Beaver in the Great Plains: habitat suitability index and flooding.

Abstract

In 1991, Kansas Department of Wildlife and Parks (KDWP) and Kansas State University (KSU) censused beaver (Castor canadensis) colonies and collected habitat information on riverine systems throughout Kansas. The resulting publication showed that the United States Fish and Wildlife Service's habitat suitability index model (HSI) does not work well in Kansas. In 1995 and 1996, portions of 9 of the rivers censused in 1991 by KDWP and KSU were recensused for beaver colonies. Several habitat characteristics were recorded in an attempt to fine tune the HSI for use in the Great Plains. River depth, river width, and river bed substrate were not shown to have statistically significant effects on beaver locations. Therefore, these variables were not considered to be valuable enough to add to the HSI for use in the Great Plains. Water fluctuation, which is used in the current HSI, was not shown to have a strong relationship with beaver colony density. Therefore, in the Great Plains, water fluctuation should be dropped from the current HSI. Two variables, slope of the river bank and river bank substrate, were shown to have statistically significant effects on beaver locations (P < 0.0001, P = 0.0001, respectively). Therefore, when calculating a beaver HSI in the Great Plains, these variables should be added. Using Geographic Information Systems (GIS), I calculated land use and related it to beaver use and non-use of sites. Beaver showed no preference for locating colonies adjacent to any particular land use. However, a comparison of combined agricultural and woodland to grassland (P =0.071) suggests that different methods for evaluating land use with GIS should be investigated. Large quantities of field time could be saved jf food requisites are quantified with GIS instead of in the field. In 1993 and 1995, major flooding occurred in the Midwest and eastern Great Plains. The effect of these floods on riverine beaver colonies was investigated by using the beaver census data collected by KDWP and Kansas State University in 1991 as a baseline. The censuses conducted in 1995 and 1996 showed no relationship between beaver colony densities and number of flood days on each respective river (P = 0.1446). Newborn beaver kits are the members of a beaver colony that are most susceptible to floods. Since young beaver do not disperse until 2 to 3 years of age, noticeable decreases in colony density may not have been detected at the time of the study.