Trends in grass pollen season in southern Spain

Transcription

1 Aerobiologia (21) 26: DOI 1.17/s ORIGINAL PAPER Trends in grass pollen season in southern Spain Herminia García-Mozo C. Galán P. Alcázar C. Díaz de la Guardia D. Nieto-Lugilde M. Recio P. Hidalgo F. Gónzalez-Minero L. Ruiz E. Domínguez-Vilches Received: 2 September 29 / Accepted: 26 December 29 / Published online: 1 January 21 Ó Springer Science+Business Media B.V. 21 Abstract The main characteristics of Poaceae pollen season at 8 sites in Andalusia were studied. Special attention was paid in the trends of grass pollen-season start and peak dates. Moreover, we analyse the intensity of the grass pollen season over H. García-Mozo (&) C. Galán P. Alcázar E. Domínguez-Vilches Departamento de Botánica, Ecología y Fisiología Vegetal, Edificio Celestino Mutis (C4), Campus de Rabanales, Universidad de Córdoba, 1471 Córdoba, Spain bv2gamoh@uco.es C. D. de la Guardia D. Nieto-Lugilde Departamento de Biología Vegetal, Facultad de Ciencias, Universidad de Granada, 181 Granada, Spain M. Recio Departamento de Biología Vegetal, Facultad de Ciencias, Universidad de Málaga, 298 Málaga, Spain P. Hidalgo Departamento Biología Ambiental y Salud Pública Facultad de Ciencias Experimentales, Campus de El Carmen Avda. de las Fuerzas Armadas s/n, 2171 Huelva, Spain F. Gónzalez-Minero Departamento Biología Vegetal, Facultad de Farmacia, Universidad de Sevilla, 4171 Sevilla, Spain L. Ruiz Departamento Biología Animal, Biología Vegetal y Ecología, Facultad de Ciencias Experimentales de Jaén, Campus Universitario Paraje Las Lagunillas, 2371 Jaén, Spain the study period as well as potential temporal and spatial patterns in these data. Statistical analysis was performed to determine the possible influence of weather-related parameters on variations in the grass pollen season. Main results show an advance in the start and peak of grass pollen season and an increase in the annual Pollen Index and in the severity of the season (days [ 25 pollen grains/m 3 ). The future consequences of these changes in grass phenology could be related with changes in land use and also in pollinosis symptoms due to the higher concentrations recorded but also to the variations on pollen season dates. Keywords Poaceae Grasses Pollen Aerobiology Phenology Climate change 1 Introduction Pollen from the different species of the grass family (Poaceae) is a major aeroallergen throughout Europe. Poaceae is a large and widespread family, although the species spectrum varies across the continent. In Spain, where 22% of the population suffer pollen allergies, an average of 8% of pollen-allergy sufferers is affected by grass pollen (Subiza 23). The percentage of grass sensitised patients from pollen-allergy sufferers ranges from 97% in northern Spain to 48% in coastal south-eastern Spain; these figures match the reported prevalence of Poaceae

2 158 Aerobiologia (21) 26: pollen allergy throughout Europe (D Amato et al. 27). There are approximately 42 species of grasses in Spain, and most are present in the south (Andalusia) (García Rollán 1985). The study of grass phenology using airborne pollen data is complicated by the stenopalynous nature of grass pollen, which gives rise to pollen counts representing an amalgam of numerous species. The species contributing most to the pollen season constitute an irregular curve with various peaks, reflecting differences in species phenology (Prieto et al. 23). The timing and intensity of the grass pollen season vary between regions, due to differences in vegetation, land use, latitude, altitude and climate (Fernández-González et al. 1999; Sánchez-Mesa et al. 23; García-Mozo et al. 29). Moreover, at local level grass phenology displays year-on-year variations depending on weather conditions (Emberlin et al. 1994; Fernández-González et al. 1999). The relationship between environmental conditions and airborne pollen counts has been widely studied in woody species, revealing a marked influence of temperature on reproductive phenology, especially in trees flowering in early spring (Peñuelas et al. 22; Menzel et al. 26). The extent of this influence has recently been examined with a view to investigating the effect of climate change on these species. In Andalusia, the effect of climate change has been analysed in depth by Sousa et al. (27), who report changes in mean temperatures consistent with those recorded throughout the Iberian Peninsula, characterised by two periods of warming in the twentieth century: one during the first half of the century and another, more marked, starting in the 197s. The thermal evolution of seasonal maximum temperatures points to an outstanding warming in spring of about 2 C, although the magnitude of changes is smaller in autumn and winter. A slight downward trend in spring and annual rainfall has been detected; the decrease is significant for spring, close to significance limits for yearly rainfall, but not statistically significant either for winter or for autumn rainfall. The effect of these changes in southern Spain has been an earlier start to the season in tree species such as Quercus (García-Mozo et al. 26) and, to a lesser extent, Olea (Galán et al. 25). However, little research has addressed grass pollen dynamics in this area over recent years (Alcázar et al. 29), and more particularly, the response of grass species to weatherrelated parameters. A recent study in Spain has shown that the variables governing the grass pollen start and peak dates are rainfall, air temperature and photoperiod (García-Mozo et al. 29). It would appear that high temperatures and low rainfall during the pollen season contribute to high airborne Poaceae pollen counts (Emberlin et al. 1999; Sánchez-Mesa et al. 23), nevertheless the factors leading to considerable year-on-year variations in phenological dates and the intensity of the grass pollen season in different climatic areas remain unclear. The main aims of this study were as follows: to seek potential temporal and spatial patterns at 8 sites in Andalusia region; to analyse the grass pollen season intensity and severity over the study period; to chart variations in grass pollen-season start and peak dates; and to determine the possible influence of weather-related parameters on variations in the grass pollen season. 