What is it about?
Stingless bee colonies were exposed to chlorpyrifos sprayed in front of the hives, with water spray and no-disturbance conditions as controls. Acoustic data were recorded during all treatments. The Hidden Markov Model successfully detected pesticide exposure periods. Model performance improved when trained and tested on each hive individually rather than on data from all hives combined. Hive identification also showed high performance, suggesting that each hive has a distinct acoustic profile even under similar conditions.
Featured Image
Photo by Logan Voss on Unsplash
Why is it important?
These findings demonstrate that acoustic monitoring using machine learning can effectively detect pesticide exposure in stingless bees. This reveals the potential of using stingless bees as sensitive bioindicators of pesticide drift, an environmental threat that impacts not only bee populations but also ecosystem health and human well-being. Early detection of chronic, low-dose pesticide exposure, which often goes unnoticed, could help prevent long-term health consequences.
Perspectives
In my perspective, the study of stingless bees is still underexplored compared to precision beekeeping for honey bees. The use of precise tools, such as machine learning applied to bioacoustic monitoring, can greatly enhance our understanding of stingless bee behavior and health, contributing to their conservation and environmental safety. This study highlights one impactful approach using these techniques, but there is still much to explore in this field.
Alex Otesbelgue
Universidade Federal do Rio Grande do Sul
Read the Original
This page is a summary of: Hidden Markov model for acoustic pesticide exposure detection and hive identification in stingless bees, PLOS One, June 2025, PLOS,
DOI: 10.1371/journal.pone.0325732.
You can read the full text:
Contributors
The following have contributed to this page
