Dumpster diving for diatom plastid 16S rRNA genes.

High throughput sequencing is improving the efficiency of monitoring diatoms, which inhabit and support aquatic ecosystems across the globe. In this study, we explored the potential of a standard V4 515F-806RB primer pair in recovering diatom plastid 16S rRNA sequences. We used PhytoREF to classify the 16S reads from our freshwater biofilm field sampling from three stream segments across two streams in south-eastern Australia and retrieved diatom community data from other, publicly deposited, Australian 16S amplicon datasets. When these diatom operational taxonomic units (OTUs) were traced using the default RDPII and NCBI databases, 68% were characterized as uncultured cyanobacteria. We analysed the 16S rRNA sequences from 72 stream biofilm samples, separated the chloroplast OTUs, and classified them using the PhytoREF database. After filtering the reads attributed to Bacillariophyta (relative abundance >1%), 71 diatom OTUs comprising more than 90% of the diatom reads in each stream biofilm sample were identified. Beta-diversity analyses demonstrated significantly different diatom assemblages and discrimination among river segments. To further test the approach, the diatom OTUs from our biofilm sampling were used as reference sequences to identify diatom reads from other Australian 16S rRNA datasets in the NCBI-SRA database. Across the three selected public datasets, 67 of our 71 diatom OTUs were detected in other Australian ecosystems. Our results show that diatom plastid 16S rRNA genes are readily amplified with existing 515F-806RB primer sets. Therefore, the volume of existing 16S rRNA amplicon datasets initially generated for microbial community profiling can also be used to detect, characterize, and map diatom distribution to inform phylogeny and ecological health assessments, and can be extended into a range of ecological and industrial applications. To our knowledge, this study represents the first attempt to classify freshwater samples using this approach and the first application of PhytoREF in Australia.