Assessing the viability of lowland tapir populations in a fragmented landscape
Durrell Institute of Conservation Ecology
2010
Thesis
Ph.D.
292
University of Kent
Canterbury, United Kingdom
This thesis aimed to assess the ecological factors that determine the long-term persistence and viability of animal populations across severely fragmented landscapes. The lowland tapir, Tapirus terrestris, and the fragmented Atlantic Forests of the Interior of the Pontal do Paranapanema Region, São Paulo State, Brazil, were used as a model to illustrate this assessment. Both empirical and modelling approaches were used. The empirical approach focused on aspects of tapir spatial ecology, intra-specific interactions, spatial and temporal interactions between tapirs and the landscape, as well as estimates of tapir abundance in Morro do Diabo State Park (370 km²) and seven smaller forest fragments (4-18 km²) where tapirs were present. The modelling approach consisted of a Population Viability Analysis (PVA) using the software VORTEX. Spatial ecology, intra-specific interactions, and interactions between tapirs and the landscape were estimated by radio-telemetry. Population sizes were derived from tapir densities obtained by radio-telemetry, nocturnal line-transect sampling, and Footprint Identification Technique (FIT). Lowland tapirs in Morro do Diabo had very large home ranges (4.7 km²) when compared to other sites, particularly contiguous habitats. Tapir home ranges had very complex internal structures, including multiple core areas of use, which comprised a very small proportion of the home range (50% core area, 17% of the home range; 25% core area, 6% of the home range). Little seasonal variation in size and location of home ranges and core areas of use were observed. These patterns were consistent for both sexes and different age classes. Telemetry results have shown that a minimum of 20 months of data collection and approximately 300 locations are necessary to determine home range size for adult lowland tapirs. Tapirs exhibited extensive home range overlap (30%), as well as overlap of core areas of use (20%). No evidence of spatial territoriality was noted. Tapirs incorporated portions of all available habitat types within their home ranges and core areas of use, but significantly selected riparian habitats, where they performed most of their main activities, particularly foraging. Tapirs avoided areas of agricultural and pastoral land, as well as secondary growth forests. It was estimated that Morro do Diabo hosts a population of 130 tapirs and, altogether, the seven forest fragments host 22 additional individuals. Tapirs have low population growth rates and so are very susceptible to threats such as road-kill, infectious disease, and fire, particularly in the small forest fragments. Results from the PVA model projected that the tapir population in Morro do Diabo has zero probability of extinction and is likely to persist over the next 100 years. However, the population is not large enough to maintain 95% of genetic diversity over the long-term. A Minimum Viable Population of 200 tapirs would be required to ensure long-term viability. The model showed that, without dispersal of tapirs from Morro do Diabo, tapirs in the small fragments will go extinct over the next 100 years. However, this study showed that tapirs in the Pontal do Paranapanema Region moved fairly easily through areas of non-natural habitat in between patches of forest, indicating a certain level of landscape functional connectivity. This provided evidence of a tapir metapopulation scenario, which proved to be a determinant factor for the persistence and viability of lowland tapirs in the Atlantic Forest of the Interior. Overall, the long-term persistence and viability of animal populations across severely fragmented landscapes appears to be dependent on the maintenance and full protection of complex landscape networks. These networks must include some large patches of habitat that can host larger animal populations and function as source areas for dispersal of individuals to smaller populations in sink habitats. Patches of forest comprising these networks must incorporate required habitat types where animals can find the resources they need in order to survive and persist. Most essentially, there must be an appropriate level of landscape connectivity, either structurally or functionally, in order to facilitate biological fluxes between patches and promote the maintenance of a demographically and genetically healthy metapopulation.