Causal network linking honey bee (Apis mellifera) winter mortality to temperature variations and Varroa mite density

Abstract

Winter season is a critical time for honey bees (Apis mellifera) colonies when individual mortalities may lead to total colony losses or diminish productivity in subsequent seasons. A deeper understanding of the causes and consequences of winter mortality is required. In this study, we analyzed winter (November–March) individual bee mortality in an apiary in Central Europe from 1991 to 2023. We observed consistency in mortality times among years, but also some systematic departures from the shared trend. We distinguished four clusters of year-specific mortality trajectories. However, we found no statistically significant differences in means of spring (March–May), autumn (October), winter (November–March) temperatures, or autumn Varroa destructor density among clusters. Nevertheless, our insights into the dynamics of individual bee mortality may be important for determining critical moments during wintering when implementing additional protective measures could prove beneficial. Hypothesis-driven path analysis indicated causal links in our study system, including both direct and indirect influences. The density of V. destructor in autumn was positively related to temperature, especially in the preceding spring, but to a lesser extent also in autumn. Increased winter mortality was related to lower winter temperatures and a higher mite infestation in autumn. We found no significant effects of individual winter mortality on honey harvests in subsequent seasons. Honey harvest was determined by bee abundance in spring, and the latter, unexpectedly, was not related to winter mortality. Our study adds to accumulating evidence of the major role of weather and climatic conditions in the resilience of honey bee colonies and improves our understanding of mortality processes. We highlighted the importance of causative factors, especially seasonal temperatures and V. destructor density, and their potential as predictive indicators of individual winter mortality, bee colony fate, and honey productivity.