1 Introduction

Since the first reported case of asthma attack caused by fungal spores in 1726 (Floyer 1745), more than 112 fungi are now recognised as allergens (Levetin et al. 2015; Twaroch et al. 2015). This list includes taxa such as Alternaria sp., Aspergillus sp., Penicillium sp., Cladosporium sp., Botrytis sp., Drechslera sp., Epicoccum sp., Leptosphaeria sp., Pithomyces sp., Pleospora sp. and Stemphylium sp. (Green et al. 2005; Kurup et al. 2000). A Pan-European survey, carried out by the Global Allergy and Asthma European Network (GA2LEN), determined that sensitisation rates towards Alternaria alternata (17.6 %) and Cladosporium herbarum (6.8 %) were the highest in the UK and Ireland compared to other northern and southern European countries (Heinzerling et al. 2005). In the UK, it is considered that exacerbation of allergic reactions may occur when concentrations of Alternaria sp. and Cladosporium sp. spores exceed 50 and 3000 spores per cubic metre of air, respectively (Frankland and Davies 1965). However, medical thresholds for exposure have yet to be established for the other fungal genera that are known to be allergenic (Green et al. 2005). A study by Denning et al. (2006) reported that asthma killed 1500 patients annually in the UK. Although the overall contribution of fungal sensitisation towards the death rate has not yet been established, it has been observed that mortality in young adults in the UK coincided with increased levels of fungal spores in the air between July and September (Khot and Burn 1984). Typical allergic responses include sneezing, running nose, wheezing, tightness of chest, coughing, shortness of breath, urticaria, angioedema, anaphylaxis, pruritus and nasal obstructions (Kurup et al. 2000, 2002; Simon-Nobbe et al. 2008). Such allergic symptoms cause a considerable negative impact on the quality of human health (Bousquet et al. 2001).

A pilot study (unpublished) with regard to fungal and pteridophyte spore levels was previously carried out at Worcester in 2005, where a non-standard spore counting method was initially used. The method enumerated fungal spores within a graticule (0.50 × 0.50 mm) from every third field of view at ×400 magnification, which resulted in the overall analysis of 0.81 % of the total microscope slide area. The correction factor (CF), calculated for specific microscope that was used, was equal to 31. This means that a single spore was multiplied by this CF in order to get a spore concentration per cubic metre of air. Often, it happened that a single spore, for example of Stemphylium sp. or Chaetomium sp., was found on the entire slide. Thus, the concentrations for these genera were greatly overestimated. In case of Stemphylium sp. such error was completely unacceptable, since the fungicide application for the brown spot, common diseases of pear, must be carried out when concentration of Stemphylium vesicarium conidia in the air exceeds 30 s m−3 (Rossi et al. 2005). Frenguelli (2003) suggested that the examined area of the microscope slide should not be smaller than 10–12 %, in relation to the studies on pollen grains. However, no previous study has reported a representative percentage of the examined area for air samples collected using the Hirst type traps in relation to fungal spores. Moreover, fungal spores were enumerated by several different operators that could possibly result in additional sources of error and inconsistency at the spore identification level. Hence, there was a necessity for an extension of data and application of a standard spore counting methods that would enable greater accuracy and comparability with other reports on fungal spore levels (Nilsson and Persson 1981; Stępalska and Wołek 2009).

Recently published studies for the Worcester area have been limited in their extent either because the analysed period of time was very short (O’Connor et al. 2014; Sadyś et al. 2015a), or it focused on a single spore type (Sadyś et al. 2014; 2015a, b) or very few fungal taxa (Sadyś et al. 2015c). Studies describing airborne concentrations of allergenic spores for other areas of the UK are also limited in their extent with the majority published between the 1950s and 1980s and nothing since 2001 (Corden and Millington 2001; Frankland and Davies 1965; Hirst 1952; Hyde and Williams 1959; Jenkins et al. 1980; Lacey 1981; Morrow Brown and Jackson 1978). A “bridge” type study that would link all previously published reports was desirable as, until now, a more detailed overview for a wider range of fungal taxa performed for a longer period of time was missing. Therefore, this study aimed to present results of a 5-year survey, where 20 clinically important fungal taxa were studied followed by comprehensive analysis of their dependence on changing weather conditions.

