Evaluating the Waterfowl Breeding Population and Habitat Survey for Scaup

Authors

Michael L. Schummer, Long Point Waterfowl and Wetlands Research Program/Bird Studies Canada & SUNY College of Environmental Science and ForestryFollow
Alan D. Afton, Louisiana State University, Baton Rouge
Shannon S. Badzinski, Environment and Climate Change Canada, Canadian Wildlife Service
Scott A. Petrie, Delta Waterfowl Foundation, Bismarck
Glenn H. Olsen, U.S Geological Survey, Patuxent Wildlife Research Center
Mark A. Mitchell, University of Illinois, Urbana

Document Type

Article

Date of this Version

2018

Citation

The Journal of Wildlife Management 82(6):1252–1262; 2018; DOI: 10.1002/jwmg.21478

Comments

U.S. government works are not subject to copyright.

Abstract

Potential bias in breeding population estimates of certain duck species from the Waterfowl Breeding Population and Habitat Survey (WBPHS) has been a concern for decades. The WBPHS does not differentiate between lesser (Aythya affinis) and greater (A. marila) scaup, but lesser scaup comprise 89% of the combined scaup population and their population estimates are suspected to be biased. We marked female lesser scaup (i.e., marked scaup) in the Mississippi and Atlantic Flyways, Canada and United States, with implantable satellite transmitters to track their spring migration through the traditional and eastern survey areas of the WBPHS, 2005–2010. Our goal was to use data independent of the WBPHS to evaluate whether breeding population estimates for scaup were biased and identify variables that might be used in the future to refine population estimates. We found that the WBPHS estimates of breeding scaup are biased because, across years, only 30% of our marked scaup had settled for the breeding period when the strata in which they settled were surveyed, 43% were available to be counted in multiple survey strata as their migration continued during the WBPHS, 32% settled outside the WBPHS area, the number of times a marked scaup was available to be counted by survey crews varied positively with the latitude that a marked scaup settled on breeding areas, the probability of a marked scaup being in a stratum while it was surveyed varied among years, and these probabilities were positively correlated with the traditional and eastern breeding population estimates for scaup. Annual population estimates derived from banding data provide a less biased and preferable method of monitoring scaup population status and trend. Development of models that include metrics such as survey stratum latitude and annual spring environmental conditions might potentially be used to improve scaup breeding population estimates derived from the WBPHS, but independent estimates from banding data would be important to evaluate such models.