Environ. Sci. Technol. 2001, 35, 292-299
Lead Concentrations and Isotopic Ratios in the Sediments of the Sea of Galilee Y I G A L E R E L , * ,† Y A E L D U B O W S K I , † LUDWIK HALICZ,‡ JONATHAN EREZ,† AND AARON KAUFMAN§ Institute of Earth Sciences, The Hebrew University, Jerusalem, Israel 91904, The Geological Survey of Israel, 30 Malchei Israel Street, Jerusalem, Israel 95501, and Department of Environmental Sciences, Weizmann institute of Science, Rehovot, Israel
The isotopic composition and concentrations of Pb in the sediments of the Sea of Galilee (Lake Kinneret) were measured. The studied sediments have been deposited in the lake since the early 1900s (ca. 1920), hence Pb data record the transition from a period when the lake vicinity was sparsely populated to the present (approximately 100 000 people living in the area around the lake). In general, there is either a constant or a relatively slow increase in Pb concentrations from 40 cm depth (3.5-4.4 µg/g; ca. 1920) to 17 ( 2 cm below the sediment-water interface (3.7-7.2 µg/g;), which was deposited in the mid-1960s. From 17 ( 2 cm below the surface, there is a much faster increase up to 7 ( 2 cm below the surface (from 6.5 to 11.5 µg/g; 1982-1983), and from 7 ( 2 cm there is a gradual decrease in Pb concentrations toward the sedimentwater interface. At station G, near the outlet of the Jordan River, Pb concentrations drop between 29 and 25 cm below the surface, probably reflecting changes in the particulate load of the Jordan River due to the drying out of the Hula Swamp in the early 1950s. 206Pb/207Pb values in all the stations record most of the shifts displayed by Pb concentrations in the sediment. The estimated value of total Pb deposited annually in the lake sediment in the early 1990s is very close to the value obtained from measurements of Pb fluxes to the lake from eolian and fluvial sources. On the basis of the linear relationship between 206Pb/207Pb (or 208Pb/207Pb) and 1/[Pb], we argue that two endmembers contribute most of the Pb to the lake sediments. Sources of Pb to the lake include (i) the weathering of basalt from the eastern Galilee and the Golan Heights contributing 2.6 ( 0.5 µg/g Pb to the sediment and (ii) anthropogenic Pb that is affecting both surface and deep (from 30 to 40 cm) lake sediments. At station S, a third source, Pb released from soils developed on carbonates, should be considered as well.
Introduction It has been widely documented that in many cases industrial Pb masks natural Pb concentrations in the environment * Corresponding author phone: 972-2-6586515; fax: 972-25662581; e-mail:
[email protected]. † The Hebrew University. ‡ The Geological Survey of Israel. § Weizmann institute of Science. 292
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 2, 2001
(1-3). It has been estimated that on a global scale anthropogenic Pb pollution accounts for 96% to greater than 99% of total atmospheric Pb deposition (4, 5). In addition, it is well-established that there exists a continuum of toxicological effects associated with Pb across a broad range of exposure (6). Therefore, much effort has been put into tracing anthropogenic Pb in different parts of the environment in general and in lakes and lake sediments, in particular (4, 5, 7-17). Some of these studies have used stable Pb isotopes to distinguish between different sources of Pb to the lake, e.g., natural (rock-derived) vs industrial (4, 5, 11, 13, 14, 16, 17). The introduction of Pb to lakes occurs from atmospheric deposition and from surface and groundwater runoff. Lead has four stable isotopes: 204Pb (2%), 206Pb (25%), 207Pb (21%), and 208Pb (52%). While 206Pb, 207Pb, and 208Pb are formed by the radioactive decay of 238U, 235U, and 232Th, respectively, 204Pb is a primordial nuclide. During the last billion years, the abundance of 207Pb has changed very little with time as compared to 206Pb since most 235U has already decayed while 238U still has a high abundance. Hence, the ratio of the high abundance isotopes 206Pb/207Pb and 208Pb/ 207Pb can be used instead of 206Pb/204Pb and 208Pb/204Pb. Anthropogenic Pb is derived from lead sulfide ore deposits and is released to the environment primarily by the combustion of gasoline (petrol lead) and as a byproduct of industrial activities such as smelting. The 206Pb/207Pb ratio of most ore bodies ranges from 0.92 to 1.20 (18) and, in many cases, is less radiogenic than rock-released Pb (usually 206Pb/207Pb values are higher than 1.20). The major source of anthropogenic Pb in Israel is petrol lead (19). The onset of extensive petrol lead consumption in the country was in the mid-1960s and reached its peak in the late 1980s (20). Since the introduction of unleaded gasoline in Israel in the early 1990s, there has been a decrease in Pb emission. The 206Pb/207Pb values of petrol lead in Israel was 1.15 in the 1960s, rose gradually to 1.22 in the 1970s and early 1980s, declined gradually to 1.18 in the late 1980s, and then declined sharply to 1.12 in the early 1990s due to the transition from U.S. to European petrol lead (19). Hence, the major goal of the current study is to determine the extent of Pb pollution of the lake sediment using Pb concentrations and isotopic composition and to relate it to major events in the recent history of the lake and its watershed.
