Effect of Nesting Cover Density on American Crow Predation of Simulated Duck Nests

Abstract

Predation by American crows (Corvus brachyrhynchos) on simulated duck nests was studied in Saskatchewan parklands during 1985. Survival of nests (>80% hidden) after nearby exposed decoy nests had been found was followed in cover of different heights and densities. Short cover was grass-dominant, whereas tallest and densest cover was shrub-dominant, particularly snowberry (Symphoricarpos spp.). Survival increased with cover height and density and reached maximum at about 70 cm height. It is postulated that tall, dense cover represents a behavioral deterrent as well as a physical barrier to crows hunting on foot. Our results indicated that the protection and enhancement of shrub cover would be a beneficial technique for increasing nesting success of certain duck species where crow predation is severe. J. WILDL. MANAGE. 51(2):481-485 The American crow is an important egg predator of nesting parkland ducks (Smith 1971), and the impact becomes more acute as habitat is converted to cropland. Less nesting habitat allows predators to concentrate foraging efforts on smaller patches of cover. Paradoxically, these same changes in land use apparently have caused crow populations to increase rather than decrease (Smith 1971, Houston 1977). To develop management practices to improve waterfowl production, information is needed on features of nesting cover that reduce predation. Previously, we concluded that predation by crows on partially concealed simulated nests was density dependent (Sugden and Beyersbergen 1986). At highest nest densities, crows found hidden nests while they searched cover on foot after a nearby exposed nest had been discovered. We predicted that maximum protection would occur when cover not only concealed nes s but also discouraged crows from searching on foot. In this paper we describe an experiment designed to test that hypothesis. We thank the farmers who allowed us to work on their land and R. G. Clark and G. H. Townsen for reviewing a draft of the manuscript. STUDY AREA AND METHODS The Proctor Lake Study Area, 50 km south of Saskatoon, Saskatchewan, and general techniques have been described (Sugden and Beyersbergen 1986). From 23 April through 3 June 1985, we established 45 5-nest plots within 400 m of an access road through the study area. There were 37 plots on grazed pasture and 8 on a wild hayfield. Each new plot was placed >100 m from any previous plot, active or terminated. Plots consisting of 5 artificial nests were placed in vegetation ranging from stands composed entirely of mixed grasses, through com-