Shijiuyang wetland, a large scale water treatment system, has been put into operation since July 2, 2008, aiming to improve the quality of micropolluted drinking source water in Jiaxing City. The wetland covers 110 ha and adopted an innovative root channel purification technology promoting the rapid formation of soil macropores at its initial operation stage. In order to evaluate the development of specific soil microbes, the rhizospheric bacterial community diversity was examined in two zones of Shijiuyang wetland. The diversity was compared with that of a natural wetland-Baiyangdian Lake, which had a similar ecological structure.
The diversity of bacterial community in soil samples was characterized and compared by denaturing gradient gel electrophoreses (DGGE) analysis of polymerase chain reaction (PCR)-amplified bacterial 16S rRNA genes. Each band detected by Quantity One software was taken as an "operational taxonomic unit" (OTU). Genotypic diversity indexes based on a comparison of combined PCR-DGGE fingerprint patterns were used to assess the bacterial community diversity. In addition, soil property and water quality were analyzed in both wetlands. Multivariate statistical analyses were conducted to compare the similarities between the bacterial communities in both wetlands and to examine the relationships between soil characteristics and bacterial community diversity.
It was found that the rhizospheric bacterial community of Shijiuyang wetland at first one-year operation stage had a close genotypic richness (S) in comparison with that of Baiyangdian Lake natural wetland. This reflected a rapid increase of bacteria species in Shijiuyang wetland. Nonetheless, the bacterial evenness index (J) and Shannon index (H') in Shijiuyang wetland were significantly lower than those of natural wetland. It was suggested that Shijiuyang wetland would take a long time to form a stable and balanced bacterial community structure pattern. One-way analysis of variance (ANOVA) showed that both wetlands had significant difference in Shannon index (P=0.003) and evenness index (P=0.001), but not in genotypic diversity. Dendrogram was derived according to Ward's hierarchical cluster analysis to classify the sampling stations in both wetlands. The dendrogram based on soil properties indicated that the sampling stations from the same wetland were closer to each other, and the stations from different wetlands were separated distinctly. In spite of that, the dendrogram based on soil bacterial genotypic diversity indexes showed a different grouping pattern. Sampling stations with relatively better water quality in Baiyangdian Lake wetland exhibited the highest level of bacterial genotypic diversity indexes, which was separated as a single branch, then followed by the other three stations in both wetlands. Baiyangdian Lake is severely contaminated as opposed to Shijiuyang wetland. Even so, the bacterial genotypic diversity showed a considerable degree of spatial variation in Baiyangdian Lake. Therefore the bacterial community diversity might be subjected to the coupled effects of pollution gradients and wetland succession period. The complex determining mechanisms need further studies.
Finally, the relationships between soil properties and bacterial community diversities were further examined by multivariate statistical analyses. It was suggested that the genotypic diversity indexes of rhizospheric bacteria would increase with the decrease of pollutant concentrations in Baiyangdian Lake. Regression analyses revealed that, of all soil variables examined, the soil bacterial diversities were significantly related to soil total nitrogen (STN). The corresponding linear expressions were as follows, S=22.98+3.11×STN (P=0.0993), J=0.78+0.066×STN (P=0.0455), H'=2.44+0.31×STN (P=0.0334). However there was no clear relationship between the bacterial diversities and other soil characteristics.