes and well-drained soils. It was hypothesized that if any biological need, other than a suitable burrow site, was satisfied by the clustering, then interburrow movements would be less frequent between clusters than within clusters. Over 600 interburrow movements were recorded by tracking in snow, trapping, visual observation, and telemetry. Movements between burrows within a cluster were more frequent than movements between clusters, although no barriers were apparent between clusters except lack of burrows. Burrow clusters and social groups may coincide.