Temporal variability of the realized eutrophication in rivers across Germany

Eutrophication, usually considered an overproduction of biomass in water bodies due to an enrichment of nutrients such as phosphorus (P) and nitrogen (N), is a common thread to riverine ecosystems. In Europe, chlorophyll a (Chl-a) concentrations, an indicator for algae biomass, have been successfully reduced from the 1980s by limiting phosphorus input into rivers. However, elevated Chl-a levels in European rivers are still found until today. As recent algae blooms were rather linked with drought periods than with particularly high phosphorus concentrations, the question is raised whether other parameters than phosphorus concentrations might be more crucial for eutrophication management in the future.

To understand the conditions under which rivers are particularly prone to an efficient conversion of phosphorus into algae biomass, we analyzed a Germany-wide dataset of Chl-a and total phosphorus (TP) concentrations with 31661 measurement pairs at 330 stations between 2000 to 2019. To quantify this conversion efficiency, we used the measure of the degree of realized eutrophication, α_realized, which is the ratio between the realized (i.e. the Chl-a measurement) and the potential eutrophication (i.e. a theoretical upper Chl-a concentration at a given TP level if all TP is converted to biomass). In a preceding study, we found that station-wise medians of α_realized are mainly controlled by water residence time with high median α_realized being either related to large rivers with a long distance to source or small rivers with close upstream lakes. As management not only asks where but also when Chl-a concentrations are at critical levels, we here analyze the temporal variability of α_realized at different stations. To that end, we calculate the coefficients of variability of α_realized, TP, and Chl-a, and statistically relate these characteristics to other instream parameters and catchment attributes at each station.

We find both stations with low α_realized variability and a positive TP - Chl-a correlation and stations with high α_realized variability and no or even a negative TP and Chl-a correlation. The former case suggests stable controls of α_realized and good predictability of Chl-a based on TP concentrations at the respective stations. The latter case implies either variable water residence times or additional controls by parameters with strong seasonality, such as light availability, water temperature, or ecological community shifts.

In this contribution, we will present whether high temporal variability of α_realized is predictable from instream parameters or catchment attributes and discuss the underlying processes. We will further conclude on the implications of the results for river management, particularly in terms of algae control and in light of climate change.