Soil texture determines the distribution of aggregate-associated carbon, nitrogen and phosphorous under two contrasting land use types in the Loess Plateau

Organic carbon (OC) and nutrient dynamics are closely related to soil texture, but how texture influences the distribution of OC and nutrients in aggregates in various land use types has not been examined. This knowledge gap precludes our mechanistic understanding of soil biogeochemical cycles at large spatial scales. Herein we compared the contents and stoichiometric ratios of OC and nutrients in both bulk soils and aggregates in cropland and woodland across a clay content gradient (7–31%) in the Loess Plateau. The soil metrics that were measured included the proportions of water-stable aggregates, and the contents of OC, nitrogen (N) and phosphorous (P) in bulk soils and each aggregate fraction. The stoichiometric ratios of carbon (C), N and P were calculated. The relationships between soil metrics and clay content were analyzed. We hypothesize that OC, N and P in aggregates increase with clay content, and these relationships are independent of land use types. In partial support of these hypotheses, the proportion of macroaggregates and the contents of OC, N and P in bulk soils and most aggregate fractions linearly increased with clay content. The slopes of these linear relationships were not affected by land use type. The C/N ratio were minimally affected, while the C/P and N/P ratios in both bulk soils and aggregates increased with clay content, and these relationships changed with land use type. Proportion of macroaggregates, contents of OC and N, and ratios of C/N, C/P and N/P were significantly higher in woodland than in cropland across or within sites. Furthermore, the distribution patterns of OC, N and P contents, and C/P and N/P ratios among aggregates varied with site and land use type. These suggested that soil texture determines the distribution of OC, N and P and their stoichiometric ratios within soil aggregates in the Loess Plateau of China, and most of these determining relationships were independent from land use types.