Gross Primary Production Six depths were sampled per CTD station ranging from near-surface to 125 m. Sample depths were based on downward fluorescence profiles and two of six samples always included both near-surface (approximately 5-10 m) and the depth of the chlorophyll maximum where applicable. Photosynthetic rates were determined using radioactive NaH14CO3. Incubations were conducted according to the method of Westwood et al. (2011). Cells were incubated for 1 hour at 21 light intensities ranging from 0 to 1200 µmol m-2 s-1 (CT Blue filter centred on 435 nm). Carbon uptake rates were corrected for in situ chlorophyll a (chl a) concentrations (µg L-1) measured using high performance liquid chromatography (HPLC, Wright et al. 2010), and for total dissolved inorganic carbon availability, analysed according to Dickson et al. (2007). Photosynthesis-irradiance (P-I) relationships were then plotted in R and the equation of Platt et al. (1980) used to fit curves to data using robust least squares non-linear regression. Photosynthetic parameters determined included light-saturated photosynthetic rate [Pmax, mg C (mg chl a)-1 h-1], initial slope of the light-limited section of the P-I curve [α, mg C (mg chl a)-1 h-1 (µmol m-2 s-1)-1], light intensity at which carbon-uptake became maximal (calculated as Pmax/ α = Ek, µmol m-2 s-1), intercept of the P-I curve with the carbon uptake axis [c, mg C (mg chl a)-1 h-1] , and the rate of photoinhibition where applicable [β, mg C (mg chl a)-1 h-1 (µmol m-2 s-1)-1]. Gross primary production rates were modelled using R. Depth interval profiles (1 m) of chl a from the surface to 200 m were constructed through the conversion of up-cast fluorometry data measured at each CTD station. For conversions, pooled fluorometry burst data from all sites and depths was linearly regressed against in situ chl a determined using HPLC. Gross daily depth-integrated water-column production was then calculated using chl a depth profiles, photosynthetic parameters (Pmax, α , β, see above), incoming climatological PAR, vertical light attenuation (Kd), and mixed layer depth. Climatological PAR was based on spatially averaged (49 pixels, approx. 2 degrees) 8 day composite Aqua MODIS data (level 3, 2004-2017) obtained for Julian day 34. Summed incoming light intensities throughout the day equated to mean total PAR provided by Aqua MODIS. Kd for each station was calculated through robust linear regression of natural logarithm-transformed PAR data with depth. In cases where CTD stations were conducted at night, Kd was calculated from a linear relationship established between pooled chlorophyll a concentrations and Kd’s determined at CTD stations conducted during the day (Kd = -0.0421 chl a * -0.0476). Mixed layer depths were calculated as the depth where density (sigma) changed by 0.05 from a 10 m reference point. Gross primary production was calculated at 0.1 time steps throughout the day (10 points per hour) and summed.

This data set was collected from a ocean acidification minicosm experiment performed at Davis Station, Antarctica during the 2014/15 summer season. It includes: - description of methods for all data collection and analyses. - marine microbial community data; Chlorophyll a concentration, particulate organic matter concentration (carbon and nitrogen), bacterial cell abundance. - phytoplankton primary productivity data; 14C-sodium bicarbonate incorporation raw data (decays per minute: DPM) and modelled productivity from photosynthesis versus irradiance (PE) curves, O2-evolution derived net community productivity, respiration, and gross primary productivity. - phytoplankton photophysiology data; community photosynthetic efficiency from PAM measurements (maximum quantum yield of PSII: Fv/Fm), PAM steady state light curve data and derived non-photochemical quenching of Chl a fluorescence (NPQ), relative electron transport rates (rETR), and effective quantum yield of PSII (delta F/Fm'). - phytoplankton carbon concentrating mechanism (CCM) data; maximum quantum yield of PSII (Fv/Fm) and effective quantum yield of PSII (∆F/Fm') from PAM measurements on size-fractionated phytoplankton samples (less than 10 microns and greater than 10 microns cells) exposed to; ethoxzolamide (EZA) which inhibits both intracellular carbonic anhydrase (iCA) and extracellular carbonic anhydrase (eCA), acetazolamide (AZA), which blocks eCA only, and a control (no inhibitor) sample. - bacterial productivity data; 14C-Leucine incorporation raw data (decays per minute: DPM) and calculated productivity.