Coupled primary production and respiration in a large river contrasts with smaller rivers and streams

Although time series in ecosystem metabolism are well characterized in small and medium rivers, patterns in the world's largest rivers are almost unknown. Large rivers present technical difficulties, including depth measurements, gas exchange (K$$ K $$, d-1$$ {\mathrm{d}}<^>{-1} $$) estimates, and the presence of large dams, which can supersaturate gases. We estimated reach-scale metabolism for the Hanford Reach of the Columbia River (Washington state, USA), a free-flowing stretch with an average discharge of 3173 m3$$ {\mathrm{m}}<^>3 $$ s-1$$ {\mathrm{s}}<^>{-1} $$. We calculated K$$ K $$ from semi-empirical models and directly estimated it from tracer measurements. We fixed K$$ K $$ at the median value from these calculations (0.5 d-1$$ {\mathrm{d}}<^>{-1} $$), and used maximum likelihood to estimate reach-scale, open-channel metabolism. Both gross primary production (GPP) and ecosystem respiration (ER) were high (GPP range: 0.3-30.8 g O2$$ {\mathrm{O}}_2 $$ m2$$ {\mathrm{m}}<^>2 $$ d-1$$ {\mathrm{d}}<^>{-1} $$, ER range: 0.8-30.6 g O2$$ {\mathrm{O}}_2 $$ m2$$ {\mathrm{m}}<^>2 $$ d-1$$ {\mathrm{d}}<^>{-1} $$), with peak GPP and ER occurring in the late summer or early fall. GPP increased exponentially with temperature, consistent with metabolic theory, while light was seasonally saturating. Annual average GPP, estimated at 1500 g carbon m-2$$ {\mathrm{m}}<^>{-2} $$ yr-1$$ {\mathrm{yr}}<^>{-1} $$, was in the top 2% of estimates for other rivers. GPP and ER were tightly coupled and 90% of GPP was immediately respired, resulting in net ecosystem production near 0. Patterns in the Hanford Reach contrast with those in small-medium rivers, suggesting that metabolism magnitudes and patterns in large rivers may not be simply scaled from knowledge of smaller rivers.