Methods Study Area. The Sea of Galilee (Lake Kinneret), a mesotrophic, monomictic lake with a surface area of 168 km2, a maximum depth of 42 m, and a mean depth of 24 m is located in the Jordan River Rift Valley (21) (Figure 1). The water level in the lake is approximately 200 m below sea level. The bedrock in the watershed is a mixture of Jurassic limestone, CretaceousEocene carbonates, and Pliocene-Pleistocene basalts. The area of the watershed is approximately 2600 km2 (22). The main sources of water to the lake include the Jordan River, springs, and small streams (22). The residence time of lake water is on the order of 5 yr (23). pH in the lake water fluctuates from 8.0 to 9.1, the ionic strength of the lake water is approximately 15 mM, and the alkalinity varies between 1.2 and 2.4 mequiv/L (24). The major dissolved ions in the lake water are as follows: Na+ (4.7-5.0 mM), Ca2+ (1.0-1.5 mM), Mg2+ (∼1.2 mM), K+ (∼0.15 mM), and Cl- (6.1-6.5 mM) (24). Epilimnion concentrations of SO42- are about 0.5 mM, whereas hypolimnion concentrations vary as a result of redox processes (25). The concentration of dissolved organic carbon is 2-4 mg/L (24). The lake sediment is composed of carbonate (47%), aluminosilicates (46%), and organic matter (7%) (26). 10.1021/es0013172 CCC: $20.00
2001 American Chemical Society Published on Web 12/09/2000
FIGURE 1. Location map of Lake Kinneret including the sediment sampling stations (A, G, and S). Nevertheless, these values vary spatially; for example, near the Jordan River inlet the amount of aluminosilicates is much higher, up to 70% (26). Approximately two-thirds of the carbonate and 90% of the organic matter in the sediment are authigenic (27). The lake has a relatively high sedimentation rate with an average value of 1000 g m-2 yr-1 (0.3 cm/yr; 26, 28, 29). Because of the high sedimentation rate and the high affinity of Pb to particulate matter in the water column (4, 8, 13, 14, 30), the sediments of Lake Kinneret should record past fluctuations in Pb pollution in the region. Sampling. Lake sediments were collected from three sites (A, G, and S; Figure 1) using a gravity corer (i.d. ) 2.5 cm). During sampling, a layer of water was left above the sediment in order to prevent damage to the top part of the core. Four cores were collected at site A in the deepest part of the lake (∼42 m depth); three of them (1, 4, and 5) were used for sediment dating, and the fourth one was used for Pb analysis. Shortly after collection, the cores were described and measured, and core 4 was X-rayed. The cores were then sliced to 1-cm slabs for dating and 2-cm slabs for Pb work. Porosity was determined only in core 4 where a small decrease from 0.94 at the sediment surface to 0.84 at a depth of 40 cm was observed. Station G is located north of station A, downstream from the Jordan River inlet, where the water depth is about 20 m. The sedimentation rate at station G is higher than in station A (0.3-0.5 cm/yr; 28, this study) and is on the order of 0.8 cm/yr (28). Station S is located in the east part of the lake where the water depth is approximately 20 m. The sedimentation rate at station S is approximately 0.5 cm/yr (28). The core slabs were dried in the laboratory (at 105 °C), and the water content was determined in some of them. Sediment Dating. Sedimentation rate in station A was determined by radiometric dating (210Pb) of one core (no. 4) and by identifying several extreme limnological events in
that core and in additional two cores (nos.1 and 5). For the radiogenic dating, 400-600-mg samples were dissolved in concentrated HNO3, H2SO4, and HClO4; evaporated to dryness; and redissolved in 0.5 N HCl. Ascorbic acid was then added, and Po was self-plated on a silver disk at 90 °C for 4 h. The activity of 210Pb was determined by R-spectrometric measurements of 210Po. The efficiency of the extraction and of the R-counting was monitored with a 208Po spike. Sedimentation rate was calculated based on the exponential decay of the unsupported 210Pb using the CIC model (31). Carbon Isotope Measurements. δ13C values of the carbonate fraction in the sediment were measured following the procedure of McCrea (32). After removing the mobile organic matter from the sediment subsamples (30-60 mg) by a dilute NaOCl solution, the residue was reacted with concentrated phosphoric acid. The δ13C value of the released CO2 was calculated (relative to PDB) using a VG-602 mass spectrometer. The samples of core 5 were analyzed at the Israeli Geological Survey using an isocarb system combined with a SIRA II (VG) mass spectrometer. The standard deviation (1σ) of the δ13Ccarbonate measurements was 0.03‰ in cores 1 and 4 and 0.13‰ in core 5. To measure δ13C values of organic carbon in the sediment, carbonate was first removed with HCl. In cores 1 and 4, the carbonate-free samples were first filtered, and the residue (∼20 mg) was heated to 540 °C in closed ampules under vacuum in the presence of CuO and silver. The δ13C value of the released CO2 was determined (relative to PDB) using a VG-602 mass spectrometer. In cores 4 and 5, the isotopic composition