Selection of landscapes by male ruffed grouse during peak abundance

We studied the habitat relationships of male ruffed grouse (Bonasa umbellus) at the landscape scale when grouse were at the peak of their population cycle. We located 290 and 230 male grouse during 2009 and 2010, respectively, on 30 landscapes encompassing 5,349 ha. We used an information-theoretic model selection approach to examine 2 sets of a priori models. We designed our first model set to evaluate the relationships between Shannon's evenness index of cover types and grouse density and our second set to evaluate other relationships between grouse density and landscape configuration. The top model in the first set was a null model (intercepts only), but a competing model indicated that Shannon's evenness index was positively correlated with male grouse density (r = 0.43). Shannon's evenness index was confounded with cover type dominance in landscapes because it was positively correlated (r = 0.55) with the proportion of the aspen cover type and negatively correlated (r = -0.79) with the proportion of the conifer cover type. The top-ranked model from our second model set included only year and road density. Road density was negatively related to grouse density (r = -0.34), which could mean either hunting pressure along roads affected density or cover types were different in landscapes with greater road densities. The year-only model indicated that male grouse density was less in 2010 than in 2009 (β_year = -0.014, 95% CI = -0.024 to -0.005). We predict that forest management practices that result in an even distribution of cover types within landscapes should result in greater densities of grouse.