The implications of climate change for biodiversity conservation and the National Reserve System: temperate grasslands and grassy woodlands

[摘要] Executive SummaryOwing to widespread fragmentation and degradation associated with agricultural and other development, temperate grassy ecosystems (TGE), including grasslands and eucalypt woodlands, represent some of the most threatened ecological communities in Australia. The goals of this study were to predict how climate change may impact these ecological communities in south-eastern Australia, and develop recommendations for adapting conservation and management efforts towards better outcomes for biodiversity under climate change. We combined an expert knowledge approach (including a workshop with over 20 TGE experts) with two types of community level spatial modelling (using artificial neural networks, ANN, and general dissimilarity modelling, GDM) to inform management responses. We focused on the following questions: (1) What are the structuring ecological processes in TGE? (2) How are ‘biotically-scaled’ environments likely to be impacted by climate change? (3) What attributes of TGE could influence the response of biodiversity to altered environments and processes? (4) What major ecological changes are likely? (5) What conservation management options are available to address these?The modelling approaches used in this study attempted to predict the “biotically important” stress that the TGE biome will be subject to under different climate scenarios. ANN models, focusing on vegetation structure, suggest very high mean stress in woodlands and grasslands across the biome, under medium and especially high 2070 climate scenarios. Spatially, predicted ANN stress is more moderate on the NSW north-west slopes, and tends to increase southwards and inland. However, interpretation of these results should be tempered by the fact that ANN predicted moderate stress even for the current climate, suggesting a poor ANN classification of environments of these structural types. Indeed, c. 34% of current TGE environments are predicted by ANN to favour other structural types, particularly Callitris, Casuarina or Eucalyptus forest or chenopod shrublands along the inland and eastern margins. Under medium 2070 scenarios, environments favouring eucalypt woodland persist predominantly on the NSW north-west slopes and along narrow margins of the Great Divide, with small expansions in coastal areas such as the Gippsland Plain. These patterns are accentuated under 2070 high scenarios. Where environments no longer favour eucalypt woodlands, environments within the current TGE biome are predicted to become increasingly favourable for chenopod shrublands from inland edges (and even parts of Tasmania where this structural type does not currently exist), with patchy areas favouring Casuarina forests. A number of other structural formations are potentially favoured in more mesic areas. Notably, the extent of environments favouring grasslands is predicted to increase. This is despite very high predicted ANN stress in current grasslands, reflecting a changing spatial location of these environments.GDM produced estimates of stress based on expected compositional dissimilarity under the present vs future climate scenarios for each 1 km2 grid cell. These showed similar patterns in grasslands and woodlands, with high stress or compositional dissimilarity predicted for reptiles, snails and plants, but low to moderate stress for birds, mammals and frogs. Lower estimates for the latter possibly reflect the wider distributions of species in these groups, but we emphasize that models did not account for changing habitat characteristics such as might be experienced with high levels of plant compositional dissimilarity. For nearly all groups, predicted stress is at the lower end of the range in Tasmania. For mammals, reptiles, snails and plants there is a tendency for lower stress in central-west Victoria and higher stress in central NSW, and sometimes South Australia. The NSW north-west slopes, which appear least stressed in the context of woodland structural

[发布日期] 2010-06-30 [发布机构] CSIRO

[学科分类] 地球科学(综合)