Only aquatic isopods and chironomid larvae in the stream instead of exuberant biodiversity? In that case, it is in poor condition. Birte Vierjahn talks to Dr. Christian Feld about the online tool for experts, developed by his team. It identifies deficiencies and therapies – and can even look into the future.
When he's not wading through bodies of water, Christian Feld works here at the Essen campus. | © Frank Preuß

How does your tool infer from insect larvae and small crustaceans how healthy a water body is from an ecological perspective?

Our biological samples tell us what we need to know, for example which caddis flies, amphipods and mussels are present, because each species has different habitat requirements. A change in the biological diversity and the community composition can therefore indicate unfavourable conditions in a water body, such as an unfavourable water temperature, oxygen content, current condition, or stream bottom habitat.

And a specialist can use these data to determine what the problem is …​

Exactly. Our tool supports the diagnosis of potential causes: Once the biological symptoms of a water body have been entered in the input section, the tool calculates the probability of possible causes in light of the symptoms, such as insufficient shade or enhanced nutrient levels, and then puts the causes into a hierarchical order. It also suggests possible management measures for improving the ecological quality of the water body. Hence, it works like a visit to the doctor: first the diagnosis, then the treatment.

So your tool helps to reverse damage to aquatic biology in the most targeted way possible?​

Yes. But in another project, we switched the direction of the tool: There, we examined the environmental deterioration of water bodies as a basis for assessing the probability of ‘harmful’ biological degradations. This tool allows us to make a prognosis of biological changes on the basis of selected environmental changes. What we mean by ‘harmful’ is, for example, the loss of biodiversity and ecosystem services. Ecosystem services are the services that ecosystems provide to humans. Streams and rivers, for example, provide food to many people. They also supply fresh water for different uses, which is supported by a process called self-purification. Floodplains contribute to climate protection, for example by storing CO2 on a large scale.

The diagnostic and prognostic tools together then help us not only to identify the most suitable management measures but also to assess their prospects for success.

Result of a diagnosis of lowland rivers with the following symptoms selected in the input section of the tool (the symptoms are derived from benthic invertebrate communities): too low a proportion of sensitive indicator species, too high a proportion of gathering collectors and flow-avoiding species. The radar plot shows the output section with indications of changing probabilities for potential causes of a deterioration. The diagnostic tool is available online at
udue.de/flussdoktor.
Your diagnostic tool becomes more precise with each water body analysed. What can it do now that it couldn't do in the beginning?​

A whole lot! First, the larger data pool makes the analyses more precise, because it can better account for the natural variability and include it in the diagnosis.

The biological symptoms of a stream can sometimes be very variable, for example with regard to the ecological implications of excessive sunlight. It is also important to consider whether a stream is located in the lowlands or highlands or whether or not it is fed by groundwater. Natural variability could be considered somewhat like an error bar in mathematics. The more streams and rivers we consider to train the tools, the better we can identify the natural variability and take it into account in the diagnosis. Second, until recently we only considered insects, small crustaceans, mussels and other invertebrates on the stream bottom. Now we have also included algae. The diagnostic information we can obtain from invertebrates is occasionally somewhat limited. The influence of nitrogen or phosphorus concentrations, for example, can be diagnosed much more reliably from plant communities.

Specifically, we just finished a project with the Baden-Wurttemberg State Environmental Agency (Landesanstalt für Umwelt Baden-Wurttemberg). We developed diagnostic tools especially for streams and rivers in this state: streams and rivers in the Alpine foothills, mountain streams and mountain rivers. We developed and fine-tuned the diagnostic tools in several workshops together with regional domain experts. The tools are about to be published and will hopefully soon be used in Baden-Wurttemberg to help plan restoration measures in streams and rivers.

Such measures have of course been conducted for a long time, not only in Baden-Wurttemberg. But they do not always take into account the actually existing biological problems and are occasionally even planned ‘on a gut instinct’. We want to help change that with the diagnostic tools.

In which direction would you like to further develop your tool?

Provided that we receive the funding, we would like to expand it to allow this diagnosis to be conducted at different levels, for example in Northern Europe or Southern Europe. It would be difficult to do so for Europe as a whole, because the ecoregions are so different, and the relationships between causes and effects cannot simply be transferred. In Southern Europe, for example, changes in water volume and discharge play a major role in dry seasons, but in Northern Europe there is no dry season. We do, however – in keeping with the doctor metaphor – want to work at a supra-regional level like a general practitioner and then go to the specialists for more specific things, thus narrowing down the level of diagnosis ever further.

Taking stock of a water sample: Feld carefully lifts out a caddisfly larva. | © Foto Frank Preuß

You can be seen wading through streams on a TV report – you have already studied more than 200 water bodies. Can you make out regional differences?

Yes, there are regional differences, even just due to the different climatic conditions. In North Rhine-Westphalia, for example, we have much more precipitation than in the east of Brandenburg. At least that was the case before 2018; then we had three years of drought in a row. Another example is the geology: Northern Germany is influenced by the last two ice ages, and many water bodies are dominated by sand and gravel. In the low mountain ranges, completely different geological formations come to light, such as shale or red sandstone, which then weather to coarser, stony substrates and shape the stream bottom. There are of course many more criteria, but climate and geology are the most striking.
Link to TV report

Do you have regional examples of a particularly healthy or sick water body?

Particularly healthy is always somewhat difficult in the Ruhr metropolitan region (laughs). Take the Emscher: Even though it is now officially free of waste water and the water quality is improving, there is still room for improvement with the structures, such as the mostly still reinforced banks, the concrete bottom, the lack of shade, and the regulated runoff. But in the middle and headwater sections, restorations have been implemented already or are still ongoing.

But there are indeed several good water bodies, such as the upper Rotbach between Grafenmühle in Bottrop and the ‘Sträterei’. That’s the upper course of that stream, located in the woods, and it’s great. It’s a classic type of sandy lowland stream in its natural state.