Validating a Florida Scrub Jay habitat suitability model, using demography data on Kennedy Space Center

Abstract

Abstract A habitat suitability index (ust) model for the Florida Scrub/ay (Aphelocoma coerulescens coerulescens] wos tested usinga geographic information system for the Tel-4 study site onKennedy Space Center, Florida. The model used suitabilitygraphs that quantify habitat preference with respect to agiven variable to produce spatial estimates of Florida Scrublay habitat suitability. Habitat suitability of each habitatpatch was dependent on its characteristics ond the charac-teristics of its sunoundings. A covetage containing threeyears of demographic data was overlaid on the HSI coverage.Areal correspondence measures and statistical testing werethen performed. Conelation coefficients between modeleddata and demographic data rcnged between 0.60 and 0.87.Spatial residual analysis also showed agteement between themodel and demography data. AII measures of model per-formance suggested that the model accurately predicted hab-itat suitability for the Tel-4 study site. lntroduction The main purpose of geographic information systems (crs)is to process spatial information (Berry, t093). Ecologistshave been modeling processes that involve spatial informa-tion since the early development of their discipline (Hold-ridge, 1,947; Whittaker, 1956; Curtis, 1g5S). However, use ofGIS technology by ecologists has been limited and repre-sents a relatively untapped potential for ecological model-ing of spatial processes (Hunsaker ef d1., 1993). With theaid of today's remote sensing and cIS technology, opportu-nities exist to further develop and test environmental mod-eling techniques that can readily be applied toenvironmental management.Many threatened and endangered wildlife species andtheir habitats are being adversely affected by human distur-bance. It is increasingly important to understand the habitatrequirements, delineate the remaining suitable habitat, andeffectively manage those units for the survival of these spe-cies. Habitat-based modeling techniques can identify remain-ing potential habitat and predict spatial habitat suitability.The modeling techniques should incorporate existing knowl-edge of species-habitat relationships, be reproducible, andprovide a quantitative measure of habitat suitability.Habitat suitability index (Hst) modeling is a common ap-proach to modeling wildlife-habitat relationships (Morrisonet al., 'l,sgz). The methodology was developed to supporthabitat evaluation procedures (ure) used by the United