The distribution of Microtus pennsylvanicus within grassland habitat

Abstract

A 4-year live-trapping study in southern Quebec demonstrated that Microtus pennsylvanicus was distributed randomly in three half acres of enclosed grassland habitat. It was not possible to establish, by correlation analysis, a strong association between number of captures at each position and environmental (mainly plant species) variables. The random distributions are attributed more to the operation of population factors than to the operation of habitat factors in these areas. A simple model is presented which shows how, at different densities, animal distribution varies in relation to the distribution of habitat structure. Specifically, it predicts that (a) as animal density increases the distribution tends toward uniformity, and (b) at a given density, animal distribution in a patchy habitat is more aggregated than in a uniform habitat.The model is tested with data from the Quebec enclosures and from an enclosure in Saskatchewan which contained 1.4 ac of grassland. First it is established that the grassland is spatially more varied (patchy), in terms of plant species composition and presumed structure, in the Saskatchewan enclosure than in any of the three Quebec enclosures. In conformance with predictions of the model, Microtus distribution was aggregated in the Saskatchewan grassland. Also in conformance with the model the distributions became less aggregated as density increased, in both Saskatchewan and Quebec enclosures, but at a given density the animals were more aggregated in the patchy Saskatchewan grassland than in the uniform Quebec grassland.