Comparing Aerial Estimates with Ground Counts of Nests in Wood Stork Colonies

Abstract

Estimates of wood stork (Mycteria americana) nests from fixed-wing aircraft in Florida colonies were calibrated using ground counts to evaluate bias and uncertainty. Inferences were based on linear regressions of log-transformed aerial estimates on log-transformed ground counts. Canopy cover at stork colonies had no effect (P = 0.737) on the relationship between aerial estimates and ground counts. Whereas no difference between ground counts and the expected value of aerial estimates was detected for colonies with a low proportion of other white-plumaged nesting birds (P = 0.241), aerial estimates underestimated (P 109 nests) stork colonies containing a high proportion of other white-plumaged nesting birds. Variability of aerial estimates was large. An approximate 95% confidence interval for the 1993 statewide nesting population, based on the statewide aerial survey, was 3,807-12,653 nests. Recommendations for improving future aerial surveys of colonial waterbirds include incorporating ground counts at selected colonies, training observers in presurvey flights, and replicating counts for each colony.