Molecular ecological network analyses revealing the effects of livestock grazing on soil microbial community in the Tibetan grassland

Abstract:

[Objective] To unveil effects of livestock grazing in Tibet on microbial interaction at the functional gene level. [Methods] We applied a recently developed network inference tool (Random Matrix Theory-based molecular ecological network) on GeoChip data related to carbon (C) and nitrogen (N) cycling with or without livestock grazing. [Results] C and N cycling gene networks in both control and grazing conditions had topological features of scale-free, small-world, modularity and hierarchy. Key genes in the grazing networks (hubs and connectors) differed substantially from those in the control. The grazing effects on soil microbial interactions were revealed by smaller, denser networks in the grazing samples, suggestive of environmental stress. In support of close linkages between aboveground plants and microbial community at this site, aboveground plant biomass was significantly (P=0.001) linked to grazing network topology. [Conclusion] Livestock grazing significantly altered microbial interaction at the functional gene level.