We present a one-year-long multi-sensor dataset collected from honey bee colonies (Apis mellifera) with rich phenotypic measurements. Data were collected non-stop from April 2020 to April 2021 from 53 hives located at two apiaries in Québec, Canada. The sensor data included audio features, temperature, and relative humidity. The phenotypic measurements contained beehive population, number of brood cells (eggs, larva and pupa), Varroa destructor infestation levels, defensive and hygienic behaviors, honey yield, and winter mortality. Our study is amongst the first to combine a wide variety of phenotypic trait measurements annotated by apicultural science experts with multi-sensor data, which facilitate a broader scope of analysis. We first summarize the data collection procedure, sensor data pre-processing steps, and data composition. We then provide an overview of the phenotypic data distribution as well as a visualization of the sensor data patterns. Lastly, we showcase several hive monitoring applications based on sensor data analysis and machine learning, such as winter mortality prediction, hive population estimation, and the presence of an active and laying queen.

In this paper, we present a multimodal dataset obtained from a honey bee colony in Montréal, Quebec, Canada, spanning the years of 2021 to 2022. This apiary comprised 10 beehives, with microphones recording more than 3000 hours of high quality raw audio, and also sensors capturing temperature, and humidity. Periodic hive inspections involved monitoring colony honey bee population changes, assessing queen-related conditions, and documenting overall hive health. Additionally, health metrics, such as Varroa mite infestation rates and winter mortality assessments were recorded, offering valuable insights into factors affecting hive health status and resilience. In this study, we first outline the data collection process, sensor data description, and dataset structure. Furthermore, we demonstrate a practical application of this dataset by extracting various features from the raw audio to predict colony population using the number of frames of bees as a proxy.

Human well-being relies on the presence and role of pollinators, as they contribute to the vitality of ecosystems, support the reproduction of wild plants, increase crop yields, and strengthen overall food security. While wild bee populations are dwindling due to climate and environmental change, there has been a notable 45% rise globally in the number of managed honey bee (Apis mellifera) colonies over the past five decades. Given their economic significance and their relative ease of tracking, honey bees have the potential to serve as bioindicators of global pollinator health. Consequently, honey bees have emerged as a keystone species requiring protection and conservation efforts. Here, we investigate the intricate relationship between air quality, environmental factors, and honey bee mortality across Canada and the United States. Using statistical and machine learning modeling, our findings underscore the honey bee’s role as a bioindicator. We found that air quality is an important predictor of honey bee mortality. The risk of honey bee mortality increased with poor air quality (ozone and Air Quality Health Index) but was substantially reduced in regions with greater vegetation availability (Normalized Difference Vegetation Index). Therefore, our study offers a beacon of hope: improving management practices by increasing greenery can significantly mitigate the impact of deteriorating air quality on honey bees, providing a vital solution to safeguard our essential pollinators.

Honey bees (Apis mellifera) play an important role in our agricultural systems. In recent years, beekeepers have reported high colony mortality rates in several parts of the world. Inadequate foraging landscapes are often cited as a major factor deterring honey bee colony health. Few studies, if any, have yet used large-scale datasets to assess the quality of landscapes encountered in commercial pollination activities. Here, we coupled a unique dataset comprising georeferenced reports on 17,743 colonies in the province of Quebec, Canada, with data derived from satellite remote sensing, to compute landscape metrics at each visited location. We ran a Cox and a random survival forests (RSF) model with time-weighted features to predict the lifespan of colonies in various landscape scenarios. Survival estimates from our RSF model indicate that colonies foraging primarily in forested areas exhibit higher survival rates, whereas those in cranberry- and maize-dominated landscapes may face lower survival probabilities. Our findings suggest that vegetation abundance could play a significant role in shaping outcomes. Additionally, landscape diversity within a 1 km radius seems to have a positive effect, with potentially greater benefits in areas where vegetation is sparse. While topography contributes valuable predictive insights, its effects are complex and challenging to fully interpret.

Self-supervised learning (SSL) has enabled the development of universal feature extractors that have redefined the performance envelope of computer vision and speech applications. Recent works have started to explore SSL in other domains, including bioacoustics, which could have significant societal benefits. Honeybees (Apis mellifera), for example, are crucial pollinators contributing to one-third of global food production. However, massive colony losses in recent years have raised concerns. Traditional hive monitoring methods rely on intrusive visual inspections by beekeepers, which can further disrupt colony dynamics. As such, IoT-based automated monitoring systems have emerged, integrating environmental and bioacoustic sensing to enable real-time, non-invasive hive assessment. In this work, we introduce a comprehensive evaluation and benchmarking of general-purpose and bioacoustic audio representations that generalize across various tasks in IoT-enabled acoustic beehive monitoring, even with limited labeled data. Herein, fourteen models are evaluated across four critical tasks: beehive state detection, beehive strength assessment, buzzing identification, and beekeeper voice activity detection. Reported results demonstrate the strong generalizability of existing representations, paving the way for advanced, scalable honeybee colony monitoring and preservation.

Honeybees (Apis mellifera) are vital to agriculture and biodiversity, serving as primary pollinators for numerous crops and wild plants. However, the decline in bee populations due to factors such as pesticides, pathogens, parasites, and climate variations poses a serious threat to food security and ecological balance. This study introduces the use of modulation spectral features, extracted from beehive acoustic signals, to better predict the strength of the honeybee colony. Experiments conducted on the public urban beehive acoustics and phenotyping dataset (UrBAN), comprised of over 3000 h of beehive audio recordings, show the proposed features offering improved predictive performance compared with traditional audio features. This work underscores the potential of automated, noninvasive acoustic monitoring systems to support sustainable beekeeping and ecological preservation.