Improved Accuracy of Optode‐based Oxygen Consumption Measurements by Removal of System Drift and Nonlinear Derivation

This method evaluation aimed to improve the accuracy and precision of the previously published method to measure oxygen consumption using optodes with integrated temperature and salinity correction in dark incubations. Significant short-term system drift currently requires a correction to remove the drift, thus reducing the precision of the oxygen consumption rates. Frequent nonlinear declines in oxygen concentration with time also call for improved data analysis and identification of its origin. Optodes in titanium casings (Aanderaa (TM) model 4330) with low oxygen binding properties showed no significant system drift in autoclaved seawater. Nonlinear oxygen dynamics fitting a quadratic polynomial occurred in 28% of 230 field samples, independent of season and water depth. Polynomial curve fit resulted in 64% higher respiration rates when derived within 1 h of the quality assured incubation, than obtained when using linear fit. Carbon substrate limitation explained the nonlinearity of oxygen decline during dark incubations. Pretreatment of the optode attached to stoppers with 0.3 mol dm(-3) hydrochloric acid resulted in the most stable performance of the sensor and simultaneously provided proper cleaning of the equipment. A conservative detection limit of 0.97 mu mol O-2 dm(-3) d(-1) was calculated for the titanium optodes, matching other methods for oxygen consumption reported in the literature. Thus, we recommend the use of model 4330 optode pretreated with HCl and the derivation of initial respiration rates by a quadratic polynomial function for best accuracy and precision of oxygen consumption in oxygenated surface waters.