PARENT SESSION
Poster Session #12: Modeling.
Tuesday, August 7, 2001. Presentation from 10:30 AM to 12:00 PM. Exhibition Hall

Beyond composition: Incorporating landscape pattern into gap analysis distribution models.

Knetter, Sonya¹, Mathews, Nancy¹, ¹

ABSTRACT- The Gap Analysis Program emerged in the late 1980's as a national effort to describe patterns of biodiversity, and at the same time provide a rapid means of assessing the conservation status of natural communities and species across broad geographic areas. Vertebrate distribution models are a fundamental component of the analysis that predict the presence of species based on the distribution of suitable habitat. The traditional Gap Analysis model defines habitat suitability in terms of localized environmental variables, but does not account for the effects of landscape pattern on species' distributions. We describe a modeling technique to predict the distribution of the Ovenbird (Seiurus aurocapillus), a common, area-sensitive, forest-interior specialist in Wisconsin. Instead of the traditional presence/absence model, we built a logistic regression model relating the probability of occurrence within suitable habitat to the amount of interior habitat (core area). We parameterized our distribution model based on data that were summarized in the scientific literature and used a sensitivity analysis to determine which parameters were most sensitive. We compared our regression model predictions to the traditional Gap Analysis model. We propose that Gap Analysis predictions are likely to result in high errors of commission for species that are sensitive to landscape-scale variables, such as habitat area and configuration. We feel our distribution model provides a defensible framework for basing empirically-derived probability statements about species' occurrence.

KEY WORDS: Gap Analysis, distribution models, Ovenbird (Seiurus aurocapillus)