was measured in the filtered carbon-free samples with an automatic carbon and nitrogen analyzer (ANCA, Europa Scientific). However, it was found that filtering the samples effected their isotopic composition. Therefore, core 5 was also analyzed without filtering. To compare the different cores, they were all calibrated as if analyzed using the ANCA without filtering where dextromed was used as an internal standard with a standard deviation of 0.2‰ (1σ). The paleolimnological results of carbon isotopes will be discussed in a separate paper (Dubowski et al., in preparation). Lead Analysis. The entire sediment core slabs were leached with Suprapure 3 M HNO3 acid at 100 °C. In addition, a few samples were completely digested by sintering with Na2O2. It was found that all the Pb in these samples was released by the 3 M HNO3 treatment. The samples from core 4 (used for 210Pb dating) were leached with 0.5 M HNO3 at room temperature. The supernate was diluted with ×2 ion exchanged water and then analyzed. Blank levels were checked periodically and never exceeded 1% of Pb concentrations in the sediment. A Perkin-Elmer Sciex Elan 6000 inductively coupled plasma mass spectrometer (ICP-MS) was used for the concentration and isotopic measurements of Pb. The sensitivity of the ICP-MS was about 2.8 × 107 counts per µg/g for 208Pb, and the limit of detection (3σ) was 2.1 ng/L. The dwell time (the amount of time spent measuring the signal at each mass during repetitive scanning over the three isotopes), which is the most critical data acquisition parameter, was 25 ms. The integration time was 5.5 s (221 sweeps). Each recording consisted of nine 5.5-s integrations of each mass peak, and the relative standard deviation (RSD) of the ratios was 0.3%. The short-term drift (up to 5 h) of the measurement was lower than the RSD of a single recording. Furthermore, the RSD of 10 recordings (during 5 h) was about 0.06-0.08%. The long-term drift (4 months) of isotopic ratios was equal to the RSD of a single measurement and was about 0.2% (33). It was observed that the isotopic values of Pb in the concentration range from 5 to 150 ng/g were independent of Pb concentrations. The isotopic measurements were checked with a 75 ng/g solution of NIST SRM 981 (common VOL. 35, NO. 2, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
293
FIGURE 2. δ13C depth profiles of (a) organic matter and (b) carbonate in the three cores from Lake Kinneret [core 1, open squares, has the shortest record (30 cm); core 4, open triangles; core 5, open circles, has the longest record (60 cm)]. In addition, horizons are shown that bear evidence of extreme climatic or limnological events whose times of occurrence are well-known. Average sedimentation rates were calculated using the shown identifiable horizons. lead isotopic standard). The mass discrimination of the SRM 981 was 0.994 for the 207Pb/206Pb ratio and 0.982 for the 208Pb/206Pb ratio. Prior to the present study, the isotopic ratios of more than 30 soil and rock samples were calibrated against SRM 981 and then compared with the results obtained with a TIMS; all the ICP-MS results agreed within 0.3% with the TIMS (33). 294
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 2, 2001
Results Sediment Stratigraphy and Dating. Five events were identified in the studied cores: (i) the flood of 1894 (34); (ii) the flood of 1969 (35); (iii) the draining of Lake Hula (an upstream swamp and a shallow lake) starting in 1950 (36); and the widespread blooms of Melosira granulata in (iv) 1982 and
TABLE 1. Lead Concentration (µg/g) and Isotopic Ratios (206Pb/207Pb, from the Water-Sediment Interface (cm)a station A
208Pb/207Pb)
of Lake Kinneret Sediments at Different Depths
station S
station G
depth
[Pb]
206Pb/207Pb
208Pb/207Pb
depth
[Pb]
206Pb/207Pb
208Pb/207Pb
depth
[Pb]
206Pb/207Pb
208Pb/207Pb
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
9.1 10.5 10.0 11.4 10.7 8.8 8.6 7.3 7.2 6.7 6.5 5.9 5.9 5.8 5.7 5.6 5.5 5.3 4.6 4.9
1.177 1.183 1.182 1.180 1.182 1.188 1.188 1.190 1.190 1.190 1.191 1.195 1.198 1.202 1.202 1.203 1.206 1.206 1.207 1.209
2.448 2.457 2.454 2.449 2.450 2.454 2.456 2.459 2.459 2.460 2.462 2.466 2.465 2.473 2.471 2.472 2.475 2.475 2.477 2.478
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37
5.2 5.9 6.8 5.5 5.3 5.0 4.7 4.8 3.0 3.7 3.7 3.4 3.3 3.7 3.5 3.4 3.4 3.5 3.6
1.176 1.180 1.181 1.180 1.181 1.185 1.187 1.186 1.207 1.199 1.196 1.202 1.204 1.194 1.194 1.202 1.200 1.196 1.200
2.450 2.458 2.456 2.454 2.455 2.461 2.462 2.466 2.483 2.480 2.472 2.481 2.479 2.473 2.474 2.476 2.478 2.472 2.480
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
7.3 7.4 7.2 7.8 7.6 7.1 7.0 6.4 5.8 4.1 4.2 3.7 3.6 4.4 5.5 5.5 5.5 5.6 4.7 4.6
1.190 1.192 1.193 1.195 1.194 1.192 1.195 1.194 1.198 1.210 1.210 1.215 1.219 1.210 1.198 1.197 1.203 1.201 1.197 1.198
2.455 2.457 2.456 2.462 2.458 2.457 2.462 2.460 2.465 2.481 2.478 2.485 2.486 2.478 2.464 2.463 2.467 2.466 2.469 2.469
a The data are from cores collected in three different locations (see text). The precision (1σ) of the isotopic data is (0.3% and of the concentration data is (2%.
