Impacts of habitat fragmentation on dispersal of native mammals

© 2011 Dr. Achim EberhartHabitat fragmentation is widely regarded as a major threat to biodiversity worldwide. In addition to the reduction of the total area of habitat available for wildlife, clearing of native vegetation creates disconnected patches of habitat immersed in a matrix, which does not provide resources for most species of wildlife. The remaining habitat patches often accommodate small populations of native animals, which suffer increased susceptibility to extinction. Movement of individuals between habitat patches, however, can interconnect such populations so that they effectively behave as larger and more stable metapopulations. Hence the traversability of any matrix for wildlife is an important attribute of fragmented landscapes and impacts on their ability to support viable populations. Further, the capability of different species to cross various matrix types varies. ‘Landscape connectivity’ describes the functional isolation of remnant areas of habitat and considers not only the distance between them but also the species-specific ability to successfully move through the matrix. In this thesis I examine the dispersal behaviour and population genetic structure of two species of marsupials in a fragmented landscape. I describe the movements of bobucks (Trichosurus cunninghami) and agile antechinus (Antechinus agilis) in the Strathbogie Ranges in north-eastern Victoria. Being forest-dependent, both species occur solely in remnants of native forest. Both also have a life-history pattern that is characterised by the dispersal of subadult males. However, they differ greatly in body size. I used GPS (for bobucks) and VHF (for antechinus) tracking to monitor individuals within their home ranges in the pre- and post-dispersal stages of their life cycle as well as during dispersal. I also used genetic parentage assignment tests to identify movements of antechinus. While the literature reports that the vast majority of subadult males of both species disperse, in my study a high proportion remained philopatric. Exploratory movements, beyond the boundaries of their home ranges, were recorded for adult and subadult bobucks; however, subadult males that later dispersed rarely exhibited this behaviour. In the months before the dispersal season, subadult bobucks that would subsequently remain philopatric appeared to occupy larger home ranges than dispersers. One possible explanation for this observation is that resource availability drives dispersal in this species and only those individuals that cannot secure access to sufficient resources in their natal home ranges disperse. In order to examine the response of dispersers to different landscape elements I monitored individuals during dispersal. Some bobucks and antechinus dispersed long distances. The very detailed dispersal pathways of bobucks, that I recorded using GPS tracking, revealed a clear pattern with all individuals dispersing along wooded landscape elements when these were available. Individuals were able to cross considerable distances across pasture using paddock trees and linear forest remnants as stepping stones and dispersal corridors. While some antechinus are likely to have crossed small gaps in linear forest remnant habitat, they appeared to be more restricted to forest during dispersal than bobucks. Genetic parentage analysis revealed that some male antechinus moved very considerable distances during the mating season, a behaviour that could significantly increase gene flow in this species. I applied genotypic analyses to infer gene flow in this system and used a landscape genetic approach to determine the connectivity of habitat in the Strathbogie Ranges for both bobucks and antechinus. I also investigated the influence of wooded landscape elements on gene flow in these two species. These analyses indicated that the distribution of forest remnants and individual trees play no role in determining gene flow in bobucks or antechinus within the Strathbogie Ranges. Surprisingly higher spatial genetic structure, which could indicate restricted dispersal, was detected in the larger, more mobile bobucks compared to antechinus. This seems to contradict my findings based on direct observations of dispersers in both species, namely the importance of paddock trees and linear forest remnants as dispersal conduits for bobucks and stronger restriction to forest of dispersing antechinus in comparison to bobucks. In fact, however, this discrepancy emphasises the difference between measurements of dispersal via direct observations versus genotypic inferences of gene flow. While observations of individual dispersers can provide detailed information on the behaviour of animals when they are traversing fragmented landscapes, data on their survival and breeding success post-dispersal is virtually impossible to obtain for many species. In contrast, molecular genetic analyses yield information on the reproductive outcome of dispersal across several generations, but usually cannot inform about dispersal pathways. The two approaches that I used are complementary, and combining them can help to avoid inaccurate conclusions about landscape connectivity for metapopulations, that are based on either technique alone. In my study, the landscape genetic model applied may have underestimated the true value of paddock trees as conduits for bobuck dispersal. Also the small number of antechinus that I was able to monitor during dispersal probably led to an underestimation of this species’ ability to leave its preferred forest habitat and traverse the matrix. Overall, the availability of linear forest remnants, scattered paddock trees and small patches of forest in the Strathbogie Ranges appeared to create high functional connectivity for populations of both species. My research emphasises the importance of maintaining and/or restoring structurally complex landscape elements, so that these can act as dispersal conduits for different species with differing mobility. This approach is an important tool to mitigate the effects of habitat fragmentation on native wildlife