Large-scale patterns and drivers of soil organic nitrogen depolymerization

Abstract

Depolymerization of macromolecular soil organic nitrogen (N) is the first step in converting high-molecular weight organic N (e.g., proteins) into inorganic N, making it a crucial driver of soil N availability and plant productivity. However, the patterns, along with the influencing factors of protein depolymerization in soils across broad geographical expanses have remained unknown. In this study, soil gross protein depolymerization rates (GPDR) across continental scale were analyzed, drawing on 275 observations sourced from 27 published papers. Within the compiled dataset, soil GPDR exhibited a parabolic relationship with latitude (ranging from 30°N to 70°N), while showing a positive linear relationship with elevation (ranging from 0 to 2000 m a.s.l.). The grand mean of soil GPDR was 92.6 ± 11.1 mg N∙kg⁻¹∙day⁻¹, with the highest rates observed in grasslands followed by forests, cropland, and other ecosystems. Soil GPDR was significantly associated with multiple factors. These included soil microbial biomass carbon and N, soil total and inorganic N, mean annual precipitation and soil water content. However, no significant associations were found with mean annual temperature or soil texture. Soil microbial biomass was identified as the most important factor regulating soil GPDR, whereas soil total N and climatic factors indirectly influenced soil GPDR likely via their effects on microbial biomass. Moreover, soil GPDR was the most important factor controlling the content of soil NH₄⁺-N, confirming the important role of protein depolymerization in regulating soil N availability. This study provides a comprehensive understanding of controls of soil N depolymerization by combining climatic, edaphic, and soil microbial factors cross continental scale. These findings emphasize the significance of considering the role of microbial biomass in regulating protein depolymerization, which could provide insights to improve global soil N cycling models.