1 Introduction

Interest in the contribution of airborne pollen to allergies has increased around the world over recent times. Allergic symptoms like asthma or rhinitis are related to the composition, timing and abundance of allergenic airborne material together with other environmental contaminants. Airborne pollen content has been monitored in a large variety of countries, and records which stretch back to the final decades of the last half century exist for different sites in Europe, Japan, and North America (D’Amato et al. 2007).

In each geographic area, there is a succession of different flowering species throughout the year, so it is important to document the timing, floral intensity and types of airborne pollen in different locations, especially in highly populated cities. To date, only a few papers have described airborne pollen in the Balearic Islands (Belmonte et al. 1995; Boi and Llorens 2008). In the Mediterranean region, pollen calendars have been developed for several cities, such as Cagliari in Sardinia (Italy) (Ballero and Maxia 2003), Zagreb in Croatia (Peternel et al. 2005), Didim and Kastamonu in Turkey (Aycan et al. 2008; Çeter et al. 2012), Chirivel, (Cariñanos et al. 2004), Malaga (Recio et al. 1998) and Nerja on the Iberian Peninsula (Docampo et al. 2007), or Oporto in Portugal (Abreu et al. 2003).

The main goal of this paper was to study the airborne pollen content of Palma de Mallorca, one of the most popular tourist resorts in Europe, which welcomes about seven million tourists per year. The study of airborne pollen in this city contributes to modelling the behaviour of airborne pollen throughout the Mediterranean. The results of this investigation, which have been illustrated in a pollen calendar, along with future research into climatic change, are expected to assist allergists in diagnosing the cause and treatment for allergy cases.

2 Materials and methods

2.1 Study area

Palma de Mallorca (39° 57′ lat. N, 02° 65′ long. E) is located on Mallorca, the largest of the Balearic Islands to the east of the Iberian Peninsula (centre of the western Mediterranean basin) (Fig. 1).

Fig. 1
figure 1

Location of the aerobiological monitoring centre of Palma de Mallorca

Full size image

Bioclimatically, Palma de Mallorca lies within the thermo-Mediterranean belt (Rivas-Martínez 1987). The annual mean temperature is 16.8 °C, and daily temperature never goes below 10 °C. The mean annual rainfall is 449 mm. The highest rainfall level is recorded in autumn and winter, and there is a hydric deficit from May to October. The mean annual humidity is 74 %, and the entire city is influenced by winds from the southeast and north.

The ombroclimate is dry or sub-dry with a vegetation of Quercetea ilicis, Rosmarinietea officinalis and Rhamno Prunetea. The area surrounding the city is a natural refuge for the Mediterranean vegetation dominated by Pinus halepensis and Olea europaea. There are natural pastures with a preponderance of ruderal plants mixed with fruit trees. On the other hand, the parks and gardens around the aerobiological station include ornamental plants like Platanus hispanica, Casuarina cunninghamiana and different species from the Cupressaceae and Palmae families.

2.2 Pollen content in the air

Airborne pollen was monitored from 2004 to 2010 using a Hirst-type volumetric pollen trap (Hirst 1952). The sampler was situated on a flat roof in the centre of Palma de Mallorca, 30 m above ground level. This monitoring station is currently included in the Spanish Aerobiology Network (REA) and has provided information about the Balearic Islands since October 2003 (Boi and Llorens 2008). Pollen sampling and data management were performed in accordance with the Spanish Aerobiology Network Management and Quality Manual (Galán et al. 2007).

Airborne pollen was expressed as daily averages of pollen grains/m3 of air (pg/m3). The annual pollen index was defined as the annual sum of the daily airborne pollen content.

The main pollen season (MPS) for the most abundant pollen types was calculated as described by Andersen (1991) and modified by Jato et al. (2006). The pollen season was defined retrospectively as the period with 95 % of the annual pollen index. The pollen season started when 2.5 % of the annual pollen index was reached, and the season ended on the day in which 97.5 % of the cumulative total of annual pollen was registered.

The pollen calendar was developed following Spieksma’s model (Spieksma 1991), which transforms 10-day mean pollen concentrations into a series of classes in accordance with Stix and Ferretti (1974). Pollen types were ordered on the basis of abundance. Subsequently, the average pollen sums were placed in exponential classes represented in a pictogram by columns of increasing height, as suggested by the International Association for Aerobiology (Spieksma 1991). In the pollen calendar, minimum 10-day mean values equal to, or greater than, 1 pollen grain/m3 of air were included.

