Theses and Dissertations

Issuing Body

Mississippi State University


Kaminski, Richard M.

Committee Member

Dinsmore, Stephen J.

Committee Member

Reinecke, Kenneth J.

Committee Member

Burger, Loren W.

Committee Member

Barras, Scott C.

Date of Degree


Document Type

Dissertation - Open Access


Forest Resources

Degree Name

Doctor of Philosophy


College of Forest Resources


Department of Wildlife and Fisheries


Estimates of abundance are critical to manage and conserve waterfowl and their habitats. Most surveys of wintering waterfowl do not use probability sampling; therefore, development of more rigorous methods is needed. In response, I designed and evaluated an aerial transect survey to estimate abundance of wintering ducks in western Mississippi during winters 2002?2004. I designed a probability-based survey using stratified random and unequal probability sampling of fixed-width transects. To correct for visibility bias inherent in aerial surveys, I conducted an experiment to model bias and incorporated correction factors into estimation procedures to produce adjusted estimates. Bias-corrected estimates were most accurate. Precision of abundance estimates of total ducks met a priori goals (CV ≤ 15%) in 10 of 14 surveys. Based on a simulation study, the implemented survey design provided the most precise estimates, yet certain refinements remained possible. I also illustrated potential applications of survey results in the context of conservation and management of wintering waterfowl populations and habitats. I described patterns of abundance within and among winters, including a comparison with surveys conducted during winters 1988?1990 that revealed mallard abundance decreased 65% from the late 1980s. I developed a method to illustrate population abundance spatially for scientific and public education. I attempted to explain temporal variation in abundance estimates relative to variables potentially representing hypotheses explaining regional distributions of ducks. I concluded the data provided stronger support for factors related to energy conservation by ducks than factors related to energy acquisition. Finally, I determined associations between duck distributions and habitat and landscape features in accordance with the habitat-complex conceptual model. Landscapes with greater interspersion and diversity of wetlands attracted increased numbers of ducks, a though other factors such as wetland area also were important. I concluded that this study advanced methodologies to survey wintering waterfowl. Although improvements were warranted, I recommend this survey design for continued monitoring of wintering ducks in western Mississippi. Furthermore, I suggest habitat management on public and private lands should include complexes of seasonally flooded cropland, moist-soil, forested, and permanent wetlands to potentially increase wintering duck numbers in western Mississippi.