Net primary production (NPP) is as a dynamic ecological indicator for evaluating land degradation and desertification. Therefore, identifying the factors affecting the changes in NPP can help better manage ecosystems, especially in arid regions. This study was conducted to evaluate the changes in the spatial-temporal patterns of NPP in the rangelands of northeastern Iran using the geographical detector model (GDM). NPP values were estimated from 2004 to 2023 using the CASA model. The spatial-temporal changes in NPP were evaluated using the Theil-Sen estimator, Mann-Kendall test, coefficient of variation, and Hurst index. The impacts of driving forces on the spatial distribution of NPP were investigated using 18 indicators and the GDM. The results showed that the estimated NPP values had strong correlation (r = 0.8) and good accuracy compared with MODIS NPP products. The 20-year average NPP in the growing season (March-September) varied from 7.74 to 192.36 (gC m???? 2 y???? 1). Annual changes in NPP had a downward trend in most of the study area based on the Theil-Sen estimator; however, the Mann-Kendall test indicated that these changes were not significant. The GDM model indicated that soil salinity, vegetation density, and soil moisture were the most important natural factors and village density and livestock density were the most important anthropogenic factors affecting NPP. According to the GDM, the interaction of soil moisture with vegetation density and soil salinity had the greatest impact on the spatial distribution of NPP. These findings demonstrated that soil salinity is a major concern, affecting rangeland productivity.

Keywords