Panagiotis Michopoulos · Athanassios Bourletsikas · Kostas Kaoukis ·Marios Kostakis · Nikolaos S. Thomaidis · Ioannis I. Passias · Helen Kaberi ·Stelios Iliakis

Abstract The distribution, quantification and fluxes of Pb were examined in an evergreen broadleaved forest in western Greece for three hydrological years. More specifically, concentrations and annual fluxes of Pb were determined in bulk and throughfall deposition as well as litterfall. The Pb concentrations were also measured in forest floor and mineral soil up to 80 cm and the isotopic ratios of 206Pb/207Pb were determined in soil layers and the parent rock material. High variability in the fluxes of the metal among the three hydrological years were found,evidence of the variability of Pb deposition in time. Litterfall fractions with a large surface area, like holm oak flowers, had high Pb concentrations. Applying a steady state model and considering the Pb amounts in throughfall and litterfall as inputs on the forest floor, the mean residence time of Pb in the forest floor was 94 years with a coefficient of variation equal to 41%. More observations are needed to lower the variability of the mean residence time.The isotopic ratio in the rock material was defined as the lithogenic ratio. The statistical tests showed that the petrol derived Pb migrated to the depth of 20 cm and its percentages in the soil pedon was in the range of 62% in the L horizon to 11% in the 10—20 cm layer. In higher depths (＞40 cm) preindustrial anthropogenic Pb affected the isotopic ratio. As the forest under consideration is remote from industrial activities, the results can serve as a baseline for future studies on Pb distribution and quantification.

Keywords Evergreen broadleaved forest · Lead ·Throughfall·Litterfall·Forest floor·Soil·Isotopic ratios

Introduction

Despite the phasing out of leaded petrol, anthropogenic lead (Pb) still dominates this metal input on earth surfaces although an appreciable decrease in atmospheric deposition has been observed(Michelutti et al.2009).Due to the long distance atmospheric transport, Pb from various human activities can be deposited in remote regions of the world.Many studies have demonstrated that pollutants(e.g.,toxic metals)from human activities could be transported even to the Polar Regions by atmospheric deposition (Hong et al.2012; Singh et al. 2013). Lead levels and isotope ratios in aerosols, ice cores, soils and lake sediments sampled in distant locations testify that Pb has contaminated all parts of the globe(Owack et al.2001;Renberg et al 2002).This is the reason that research has been carried out in remote forests. Foliage drops as litterfall which decomposes and heavy metals end up in soils or taken up by plants just like nutrients.One of the first studies on Pb deposition in forests was that of Heinrichs and Mayer(1977).The authors dealt with the distribution and cycling of Pb and other metals in beech and spruce forests in central Germany. In another early study, Van Hook et al. (1977) worked with the distribution of Cd, Pb and Zn in a mixed deciduous forest in eastern Tennessee (USA). Zo¨ttl (1985) reviewed the distribution and cycling of Pb and other metals in the Black Forest in Norway spruce stands in Germany. In a more recent article, Tang et al. (2015) examined the Pb and Cd concentrations in forest soils, in the east of the Tibetan Plateau in China. Other authors worked exclusively with Pb. Friedland and Johnson (1985) examined the distribution of Pb in forests of red spruce, balsam fir and white birch in in the southeastern USA and so did Turner et al.(1985)in forests of pitch pines and white and black oaks in New Jersey,USA.Watmough and Dillon(2007)dealt with the Pb biogeochemistry in forests of white pine, eastern hemlock, red oak and red maple in central Ontario, USA.