2 Materials and methods Airborne pollen concentrations were monitored at 8 sites in Andalusia (southern Spain) belonging to the Andalusia Aerobiology Network (RAA) uco.es/raa/. The studied sites were as follows: Huelva, Cádiz, Málaga, Almería, Granada, Jaén, Córdoba and Sevilla. Volumetric spore traps (Hirst 1952) were used, following the protocol developed by Galán et al. (27). Although aerobiological monitoring started in the 198s in Córdoba city, at most of the study sites it started in Daily average concentrations of grass pollen were expressed as pollen grains per cubic metre of air sampled. The pollen-season start date was defined as the first day on which 5 pollen grains/m 3 were recorded with the 3 following days recording 5 or more pollen grains/m 3 (García-Mozo et al. 29). The end of the season was determined by the first day on which pollen grains/m 3 were recorded, not as a result of rainfall conditions. The peak date was defined as the day on which maximum pollen counts were recorded, indicating that most plants in a given population were in full bloom. The severity of the season was assessed as the number of days on which daily average grass pollen counts exceeded the threshold of 25 grains/m 3,

3 Aerobiologia (21) 26: indicated by the Spanish Aerobiology Network (REA) as the threshold over which pollen-allergy sufferers start to show serious symptoms (Galán et al. 27). The 8 pollen traps are located in the capital cities of the eight provinces of Andalusia, all within the Mediterranean climate region (Fig. 1). Although all the sites themselves are urban, they differ in the nature of the surrounding built-up areas and in land use in the natural environs. The main geographical and climatic characteristics are shown in Table 1. Cádiz and Huelva, on the Atlantic coastal plain of south-western Spain, and Málaga and Almería, on the Mediterranean sea belong to the Thermomediterranean bioclima with mild temperatures due to a strong maritime influence. Córdoba and Sevilla cities are located in the Guadalquivir valley, with a high thermomediterranean climate, characterised by very hot summers. Finally, Granada and Jaen are located inland in eastern Andalusia, at the highest altitude of the 8 cities; climate is Mesomediterranean, and their inland location leads to a larger range of temperatures, with colder winters and warmer summers than at coastal sites. Meteorological data were supplied by the Spanish Meteorology Agency (AEMET) and the Andalusian Agroclimatic Information Network (RIA). For each pollen season characteristics, statistical results are shown in Table 2. We calculated the slope of the linear regression plot for recorded observations, the t value (Student s t distribution) as an Fig. 1 Location of sampling sites estimate of the deviation from the population mean (t values between?1 and -1 with p [.5 indicate no apparent change), and the p value as a measure of the statistical probability of fit to the regression line (p \.5 indicates a significant fit to linearity). Trend regression analysis give us information about the advance or delay (? or -), the degree (slope), and the significance of the analysis (fit to lineal line) if p \.5 or.1. This degree of significance speak us about the probability of the following event to fit the line, so it is possible to detect high slopes not significant due to high variability of dates, usually occurring in biological events such as phenology. t value (between -1 and 1) indicates us the degree of change to detect apparent changes in phenology (Bradley et al. 1999). Phenological events and meteorological data recorded over previous months were subjected to a Spearman correlation test (Table 3). Analysed meteorological periods were chosen in base to the results of previous analyses performed by authors (García- Mozo et al. 29; García-Mozo et al. in press) and the indicated in other related studies (Emberlin et al. 1994; Peñuelas et al. 22; Recio et al. in press). The STATISTICA 6. software package was used. Results were considered significant at a 95% confidence interval. 3 Results 3.1 Seasonal counts and severity Variations in Pollen Index (PI), Peak Value (PV) and severity (days [ 25 grains/m 3 ) are shown in Fig. 2, while associated linear trend equation statistics are provided in Table 2. The correlation results for meteorological parameters and pollen count values are shown in Table 3. Generally speaking, the Pollen Index for the western provinces (Córdoba, Sevilla, Huelva and Cádiz) displayed the highest cumulative counts. Daily airborne grass pollen counts reached a maximum of 1,8 grains/m 3 (Cádiz 23). Other sites recording high peak counts were Huelva, Córdoba and Málaga. The severity of the grass season ranged from days to over 6 days depending partly on the site, but mostly on annual climate characteristics, especially previous cumulated rainfall (Table 3).

