On modelling the relationship between vegetation greenness and water balance and land use change

Here we sought a biologically meaningful, climate variable that captures water-energy availability and is suitable for high resolution (250 m × 250 m) modelling of the fraction of photosynthetically active radiation intercepted by the sunlit canopy (F V ) derived from a 10-year (July 2000 – June 2010) time series of Moderate Resolution Imaging Spectroradiometer (MODIS) Normalized difference vegetation index (NDVI) satellite imagery for Australia. The long-term mean annual evaporation deficit, and mean annual water availability indices all yielded strong linear relationships with mean F V ( $$\overline{{{\boldsymbol{F}}}_{{\boldsymbol{V}}}}$$ F V ¯ , %). We hypothesised whether some of the scatter about the relationships was related to land-use changes that have disrupted the vegetation-climate-soil equilibrium. Using continental-scale spatial data layers of protected area status and vegetation condition classes we repeated our analyses with restricted datasets. $$\overline{{{\boldsymbol{F}}}_{{\boldsymbol{V}}}}$$ F V ¯ of intact native vegetation within protected areas was greater than all modified vegetation classes. There was a consistent decline in the slopes of the regression relationships with increasing intensity of woody vegetation clearing and livestock grazing. Where native vegetation has been transformed by land use there was a 25% reduction in predicted $$\overline{{{\boldsymbol{F}}}_{{\boldsymbol{V}}}}$$ F V ¯ .