Direct measurement of pCO2 in cultures of marine phytoplankton: How good is the estimate from pH(NBS) and single point titration of alkalinity?

Abstract

In physiological studies of marine phytoplankton in culture, the CO2 equilibrium is typically calculated from measurement of pH(NBS) (calibrated in dilute National Bureau of Standards buffers) and a single point titration of alkalinity (A(T)). This approach has widespread appeal because it is simple, inexpensive, and requires very low sample volumes. However, its continued application ignores advances in analytical and theoretical oceanic CO2 chemistry that suggest fundamental flaws in the assumption pH(NBS) = -log[H+] as a consequence of the high ionic strength of seawater and liquid residual junction errors in the electrode. Here we compare directly measured pCO2 in sterile artificial seawater with calculated values, adopting the usual assumptions and commonly used equilibrium constants. Calculated values either overestimated (ca + 15 to + 23 % error) or underestimated (-9 to -11% error) directly measured pCO2 depending upon whether constants were derived, respectively, on the pH(NBS) or 'seawater' (pH(SWS)) scales. With the currently accepted equilibrium constants on the 'total hydrogen ion' pH scale (pH(TOT)), and converting pH(NBS) to pH(SWS) using an apparent activity coefficient f(H) (optimum value 0.85) and then to pH(TOT), excellent agreement was achieved between calculated and measured pCO2, both in sterile seawater and in cultures of the diatom Thalassiosira pseudonana. However, f(H) is generally unknown and is specific to electrode and electrode condition, making calculated pCO2 a rather nebulous concept. Without knowledge of f(H), calculated pCO2 had a total uncertainty of an order (~120 ppmv at atmospheric equilibrium 360 ppmv) similar to the variation in atmospheric pCO2 between glacial periods and the present (~160 ppmv). This method therefore clearly lacks the resolution required to address the biogeochemical significance of key physiological questions. Future studies should measure pCO2 directly, and then if required calculate CO(2(aq)) from pCO2 and the solubility coefficient. Alternatively, since analytical precision of calculated pCO2 was excellent, and accuracy is potentially good when f(H) is known, we advocate improved interdisciplinary collaboration in order to improve this pH and A(T) approach as a simple but effective tool for the study of marine phytoplankton physiology under controlled conditions.