2 Materials and methods

2.1 Site location

The city of Worcester is located in Worcestershire, in the West Midlands, England (O’Connor et al. 2014). The city population was estimated to be 98,800 people with population density at 2969 people per km2 (Rice 2011). The composition of green urban areas includes public parks, allotments, racecourse, cricket field, cemetery and grasslands. An important element in the landscape of the city is the River Severn that divides the city. Overall, the green urban areas occupy 1.08 %, while non-irrigated areas 11.84 % and pastures 11.12 % of the city area (Fig. 1).

Fig. 1
figure 1

Spatial plan of the city of Worcester, England, showing the location of the sampling site (Corine Land Cover 2006)

Full size image

2.2 Air sample collection

The sampling of airborne fungal spores was conducted continuously from 1 Jan 2006 to 31 Dec 2010. The 7-day volumetric spore trap (Burkard Manufacturing Co. Ltd., Rickmansworth, Hertfordshire, UK) was situated on the roof of the University of Worcester (52° 11′ N, 2° 14′ W) approximately 10 m above ground level. The air sampler worked by impacting particles present in the air, directly on to collection tape, which was tightly fastened to the drum inside the sampler and covered by an adhesive medium (Hirst 1952). The pump located in the lower part of the sampler produced an air pressure with a suction power equal to 10 l of air per minute. Drum rotated with a speed of 2 mm h−1 due to a built-in clock mechanism and was changed weekly, on the same day at 09:00 UTC. Tape was then removed from the drum in the laboratory and cut into seven 48-mm long segments that corresponded to each of the 7 days of sampling. Each sample was mounted with glycerogelatin mountant and lactophenol Cotton Blue stain and sealed using nail varnish to prevent drying of the samples.

2.3 Spore identification and taxa selection

Fungal spores were identified at the genus level with the aid of the light microscope (Nikon Eclipse E400) and counted along one central longitudinal sweep under ×400 magnification (Nilsson and Persson 1981; Stępalska and Wołek 2009). This counting method is very popular among previously published spore work, and thus, it was chosen in order to make current study comparable with others (e.g. Skjøth et al. 2015). Then spore counts were multiplied by a correction factor resulting in an estimation of number of spores per cubic metre of air (s m−3) after Lacey and West (2006). The concentration of 20 types of fungal spores were selected upon (1) their allergenic properties and (2) abundance in the air based upon the results of a pilot study conducted in 2005 and comprised: Alternaria sp., Aspergillus sp./Penicillium sp., Blumeria sp. (previously Erysiphe sp.), Botrytis sp., Cladosporium sp., coloured basidiospores, Didymella sp., Drechslera sp., Entomophthora sp., Epicoccum sp., Ganoderma sp., Leptosphaeria sp., Periconia sp., Pithomyces sp., Pleospora sp., Polythrincium sp., rusts, smuts, Stemphylium sp., Torula sp. Some fungal spores were counted jointly because visual distinction was not possible between genera, i.e.: Aspergillus sp. and Penicillium sp., or they were classified as a “group”, i.e.: coloured basidiospores, rusts and smuts. This is established practice by many aerobiologists (e.g. Lacey and West 2006). Occasionally, when classification became problematic, several different fungal spore atlases were used to determine affinity to a certain genus or group (Ellis 1971; Grant Smith 1990; Lacey and West 2006; Ogden et al. 1974) under ×1000 magnification. The names of investigated fungal genera (groups) were abbreviated to their names only in the text. The data analysis and control was performed under the ISO/IEC 17025. Discussion on the limitations of aerobiological surveys and possible errors encountered at various stages of data sampling and analysis can be found in Comtois (1998), Galán et al. (2014), Oteros et al. (2013), Sterling et al. (1999), Stępalska and Wołek (2009).

2.4 Weather data

The weather data were collected using the WeatherLink (9.0) Vantage Pro2 station (Davis, Davis Instruments, Hayward, California, United States), which was co-located with the spore trap. The influence of the following parameters was studied: maximum temperature, minimum temperature, mean temperature, dew point temperature, wind direction, relative humidity, rain, air pressure, leaf wetness and leaf temperature. As the wind was initially recorded in 16 cardinal directions, it had to be re-calculated to the degree format to enable statistical examination (Sadyś et al. 2015c). Due to changes applied in the operation of the weather station, leaf wetness and leaf temperature were not recorded in 2010.

