Peer Review #1 of "Co-occurrence patterns of litter decomposing communities in mangroves indicate a robust community resistant to disturbances (v0.2)"

Abstract

Background. Mangroves are important coastal ecosystems known for high photosynthetic productivity and ability to support marine food chains through supply of dissolved carbon or particular organic matter. Most of the carbon found in mangroves is produced by its vegetation and is decomposed in root associated sediment. This process involves a tight interaction between microbial populations, litter chemical composition, and environmental parameters. Here, we study the complex interactions found during litter decomposition in mangroves by applying network analysis to metagenomic data. Methods. Leaves of three species of mangrove trees typically found in the southeast of Brazil (Rhizophora mangle, Laguncularia racemosa, and Avicennia schaueriana) were collected in separate litter bags and left on three different mangroves for 60 days. These leaves were subsequently used for metagenome sequencing using Ion Torrent technology. Sequences were annotated in MG-RAST and used for network construction using MENAp. Results. The most common phyla were Proteobacteria (classes Gamma and Alphaproteobacteria) followed by Firmicutes (Clostridia and Bacilli). The most abundant protein clusters were associated with the metabolism of carbohydrates, amino acids, and proteins. Non-metric multidimensional scaling of the metagenomic data indicated that substrate (i.e., tree species) did not significantly select for a specific community. Both networks exhibited scale-free characteristics and small world structure due to the low mean shortest path length and high average clustering coefficient. These networks also had a low number of hub nodes most of which were module hubs. Discussion. This study demonstrates that under different environmental pressures (i.e., plant species or mangrove location) the microbial community associated with the decaying material forms a robust and stable network.