Abstract

Human population growth and development patterns have led to the accelerated loss of farmland and forest since the 1950s and has raised considerable concern over the loss of biodiversity and its impact on basic ecosystem functions that support ecological services needed by society for sustainability (EPA, 2004 and Conservation Fund, 2006). Traditional approaches to wildlife conservation have relied on a reactive species-by-species approach that is often prohibitively expensive, biased toward game or charismatic species and often ineffective (Pitelka 1981, Noss 1991). I have developed a rapid habitat assessment tool that uses an integrated series of databases that link forest composition, vegetative development stage, forest and non – forest structure, and non-forest habitat features with maps of known vertebrate distribution. The tool is intended to aid in the evaluation of habitat impacts associated with changes in land use and natural resource management within Maryland.

I tested the habitat assessment tool using presence/absence data for 29 commonly occurring herpetofauna collected by the Maryland Department of Natural Resources over a two year period on seven sites within the Piedmont Plateau physiographic province. The initial run of the assessment tool correctly predicted the presence of 86% of the herpetofauna. The average omission error rate was 14%. The average commission error rate was 43%. I found a significant goodness of fit between the observed and predicted herpetofauna at the 7 inventory sites (χ²_{0.05, 6df} = 0.9, P>.975)