THE INFLUENCE OF METEOROLOGICAL FACTORS ON THE HAZEL ( CORYLUS L . ) POLLEN CONCENTRATION IN SOSNOWIEC IN THE YEARS 1997-2007

An analysis of hazel pollen seasons in Sosnowiec was presented on the basis of data from the years 1997–2007. The research was conducted by means of the volumetric method using a Burkard-type spore trap. The duration of pollen seasons was determined by means of the 98% method. The research demonstrated statistically signifi cant correlations between the average, maximum and minimum temperature, relative humidity as well as the number of days with sub-zero temperature and the beginning of the hazel pollen season. It was demonstrated that the duration of the pollen season depended on air relative humidity, insolation and precipitation during the season and the annual sum depended on the weather conditions of the year preceding pollen production and also the weather conditions two years earlier. Signifi cant correlations were also found between weather conditions and the daily concentration of pollen grains. The daily concentration decreased when relative humidity was high and increased with high insolation and air temperature.


INTRODUCTION
Hazel belongs to a group of early blossoming trees.The moment when hazel blossoms is considered to be the beginning of botanical early spring.In Poland one species grows in the wild, that is, the Common Hazel (Corylus avellana).There are some other species in horticulture.Among arborescent plants, hazel pollen grains, apart from birch and alder, are the most common cause of pollinosis (H o f m a n and M i c h a l i k , 1998).It is assumed that the fi rst symptoms of pollinosis appear at a concentration of 35 grains per cubic metre of air (R a p i e j k o et al. 2004).Hazel pollen contains strong allergens, but it does not cause severe allergic symptoms, because the pollen concentration in Poland's urbanized areas does not usually exceed medium values (usually 20-30 grains per 1 cubic metre of air).Hazel pollen allergy is often accompanied by hypersensitivity to alder and birch pollen allergens as well as allergic symptoms appear after consumption of hazel nuts.
As hazel blossoms at the end of winter and the beginning of spring, it is exposed to sudden weather changes, especially to large temperature fl uctuations.This has an impact on considerable differences in the dates of the beginning of pollen seasons and their duration, the periods of maximum pollen concentrations as well as the maximum values of hazel pollen concentrations and also pollen concentrations of other trees which blossom in early spring (D e t a n d t and N o l a r d , 2000; R a p i e j k o , 2003).According to many authors, air temperature during winter and spring is the most important weather element infl uencing hazel pollen seasons (K a s p r z y k , 1997; G a l a n et al. 2000; P u c et al. 2006).
The aim of this study was to demonstrate the impact of meteorological conditions on the dates of the beginning and end of pollen seasons, their duration, annual sums and the daily concentration of hazel pollen grains in Sosnowiec over 11 years.

MATERIALS AND METHODS
An analysis of the hazel pollen concentration in the air of Sosnowiec was performed on the basis of data from the years 1997-2007.The research was conducted by means of the volumetric method using a Burkard-type spore trap.The trap site was placed at a height of approximately 20 metres above the ground on the premises of the Faculty of Earth Sciences at the University of Silesia in Sosnowiec (50° 17' 50"N and 19° 08' 20"E, 262.5

metres above sea level).
A microscopic analysis, after slides had been stained with alkaline fuchsin, was performed on the surface of 4 horizontal strips (M a n d r i o l i et al. 1998).The result was expressed as the average daily number of pollen grains per 1 cubic metre of air.The duration of pollen seasons was determined by means of the 98% method, according to which the beginning of the season is on the day when 1% of the cumulative sum of a given taxon's pollen is recorded and the end when 99% of pollen is found (E m b e r l i n et al. 1994; S p i e k s m a and N i k k e l s , 1998).
In order to assess the impact of meteorological factors on the selected properties of pollen seasons -the beginning of the season, the annual sum of the pollen grain concentration, the duration of the season and the daily concentration -the average monthly and daily values of 7 weather elements were taken into account: average, minimum, maximum and near-ground temperature, precipitation, relative humidity and insolation.The correlations between individual meteorological conditions and various properties of the hazel pollen season were established by means of Pearson correlation coeffi cients and multiple linear regression (S t a n i s z , 2007).

