Abstract
Historically, estimates of pelagic primary production in lake ecosystems were made by measuring the uptake of carbon-14 (C-14)-labeled inorganic carbon in samples incubated under laboratory or in situ conditions. However, incubation approaches are increasingly being replaced by methods that analyze diel changes in high-frequency in situ data such as free-water dissolved oxygen (O-2). While there is a rich literature on the comparison of approaches for estimating primary production using incubations (e.g., C-14 and O-2 bottle experiments), as well for approaches using high-frequency data (e.g., diel O-2 and CO2 metabolism models), there are few direct comparisons of C-14 incubations and free-water O-2 approaches for estimating primary production. We used 20 lake-years of concurrent measurements of primary production quantified from high-frequency free-water O-2 data and C-14 incubations in four different lakes (4-7 years per lake) to compare these different approaches. Across all lakes, 61% of the C-14 production estimates were within the 95% credible intervals of the free-water O-2 production estimates. Error-in-variable regressions support the assumption that C-14 methods estimate a production value between gross primary production and net primary production and the bottle effect is constant across the entire range of production values considered here. There was little evidence that daily pelagic, epilimnetic estimates of primary production differed substantially based on the selection of free-water O-2 or C-14 approaches in these lakes during summer stratified conditions.