Assessing meteorological key factors influencing crop invasion by pollen beetle (Meligethes aeneus F.) – past observations and future perspectives

The pollen beetle, Meligethes aeneus F. (Coleoptera: Nitidulidae), is a severe pest of winter oilseed rape. A phenological model to forecast the first spring invasion of crops in Luxembourg by M. aeneus was developed in order to provide a tool for improving pest management and for assessing the potential effects of climate change on this pest. The model was derived using long-term, multi-site observational datasets of pollen beetle migration and meteorological data, as the timing of crop invasion is determined mainly by meteorological variables. Daily values of mean air and soil temperature, accumulated sunshine duration and precipitation were used to create a threshold-based model to forecast crop invasion. Minimising of the root mean squared error (RMSE) of predicted versus observed migration dates was used as the quality criterion for selecting the optimum combination of threshold values for meteorological variables. We identified mean air temperature 8.0 °C, mean soil temperature 4.6 °C, and sunshine duration of 3.4 h as the best threshold values, with a cut-off of 1 mm precipitation and with no need for persistence of those conditions for more than one day (RMSE=9.3days$RMSE=9.3\,\text{days}$). Only in six out of 30 cases, differences between observed and predicted immigration dates were >5$>5$ days. In the future, crop invasion by pollen beetles will probably be strongly affected by changes in air temperature and precipitation related to climate change. We used a multi-model ensemble of 15 regional climate models driven by the A1B emission scenario to assess meteorological changes in two 30‑year future periods, near future (2021–2050) and far future (2069–2098) in comparison with the reference period (1971–2000). Air temperature and precipitation were predicted to increase in the first three months of each year, both in the near future and the far future. The pollen beetle migration model indicated that this change would be associated with onset of crop invasion 10 days and 23 days earlier in the near future and far future, respectively, despite the inhibitory effect of increased precipitation. Our phenological model should improve pollen beetle management by provision of more accurate warnings of crop invasion and help stakeholders to recognise and plan for future challenges in pest management