Discrimating the biophysical signal from human - induced effects on long - tern primary production dynamics. The case of Patagonia.

Abstract

The temporal trend of aboveground net primary production (ANPP) is frequently used to estimate the human effect on ecosystems. In water-limited ecosystems, like most grazing areas in the world, human effects act upon ANPP in combination with environmental variations. Our main objective was to quantify long-term (1981-2012) changes of ANPP and discriminate the causes of these changes between environmental and human at a subcontinental scale, across vast areas of Patagonia. We estimated ANPP through a radiative model based on remote sensing data. Then, we evaluated the relation between ANPP and environmental interannual variations of two hierarchically related factors: ENSO, through the Southern Oscillation Index, SOI, and precipitation. We described the human effect through the shape of the temporal trends of the residuals of the environmental model (RESTREND) and quantified human relative impact through the RESTREND: ANPP trend ratio. ANPP interannual variation was significantly explained by ENSO (through SOI) and precipitation in 65% of the study area. The Southern Oscillation Index had a positive association with annual precipitation. The association between ANPP and annual precipitation was positive. RESTREND analysis was statistically significant in 92% of the area where the tested environmental model worked, representing 60% of the study area, and it was mostly negative. However, its magnitude, revealed through the RESTREND: ANPP trend ratio, was relatively mild. Our analysis revealed that most of ANPP trends were associated with climate and that even when human density is low, its incidence seems to be mainly negative.