The relationship between the spatial distribution of turtle nest sites and nest predation was investigated in a landscape encompassing a 12-ha freshwater wetland in west-central South Carolina, USA. The goals of the study were to determine if nest predation is a function of nest density, and if two other spatial factors, habitat type and distance of nests from the wetland, influence survival rate. Gravid females of three turtle species (Kinosternon subrubrum, Pseudemys concinna floridana, and Trachemys scripta) inhabiting the wetland were monitored from the time they exited the wetland to nest until they returned from their nesting forays. A total of 145 nests were located during three nesting seasons (1992-1994). Nests were monitored from the day of oviposition until they were deemed to have successfully produced hatchlings or depredation was documented.
The three-year mean predation rate was 84.2% (80%, 88%, and 81% for 1992, 1993 and 1994, respectively). Modeling of the point pattern of nest locations suggests that nests were distributed at varying densities and that density changed with habitat type and distance to the wetland. Logistic regressions of predation rates against two crowding indices demonstrate no significant evidence for density-dependent nest predation. Furthermore, autologistic regression provided no evidence indicating that nest survival is dependent on the survivorship of neighboring nests. Logistic regressions of predation rates against habitat type and distance from the marsh of origin suggest that predation rate is not a function of these two spatial characteristics. In summary, no evidence was found that linked predation rate to any of the spatial characteristics of nest distribution that were examined. The lack of density dependent nest predation is contrary to findings from some previous avian and chelonian studies, but supports the findings of other researchers. It is suggested that predation during the current study was either density independent or that any local density effects were masked by an ideal free distribution of turtle nests. Our data support the findings of several researchers that suggest density, as a single factor, is not a generalizable predictor of nest predation probability.