Effects of rainfall on the spatial variability of gross primary productivity and evapotranspiration across USDA Long-Term Agroecosystem Research Sites

Quantifying how patterns of gross primary productivity (GPP) and evapotranspiration (ET) vary along precipitation gradients is important for carbon and water cycle studies. The amount of rainfall determines the change in GPP and ET in time and space. We investigated the effects of rainfall on GPP and ET across 19 United States Department of Agriculture (USDA) Long-Term Agroecosystem Research (LTAR) sites by using 19 years of Moderate Resolution Imaging Spectroradiometer (MODIS) retrieved GPP and ET data. We also explored the variability in GPP and ET across the longitudinal gradient of study sites. A strong relationship of spatial variability in annual sums of rainfall with annual sums of GPP (R2= 0.83) and ET (R2=0.76) were found across LTAR sites. Results indicated that rainfall was one of the major drivers for spatial variability of GPP and ET. Annual sums of GPP and ET were also strongly coupled (R2=0.93). Longitude demonstrated a strong effect on annual sums of GPP and ET with R2 of 0.73 and 0.74, respectively whereas there was no effect of latitudes on GPP and ET. Annual rainfall based empirical models estimated the annual sums of GPP and ET for different LTAR sites across time. The developed models showed good agreement (slope= 0.85-0.99 and R2= 0.83- 0.92) with MODIS derived GPP and ET. The analyses of rainfall controls on ecosystem carbon and water fluxes will provide a better understanding of how rainfall gradients influence carbon and water budgets at the regional scale.