We evaluated a habitat suitability (HSI) model developed for mallards (Anas platyrhynchos) wintering in the Lower Mississippi Valley by comparing mallard densities obtained from aerial surveys with habitat suitability indices derived from satellite imagery for 25, 256-km2 sampling units. Regression models that related mallard densities to habitat suitability indices accounted for only 29% of the variability in the data and the 95% confidence interval of predicted mallard densities included zero for most habitat suitability indices evaluated. Thus, we conclude that the published HSI model is a poor predictor of wintering mallard density in the Lower Mississippi Valley. We suggest model revision to allow users to remotely obtain model inputs for habitat characteristics at landscape scales. Further, we suggest the model be revised to consider yearly variation in habitat and flood conditions that better reflect the ability of an area to support wintering mallards.