2.5 Statistical analyses

Fungal spore seasons were calculated using the 90 % method introduced by Nilsson and Persson (1981), which defines beginning of the season as a date when 5 % of the accumulative daily spore concentration was exceeded, and the end of the season, when 95 % of the accumulative daily spore concentration was reached. Seasonal Fungal Index (SFI), which constitutes a sum of daily mean spore concentration recorded within a spore season, was also calculated. Distribution of the spore data was analysed by application of three tests for normality, i.e.: Kolmogorov–Smirnov test, Kolmogorov–Smirnov test with Lilliefors correction and Shapiro–Wilk test. The nature of the species-environment dependences was tested by visual evaluation of the plotted correlation graphs. A Spearman’s rank test was used to examine the relationship between concentration of investigated fungal spores and selected meteorological parameters. Since none of the investigated fungal spore seasons were normally distributed, the Kruskal–Wallis test was therefore used to examine the difference between annual sums of daily mean spore concentration, after Del Mar et al. (2000). The statistical analyses were performed with the aid of Microsoft Excel (2010), Statistica StatSoft (2012) and GenStat (17) software. The spore calendar was constructed upon a 5-year mean of monthly sums of daily spore concentration for each studied taxon. Spore counts were then presented on a logarithmic scale in order to enable comparison of extremely low or high spore counts simultaneously (Nikkels et al. 1996). Finally, a simple linear regression was used in order to detect any trend in the overall 5-year SFI.

3 Results

3.1 Overall fungal spore contribution and annual changes

The list of 20 types of fungal spores that were analysed in the air samples collected in Worcester and their 5-year mean distribution is presented on a logarithmic scale in Fig. 2. Cladosporium spores were the most highly occurring type out of the 20 selected taxa, with the peak recorded in August. The second largest group was coloured basidiospores, followed by Didymella ascospores with maximum concentrations observed in July and September, respectively (Fig. 2). The spore load of the most allergenic fungal spores (Alternaria) was comparable in number to Aspergillus/Penicillium and Botrytis spores (Fig. 2), while Ganoderma basidiospores were observed at twice this concentration and Leptosphaeria and smuts at four times as much (Fig. 2). Other taxa, such as Periconia, Pithomyces, Pleospora, Stemphylium and Torula, occurred in much lower concentrations (Fig. 2).

Fig. 2
figure 2

5-year mean distribution of examined spore types presented on a logarithmic scale. Peak months for each taxon are highlighted in red

Full size image

The Kruskal–Wallis test results (Table S1) show that, except for coloured basidiospores and Torula, all examined spores showed annual fluctuations in their distribution patterns. The simple linear regression analysis results (Table 1) revealed that 55 % of all examined taxa showed a reducing trend, 20 % remain unchanged and only 25 % exhibited increasing trend. However, only one trend, for Entomophthora, was found to be statistically significant (Table 1).

Table 1 Linear regression trend analysis for annual changes in Seasonal Fungal Indices
Full size table

3.2 Fungal spore season characteristics

The summary statistics of main spore seasons were calculated for all examined fungal genera (Table 2). The three most frequently found types of those investigated in the air of Worcester during the study period are as follows: Cladosporium, coloured basidiospores and Didymella.

Table 2 Characteristics of investigated fungal spore types
Full size table

Cladosporium occurred in quite short spore seasons as their duration varied from 111 to 186 days (Table 2, Fig. 2). The daily peak values were approximately between 20,000 and 47,000, and they mainly occurred in July. In years 2006 and 2010, there were marked differences in comparison with the remaining 3 years of observations.

Coloured basidiospores showed stable and uniform seasons, which started in May and ended in November, with a duration lasting approximately 180 days (Table 2; Fig. 2). The only exception was found in 2010, as the length of that particular season was shorter by 40 days. Maximum spore concentrations were recorded mostly in September and October, with daily peaks oscillating between 3000 and 4600 spores per cubic metre of air. Coloured basidiospores occurred in a bi-annual pattern in their distribution, as they were found in significantly greater quantities in the years 2006, 2008 and 2010.

