Distribution and abundance of Minnesota-breeding Ring-necked Ducks Aythya collaris

Concerns about Ring-necked Duck Aythya collaris breeding numbers in Minnesota suggested a population survey was needed; however, knowledge of the species' distribution and abundance was poor. Existing waterfowl survey data were used to define an aerial sampling frame, and a habitat model incorporating land-cover imagery was used to develop a survey to describe breeding pair distribution and abundance in 2004-2007. In 2004, a stratified random sample was used to estimate pair numbers for plots having moderate to high potential for breeding ducks, and a sensitivity analysis was used to estimate pair numbers for the remaining un-surveyed plots. Pairs occurred throughout the survey area, primarily on small (median = 8.3 ha) wedands with open, bog-like margins, with greatest concentrations in the west. The 2004 estimate was of ∼20,000 breeding pairs, with sensitivity analysis suggesting that half of these occupied plots assumed to have few if any Ring-necked Ducks. Incomplete habitat model definitions and shortcomings in land-cover imagery contributed to stratification failure in 2004. Habitat definitions were refined and survey plots were re-stratified in 2005-2007. A second survey replaced the sensitivity analysis to better estimate pairs in plots assumed to have few ducks. Breeding pair estimates from both surveys combined were of 11,000-15,000 pairs in 2005-2007. Survey development is used to illustrate how an adaptive approach can better ensure survey success. By focusing first on development, shortcomings in the untested habitat model, land-cover imagery, and initial stratification were revealed and changes leading to more precise population estimates were made. Issues encountered in survey development likely typify those experienced when initiating large-scale monitoring efforts of little studied species.