2.3 Statistical analysis

Statistical analysis was performed using the Statistica program version 8.0 (StatSoft, Inc. 1984–2008). Makensens application version 1.0 (freeware copyright Finnish Meteorological Institute 2002) was used to perform the nonparametric Mann–Kendall test (Hollander and Wolfe 1999) of a linear regression on the pollen time series as suggested and modified by Sen (1968) to verify the null hypothesis (HO) that no temporal trend occurred in the pollen concentration time series. This nonparametric method was employed for the pollen concentrations series because the trend was considered to be linear.

3 Results

During the study period (2004–2010), 13 pollen types with the highest presence in the air of Palma de Mallorca were monitored. The types collectively accounted for 91.85 % of the total pollen count, of which 67.63 % represented woody plants (Betula spp., Castanea sativa, Cupressaceae, Olea europaea, Pinus spp., Pistacia lentiscus, Platanus hispanica, Quercus ilex) and 24.22 % represented herbs and grasses Chenopodiaceae/Amaranthaceae, Plantago spp., Poaceae, Parietaria judaica (including a species of Urtica) and Urtica membranacea.

The mean annual pollen index was 20,027. The highest annual concentration was registered in 2005 (25,870) and the lowest in 2009 (14,726).

The seasonal pollen dynamics of the studied taxa varied. The pollen curve built on the basis of mean annual concentrations showed the general course of the pollen concentration during the year, with the highest levels in the months of March, April and May. The airborne pollen concentration had a pronounced seasonality. From February to June, a higher pollen concentration was detected, representing 88.74 % of the annual average pollen count. The monthly distribution was as follows: February, 9.06 %; March, 26.94 %; April, 19.10 %; May, 23.06 %; and June, 10.58 %. In contrast, lower concentrations occurred from July to January with a cumulative value of less than 3 % of the annual pollen count average (Fig. 2).

Fig. 2
figure 2

Monthly mean values

Full size image

As the year progressed, several pollen peaks were detected due to the successive flowering of the different taxa. Some pollen types, like Cupressaceae, Poaceae, Chenopodiaceae/Amaranthaceae, were detected over longer periods of time, because, in the main, they represented several species that bloomed at different times. The first peak was detected during winter, owing to the pollination of species such as Cupressaceae, Parietaria judaica and Urtica membranacea. A second peak arose in early spring from species like Pinus, Platanus and Quercus. During the late spring, pollen concentrations from Olea europaea, Plantago, Poaceae and Parietaria reached their highest values, whereas pollen from Castanea sativa and Chenopodiaceae/Amaranthaceae were mainly present in summer. During late summer, autumn and early winter, the daily pollen counts fell significantly.

The pollen types which were most abundant in the air of Palma de Mallorca during the period studied were, in order of decreasing pollen index: Cupressaceae, Olea, Platanus, Pinus, Parietaria, Urtica membranacea, Quercus, Poaceae, Chenopodiaceae/Amaranthaceae, Plantago, Castanea, Pistacia and Betula (Table 1).

Table 1 Seasonal pollen index for the main pollen season types from 2004 to 2010 and the mean values over the entire period
Full size table

The pollen which was most prevalent during the study period was that of the natural woody plants of the surrounding areas and from the ornamental species of Cupressaceae and Platanus. Cupressaceae, Olea, Platanus and Pinus represented 59.82 % of the total pollen counts recorded during the period. Quercus, Pistacia, Betula and Castanea make up 7.1 %, and herb pollen mostly associated with Parietaria and Urtica accounted for 24.21 % of the total count.

The highest daily pollen peak registered in Palma de Mallorca by the different pollen types arose from Platanus (1,962 pg/m3, 28 March 2006), followed by Olea (1,101 pg/m3, 20 May 2007), Cupressaceae (967 pg/m3, 11 March 2005), Pinus (710 pg/m3, 26 March 2010), Quercus (297 pg/m3, 10 May 2005), Castanea (108 pg/m3, 29 June 2004), Betula (95 pg/m3, 27 April 2007) and Pistacia (75 pg/m3, 6 April 2006). In the case of herbaceous plants, the highest daily concentrations came from Urtica (203 pg/m3, 16 March 2005), Parietaria (118 pg/m3, 21 April 2004), Poaceae (91 pg/m3, 4 May 2005), Plantago (54 pg/m3, 19 April 2005) and Chenopodiaceae/Amaranthaceae (31 pg/m3, 29 August 2010) (Table 2). Allergy risk thresholds depend on the pollen type. Moderately allergenic values are 25–50 pg/m3 in the case of herbaceous taxa and 50–00 pg/m3 in the case of arboreal taxa, whereas highly allergenic levels are >50 pg/m3 and >100 pg/m3 for herbaceous and arboreal taxa, respectively (Galán et al. 2007). Given these thresholds, there was a considerable variation in the number of days during which each taxon presented risks for allergy sensitivity (Table 2).

