Factors Controlling Microbial Community Structure in An Agroecosystem.

A replicated agroecological experiment was run in the field for three years in order to identify the major factors controlling soil microbial biomass and community structure. Tomato was the main crop and nine treatments were included: bare soil, black polyethylene, white polyethylene, hairy vetch cover crop, hairy vetch above-ground biomass, hairy vetch below-ground biomass, rye cover crop, rye above-ground biomass, and rye below-ground biomass. The following hypotheses were tested: (1) Temperature and moisture differences between polyethylene covered and cover cropped treatments are partly responsible for treatment effects on soil microbial community composition, and (2) Different species of cover crops have unique root and shoot effects on soil microbial community composition.  Soil factors measured included pH, texture, temperature, and moisture. Tomato rhizosphere and bulk soil were sampled and analyzed for phospholipid fatty acids (PLFA). Soil DNA was extracted from a subset of samples and analyzed by T-RFLP and pyrosequencing. Cover cropping increased the absolute amount of all microbial groups, but decreased the proportion of Gram-positive bacteria, which we attribute to increased readily available carbon under cover-cropped treatments. Vetch shoots increased the amount and proportion of Gram-negative bacteria, fungi, and arbuscular mycorrhizal fungi in the rhizosphere of tomato plants. T-RFLP revealed differences in phylum-level responses to season, cover crop, and rhizosphere. Pyrosequencing of the 16s rRNA gene indicated positive effects of vetch on the proportion of Bacteroidetes, Chloroflexi, Nitrospira, and Planctomycetes, while negative vetch effects were observed on the proportion of Actinobacteria, Gemmatimonadetes, and Firmicutes. The lower proportion of Firmicutes and Actinobacteria under vetch was consistent with the PLFA results. The imposed treatments were much more significant than soil temperature, moisture, pH, and texture in controlling microbial biomass and community structure. Our first hypothesis was rejected while the second hypothesis was confirmed.