Differential responses of soil microbial biomass, diversity and interactions to land use intensity at a territorial scale

Abstract

Impact of land use intensification on soil microbial communities across a territory remains poorly documented. Yet, it has to be deciphered to validate the results obtained at local and global scales by integrating the variations of environmental conditions and agricultural systems at a territorial scale. We investigated the impact of different land uses (from forest to agricultural systems) and associated soil management practices on soil molecular microbial biomass and diversity across a territory of 3300 km2 in Burgundy (France). Microbial biomass and diversity were determined by quantifying and high-throughput sequencing of soil DNA from 300 soils, respectively. Geostatistics were applied to map the soil macro-ecological patterns and variance partitioning analysis was used to rank the influence of soil physicochemical characteristics, land uses and associated practices on soil microbial communities. Geographical patterns differed between microbial biomass and diversity, emphasizing that distinct environmental drivers shaped these parameters. Soil microbial biomass was mainly driven by the soil organic carbon content and was significantly altered by agricultural land uses, with a loss of about 71 % from natural to agricultural ecosystems. The best predictors of bacterial and fungal richness were soil texture and pH, respectively. Microbial diversity was less sensitive than microbial biomass to land use intensification, and fungal richness appeared more impacted than bacteria. Co-occurrence network analysis of the interactions among microbial communities showed a decline of about 95 % of network complexity with land use intensification, which counterbalanced the weak response of microbial diversity. Grouping of the 147 cropland plots in four clusters according to their agricultural practices confirmed that microbial parameters exhibited different responses to soil management intensification, especially soil tillage and crop protection. Our results altogether allow evaluating the different levels of microbial parameters' vulnerability to land use intensity at a territorial scale.