Long-term variability in the metabolism of an aquatic ecosystem: results from 25 years of continuous dissolved oxygen monitoring in a large urban river.

The availability of Dissolved oxygen (DO) sensors and open-source modelling tools has greatly expanded the breadth of knowledge regarding the metabolism of aquatic systems. River metabolism is one of the most integrative indicator of ecological health, it can be applied across diverse water systems, through continuous measurement systems, and is highly sensitive to environmental stressors as it reflects shifts in anthropogenic disturbance or management actions. While former studies showed that the estimation of aquatic ecosystem metabolism can be used as a relevant tool for urban water management, long-term metabolism investigations remain currently scarce, especially in large rivers.

Relying on the MeSeine Observatory, a sensor network of 8 sites measuring  continuously DO and water temperature in the Seine River from upstream to downstream of Paris region, we examined 25 years of spatial and temporal variability of the river metabolism and its underlying drivers. Daily metabolic rates, including gross primary production (GPP), ecosystem respiration (ER), which both contribute to the net ecosystem production (NEP), were estimated using the single-station open-channel method, integrating hourly DO, temperature, and river discharge data.

Preliminary results reveal distinct seasonal and inter-annual patterns in GPP, ER and NEP rates. ER and GPP peaked in late spring and in summer, driven by high light availability, warm temperatures and low discharge, while the lowest rates occurred in winter across all sites. Inter-annual variations were primarily influenced by hydroclimatic conditions and sewage inputs. Annual mean rates of GPP, ER and NEP across all sites ranged from ~0.29 to ~2.3, ~-5.65 to ~-0.53, and ~-4.16 to ~0.51 gO₂.m⁻³.d⁻¹, respectively. ER consistently exceeded GPP, indicating a predominantly heterotrophic status downstream of Paris. Notably, the upstream site of the observatory (the only one before Paris) exhibited several years with positive annual NEP values. Along the river, net heterotrophy increased downstream, likely due to urban organic matter inputs. While GPP and ER displayed similar temporal patterns, NEP followed a distinct trajectory, aligning more closely with water quality, organic matter and nutrient concentrations. Moreover, in recent years, NEP tends to increase, reflecting a decrease of the heterotrophy and nutrient concentrations in the river.

The analysis demonstrates that the Seine River's metabolism has exhibited distinct seasonal and upstream-to-downstream trends over the past 25 years, driven by urban impacts, seasonal dynamics, and hydroclimatic conditions. Future work will focus on refining these observations to uncover long-term trends and establish clear relationships between metabolic rates and key environmental stressors across shorter timescales, with special focus on the structural components, hydrological information and sewage effluents. To assess ecological status and develop effective management tools, it is essential to understand the interactions between river components and the cause-effect mechanisms underlying aquatic ecosystem alterations.