An improved method for determining the distribution of swift fox in Kansas

During 1997 and 1998 we tested a new method for determining the distribution of swift foxes (Vulpes velox) in Kansas. From a sampling frame of 30 counties in western Kansas, we selected a systematic sample of alternate townships in a checkerboard pattern. During September and October 1997 and August 1998, experienced observers delineated suitable swift fox habitat within each sample township and searched it for evidence of occupancy (tracks, dens, and the animals themselves) by swift fox and other furbearers. Each township was searched for a minimum of 30 minutes, with searches continuing until swift foxes were either detected or for 120 minutes. Of the 288 townships selected in 1997, 271 (94.1%) were searched effectively with swift fox detected in 40.5% of the townships. Adverse weather conditions prevented surveys in two northwestern counties of our sample frame. In 1998, 245 township were searched effectively. Swift fox were detected in 27 counties searched to date. We did not detect swift fox in Seward, Meade and Ford counties, where the species is thought to be uncommon or absent. Tracks were difficult to discern in areas with hard or sandy soils and were sometimes obliterated by adverse weather, vehicle traffic, and agricultural activities. To determine how frequently we failed to detect swift foxes that were present, we plan to repeat searches in 1999 in townships where swift foxes were not detected previously. Nevertheless, preliminary results suggest our method to be a practical means for conducting landscape-scale presence/absence surveys of swift fox. Restricting searches to habitat judged best for swift foxes and most favorable for track detection helped control costs and achieve high detection rates.