Processing and analyses of 16S rRNA gene amplicons
The amplicons, originating from two different library designs, were processed as two independent data sets. The raw data were quality checked in FastQC (Andrews, 2010) and trimmed using USEARCH v9.0.1001 (Edgar, 2013). The reads were further processed and clustered into OTUs (operational taxonomic units) according to the workflow implementing USEARCH v9.0.1001 scripts. All putative contaminants found in negative controls were discarded and highly abundant, potentially symbiotic, OTUs were identified. Details on the read processing, decontamination and OTU filtering, including the accession links to the raw data, are provided in the SupplementaryInformation/SupplementaryText.