Factors Influencing Use of Fecal Line Surveys For Estimating Trends in Predator Populations in Western South Dakota

Development of a technique to estimate relative densities of coyote (Canis latrans), red fox (Vulpes vulpes), and striped skunk (Mephitis mephitis) populations is necessary to monitor population changes in western South Dakota. Fecal line surveys were used to estimate relative densities because they are easy to use, inexpensive, do not elicit responses of predators, and provide the most potential for estimating trends. The objectives for this study were to determine the minimum number of transects needed and probability of finding a coyote , red fox , and striped skunk fecal pile on fecal line transects placed in various land uses and road types in western South Dakota , to determine if temporal variation of these populations occurred within study areas , to examine relative density indices of these populations within and across study areas , and to determine the relationship between relative density indices, food habits, and home ranges of coyotes in western South Dakota. Minimum number of transects needed and the best placement of transects for detecting trends was ascertained by establishing a total of 57 transects in various land uses (e.g., cropland, forested, grassland) and road types (e.g., tractor paths, dirt roads with limited vehicle travel) with in each study area. Fecal line surveys were conducted continuously every 2-3 weeks in the northwest (Harding County), westcentral (Haakon and N. Jackson counties), and the B lack Hills (Lawrence, Pennington, and Custer counties), South Dakota study areas from February to October 1 998 and from February to mid-May and September to October 1 999. Food habits of coyotes were determined from fecal analysis of samples collected from study areas. Four coyotes (3 F, 1 M) were radio-collared in Custer State Park (CSP) and Wind Cave National Park (WCNP) to determine home ranges and habitat use. Minimum number of transects necessary to stabilize mean relative density indices was 11 for coyote, red fox , and striped skunk in all 3 study areas. Probability of finding a fecal pile in various land uses and road types differed with each predator species and study area. For instance, land use was an important contributing factor in the probability of finding a coyote fecal pile in the northwest area (X2=16.967; df = 1; P\u3c0.001) whereas road type was more important in the westcentral area (X2=13.794; df=2; P\u3c0.001). Land use was the only important factor in finding a red fox fecal pile in the northwest area (X2=5.552; df=1; P\u3c0.001). Landu use and road type were not important factors in the probability of finding a striped skunk fecal pile (P\u3e0.05). The overall mean relative density index for coyotes was significantly lower (F=19.19; df = 2,42; P\u3c0.001) in the northwest area than in the Black Hills and westcentral areas. Mean relative density indices were significantly higher (F=7.00; df = 1,28; P=0.013 and F=9.41; df=1,28; P=0.005, respectively) in the northwest area than the westcentral are for both red fox and striped skunk in 1998. Temporal variation (P\u3c0.05) in coyote, red fox, and striped skunk populations was observed within study areas. Fecal pile detection/sightability differed with each observer (F=4.114; df-2,34; P=0.025). Relative density indices were corrected using the mean sightability, which was estimated at 0.775. Using Program Scat 1.5, correction factors were used to associate relative frequencies of food items in coyote fecal samples to the actual amount of prey consumed. Food habits were quantified by percent-of-scat (POS) as well as by percent-fresh-weight-of-prey (%FWP). Mammals were the most common food item in coyote diets in all study areas. Deer/pronghorn (64.4 %FWP) comprised the greatest part of coyote diets in the Black Hills/Wind Cave National park region, whereas small-sized mammals (34.6 %FWP) were the greatest item in the prairie region (northwest and westcentral areas). Using the 95% minimum convex polygon home range estimator, mean home range of radio-collared coyotes in CSP/WCNP was 38.12 km2. It is recommended that a fecal line survey program be implemented in western South Dakota for monitoring trends in predator populations. A minimum of 11 transects should be distributed across land uses and road types in each areas. Furthermore, fecal line surveys should be conducted 2 times per year, once during March and again in September to monitor predator populations in western South Dakota