Depth-integrated, continuous estimates of metabolism in a clear-water lake

High-frequency dissolved oxygen (DO) measurements have been used for estimating gross primary production (GPP) and respiration (R) in lake ecosystems. Most researchers have determined GPP and R only in surface waters, a practice that may underestimate R in general and GPP in clear-water lakes in particular. We deployed oxygen sondes at multiple sites and depths in a clear-water lake. Rates of GPP or R were similar horizontally over the surface waters of the lake. Diel DO signals weakened with depth; however, removing noise from the data, by either wavelet transforms or moving averages, enhanced our ability to resolve diel metabolic signals. While GPP declined sharply with depth, R was unrelated to depth. The majority of GPP and R occurred in the upper mixed layer, but deeper water accounted for 14%-28% of GPP and 20%-43% of R, depending on the statistical filtering technique used. GPP and R were nearly in balance in the surface waters, but for the entire lake R exceeded GPP, and net ecosystem production was negative. Deployment of oxygen sondes in various habitats and at multiple depths allows for a more complete estimate of whole-lake metabolism and a better understanding of the spatial and temporal complexity of lakes.