Table 2 Length of the pollen season for the most abundant pollen types
Full size table

Variability in the timing of phases in consecutive pollen seasons was analysed for each taxon. The pollen seasons started with Urtica, followed by Cupressaceae, Parietaria, Betula, Pinus, Pistacia, Platanus, Plantago, Poaceae, Chenopodiaceae/Amaranthaceae, Quercus, Olea and ending with Castanea. The beginning of the season for each type of pollen was more homogeneous than the subsequent phases. The irregular progression of the flowering phase and the pollen stage for each taxon caused greater variability in the 50, 75 and 97.5 %, compared to the first percentiles (5, 25) (Fig. 3).

Fig. 3
figure 3

Pollen season variability

Full size image

The mean daily average pollen concentration in Palma de Mallorca shows a downward trend over the study period observed. The lowest values occurred in 2008–2010. The trend was significant according to a Mann–Kendall test (n = 20; S = 3.11; p = 0.017). Given the p value, the null hypothesis of no trend was rejected. Sen’s method showed a decrease in the mean daily pollen of about 5 grains every year (slope value Q = −4.64, intercept value B = 70.98) (Fig. 4).

Fig. 4
figure 4

Mean daily airborne pollen concentration in Palma de Mallorca over the period 2004–2010

Full size image

The pollen calendar for Palma de Mallorca developed for the 13 most abundant pollen types during the period between 2004 and 2010 was produced with an exponential scale where each step was approximately twofold greater than the previous. The calendar showed that March, April and May were the months with the highest pollen content in the air. Longer pollen seasons usually include pollen types from different species, whereas shorter pollen seasons usually represent pollen from only one species or from different species that flower at the same time. In general, the pollen season for arboreal taxa, such as Betula, Castanea, Pistacia, Platanus and Quercus, was shorter than the season for Cupressaceae, Olea and Pinus. Cupressaceae, the most abundant pollen type, had two different peaks each year. One of these occurred from February to March, and another peak with lower intensity occurred from October to November. Olea was the second most abundant pollen type in the air of Palma de Mallorca with a high concentration during the month of May. Pinus pollen during the first peak of April–May corresponded to the local species Pinus halepensis and to other species (P. canariensis, P. sylvestris, P. pinea and P. pinaster) during the second peak. The herbs, Parietaria judaica, Chenopodiaceae/Amaranthaceae and Poaceae, had longer pollen seasons than Urtica membranacea or Plantago (Fig. 5).

Fig. 5
figure 5

Pollen calendar for Palma de Mallorca 2004–2010

Full size image

4 Discussion

One of the main goals of aerobiological observations is to define the pollen spectrum in a given region. In general, our study offers relevant information on the pollen content in the air of Palma de Mallorca. Medium–high levels of different types of airborne pollen with considerable allergenic capacity were found, consistent with other reports (Belmonte et al. 1995; Boi and Llorens 2008). This information is useful in the treatment and diagnosis of hay fever (Marshall 2004) or in studies of local and long-range airborne pollen transport (Makra et al. 2010).

The annual pollen index in Palma de Mallorca is lower than in other cities of the Mediterranean region, such as Cagliari (Italy) (Ballero and Maxia 2003), Nerja (Spain) (Docampo et al. 2007), Didim or Kastamonu (Turkey) (Aycan et al. 2008; Çeter et al. 2012). The pollen compositions broadly reflect the contribution of Mediterranean local vegetation, which consists of natural woodlands surrounding each town and anthropogenic ornamental flora in the cities.

The exceptions are Castanea and Betula pollen, which were detected at low levels, a finding which is noteworthy because these species are not present in the Balearic Islands (Castroviejo et al. 1993). This pollen was probably carried across to Mallorca by dominant winds from Europe, demonstrating long-range, distant transport and deposition in areas far from the source (Peeters and Zooler 1988; Skjøth 2007).

Analysing the pollen season dynamics, the highest concentrations during March or May coincided mostly with the flowering of woody plants. The allergenic pollen of Cupressaceae was the most predominant type, followed by Olea, Platanus, Pinus and Parietaria in terms of peak values and daily pollen count. These species (excluding Pinus) are the most important taxa in the whole of Mediterranean region, as regards pollen severity and allergenic potency. Urtica, Quercus, Poaceae, Chenopodiaceae/Amaranthaceae, Plantago, Castanea, Pistacia and Betula, in order of abundance, make up the rest of the pollen types which are considered important for human health (D’Amato et al. 2007).

