Diel O2 data from JMS99 and JMS75 were analyzed to assess differences in ecosystem metabolism above and within the zone of the CHLa maximum. Diel amplitudes for dissolved O2 were consistently larger at the site of the CHLa maximum (mean=3.8 mg L−1) relative to the upriver site (mean=1.4 mg L−1). Atmospheric exchange was a small component of O2 fluxes at both sites with average values corresponding to 2% and 19% of biologically driven O2 fluxes. Atmospheric exchange was greater at JMS75 due to
its shallower depth (0.08±0.02 mg L−1 day−1 and −1.27±0.29 mg L−1 day−1 at JMS99 and JMS75, respectively); negative fluxes at this site were a result of persistent supersaturation. By comparison, the upriver site was typically under-saturated except in late summer during the afternoon O2 peak. Ecosystem production and respiration values derived from diel O2 data were also higher at JMS75 relative to JMS99 (Fig. 5). At JMS75, NPPe was on average three times higher (3.98±0.46 mg L−1 day−1 vs. 1.11±0.19 mg L−1 day−1) and Re was twofold higher (5.18± 0.62 mg L−1 day−1 vs. 2.31±0.42 mg L−1 day−1) relative
to JMS99 over the period of study (Table 1). Net ecosystem metabolism was positive at JMS75 (1.39±0.62 mg L−1 day−1) but near zero at JMS99 (−0.17± 0.62 mg L−1 day−1). Spatial and temporal variation inNPPe was strongly related to average underwater irradiance with a non-linear model (Jassby and Platt 1976) accounting for 89% of the variation in NPPe (N=22; p< 0.0001). The model depicted a saturation response when I x-sec exceeded 10 E m−2 day−1. CHLa was the best predictor for both Re and NEM (R2=0.46 and 0.45, respectively; p