SEARCH

SEARCH BY CITATION

Keywords:

  • climate change;
  • climatic variability;
  • cryptic refugia;
  • landscape scale;
  • microclimate;
  • topoclimatic grids

Abstract

Ecologists are increasingly recognizing the conservation significance of microrefugia, but it is inherently difficult to locate these small patches with unusual climates, and hence they are also referred to as cryptic refugia. Here we introduce a new methodology to quantify and locate potential microrefugia using fine-scale topoclimatic grids that capture extreme conditions, stable climates, and distinct differences from the surrounding matrix. We collected hourly temperature data from 150 sites in a large (200 km by 300 km) and diverse region of New South Wales, Australia, for a total of 671 days over 2 years. Sites spanned a range of habitats including coastal dune shrublands, eucalypt forests, exposed woodland ridges, sheltered rainforest gullies, upland swamps, and lowland pastures. Climate grids were interpolated using a regional regression approach based on elevation, distance to coast, canopy cover, latitude, cold-air drainage, and topographical exposure to winds and radiation. We identified extreme temperatures on two separate climatic gradients: the 5th percentile of minimum temperatures and the 95th percentile of maximum temperatures. For each gradient, climatic stability was assessed on three different time scales (intra-seasonal, intra-annual and inter-annual). Differences from the matrix were assessed using a moving window with a 5 km radius. We averaged the Z-scores for these extreme, stable and isolated climates to identify potential locations of microrefugia. We found that our method successfully predicted the location of communities that were considered to occupy refugia, such as rainforests that have progressively contracted in distribution over the last 2.5 million years, and alpine grasslands that have contracted over the last 15 thousand years. However, the method was inherently sensitive to the gradient selected and other aspects of the modelling process. These uncertainties could be dealt with in a conservation planning context by repeating the methodology with various parameterizations and identifying areas that were consistently identified as microrefugia.