Comparison of Microscopic and PCR Amplicon and Shotgun Metagenomic Approaches Applied to Marine Diatom Communities

Abstract

Metagenomic analysis is recognized as an effective tool to monitor marine organisms. However, to date, the accuracy of the phytoplankton composition determined based on metagenomics has not been fully verified. Among phytoplankton, diatoms can be identified by silica cell wall morphology and enumerated using light microscopy (LM), although identification using either scanning electron microscopy (SEM) or DNA sequencing is required in some species. We examined the species composition of diatoms during the spring bloom in Sendai Bay using metagenomic analysis (whole metagenome, ribosomal small subunits (SSUs) and large subunits from shotgun metagenomic data, and PCR amplicons of the SSU V4 region) and compared the results with those determined by using morphological identification by microscopic observation. For all the metagenomic analysis methods, the abundance estimated by SSU sequences was closest to that revealed by LM. However, an adequate reference database and sufficient numbers of metagenome sequences (>10⁷ reads) would improve estimation of species composition based on the SSU metagenomic data. Furthermore, seasonal dynamics based on microscopic observation and the SSU metagenomic approach were compared in the diatom communities at Sendai Bay from March 2012 to June 2014, revealing different tendencies in the annual changes both quantitatively and qualitatively. On the other hand, the metagenomic analysis showed changes in species composition in the genus Skeletonema, which could not be recognized by LM. These results indicate that the metagenomic approach can be a powerful tool for classification of species, but it is still ineffective for quantitation and precise identification of species. For phytoplankton monitoring, it is preferred to use morphological observation in combination with metagenomic analysis.