PREDICTING RARE ORCHID (SMALL WHORLED POGONIA) HABITAT USING GIS

Isotria medeoloides (Pursh) Raf., commonly known as the small whorled pogonia, is the rarest orchid in eastern North America, north of Flori da. Although it has a wide ranging distribution, more than 80 percent of the known plants are found in New Hampshire and Maine. A geographic information system (GIS) was used to facilitate locating potential ha bitat for Isotria medeoloides in the two states. Field characteristics , including physiography, soils, and forest and herbaceous cover, were evaluated at 27 extant locations of the orchid. Site locations were d igitized, and general habitat characteristics at each site were identi fied using the cls. Digital habitat information included topographic d ata derived from the USGS Digital Elevation Model, soil types from the USDA Natural Resource Conservation Service, and spectral data from La ndsat Thematic Mapper. Intervals within each habitat parameter were de fined and weighted according to their frequency of occurrence at extan t sites. Two predictive GIS overlay models were developed: an equal we ight model and a chi-square model. In the first scenario, each habitat characteristic was weighted equally. in the second, the importance of each habitat parameter was evaluated with a chi-square test of habita t features at sites with and without orchids. As a result of the chi-s quare evaluation, the following general characteristics were determine d to be positively associated with small whorled pogonia sites: soils with a pan layer percent slopes between 11 and 17 percent, and a digit al reflectance greater than 68 for the near infra-red band of Landsat Thematic Mapper data. To determine the effectiveness of each model in predicting potential habitat, field surveys of eight USGS topographic quadrangles in New Hampshire and the town of Acton, Maine were conduct ed. Nine previously undiscovered populations of Isotria medeoloides we re found. The accuracy of the models was assessed at 23 Isotria medeol oides populations, 13 of which were excluded from model development. T he equal weight model correctly predicted 57 percent of the sites as p otential habitat, and the chi-square model correctly predicted 78 perc ent of the sites. The GIS provided a useful tool for combining ecologi cal habitat characteristics to successfully predict potential habitat for Isotria medeoloides.