RESULTS
During the eleven-year period examined, the hazel pollen season started earliest in 2007 (10.01) and latest in the years 2005 (14.03) and 2003 (8.03) (Tab.1).The beginning of the pollen season in the years examined was infl uenced by average, minimum and maximum temperature and relative humidity 40 days before the season, which is confi rmed by correlation coefficients between meteorological conditions and the date of the beginning of the season (for the average temperature r = -0.78,for the minimum temperature r=-0.81, for the maximum temperature r = -0.73,for relative humidity r=-0.62 respectively).High temperatures and high relative humidity caused the examined taxon's pollen to be present earlier in the air in the years 1998, 2000, 2002, 2004and 2007 (Figs 1-3) (Figs 1-3).The temperatures directly preceding the beginning of the season are also important.In all the years, the hazel pollen season started after a few or between ten and twenty days with the maximum temperature above zero.The year 1998 was an exception -this year, the maximum temperature fell below zero one day before the beginning of the pollen season (Fig. 1).The number of days with sub-zero average temperature also infl uenced the beginning of the hazel pollen season.The greater the number of such days from the beginning of January, the later the pollen season started.In 2005, the number of days with sub-zero temperature amounted to 48, in 2003 -46, whereas in 2007 the average temperature remained above zero until 22 January (Fig. 3).
Moreover, the impact of cumulative average temperature calculated from 1 January (over 0°C and over 5.5°C) on the date of the beginning of the pollen season was examined, but this method did not give positive results.
The shortest pollen season was in 2005 -24 days, the longest seasons in 2006 and 2007 -67 days.The correlation coeffi cients between weather elements and the duration of hazel pollen seasons calculated for individual pollen seasons indicate that there is a statistically signifi cant negative correlation between the season duration and the average (r=-0.62)and maximum (r=-0.65)daily temperature and insolation (r=-0.63)during the season.A statistically signifi cant positive correlation was found for relative humidity (r=0.73) and precipitation (0.69) during the season as well as for precipitation in March (0.65).This indicates that the hazel pollen season is extended when relative humidity and precipitation is high.On the other hand, high temperature and insolation cause pollen of this taxon to be present in the air over a shorter period of time.However, after conducting the r coeffi cient signifi cance test, that is, after calculating the coeffi cient of determination r 2 , it turned out that only a small percentage of the variability in the duration of pollen seasons could be explained by the correlation of this property with the weather elements examined (average temperature -38%, maximum temperature -42%, insolation -40%, relative humidity -53%, precipitation during the season -48%, precipitation in March -42%).
A model of the dependence of the pollen season duration on weather conditions developed by means of the multiple linear regression method is as follows: Y (duration) = 0.835 T aver -1.590 T max + 9.164 Humidity + 0.537 Precipitation + 2.067 Insolation -20.61 ± 0.052 Relative humidity, precipitation and insolation are the statistically signifi cant structural parameters of the model.Because these parameters are positive, they will have an impact on the pollen season extension.The above regression model is characterized by an estimation error of 0.052 and a corrected coeffi cient of determination r 2 of 0.91.Thus, the developed model accounts for approximately 91% of the variability in the duration of hazel pollen seasons.This indicates that there is a strong linear dependence between weather elements and the duration of the pollen seasons of the taxon in question.
The annual sums of hazel pollen ranged from 947 in 2007 to 329 grains in 1998.In 2001 and 2004, the annual sums were also high and they amounted to 875 and 833 pollen grains, respectively (Fig. 4).An analysis of the impact of the selected meteorological factors on the annual sum revealed the highest positive correlation for the average temperature in May in the preceding year and the highest negative correlation for precipitation in June in the preceding year.A statistically signifi cant correlation with the meteorological conditions two years before the season was also found, that is, with the average and maximum temperature in September.An analysis of the annual sums during individual years did not result in fi nding two-year periodicity, which means that the years with the high values of annual sums are separated by the years with low annual sums (Fig. 4).
The period of maximum hazel concentrations started in all the study years after the snow cover had totally disappeared.The days with maximum pollen concentrations occurred in the years examined between 2 February and 9 April (Figs 1-3).There were few days with the threshold hazel pollen concentration, that is, one which results in the onset of clinical symptoms in people with hypersensitivity: from 2 days in 1998 to 10 days in 2007 (Tab.1).
The correlation coeffi cients between weather elements and the daily concentration indicated that the daily concentration of hazel pollen grains increased together with air temperature and insolation, whereas it decreased when relative humidity was high (Tab.2).A statistically signifi cant negative correlation with precipitation was not found.

