Simon Ferrier (1), Thomas D Harwood (1), Kristen J Williams (1), Andrew Hoskins (1), Karel Mokany (1), Alex Bush (1), Chris Ware (1), Glenn Manion (2)
1 CSIRO Land and Water, PO Box 1600, ACT 2601, Australia
2 NSW Office of Environment and Heritage, University of New England 2351, Australia
Two broad analytical approaches have dominated efforts to assess potential impacts of climate change on the spatial distribution of biodiversity, and to thereby inform policy formulation, planning and management aimed at addressing these impacts. The first, and most widely applied, approach focuses on modelling shifts in the distribution of particular biological entities – mostly individual species, but also higher-level aggregations such as species assemblages or functional groups. The second approach focuses instead on analysing spatiotemporal patterns in climate alone – e.g. projections of climatic stability, velocity of climate change, and novel and disappearing climates, along with consideration of such patterns in adaptation strategies aimed at “conserving nature’s stage”. An arguable strength of this approach is its utility for addressing regions and/or components of biodiversity where the data and understanding required to explicitly model biological responses are lacking. However, analyses of climate alone do not recognise that the level of biological change expected to be associated with a given amount of change in a climatic attribute can vary markedly between biological groups, environments, and biogeographic regions. We here describe how these sources of variation can be accommodated by combining best-available location records for large numbers of species, with statistical modelling of spatial turnover in species composition, to scale (transform) multidimensional climate space, such that distances within this transformed space correlate as closely as possible with observed levels of biological turnover. We then use recent analyses underpinned by this approach to demonstrate how it can serve as a third major option for assessing and addressing climate-change impacts on biodiversity, effectively occupying the middle ground between explicit modelling of shifts in biological distributions, and analyses based on spatiotemporal patterns in climate alone.