Comparative performance of CO2 measuring methods: Marine aquaculture recirculation system application

Abstract

Many methods are available for the measurement of dissolved carbon dioxide in an aqueous environment. Standard titration is the typical field method for measuring dissolved CO2 in aquaculture systems. However, titrimetric determination of dissolved CO2 in marine water aquaculture systems is unsuitable because of the high dissolved solids, silicates, and other dissolved minerals that interfere with the determination. Other methods used to measure dissolved carbon dioxide in an aquaculture water included use of a wetted CO2 probe analyzer, standard nomographic methods, and calculation by direct measurements of the water's pH, temperature, and alkalinity. The determination of dissolved CO2 in saltwater based on partial pressure measurements and non-dispersive infra-red (NDIR) techniques with a CO2 gas analyzer are widely employed for oceanic surveys of surface ocean CO2 flux and are similar to the techniques employed with the head space unit (HSU) in this study. Dissolved carbon dioxide (DC) determination with the HSU using a infra-red gas analyzer (IRGA) was compared with titrimetric, nomographic, calculated, and probe measurements of CO2 in freshwater and in saltwater with a salinity ranging from 5.0 to 30ppt, and a CO2 range from 8 to 50mg/L. Differences in CO2 measurements between duplicate HSUs (0.1-0.2mg/L) were not statistically significant different. The coefficient of variation for the HSU readings averaged 1.85% which was better than the CO2 probe (4.09%) and that for the titrimetric method (5.84%). In all low, medium and high salinity level trials HSU precision was good, averaging 3.39%. Differences existed between comparison testing of the CO2 probe and HSU measurements with the CO2 probe readings, on average, providing DC estimates that were higher than HSU estimates. Differences between HSU and titration based estimates of DC increased with salinity and reached a maximum at 32.2ppt. These differences were statistically significant (P<0.05) at all salinity levels greater than 0.3ppt. Results indicated reliable replicated results from the head space unit with varying salinity and dissolved carbon dioxide concentrations. © 2010.