Analysis of microbial diversity of submerged biofilters in recirculating aquaculture system for grouper based on high-throughput DNA sequencing

Abstract: Nitrification biofilters were widely used to remove ammonia and other metabolic waste products in recirculating aquaculture systems, in which the biofilms played a crucial role in the biotransformation of contaminants. The functional dynamics of the microbial communities were not thoroughly characterized, which provided the impetus for efforts to characterize their species composition and activity. Microbial communities in various biofilters have been intensively studied by various molecular methods, such as polymerase chain reaction - denaturing gradient gel electrophoresis (PCR-DGGE), terminal restriction fragment length polymorphism (T-RFLP), sequencing of clones in DNA libraries and fluorescent in situ hybridization (FISH). However, these tools tend to underestimate the microbial diversity of the biofilter. So-called metagenomic studies use either "shotgun" or PCR-directed sequencing to characterize largely unbiased samples of genes from all the members of sampled communities. Compared to traditional methods, application of the metagenomics approach - in combination with next-generation DNA sequencing, which can generate millions of DNA sequence reads with an average length of over 400 bp - can provide insights of unprecedented depth into microbial community composition and diversity. Against this background, we applied polymerase chain reaction-mediated amplification of the 16S rRNA gene and high-throughput Illumina-MiSeq sequencing to seek insights into the microbial community composition of a series of submerged biofilters in commercial recirculating aquaculture systems for grouper (Epinehelus moara). Our results demonstrated the usefulness of the approach for elucidating bacterial community structure in the series of biofilter; we detected a mean of 27,737 usable DNA sequence reads and 737 operational taxonomic units (OTU) for each of the 3 submerged biofilters and 488 OTUs were shared among the series of submerged biofilters. Rarefaction curves indicated that further sequencing would not result in identification many more OTUs within each sample. Biofilter 3 had the highest microbial diversity than biofilter 1 and biofilter 2 based on the value of Chao1 index and Simpson index. Bacterial community compositions were characterized at the phylum and genus levels, respectively. Among the 14 most frequently observed phyla, the most dominant was Proteobacteria, representing about 50% of reads in the overall data set for a given filter. The other dominant phyla were Bacteroidetes (23.3%-35.4%), Verrucomicrobia (2.0%-8.8%), Chloroflexi (2.5%-3.5%), Planctomycetes (1.6%-6.3%), Actinobacteria (2.0%-2.7%) and Nitrospirae (0.2%-7.1%), which were similar with the previous study. At the genus level of taxonomic classification, more than eighty taxa were observed, including 2 genera which were important to the process of nitrification. The nitrite-oxidizing genus Nitrospira was abundant in biofilter 3 (7%) and less abundant in biofilter 2 (0.8%) and biofilter 1 (0.2%). The ammonia-oxidizing genus Nitrosomonas varied ranging from 0.1 % to 2.3%. These results suggested that Nitrosomonas and Nitrospira played important role in nitrification in the series of submerged biofilters. The UPGMAM dendrogram created by hierarchical cluster analysis of our results showed the biofilter1 and biofilter 2 closely clustered. The results showed that high-throughput DNA sequencing technology could provide unprecedented depth into microbial community and the findings in this work would data and theoretical basis for design and optimization of series of submerged biofilters in recirculating aquaculture system.