Abstract
Understanding how aquatic ecosystems respond to perturbations has emerged as a crucial way to predict the future of these ecosystems and to assess their capacity to produce oxygen and store atmospheric carbon. In this context, in situ mesocosm experiments are a useful approach for simulating disturbances and observing changes in planktonic communities over time and under controlled conditions. Within mesocosm experiments, the estimation of fundamental parameters such as gross primary production (GPP), net community production (NCP), and respiration (R) allows the evaluation of planktonic metabolic responses to a perturbation. The continuous estimation of these metabolic parameters in real time and at high frequency is made possible by employing noninvasive automated sensors in the water column. However, some uncertainties and methodological questions about the estimation of daytime respiration remain to be addressed for this method, and notably to address the fact that respiration could be significantly higher during the day than during the night. In this study, data from two in situ mesocosm experiments performed in fall and spring in a coastal Mediterranean area were used to develop a new method of estimating daytime respiration, and in turn daily GPP, R, and NCP, by considering the maximum instantaneous R, and that takes into account the variability of the coupling between day-night and dissolved oxygen cycles. This new method was compared with the Winkler incubation technique and with another existing method. Results showed that using this existing method, daytime R was significantly underestimated relative to estimates obtained with the newly proposed method.