THE INFLUENCE OF CHEMICAL CHARACTERISTICS OF PRECIPITATION ON TREE HEALTH IN BANJICA FOREST ( BELGRADE , SERBIA )

The most represented tree species in the Banjica Forest are Acer negundo, Quercus robur, Acer pseudoplatanus, Populus nigra, Fraxinus pennsylvanica, Fraxinus ornus and Robinia pseudoacacia. According to the ICP Forests combined assessment (degree of defoliation and decolorization), endangered species are Populus nigra (64.3% of heavily damaged trees), Quercus robur (45.5%), Fraxinus pennsylvanica (37.0%) and Acer negundo (26.6%), while the situation is much better for Acer pseudoplatanus and Fraxinus ornus. For Robinia pseudoacacia, 83% of trees are without decolorization, however, defoliation is established. In the period from April to October 2009, the average pH of rainwater was 5.46, and 5.18 in the period from November 2009 to March 2010. The concentration of SO4 in the period from April to October 2009 amounted to an average of 24.21 mg/l, and 28.87 mg/l in the period from November 2009 to March 2010. The concentration of SO4 and pH values is a possible explanation for the condition of the trees.


INTRODUCTION
"Banjica Forest" is located in Belgrade and is a natural memorial park (protected from 1993).The total area is 41.59 hectares (Management Unit "Banjica"), while the forest covers about 39 ha.The specificity of Banjica Forest is that it is completely surrounded by urban areas, and is also a habitat for about 70 species of birds.In 2005, signs of defoliation and decolorization were observed in a number of the trees in Banjica Forest.Conducted entomological and phytopathological analyses could not fully explain this phenomenon (Karadžić et al., 2007).Since the studied area is affected by a number of anthropogenic sources of pollution, it was presumed that this could be the framework for further research into the causes of the above-mentioned phenomenon.Therefore, research into the chemical characteristics of precipitation was undertaken.Primarily it was thought that the larger share of tree damage could be explained by so-called "acid rain" (pH below 5).Bini and Bresolin (1998) emphasize the importance of anthropogenic components in the acidification of rainwater in northern Italy, where the average pH value is 5.2.Lara et al. (2001) found that anthropogenic activities (dust from the soil, burning of sugar cane and industrial emissions) affected the chemical composition of rainwater in the Piracicaba Basin in southeastern Brazil.The average pH value was 4.4 to 4.5, and the acidity of rain was significant at all tested points.However, Gregory et al. (1996) considered acidic pollutants to be only one of the causes of damage to forests, and their importance was greater in areas with strong local emissions.The author pointed out the importance of bad breeding measures and climate impacts.Cape (1993) states that the indirect influence of wet deposition on vegetation over processes in the soil is usual, but that leaves exposed to fog and clouds, which often contain much higher concentrations of pollutant ions than rain, can also be directly damaged.It has been established that there is a risk to vegetation in mountain areas, occurring near the base of the clouds, where the concentrations of ions are the highest.
The large number of trees with signs of damage, location of sites and the presence of pollution sources were taken into account in the case of Banjica Forest.The research objectives were: (i) to determine the degree of health deficiency of the most represented tree species on the investigated site, and (ii) to carry out chemical analysis of rainwater over a period of 12 months, and based on the results obtained, to assess the possibility of the impact of certain chemical characteristics on the forest trees.

MATERIALS AND METHODS
Methods based on visual assessment of the state of treetops were used to assess the health of the trees in Banjica Forest.In the studied area, 19 test areas were singled out, covering 15 m 2 each.All the trees were marked with numbers.Assessment of defoliation was carried out during June and July 2007, while the assessment of decolorization was performed in August of the same year.After that, in 2008 and 2009 a check of the state of trees was made, which showed that there were no changes.In this part of the research, methods in the ICP Forests program were used (Tables 1-3) (Innes, 1990).Assessments of defoliation and decolorization are shown as a percentage of the ideal treetop appearance for the observed species (Tables 1 and 2).
The examination also included the biological status of the trees in order to separate the impact of  Collecting samples of rainwater and snow water was done during the period 1 April 2009 -31 March 2010.Analyses were undertaken at the Faculty of Chemistry in Belgrade.The pH and conductivity were examined, and methods for determining the concentrations of some ions in the samples are shown in the Table 4.
The results are shown by dates of sampling, as well as by periods: April-October (roughly coinciding with the vegetation period) and November-March (outside the vegetation period).

