A prediction of the beginning of the flowering of the common hazel in the Czech Republic

Abstract

Timely information on the beginning of the flowering of important plant species of pollen allergens is consequential for the entire population due to pollen allergy and its extensive clinical impact worldwide. This paper examines the prediction of the beginning of the flowering of the common hazel (Corylus avellana) based on the PhenoClim phenological model using long-term phenological observations (1991–2020) in the Czech Republic. Furthermore, temporal and spatial evaluations of the beginning of the flowering of C. avellana were examined in different climate zones in the Czech Republic within the same period. In total, 40 phenological stations at altitudes from 155 to 743 m asl located in warm, medium warm, and cold climate zones were evaluated using the Mann–Kendall test. The beginning of the flowering of C. avellana changed progressively in timing, and the difference in the rate of shifts was between −33 and + 15 days per the entire period. An extreme shift to an earlier date was detected at stations located in a warm region (W2). In contrast, the highest shift to a later date was found at stations located in the cold climate regions (C4, C6, C7). Using the PhenoClim, the base temperature and temperature sums were calculated for the beginning of the flowering of the common hazel. As the most accurate predictor for this phenological phase and species, the maximum air temperature was determined as the best predictor based on the combination of RMSE and R² values. The optimal start day for calculation was January 1st; the threshold (base temperature) was 2.7 °C with a temperature sum of 155.7 °C. The RMSE value was 5.46, and the MBE value was −0.93. The simulated data showed an excellent correlation with the observed data—the correlation coefficient was 0.932. The PhenoClim model results can be used in the forecast modelling of the beginning of the flowering of the common hazel in the Czech Republic.