Didymella ascospores occurred during the shortest time periods of between 64 and 90 days (Table 2; Fig. 2). Seasons started either by the end of June or at the beginning of July, and finished mainly in September of each year. Although duration of the seasons was short, their intensity was revealed to be substantial, as the SFI scored from 26,000 to 190,186. Years 2007 and 2008 had the greatest concentration of Didymella ascospores during the 5-year monitoring at Worcester. The maximum spore concentration of Didymella ascospores recorded within a 24-h period was approximately 20,000 s m−3.

3.3 Spore co-occurrence and high spore count days

The daily average concentrations were examined towards the co-occurrence between all studied taxa (Fig. 3). Two major groups of co-occurring spore types were identified: (a) Alternaria, Cladosporium, Drechslera, Epicoccum, Polythrincium, and (b) Didymella and Leptosphaeria (Fig. 3). The strongest associations were found between Ganoderma and Polythrincium (r = 0.71), as well as between Alternaria and Epicoccum (r = 0.69) spores. In contrast, unique distribution patterns showed Aspergillus/Penicillium group and Pleospora spores, which correlation coefficient values did not exceed r = ±0.12.

Fig. 3
figure 3

Co-occurrence of all examined fungal spore types; average daily spore concentrations were tested. Colour indicates level of correlation between fungal taxa, where brown colours stand for r s  = 1, and navy blue for r s  = 0

Full size image

The number of days when established clinically allergenic thresholds were exceeded for Alternaria (≥50 s m−3), Cladosporium (≥3000 s m−3) and for both spore taxa simultaneously, are presented in Fig. 4. The overall contribution of high spore count days towards the spore seasons varied annually, and in all cases was found to be the highest in 2006 (Fig. 4c). Alternaria high spore count days constituted between 25 % and 45 % (Fig. 4a) and Cladosporium 32–74 % of the 5-year sampling period (Fig. 4b).

Fig. 4
figure 4

Number of days within each year when allergenic thresholds were exceeded during the spore survey for: (a) Alternaria ≥50 s m−3, (b) Cladosporium ≥3000 s m−3 and (c) both taxa

Full size image

3.4 Spore dependence on weather

Overall, the Spearman’s rank test (Table 3) showed that statistically significant (p ≤ 0.05) relationships between spore concentration and meteorological parameters were mainly found below r s  = 0.50.

Table 3 Spearman’s rank correlation coefficients between fungal spore concentrations and meteorological parameters
Full size table

The presence and high quantities of Cladosporium spores trapped at Worcester were shown to be correlated with all temperature parameters. The vector of these relationships was directly proportional (Table 3). Relative humidity was negatively correlated with spore counts.

Relative humidity was positively correlated with the concentration of coloured basidiospores in the air at the sampling site (Table 3). Statistically significant correlations were also observed with temperature, leaf wetness and air pressure (Table 3).

Didymella ascospores were strongly correlated with relative humidity, rainfall and leaf wetness (Table 3). The Spearman’s rank coefficients varied between r s  = 0.35 and r s  = 0.68, while the influence of all temperature parameters were much lower if statistically significant at all. The direction of prevailing winds recorded in 2006 influenced to some extent the Didymella ascospore counts observed in Worcester (r s  = 0.22).

4 Discussion

This study was performed throughout a 5-year period, including two of very contrasting weather conditions in 2006 and 2007 that had a great impact on observed spore levels (Sadyś et al. 2015d). Several fungal spore types did not show any statistically significant correlation with prevailing wind directions recorded at Worcester, i.e. Blumeria, Didymella, Leptosphaeria, Periconia, Pithomyces, smuts and Stemphylium (Table 3). Based on the results, there is a possibility that some of those spores could originate from local sources. This hypothesis could also be supported by the high values of Spearman’s rank coefficients, e.g. the Blumeria correlation with maximum temperature, Didymella and Leptosphaeria correlation with rainfall, or smuts with relative humidity (Table 3). In contrast, the origin of Periconia, Pithomyces, smuts and Stemphylium cannot be fully explained in a similar manner, as further analysis showed that most of the associations were not strong or even statistically significant. However, this could be caused by low spore counts recorded in the air samples (Fig. 2); hence, outcomes of the association test were likewise poor. Perhaps, due to certain features, like size, shape or surface structures some spores may not become easily airborne. For instance, the size of Stemphylium spores may be between 50–200 and 3–10 μm, and their shape varies from oblong to ellipsoidal and often the wall surface is rough, as it contains verruca or prickles (Ellis 1971). Another explanation may be the fact that an elongated shape accelerates the deposition, as they fall horizontally with the long axis, which has been reported for Alternaria spores (Lacey 1991).

