Time linkages between pollination onsets of different <i>taxa</i> in Perugia, Central Italy – an update

Abstract

Introduction and objective. In the last decades, increasing attention has been paid to pollinosis. Numerous studies have been carried out concerning the pollination timing of allergenic plant species and the possibility to forecast its beginning and intensity using several statistical methods and models. This study proposes a simple and fast method to identify in advance the time lapse in which the pollination of some allergenic taxa should start. Materials and methods. The times of pollination of 14 taxa were recorded in the area of Perugia (Central Italy) by means of a 7-volumetric Hirst-Type pollen trap. For a 30-year period (1984-2013), annual starting dates were calculated for each taxa, using the 5% method (Lejoly-Gabriel). The time linkages between these starting dates were then estimated, considering them in pairs and calculating linear regression coefficients. Results. For the significantly linked species, forecasting models were obtained by means of linear regression analysis. To apply these models to the ongoing pollen season, pollination beginning of the earlier species has to be calculated using a sum-based method. From this date, through the obtained equations, it is possible to predict the approximate period in which the pollination of the second linked taxa should start. Conclusions. The possibility to predict the start of the pollen season of these taxa could be of great importance from the allergological point of view. In fact, an early or delayed flowering can have considerable effects in the prophylaxis programming and efficacy.

Recent developments in monitoring and modelling airborne pollen, a review

“…Nevertheless, due to the relative ease of construction regression models remain popular in aerobiological studies and a number of papers have been published in recent years that include simple linear regression analysis (Piotrowska-Weryszko 2013b;Bonini et al 2015;García-Mozo et al 2015;Frenguelli et al 2016;Picornell et al 2019b) as well as polynomial and multiple regression analysis (Sabariego et al 2012;Ocaña-Peinado et al 2013;Oteros et al 2013a;de Weger et al 2014;Donders et al 2014;Novara et al 2016;Ritenberga et al 2016;Tseng et al 2018). Approaches include the more traditional stepwise or backwards elimination multiple regressions (Sicard et al 2012;Aboulaich et al 2013;Myszkowska 2013;Howard & Levetin 2014;Malkiewicz et al 2014;Murray & Galan 2016;Janati et al 2017;Robichaud & Comtois 2017;Galera et al 2018;Volkova & Severova 2019), which are by far the most frequently used, as well as logistic or 'logic' regressions (Escabias et al 2013;Myszkowska 2014b;Myszkowska & Majewska 2014;Katz & Batterman 2019) , Partial Least Squares (Brighetti et al 2014;Oteros et al 2014;Aguilera et al 2015a;Bogawski et al 2019b;Lara et al 2019) and Generalized Linear Model (Charalampopoulos et al 2018).…”

“…Regression analysis continues to be used to model daily average airborne pollen concentrations (Janati et al 2017) in addition to different characteristics of the pollen season including start date (Myszkowska 2014a;Novara et al 2016), peak day (Myszkowska 2013), duration (Zhang & Huang 2015) and intensity (Oteros et al 2013a;Bonini et al 2015). Regression models have been constructed for different pollen types including Alnus (Piotrowska-Weryszko 2013b;Myszkowska 2014b;Novara et al 2016), Ambrosia (Howard & Levetin 2014;, Artemisia (Piotrowska-Weryszko 2013a;, Betula (Myszkowska 2013;Inatsu et al 2014;Robichaud & Comtois 2017;Tseng et al 2018;Bogawski et al 2019b), Corylus (Myszkowska 2014b;Frenguelli et al 2016;Novara et al 2016), Cupressaceae (Sabariego et al 2012Ocaña-Peinado et al 2013;Charalampopoulos et al 2018;Picornell et al 2019b), Olea (Sicard et al 2012;Oteros et al 2013a;Frenguelli et al 2016;Charalampopoulos et al 2018;Picornell et al 2019b), Platanus (Frenguelli et al 2016;Charalampopoulos et al 2018;Picornell et al 2019b), Poaceae (Piotrowska 2012;Aboulaich et al 2013;de Weger et al 2014;Janati et al 2017;Picornell et al 2019b…”