Bioacoustic Dataset of Sudden Queen Loss in an Apis mellifera L. Honeybee Colony

The queen bee plays a crucial role in maintaining colony well-being. However, traditionally determining her presence within a hive requires invasive methods, causing stress among the bees and potential nest damage. Audio monitoring offers a simpler, cost-effective, and non-invasive approach to detect various conditions using colony bioacoustics. In this study, we recorded an Africanized honeybee colony for six days, removing the queen on the final day. From these recordings, we sampled 300 frames of 3 seconds each, extracting three time-domain features, five spectral features, and 13 Mel-Frequency Cepstral Coefficients (MFCCs). This resulted in a dataset of 5,400 samples capable of distinguishing queen presence or absence, even with class imbalance.