4 16 Aerobiologia (21) 26: Table 1 Characteristics of study sites Site Sampling period Coordinates Thermoclimate Alt. T Rf Cordoba ( ) 37 5 N, 4 45 W Thermomediterranean Sevilla ( ) N, 5 59 W Thermomediterranean Jaén ( ) N, 3 47 W Mesomediterranean Granada ( ) N, 3 35 W Mesomediterranean Huelva ( ) N, 6 57 W Thermomediterranean Cádiz (21 28) N, 6 18 W Thermomediterranean Almería ( ) 36 5 N, 2 28 W Thermomediterranean Málaga ( ) N, 4 19 W Thermomediterranean Alt., altitude (m a.s.l.); T, annual mean temperature ( C); Rf, annual rainfall (mm) Córdoba The cumulative annual PI varied considerably, ranging from a minimum of 1,15 in 1995 to a maximum of 1,83 in 27. Daily PV also displayed year-onyear differences, varying between 71 grains/m 3 in 1995 and 863 grains/m 3 in 27. Although both values showed a cyclic fluctuation over the study period, a significant rising trend was detected indicated by the positive and significant slope of the trend regression equation showed in Table 2. A similar rising trend was found for severity. Over recent years, seasons with more than 5 days exceeding 25 grains/ m 3 were recorded. Table 3a shows the positive influence of the previous months rainfall on all values relating to the intensity of the grass pollen season Sevilla In this city, neither annual PI nor PV displayed any apparent rising trend over the study period. Indeed, this site recorded the lowest increase in grass pollen counts. By contrast, although severity varied over the study period, a significant rising trend was detected in later years with more than 6 days exceeding 25 grains/m 3. Statistical analysis revealed a significant correlation of PI, PV and severity with rainfall recorded over previous months Jaén Records showed a rising trend for PI and PV, although PI varied considerably over the study period, ranging from 798 in 1999 to 4336 in 27, while PV varied between 42 grains/m 3 in 1995 and 476 grains/m 3 in 23. Seasonal severity in Jaen was the lowest of all the Andalusia inland capitals; although a trend towards increasing severity was noted over later years of the study. Rainfall recorded before full flowering dates influenced both PI and PV Granada Here, annual cumulative counts increased over the study period, although fluctuations associated with dry periods were observed. The lowest PI was detected in 1995 (535) and the highest in 27 (3288). PV values ranged from 34 grains/m 3 in 1995 to 34 grains/m 3 in 1996, with a less marked rising tendency in later years. Severity here was less intense than in the other inland cities, ranging from 1 day in 1995 to 46 in 24, although a rising trend was apparent. Table 3a shows, yet again, that cumulative previous rainfall was main factor influencing these values Huelva Huelva city, on the Atlantic south coast, displayed remarkable fluctuations in both PI and PV over the study period. PI ranged from 953 in 25 to 7773 in 26, while PV varied between 34 pollen grains/m 3 in 25 and 1274 in 21 A significant increase in PI values was revealed by the high slopes and t values of trend equations. Severity also increased over the study period. In the last 3 years, seasons with around 5 days exceeding 25 grains/m 3 were detected. Previous months rainfall had a clear influence on PV and severity, but not on PI.

