A new change-in-ratio procedure robust to unequal catchability of types of animal.

The change-in-ratio technique is a useful practical procedure for the estimation of game animal population sizes. The major problem with this technique is failure of the assumption that both types of animals are captured or sighted with equal probabilities. Here we extend the change-in-ratio technique to the case where there are two removals with emphasis on the special situation where there are two consecutive single-type removals. The advantage of this extension is that it allows an estimation procedure which is robust to unequal capture or sighting probabilities. It is also possible to test the assumption of equal sighting probabilities. Some numerical results on mean squared error of the population size estimator for the new design and the traditional design are given. The procedure is illustrated on some juvenile grass carp data collected in a small pond where the population size is known. We believe this technique is potentially useful to wildlife and fisheries biologists and that more statistical research would be beneficial.