The Accuracy of a Population Estimation from Multiple Recapture Data

Abstract

Several methods have been developed for estimating population size from multiple markrelease-recapture data, which vary to suit the animal and environment under study (for example: Fisher & Ford 1947; Jackson 1939; Leslie, Chitty & Chitty 1953 and earlier; MacLeod 1958; Jolly 1965). In some cases a method is available whereby confidence limits may be found for certain of the parameters estimated (Leslie & Chitty 1951; Bailey 1952; etc.). The usefulness of such variances naturally depends on the validity of the assumptions upon which they are based. Under field conditions it is not uncommon to be able to show that some of the assumptions do not hold, so that any estimated variance would accordingly be invalidated. It is much rarer to possess sufficient independent information about the size of a population to be able to make a direct comparison between the estimate and a figure known with some degree of certainty to be correct. In 1964 an experiment was performed in the field (Brower, Cook & Croze 1966) which, although designed for a different purpose, allows a comparison of this kind to be made: in effect a multiple recapture census was conducted on a population of known size. The results of the comparison have been calculated because the question of accuracy is of very great importance to the study of wild populations, and it is particularly difficult to assess accuracy when the population is small. THE EXPERIMENT In an experiment to test the effectiveness of mimicry, male individuals of the North American moth Hyalophora promethea (Drury) were painted with different patterns, liberated in Trinidad where they do not naturally occur, and subsequently recaptured. Three colour classes were used but the differences in recapture rates were relatively small and do not seriously affect the accuracy of the analysis to be made. The different types have therefore been grouped together. A total of 1093 individuals was liberated for the first time, this being the entire population of the moth in Trinidad at the time. Recaptures were made using a trap containing caged promethea females, which attract the males over great distances by means of the scent they emit. If an individual caught in this way was in good condition, it was re-released and stood a chance of being recaptured a second time. Thus, by regarding the individuals trapped for the first time as the captured sample, the ones from this group which were liberated again as the releases, and the members of that category which were subsequently trapped as the recaptures, we have sufficient information to make an estimate of the total population available in the area. The basic data are given in Table 1.