Slides were digitally imaged using a Zeiss Axiovert 200 M inverted compound microscope equipped for high-throughput epifluorescence microscopy with a motorized focus drive, stage, objective and filters. Digital images were acquired with a Zeiss AxioCam MRc black and white 8-bit CCD camera. SV samples (50 mL aliquots) were viewed at 630X magnification, and LV samples (450 mL aliquots) were viewed at 200X magnification. A minimum of 20 random positions were imaged for each slide, with each position consisting of three to four fluorescent channels: Chl a, DAPI, FITC (SV and LV samples) and phycoerythrin (SV samples only).  

Images were analyzed using ImagePro software to semi-automate the enumeration of eukaryotic cells larger than 1.5 um in length (Taylor et al., 2012). Whenever possible, 20 positions and >300 cells were counted for each slide. Each cell was manually identified and grouped into identifiable taxonomic groups. Autotrophic cells were identified by the presence of chlorophyll a (red autofluorescence under blue light excitation), generally clearly packaged in defined chloroplasts, and obvious heterotrophic cells with recently consumed prey were manually excluded from the autotroph classification. 

Biovolumes (BV; um3) were calculated from the length (L) and width (W) measurements of each cell using the geometric formula of a prolate sphere (BV = 0.524*LWH), assuming H = W. Biomass was calculated as carbon (C; pg cell-1) using the equations of Menden-Deuer and Lessard (2000): C = 0.288 BV^0.811for diatoms and C = 0.216 BV0.939 for non-diatoms.

BCO-DMO Processing Notes:
- size_class and taxon columns were transposed from columns into rows.