FIGURE 3. Unsupported 210Pb (210Pbexcess) specific activity (dpm/g) along core 4 and sedimentation rate as calculated from the exponential decay of 210Pb. On the basis of stratigraphic dating, the deepest sample (38 cm, ∼71 yr old) is expected to contain only ∼10% of the original unsupported 210Pb. Therefore, 210Pbexcess values were defined as follows: 210Pbexcess ) 210Pbmeasured - {210Pbmeasured at 38 cm depth - K}, where K is the correction for these 10% [equal to 2.14 dpm/g as determined by iteration from the best straight line fitted through a plot of ln(210Pbexcess) vs depth]. (v) 1988 (37). The flood events of 1969 and 1894 are represented by high content of silicate clays, lower δ13C values of the carbonate, and higher δ13C values of the organic carbon in the sediment at 14 cm in cores 1 and 5 and at 51 cm in core 5 (Figure 2). The imprint of the 1969 flood was better observed in core 1 probably because this core was sampled farther north than the other two cores, closer to the Jordan River inflow. Core 5 is the only core long enough to document the 1894 flood (Figure 2). The reasoning for the observed shifts in δ13C values as a result of flood events is given elsewhere (26, 31, 38). The draining of the Hula Swamp at the beginning of the 1950s introduced large amounts of organic-rich soil-derived sediment with high δ13C values (39; Figure 2). The two extensive blooms of M. granulata in 1982
and 1988 (37) were clearly detected by unusually high concentrations of the diatom fractures in all three cores. The measured specific activity of 210Pb in the sediments varies from 10 dpm/g at the sediment surface to 3 dpm/g at the base of the core. On the basis of the stratigrafic ages, it was estimated that the deepest sample (38 cm) is 71 yr old and therefore should contain only approximately 10% of the original excess 210Pb. The latter was determined to be 2.14 dpm/g using the best fit line in a plot of ln(210Pbexcess) vs depth (Figure 3). The excess 210Pb in each sample was then calculated by subtracting the 210Pb activity of the deepest sample (after a correction for its excess 210Pb content) from its measured value. The exponential decrease of excess 210Pb with depth (Figure 3) suggests that the sedimentation rate was more or less constant during that time period (40). Lead Concentrations and Isotopic Composition. At station A, in the center of the lake, there is a continuous increase in Pb concentrations in the sediment starting at the bottom of the core (i.e., 39 cm; ∼ 4.5 µg/g; Table 1and Figure 4). At station S, in the eastern part of the lake, the concentrations of Pb in the sediment remain constant from 37 to 17 cm (∼3.5 µg/g; Table 1 and Figure 4). There is a significant increase in Pb concentrations from 17 to 5 cm below the surface (∼6.5 µg/g) that is more or less parallel to the one observed at station A. At station G, in the northern part of the lake, the concentration of Pb is constant from 39 to 29 cm (∼5 µg/g; Table 1 and Figure 4). From 29 to 25 cm, Pb concentrations drop to ∼3.5 µg/g, and from 25 cm, they start to increase (Table 1; Figure 4). In all stations, maximum Pb concentrations were found 5-7 cm below the sedimentwater interface (Table 1; Figure 4). 206Pb/207Pb values in all the stations record most of the shifts displayed by Pb concentrations (Figures 4 and 5).
Discussion Sedimentation Rate. On the basis of δ13C values of the organic matter in the sediment, the stratigraphic correlation between the cores was established (Figure 2). From the sediment thickness and the time interval between each two identified events (datum planes), average sedimentation rates were determined (Figure 2). According to the stratigraphic data, the sedimentation rate was 0.40-0.55 cm/yr except for the VOL. 35, NO. 2, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
295
FIGURE 4. Profiles of Pb concentration (µg/g) in the three cores of the lake sediment.
FIGURE 5. Profiles of
206Pb/207Pb
ratio in the three sampled cores of the lake sediment.
