An integrated method to quantify human appropriation of net primary production in grasslands of the Qinghai-Tibet Plateau

Abstract

Human appropriation of net primary production (HANPP) is an important indicator for measuring anthropogenic influences on ecosystems. However, it is a challenge to accurately delineate the spatial heterogeneity of the HANPP, mainly due to great uncertainties in the simulation of potential net primary production (NPPp). Here, we established a novel method for simulating grassland NPPp by integrating ecoregion delineation and a random forest (RF) model and derived HANPP calculated by the NPPp minus NPP remaining in the ecosystem after harvest (NPPeco). In a case study of the Qinghai-Tibet Plateau (QTP), we evaluated the proposed method via validation and comparison with the NPPp and HANPP results in 2010 based on the Thornthwaite Memorial (TM) model and LPJ-GUESS model. The results indicated that the grassland NPPp values simulated by our method were closer to the observed field data in the two ecoregions, while the TM model generally overestimated NPPp. Moreover, the spatial distribution of the grassland HANPP based on the proposed method showed a high correlation coefficient (0.704) with grazing intensity, which was 0.669 higher than that of the TM model and 0.21 higher than that of the LPJ-GUESS model. The proposed method can provide a valuable basis for grassland sustainability assessment and disaggregating livestock data from administrative units into grid cells.