The role of population monitoring in the management of North American waterfowl

Despite the effort and expense devoted to large-scale monitoring programs, few existing programs have been designed with specific objectives in mind and few permit strong inferences about the dynamics of monitored systems. The waterfowl population monitoring programs of the U.S. Fish and Wildlife Service, Canadian Wildlife Service and state and provincial agencies provide a nice example with respect to program objectives, design and implementation. The May Breeding Grounds Survey provides an estimate of system state (population size) that serves two primary purposes in the adaptive management process: identifying the appropriate time-specific management actions and updating the information state (model weights) by providing a basis for evaluating predictions of competing models. Other waterfowl monitoring programs (e.g., banding program, hunter questionnaire survey, parts collection survey, winter survey) provide estimates of vital rates (rates of survival, reproduction and movement) associated with system dynamics and variables associated with management objectives (e.g., harvest). The reliability of estimates resulting from monitoring programs depends strongly on whether considerations about spatial variation and detection probability have been adequately incorporated into program design and implementation. Certain waterfowl surveys again provide nice examples of monitoring programs that incorporate these considerations.