period between 1950 and 1969 (∼0.7 cm/yr, Figure 2). Using these sedimentation rates and the stratigraphic correlation between the cores, stratigraphic ages were determined for core 4 (Figure 2). The age of the deepest part of the core (below 25 cm) has the largest uncertainty since the age of the bottom of core 4 was estimated only by stratigraphic correlation with core 5, in which a constant sedimentation rate was assumed between 1894 and 1950. Recent independent 137Cs measurements on other cores from station A are in agreement with the stratigraphic ages, where depths of 17 and of 4-5 cm are from 1963/1964 and 1986, respectively (41). An average sedimentation rate of 0.45 cm/yr was calculated from a plot of excess 210Pb activity versus depth (Figure 3). Previous studies have suggested that sedimentation rates at the center of Lake Kinneret range from of 0.2 to 0.6 cm/yr (26, 28, 41-43). Our result (0.45 cm/yr) is therefore in good agreement with previous studies and with the stratigraphic dating (Figure 2). Nevertheless, 210Pb data do not indicate the increase in sedimentation rate between 1950 and 1969 suggested by the stratigraphic dating. This can be the result of an insufficient number of 210Pb data samples at this part of the core. Accumulation of Anthropogenic Pb in the Sediment. From a concentration of 4.4 ( 0.6 µg/g Pb at a depth of 296
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 2, 2001
approximately 40 cm, Pb concentrations rise to maximum values of 6.8, 7.8, and 11.4 µg/g in stations S, G, and A, respectively, at 5-7 cm from the surface (Table 1; Figure 4). The highest concentrations of Pb are recorded in the sediment of station A, the closest among the studied stations to Tiberias (the largest city around the lake). The average concentration of Pb in the top 39 cm of the lake sediment in the three cores is 5.9 ( 2 µg/g (Table 1), and the average sedimentation rate in the lake is approximately 1000 g m-2 yr-1 (26). From these values, the average flux of Pb to the lake area (168 km2) is estimated as ∼1000 kg/yr. The amount of total Pb added to Lake Kinneret sediment in the last 80 yr is therefore 80 t. On the basis of the background concentration of Pb in the sediment (2.6 ( 0.6 µg/g; see below), about 45% of the total Pb added during the last 80 yr has been of natural origin. Similarly, during the 1990s (approximately the top 3 cm of the sediment) (Figure 2), the average concentration of Pb in the sediment was 7.7( 2 µg/g; hence, the annual flux of Pb was ∼1200 kg. During this period, anthropogenic Pb made up approximately two-thirds of the total Pb added to the lake sediment annually (∼800 kg/yr). This estimated value of total Pb deposited annually in the sediment during the 1990s (1200 kg) is very close to the value obtained from measurements of Pb fluxes to the lake from eolian and fluvial sources in 1993-1995 (44).
The average annual consumption of alkyl lead in Israel in the early 1990s was on the order of 600 t/yr (as tetraalkyl lead, TAL) (20). This means that approximately 400 t of Pb has been emitted annually to the atmosphere in Israel during the early 1990s (about 65% of Pb in TAL). It is estimated that traffic in the watershed of Lake Kinneret and in the lake vicinity accounts for 15% of the yearly automobile traffic in Israel (45); hence, about 75 t of petrol lead has been emitted in the area directly affecting the lake. Since we estimated that approximately 800 kg/yr of anthropogenic Pb has been deposited annually in the lake during the 1990s, roughly 1% of the emitted petrol lead has been deposited in the lake sediment (although higher Pb concentrations might be expected closed to major population centers such as Tiberias). Therefore, most of the petrol lead has not traveled long distances and still resides in soils along roads in the area. This conclusion is supported by the well-established observation that most of the Pb emitted from cars is accumulated along roads (6) and is likely to be transported slowly through the watershed to the lake, a process that might take centuries. Sources of Pb to the Lake Sediment. Sources of Pb to the lake should include (i) natural Pb released by rock and soil weathering and (ii) anthropogenic Pb, mostly petrol lead (since there are no major industrial activities in the lake vicinity). Both kinds of Pb are introduced to the lake via the Jordan River and other smaller streams, eolian dust and aerosols, and springs. As indicated previously, there are two major kinds of rock types in the lake watershed: basaltic rocks and carbonates. The average Pb isotopic value of basaltic rocks from the Golan Heights and the Galilee (206Pb/207Pb ) 1.232 ( 0.008, 208Pb/207Pb ) 2.499 ( 0.006) (46), plotted in Figure 6a,b, is assumed to delineate the isotopic composition of Pb released by basalt weathering to the streams and springs that feed the lake. Lead released from the weathering of carbonate rocks is originated either in the residue (mostly aluminosilicates) or in the acid-soluble fraction (carbonate) of the rock (19). The 206Pb/207Pb ratio of Pb in the residue of Cretaceous carbonate rocks in Israel ranges between 1.198 and 1.206, and the 208Pb/207Pb ratio ranges between 2.475 and 2.492 (19). The 206Pb/207Pb ratio of the carbonate fraction of Cretaceous carbonate rocks in Israel ranges between 1.24 and 1.60, and the 208Pb/207Pb ratio ranges between 2.46 and 2.48 (19). The isotopic compositions of both types of carbonate-rock lead (insoluble-residue lead and carbonate-fraction lead) are different from the isotopic composition of basalt lead (Figure 6a,b). Mount Hermon, located at the headwater of the Jordan River, is composed of Jurassic carbonates but has a high abundance of sulfide minerals, including galena (PbS) (47). The isotopic composition of galena lead from Mount Hermon (206Pb/207Pb ) 1.174 ( 0.005, 208Pb/207Pb ) 2.471 ( 0.009) (48) is plotted in Figure 6a. Anthropogenic Pb in Israel has had a wide range of values (19). 