Time series analysis of Varroa destructor counts in Ontario honey bee colonies and their association with weather variables

The parasitic mite, Varroa destructor, is widely considered the most important risk factor for honey bee colony health in Canada, consistently associated with colony loss and poor colony health. Examining the temporal epidemiology of Varroa mites is crucial for identifying population-level trends, understanding seasonal patterns, and evaluating potential associations with external risk factors. This study examines the temporal patterns of observed Varroa in Ontario, Canada, over a five-year period (2015-2019), using provincial ministry inspection data. Through time-series decomposition and regression modelling, seasonal patterns and long-term trends in mite counts were described, with tests for associations with historical weather data, both instantaneous and lagged, to account for delayed effects on mite counts. A repetitive seasonal pattern and a slight decreasing trend were observed in mite counts throughout the duration of the study. Associations with ambient temperature and dew point temperature were observed when a seven-week lag was applied. These results provide an epidemiological perspective on Varroa mite infestations over time, offering valuable insights for surveillance by establishing a reference for expected mite levels.