5 Aerobiologia (21) 26: Table 2 Trend regression analyses of the changes in pollen season phenology n Linear trend regression Slope t p Pollen index Peak value Cordoba N [ Start date Peak date Season length Pollen index Peak value Sevilla N [ Start date Peak date Season length Pollen index Peak value Jaen N [ Start date Peak date Season length Pollen index Peak value Granada N [ Start date Peak date Season length Pollen index Peak value Huelva N [ Start date Peak date Season length Pollen index Peak value Cadiz N [ , Start date Peak date Season length Pollen index Peak value Malaga N [ Start date Peak date Season length Table 2 continued Cádiz Cádiz, also on the Atlantic coast, displayed the highest PI and PV of the coastal cities studied; despite some fluctuations, the trend remained generally stable over time. PI ranged from 18 in 1999 to 9366 in 23, and PV from 82 grains/m 3 in 1999 to 1842 grains/m 3 in 23. Severity presented interannual variations showing a slight negative trend. Cádiz was the only site where a slight negative trend was detected over the study period, with an average of 27 days exceeding 25 grains/m 3. Weather-related parameters over the previous months had no significant influence on cumulative pollen counts, PV or severity Málaga n Linear trend regression Slope t p Pollen index Peak value Almeria N [ Start date Peak date Season length n represents the number of observations. Slopes of the regression are reported in unit/year, p \.5 indicates a significant fit to linearity, t values are given as Student s t distribution (t values between?1 and -1 with p [.5 indicate no apparent change) Phases: Start date: start of pollen season; Peak date: date when maximum pollen concentration is recorded; Season length: number of days of pollen season; Pollen index: sum of annual pollen concentrations; Peak value: concentration recorded in the Peak date; N [ 25: number of days with concentrations exceeding 25 pollen grains/m 3 Málaga city, located on the Mediterranean coast, shows one of the lowest pollen records of Andalusia and a general increase in PI, PV and severity is detected in last years. PI ranged from 887 in 25 to 457 in 27, PV from 54 grains/m 3 in 1992 to 815 grains/m 3 in 23. The severity of the season shows an average value of 23, from 3 days[25 grains/m 3 in 1992 to 38 days in 27. In Table 3b, correlation

6 162 Aerobiologia (21) 26: Table 3 Spearman correlation analyses for different phenophases n Start date Peak date Season length Pollen index Peak value N [ 25 R t(n-2) p R t(n-2) p R t(n-2) p R t(n-2) p R t(n-2) p R t(n-2) p (a) Inland Cordoba T J-A T J-Jn P Rf J-A P Rf J-Jn T J-A Sevilla T J-Jn P Rf J-A P Rf J-Jn T J-A Jaen T J-Jn P Rf J-A P Rf J-Jn T J-A Granada T J-Jn P Rf J-A P Rf J-Jn (b) Coastal Huelva T J-A T J-Jn P Rf J-A P Rf J-Jn Cadiz T J-A T J-Jn P Rf J-A P Rf J-Jn Malaga T J-A T J-Jn P Rf J-A P Rf J-Jn Almeria T J-A

7 Aerobiologia (21) 26: Table 3 continued n Start date Peak date Season length Pollen index Peak value N [ 25 R t(n-2) p R t(n-2) p R t(n-2) p R t(n-2) p R t(n-2) p R t(n-2) p T J-Jn P Rf J-A P Rf J-Jn Higher probabilities than 95% are in bold to highlight the most important factors affecting phenology Phases: Start date: start of pollen season; Peak date: date when maximum pollen concentration is recorded; Season length: number of days of pollen season; Pollen index: sum of annual pollen concentrations; Peak value: concentration recorded in the Peak date; N [ 25: number of days with concentrations exceeding 25 pollen grains/m 3 Meteorological factors: T J-A: Mean Temperature from January to April; T J-Jn: Mean Temperature from January to June; P P Rf J-A: Cumulated Rainfall from January to April; Rf J-Jn: Cumulated Rainfall from January to June analyses indicate the significant influence of previous month s rainfall on all these values Almería The province of Almería is located in the eastern area of the Mediterranean south coast, and has lowest rainfall of all Andalusia, as well as the lowest grass pollen counts. Mean PI was 1178, and displayed a rising trend over the later years of the study period. PV and severity were also the lowest in Andalusia. PV ranged from 18 grains/m 3 in 25 to 29 in 22 and showed no remarkable increase in later years. The average number of days with counts exceeding 25 grains/m 3 was 8. Moreover, in dry years (i.e. 1999, 25) no daily pollen counts reached this value, although in later years there was a trend towards increased severity. Spearman s test revealed no significant influence of meteorological variables on these data. 3.2 Pollen season phenology Three aspects of grass reproductive phenology were analysed: start date, peak date and pollen-season length. Mean results refer to non-leap years. Data are shown in Fig. 3, while the associated linear trend equation statistics are provided in Table 2. The correlation results for weather-related parameters and phenological data are shown in Table Córdoba The range in pollen-season start dates during the period 1982 to 28 was 62 days; the mean pollenseason start date was 17 April. The years with earliest starts (i.e. 1986, 1988, 27 and 28) had relatively warm and wet early springs, while years with late starts (i.e. 1983, 1993 and 24) usually had cooler and drier early springs. Overall, there was a weak but significant trend towards earlier start dates, although in the last 5 years a trend towards latter flowering is observed. Peak dates showed a weaker advancing trend, with less year-on-year variability; the mean peak date was 19 May. Statistical analysis detected a relationship between start dates and previous months rainfall. Mean pollen-season length was three months (96 days), with year-on-year fluctuations (ranging

