Assessing nest sites, survival rates and population dynamics of free-living honeybee colonies in Germany: A comparative study using personally collected and Citizen Science data

Our understanding of the western honeybee ( Apis mellifera ) predominantly stems from studies conducted within beekeeping environments, leaving the presence and characteristics of honeybees outside managed settings largely unexplored. This study focuses on free-living colonies, examining their habitats, nesting sites, and survival rates, thereby highlighting the liminal state of the species in Europe. Through the BEEtree-Monitor project, we personally monitored (PM) nest sites in Munich (N=107) and coordinated Citizen Science (CS) monitoring across Germany (N=423), resulting in 2,555 observations on 530 colonies from 311 participants over 7 years. While habitat preferences differed between urban, rural and forested areas, we found that 31% of the occupied nest sites were in buildings and 63% in mature trees with cavities, including lime ( Tilia spp., 18%), beech ( Fagus sylvatica , 14%), oak ( Quercus spp., 13%), and ash ( Fraxin us excelsior, 11%). On average, only 12% of the PM colonies in Munich survived annually, according to model predictions based on observed data. Consequently, we found a minimum free-living colony density per km² of 0.06 in spring but 0.42 in summer, accounting for at least 4% of the total Munich population during the summer. Comparing the two monitoring approaches (PM vs. CS) and published data revealed significant discrepancies: survival rates reported by CS were markedly higher (model prediction: 28%), than PM and other published studies. We found that CS not only yielded significantly fewer updates per colony, but also that 76% of CS reports noted active colonies compared to 42% from PM, indicating an underreporting of abandoned sites. To ensure the reliability of survival data in CS projects, regional swarming should be monitored, and the timing of reports needs particular attention as 46% of reports about overwintering were too late (i.e. after the onset of the swarming season; 11% in PM). CS data proved to be highly valuable for nest site and habitat analysis but require a strict validation protocol for survival statistics. This study enhances our understanding of the ecological dynamics and conservation needs of free-living honeybee cohorts, addresses potential monitoring biases, and suggests standardized data collection protocols for future monitoring projects.