Lead (Pb) presents a unique characteristic, i.e., the contribution of anthropogenic Pb to the total Pb concentration can be quantified (Bindler 2011). Lead has four stable isotopes:208Pb,207Pb,206Pb and204Pb;three of them(208Pb,207Pb,206Pb) derive from radioactive decay of232Th(208Pb),235U(207Pb)and238U(206Pb),while204Pb is the only non-radiogenic isotope. Natural materials have characteristic Pb isotope ratios resulting from the different half-lives of Th and U and geological evolution. Anthropogenically emitted Pb retains the isotopic composition of Pb of the ore from which it was derived(Sturges and Barrie 1987). The isotopic composition of Pb in these emissions differs (in most cases) from the native Pb isotope ratios in soil (Erel 1998) and this fact provides the opportunity to determine the contribution of deposited Pb from anthropogenic sources to Pb concentrations in soils. The most widely used isotope ratio is the206Pb/207Pb one. This ratio has often been used to quantify Pb in forest ecosystems for soils and/or vegetation (Bing et al. 2014; Klaminder et al.2005; Michopoulos et al. 2018; Wilcke et al. 2001).

Evergreen plants are better heavy metal collectors than deciduous ones because they keep their leaves throughout a whole year. As mentioned above, most of the studies dealing with heavy metals focused on evergreen coniferous forests in central Europe.However,evergreen broadleaved forests occupy a substantial area in Mediterranean countries. Nearly all studies on the heavy metals content in broadleaved forests have dealt with holm oak (Avila and Rodrigo 2004; De Nicola et al. 2015, 2017). It often happens that evergreen broadleaved forests are mixed and so far there has not been any work on such forests with regard to heavy metals.

The goals of this study were to examine the distribution and quantification of Pb in a remote from big cities and industrial activities forest ecosystem of mixed evergreen broadleaved species in western Greece. More specifically,the Pb concentrations were determined in bulk and throughfall deposition as well as litterfall, for three consecutive hydrological years. In addition, total Pb concentrations were measured in organic and mineral soil layers to a depth of 80 cm. Moreover, the quantification of anthropogenic Pb in soils was carried out through the use of the isotope ratio206Pb/207Pb. As that forest is a remote one, it can serve as a baseline for other forests (rural or urban)with regard to Pb pollution.

Materials and methods

Study area

The experimental plot is situated in the area of the city of Amfilochia in western Greece (Fig. 1), at an altitude of 360 m.The aspect is northeast and the slope is moderately steep(19%).It has an area of 0.274 ha,and is enclosed in a catchment of a total area of 117 ha. The average annual rain height is 1213 mm derived from 20 years of observation (1996—2016). The last 4 years, in which deposition sampling took place, were rainy heavily. The age of the stand has a range of 60—100 years. The canopy closure is about 1.4 as there is overlapping of canopies of the various species. The vegetation cover is an evergreen forest consisting of the species holm oak (Quercus ilex L.), strawberry tree (Arbutus unedo L.), Kermes oak (Quercus coccifera L.), tree heaths (Erica arborea, L.), march heath(Erica verticillata L.) and green olive tree (Phillirea latifolia L.).

The soil was developed on sandy flysch, it is deep, well drained and classified as Haplic Luvisol (FAO 1988).

Deposition sampling

Deposition (bulk and throughfall) sampling was done on a weekly basis and a composite sample was formed every month according to weekly volumes. Bulk deposition was collected in a nearby forest clearing using two polyethylene funnels having a diameter of 18 cm, connected to a 5-L polyethylene bottle through a 1.05 m PVC pipe with an internal diameter of 2.5 cm. Throughfall was collected using 20 collectors,identical to those for bulk precipitation,placed randomly in the plot. The volume of water samples was measured and all samples were stored immediately in a fridge at 4 °C for a maximum of two weeks. If longer storage periods were needed, the samples were stored at— 2 °C. Sampling started at the start of October and ended approximately the same month of the next year. These periods happen to be the hydrological years in Greece.Three sampling hydrological years were sampled, involving 2012—2013, 2013—2014 and 2014—2015.

Fig. 1 Location of the experimental plot

Litterfall collection

For litterfall collection, 10 L traps in the form of cylindrical plastic buckets, each having a collecting area of 0.242 m2, were placed systematically along a line in both plots, approximately 0.50 m above ground at a distance of 10 m from each other. The bottom of each bucket was perforated so that rain could drain out. A plastic net was put at the bottom of the litter traps to avoid loss of small material. Collection of litterfall was done approximately every 3 or 4 months to have enough material for chemical analysis. Initial and final samplings were carried out the same dates as in deposition. A composite sample from the 10 traps was formed and transported to the laboratory for analysis. Leaves, twigs, flowers, fruits and fruit cups were separated and dried at 80 °C for 48 h. For each collection,the leaves formed one sample whereas for the other fractions we had a composite sample per year.Accordingly,we had three or four leaves samples per year and one sample from the other fractions.All litterfall fractions were ground in a special stainless mill and stored for analysis.