8 164 Aerobiologia (21) 26: (a) Number of days gr/m3 Pollen Index (c) (e) Sevilla Jaen Córdoba Granada Sevilla Jaen Córdoba Granada Sevilla Córdoba Jaen Granada Number of days gr/m3 Pollen Index Huelva Málaga Cádiz Almería Huelva Málaga Cádiz Almería Huelva Cádiz Málaga Almería Fig. 2 Pollen count results through the studied period. Pollen Index in inland (a) and coastal sites (b); Peak Value in inland (c) and coastal sites (d); Severity of the pollen season (number of days [25 grains/m3) in inland (e) and coastal sites (f) (b) (d) (f) from 43 days in 1995 to 179 days in 1986 and 153 in 28), and a slight rising trend was observed over the study period. Wetter and warmer springs lead to earlier start dates but also favoured later end dates, giving rise to longer pollen seasons, although the increase in length did not prove statistically significant Sevilla The mean start date occurring was 2 April, with an 89-day range, from 18 February in 23 to 18 May in 1999; a trend towards earlier starting was observed. A similar advancing trend was noted for peak dates, although it lacked statistical significance. There was also less variation (range 31 days). Earlier flowering dates were associated with higher temperatures and greater rainfall in earlier months, although there was no significant correlation. Sevilla had the longest grass pollen season, albeit with marked fluctuations (from 124 days in 1993 to 256 days in 24). Although Sevilla season lengths displayed the highest regression slope of the inland sites studied, the detected increase during the study period did not prove statistically significant.

9 Aerobiologia (21) 26: (a) 2 (b) Sevilla Córdoba 6 4 Huelva Cádiz (c) 2 Jaen Granada (d) 2 Málaga A lmería Sevilla Córdoba 6 4 Huelva Cádiz 2 Jaen Granada 2 Málaga Almería (e) 25 Sevilla Córdoba (f) 25 Huelva Cádiz 2 Jaen Granada 2 Málaga Almería Fig. 3 Phenological results through the studied period. Start of the season in inland (a) and coastal sites (b); Peak date in inland (c) and coastal sites (d); Length of the pollen season in inland (e) and coastal sites (f) Jaén Start dates between 1993 and 28 had a range of 7 days; the mean start date was 19 April. There was a strong trend towards earlier start dates. Peak dates fluctuated with a smaller range (37 days), and showed a slight advancing trend. Spearman correlation confirmed the influence of spring rainfall on earlier start dates in Jaén. The mean season length was 3 months (91 days), varying from 54 days in 1999 to 161 days in 27, and increasing slightly in the later years of the study. The significant influence of rainfall in longer seasons, shown in Table 3a, is probably due to the influence of spring rainfall on earlier starts Granada The earliest start date was 9 March (23) and the latest 14 May (1993), with a mean date of 14 April, showing a trend towards earlier start dates. Granada recorded a strongest and most significant trend towards earlier peak dates of all the inland sites studied; the range was of 8 days (23 March in 1999 to 11 June in 1998). Season length fluctuated with a range of 95 days, with a mean of 112 days; there was a general trend towards increased season length over the study period. No significant correlation was detected between climatic variables and phenology data, although earlier flowering was associated with warmer and wetter early springs.

10 166 Aerobiologia (21) 26: Huelva There was a marked trend towards earlier start dates. Start dates ranged from 14 May (2) to 23 March (27), the mean date being 7 April (Fig. 3b). Spearman s test results indicated a significant influence of spring temperature on earlier start dates. Peak dates (mean 22 May), also showed a significant trend towards earlier dates, a trend which also correlated with temperature. Season duration, with a mean of 3.5 months (15 days), increased considerably over the later years of the study, though varying between a minimum of 6 days in 2 and a maximum of 17 days in Almería The most marked trend towards earlier start dates was recorded here; start dates ranged from 28 May in 1998 to 7 March in 28. There was also a less marked advance in the peak date, as shown in Fig. 3d, which displayed less variation (11 July in 21; 29 April in 25). The correlation test (Table 3b) confirmed the influence of higher temperature in the previous months on the earlier peak dates recorded over the study period. Season length ranged from a mean of 3 days in the years to a mean of 11 days