Assessment and statistical modelling of airborne microorganisms in Madrid

© 2020 Elsevier Ltd. All rights reserved. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/. This document is the Accepted version of a Published Work that appeared in final form in Environmental Pollution. To access the final edited and published work see https://doi.org/10.1016/j.envpol.2020.116124

The limited evidence available suggests that the interaction between chemical pollutants and biological particles may intensify respiratory diseases caused by air pollution in urban areas. Unlike air pollutants, which are routinely measured, records of biotic component are scarce. While pollen concentrations are daily surveyed in most cities, data related to airborne bacteria or fungi are not usually available. This work presents the first effort to understand atmospheric pollution integrating both biotic and abiotic agents, trying to identify relationships among the Proteobacteria, Actinobacteria and Ascomycota phyla with palynological, meteorological and air quality variables using all biological historical records available in the Madrid Greater Region. The tools employed involve statistical hypothesis contrast tests such as Kruskal-Wallis and machine learning algorithms. A cluster analysis was performed to analyse which abiotic variables were able to separate the biotic variables into groups. Significant relationships were found for temperature and relative humidity. In addition, the relative abundance of the biological phyla studied was affected by PM10 and O3 ambient concentration. Preliminary Generalized Additive Models (GAMs) to predict the biotic relative abundances based on these atmospheric variables were developed. The results (r = 0.70) were acceptable taking into account the scarcity of the available data. These models can be used as an indication of the biotic composition when no measurements are available. They are also a good starting point to continue working in the development of more accurate models and to investigate causal relationships.