Soil samples collection

The soil samples collection was carried out by means of systematic sampling. Inside the plot, along three lines distancing 25 m from each other six soil pits were excavated. Each pit was 5 m away from each other.From each soil pit, the samples collected were the organic horizons L and the FH by means of a frame 15 cm × 15 cm and mineral soil layers from the depths 0—10,10—20,20—40 and 40—80 cm. There was mixture of six samples of equal volume per horizon and soil depth (taken at random) to have three pooled samples per horizon and depth. The samples of the L horizons were weighed, as they were those of the FH after sieving. The samples of the FH horizon and mineral layers were air dried and passed through a 2 mm sieve stored for analysis to determine texture, pH, cation exchange capacity (C.E.C.) and exchangeable cations. Like litterfall, soil samples were dried at 80 °C for 48 h. The samples of the L horizon at their initial conditions and those of the FH horizon and mineral layers after sieving were pulverized in a ball mill and stored for analysis.

Soil parent material collection

Fragmented flysch rock was collected from three places(randomly selected) inside the plot after digging to the point that parent material appeared.To remove Pb from the rock surfaces, the rock fragments were washed with soap and tap water and immersed in a DTPA solution for 24 h.The DTPA solution was identical to the one used to extract(and then determine) trace elements from soils (Lindsay and Norvell 1978). The rocks were rinsed thoroughly with deionized water and dried at 80 °C for 48 h. Before the isotope analysis rock subsamples were pulverized in a ball mill.

Chemical analysis

The soil pH was determined by a glass electrode in a mixture of soil and water at a ratio of 1:5 per volume. The soil texture was measured by the hydrometer method. The organic C was determined by a C analyzer.

Exchangeable cations in soils were extracted with a 0.1 M BaCl2solution and cation exchange capacity(CEC)was calculated by adding the calculated exchangeable cations (Ca2+, Mg2+, K+, Na+, Al3+, Mn2+, Fe3+).

Concentrations of metals in BaCl2extracts and deposition were determined with an ICP-MS instrument(Thermo ICAP Qc). Pulverized soil and ground litterfall samples were digested in a microwave oven with HF and aqua regia(mixture of HCl and HNO3, 3:1) at a temperature range of 160—170 °C for 20 min.Likewise,the metal concentrations were measured with the same ICP-MS instrument. The Pb isotope ratios were determined with an X series 2 ICP-MS(Thermo).

The quality assurance of the metal analysis was checked with the use of sediment samples(TAQC!-River Sediment,training on Analytical Control-Water Framework directive)and that of the isotopes with the use of certified reference material (BCR-2 (United States Geological Survey).

All results are expressed on a dry weight basis at 80 °C.

Calculations and statistics

The coefficients of variation (%) were calculated as the percentages of the standard deviations over the means for the Pb concentrations in leaves litterfall,the isotopic ratios in soils,soil pH,organic soil C,CEC and clay percentages in soils.

Annual volume weighted means of Pb in bulk precipitation and throughfall for each hydrological year were calculated taking into account monthly Pb concentrations and deposition volumes.

Annual fluxes of Pb in bulk precipitation and throughfall for each hydrological year were calculated by multiplying the measured volume in the samplers with the concentrations on a monthly basis and summing up the monthly amounts.

Annual fluxes of Pb in litterfall for each hydrological year were calculated by multiplying Pb concentrations with the amount of litterfall every time of collection, summing up the amounts and dividing by the total area of the litter traps.

Pb pools in the forest floor were found by multiplying the measured concentrations of Pb in the L and FH horizons by the dry weight of the horizons and taking into account the area from which the samples were collected.

