Using Lead Concentrations and Stable Lead Isotope Ratios to Identify Contamination Events in Alluvial Soils

1 Géographie et Laboratoire de recherche en géomorphologie fluviale et sols, Université du Québec à Trois-Rivières, 3351, boul. des Forges, C.P. 500, Trois-Riviéres, QC, Canada G9A 5H7 2 Environnement et Développement durable, Ville de Trois-Rivières, C.P. 368, Trois-Rivières, QC, Canada G9A 5H3 3 Département de Chimie et Biologie, Sciences de l’Environnement, Université du Québec à Trois-Rivières, 3351, boul. des Forges, C.P. 500, Trois-Rivières, QC, Canada G9A 5H7 4 GEOTOP, Université du Québec à Montréal, 210 Ave. Président-Kennedy, PK-7150, Montréal, QC, Canada H2X 3Y7


Introduction
In the recent years, a growing number of environmental science studies on soil and sediment contamination were characterized by the use of lead stable isotopes to determine the source and origin of this element (geogenic or anthropic) and to evaluate its persistence in the environment [1][2][3].Mining and industrial activity have introduced various forms of lead pollutants into the environment and the intensive use of fossil fuels has also resulted in lead and several other heavy metals affecting the environment to varying degrees [4,5].In the past two decades, research has examined lead isotope signatures to trace emission sources and assessed spatial and temporal changes of recent lead pollution originating from lead smelters and manufacturing plants and from the use of alkyllead in petroleum products, particularly before 1990 [6,7].The atmosphere is recognized as major means of transport [8], but fluvial transport also constitutes a vector of pollution [9][10][11].Studies also examined lead contamination in soils, sediments and groundwater aquifers along with other metals that are generally geochemically associated (e.g., Cd, As and Zn) [12,13].Results showed that lead isotopes serve as indicators of environmental contamination events affecting a large spectrum of environmental media.
Other researchers have drawn on a combination of lead and lead isotope ratios ( 207 Pb/ 206 Pb, 208 Pb/ 206 Pb) as geochemical tracers to recognize the succession of atmospheric and pedogeochemical changes related to industrial activities in the mining area [11,14].Lead is widely considered immobile and persistent in soils, though its mobility in soil can increase with a high concentration of organic matter complex or chelate, which can play a key role in the movement of iron and other elements in the profile [14,15].Stable lead isotope ratios can be used as a complementary tool in dating contamination events and for the evaluation of sedimentation rates, including the use of lead-210 [3,16,17].Lead has four stable isotopes ( 204 Pb, 206 Pb, 207 Pb and 208 Pb) while the radioactive decay of 238 U, 235 U and 232 Th produces radiogenic lead isotopes (206, 207, and 208 Pb) [8,16].Although the various sources of lead have specific isotopic signatures, stable lead isotopes have been used generally to distinguish between lead originating from natural or anthropogenic sources [2,18].Lead isotope analysis has proved to be an effective technique for identifying the origin of lead in different terrestrial, marine and aquatic ecosystems [1,12,13].Also, the isotopic composition of lead is not affected to any measurable extent by physical or chemical processes [6].
This study focuses on a case of contamination along the riverbanks of the Saint-Franc ¸ois and Massawippi rivers (Québec, Canada) that was never officially reported by the government.This recent discovery led to the identification of soils contaminated by C 10 -C 50 hydrocarbons and other pollutants (e.g., heavy metals).The contamination extends more than over more than 100 kilometers of riverbank between the municipalities of Eustis and Drummondville (Southern Québec).The objective of this study is to attempt to identify the contamination events through the use of pedologic and radiogenic methods, including an analysis of heavy metals (e.g., Cd, Cu, Ni, Zn) and lead concentrations, stable lead isotope ratios ( 204 Pb/ 206 Pb, 207 Pb/ 206 Pb, 208 Pb/ 206 Pb) and sedimentation rates ( 210 Pb) in alluvial soils.This study focuses on the lead concentration and the pattern of lead isotopic composition found in floodplain soils.The use of lead in petroleum products (e.g., gasoline, heating oil) before 1990 and its prohibition in Canada after this date (the phasing out of leaded gasoline in Canada started in the 1980s and leaded gasoline was finally prohibited in 1990 [6,19].) provides a relatively good indicator of hydrocarbon contamination (presence of lead) in the overbank sediments, and helps estimate the date of the contamination events by using radiogenic data (lead isotopic signatures and lead-210).