The only long-term spore data set collected in the UK based upon volumetric sampling is available for Derby, located in the East Midlands of England (MAARA 2004). Results of spore surveys carried out both in Derby and Worcester showed the dominance of Cladosporium spores in the air, with maximum concentrations recorded between July and August (Fig. 2; MAARA 2004). Similar trends in spore distributions were found for Alternaria and Leptosphaeria with peaks observed in August. Smuts (Ustilago) peaks occurred in June and Blumeria maximum concentrations were recorded between June and August. Simultaneously, other taxa showed 1–2 month differences in their occurrence between sites. In Worcester, taxa such as Aspergillus/Penicillium type, Didymella and Polythrincium were sporulating later compared to Derby, while at the same time Epicoccum was sporulating earlier. Some of these differences could be caused by the distance between spore source and location of the aerobiological station, where these spores were trapped. An application of atmospheric models, such as Hybrid Single-Particle Lagrangian Trajectory (HYSPLIT), has already demonstrated and explained potential transport of bioaerosols at the regional and long-distance scale along with moving air masses (Sadyś et al. 2014; 2015b; Skjøth et al. 2012), although the differences in local and regional climate and land use should not be neglected (e.g. Kasprzyk et al. 2015).

The prevalence of Cladosporium spores above other taxa was also found in other European sites, for example in Rzeszów (Poland), Timişoara (Romania) and Leiden (Netherlands) with a contribution varying between 41 and 93 % (Ianovici and Tudorica 2009; Kasprzyk 2008; Nikkels et al. 1996). This is a characteristic feature for the temperate climate (D’Amato and Spieksma 1995) in contrast to tropical climates, where the key role is played by Aspergillus/Penicillium taxa (Desbois et al. 2006).

Results of the trend analysis showed that although only 25 % of investigated taxa had an increase in their spore count throughout a 5-year survey (Table 1, Fig. 4), the Cladosporium genus was found among this group. Taking into account that Cladosporium along with Alternaria are the most allergenic fungal genera (D’Amato and Spieksma 1995), this is a trend that should be carefully monitored. The result of Ganoderma distribution fits well with previous findings reported by Sadyś et al. (2014), who examined changes in Ganoderma spore distribution in combination with air mass and land cover analysis. They found that 78 % of the air masses reached Worcester from a 180° arc direction from the East to the West, where large complexes of broadleaf and mixed forests (Ganoderma hosts) were located. Assuming that forest areas did not experience any significant damage in the form of a fire, pest, etc. during the period under investigation and the overall wind direction remained constant due to the terrain orography, it would be expected that there would be no significant change in the overall Ganoderma SFI (Table 1).

A limited number of studies have focused on an analysis of the co-occurrence between pollen and spore taxa in the air (Damialis et al. 2015; Kasprzyk 2008). However, an internal cross-examination for spores has often being overlooked, although all spore season characteristics are reported (e.g. Kasprzyk et al. 2004). This study has, for the first time, included this type of analysis (Fig. 3). A continuation of this study would be beneficial as differing results were found for Alternaria spores compared with previous findings, which suggested a potential increase in spore production (Sadyś et al. 2015d). In addition, some caution would be advisable as these results are based on 5-year data and an extension of the observation period would be recommended (Nikkels et al. 1996).

5 Conclusion

Allergenic fungal spores can be found throughout the year in the atmosphere of the studied area. The analysis depicted time periods when allergenic spore types occurred in the air with a maximum concentration. This information hopefully will be used by local allergists while setting up a treatment for patients sensitised to fungal spores and will help sufferers to avoid exposure to fungal spores during the most dangerous hours and days by planning their activities and holidays accordingly. The maximum monthly sum of mean daily spore concentration varied between taxa, and the earliest peaks were recorded for Pleospora in April and Ustilago fungi in June. However, the majority of investigated spore types occurred in the greatest concentrations between August and September. The most dominant genus was Cladosporium, which exceeded an allergenic threshold of 3000 s m−3 40 times during very rainy years and twice as much during dry years.