The percentage of anthropogenic Pb was calculated according to the following equation (Emmanuel and Erel 2002):

where Panis the percentage of anthropogenic Pb, Rliis the ratio of206Pb/207Pb lithogenic, Rsois the ratio of206Pb/207Pb soil sample, and Ranis the ratio of206Pb/207Pb anthropogenic.

The lithogenic Pb can be considered the one found in the minerals of the flysch rock of the area. Accordingly, the isotopic ratio of206Pb/207Pb in the rocks is the ratio of the lithogenic Pb.

The average value of206Pb/207Pb(1.13)in the European atmosphere (Saether et al. 2011) was used as the isotopic ratio of the anthropogenic Pb.

The isotope ratios206Pb/207Pb of the soil horizons and layers were compared statistically through a t-test with the value of the isotope ratio206Pb/207Pb in the flysch rock samples.

The mean residence time (in years) of Pb in the forest floor was calculated as the ratio of the metal stocks in the forest floor(LF + H) over the Pb inputs to the forest floor expressed in fluxes of throughfall plus litterfall.

Results

The average bulk deposition and throughfall heights in the three hydrological years were 1401 and 896 mm, respectively. They were rather uniform in the three hydrological yeas as the coefficients of variation were 4% for the bulk deposition and 9% for the throughfall one. The second hydrological year had higher Pb concentrations in both bulk and throughfall deposition (Table 1) and that caused elevated values for fluxes for this year as the rain heights were more or less the same in the 3 years.

The Pb concentrations in litterfall (Table 1) had high variability for the first and second hydrological years in the leaves fraction. The flowers fraction had high Pbconcentrations, which, however, were not reflected in the fluxes values as the mass of flowers was not as high as the leaves mass.

Table 1 Volume weighted means of concentrations and fluxes of Pb in deposition and litterfall

The sum of the throughfall and litterfall fluxes are the Pb inputs on the forest floor (Table 2). The Pb load in the forest floor(sum of the Pb loads in the L and FH horizons)was found 674 g ha-1.Each hydrological year has its own total Pb input (Table 2) so three different mean residence times of Pb in the forest floor were calculated.The average value was 94 years with a 41% coefficient of variation.

The highest variability of Pb concentrations in soils was found in the L horizon and then fall abruptly in all other layers (Table 3). The FH horizon had the highest Pb concentration, which went decreasing further down the soil profile. In all mineral soil layers the Pb concentrations(Table 3) followed the organic carbon distribution. As the clay percentages were approximately constant through the mineral soil profile,it can be inferred that organic matter was the most important factor for Pb distribution in the soil.

The206Pb/207Pbratio in the parent rock material(flysch)was found 1.189(Table 4)with a coefficient of variation of only 0.09%.For this reason,this ratio was chosen to be the single value with which the rest of the ratios of the soil layers had to be compared statistically (t test). The comparison showed that all layers differed significantly(p ＜ 0.05) apart from the 20—40 cm layer. In the mineral soil layers the206Pb/207Pb values increased with depth withthe exception of the 40—80 cm layer (Table 4). The values of the206Pb/207Pb ratios in soils in the present work ranged from 1.153 in the L horizon to 1.193 in the 20—40 cm layer.In the 40—80 cm layer, the ratio was 1.177. The percentages of petrol derived Pb ranged from 62%in the L horizon and went decreasing reaching 11%in the 10—20 cm layers.

Table 2 Inputs of Pb to the forest floor

Discussion

Hydrological cycle

Lead was enriched in throughfall in all the volume weighted means of concentrations and in the flux of the first hydrological year (Table 1). Lower Pb fluxes in throughfall than bulk deposition were also measured in some coniferous and beech forests in central and northern Europe (Bringmark et al. 2013). One might expect that in all studies higher Pb fluxes in throughfall than bulk deposition should be observed as dry deposition is the prevalent deposition type for Pb (Azimi et al. 2003). Indeed, Pb dry deposition on leaf surfaces can be dissolved by dissolved organic carbon and thus enrich throughfall with Pb (Hou et al. 2005). It is probable that in winter months the high amount of precipitation after removing dry deposition can produce fluxes in bulk deposition higher than those in throughfall due to the high volume amount of precipitation.

