engleski

Two-way insight into diatom diversity of the Northern Adriatic. HighThroughput Sequencing in marine phytoplankton research

For the European Marine Strategy Frame Directive (MSFD), marine microphytoplankton is monitored for community structure, overall abundance, toxic and invasive species. Identification of diatom species (the major component in Adriatic phytoplankton communities) based on morphology (light microscopy) requires expert taxonomic knowledge and is limited by the resolution of light microscopy. As demonstrated in freshwater systems, metabarcoding and High-Throughput Sequencing (HTS) are promising tools that might increase taxonomic resolution (Vasselon & al. 2017, Zimmermann & al. 2015). Here we present diatom diversity data of the Northern Adriatic (Mediterranean Sea) assessed by HTS of ribosomal 18S and plastid rbcL gene and compared with light microscopy results. Sampling stations represent a transect from the coastal area to the open waters that are included in monitoring programmes of monthly diatom diversity and abundance assessments (Godrijan & al. 2013). Two DNA barcodes (rbcL and 18S V4) were amplified for samples at every station and HTS was performed using the PGM Ion Torrent technology. Around 14% of the known diatom species in the study area are represented in sequence reference databases (rbcL and 18S V4) (Rimet & al. 2016, Quast & al. 2012). In the studied samples, around 35% of the species observed by light microscopy (LM) are represented in sequence reference databases. Around 30% of the species identified by LM were not detected by HTS. This could be the result of incompleteness of the reference database but might also be an effect of DNA extraction and PCR amplification biases. 21% of the species uncovered by HTS (which amounts up to 52% of the species uncovered by LM) are species not identifiable by LM (cryptic species). For unrestrained OTU formation, HTS uncovered up to 3 times more potential species (OTUs) (within diatoms) than with LM. The known diatom species for the region are not well represented in sequence reference databases (around 14%). Even though the tested samples showed better coverages, our results indicate a negative effect of this low reference database coverage on the alpha diversity (at species level) recovered by HTS. However, unrestrained OTU formation demonstrated that HTS does allow the detection and monitoring of significantly larger diatom diversity if compared to LM. The detection of large numbers of cryptic species additionally demonstrates the potential of HTS for the monitoring of invasive and/or cryptic species. With the correlation between observed cell numbers and read numbers still under discussion, it appears safe to assume that some species might be over- or underrepresented in the HTS results. Future research will have to demonstrate, whether HTS will be helpful in the early detection of low abundance (e.g. invasive) species.

phytoplankton ; Adriatic Sea ; diatoms ; metabarcoding

nije evidentirano