I worked with a research project that was sensing, collecting and analyzing types of mussels in rivers that were used as an indirect measurement of water quality in the river. The assumption is that the mussels can act as a proxy for certain contaminants. In fact, they actually filter a lot of water daily so you can occasionally analyze one as to what sort of contaminants it absorbed as a proxy for whats probably in the water.
Meanwhile, water intakes typically only monitor for known and suspected contaminants. They have to rely on other indirect measurements like changes conductivity as a signal to analyze the water and these are noisy and fairly unreliable compositional change detection proxies.
Neat stuff.