RESULTS AND DISCUSSION
Table 5 presents data on tree species, the presence of which is confirmed in the test areas.
The most represented species are Acer negundo, Quercus robur, Acer pseudoplatanus, Populus nigra, Fraxinus pennsylvanica, Fraxinus ornus and Robinia pseudoacacia.Other species comprise less than 5% of the share of the total number of labeled trees.Therefore, in this paper attention was focused primarily on the seven most represented species.
Data on the distribution of the seven most represented tree species according to the degree of defoliation (Table 6) indicate that the best situation is that for Acer platanoides, of which 85% did not exhibit defoliation.A good state was also recorded for Acer pseudoplatanus, 70% of which did not exhibit defoliation.However, there were no Populus nigra without defoliation: 57% have severe and 40% moderate defoliation.For Robinia pseudoacacia, most trees were with moderate (56%) and severe (22%) defoliation.Quercus robur is considered to be an endangered species, since 64% of the trees have moderate and 11% severe, defoliation.Fraxinus pennsylvanica had 63% with moderate defoliation, so that in this species, as in the past one, the transition of these trees to the group with severe defoliation can be expected.A similar, but somewhat more favourable situation was that of Fraxinus ornus, of which 46% of the trees have moderate defoliation.Note: The data on the mean daily air temperature and precipitation are given according to the Republic Hydrometeorological Institute of Serbia.The precipitation value deviations occurred due to differences in the time of sampling and different locations.
Combined assessment includes defoliation, which is permanent, and decolorization, which is more a result of the seasonal influences of biotic (phytopathogenic fungi and injurious insects) and abiotic factors (lack of moisture, extreme temperatures).According to the combined assessment (Table 8), endangered species are Populus nigra (64.3% of heavily damaged trees), Quercus robur (45.5%),Fraxinus pennsylvanica (37.0%) and Acer negundo (26.6%).The situation for the species Acer pseudoplatanus (66.7% of trees without damage) and Fraxinus ornus (46.2%) was much better.A significant share of trees without damage could be seen in the less represented species of Tilia spp.(66.7%) and Juglans regia (50%).
Regarding the test areas, it can be concluded that both defoliation and decolorization are fairly evenly represented throughout the studied area.Due to the cumulative effect of moderate defoliation and moderate decolorization, severe treetop damage was recorded in seven of the test areas.
Research results related to precipitation are shown in Tables 9 and 10.
In 6 cases out of 34, individual samples (Table 9) had a pH below 5.The lowest recorded value was 4.65 (25 June 2009), while the highest value of 6.1 was recorded on 23 May 2009.Extreme values appeared at an interval of 33 days.
In the studied area, the dominant anion is SO 4 2- (Tables 9 and 10).Türküm et al. (2008) found that significant amounts of SO 4 2-arrive to Turkey from northern Europe, Ukraine, Russia and some parts of the Balkans.Başak and Alagha (2004) state that the average pH of rainwater in the area of Istanbul is 4.81 and the main cause is sulfur emissions.Millet et al. (2001) found for Thann (Alsace, France) a pH in the range of 3.60 to 6.58, while in Tours (Indre et Loire, France) values exhibited a narrow range (5.49-7.01).In both cases, the dominant anion was SO 4 2-.Momin et al. (2005) found surprisingly high concentrations of SO 4 2-in a rural area of India during the monsoon season.Akoto et al. (2011) found that the pH of rainwater in a mining area in Ghana was between 4.0 and 5.6, where the main anions were SO 4 2-and Cl -.Leal et al. (2004) studied the chemical composition of rainwater in Sao Paulo (Brazil), where the largest contribution to free acidity was from SO 4 2-(28.8%).Based on research in northeastern Uruguay, Zunckel et al. (2003) stated that the presence of NO 3 -and SO 4 2-in rainwater was characteristic for agricultural areas.The link between damage to treetops and SO 2 concentration was determined for spruce in northern Bohemia (Ardo et al., 2000), and for Pinus brutia and Pinus nigra in the area around Izmir (Turkey) (Kantarci, 2003).