A pot experiment with spruce plants using spiked Pb showed that only 2%of this element in the plants originated from the soil,with the remainder originating from deposition(Hovmand et al. 2008). For these reasons, the sum of throughfall and litterfall fluxes of Pb represent the total deposition of this metal(Bringmark et al.2013).In terms of Pb concentrations, the ranges in bulk deposition and throughfall were 0.044—0.552 and 0.171—0.731 μg L-1,respectively.These concentrations are far lower than thosefound in forested areas in the 1970s and 1980s.Heinrichs and Mayer (19 77) found a Pb concentrations of 38 and 48 μg L-1in the bulk and throughfall deposition, respectively, in a beech f orest in central Germany and a flux of 365 g ha-1year-1in throughfall . Friedland and Johnson(1985) found 17 and 51 μg L-1in bulk and throughfall deposition,respectively,in conif erous fo rests in northeastern USA and a flux of 400 g ha-1year-1in throughfall. The maximum v alue o f Pb flux in throughfall was only 6.22 g ha-1year-1in our plot. As years passed, the Pb concentrations in both bulk and throughfall deposition in forests was reduced as Pb was excluded from petrol. In a remote holm oak forest i n Spain Rodrigo et al.(1999)found 0.59 and 0.99 μg L-1Pb concentrations in bulk and throughfall dep osition, respectively.The Pb fluxes were 5.9 and 6.9 g ha-2year-1for bulk and throughfall deposition,almost similar to those found in our work in the second hydrological year(Table 1).In three rural forests of Japanese cedar (Cryptomeria japonica) in Japan Itoh et al. (2006)found a little higher Pb concentrations than those in our work in both bulk precipitation(0.66—1.35 μg L-1)and throughfall(0.50—1.01 μg L-1).In 14 sites in central and northern Europe Bringmark et al.(2013)found higher f luxes(in most sites) in bulk deposition ( 6.3—24.3 g ha-1year-1) and throughfall(6.5—18.2 g ha-1year-1)than those in the present work. Probably these sites have higher values of precipitation amounts and therefore higher fluxes of Pb.

Table 3 Lead concentrations and selected properties of the various soil layers in the plot

Table 4 Isotope ratios of 206Pb/207Pb and percentages (%) of petrol derived anthropogenic Pb in the various soil layers (organic and mineral)

Litterfall

Most studies dealing with metals refer to litterfall fluxes and not concentrations. The reason is that littefall fluxes together with the throughfall ones constitute the metal inputs to forest floors.This is especially important for such metals as Pb that deposit mainly as dry particles and not as leachates from leaves. However, the Pb concentration has its own importance as its variation has to do with the corresponding variation in atmospheric deposition. In Table 1, one can observe that the coefficients of variation with regard to Pb concentrations in leaves in the last two hydrological years are very high. In terms of the Pb concentrations magnitude in the leaves fraction, the values were low to moderate according to literature.Kapusta et al.(2003) referring to leaves litterfall in mixed forest ecosystems in Poland along a transect from the city of Cracow found much higher concentrations of Pb close to the city. In high distances from Cracow (35 km) the metal concentrations were a little higher than those in our work.Watmough and Dillon (2007) found similar Pb concentrations to our work in the leaves of litterfall in mixed forests in northern Ontario. The highest Pb concentrations in our work were found in the fraction of flowers. These flowers came from the holm oak plants. They have a high surface and therefore act as scavengers for dry deposition.The twigs fraction also had elevated concentrations of Pb.In most cases, twigs have rough area and trap some dry deposition. The least concentration was in fruits and fruit cups,both also belonging to holm oak plants.Fruits have a smooth surface, and therefore, they are not very adhesive material for dry deposition. We can hypothesize that fruits accumulate elements from the soil and their Pb concentration does not reflect the atmospheric pollution as in the case of trunk wood (Rossini Oliva and Mingorance 2006).The sum of the Pb fluxes in all litterfall fractions (together with the throughfall inputs) are showed in Table 2. In the above-mentioned work of Ka pusta e t al. (2003) the leaves litterfall was abo ut 4 g ha-1year-1in remote forests and about 20 g ha-1year-1in forests close to Cracow. Ukonmaanaho e t al. (2001) found rather high fluxes of Pb(22 g ha-1year-1) in Sitka spruce forests in Finland (no information about the fraction), whereas, Watmough and Dillon (2007) in mixed forests i n Ontario, Canada, found similar Pb fluxes (6.9 g ha-1year-1) to our second hydrological year (also mixed litterfall fractions). In general, the litterfall fluxes of any element depend on both concentration and litter quantity.

