A new methodological approach to analyse intradiurnal pollen patterns in protected areas: the case of Sierra de las Nieves National Park (southern Iberian Peninsula)

Abstract Analysing pollen dispersal dynamics in natural ecosystems is essential to unravel the effects of meteorological conditions on pollen dispersion and decipher the reproductive strategies of threatened species. However, most aerobiological studies have been conducted in cities. In this study, the intradiurnal airborne pollen dynamics of Sierra de las Nieves National Park (southern Spain) were studied by applying a novel methodology based on the combination of decision trees and clustering to identify the meteorological drivers of intradiurnal pollen dispersion. To that end, pollen patterns were studied in the main pollen types of this protected area during 2018–2024. The days with high pollen detection for each pollen type were clustered according to their intradiurnal pattern. The meteorological conditions of the grouped days were analysed using decision tree algorithms to identify possible causes of their intradiurnal pollen pattern. The highest pollen detection usually occurred around 12:00 and 14:00 h. Most pollen types exhibited a peak during daylight, corresponding to the typical pattern of local pollen sources. Some pollen types, such as Castanea and Urticaceae, exhibited a nocturnal peak characteristic of distant pollen transport. Most pollen types had two different intradiurnal patterns triggered by different meteorological conditions, except Plantago, Poaceae and Quercus . The most relevant variables determining the intradiurnal patterns observed were the frequency of winds blowing from the northwest and northeast quadrants, relative humidity and maximum temperatures. Combining cluster analysis with decision trees proved to be of great utility to analyse the influence of weather conditions on intradiurnal pollen patterns.