Integrating SAS and GIS software to improve habitat-use estimates from radiotelemetry data

Radiotelemetry has been used commonly to remotely determine habitat use by a variety of wildlife species. However, habitat misclassification can occur because the true location of a radiomarked animal can only be estimated. A nalytical methods that provide improved estimates of habitat use from radio telemetry location data using a subsampling approach have been proposed pre viously. We developed software, based on these methods, to conduct improved habitat-use analyses. A Statistical Analysis System (SAS)-executable file generates a random subsample of points from the error distribution of an es timated animal location and formats the output into ARC/INFO-compatible coo rdinate and attribute files. An associated ARC/INFO Arc Macro Language (AML ) creates a coverage of the random points, determines the habitat type at e ach random point from an existing habitat coverage, sums the number of subs ample points by habitat type for each location, and outputs tile results in ASCII format. The proportion and precision of habitat types used is calcul ated from the subsample of points generated for each radiotelemetry locatio n. We illustrate the method and software by analysis of radiotelemetry data for a female wild turkey (Meleagris gallopavo).