Lead residence time in the forest floor

Understanding the long term fate of Pb in forests floors is essential to predict possible migrations of the metal further down the mineral soil horizons.There are various methods to calculate the mean residence time of Pb in the forest floor. They were briefly described by Richardson et al.(2014). In this work, the ratio of the metal stock in the forest floor (LF + H) over the Pb inputs as fluxes(throughfall + litterfall)to the forest floor was used to find the metal residence time (in years) in the forest floor. This equation was first used to find the mean residence time of major nutrients in forest floors (Gosz et al. 1976). The presupposition for applying this equation is that an ecosystem is at a steady state. In forests, this stage can be more or less accomplished when the canopy closure is complete. In the experimental plot in our work, this ecological stage has been achieved. Miller and Friedland(1994)argued that in order to apply the steady state model,the Pb amount in the forest floor should not change with time.If the Pb inputs in forests are not constant in time,the Pb amount in the forest floor can fluctuate. Due to these fluctuations in the Pb inputs in our work (Table 2) in the three hydrological years the coefficient of variation for the Pb residence time in the forest floor reached a value of 41%. However, the steady state model can be improved if more years of observation are used to reduce variability.Furthermore, the phasing out of the petrol derived Pb will stabilize (more or less) deposition inputs.

Applying this equation for the three hydrological years in our experimental plot, taking into account the total Pb input to the soil for each hydrological year (Table 2) and the total amount of Pb in the forest floor (674 g ha-1) the average residence time of Pb in the forest floor was found 94 years.In general,the range of the Pb residence times in forest floors has been found to vary a lot. Ukonmaanaho et al. (2001) applied the steady sate model in a forested catchment in Finland and found 129 years of residence time of Pb in the forest floor.Klaminder et al.(2006)using Pb-210 as a tracer for Pb migration found 250 years of residence in the mor layer of a mature forest in northern Sweden.

Soils

Estimates of Pb content in soils are strongly dependent upon the method used to digest soils and for this reason,comparisons between studies have to take into account the differences. It was found that total Pb burdens in soil estimated using HF digests were approximately 2.5 times greater than Pb burdens estimated using a HNO3/H2SO4digest(Watmough and Dillon 2007).The reason for this is that HF acid dissolves silicate bound Pb, which is likely natural in origin.The use of X-ray fluorescence and fusion with alkali carbonates or metaborate salts are also supposed to give total concentrations. In any case, the method must be mentioned.