in the later years of the study, showed a significant increase during the studied period (Table 2) Cádiz The mean start date was 7 April, with variations within a range of 2 months, from 14 May in 2 to 14 March in 24; a trend towards earlier starts was recorded over the later years of the study. Peak dates remained fairly stable (mean 22 May). Spring temperatures influenced the advance in pollen-season start and peak dates. The smallest increase in season length was recorded here, although there were marked year-on-year variations, from 6 days in 2 to 17 days in Málaga Fluctuations in start dates were less marked than at other study sites, especially those in inland areas. The mean start date was 3 April, and variations ranged from 2 May in 1993 to 8 March in Figure 3b shows a very weak but stable advance in the start date over the study period. Early spring temperature proved to be the main factor affecting start date variations, while spring rainfall was the factor most favouring later end dates. Peak dates were also quite stable; mean date was 2 May, and peaks ranged from 8 May in 1994 to 3 June in 21. The pollen season lasted an average of 14 days, showing an increasing trend in the later years of the study. In the early 199s, the season lasted around 7 days, whereas in the later years the average duration was almost 11 days. 4 Discussion Although year-on-year variations undoubtedly hinder inter-site comparisons, the results obtained here revealed similar trends for the grass pollen season. Trends in pollen counts (except for the severity of the Cadiz grass pollen season) displayed positive slopes, indicating an increase in airborne grass pollen in Andalusia, especially in inland areas. Moreover, Student s t values suggested that in most cases this increase was statistically significant. The rate of increasing severity of the pollen season was particularly striking. The considerable increase in airborne grass pollen counts contrasts with the decline in grassland areas around the study sites over recent years, as a result of changing land use in Andalusia (Coves 27). Since 1956, the built-up area in Andalusia has increased by 286%. The study cities have grown exponentially over recent years, and green areas both within the cities and in the surrounding areas have been widely built up. Increasing tourism in coastal sites and migration from rural to urban areas has prompted constant urban growth and a consequent decline in grasslands. Moreover, the grassland category is not limited to flowering grass, but extends to cereal crops. In aerobiological studies, however, the area under cereal crops is not considered an important grass pollen production area because pollen grains from cereals tends to be larger than those from ordinary grasses, presenting a more limited dispersal pattern (Emberlin et al. 1999). All this serves to highlight the

11 Aerobiologia (21) 26: influence of environmental factors on the increase in pollen counts. A number of studies report that reproduction patterns in this family vary according to latitude, ecology and climatology (Raju et al. 1985; Connor 1986; Emberlin et al. 1994). Emberlin et al. (1999) studied regional variations in grass pollen seasons in UK, reporting that warmer weather in the early months of the year would influence net productivity and hence pollen production. Our results show a stronger relationship between pollen production and rainfall over the previous months, confirming that the development of grasses is positively influenced by water availability, especially in Mediterranean areas (Clary et al. 24). This influence was stronger in inland sites where evapotranspiration is higher and reduces water availability in spring months. On the other hand, wind influence can be detected in hourly/ daily aerobiological data compared with hourly/daily wind data. Given that the main objective of the work is not this sort of analysis but offer a general idea of temporal trend in grass records in South Spain, wind influenced was not analysed. In our opinion, the punctual influence of wind in some dates would not be a significant factor to control temporal changes (phenological and severity changes) during a so long period. Camacho and Subiza (24) have underlined the influence of previous winter rainfall and temperature on grass pollen records in Madrid. Other studies, in various regions of the world, have shown increasing trends in pollen counts over time, and have related these to trends to climate changes over the same periods. Spieksma et al. (1995) examined airborne birch (Betula spp.) pollen data from five European stations in northern Europe, observing a rise in annual counts. The same trend for birch, but also hazel and grasses, was found in Switzerland by Frei and Leuchner (2). Teranishi et al. (26) revealed earlier pollen season for the Japanese cedar (Cryptomeria japonica) due to warmer temperature from 1983 to 23 at an urban location in Japan. Another factor that could be influencing the increase in pollen production per plant as well as subsequent airborne pollen release is the increase