Bénard, Robin; Levasseur, Maurice; Scarratt, Michael Grant; Lizotte, Martine; Michaud, Sonia; Starr, Michel; Tremblay, Jean-Éric; Kiene, Ronald P; Kameyama, Sohiko: St. Lawrence Estuary phytoplankton spring bloom and associated dimehtylsulfide cycling sensitivity to acidification in a microcosm experiment [dataset]. PANGAEA, https://doi.pangaea.de/10.1594/PANGAEA.915942 (dataset in review)

The potential impact of ocean acidification (OA) on the concentrations of dimethylsulfide (DMS) and dimethylsulfoniopropionate (DMSP), as well as the processes governing the cycling of sulfur compounds by microbial organisms were investigated in a microcosm experiment during summer 2013. A natural planktonic community sourced from the surface waters of the Lower St. Lawrence Estuary (LSLE) was monitored over 12 days under three pCO2 targets: 1 × pCO2 (775 µatm), 2 × pCO2 (1850 µatm), and 3 × pCO2 (2700 µatm). The effects of heightened pCO2 were assessed on phytoplankton biomass, taxonomy and productivity, as well as on DMSP and DMS concentrations and microbial cycling via 35S-DMSPd radioisotope uptake assays. A mixed phytoplankton bloom comprised of the diatoms Chaetoceros spp., Skeletonema costatum, and Thalassiosira spp., as well as unidentified flagellates (2-20 µm) developed over the course of the study period. Concentrations of chlorophyll a (Chl a) peaked on day 6 from an initial 1.26 µg L-1 to 20 ± 4 µg L-1 and 17 ± 1 µg L-1, at 1 × pCO2 and 2 × pCO2, respectively. However, the timing and the magnitude of the biomass build-up were altered in the 3 × pCO2 treatment reaching only 11 ± 4 µg L-1 by day 8 of the experiment. Variations in the concentrations of total DMSP (DMSPt) were largely related to variations in Chl a (Spearman's rho correlation (rs) = 0.77; p < 0.001; n = 61) and peak values of DMSPt were reached consecutively on day 8 in the 1 × pCO2 treatment (264 nmol L-1), on day 10 in the 2 × pCO2 (245 nmol L-1), and on day 12 in the 3 × pCO2 (192 nmol L-1). Concentrations of dissolved DMSP (DMSPd) increased irregularly from 2 ± 1 nmol L-1 to an overall average of 5.2 ± 0.8 nmol L-1 by the end of the experiment; the fluctuations being broadly associated with those observed in DMSPt (rs  = 0.70; p < 0.001; n = 61). Neither concentrations of DMSPd nor the microbial scavenging efficiency of DMSPd (kDMSPd) seemed to be strongly affected by increasing pCO2 suggesting that acidification did not alter the availability and the bacterial uptake of this substrate. However, our results show a reduction of the mean microbial yield of DMS by 34 % and 61 % in the 2 × pCO2 and 3 × pCO2 treatments. Doubling and tripling the pCO2 respectively resulted in a 15 % and 40 % decline in average concentrations of DMS compared to the control. DMS concentrations were positively correlated with microbial yields of DMS (rs = 0.65; p < 0.001; n = 45), suggesting that in diatom-dominated systems, the impact of OA on concentrations of DMS may be strongly linked with alterations of the microbial breakdown of DMSPd. Findings from this study provide the first empirical evidence of the sensitivity of the microbial DMSP switch under OA.