The Pb concentrations in the soil layers (Table 3) are considered moderate and rather low even for a remote forest. Higher Pb concentrations (HF digests) were found in the soil layers of a mountainous fir forest in central Greece having the same soil parent material (flysch) as in our plot(Michopoulos et al.2018).In forest soil pr ofiles in Denmark the concentrations were 15 mg kg-1in the 0—10 cm lay er (HNO3digests) but dropped abruptly to 3.2 mg kg-1in the 10—20 cm (Hovmand et al. 2008).These soils were sandy, whereas the soils of our plot are characterized as silty loams. Table 3 also contains the values of some selected soil properties in our plot. It is remarkable that the layer 10—20 cm of our plot and that of Danish soils had approximately the same amount of organic C but different amounts of clay. In other cases,organic matter plays a significant role. Kulander et al.(2008) found high concentrations of total Pb (5 5 mg kg-1in a horizon sand an average of 25 mg kg-1in the rest mineral soil layers) in sandy soils albeit with high concentrations of organic C (3.2—6.9%) in mountains of northwestern Spain (X-ray-fluorescence method). Blaser et al.(2000)quoted the concentrations of total Pb and other metals in 23 soil profiles of Swiss forests (X-ray-fluorescence method). One of them had flysch rock as parent material.The soil in our plot had lower concentrations than those in flysch and more resembled (in terms of Pb concentrations) the soils derived from mica schist and marl. It seems that Pb deposition plays a far more important role than parent material in affecting the Pb concentrations especially in the surface soil horizons. In the past, before the phasing out of leaded petrol, the Pb enrichment of the forest floors was app reciable.Friedland and Johnson(1985)found 219 mg kg-1of Pb in the forest floors (HF acid digests) of co niferous forests in the northeastern USA and 100 mg kg-1in forested watersheds in the New Jersey,USA(HF digests) (Turner et al. 1985). Since that time, the Pb concentrations have dropped considerably in the forest floors of USA forests (Richardson et al.2014) (HCl/HNO3digests). Even for European forests, the Pb concentrations in the forest floor in our plot are rather low. In a large soil survey of European forests (including both coniferous and broadleaves species) in 2007 the range values for Pb concentrations for the L and FH horizons (HCl/HNO3digests)were 5—10 and 20—30 mg kg-1, respectively (De Vos and Cools 2011). The low Pb concentration found in the L horizon in our plot (Table 3) is probably due to the broadleaved nature of the forest canopies.In the L horizon of a fir forest in Greece t he Pb concentration was found approximately 5 mg kg-1(Michopoulos et al. 2018) justifying that coniferous litter is a very good metal scavenger.

There are places where the Pb concentrations in soils increase with time. Tang et al. (2015) found that the Pb concentrations (HF digests) in soils of in the eastern Tibetan plateau in China had higher concentrations in 2012 than 1990.

Isotopes ratios

Direct information on the206Pb/207Pb ratio of petrol in Greece since its introduction in the 1920s is not available.Kersten et al. (1997) set the ratio closer to that in British petrol (1.07 ± 0.01). However, the ratio values in the Pb dry deposition can be modified by other factors as not all the quantity of anthropogenic Pb found in the environment comes from petrol.In eight Swiss peat bog profiles Shotyk et al. (2000) found that 50—75% of anthropogenic Pb was preindustrial Pb. At a local scale, the introduction of Pb was done about 1000 years ago when the Cu mining became widespread (Mighal al. 2009). In a compilation of Pb values for old Greek copper ores in the Aegean area the range of the206Pb/207Pb values was between 1.19 and 1.20(A°berg et al. 2001). Also, the206Pb/207Pb ratio in Western Greece is influenced by other factors such as the combustion of lignite,having an average206Pb/207Pb value of 1.20(A°berg et al.2001).The same authors found that in Kozani,a city in northern western Greece, the air filters had a206Pb/207Pb value of 1.12,which was closer to Greek petrol than to lignite (1.20).

The remark made for total dissolution of soil minerals by strong acids with regard to heavy meal concentrations,also stands for the Pb isotopic ratios. The total dissolution involving HF gives higher values of206Pb/207Pb ratios.The reason is the existence of resistant minerals to weak acids like zirconium, which tend increase the206Pb/207Pb ratio(Erel et al. 2004).

The values of206Pb/207Pb in the various mineral soil layers had very low coefficients of variations, far lower than those in the Pb concentrations (Table 4). That means that the anthropogenic Pb migrated more or less uniformly down the soil profiles.The variability of the of206Pb/207Pb ratio in the flysch rocks was even lower than those in soils even in the deeper soil layers, which means that some mixture of the lithogenic and other Pb of different origin took place in the soil profiles.In forest soils in Sweden the isotopic ratio range was found 1.3—＞2 in the C horizons,whereas in surface soils the range was 1.14—1.18 (Bindler 2011). Similar distributions of206Pb/207Pb with depth was found by Wilcke et al. (2001)and Steinnes et al.(2005)in soils under Scotch pine in forest soils in Slovakia (HNO3/HClO4digests) and Norway (HNO3digests), respectively.In our work, all ratio values increased with depth apart from the 40—80 cm layer for reasons discussed below.

