Rice to vegetables: short- versus long-term impact of land-use change on the indigenous soil microbial community.

Land-use change is known to have a significant effect on the indigenous soil microbial community, but it is unknown if there are any general trends regarding how this effect varies over time. Here, we describe a comparative analysis of microbial communities from three adjacent agricultural fields: one-century-old paddy field (OP) and two vegetable fields (new vegetable field (NV) and old vegetable field (OV)) that were established on traditional paddy fields 10 and 100 years ago, respectively. Soil chemical and physical analysis showed that both vegetable fields were more nutrient rich than the paddy field in terms of organic C, total N, total P, and available K. The vegetable fields possessed relatively higher abundance of culturable bacteria, fungi, and specific groups of bacteria (Actinomyces, nitrifying bacteria, and cellulose-decomposing bacteria) but lower levels of microbial biomass C and N. Notably, the decrease of biomass was further confirmed by analysis of seven additional soils in chronosequence sampled from the same area. Next we examined the metabolic diversity of the microbial community using the EcoPlate(TM) system from Biolog Inc. (Hayward, CA, USA). The utilization patterns of 31 unique C substrates (i.e., community-level physiological profile) showed that microorganisms in vegetable soil and paddy soil prefer to use different C substrates (polymeric compounds for NV and OV soils, phenolic acids for OP soil). Principal component analysis and the average well color development data showed that the NV is metabolically more distinct from the OV and OP. The effect was likely attributable to the elevated soil pH in NV soil. Furthermore, we assessed the diversity of soil bacterial populations using the cultivation-independent technology of amplified ribosomal DNA restriction analysis (ARDRA). Results showed that levels of bacterial diversity in OP and NV soils were similar (Shannon's diversity index H = 4.83 and 4.79, respectively), whereas bacteria in OV soil have the lowest score of diversity (H = 3.48). The low level of bacterial diversity in OV soil was supported by sequencing of ten randomly selected 16S rDNA clones from each of the three rDNA libraries. Phylogenetic analysis showed that all the ten OV clones belonged to Proteobacteria with eight in the gamma-subdivision and two in the alpha-subdivision. In contrast, the ten clones from NV and OP soils were classified into four and eight bacterial classes or unclassified groups, respectively. Taken together, our data suggest that land-use change from rice to vegetables resulted in a decrease of bacterial diversity and soil biomass despite an increase in the abundance of culturable microorganisms and, moreover, the decrease of bacterial diversity occurred during long-term rather than short-term vegetable cultivation.