Predator Activity Related to Landscape Features in Northern Iowa

Abstract

We studied mammalian predator activity in relation to patches of grassland habitat in the agricultural landscape of northern Iowa to understand the potential interactions with ground-nesting birds, including waterfowl and ring-necked pheasant (Phasianus colrhicus). We quantified presence and movement direction of striped skunk (Mephitis mephitis), raccoon (Procyon lotor), and red fox (Vulpes vulpes) using unbaited track stations placed along the edges of 100-ha blocks of grassland and at other locations throughout the surrounding landscape. We used logistic regression with repeated measures and Akaike weights to develop predictive models of predator presence as a function of landscape variables including distance from a grassland block, shape of grassland edges, and presence of woodlands, farmsteads, and wetlands. Predators were detected at track stations in the landscape near (≤500 m) grassland blocks 33.5% of the time, more frequently than at stations immediately adjacent to edges of blocks (22.6%), and much more frequently than at stations distant (>500 m) from blocks (13.6%). Striped skunk presence at a station decreased as distance from grassland patches increased and was positively related to the number of farmsteads; raccoon presence was positively related to presence of woody cover; and red fox presence increased with greater area of pastureland and greater isolation from farmsteads, and decreased with increasing amount of strip habitat in the landscape. Predicted predator presence at locations where duck nests were found ≤500 m from blocks of grassland (23.1%) was within 1 SE of the Mayfield nest mortality rate, whereas predicted presence at isolated nest locations (12.0%) was greater than observed nest mortality. Track stations indicated that predators moved into and out of grassland patches at corners of blocks (80% of the time) much more frequently than when they traveled along the straight sides of blocks (7%). If presence of predators is directly related to predation rate, our models predict that risk to nesting birds would be greatest in patches near large grassland blocks where corridors, corners, and smaller patches focus predator activity. We envision that wildlife biologists could use models of predator activity to predict the potential influence of landscape configuration on predation risk to nesting birds.