Variability in Nest Survival Rates and Implications to Nesting Studies

Abstract

Four resonably large samples (83-213) of mallard (Anas platyrhynchos) and blue-winged teal (A. discors) nests on an interstate highway right-of-way in south-central North Dakota [USA] were used to evaluate potential biases in hatch-rate estimates. Twelve weekly searches for nests were conducted with a cable-chain drag in 1976 and 1977. Nests were revisisted at weekly intervals. Methods used to estimate hatch rates for the 4 data sets were: the traditional method, the Mayfield method and 2 modifications of the mayfield method that are sometimes appropriate when daily mortality rates of nests are not constant. Hatch rates and the average age of nests at discovery declined as the interval between searches decreased, suggesting that mortality rates were not constant in the samples. An analysis of variance indicated that daily mortality rates varied with the age of nests in all 4 samples. Mortality was generally highest during the early laying period, moderately high during the late laying period and lowest during incubation. This relationship of mortality to nest age might be due to the presence of hens at nests or to differences in the vulnerability of nest sites to predation. A modification of the Mayfield method that accounts for age-related variation in nest mortality was most appropriate for the samples. Methods for conducting nesting studies and estimating nest success for species possessing similar nesting habits are suggested.