Water samples were collected monthly or bimonthly (September 2019 to October 2020) at the Martha's Vineyard Coastal Observatory (MVCO) from about 2 meters below the surface using bucket casts or a CTD deployed from R/V Tioga. Samples were transferred to a carboy in a cooler for transport back to the laboratory. Water was processed within 2 hours of collection. One liter of whole seawater was collected onto 47-millimeter (mm) 0.2-micrometer (µm) Millpore Isopore filters and stored at -20 degrees Celsius (°C) for DNA extraction and amplicon sequencing. Nucleic acids were extracted from half of a 47mm filter that was cut into small pieces using scissors and cleaned with alcohol in between samples. The lysis method followed a hot detergent plus bead disruption protocol (Gast et al., 2004). Filter pieces were placed in a sterile 2-milliliter (ml) microfuge tube and 2 x lysis buffer was added, along with approximately 50 microliters (µl) of beads. The tubes were vortexed for 30 seconds and then placed at 65°C for 5 minutes, followed by two cycles of vortex and heat incubation. Sodium chloride and CTAB were added, and the samples were incubated at 70°C for 10 minutes. Following extraction with an equal volume of chloroform, the aqueous phase was removed and precipitated overnight at -20°C with 0.6 volume of isopropanol. The recovered DNA pellet was resuspended in 20 µl of sterile water.

Amplification of the V4 region of the 18S ribosomal RNA gene was accomplished using the primers 574V4F (5’ [TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG]CGGTAAYTCCAGCTCYV) and 1132V4R (5’ [GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG]CCGTCAATTHCTTYAART), described in Hugerth et al. (2014) and modified to include 5’ adapter sequences (indicated by square brackets). PCR reactions were performed in triplicate for each sample using 1 µl template DNA, 1.25 units AmpliTaq DNA polymerase, 2 millimoles (mM) MgCl2, 2 µl 2.5 µM dNTPs, and 2.5 µl 10X reaction buffer (25 µl total volume) with the conditions: 95°C for 5 minutes; 35 cycles of 95°C for 30 seconds, 58°C for 30 seconds, 72°C for 90 seconds; 72°C for 5 minutes; 4°C hold. No template negative controls were included. Each sample reaction was examined to confirm the correct product size of approximately 500 base pairs (bp). Triplicate reactions were pooled and sent to the RI-INBRE Molecular Informatics Core for library preparation and Illumina MiSeq (250 bp paired end; 500 cycle kit V2) sequencing.

Amplicon data collected in this work was combined with other MVCO time series amplicon data to examine the temporal variation in Syndiniales types. The prior sequencing effort covered samples collected between 2013-2019 and 2020-2021, and used the same primer set reported here. Trimmed and demultiplexed raw reads were imported into qiime2 and the forward reads were analyzed for high quality (Q value 30 over 90% of read), removal of chimeric sequences, identification of amplicon sequence variants (ASVs) at 100% similarity, removal of ASVs that occurred in fewer than 2 samples, and assignment of taxonomy using PR2 database.