Sites Location.
The sites under study are located along the Saint-Franc ¸ois and Massawippi rivers (Figure 1) and form part of the Saint-Franc ¸ois watershed basin, which covers a surface area of 10,230 km 2 .Ranging from an altitude of 304 to 762 m above sea level, it reaches its peak on the American side of the Appalachian Mountains (VT).In the upstream part of the basin, the relief is characterized by mountains, hills and valleys dominated by wooded areas and agricultural lands, while the downstream part of the basin is a flat landscape of farmland, small towns, and the city of Sherbrooke.The riverbanks are characterized by different distinct geomorphological sections comprised of glacial, glaciofluvial outwash or glaciolacustrine materials and fluvial deposits.The middle section of the Saint-Franc ¸ois and Massawippi rivers (sampling sites) is characterized by low floodplains (less than 3 m high) covered by overbank sediments composed mainly of silty and fine sandy loam materials and immature soils like regosolic and cumulic regosol soils [20].Subject to periodic flooding.These rivers may experience more than one flood per year [17].They go through former industrial and mining areas and agriculture lands that are major factors in the degradation of water quality.Traces of contamination (e.g., hydrocarbons, PAHs and heavy metals) have in fact been observed in alluvial soils along the riverbanks [17].In addition to the many industrial activities, mining has long been one of the main sources of pollutant discharge in area watercourses, primarily the Massawippi River.Several thousand tons of copper, nickel and tungsten were produced at the Eustis-Capelton mines.A study showed that these mines were responsible for copper concentration in the Saint-Franc ¸ois and Massawippi rivers [21,22].

Sampling and Laboratory
Analysis.Field investigations were undertaken on overbank sediments found at four sites (Figure 1).These sampling sites are located along riverbanks on recent alluvial soils (less than one thousand years).For each site, a trench was dug along the riverbank in order to reach the contaminated layers (the hydrocarbon layer), which in some cases were found at more than one metre in depth.In all, 90 soil samples were taken from the five profiles (MAS-13, STE-1, STE-1-2, STO-6 and RIC-9).A detailed morphological description of the profiles was used to characterize the texture, color horizon (Munsell chart) and depth of the hydrocarbon layer.In addition, soil horizons were identified in the field using the criteria of the Canadian System of Soil Classification [23,24].Soil samples were collected in the profile in 20-cm-deep sections until the contaminated hydrocarbon layer was reached.Soil samples were analyzed in the laboratory to characterize the textural composition, pH and total organic carbon (TOC).For the grain size analysis, the dry sandy fraction was obtained by sieving, while the finer fractions were obtained using a hydrometer (Boyoucos method).The methods used for the chemical analyses consisted of determining the pH in using a 1 : 2 soil-solution ratio (CaCl 2 :0.01 M) and the TOC content methods [25,26].