Cupressaceae pollen was the most prevalent in air, consistent with reports from some eastern Mediterranean areas such as Sardinia (Ballero and Maxia 2003) and Greece (Gioulekas et al. 2004). Cupressaceae has a longer pollen season, and different species have distinct timing of pollination. For example, Juniperus phoenicea L. subsp. turbinata (Guss.) Nyman and Cupressus arizonica E.L. flower in autumn, while Cupressus sempervirens L. and Thuya orientalis L. release pollen in winter.

Olea is the second most important taxa analysed. It is considered one of the most important causes of respiratory allergic disease in the Mediterranean region (Florido et al. 1999) and the major cause of allergies in southern Spain (Domínguez-Vilches et al. 1993; Vázquez et al. 2003, Alba et al. 2006).

Platanus is the third pollen type in prevalence. It is an allochthonous plant which is common in the gardens, streets and parks of Palma de Mallorca. It is a probable source of allergy problems for the population, given that these taxa have allergenic importance in some Mediterranean areas (Alcázar et al. 2011).

Pinus, the fourth pollen type in abundance, is the most significant pollen-producing taxa in other Mediterranean areas, like Didim or Nerja (Aycan et al. 2008; Docampo et al. 2007), and it has exhibited a positive trend in pollen intensity over time.

Parietaria judaica, as the fifth most important pollen type, is seen as one of the most common causes of allergies in the Mediterranean region (D’Amato et al. 1991). Parietaria has a longer blooming period and a longer pollen season in warm and wet conditions (Galán et al. 2000).

According to Szczepanek (1994), herbaceous taxa, like Chenopodiaceae/Amaranthaceae or Plantago, usually show longer pollen seasons (60–120 days). Poaceae is one of the most important causes of pollinosis in many parts of the world, but low levels have been observed in Palma de Mallorca compared to other places such different sites in Spain (García-Mozo et al. 2009), or Thessaloniki in Greece (Gioulekas et al. 2004), and Zagreb in Croatia (Peternel et al. 2005).

Our pollen calendar offers a preliminary view of the spectrum of dominant pollen types during a 7-year period (2004–2010), which is considered to be the minimum time required to obtain a representative view of the behaviour of airborne pollen in any area (Rodríguez et al. 2004).

Knowledge of the pollen season variability, in terms of timing and intensity, has an important role in public health. All the species examined showed variability in the progression of the pollen season. Depending on the taxa, the consecutive phases (percentiles) of the pollen season are achieved at different times. The start of the pollen season (the 5 %) is the most important phase to predict because it is associated with the appearance of allergy symptoms, although aeroallergen sources may occur even before a pollen season starts (Madeja et al. 2005).

The downward trend in the airborne pollen in Palma de Mallorca during the period studied conflicted with reports from other sites where the mean pollen count has increased (Cristofori et al. 2010; Myszkowska et al. 2011). Although we detected a declining trend in average daily pollen levels, no significant modification in the composition of local vegetation has been observed. Future studies with longer data sets, integrating measurements of local climate parameters may provide useful insights which would show whether these results were long-term trends, or just short-term fluctuations.

Global warming over the last decades can significantly impact animal and plant populations (Root et al. 2003) and affect the reproductive processes that underpin pollen production (Mayer et al. 2011). Variation in meteorological conditions affect plant responses in terms of phenology and floral intensity (Ziska et al. 2008; Davies and Smith 2006), so future research of the presence of airborne pollen takes on great importance.

5 Conclusions

Documentation of the composition of dominant airborne pollen in the city of Palma de Mallorca contributes to understanding the flowering behaviour of local vegetation. It also assists allergologists and allergy patients. The mean annual pollen index over the 7-year study period was 20,027. The highest daily counts occurred between February and June, in particular in March and May, while the lowest counts were in July–January. The most important airborne pollen types in Palma de Mallorca are Cupressaceae, Olea, Platanus, Pinus and Parietaria. Pollen from Olea or Poaceae did not reach the significant levels seen in other parts of Europe. We also found that the average daily pollen count, inclusive of all taxa, showed a clear, declining trend over time.

The pollen calendar of Palma de Mallorca is the first one to have been produced for the Balearic Islands. It shows the species which are typically Mediterranean and included a range of taxa, many of which have a long pollen season.

Future research will focus on the production of pollen and pollen allergens, and their relationship to climate change may help us to understand the modification in the island’s vegetation.