Northern Prairie Wildlife Research Center
We are reasonably comfortable with the data on clutch destruction; methods for gathering and analyzing such data are well developed (Klett et al. 1986). The predator indices were developed for this study (A. B. Sargeant, R. J. Greenwood, M. A. Sovada, and T. L. Shaffer, unpublished data) and had not been tested extensively until now. Nonetheless, we believe they provided a realistic representation of predator activity on each quarter.
Interpretation of the findings
The red fox emerged as the principal predator on both early and late nests. This result was not surprising for the red fox is recognized as an important predator on adult ducks in the prairie region (Sargeant 1972; Sargeant et al. 1984); remains of numerous ducks were found at fox dens on the study areas (A. B. Sargeant, unpublished data). Further, experimental evidence (A. B. Sargeant, unpublished data) indicates that red foxes will take large numbers of duck eggs and cache them for future use. Several studies of nesting ducks have implicated the red fox as a major predator of eggs (e.g., Martz 1967; Duebbert and Lokemoen 1976; Higgins 1977), although the identity of responsible predators was determined subjectively.
The American crow was significantly implicated in destructions of early, but not late, nests. This species is also widely recognized as an important predator of duck eggs (Kalmbach 1937; Hammond 1940; Smith 1971; Stoudt 1971). Crow predation on duck eggs has been reported to decline after crow clutches hatch and as the concealment quality of nesting cover increases (Hammond 1940; Hilden 1964; Smith 1971; Stoudt 1971; Sugden and Beyersbergen 1987).
The badger index was significantly correlated with predation rates of early nests. The badger has not been implicated as a major predator of duck eggs (e.g., Keith 1961; Smith 1971; Stoudt 1971), but little is known about how the species interacts with waterfowl. Predation by badgers may have been misinterpreted in previous nesting studies. We have no explanation for the difference between early and late nests.
The striped skunk index correlated strongly with nest predation rates in both early and late samples, but was retained in the regression only for late nests. We suspect that skunks are important predators throughout the season, but that the correlation between skunk and fox indices clouded the picture. Numerous investigators (e.g., Kalmbach 1938; Ellig 1955; Keith 1961; Stoudt 1971; Sargeant and Arnold 1984) have indicated that the striped skunk is a major predator of duck eggs in the Prairie Pothole Region. Greenwood (1986) found that removal of skunks from duck nesting areas in North Dakota increased nest success rates from an average of 5 to 15%, although results were not consistent among areas. Kalmbach (1937) and Ellig (1955) also reported increased duck nesting success following removal of striped skunks.
Other predators did not figure significantly in the regressions, but may have been important, especially in local situations. Densities of Franklin's ground squirrels did not relate significantly to the daily predation rate of nests; even univariate correlations were negligible. This rodent was fairly abundant in only four study area - years, however. Also, because it lives almost exclusively in dense grass, forb, and shrub vegetation (Choromanski-Norris 1983), only a portion of duck nests in quarters with Franklin's ground squirrels were at risk to predation by the species. The species has been recognized as a predator on duck eggs (Sowls 1948; Sargeant et al. 1987b), but it may take only some eggs out of a nest, resulting in clutch reduction rather than clutch destruction.
Black-billed magpies were common on some study areas, but abundant in none. Smith (1971) and Stoudt (1971) thought that magpies, although common on their Alberta and Saskatchewan study areas, were of minor significance compared with crows as predators of duck eggs. Our data also suggest that magpies were of minor consequence to the survival of duck clutches.
Nest predation rates tended to be lower in years and study areas with higher proportions of seasonal or semipermanent wetlands containing water in May. Several reasons for this relationship are plausible, including greater dispersal of nests when ponds are plentiful and better security afforded nests by lusher vegetation occurring when precipitation is greater, but we lean toward the role of buffer prey. A high proportion of basins holding water implies that precipitation previous to the time of measurement was adequate to insure vigorous plant growth (Boyd 1981) and to attract and stimulate nesting by a variety of other birds. High plant productivity probably increases the abundance of alternative prey for the predators we considered. Buffer prey can influence predation rates on birds and their clutches (e.g., Larson 1960; McInvaille and Keith 1974; Pehrsson 1985; Summers 1986; Beintema and Muskens 1987).