206Pb/207Pb values range from 1.15 before 1967 to 1.22 in the 1980s (19, 49). Currently, most cars in Israel use unleaded gasoline, and the ones that do, use alkyl lead produced in Europe with 206Pb/207Pb values of 1.11-1.12 (19). Although the number of cars in Israel that use leaded gasoline is declining, they are still the major source of anthropogenic Pb in the atmosphere of Israel (19). The average 206Pb/207Pb value of anthropogenic Pb in Israel in the early 1990s was estimated to be 1.185 (19). The linear relationship between 206Pb/207Pb (or 208Pb/ 207Pb) versus 1/[Pb] suggests that only two end-members contribute significant amounts of Pb to the lake sediment (50; Figure 6c). The values of Pb from station G overlap with Pb from A, but Pb at station S forms a distinct linear array. An inspection of 206Pb/207Pb versus 208Pb/207Pb values of Pb in the three sediment cores reveals that the weathering of basalt is indeed one source of Pb to the lake (Figure 6a). It
FIGURE 6. (a) 206Pb/207Pb vs 208Pb/207Pb of the sediment samples and of several possible sources of Pb to the lake. (b) 206Pb/207Pb vs 208Pb/207Pb of the deepest samples from stations S and G. (c) 206Pb/207Pb vs 1/[Pb] of the same profiles. is even more evident when only the deeper samples at station S and G (17-37 cm in S, and 19-39 cm in G) are included (Figure 6b). Lead in the deeper part of the sediment in stations S and G has alternating isotopic values ranging from 206Pb/ 207Pb of 1.193-1.219 and 208Pb/207Pb values of 2.463-2.485 (Table 1). The more radiogenic values are close to the average isotopic value of basaltic rocks from the Golan Heights and east Galilee (Figure 6b). The relative contribution of natural Pb from the insoluble residue of carbonate bedrock to the Pb budget of the lake is unclear, but it seems that this Pb cannot be the other end-member of Pb in the deeper lake sediment (Figure 6b). It is also clear that Pb released from the carbonate fraction in carbonate rocks or from galena cannot be a major source of Pb to the lake (Figure 6a,b). These observations led us to the following conclusions: (i) Pb released from the weathering of basalt makes up the radiogenic end-member of Pb in the lake, and it affects all stations, albeit in different proportions. Extrapolating the VOL. 35, NO. 2, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
297
best fit lines in Figure 6b,c to the measured Pb isotope ratios of the basalt end-member, it can be estimated that the concentration of basalt-derived Pb in the lake sediment is 2.6 ( 0.5 µg/g. This value is probably the background, natural concentration of Pb in the lake sediment. (ii) The other endmember of Pb in the lake can only be anthropogenic Pb based on its estimated isotopic composition and the high concentration of Pb. It seems that anthropogenic Pb in Israel is affecting both surface and deep (30-40 cm) lake sediment. The nonradiogenic (anthropogenic-like) end-member, found in the deeper lake sediment, could be Pb released from both Israeli and Syrian sources prior to 1967 when the lake was on the border between the two countries. Therefore, even prior to the extensive human activity in the region, i.e., prior to 1967, Pb in the lake sediment was contributed by two major end-members: basaltic rocks and anthropogenic sources. (iii) Station G, located near the inlet of the Jordan River, has Pb characteristics that are similar to the ones at station A located at the center of the lake, reflecting the fact that the Jordan River is the major source of water and solutes to the lake. (iv) The cyclic trend in the deeper part of the lake sediment (at least at stations S and G) records shifts between Pb contributed by basaltic rocks and anthropogenic sources, which in turn probably reflect changes in the hydrology of the Kinneret Watershed. On the basis of the average sedimentation rate in the studied stations, each cycle lasted for approximately 10 yr (Table 1; Figure 2). The exact nature of these cycles requires further investigation. (v) The trend of Pb isotopic values at station S plots closer to the isotopic values of Pb released from the insoluble residue of carbonate bedrock. This is because of the fact that station S is located near the inlet of Vaddi Samach that drains the southern part of the Golan Heights. Although, in general, basalt is the main bedrock type in the Golan Heights, Eocene and Miocene carbonates, sandstone, and shales are exposed in most of the drainage area of Vaddi Samach (51). Events in the Lake History Recorded by Pb. Currently we have stratigraphy and chronological data only for cores from