An evaluation of light intensity functions for determination of shaded reference stream metabolism

The performance of three single-station whole stream metabolism models were evaluated within three shaded, seasonally hypoxic, Missouri reference streams using high resolution (15-minute) dissolved oxygen (DO), temperature, and light intensity data collected during the summers (July–September) of 2006–2008. The model incorporating light intensity data consistently achieved a lower root mean square error (median RMSE = 0.20 mg L−1) relative to models assuming sinusoidal light intensity functions (median RMSE = 0.28 mg L−1) and constant diel temperature (median RMSE = 0.53 mg L−1). Incorporation of site-specific light intensity into metabolism models better predicted morning DO concentrations and exposure to hypoxic conditions in shaded study streams. Model choice significantly affected (p < 0.05) rate estimates for daily average photosynthesis. Low reaeration (pooled site mean 1.1 day−1 at 20 °C) coupled with summer temperatures (pooled site mean = 25.8 °C) and low to moderate community respiration (site median 1.0–3.0 g O2 m−2 day−1) yielded diel dissolved oxygen concentrations near or below critical aquatic life thresholds in studied reference streams. Quantifying these process combinations in best-available or least-disturbed (i.e., reference) systems advances our understanding of regional dissolved oxygen expectations and informs environmental management policy. Additional research is warranted to better link landscape processes with distributed sources that contribute to community respiration.

► We model the importance of light intensity in determining stream metabolism.
► Light intensity data are needed to model dissolved oxygen in shaded streams.
► We estimate metabolism parameters yielding hypoxic dissolved oxygen concentrations.
► Characterizing metabolism in reference streams informs management decisions.