Monitoring infestation levels using an AI-based Chironomid quantification tool (CQT) in a laboratory Xenopus laevis facility

Objective Insect infestations in laboratory animal facilities represent a potential risk to animal welfare, staff hygiene, and experimental integrity. In this study, we developed and applied a custom AI-assisted image analysis workflow, the Chironomid Quantification Tool (CQT-AI), to detect and quantify infestations of non-biting midges ( Chironomidae ) in an aquatic housing system for Xenopus laevis . To evaluate the applicability of CQT-AI for monitoring infestation dynamics and treatment effects, different control strategies, including increased water exchange and biological control using Bacillus thuringiensis israelensis (Bti), were tested in parallel. Results CQT-AI proved to be a reliable tool for detecting and quantifying infestation dynamics of adult Chironomidae , enabling objective assessment of treatment effects within 1 min. Increased water exchange resulted in measurable changes in water chemistry but only moderately reduced infestation levels. In contrast, treatment with Bti led to rapid and near-complete eradication of adult midges within six weeks without detectable adverse effects on animal health. Overall, the study demonstrates that CQT-AI provides an efficient and objective approach for monitoring infestation dynamics and evaluating control strategies in aquatic laboratory systems.