DISCUSSION
Exposure to hazel pollen allergens varied considerably in the years examined.The differences are visible in the dates of the beginning and duration of pollen seasons, maximum daily concentrations of pollen grains and annual sums.Meteorological conditions have a signifi cant impact on the time of pollen production by plants, especially those which blossom in early spring (K a s p r z y k , 1997; W e r y s z k o -C h m i el e w s k a a nd R a p i e j k o , 1997).The research conducted confi rms that the average, minimum, maximum and near-ground temperature, relative humidity, insolation and precipitation infl uence some properties of pollen seasons.Similar correlations were found in Poznań (S t a c h , 2000), Spain (J a t o et al. 2002) and Szczecin (P u c , 2007).
The average, minimum and maximum temperature as well as relative humidity 40 days before the season had an impact on the beginning of the pollen season in the years.Moreover, a signifi cant correlation between the beginning of the hazel season and the number of days with sub-zero temperature from the beginning of the year was found, which was also demonstrated by U r u s k a ( 2003 2000) emphasize that trees which blossom in early spring start to produce pollen when the cumulative temperature from 1 January reaches a certain threshold specifi c for every species.However, the research conducted in Sosnowiec did not confi rm the effectiveness of this method (the sums of the values of average daily temperatures over 0°C and over 5.50°C were taken into account).
The duration of the hazel pollen seasons in the years examined ranged from 24 to 67 days.The statistical analysis demonstrated that relative humidity, precipitation and insolation during the pollen season had the most signifi cant impact on this property of the pollen season of the studied taxon.An increase in the above parameters should be accompanied by an extension of the season.Additionally, grains which are transported from distant places can have an impact on the extension of the hazel pollen season (H j e l m r o o s , 1992; P u c , 2006).
A great deal of research proves that the annual sum of pollen grains of some tree species is dependent on the weather elements of the year preceding the season or even two years earlier (S a r v a s , 1972; R o d k i e w i c z et al. 1996; N o r r i s -H i l l , 1998).In Sosnowiec the average temperature in May and precipitation in June in the year preceding the season had an impact on the annual sum of hazel pollen.A statistically signifi cant correlation with the average and maximum temperature in September two years earlier was also found.An analysis of the annual sums in individual years resulted in fi nding that there was no twoyear periodicity, that is the years with high and low values did not occur alternately.Furthermore, it was found that there was a slight tendency for the annual sums of hazel pollen grains to increase, but they did not exceed 1000 grains in all the years examined.The relatively low values of annual sums, in spite of the large production of pollen by the infl orescence, might be caused by light frosts, which frequently occur during the fl owering of hazel (D y a k o w s k a , 1937; P i o t r o w s k a -W e r y s z k o , 2001).
The hazel pollen season can be classifi ed as a compact one (H y d e , 1956).The number of pollen grains in the air increased, reached one distinct maximum and then decreased.The years 2001The years , 2004The years and 2007 were an exception -more than one maximum was then found (Figs 1-3).

CONCLUSIONS
1.The average, minimum and maximum temperature as well as relative humidity 40 days before the season and the number of days with sub-zero temperature had an impact on the beginning of the hazel pollen season.2. The average temperature in May and precipitation in June in the preceding year as well as thermal conditions in September two years before the season infl uenced the annual sums of hazel pollen.3. The duration of the pollen season depended on air relative humidity, insolation and precipitation during the season.4. The daily concentration of hazel pollen grains increased when the air temperature and insolation were high and fell with an increase in air relative humidity.

Fig. 1 .
Fig. 1.Correlation among the number of hazel pollen and meteorological factors in the years 1997-2000.

Fig. 2 .
Fig. 2. Correlation among the number of hazel pollen and meteorological factors in the years 2001-2004.
) in her research conducted in Gdańsk.Many authors (F r e n g u e l l i et al. 1992; R o d k i e w i c z et al. 1996; M i n e r o et al. 1999; G a r c i a -M o z o et al.

Table 1
Characteristics of hazel pollen seasons in Sosnowiec.