Autumn Roosting Flight Counts as an Index to Wood Duck Abundance

Abstract

Roosts and roosting flights of wood ducks (Aix sponsa) were studied during autumn, 1959-64. Observations were made during 768 flight periods at 52 roosts. The primary study area was a 100-mile segment of Mississippi River bottoms in northeastern Iowa. Roosts were in many species of emergent vegetation. Barring major changes in habitat, autumn roost sites were usually the same each night and each year. Autumn migrants used roosts established in late summer by local birds. Seasonal changes in roosting populations were similar each year. September 20 is recommended for making roosting flight counts to be used as an index to annual changes in abundance of wood ducks. Evening counts were superior to morning counts. Numbers at a roost ranged up to 5,400. Sampling units, sampling schemes, sensitivity, and use of roosting flight counts for estimating abundance of wood ducks are discussed. The Mississippi Flyway Council stated that adequate information on population levels of wood ducks was not provided by techniques used to census other waterfowl (Yancey et al. 1958). In response, the Iowa Cooperative Wildlife Research Unit initiated a study of some methods of censusing wood ducks. This is a report of an investigation of wood duck roosting phenomena in northeastern Iowa and adjacent areas during August-November, 1959-64. The main objective was to evaluate roosting flight counts as a method of estimating abundance of wood ducks. We are grateful to Refuge Manager Don Gray and his staff on the Upper Mississippi River Wildlife and Fish Refuge for their assistance and support of this study.