Plants on the ground, pollen in the air: how much do they match?

Urban green spaces are a vital element of sustainable cities. Nonetheless, there are associated disservices, one most important being pollen induced allergies. To examine how much vegetation analysis of urban green spaces can be an efficient indicator of the pollen-related qualitative and quantitative features of their atmospheric environment, we studied six such spaces, in Thessaloniki, Greece. We made a full analysis of their woody vegetation and collected aerobiological data, with sampling at breast height. Cupressaceae, Platanus, Quercus, Pinaceae and the herbaceous Urticaceae were the main pollen providers in almost all of them, when the main woody components of their vegetation were Cupressaceae, Rosaceae, Pinaceae and Fabaceae, with Quercus having only sporadic occurrence. The number of taxa represented in pollen and vegetation were not correlated, and pollen from external sources was detected even at high concentrations. Pollen similarity was higher than vegetation similarity, with taxa identity being more important than abundance in differentiating the green spaces. Pollen incidence was synchronized in many cases but, like concentration, duration of the pollen season also varied largely among green spaces, even when in proximity. Positive relationships between pollen concentration and vegetation abundance were detected for a few taxa, primarily Cupressaceae, and for green spaces that covered a large area (around 40 ha) or had an element of isolation. Vegetation analysis is not a reliable indicator of the pollen related atmospheric environment at the local scale. Aerobiological surveys are additionally needed locally to provide the necessary information regarding the prevailing conditions and the associated risks.