Estimating autotrophic respiration in streams using daily metabolism data

The fraction of gross primary production (GPP) that is immediately respired by autotrophs and their closely associated heterotrophs (AR(f)) is unknown. This value is necessary to calculate the autotrophic base of food webs, which requires knowing production available for grazers. AR(f) is also necessary for estimating heterotrophic respiration (HR) which is needed to calculate C spiraling in streams and rivers. We suggest a way to estimate AR(f) from daily metabolism data using quantile regression between GPP and 90% quantile of ecosystem respiration (ER). We reasoned that autotrophic respiration represents the lower limit for ER on any one day and used quantile regression to estimate the relationship of the lower quantile of ER with respect to GPP. We examined this approach with simulation modeling and application of quantile regression to estimates of continuous GPP and ER from > 20 streams. Simulation modeling showed that low-uncertainty estimates of AR(f) required large variation in daily GPP. Covariance between HR and GPP, which might be observed if the processes were temperature controlled, biased estimates of AR(f). Seasonal estimates of AR(f) were robust to daily variation in AR(f) as a function of GPP. AR(f) calculated from previously published estimates of daily metabolism from streams averaged 0.44 (SD = 0.19) with high variation among streams. This value is higher than most physiological measurements, probably because of light limitation of algae and from HR closely associated with daily GPP. How much of AR(f) was from algal respiration vs closely associated heterotrophic respiration is not known, but we suggest that the value (1 - AR(f))GPP represents the amount of C available to animals.