sample bottle, half fill, 5x inversion, fill remainder) was used to fill all subsample bottles 
detailed below. 
Water samples containing organisms and particles < 50 pm (10-16 L filtrate samples) 
were processed to generate the fractions required for the remaining analyses (< 50 pm for flow 
cytometry, 10-50 pm for all other methods). The subsample bottle for flow cytometry was filled 
first, and the remaining water was filtered on a 10 pm (pore diameter) Sterlitech polyester 
track etch (PETE) membrane filter. The retained particles were resuspended in filter-sterile sea 
water with a final concentration up to 16x times the original concentration (see Table A2). The 
concentrated sample was split into subsample bottles for analysis of the 10-50 pm size class (8 
subsamples, volume 25-350 ml). For most analytic methods, there was no further assessment 
of the size of organisms in the size-fractionated samples (i.e. all organisms contained in a given 
sample were considered to be within the relevant size class). However, the Satake Pulse 
Counter uses pulse strength to estimate organism size (see Appendix B for details), and 
microscopists used photomicrographs of 50 pm calibration beads (> 50 size class) and 
Sedgewick-Rafter grid widths (10-50 size class) as size references. 
After each trial, all sampling gear, sample carboys, and subsample bottles were cleaned 
in a dilute (100-200 ppm) or concentrated (2500 ppm) bleach solution (depending on 
equipment robustness) made using the ship's potable water supply to prevent cross 
contamination of living organisms between tests. After bleaching, all equipment was rinsed 
with MilliQ water three times; plankton nets were rinsed with potable water three times before 
being rinsed once with MilliQ water and hung to dry. Prior to re-use, all sample carboys and