Michael W. Brown James R. Nassar Kenneth J. Reinecke 1992 <p>Winter waterfowl surveys rarely use sampling methods, and little is known about the precision and biases of their population estimates. Consequently, we developed aerial transect surveys (n=5) in 4 strata comprising 16 substrata in the lower Mississippi Alluvial Valley during winters 1987-88 through 1989-90 to estimate mallard (<i>Anas platyrhynchos</i>) population indices and determine regional patterns of habitat use. Mallard population indices ranged from 1,147,628 (SE=192,341) in December 1988 to 1,790,708 (SE=179,406) in January 1988. Coefficients of variation (CV's) for early winter surveys averaged 0.15 and those for late winter surveys averaged 0.10. During early winter, 59-69% of mallards were on wetlands with water regimes managed for waterfowl; whereas in late winter, 52-79% used wetlands with unmanaged water regimes. Late winter was wet during 1987-88 and 1988-89, and most mallards (62-68%) were on naturally flooded croplands. Use of forested wetlands (3-11%) and moist-soil habitats (3-29%) varied among surveys but was not correlated with water conditions. The number of mallards using naturally flooded croplands (e.g., &gt;1,100,000 in Jan 1988) illustrated the extent of habitat use on private lands. We recommend transect surveys (e.g., 5-yr intervals) for evaluating responses of mallard populations to management programs and as a sampling framework for integrating regional waterfowl research and management data.</p> application/pdf 10.2307/3808867 en Wiley Evaluation of aerial transects for counting winter mallards article