Bioclimatic Predictors of Agro-Ecological Classes and Projected Shifts Under Climate Change.

Land use classification and species distribution studies often rely on biophysical variables hypothesized to be key drivers or determinants of land use/cover. Weak relationships, however, can occur between delineated land use classification and actual land use. In contrast, classification based on land use/cover that has emerged as a consequence of determinants may be advantageous as the actual land use can then be used for selecting important driving variables. The National Agricultural Statistics Service (NASS) cropland data-layer of actual land use/cover was used to classify the REACCH (Regional Approaches to Climate Change) study region, in the inland Pacific Northwest, into four major agro-ecological classes (AECs): (1) annual cropping (limited annual fallow); (2) annual crop-fallow transition (e.g. 3-yr rotations with fallow every 3rd year); (3) grain-fallow, 2-yr; and (4) irrigated (Huggins et al., 2011). Climatic parameters and bioclimatic variables (30) important in Mediterranean climates were calculated for the REACCH region at a 4 by 4 km scale using 30 year (1981-2010) precipitation and temperature data. Bioclimatic variables useful for classifying the four AECs will be identified using discriminant analysis. Identified bioclimatic variables will then be combined with future climate change scenarios to predict future shifts in the geographic distribution of AECs in the REACCH region.