Growth as a phenotypic endpoint for the assessment of the impact of pollutants on water fleas - Comparisons between manual, online and AI assisted approaches.

Environmental risk assessment relies on the detection of contaminants in the environment. However, these methods even if they are reliable for known chemicals in the environment, yet, they fall short in realism to capture complex, real-world ecological conditions and provide mechanistic insight. The use of bioindicator species such as Daphnia magna provides a more holistic view on the interactions between common freshwater pollutants and the biological inhabitants of the ecosystem. Daphnids exhibit phenotypic characteristics which can serve as easily quantifiable endpoints to monitor physiological changes and among them, growth serves as a sensitive and non-invasive endpoint. In this study, we developed a miniaturised growth assay which we applied also with the implementation of AI approaches to automate toxicological assessment. The miniaturised method proved effective and captured the physiological alterations caused by leachates of tobacco and smoked cigarette filters as indicative complex mixtures of pollutants.