Mass flow controllers were used to deliver a gas mixture of N2, O2, and CO2 to gas membrane contactors (Liqui-Cel Part No. G501), which in turn delivered the gas mixture to the seawater to maintain the desired CO2 and O2 concentrations in each treatment replicate (Bockmon et al. 2013, Biogeosciences).
A pH sensor (Honeywell Durafet III) and Aanderaa oxygen optode were used to continuously record from one replicate per treatment, per day. Along with daily discrete samples measured for pH and [O2], this sensor monitoring confirmed the stability of conditions, with the average offset of replicates to be 0.01 pH units and 3µmol O2 kg-1 (Frieder et al. 2014, GCB). Discrete pH measurements were determined spectrophotometrically and the discrete [O2] values were determined using a Winkler Titration method (as described in Frieder et al. 2014, GCB).
Discrete samples were taken at the beginning and end of each experiment for the determination of total alkalinity and salinity. Total alkalinity measurements were determined using an open-cell, potentiometric titration with an accuracy of +- 2 µmol kg-1 (Dickson et al., 2007). Salinity was calculated from density measured at 20°C on a Density Meter (Mettler Toledo DE45) with an accuracy of 0.05 salinity units. HOBO Pendant temperature data loggers with an accuracy of +-0.5°C continuously measured temperature every 5 min in all replicates.