Abstract

Introduction Urticaceae family comprises herbs and small shrubs mostly represented in Spain by two genera: Urtica and Parietaria. Both taxa are wind-pollinated and are found in the countryside and urban areas, growing on walls and soils rich in nitrogen. Urticaceae pollen is present in the air near all year round with the highest concentrations between March and November. This taxon is causing about 16 % of the pollinosis in Spain; 23 % in the Mediterranean Spanish region1. Objectives Our aim is to establish logistic regression models to predict the presence of airborne Urticaceae pollen and the exceedance of significant concentration thresholds. Material and methods Our study is based on data from years 1995-2013 of eight aerobiological monitoring stations in Catalonia (NE Spain). Daily maximum and minimum temperatures, rainfall and Urticaceae pollen concentrations of the period 1995-2012 are used to establish four logit regression equations applied to predict the exceedance of thresholds 1, 4, 8 and 12 pollen/m³. Data from 2013 are used to validate the models. Results Pollen levels correlate positively with minimum and range temperatures and cumulative rainfall, and negatively with rainfall in the same day. The logistic regression models applied predict with 60-78 % confidence the presence of airborne Urticaceae pollen; levels ≥4 pollen/m3 with 47-79 %; levels ≥8 with 65-79 % and levels ≥12 pollen/m3 with 74-88 %. Conclusions The prediction of airborne Urticaceae pollen thresholds by means of logistic regression models is a simple and practical process, with reliable estimation values, recommended to be used in aerobiological information systems.