Abstract/Summary

Species conservation is expected to become more challenging in a changing climate, and this may be most acute in the speciose tropics, where species are generally adapted to a narrow temperature range from which novel climates are projected to emerge this century. Island endemic species are expected to show particularly low adaptive capacity. Yet, climate change vulnerability assessments are constrained by a shortage of detailed and long-term datasets, uncertainty over the impacts of extreme events, and poor understanding of implications for tropical animal phenology. These knowledge gaps limit our capacity to design management strategies to alleviate the impacts of future climate risks. This thesis explores these issues for species in Mauritius, where successful recovery programmes have involved the collection of detailed individual-based datasets. Long-term climate trends, including for extremes, were quantified for areas of high conservation importance. Detailed analyses were carried out into climate impacts on the breeding phenology and demography of the Mauritius kestrel (Falco punctatus) and echo parakeet (Psittacula eques). Climate trend analyses reveal local-scale variation in long-term trends, with nearby populations of conservation concern under different pressures from changes in rainfall. Detailed analysis for the Mauritius kestrel shows that its breeding phenology is tracking shifts in a seasonal climate window. Individual birds show plasticity in response to temperature increases, despite the effects of a severe population bottleneck. In contrast, echo parakeets show maladaptive adjustments to climate variability, with delayed breeding in drier years increasing exposure to tropical cyclones. Cyclones appear to mediate reproductive senescence, but this risk is reduced through the re-scheduling of breeding by supplementary food provision. Based on these results, vulnerability assessments and management plans for restricted-range species, such as those endemic to tropical islands, should incorporate local-scale climate trends, meteorological indices for extremes, and analysis of phenology and temporal exposure to climate risks.