Contribution of hydrologic opportunity and biogeochemical reactivity to the variability of nutrient retention in river networks

In-stream nutrient retention results from the interaction between hydrological and biogeochemical processes involved in downstream transport. While hydrological processes set the opportunity for nutrient retention to occur, metabolic activity and abiotic processes determine the potential biogeochemical reactivity of streams. Yet, a comprehensive assessment of the relevance of hydrological opportunity versus biogeochemical reactivity on the variability of nutrient retention across streams is still missing. We compiled an extensive data set of existing studies on nutrient retention for ammonium, nitrate, and soluble reactive phosphorus to explore how variability in hydrological opportunity and biogeochemical reactivity explain nutrient retention. We quantified the relative contribution of hydrological opportunity and biogeochemical reactivity to the observed variability in stream nutrient retention using a linearization of the retention equation, which allows for an exact partitioning of the variance associated with residence time and nutrient uptake rate. Finally, we explored potential patterns of nutrient retention along the river network resulting from the interaction between hydrological opportunity and biogeochemical reactivity. Our results show that biogeochemical reactivity has a more relevant role on nutrient retention variability than previously thought, explaining over 66% of the variability in nutrient retention. Among the studied nutrients, retention variability of ammonium was the most subjected to biogeochemical reactivity controls. Furthermore, our results provide insights on controls of longitudinal patterns of nutrient retention along river networks, indicating that retention per unit length for the three nutrients will most likely decrease from headwaters to river mouth because of a decrease in water residence time.R.M. participation was funded by MINECO through the project REMEDIATION (CGL2014-57215-C4-2-R). S.B. work was funded by MINECO through the project NICUS (CGL-2014-55234-JIN). E.M. participation was funded through the project MEDSOUL (CGL2014-59977-C3-2-R). D.v.S. was funded by MINECO through the project DIVERSION (CGL2016-77487-R) and by the Basque Government through a Grant for Research Groups of the Basque University System (IT-951-16).Peer reviewe