MODELING THE VIABILITY OF METAPOPULATIONS OF THE ENDANGERED LEADBEATERS POSSUM IN SOUTHEASTERN AUSTRALIA

The population simulation computer program, ALEX, was used to estimate the probability of extinction of the endangered species Leadbeater's possum, Gymnobelideus leadbeateri McCoy, in complex network:; of remna nt habitat patches within south-eastern Australian forests. Forest inv entory data stored in a Geographic Information System (GIS) were inter rogated to derive information on the spatial location and size of pote ntially suitable habitat patches within four management area:, in the central highlands of Victoria: the O'Shannassy Water Catchment and the Murrindindi, Ada and Steavenson Forest Blocks. This information was c ombined with data on the life history attributes of G. leadbeateri, as well as sub-models that tracked the dynamics of key habitat attribute s, to simulate the behaviour of metapopulations of the species over th e next 150 years. Analyses were completed for simulated populations of animals in: (i) hypothetical isolated patches of old growth forest of varying size, and, (ii) assemblages of real patches of potentially su itable habitat within each of the four study areas. In addition, a det ailed investigation was completed for the Ada Forest Block in which ar eas of habitat of different size and spatial isolation were deleted fr om the patch structure:. The results of our analyses indicated that si ngle isolated populations inhabiting patches of less than or equal to 20 ha were highly susceptible to extinction. For areas that were large r than this, the probability of survival increased with the patch size and approached 1% for patches that were > 150 ha. Our analyses of com plex ensembles of patches indicated that G. leadbeateri is vulnerable to extinction during the next 150 years within the Murrindindi and Ada Forest Blocks. The species was predicted to have better prospects of survival in the Steavenson Forest Block and the O'Shannassy Water Catc hment where there are relatively large remaining areas of old growth f orest (approximately 250 ha and 6420 ha respectively), although we did not include the impacts of catastrophic events such as logging and wi ldfire in our investigation. The retention of the few remaining relati vely large habitat patches (i.e. those > 12 ha) was predicted to be cr itical for the persistence of G. leadbeateri in the Ada Forest Block. Conversely, the deletion of relatively small (< 3 ha) and/or isolated patches of old growth had a minor impact on the predicted probability of extinction. We discuss the implications of our findings for the dev elopment of conservation strategies for the species, particularly the identification of areas to be excluded from logging.