The statistical comparison of the206Pb/207Pb ratios in the soil layers with that of the parent rock material showed some very interesting results. It is expected that as we move further down the soil profile the differences between the ratios in the rocks and soils would be become nonsignificant. That is true for the 20—40 cm layer. However,the 40—80 layer has a ratio value that differs significantly from that of the parent rock material.In fact,the 40—80 cm layer has a lower value than that of the rock material.Similar result with regard to lower ratio values in deeper layers were found by Bra¨nnvall et al. (2001) in Swedish boreal soils(HNO3/HClO4digests)and Bacon et al.(2006)in soils of an upland catchment in Scotland (HF digests).The explanation given by Bacon et al. (2006) was that in the deeper soil layers it is not only the lithogenic Pb that dominates the206Pb/207Pb ratio but also the anthropogenic Pb of the preindustrial period.The much longer time space of the preindustrial Pb use allowed the latter to move deeper in soil profiles than the petrol Pb. Investigations by Hong et al. (1994) showed evidence of the Hellenic—Roman silver coinage production as far away as in Greenland ice cores. The question that arises is whether the preindustrial Pb had a206Pb/207Pb ratio value close to the 1.177 found in the 40—80 cm layer. From about 400 BC to 400 AD the206Pb/207Pb signature was about 1.18—1.19 in the ice cores mentioned above(Rosman et al.1997).In the UK the same ratio in atmospheric deposition prior to 1900 was about 1.17 (Farmer et al. 2002). So preindustrial values close to 1.177 determined in the 40—80 cm layer have been found in soils and cannot be excluded for the justification of the lower ratio values in that layer.

The method of determination the percentages of anthropogenic Pb in soils is based on the stable isotopic ratio206Pb/207Pb in petrol(related to its mineral origin)and its relation to that of lithogenic origin which also is stable depending of course on the rock type (Sturges and Barrie 1987). It was decided to use the average ratio value found in the European atmosphere, which was 1.13 (determined in humus samples) (Saether et al. 2011). In order for the percentages to be meaningful,the ratio values must differ significantly from that in the rock minerals. So the equation of Emanuel and Erel (2002) was applied to the L and H horizons and the 0—10, and 10—20 cm layers. The 20—40 cm layer has ratio values that did not differ significantly from that in the flysch rock. We can conclude that Pb derived from petrol migrated up to the 20 cm depth.As mentioned, the 40—80 cm layer did not contain Pb from petrol. Likewise, the percentages of the petrol derived Pb are meaningful up to the 20 cm depth and not to the depths of 20—40 cm where the ratios did not differ significantly and neither to the 40—80 cm depth where the dominating Pb is considered of preindustrial origin. The calculated percentages for the first four soil layers varied from 62%in the L horizon to 10.7%in the 10—20 cm layers(Table 4).In comparison with values found in other forest soils, the percentages are rather similar. Watmough and Hutchinson(2004) found that approximately 60% of Pb in the top soil under conifers in Ontario of Canada was anthropogenic(HNO3digests). Itoh et al. (2007) found 55% of anthropogenic Pb in the forest floor under Japanese cedar 120 km from Tokyo (HCl/HNO3digests). Other researchers have found higher values. In pristine areas of northern Norway more than 80% of the humus layer was found to be of anthropogenic origin (Steinnes et al. 2005). At 16 sites in northeastern United States the percentages of the petrol derived Pb were found about 90%in the forest floor and in the range of 52—72% in the lower mineral horizons(Richardson et al. 2014).

Conclusion

The 3 years of monitoring were not enough to have a mean residence time of Pb in the forest floor with low variability.The main reason is that each year has its own inputs of Pb(throughfall + litterfall) on the forest floor and therefore more observations are needed. As expected, the FH soil horizon had the highest Pb concentration among all parts of the ecosystem. Another important conclusion was that the anthropogenic preindustrial Pb is probably present in deeper soil horizons. The206Pb/207Pb ratio in the parent rock material can serve as a criterion to determine the percentage of petrol derived Pb in soils.