The relative importance of microclimate and land use to biodiversity

The extent of species' distributions, and their abundance within them, can be affected by many factors. At coarse resolutions and large spatial extents, climatic variables are often found to be important in determining the boundaries of species' ranges. However, several authors have postulated that variables such as land cover or species interactions are more important in determining where species will be found at finer resolutions within a landscape. Climatic variables are often not available at the fine resolutions necessary to test this hypothesis. In this study, I first show that topography is important in influencing the temperature that will be experienced at a particular location. Temperatures generally decrease with elevation (known as the lapse rate). However, extremes of temperature, such as absolute maxima and minima, can also be influenced by the angle and aspect of a slope. Whilst high resolution elevation data are occasionally included in climatic models of species' distributions, the effects of slope and aspect have so far been ignored. My second data chapter uses all three topographical variables (elevation, slope and aspect) to interpolate temperatures for three landscapes. These interpolated temperatures, along with land cover variables and an index of soil moisture, are included in exploratory analyses which suggest that microclimate is at least as important as land cover in determining spatial patterns of abundance for 20 species of ground beetle. Given this potential importance of microclimate in determining where species will be found at fine resolutions, I then show that different resolution data layers generate different predictions about the amount of a landscape that will remain suitable following climatic warming. Fine resolution layers predict a slower loss of thermally suitable habitat for a northerly-distributed species in a reserve located at its southern range margin. I therefore conclude that, in topographically diverse areas, where a wide range of temperature conditions can be encountered, climatic variables should be considered at spatial resolutions relevant to individuals of species whose distributions are modelled using the 'climate-envelope' approach rather than the coarse-resolution data that are currently used.