|
Résumé :
|
Design is the most influential element in the pathway to statistical inference. Designs for capture-mark-recapture (CMR) studies have become more diverse and complex during the past 50 years. I review the history of CMR designs developed for discrete time models with a focus on the assumptions of those designs, the parameters that can be estimated, the question that can be addressed, and the interrelationship among designs. CMR designs can be used to address questions in a variety of disciplines including population dynamics, evolutionary ecology, life-history, and community ecology. Methodological advances have also enabled investigators to consider designs that include nearly all common sources of CMR data simultaneously. CMR designs have been generalized to account for many assumptions; however, more emphasis needs to placed on use of auxiliary variables to increase precision and address model assumptions. CMR designs also need to be placed in the larger context of experimentation. Finally, development of more specific designs to fully utilize data resulting from emerging technologies is needed.
|