The aim of the study was to determine functional relationships between composition of air spora and meteorological factors, using multivariate statistical technique: canonical correspondence analysis (CCA). Analyses were conducted for the data collected during the 4 year (2007–2010) and, in order to show the dynamics of such relationships, for each year separately. The CCA results indicated that all statistically significant variables accounted for 15.3 % of the total variance in the spore data in the 4 years. The largest amount of the total variance was explained in this period by the mean air temperature (9.2 %). The meteorological factors impacted spore composition differently in different years, when analysis was done for each year separately. The highest values of the total variance in the spore data, explained by the statistically significant variables, were found in 2010 (32.3 %), with the highest contribution of mean air temperature (23.8 %). In that year, the above-mentioned parameter had the lowest value in comparison to other years. Canonical correspondence analysis provides not only a comprehensive assessment of the impact of meteorological factors on specific spore combinations in the air, but also informative graphical presentations of the results, illustrating the correlation between the occurrence of particular spore taxa and meteorological variables.

Fungal spores are known to cause allergic sensitization. Recent studies reported a strong association between asthma symptoms and thunderstorms that could be explained by an increase in airborne fungal spore concentrations. Just before and during thunderstorms the values of meteorological parameters rapidly change. Therefore, the goal of this study was to create a predictive model for hourly concentrations of atmospheric Alternaria and Cladosporium spores on days with summer storms in Szczecin (Poland) based on meteorological conditions. For this study we have chosen all days of June, July and August (2004–2009) with convective thunderstorms. There were statistically significant relationships between spore concentration and meteorological parameters: positive for air temperature and ozone content while negative for relative humidity. In general, before a thunderstorm, air temperature and ozone concentration increased, which was accompanied by a considerable increase in spore concentration. During and after a storm, relative humidity increased while both air temperature ozone concentration along with spore concentrations decreased. Artificial neural networks (ANN) were used to assess forecasting possibilities. Good performance of ANN models in this study suggest that it is possible to predict spore concentrations from meteorological variables 2 h in advance and, thus, warn people with spore-related asthma symptoms about the increasing abundance of airborne fungi on days with storms.

Air sampling was conducted in Szczecin (Poland) throughout April–September 2013. The final data set included 177 daily and 4248 hourly samples. The total of 21 types of spores, which occurred in a number >10 in the season, were taken into account. The following meteorological parameters were analyzed: air temperature, relative humidity, precipitation and wind speed. Effects of individual weather parameters on hourly and daily concentrations of different fungal spore types were examined using Spearman’s rank association test, whereas effects of complex of meteorological factors on hourly and daily compositions of spore were assessed using detrended correspondence analysis (DCA) and redundancy analysis (RDA). Airborne fungal spore distribution patterns in relation to meteorological variables were determined by RDA, after DCA results detected a linear structure of the spore data. The RDA results obtained indicated that all the applied variables accounted for 20 and 22% of the total variance in the hourly and daily spore data, respectively. The results of stepwise forward selection of variables revealed all included hourly and daily meteorological variables were statistically significant. The largest amount of the total variance in the spore composition was explained by the air temperature in both cases (16%). Multivariate ordination did not show large differences between the hourly and daily relationships (with exception of wind speed impact), while the differences between simple hourly and daily correlations were more clear. Correlations between daily values of variables were in most cases higher than between hourly values of variables.