The hydrocarbon sample analysis was done using gas chromatography coupled with a flame ionization detector (GC-FID).The detection limit for the C 10 to C 50 hydrocarbons was 60 mg kg −1 of dry matter.For metal concentrations (Pb, As, Cd, Cu and Zn) and lead isotopic analysis, soil samples were mixed in a receptacle to ensure homogeneity, stored in sterile containes, and refrigerated at the site.Sampling methods used are in accordance with standards established by the Ministère de l'Environnement du Québec [27,28] and the Institut national de la recherche scientifique Centre -Eau Terre Environnement (INRS-ÉT É), Centre géoscientifique de Québec (CGQ), and BIOLAB laboratory  protocols.The samples were used to analyze heavy metals and lead concentrations as well as isotope ratios ( 204 Pb/ 206 Pb, 207 Pb/ 206 Pb, 208 Pb/ 206 Pb).Metal concentration levels were determined using an extraction procedure adopted by the Ministère de l'Environnement [28].At the INRS geochemistry laboratory (INRS-ETE/Delta Lab), the soil samples were dissolved using HNO 3, and the resulting liquid residue analyzed by inductively coupled plasma-mass spectrometers (ICP-MS) to obtain trace metal concentrations and isotopic compositions.Blank samples were prepared for every batch of 10 or 12 samples.For Pb concentration analysis, the complete procedures, including the duplicate analyses required to validate the laboratory tests, are available in government documents [28].The quality control and acceptability criteria (QA/QC) had to conform to document DR-12-SCA-01 from the Ministère de l'Environnement [28] and applied as follows: (i) the curve determination coefficient should be at least 0.995; and (ii) the duplicate analytical method must be under the methodological quantification limit (MQL).The randomly tested check standards should not exceed 20% of the target value, except for the level 1 standard, which is used to ensure adequate sensitivity.[3,4,29] given that excess 210 Pb activity diminishes theoretically in an exponential manner, thus determining the sedimentation rate [29].There are two types of 210 Pb: the first originating directly from the atmosphere (supported or unsupported excess 210 Pb) and the second produced in situ as a result of radium disintegration ( 226 Ra), originating from 238-Uranium in the analyzed sediments (supported lead).It is assumed that the flow of the lead originating from the atmosphere is constant from year to year, and it becomes possible to estimate the sedimentation rates over relatively long periods [29].The work by Nriagu [30] in fact shows greater use of lead after 1800, which constitutes an additional indicator in the interpretation of anthropogenic sources of pollution.
The method used for 210 Pb age dating is alpha spectroscopy, a technique that offers greater flexibility compared to beta spectroscopy and more precision compared to gamma ray spectroscopy.The alpha spectroscopy method is used by the GEOTOP-UQAM-McGILL laboratory where the analyses were performed.A more complete description of the methods was provided elsewhere [17].

Alluvial Soils and Hydrocarbon-Contaminated Layers.
The alluvial soils analyzed have been classified in the Cumulic Regosol (CU.R) subgroup of the Canadian System of Soil Classification [24].The soils show little development and are characterized by the absence of Ah horizons and accumulation horizons (B), as well as by little chemical alteration.The soils that were analyzed generally show pH values ranging from 3.45 to 7.09 with an average value of 4.63 and TOC rates ranging from 0.12 to 1.48% (Tables 1 and 2).The pH values in the soil (MAS-13) of the mining area (Eustis-Capelton) were more acidic at the bottom of the profile (120 cm and more), probably due to the presence of acidified toxic waste [21] found at the base of the soil profile, which come from the former Eustis mining site [22] located a few kilometres upstream.Most of the alluvial soils examined are made up of fine material, mainly fine sandy loam, fine loamy sand or sandy loam (Tables 1 and 2), which is a common feature of flood or fluvial deposits in these rivers [17,20].It could be said that this textural uniformity (basically fine material) is the dominant characteristic of the soils that form the floodplains of the rivers under study.It should be noted that the fine-textured soils retain more contaminants than the coarse matrix [31,32].
The field observations and the samples taken from the riverbanks reveal that the most highly contaminated soils extend from Eustis to Windsor (Figure 2).The banks of the Massawippi River contain contaminants that can be traced back to the former Eustis mine, while no contamination was observed in the soil materials upstream of the site.From Eustis to Richmond, however, both riverbanks are contaminated with hydrocarbons (C 10 -C 50 ) and heavy metals (see also [33]).Considering the thickness of the contaminated soil, the most problematic sites are those in the Massawippi area.The contaminated layers may extend over more than one metre in depth in the riverbank deposits and the coloration and odor are more pronounced.Based on these observations, the origin of the discharge would appear to be Eustis.The former mining site is still being used as a storage area for pulp and paper waste and served as a landfill for mining waste containing copper, lead and zinc [21].The contaminated soil sample taken at the Eustis mine site (EUS) showed a relatively high C 10 -C 50 hydrocarbon contamination rate, that is, 380 mg kg −1 of dry matter.At the Massawippi site (MAS-13), the rate is <60 mg kg −1 , and for the Windsor (STE-1, STE-1-2) and Richmond (STO-6) sites the values varied from 660 to 80 mg/kg (Table 1).The deepest contaminated hydrocarbon layers (C 10 -C 50 ) were found at the Massawippi (MAS-13 profile) and Windsor a : FSL: Fine sandy loam; FLS: Fine loamy sand; SL: Sandy loam; MFS: Medium fine sand (see [24]).
