The biological sink of atmospheric H2 is more sensitive to spatial variation of microbial diversity than N2O and CO2 emissions in an agroecosystem

The relationship between soil microbial diversity and agroecosystem functioning is controversial due to the elevated diversity level and the functional redundancy of microorganisms. A field trial was established to test the hypothesis that enhanced crop diversity with the integration of winter cover crops (WCC) in a conventional maize-soy rotation promotes microbial diversity and the biological sink of H2 in soil, while reducing N2O emissions to the atmosphere. Vicia villosa (hairy vetch), Avena sativa (oat), and Raphanus sativus (Daikon radish) were cultivated alone or in combinations and flux measurements were performed throughout two subsequent growing seasons. Soil acted as a net sink for H2 and as a net source for CO2 and N2O. CO2 flux was the most sensitive to WCC whereas a significant spatial variation was observed for H2 flux with soil uptake rates observed in the most productive area two-fold greater than the baseline level. Sequencing and quantification of taxonomic and functional genes were integrated to explain variation in trace gas fluxes with compositional changes in soil microbial communities. Fungal communities were the most sensitive to WCC, but neither community abundance nor beta diversity were found to be indicative of fluxes. The alpha diversity of taxonomic and functional genes, expressed as the number of effective species, was integrated into composite variables extracted from multivariate analyses. Only the composite variable computed with the inverse Simpson9s concentration index displayed a reproducible pattern throughout both growing seasons, with functional genes and bacterial 16S rRNA gene defining the two most contrasting gradients. The composite variable was decoupled from WCC treatment and explained 19-20% spatial variation of H2 fluxes. Sensitivity of the trace gas exchange process to soil properties at the local scale was inconsistent among H2, N2O and CO2, with the former being the most related to microbial diversity distribution pattern.