A SIMULATION STUDY OF THE IMPACTS OF POPULATION SUBDIVISION ON THE MOUNTAIN BRUSHTAIL POSSUM TRICHOSURUS-CANINUS OGILBY (PHALANGERIDAE, MARSUPIALIA) IN SOUTH-EASTERN AUSTRALIA .2. LOSS OF GENETIC-VARIATION WITHIN AND BETWEEN SUBPOPULATIONS

The subdivision of populations that results from habitat fragmentation can impact the amount and pattern of genetic variation in metapopulations of organisms. Subdivision can lend to a loss of heterozygosity, and increased inbreeding within subpopulations is one factor that may contribute to an extinction vortex. However, a number of genetic models have also shown that, under some conditions, subdivision of populations can protect heterozygosity and allelic diversity, and that small subpopulations can become adapted to inbreeding. In this study we further investigated the relationships between habitat fragmentation, population subdivision, and various measures of genetic variability. VORTEX, a computer program for Population Viability Analysis (PVA), was used to simulate the impacts of fragmentation and subdivision on the genetic variation within and between subpopulations of the mountain brushtail possum Trichosurus caninus Ogilby, a forest-dependent species of arboreal marsupial that inhabits wet sclerophyll forests and rainforests in eastern Australia. For this study and a related investigation of the demographic stability of populations of T. caninus, hypothetical populations of 100, 200 and 400 animals were partitioned into one to 10 subpopulations that were, in turn, linked by varying rates of migration between subpopulations. Our application of PVA allowed an examination of the effects on population dynamics of demographic fluctuations, genetic drift, and interactions between demographic and genetic processes. The results of our analysis showed that both gene diversity (heterozygosity) and allelic diversity of subpopulations and metapopulations were lost rapidly when a population was subdivided. The increased levels of demographic stochasticity that resulted from population subdivision caused a decrease in the effective population sizes of the metapopulations that were modelled as well as the ensemble of subpopulations in each case. This, in turn, resulted in move rapid losses of genetic variation than would occur in the absence of demographic fluctuations. The extinction of some subpopulations further accelerated the loss of gene diversity and alleles. In addition, the genetic and demographic instability of small, fragmented populations diminished the effectiveness of selection against recessive lethal alleles. Migration among subpopulations partly reversed the impacts of population subdivision. However, even high rates of migration did not eliminate demographic fluctuations or prevent subpopulation extinction. Therefore, although gene flow largely prevented genetic divergence between subpopulations, migration did not prevent subdivided populations from losing genetic variation more vapidly than single populations of the same total size. The dynamics of small, fragmented populations were shown to be critically dependent on the interactions between demographic and genetic processes. Thus, our findings ave in striking contrast to the predictions made by several previous models of genetic changes in metapopulations that exclude consideration of the impacts of demographic stochasticity. These results demonstrated the value of PVA simulation models in revealing some of the consequences of fragmentation that have previously been overlooked. Our investigation has indicated that populations of forest-dependent taxa and other species that depend on habitats undergoing rapid change due to human activities may have to be relatively large and continuous to avoid significant losses of genetic variation.