Distinct environmental controls on above- and below-ground net primary productivity in Northern China’s grasslands

Abstract

Grasslands, which cover approximately 40 % of the global land surface, are crucial to the global carbon cycle due to their substantial carbon storage capacity and sensitivity to environmental changes. Net primary productivity (NPP) serves as a critical indicator of ecological function, representing the net carbon input to terrestrial ecosystems. Accurately quantifying and understanding grassland NPP, particularly below-ground net primary productivity (BNPP), and its responses to environmental changes are essential for assessing carbon sequestration potential and predicting future dynamics under varying environmental conditions. However, this remains challenging at large spatial scales. In our study, we integrated observations of above-ground net primary productivity (ANPP) with satellite-derived total net primary productivity (MODIS NPP) products across 508 individual locations in the diverse grasslands of Northern China. Plot-level BNPP was derived by subtracting field-measured ANPP from MODIS NPP and has been rigorously validated using field-measured BNPP from 68 ground sites. We then explored how climate, soil characteristics, topography, and grassland types influence ANPP and BNPP using machine learning models. Our results indicate that ANPP and BNPP averaged 154 g m−2 and 273 g m−2, respectively, exhibiting significant variations across different grassland types. Furthermore, national mapping revealed regional disparities, with higher ANPP and BNPP values in the northeast compared to the southwest of Northern China. Climate was identified as the predominant driver of ANPP, explaining 41 % of the variance, followed by topography (26 %) and soil conditions (25 %). Conversely, topography had the most substantial impact on BNPP, accounting for 44 % of the explained variance, with climate and soil each contributing 25 %. These findings underscore the vulnerability of ANPP to climate change impacts such as droughts and temperature increases, and highlight how landscape features, such as elevation, slope, and aspect significantly influence BNPP, with crucial implications for soil carbon inputs and sequestration. Overall, this comprehensive study advances our understanding of the spatial patterns and environmental determinants of ANPP and BNPP in Northern China’s grasslands. It offers valuable insights for developing sustainable management practices and for the preservation and effective management of these crucial ecosystems, especially in the context of ongoing environmental changes.

Keywords

• NPP
• Remote sensing
• Grasslands
• Machine-learning model
• Mapping
• Northern China