High concentrations of Na + and K + were detected in the studied area, but the question arises as to their origin.
It can be seen in Table 10 that the reaction of atmospheric water in both tested periods was below  (2002), who found that in the vicinity of Ankara (Turkey) the average pH value of rainwater was 6.3 -7.0.Only about 4% of the rainwater samples had a pH below 5.0, while in only 15% the pH was lower than 5.6.
The slightly lower pH in the examined area during the winter period (Table 10) could be explained by anthropogenic influence, i.e., an increased consumption of fossil fuels.Sulfurous, sulfuric and hydrochloric acid, as well as the hydrolysis of ammonium sulfate, are important for the reduction of the pH of rainwater.The importance of sulfuric acid was confirmed by this research, given the high values obtained for the SO 4 2-.
Although the results obtained indicate the anthropogenic influence, it should be noted that the lowest pH (4.65) was found in early summer (25 June 2009), when the consumption of fossil fuels decreases and the number of people is lower compared to the rest of the year due to the beginning of the holiday season.Research results in other states also indicate the complexity of problems.Sanusi et al. (1996) investigated the chemical composition of rainwater in eastern France and found lower pH values (4.4) in rural compared to urban areas (Strasbourg 5.0 and Colmar 5.7); they explained this by the presence of CaCO 3 in the loess, which is the main constituent of the land in this area.Arsene et al. (2007) found that the average pH value of rainwater in northeastern Romania was 5.92 and that the main neutralizers were CaCO 3 and NH 3 .Lara et al. (2010) found that the average pH value in Monterrey, an industrial centre in northeastern Mexico, was 6.58, which they explained by alkalization affected by high concentrations of Ca 2+ and Mg 2+ .
These results indicate the need to reduce emissions of pollutants.The example of Portugal should be kept in mind, where the reduction of SO 2 emission came from the reduced quantity of sulfur in fuels (Santos et al., 2011).One should also examine the possibility that rainwater is a source of nutrients for plants, which is indicated by the results obtained in the region of Rio Grande do Sul, Brazil (Calil et al., 2010).
The results presented in this paper highlight the complexity of the problem and pose the question of the origin of pollutants.Bearing in mind the results of Radovanović et al. (2003), Milovanović and Radovanović (2009) and Stanojević (2010), the impact of regional atmospheric circulation on this phenomenon should also be investigated.However, the impact of the chemical characteristics of precipitation in the area of Banjica Forest will only be fully understood if long-term research is continued.

Table 1 .
Assessment of defoliation

Table 3 .
Combined assessment

Table 4 .
Methods for determining ionic concentrations + CalculationAll anions are calculated and expressed in milliequivalents, as well as cations Ca 2+ and Mg 2+ .The sum of all anions should be equal to the sum of all cations.

Table 5 .
Number and presence of tree species in the test areas

Table 6 .
Distribution of trees by degrees of defoliation (%)

Table 7 .
Distribution of trees by degrees of decolorization (%)

Table 8 .
Distribution of trees according to the degrees of combined assessment

Table 9 .
Mean daily air temperature, precipitation, pH, conductivity and ion content in mg/l by dates(April 2009 -March 2010)

Table 10 .
Average values of ions (mg/l) in rainwater for the periodsApril -October 2009 and November 2009 -March 2010