• autologistic regression;
  • Bonasa umbellus;
  • Breeding Bird Atlas;
  • forest maturation;
  • New York State;
  • occupancy;
  • ruffed grouse


In New York State, USA, the abundance of ruffed grouse (Bonasa umbellus) has declined >75% since the 1960s. We hypothesized that range contraction of grouse in New York State was associated with broad-scale spatial patterns relating to forest maturation and that these patterns would be evident at the landscape scale. We evaluated data available from New York Breeding Bird Atlases conducted in the early 1980s and again in the early 2000s, which included surveys of >5,000, 25-km2 sample blocks across a landscape of 125,384 km2. We related detection and non-detection of grouse within Breeding Bird Atlas blocks to forest amount, landscape configuration, and presence of neighboring grouse populations via autologistic regression. Loss of early successional forest was inversely related to the continued persistence of grouse (R2 = 0.61). Models showed that ruffed grouse were no longer detected in marginal areas of their distribution, blocks with low amounts of forest, and blocks with landscape configurations potentially advantageous to nest predators. We conclude that declines in both habitat quality and quantity have contributed to range contraction within New York State and that the factors affecting grouse populations have acted at broader spatial scales than are typically considered in habitat evaluation. Our work demonstrates that large landscape-scale data sets are valuable in assessing changes in distribution and abundance of wildlife populations. Landscape analyses at large geographic scales (>100,000 km2) are of particular importance to managers because they represent ways of quantifying the influence of economic incentives and land-use policy on habitats across large regions. © The Wildlife Society, 2012