Classifying the Biological Status of Honeybee Workers Using Gas Sensors
Jakub T Wilk1, Beata Bąk1, Piotr Artiemjew2
1Apiculture Division, Faculty of Animal Bioengineering, University Warmia and Mazury in Olsztyn, Sloneczna 48, 10-957 Olsztyn, Poland.
Sensors (Basel, Switzerland)
|January 1, 2021
Summary
Honeybee worker bees have distinct odors based on age and colony status. Gas sensors successfully differentiated between young, old, and queenless colony bees, offering potential for practical beekeeping applications.
Area of Science:
- Apiculture
- Animal Behavior
- Biosensors
Background:
- Honeybee (Apis mellifera) worker bees exhibit olfactory cues that vary with age and colony health.
- Understanding these olfactory differences is crucial for monitoring colony status and bee welfare.
Purpose of the Study:
- To investigate the potential of gas sensor technology to differentiate honeybee workers based on age and colony biological status.
- To evaluate the effectiveness of various classifiers in analyzing sensor data for accurate bee classification.
Main Methods:
- Laboratory experiments utilizing two prototype multi-sensor gas detectors.
- Exposure of sensors to odors from young bees (0-1 days), old bees (3-6 weeks), and bees from queenless colonies.
- Analysis of sensor responses using 10 different classification algorithms, including Euclidean.1nn.
Main Results:
- The gas sensor system demonstrated an ability to distinguish between different biological statuses of honeybee workers.
- The Euclidean.1nn classifier achieved a balanced accuracy of 0.763 and a true positive rate of 0.742.
- Classification effectiveness is influenced by environmental factors within the test chamber and the choice of classifier.
Conclusions:
- Gas sensor technology shows promise for non-invasive monitoring of honeybee colony status in beekeeping.
- Further optimization of classifiers and environmental controls can enhance the accuracy and practical applicability of this technology.
