Predicting tree pollen season start dates using thermal conditions.

Dorota Myszkowska

doi:10.1007/s10453-014-9329-3

Abstract

Thermal conditions at the beginning of the year determine the timing of pollen seasons of early flowering trees. The aims of this study were to quantify the relationship between the tree pollen season start dates and the thermal conditions just before the beginning of the season and to construct models predicting the start of the pollen season in a given year. The study was performed in Krakow (Southern Poland); the pollen data of Alnus, Corylus and Betula were obtained in 1991–2012 using a volumetric method. The relationship between the tree pollen season start, calculated by the cumulated pollen grain sum method, and a 5-day running means of maximum (for Alnus and Corylus) and mean (for Betula) daily temperature was found and used in the logistic regression models. The estimation of model parameters indicated their statistically significance for all studied taxa; the odds ratio was higher in models for Betula, comparing to Alnus and Corylus. The proposed model makes the accuracy of prediction in 83.58 % of cases for Alnus, in 84.29 % of cases for Corylus and in 90.41 % of cases for Betula. In years of model verification (2011 and 2012), the season start of Alnus and Corylus was predicted more precisely in 2011, while in case of Betula, the model predictions achieved 100 % of accuracy in both years. The correctness of prediction indicated that the data used for the model arrangement fitted the models well and stressed the high efficacy of model prediction estimated using the pollen data in 1991–2010.

Highlights

Temperature is one of the major environmental factors that determines the readiness of plants to flower
Betula pollen seasons are more stable in comparison with Alnus and Corylus pollen seasons, which is confirmed by the coefficients of variation (7.3, 46 and 46.2 %, respectively)
Compared to the models proposed for Alnus and Corylus, the model for the Betula accurately predicted that the season would start in 90.41 % of cases (Table 3)

Summary

Introduction

Temperature is one of the major environmental factors that determines the readiness of plants to flower. The observation-based models refer to the relationship between the pollen season start day and many meteorological elements, such as air temperature, rainfall, sunshine and snow cover occurring both before the pollen season and in the preceding season (Emberlin et al 1993; Norris-Hill 1998; Corden et al 2000; AdamsGroom et al 2002; Rodriguez-Rajo et al 2004). These models are constructed without knowledge of the sources, emission or calculations of diffusion (NorrisHill 1995). Some papers on the thermal conditions required for the pollen season start dates of Corylus (Frenguelli et al 1991; Frenguelli and Bricchi 1998), Alnus (Jato et al 2000; Clot 2001; Rodriguez-Rajo et al 2006; Emberlin et al 2007) and Betula (Linkosalo et al 2010) were published

Objectives

Methods

Results

Discussion

Conclusion