Biogeography of Tasmanian native land snails

The Australian island state of Tasmania has a well-sampled land snail fauna consisting mostly of endemic species, mostly confined to particular portions of the state. This thesis analyses the known distribution of different species within the state with the aims of: (i) describing and summarising these distributions, (ii) examining the applicability of models used for other Tasmanian taxa to land snail distributions, (iii) categorising and tentatively explaining snail distributions where possible, and (iv) assessing some ramifications of snail distributions for conservation planning. One hundred and six species/morphospecies are discussed, of which approximately 39 were not formally or informally recognised before this project commenced. Undescribed species (and genuine species incorrectly synonymised) are here identified chiefly through qualitatively significant shell-feature distinctions, or through reliable quantitative differences. Species distribution is analysed primarily at a resolution of 10x10km grid squares, at which level 4272 records (including project specific fieldwork aimed at improving the representativeness of sampling) are used. Discussion is chiefly inferential in the absence of adequate statistical models, but overall database statistics are used to comment on whether apparent gaps in species ranges are likely to be meaningful. The single most significant pattern in the Tasmanian mainland snail fauna distributions separates the west and far south from the remainder, a pattern to which geology, rainfall, vegetation and substrate may all contribute. The manifestation of this pattern varies between different contributing species. This result is practically identical to results previously obtained for other invertebrates, but another broad-scale pattern, the influence of the face of the Great Western Tiers in the central north, is not. Congruence with bioregional models based on trees and vertebrates is only approximate and piecemeal, showing potential limitations for conservation planning of models that do not include poorly dispersing taxa (such as non-flying invertebrates) in their source data. Known facets of local endemism such as island endemism, karst endemism, parapatric species mosaics, alpine endemism, \toeholding\" and glacial refuge endemism are all reflected in the fauna to some degree. The value of these for predicting the likely locations of undescribed or as yet undifferentiated twha is however undermined by the frequency of distributions not explicable by any known direct cause and by the relatively small and spatially variable contribution of each specific known cause. The potential merits of limiting the spatial scale of comprehensive habitat loss especially in poorly-surveyed areas are therefore discussed."