Translocation of an imperilled woodrat population: integrating spatial and habitat patterns

Many species have strong habitat preferences that directly influence population viability. For successful reintroduction of threatened populations that rely on habitat structures, the correct placement of artificial structures is also important to population persistence. In this paper, we present a hierarchical approach to the problem of translocating animals that rely on permanent habitat structures, in which we first use population dynamics data to identify areas of suitable habitat, and then identify optimal configuration for habitat structures. We use data collected from a non-endangered, conspecific population of the endangered riparian woodrat (Neotoma fuscipes riparia) to examine the degree to which the distribution of dens in translocation sites might influence the likelihood that animals persist in their new environment. We characterize the habitats in which dens occur, analyze their spatial clustering, and compare them to temporal changes in population status for sex and age classes. We compare the potential efficacy of translocation efforts using spatial analysis versus solely habitat-based approaches and identify the optimal spatial configuration of dens that should be considered in this translocation effort. We found that patterns of habitat use were positively correlated with overstorey cover, and animal weight was positively correlated with understorey cover. Woodrats appear to select den locations on the basis of understorey cover, but also benefit from dense overstorey cover and distance to nearest tree. Our results suggest that in translocation efforts, artificial dens should be placed in clusters within a radius of 15 m, as values above and below this value showed negative correlations with body mass. Translocations should occur after reproductive events, which occur in April and August for woodrats in southern California. Our analyses provide practical guidelines in determining appropriate timing and spacing for translocation events in the context of animal condition, minimizing disease transmission, and reproduction. All correspondence to: L. R. Gerber. Tel: 480 727 3109; Fax: 480 965 2519; E-mail: [email protected]. Animal Conservation (2003) 6, 309–316 © 2003 The Zoological Society of London DOI:10.1017/S136794300300338X Printed in the United Kingdom Empirical evidence suggests that several factors may influence the efficacy of species translocations. First, efforts to reintroduce mammals appear to have been more successful than efforts for other taxonomic groups (Scott, Murray & Griffith, 1999). The release of animals into high-quality habitat and the use of animals from non-endangered or increasing subpopulations contribute to translocation success (Griffith et al., 1989). Reintroductions are thought to be less successful in the absence of high-quality habitat or in areas outside of the historical range. For example, efforts to reintroduce the thick-billed parrot at the northern extreme of its range failed, as did efforts to reintroduce whooping cranes in Idaho outside their historical range (Scott et al., 1999). Conservation strategies should also consider the spatial patterns of critical structures such as artificial reefs and bluebird nest boxes for habitat and species recovery (e.g., Newton, 1994; Pickering & Whitmarsh, 1997). When translocating animals and creating structures, it is also important to consider the risk of transmitting diseases. For example, the translocated and regionally endangered Allegheny woodrat (Neotoma magister) collects faeces from exotic raccoons for nestbuilding material and, as a result, suffers such high mortality that woodrats apparently cannot persist where raccoons are common (LoGiudice, 2000). We present a hierarchical approach to the problem of translocating animals that rely on permanent habitat structures, in which we first use population dynamics data to identify areas of suitable habitat, and then identify optimal configuration (using Ripley’s L-function analyses) for habitat structures within this habitat. Using data from an unimperilled population, we adopt this approach to make recommendations for the translocation of endangered riparian woodrats (Neotoma fuscipes riparia) in California using artificial dens. This population, a subspecies of the dusky-footed woodrat (Neotoma fuscipes), is limited in range to a 250 acre habitat in central California, includes fewer than 500 animals, and appears to be declining in abundance (Williams, 1993). Loss and fragmentation of habitat are the principal reasons for the decline of the riparian woodrat population. Much of this loss was the result of the construction of large dams and canals that diverted water for the irrigation of crops and permanently altered the hydrology of valley streams (Close & Williams, 1988). Currently, fire and flood pose severe threats to the only major remnant riparian community in the San Joaquin valley and the last refuge for the riparian woodrat. To reduce the potential risks associated with floods, wildfires and disease (Williams, 1993), the Endangered Species Recovery Program (ESRP) has recently proposed establishment of this subspecies in other locations. Prior to this proposed translocation effort, a number of factors should be evaluated, such as food availability, nesting sites and travelling corridors, and competition and interbreeding with other woodrats (Williams, 1993). Our study is intended to complement current research efforts being conducted by ESRP that will serve as a basis for the translocation of this subspecies. Woodrats (genus Neotoma) construct complex dens composed of sticks of various sizes (Lindsdale & Tevis, 1951). They rely on these dens for shelter, protection from predators, and food storage, and the dens are often inherited between generations (Kelly, 1989). A significant amount of time and energy goes into constructing and maintaining these dens, and their spatial patterns and location in relation to construction material are important to population viability (Olsen, 1973; McGinley, 1984). For example, woodrats require habitat that provides cover and food resources, and dens are often located in or near a large core element such as a rock or a tree (McGinley, 1984). Experiments have shown that woodrats exhibit a preference for low trees with overhanging branches over other configurations (Olsen, 1973). Because woodrats have strong habitat preferences (Cameron & Rainey, 1972; Wright, 1973) and these preferences affect population viability, successful reintroduction of threatened populations, such as the endangered riparian woodrat (N. f. riparia), relies on the correct placement of artificial dens. Even though woodrats are prolific den builders once they become established in new habitats (Bonaccorso & Brown, 1972), artificial dens are likely to increase the probability that animals occupy new habitats in translocation efforts. Despite the importance of this information, we are not aware of previous studies of the spatial patterns of woodrat dens. In this paper, we use data collected from a nonendangered, conspecific population of woodrats to estimate how the distribution of dens in translocation sites might influence the likelihood that animals persist in their new environment. We characterize the habitats in which dens occur, analyze their spatial clustering, and compare changes in weight over time for sex and age classes. We use our results to identify ideal conditions for establishing woodrat dens in translocation sites. We compare outcomes for translocation using spatial analysis versus solely habitat-based approaches and identify the optimal spatial configuration of dens that should be considered in translocation efforts.