in atmospheric CO 2 in recent years. There is experimental evidence to suggest that increased CO 2 prompts not only the increase and acceleration of vegetative growth in plants but also an increase in pollen production mostly due to the increase in the number of flowers (Ladeau and Clark 26; Rogers et al. 26; Ziska and Caulfield 2). With regard to phenological trends, a number of authors have reported that warmer weather in the preseason months of the year would encourage grass growth, with earlier maturation of vegetative and reproductive structures (Emberlin et al. 1994; Peñuelas et al. 22; Recio et al. in press). The present results for grass reproductive phenology show a significant advance in the pollen-season start date and longer pollen seasons; similar trends have been detected for other herbaceous species in Europe (Stach et al. 27). Nevertheless, these changes for grass species were always less marked than those reported for woody species (Clot 23; Peñuelas et al. 22; Menzel et al. 26). In Andalusia, a more pronounced advance in the start of the Quercus and Olea pollen season has been reported in last decades (Galán et al. 25; García-Mozo et al. 26). In northern latitudes, this finding has been associated with longer periods of low counts at the beginning of the season, before the onset of the main grass flowering period (Emberlin et al. 1999). The present results for southern Spain appear to indicate that Peak Dates have advanced slightly over recent decades, although most of the values for the Student t distribution indicated no significant changes. Season length in all cases showed a slight but not significant rising trend over the years, influenced by the earlier start dates; the pollen-season end dates remained quite stable through the period (data not shown). The clinical consequences of the findings presented here, based on data for southern Spain are therefore clear: grass-allergy sufferers are exposed to longer periods of airborne pollen and, especially, to higher counts over a longer period; this may have negative effect on the quality of daily life. Moreover, changes in the phenology of species producing allergenic pollen in southern Spain are leading to an overlap of pollen seasons (e.g. of olive and grass pollen seasons), thus increasing the risk for patients sensitised to both pollen types, who account for a high percentage of the allergic population in Andalusia. Climate change is modifying the pollen season phenology and the pollen severity of plants of allergenic interest; the negative impact on health could be intensified due to the more frequent occurrence of heavy rainfall events and to an increase

12 168 Aerobiologia (21) 26: in the number of high urban pollution episodes, and the subsequent impact on the prevalence of respiratory disease (D Amato and Cecchi 28; Beggs 24). Acknowledgments The authors are grateful to the European Social Fund for co-financing, to the Spanish Ministry of Science for the Ramón y Cajal contract awarded to Dr. García-Mozo and to the Andalusian Regional Government Department of Science and Technology for funding the project entitled Análisis del polen atmosférico como bioindicador de la calidad del aire y de los efectos del cambio climático en la fenología y biodiversidad de los ecosistemas andaluces P6- RNM-234. References Alcázar, P., Stach, A., Nowak, M., & Galán, C. (29). Comparison of airborne herb pollen types in Córdoba (Southwestern Spain) and Poznan (Western Poland). Aerobiologia, 25, Beggs, P. J. (24). Impacts of climate change on aeroallergens: Past and future. Clinical and Experimental Allergy, 34(1), Bradley, N. L., Leopold, A. C., Ross, J., & Huffaker, W. (1999). Phenological changes reflect climate change in Wisconsin. Proceedings of the National Academy of Science, 17(96), Camacho, J. L., & Subiza, J. (24). Relación entre factores meteorológicos y recuentos de pólenes de gramíneas causantes de alergias In García Codron, J. C., Diego Liaño, C., Fdez de Arróyabe Hernáez, P., Garmendia Pedraja, C., y Rasilla Álvarez, D. (Eds.) (24). El Clima entre el Mar y la Montaña. Asociación Española de Climatología y Universidad de Cantabria, Serie A, n8 4, Santander. Clary, J., Savé, R., Biel, C., & de Herralde, F. (24). Water relations in competitive interactions of Mediterranean grasses and shrubs. Annals of Applied Biology, 144(2), Clot, B. (23). Trends in airborne pollen: An overview of 21 years of data in Neuchâtel (Switzerland). Aerobiologia, 19(3 4), Connor, H. E. (1986). Grass systematics and evolution. Systematics and evolution. Washington, DC., London: Smithsonian Institution Press. Coves, F. (27). Evolución de los usos y coberturas vegetales del suelo en Andalucía de 1956 a la actualidad. International Conference on Forestal Fire. Seville, 27. D Amato, G., & Cecchi, L. (28). Effects of climate change on environmental factors in respiratory allergic diseases. Clinical and Experimental Allergy, 8, D Amato, G., Cecchi, L., Bonini, S., Nunes, C., Annesi-Maesano, I., Behrendt, H., et al. (27). Allergenic pollen and pollen allergy in Europe. Allergy, 