Hawaiian Ocean Time-series Flow Cytometry, November 2005 - December 2021

Flow cytometry samples were collected according to HOT sampling protocols (https://hahana.soest.hawaii.edu/hot/protocols/protocols.html). Briefly, all samples were collected prior to daylight (02:00-04:00 local time) at 8 standard depths (5, 25, 45, 75, 100, 125, 150, 175 m). Aliquots (10-12 mL) of seawater were collected from rosette Bullister bottles into thrice-rinsed 15 ml polypropylene centrifuge tubes. 2 mL was then subsampled into cryovials containing 30 µL of 16% paraformaldehyde (0.24% final concentration). Samples were mixed well, allowed to fix, and flash frozen using liquid nitrogen. Samples were then stored at -80 °C and analyzed ashore. An InFlux flow cytometer (formerly Cytopeia Inc., Seattle, WA, now Becton Dickinson) equipped with a 488 nm (100 mW) laser was used to identify Prochlorococcus and Synechococcus using the following detectors: red fluorescence (proxy for chlorophyll a concentration per cell, 692/40 nm bandpass filter), orange fluorescence (proxy for phycoerythrin concentration per cell, 585/40 nm bandpass filter), green fluorescence (542/27 nm bandpass filter) and forward light scatter (488 nm blocking). Fluorescent beads (Ultra Rainbow Fluorescent Particles, 1.0-1.5 µm size. P/N URFP-10-5; Spherotech) were used for alignment and cross-comparison of optical properties, with 2 µl added to each sample. Prochlorococcus was gated as low forward scatter, high red fluorescence, then as above the detection limit for orange fluorescence. Synechococcus was gated as above noise in green-orange fluorescence space, and then high forward scatter and high red fluorescence. Forward scatter and red fluorescence were normalized using mean bead values in logarithm 10 space. Initial analysis of the population-specific optical properties revealed offsets in bead normalized red fluorescence for Prochlorococcus and Synechococcus that we attribute to the changes in instrument laser settings coincident with changes in the primary HOT flow cytometry technicians and their chosen instrumental settings in February 2009 (HOT-209) and February 2015 (HOT-269). The primary result of the settings changes was that the linear log10 relationship between beads, Prochlorococcus, and Synechococcus were not comparable before and after these dates, even though cell concentrations were consistent. As such, we have introduced a correction factor to shift bead normalized red fluorescence values for both Prochlorococcus and Synechococcus to maintain the same mean value across each of the three time periods. The correction factor is identical for all cruises and depths within each time period. Only bottles at 45 m were used for each time period to avoid any potential artifacts from surface, DCM, or deeper samples. A different corrective factor was used for each time period and for both Prochlorococcus and Synechococcus. The middle period (2009-2014) was set as zero. For the first period (2005-2008), 0.3021 and 0.3588 were subtracted from all log10 bead normalized red fluorescence values for Prochlorococcus and Synechococcus respectively. For the third period (2015-2021), 0.9138 and 0.8151 were added to all log10 values of Prochlorococcus and Synechococcus respectively. The resulting corrected values contained no discontinuity between the previously separated time periods. Uploaded values for FSC are bead normalized and for red fluorescence are both bead normalized and corrected.