CAPTURE-RECAPTURE METHODS WITH UCA

¹ Department of Science Education and Department of Oceanography, Florida State University, Tallahassee, 32306

1. Estimates of the size of a population of sand fiddler crabs, Uca pugilator (Bosc), were made with the Lincoln Index and Jackson's "positive" and "negative" models in the mobile phase of the activity cycle, and by stratified sampling of the immobile phase.

2. Restricted random sampling was employed because of extremely contagious distribution of the crabs during the mobile phase and of desire to achieve better coverage of the burrowing area than unrestricted random sampling might have yielded during the immobile phase.

3. During the mobile phase, the overall population was divided into four mobs, one separated from the rest by a physical, and the others from each other by a behavioral barrier. Only three out of 8967 marked individuals were recaptured outside their home territories. Some evidence suggesting that the mobs are derived from subpopulations occupying circumscribable burrowing areas was obtained.

4. Appropriateness of the Lincoln Index was supported by immediate random dispersal of marked crabs among the unmarked mobs, and by failure of the Jackson methods to reveal evidence of population size change during the sampling period.

5. Recognition of the inhomogeneity of variances of estimates obtained by the several methods precluded statistical testing of the estimates on the assumption that they arise from the same population. While it is manifest that all the samples were drawn from the same over-all, isolated population of crabs, it is also obvious that the variance of one estimate was based on spatial distribution, while those of the others were based on behavior. Nevertheless, it was concluded that the assumptions implicit in capture-recapture methods are satisfied during the mobile activity phase of Uca pugilator and that for reasons of practicality and preservation of the habitat, these are preferable to estimates of population size by excisive random, areal sampling during the immobile phase.

6. It is recognized that larger samples would have contributed to more satisfactory validation of the methods, to reduction of variance, and to positive bias, But it was concluded that those which formed the basis of this study were atleast adequate, or even preferable, because of probable effects on the crabs' behavior of extended disturbance of mobbing and of the burrowing area.