The use of forest maps for the description of wildlife habitats: limits and recommendations

We evaluated the reliability of forest maps for describing wildlife habitat s. During the summer of 1997, we sampled 186 boreal forest stands located i n Jacques-Cartier Park, Quebec. In each stand, we measured slope, crown clo sure, basal area, as well as tree height and age. We determined if map clas sifications, with regard to dominant species composition, density, tree hei ght, tree age, and slope, correlated with field observations. We also measu red lateral cover and deciduous browse availability, variables that are con sidered useful for the characterization of wildlife habitats, to examine ho w these habitat features were related to map classification. Age (57% of th e sites correctly classified) and density (34%) were the variables for whic h map classification had the best and worst correspondence with field measu rements, respectively. Dominant species on maps were correctly identified i n < 74, < 55, and < 40% of the sites in coniferous, mixed, and deciduous st ands, respectively. The use of a simple classification method based on cove r type alone resulted in improved correlations, since 94, 60, and 29% of th e coniferous, mixed, and deciduous stands, respectively, were properly iden tified on maps. We related lateral cover and food availability to stand cat egories using the most reliable map variables. We conclude that forest maps are useful for describing major habitats at the stand level. When forest r esource maps are to be used for studying habitat suitability, we recommend sampling a subset of stands to assess if important wildlife habitat feature s, which reflect species requirements, can be related to habitat characteri stics as determined by the maps.