Integrating climate change into conservation and restoration strategies : the case of the Tasmanian eucalypts

Global climate change will negatively impact major components of the world's biodiversity over the next century, and there is an increasing need for strategies and tools to guide species conservation and management. This issue is particularly relevant to species and provenance choices in ecological restoration where diverse strategies have been proposed to ensure future climate resilience. This thesis focuses on the Eucalyptus tree flora of the island of Tasmania, Australia, assessing the likely change in species modelled suitable habitat, and strategies for identifying suitable species and provenances for restoration tree plantings. This is a topical issue as the long-standing 'local-is-best' paradigm in ecological restoration is being increasingly questioned in the face of global climate change. Habitat suitability models were used to determine whether a eucalypt species susceptibility to future climate change (based on 2020, 2050 and 2080 projections) can be predicted by evolutionary history, habitat type or distribution pattern. A large extent of the modelled current climate habitat will be lost by the end of this century and 63% of records will be outside of modelled suitable habitat. Species from subalpine habitats were most at risk of maladaptation under future climate change. Susceptibility was not linked to evolutionary history or distribution pattern. Despite the loss of suitable climate habitat, many of the current conservation reserves still maintained populations of species that will not be at risk of climate maladaptation by 2080. To assist in guiding species choice for our target restoration region (the Midlands), we used the same climate projections to model the suitability of regionally local and non-local species through space and time. We identified one non-local and ten regionally local candidate species that would not be maladapted to the future climate habitat of the Midlands. The predicted decrease in the suitability of most local species in the future was compensated by the northern expansion of predicted suitable habitat for a regionally local and non-local species well outside their current distribution range. A framework was subsequently developed to identify provenances of targeted restoration species likely to be best adapted to the future climates of restoration sites to test and implement climate-adjusted provenancing. This framework accounted for inbreeding risk through habitat fragmentation and was translated into software (Provenancing Using Climate Analogues [PUCA]). This software matched current and future climate predictions for restoration sites to known provenances of the target species that are already growing in analogous climates for 2020, 2050 and 2080 projections. I demonstrate the application of PUCA and identified key outstanding assumptions which need empirical validation. Experimental trials were established to test the assumptions of the PUCA model and identify the climate components which have shaped adaptive variation among the provenances of a key restoration species E. ovata. A germination trial showed that provenance home-site climate effects did not affect seed germination characteristics. A glasshouse and a common garden field trial were established using range-wide collections of open-pollinated seed from 45 provenances (collected from 312 individuals), to study how home-site climate affects seedling functional traits and early field performance. Two orthogonal climate vectors (representing aridity and frost gradients) were shown to shape adaptive variation among provenances in seedling functional traits. Predicted changes in these vectors from contemporary to 2080 were spatially mapped as an alternative approach to predict (i) the adaptive limits of the species, (ii) the change in the climate selective surface, and (iii) provenance transfer functions. Only two-year performance data was available for the field trial, and this revealed little evidence of maladaptation of provenances from homes-sites spanning the selective gradient. The demonstration that provenances from future analogous environments have a broad transfer function and can be successfully established in restoration plantings is important for long-lived trees, as future-adapted individuals must establish under current climate regimes.