62(9), Emberlin, J., Jones, S., Bailey, J., Caulton, E., Corden, J., Dubbels, S., et al. (1994). Variation in the start of the grass pollen season at selected sites in the United Kingdom Grana, 33, Emberlin, J., Mullins, J., Corden, J., Jones, S., Savage, J., Millington, W., et al. (1999). Regional variations in grass pollen seasons in the UK, long term trends and forecasts models. Clinical and Experimental Allergy, 29, Fernández-González, D., Valencia-Barrera, R. M., Vega, A., Díaz de la Guardia, C., Trigo, M. M., Cariñanos, P., et al. (1999). Analysis of grass pollen concentrations in the atmosphere of several Spanish sites. Polen, 1, 132. Frei, Th., & Leuschner, R. M. (2). A change from grass pollen induced allergy to tree pollen induced allergy: 3 years of pollen observation in Switzerland. Aerobiologia, 16(3 4), Galán, C., Cariñanos, P., Alcázar, P., & Domínguez-Vilches, E. (27). Spanish aerobiology network (REA): Management and quality manual. Córdoba, Spain: Servicio publicaciones de la Universidad de Córdoba. Galán, C., Garcia-Mozo, H., Vazquez, L., Ruiz, L., Díaz de la Guardia, C., & Trigo-Perez, M. (25). Heat requirement for the onset of the Olea europaea L. Pollen season in several places of Andalusia region and the effect of the expected future climate change. International Journal of Biometeorology, 49(3), García Rollán, M. (1985). Claves de la flora de España : (Península y Baleares). Ed. Mundi-Prensa. Madrid. García-Mozo, H., Galán, C., Belmonte, J., Bermejo, D., Candau, P., Díaz de la Guardia, C., Elvira, B., Gutiérrez, M., Jato, V., Silva, I., Trigo, M. M., Valencia, R., & Chuine, I. (29). Predicting the start and peak dates of the Poaceae pollen season in Spain using process-based models. Agriculture and Forest Meteorology, 149, García-Mozo, H., Galán, C., Jato, V., Belmonte, J., Díaz de la Guardia, C., Fernández, D., et al. (26). Quercus pollen season dynamics in the Iberian Peninsula: Response to meteorological parameters and possible consequences of climate change. Annals of Agricultural and Environmental Medicine, 13, García-Mozo, H., Mestre, A., & Galán, C. Phenological trends in southern Spain: A response to climate change. Agriculture and Forest Meteorology (in press). Hirst, J. (1952). An automatic volumetric spore-trap. Annals of Applied Biology, 36, Ladeau, S. L., & Clark, J. S. (26). Pollen production by Pinus taeda in elevated atmospheric CO 2. Functional Ecology, 2, Menzel, A., Sparks, T. H., Estrella, N., Koch, E., Aasa, A., Ahas, R., Alm-Kübler, K., Bissolli, P., Braslavská, O., Briede, A., Chmielewski, F. M., Crepinsek, Z., Curnel, Y., Dahl, Å., Defila, C., Donnelly, A., Filella, Y., Jatczak, K., Måge, F., Mestre, A., Nordli, Ø., Peñuelas, J., Pirinen, P., Remišová, V., Scheifinger, H. M., Striz, A., Susnik, A., Van Vliet, J. H., Wielgolaski F. E., & Zust A. S. Z. (26). European phenological response to climate change matches the warming pattern. Global Change Biology, 12(1), Peñuelas, J., Filella, I., & Comas, P. (22). Changed plant and animal life cycles from 1952 to 2 in the Mediterranean region. Global Change Biology, 8, Prieto-Baena, J. C., Hidalgo, P., Domínguez, E., & Galan, C. (23). Pollen production in the Poaceae family. Grana, 42,

13 Aerobiologia (21) 26: Raju, M. V. S., Jones, E. J., & Ledingham, G. F. (1985). Floret anthesis and pollination in wild oats (Avena fatua). Canadian Journal of Botany, 63, Recio, M., Rodriguez-Rajo, F. J., Jato, V., Trigo, M. M., & Cabezudo, B. The effect of recent climatic trends on Urticaceae pollination in two bioclimatically different areas in the Iberian Peninsula: Málaga and Vigo. doi 1.17/s Climatic Change (in press). Rogers, C. A., Wayne, P. M., Macklin, E. A., Muilenberg, M. L., Wagner, C. J., Epstein, P. R., et al. (26). Interaction of the Onset of Spring and Elevated Atmospheric CO2 on Ragweed (Ambrosia artemisiifolia L.) pollen Production. Environmental Health Perspectives, 114(6), Sánchez-Mesa, J. A., Smith, M., Emberlin, J., Allitt, U., Caulton, E., & Galan, C. (23). Characteristics of grass pollen seasons in areas of southern Spain and the United Kingdom. Aerobiologia, 19(3 4), Sousa, A., García-Barrón, L., & Jurado, V. (27). Climate change in Andalusia: Trends and environmental consequences. Edita Consejería de Medio Ambiente de la Junta de Andalucía.Sevilla. ISBN: p. Spieksma, F. T., Emberlin, J. C., Hjelmroos, M., Jäger, S., & Leuschner, R. M. (1995). Atmospheric birch (Betula) pollen in Europe: Trends and fluctuations in annual quantities and the starting dates of the seasons. Grana, 34, Stach, A., Garcia Mozo, H., Prieto Baena, J. C., Czarnecka- Operacz, M., Jenerowicz, D., Silny, W., et al. (27). 1 years term trends, seasonal variation and sensitization to aeroallergen of Artemisia in Poznań, (Poland). Investigational Allergology and Clinical Immunology, 17(1), Subiza, J. (23). Gramíneas: Aerobiología y polinosis en España. Allergology and Clinical Immunology, 18(3), Teranishi, H., Katoh, T., Kenda, K., & Hayashi, S. (26). Global warming and the earlier start of the Japanese-cedar (Cryptomeria japonica) pollen season in Toyama, Japan. Aerobiologia, 22(2), 9 94.