station A (Table 1; Figures 2 and 3), hence our discussion will focus on this station. There are two general shifts in both Pb concentration and isotopic composition that are recorded in all the stations. From 17 cm below the sediment-water interface (19 at station G), there is an increase in Pb concentration accompanied by a decrease in 206Pb/207Pb values (Table 1; Figures 4 and 5). This trend ends at 7 ( 2 cm below the sediment-water interface, where Pb concentrations start to decline (or remain constant at station G; Figures 4 and 5). On the basis of our sediment dating at station A, the 17 cm deep layer represents the mid-1960s. Hence, the increase in Pb concentrations at this depth records the 1967 war, when the Golan Heights was seized by the Israeli army and, as a result, the numbers of motorized vehicles and tourists in the region have increased dramatically. There are some differences between station A and the other two stations. Whereas at station A the increase in Pb concentration in the sediment start from the bottom of the core, at station G, Pb concentrations are constant from 39 to 29 cm, and they start to increase only at depth of 25 cm (Table 1). At station S, Pb concentrations are constant from 37 to 17 cm and then start to increase (Table 1). Therefore, we do not argue that the 17 or 19 cm deep layer in the different stations is an unequivocal marker of the year 1967. We merely say that the increase in human activity around the lake in the 1960s is recorded in the profiles of Pb concentration and isotopic composition in the sediment. The decline in Pb concentration and the shift in Pb isotopic composition at 7 ( 2 cm below the surface (Figures 4 and 5) might record the decline in petrol lead maximum allowance in 1980s (20). At station G, an additional event is recorded by Pb concentration and isotopic composition in the sediment 298
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 2, 2001
FIGURE 7. 206Pb/207Pb of 0.5 M HNO3 leaches vs 206Pb/207Pb of 3 M HNO3 leaches of lake sediment collected in station A. The numbers represent sample depth (cm) below the surface. (Table 1; Figures 4 and 5). From 29 to 25 cm below the surface there is a decline in Pb concentration accompanied by an increase in 206Pb/207Pb values (Table 1). From 25 to 19 cm there is a slow increase in Pb concentration and a gradual decrease in 206Pb/207Pb values (Table 1). This indicates that the samples from 29 to 19 cm below the surface at station G contain less anthropogenic Pb and that more Pb is released from the basaltic, natural end-member. This event is not recorded in a simple way in the dated sediment of station A, although at these depths there is a slower increase in Pb concentrations (Table 1; Figure 4). However, when plotting the 206Pb/207Pb values of the 0.5 M HNO3 versus the 3 M HNO3 leaches of the sediment at station A, 206Pb/207Pb ratio at this depth range behave differently than in other samples (Figure 7). In the whole profile, except for samples between 31 and 21 cm there is a positive linear relationship between 206Pb/207Pb values of both leaches. Only samples from 31 to 21 cm display a negative relationship. The reason for this behavior is that, in samples that contain appreciable amounts of anthropogenic Pb, there is a linear relationship between the two types of acid treatments where the stronger acid releases more radiogenic Pb (less anthropogenic) than the weak acid (19). In pristine, relatively uncontaminated samples (like the samples from 29 to 19 cm below the surface at station G or from 31 to 21 cm at station A), there is a completely different mechanism. The isotopic composition of Pb released from the rock is a mixture of Pb released from different minerals. In many cases, the more radiogenic minerals are the first to release Pb (weak acid treatment), and common Pb is released by stronger treatment, implying a negative correlation between the two leaches (52, 53). According to Figure 2, the 29-19-cm interval at station A represents the time between the mid-1940s and the mid1960s. Therefore, the shift in Pb isotopes and concentration might record the parching of the Hula Swamp along the Jordan River, upstream from the lake, which started in 1950. The decrease in Pb concentration downstream from the swamp can be accounted for by the recorded changes in the redox properties of the Jordan Valley in this region (54). It has been extensively recorded that this event affected the nutrient supply to the lake (21, 39, 55). One of the outcomes of the draining of the swamp was the oxidation Fe(II) and Mn(II) and the precipitation of iron and manganese oxides and hydrous oxides which then might have adsorbed Pb from the Jordan River water (54, 56). In addition, the draining of the swamp reduced the residence time of Jordan River water in the Hula Basin and affected the sedimentation rate in the northern part of Lake Kinneret (27). Hence, these processes could have caused a decrease in Pb concentration in the Jordan River and in the lake near its inlet.
