Seasonal changes in rates of gross primary production (GPP), net ecosystem production (NEP), and respiration (R) were determined from frequent automated profiles of dissolved oxygen (DO) and temperature in a clear-water polymictic lake. Metabolic rate calculations were made using a method that integrates rates across the entire depth profile and includes DO exchange between depth layers driven by mixed-layer deepening and eddy diffusivity. During full mixing, NEP was close to zero throughout the water column, and GPP and R were reduced 2-10 times compared to stratified periods. When present, the metalimnion contributed 21% and 27% to whole-lake areal rates of GPP and R, respectively. Net autotrophy prevailed in the epilimnion (NEP = 11 +/- 14 mmol O-2 m(-3) d(-1); mean 6 +/- SD) compared to balanced production in the metalimnion (NEP = 2 +/- 19 mmol O-2 m(-3) d(-1)) and net heterotrophic conditions in hypolimnic waters (NEP = -15 +/- 24 mmol O-2 m(-3) d(-1)). Positive NEP occurred in the metalimnion during periods when the photic depth extended below the mixed-layer depth. Although the single-sonde method estimated higher areal GPP (19%) and R (14%) compared to the two depth-integrated approaches, differences were not significant. During stratification, daily variability in epilimnetic DO was dominated by metabolism (46%) and air-water gas exchange (44%). Fluxes related to mixed-layer deepening dominated in meta- and hypolimnic waters (49% and 64%), while eddy diffusion (1% and 14%) was less important. Although air-water gas exchange rates differed among the three formulations of gas-transfer velocity, this had no significant effect on metabolic rates.