Molecular biology and AI in biodiversity research (IQ-Water)

Research project on AI-supported monitoring and forecasting of biodiversity and water quality in drinking water reservoirs

Drinking water reservoir in the Black Forest - a place of natural biodiversity and ecosystem services

Due to climate change, pollution and invasive species, reservoirs and lakes are confronted with a loss of near-natural biodiversity.  The IQ-Water project uses modern molecular biology methods and AI-supported models to gain a better understanding of these ecosystems.

Surface waters such as lakes and reservoirs are complex biological systems with a large biological diversity of different organisms such as bacteria, algae, protozoa, fungi and higher plants and animals. The ecosystem services provided by this biological diversity enable, among other things, the near-natural production of over 12% of Germany's drinking water. Water quality is closely interrelated with biodiversity. Driven by climate change, environmental pollution and the spread of invasive species (neobiota), there has been a significant loss of biodiversity and a change in species composition in the majority of German water bodies. Significant for humans and the environment are, for example, the increased occurrence of toxic algal blooms, the introduction of faecal contamination, the spread of antibiotic resistance, the change in species composition due to invasive neobiota or altered life cycles of the plankton community due to higher water temperatures. All of these changes have a direct impact on the ecosystem services of biotic communities and therefore ultimately on water quality.

Basically, the analysis of the water quality of drinking water reservoirs is based on the recording of classic microbial indicators and physico-chemical measurement parameters. This extensive database goes back many years. However, lakes and reservoirs are complex ecosystems whose vulnerability to climate change, sources of pollution and neobiota have not been sufficiently researched to make complex predictions regarding their future biodiversity development under changing environmental conditions.

Microorganisms in particular are among the groups of organisms that have been insufficiently studied to date. Molecular methods have the potential to close this knowledge gap in aquatic biodiversity. The analyses of metagenomes and so-called environmental DNA (eDNA) allow direct information on the entirety of genetic information in an ecosystem and thus the representation of genetic biodiversity. The highly complex data of different origins represent a hurdle for a comprehensive evaluation, especially when including historical time series.

The aim of the IQ-Water project is the joint analysis of biological, chemical and physical data to model biodiversity and biodiversity loss using modern AI-supported methods. The use of AI to analyze biodiversity has hardly been investigated so far and is limited, for example, to video monitoring of marine ecosystems or pure potential analyses. The main obstacles to the use of these methods for complex physical and biological systems are the time-consuming data collection and the consideration of complex models and expert knowledge about the fundamentals of the systems.

The AI-supported evaluation of the data should enable future biodiversity development, the proliferation of cyanobacteria or coliform bacteria, the input of faecal contaminants and their origin as well as the occurrence of neobiota to be identified or predicted at an early stage. Based on a better understanding of the loss of biodiversity, measures to preserve the natural ecosystem and protect water quality are to be introduced at an early stage.