Functional composition of sediment microbial communities varies between floodplain wetlands with differing restoration histories

Background/Question/Methods

The widespread degradation of wetlands due to land use change can impair critical ecosystem services, and has prompted considerable restoration effort in recent decades. Many wetland ecosystem services are supported by microbial communities in sediments. Thus, understanding the effects of wetland restoration on microbial communities and their function is an important aspect of evaluating restoration success. To determine how microbial functional composition varies between wetlands with differing restoration histories (i.e., hydrological management histories), we compared communities in a degraded wetland, a restored wetland and a natural reference wetland within the Shiawassee River floodplain near Saginaw, MI, USA. Genomic DNA was extracted from sediment collected at three replicate plots in each wetland, and shotgun metagenomic libraries were created using Illumina HiSeq 2500 high-throughput sequencing. The metagenomes were uploaded to the Metagenomics Rapid Annotation using Subsystem Technology (MG-RAST) server for functional gene annotation. Abundances of reads assigned to functional classifications were normalized in MG-RAST and used for comparison of microbial functional composition between wetlands.

Results/Conclusions

The overall functional composition of microbial communities in each wetland was distinct (PerMANOVA; Pseudo-F = 23.60; P = 0.004). Additionally, the abundance of microbial functional genes that mediate ecologically important processes varied significantly between wetlands. For example, the normalized abundance of sequencing reads associated with denitrification was greater in the restored wetland than the degraded wetland (one-way ANOVA; F = 11.60; P = 0.007), but not different from that of the natural reference wetland (F = 11.60; P = 0.087). In contrast, the abundance of methanogenesis-associated reads in the restored wetland was greater than in the degraded wetland (F = 224.51; P < 0.001), but lower than in the natural wetland (F = 224.51; P = 0.001). Our results indicate that, as of the time of this study, restoration has not produced a microbial community with natural functional composition in the restored wetland. Slow recovery or alternative stable states may explain differences in microbial functional composition between restored and natural wetlands; our results suggest either explanation may vary with respect to specific functions.