Literature Cited (1) ) Murozumi, M.; Chow, T. J.; Patterson, C. Geochim. Cosmochim. Acta 1969, 33, 1247. (2) Meybeck, M.; Helmer, R. Palaeoclimatol., Palaeoecol. (Global Planet. Change Sect.) 1989, 75, 283. (3) Nriagu, J. O. Nature 1989, 338, 47. (4) Shirahata, H.; Elias, R. W.; Patterson, C. C. Geochim. Cosmochim. Acta 1980, 49, 149. (5) Flegal, A. R.; Nriagu, J. O.; Niemeyer, S.; Coals, K. H. Nature 1989, 339, 455. (6) Air Quality Criteria for Lead; U.S. Environmental Protection Agency, Office of Research and Development: Research Triangle Park, NC, 1986; EPA/600/8-83/028bF. (7) Edgington, D. N.; Robbins, J. A. Environ. Sci. Technol. 1976, 10, 266. (8) Nriagu, J. O.; Kemp, A. L. W.; Wong, H. K. T.; Harper, N. Geochim. Cosmochim. Acta 1979, 43, 247. (9) Goldberg, E. D.; Hodge, V. F.; Griffin, J. J.; Koide, M. Environ. Sci. Technol. 1981, 15, 466. (10) Griffin, J. J.; Goldberg, E. D. Environ. Sci. Technol. 1983, 17, 244. (11) Petit, D.; Mennessier, J. P.; Lamberts, L. Atmos. Environ. 1984, 18, 1189. (12) Renberg, I.; Persson, M. W.; Emteryd, O. Nature 1994, 368, 323. (13) Ritson, P. I.; Esser, B. K.; Niemeyer, S.; Flegal, A. R. Geochim. Cosmochim. Acta 1994, 58, 3297. (14) Graney, J. R.; Halliday, A. N.; Keeler, G. J.; Nriagu, J. A.; Robbins, J. A.; Norton, S. A. Geochim. Cosmochim. Acta 1995, 59, 1715. (15) Birch, L.; Hanselmann, K. W.; Bachofen, R. Water Res. 1996, 30, 679. (16) Moor, H. C.; Schaller, T.; Sturm, M. Environ. Sci. Technol. 1996, 30, 2928. (17) Camarero, L.; Masque, P.; Devos, W.; Ani-Ragolta, I.; Catalan, J.; Moor, H. C.; Pla, S.; Sanchez-Cabeza, J. A. Water Air Soil Pollut. (in press). (18) Sturges, W. T.; Barrie, L. A. Nature 1987, 329, 144. (19) Erel, Y.; Veron, A.; Halicz, L. Geochim. Cosmochim. Acta 1997, 61, 4495. (20) Israeli Oil Refinery, The Ministry of Infrastructure, archive; Haifa, Israel, 1998. (21) Serruya, C. Lake Kinneret, Monographiae Biologicae; Dr. W. Junk B. V. Publishers: The Hague, The Netherlands, 1978. (22) Michelson, H. In Lake Kinneret, Monographiae Biologicae; Serruya, C., Ed.; Dr. W. Junk B. V. Publishers: The Hague, The Netherlands, 1978; p 17. (23) Mero G.; Ben Zvi A., The Hydrological Survey of Israel, Jerusalem, unpublished data, 1975. (24) Lake Kinneret Data Base, The Laboratory of Lake Kinneret, Tiberias, Israel, 1996. (25) Eckert, W.; Truper, H. G. Biogeochemistry 1993, 21, 1. (26) Serruya, C. In Interaction between sediment and freshwater; Golterman, H. L., Ed.; Dr. W. Junk B. V. Publishers: The Hague, The Netherlands, 1977; p 48. (27) Stiller, M. Isr. J. Earth Sci. 1979, 28, 1. (28) Stiller, M.; Assaf, G. IAHS-AISH symposium; Helsinki, 1973. (29) Stiller, M.; Imboden, D. M. In Sediments and water interactions; Sly, P. G., Ed.; Springer-Verlag: New York, 1986; p 501. (30) Erel, Y.; Patterson, C. C. Geochim. Cosmochim. Acta 1994, 58, 3289.
(31) Dubowski, Y. Stable isotopes paleolimnology of Lake Kinneret. M.Sc. Thesis, The Hebrew University of Jerusalem, 1996. (32) McCrea, J. M. J. Chem. Phys. 1950, 18, 129. (33) Halicz, L.; Erel, Y.; Veron, A. At. Spectrom. 1996, 17, 186. (34) Ashbel, D. Regional Climatology of Israel; The Hebrew University: Jerusalem, 1951; 244 pp. (35) Gophen, M., Itzhaki, G. Lake Kinneret; Ministry of DefenseIsrael Press: 1992; 335 pp. (36) Stiller, M. Rates of transport and sedimentation in Lake Kinneret. Ph.D. Thesis, Weizmann Institute of Science, 1974. (37) Berman, T.; Yacobi, Y. Z.; Pollinger, U. Aquat. Sci. 1992, 54, 2, 104. (38) Stiller M. In Interaction between sediment and freshwater; Golterman, H. L., Ed.; Dr. W. Junk B. V. Publishers: The Hague, The Netherlands, 1977; p 57. (39) Selinger Y.; Phosphate geochemistry in Lake Kinneret watershed. Ph.D. Thesis, Technion, 1991. (40) Appleby, P. G., Oldfield, F. In Uranium disequilibrium applications to earth, marine and environment sciences; Ivanovich, M., Harmon, R. S., Eds.; Clarendon Press: Oxford, U.K., 1992; p 731. (41) Kirchner, G.; Stiller, M.; Nishri, A.; Koren, N. A Simultaneous measurement of 137Cs, 210Pb, and 226Ra in sediments of the Dead Sea and Lake Kinneret by γ-spectroscopy. Manuscript in preparation. (42) Serruya, C. Verh. Int. Ver. Limnol. 1974, 19, 1357. (43) Nishri, A., Koren, N. Verh. Int. Ver. Limnol. 1993, 25, 290. (44) Foner, H. The Geological Survey of Israel, Jerusalem, unpublished data. (45) Central Bureau of Statistics. Traffic counting on non-urban roads 1986-1996; The State of Israel: 1997. (46) Stein, M.; Hofmann, A. W. Earth Planet. Sci. Lett. 1992, 114, 193. (47) Shemron, A. E. Geochemical exploration and new geological data along the SE flanks of the Hermon range; GSI 32; The Geological Survey of Israel: Jerusalem, 1989. (48) Rosenbaum, J.; Shemron, A. E. The Geological Survey of Israel, unpublished data. (49) Sherrell, R. M.; Boyle, E. A.; Hamelin, B. J. Geophys. Res. 1992, 97C, 11,257. (50) Faure, G. Principles of Isotope Geology, 2nd ed.; Wiley: New York, 1986; pp 1-589. (51) Michelson, H.; Mor, D. Gamla, Geological Map; The Geological Survey of Israel; Jerusalem, 1985. (52) Erel, Y.; Harlavan, Y.; Blum, J. D. Geochim. Cosmochim. Acta, 1994, 58, 5299. (53) Harlavan, Y.; Erel, Y.; Blum, J. D. Geochim. Cosmochim. Acta, 1998, 62, 33. (54) Bein, A.; Nielsen, H. J. Geol. Soc. London 1988, 145, 133. (55) Avnimelech, Y. Water Air Soil Pollut. 1980, 14, 451. (56) Erel, Y.; Morgan, J. J. Geochim. Cosmochim. Acta, 1992, 56, 4157.
Received for review May 31, 2000. Revised manuscript received October 23, 2000. Accepted October 26, 2000. ES0013172
VOL. 35, NO. 2, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
299
