Biology and conservation ecology of selected saproxylic beetle species in Tasmania's southern forests

Many saproxylic beetle species (those associated with dead wood) are considered to be threatened by intensive forest management, yet few have been subjected to autecological studies that might help in their conservation. For this thesis, two lines of research were followed to investigate the autecology of six species thought to be vulnerable to intensive forestry. The study was carried out in the wet-eucalypt production forests of southern Tasmania. The first line of research explored methodologies that could be used to help understand life-history and feeding traits of log-dwelling beetles, in order that the resultant autecological information might be used as input into models of species persistence. Study 1a investigated the sexual development and life-history strategies of two sympatric species of stag-beetle, Lissotes menalcas and L. cancroides; and study 1b used molecular approaches to explore the feeding relationship of a mycophagous saproxylic beetle Prostomis atkinsoni and its putative food source, ‚ÄövÑv=gingerbread rot'. Both studies in this line of research provided useful methodological insights. Additionally, the findings from study 1asuggest that coexistence of L. menalcas and L. cancroides may be enabled by their differing patterns of sexual development across seasons; while the findings of study 1b were equivocal, in that numerous fungal species were detected in the gut-contents of adult and larval P. atkinsoni but few were found in both; furthermore, there was little overlap in the mycota of beetles' guts and gingerbread rot. Unexpectedly, some species of basidiomycete known to be wood-rotters were found in the guts but not in the gingerbread rot itself. The second line of research comprised predictive modelling approaches at the level of (a) microhabitat (coarse woody debris) and (b) site and landscape, both aimed at identifying the habitat requirements of all six species and exploring the possibility of making reasonable predictions about the species' occurrence at each of these spatial scales. In study 2a, models performed well in relating the probability of occurrence of each of the six beetle species to the presence or absence of particular suites of rotten-wood types. The addition of site-scale variables did not greatly enhance the models' abilities to predict occurrence of the six species of beetles, but at this level, forest stand structure and age of regeneration generally contributed the most explanatory power. In study 2b, geology, aspect, and time since last fire were good predictors of beetle occurrence across the study-area, although predictive ability may have been inflated by overfitting and autocorrelation. Overall, the models indicate that in unharvested forest these beetles tend to preferentially occur (within their chosen rotten-wood types) in multi-aged stands with a mature eucalypt element, while their continued occurrence in forest regenerating following clearfelling may be explained by their occupation of ‚ÄövÑv=legacy' coarse woody debris derived from the previous unharvested stand. The findings from both lines of research provide clear pathways for future exploration and modelling of species persistence in production-forest landscapes, and could help guide management practices accordingly, particularly in relation to the required spatial and temporal arrangement of mature forest.