Survival and Recruitment in a Human-Impacted Population of Ornate Box Turtles, Terrapene ornata, with Recommendations for Conservation and Management

—Alteration and loss of habitat is a major factor in the recent declines of many turtle populations. However, there are few studies of turtle populations in areas that are used intensively by humans. We used temporal symmetry modeling and an information-theoretic approach to model selection to estimate survival and recruitment in a population of Ornate Box Turtles, Terrapene ornata, in fragmented, isolated habitat over an eight-year period. Apparent annual survival was high during this period (0.97, SE 5 0.06), as was the seniority probability (0.9560.04). Recruitment into the adult population (k) was estimated at 1.02 (6 0.06). Our results suggest a healthy population, but we note several reasons for a cautious management approach. These include a vulnerability of k to the removal of adults, the need for increased recruitment to offset loss of genetic diversity, and the uncertainty of our estimates resulting from the sampling and modeling processes. Habitat loss and fragmentation increase risk of extirpation for populations of a variety of taxa (Burkey, 1995; Fahrig, 2002; Hokit and Branch, 2003a). Once a patch of habitat becomes small and isolated, stochastic and deterministic factors can lead to a decline in the vital rates of populations inhabiting that patch (Lande, 1993; Hokit and Branch, 2003a,b). Even when populations persist, severe declines in vital rates can result in bottlenecks and a loss of genetic diversity (Hoelzel, 1999; Kuo and Janzen, 2004). From a conservation and management perspective, it is, therefore, crucial to assess the demographic and genetic health of populations in areas where human activity has reduced the amount of available habitat. Habitat alteration is implicated as a critical factor in the widespread decline of turtle populations (Mitchell and Klemens, 2000), but there are few attempts to rigorously estimate parameters of turtle populations in areas that are used intensively by humans (Doroff and Keith, 1990; Kazmaier et al., 2001). Studies of this kind are needed to guide management decisions given current rates of habitat destruction and the conservation status of many turtle species. Furthermore, there are few studies that use data from multiple sources (i.e., population ecology and molecular genetics) to gauge the health of turtle populations in degraded areas (Rubin et al., 2001, 2004). Assessing demographic health of turtle populations begins with estimating adult survival and recruitment into adult populations. Growth rate of turtle populations is most sensitive to adult survival (Heppell, 1998), and loss of adults can lead to population declines regardless of reproductive rates (Congdon et al., 1993, 1994; Heppell et al., 1996). Although many studies of turtle demography use mark-recapture models to estimate population parameters, few authors consider the assumptions of these models (Lindeman, 1990), and variance (uncertainty) of population estimates is rarely reported (Langtimm et al., 1996). Furthermore, models such as the Cormack-Jolly-Seber (CJS) model for open population survival estimates (Cormack, 1964; Jolly, 1965; Seber, 1965) and information-theoretic approaches to model selection such as Akaike’s Information Criterion (AIC; Akaike, 1973) are used infrequently in studies of turtle demography (but see Kazmaier et al., 2001; Fonnesbeck and Dodd, 2003). These methods are useful because they allow researchers to construct models of population processes and estimate population parameters based on knowledge of the study organism and study methods. In turn, researchers can determine what models are most appropriate for the data, retain the ‘‘best’’ model or models, and calculate reliable estimates of uncertainty for parameter estimates (Anderson et al., 2000; Burnham and Anderson, 2002; Fonnesbeck and Dodd, 2003). In this study, we used a CJS framework and AIC to estimate annual adult survivorship and recruitment in a fragmented and isolated population of Ornate Box Turtles (Terrapene ornata) on the periphery of the species’ geographic range. 1 Corresponding Author. E-mail: kbowen@iastate. edu Journal of Herpetology, Vol. 38, No. 4, pp. 562–568, 2004 Copyright 2004 Society for the Study of Amphibians and Reptiles