On catch-effort methods of estimating animal abundance

By making a clear distinction between catch and removal, catch-effort methods can be applied in broader circumstances than generally believed. For instance, animals can be added to a population rather than removed, data on animal sightings can be treated as 'catch,' and more than one bind of data (e.g., data on both sightings and kill) can be used jointly to produce population estimates. Models and estimation procedures are presented which illustrate these possibilities. Also, least squares estimation (generally nonlinear) can be used in conjunction with any likelihood model, and we illustrate this procedure in a simulation study involving both natural and catch hazard functions. Properly formulated, least squares estimates appear to be as accurate and precise as corresponding maximum likelihood estimates.