sites (STE-1 profile), with depths of 200-220 cm and 160-180 cm, respectively.In some other soil profiles found in the Massawippi and Saint-Franc ¸ois rivers [33], the hydrocarbon layers can be found at the surface (5-10 cm) of the profiles and a depths from 60 to 90 cm.This points to either a remobilization of contaminated layers during flooding or local and site-specific spills.Lastly, profile RIC-9, found more than 30 metres from the riverbank, does not show any trace of hydrocarbons, unlike the other profiles.
In reviewing the information recorded in the archives of the Société d'Histoire de Sherbrooke (e.g., regional newspapers), the pollution events seem to have occurred on two separate occasions in the Saint-Franc ¸ois River, that is, October 25, 1955 andMay 17, 1963, though no details on the events or their causes are provided.The 1963 newspaper article mentions the presence of large expanses of oil on the river in the Sherbrooke area (Figure 1).There were no newspaper articles on cases of river contamination (e.g., oil discharge) for the Massawippi River.However, it is probable that other contamination events (e.g., local contamination) occurred but were not all mentioned in the documents that were consulted.However, the presence of over more than 100 km of riverbank contaminated by hydrocarbons (C 10 -C 50 ) points to one or more major spills in the river.

Heavy Metals Concentrations and Lead Isotope Ratios in Soil
Profiles.The results obtained for lead and other metal concentrations of soil profiles for the selected sites (Richmond, Windsor and Massawippi) are presented in Figure 3, and Tables 2 and 3.The concentration levels of the metal elements analyzed must comply with the Environment Ministry's standards for contaminated sites [27,28] developed for criteria A, B and C, that is, 6, 30 and 50 mg kg −1 for arsenic (As), 1.5, 5 and 20 mg kg −1 for cadmium (Cd), 40, 100 and 500 mg kg −1 for copper (Cu), 50, 500 and 1500 mg kg −1 for lead (Pb) and 110, 500 and 1500 mg kg −1 for zinc (Zn), respectively.In general, Pb concentrations  and other metal or metalloid elements (e.g., As, Cd) are relatively low, except for the MAS-13 profile, which exceeds the concentration of criterion A for arsenic (37.9 mg kg −1 ) and lead (121.8 and 149.1 mg kg −1 ).This profile shows distinctive peaks (160 and 220 cm depth) of contamination (Figure 3).The As and Pb concentrations are more elevated deeper in the profile than on the surface, which is usually the case in contaminated soils where the sources of lead and other elements are inputs by atmospheric pollutants [10,16].As the highest concentrations are found at the bottom of the soil profile (more than one metre), it is likely that contamination (hydrocarbon layers) from pollutants was carried off by the river at low levels.In fact, virtually all the soil profiles observed in the Massawippi area contained hydrocarbon layers at the bottom of the profiles.Samples of the contaminated layers can only be taken when the Massawippi River is at low levels, that is, around 3.76 and 6.61 m 3 /s, most often during September and early October [34].Soil profile STE-1 at 180 cm containing lead showed concentration levels that slightly exceeded criterion A with a value of 50.82 mg kg −1 (Figure 3).This sample was taken directly from the contaminated hydrocarbon layer (160-180 cm depth).In the same profile (STE-1), with respect to other metal element concentration levels (e.g., Cu and Zn), some samples had values that exceeded criterion A (Table 3), except for the upper horizons (0-40 cm).These results can serve as indicators of the contamination events (oil discharges most likely) that occurred probably at different times along the Saint-Franc ¸ois River.For profile STO-6, the concentration of heavy metal and metalloid elements analyzed (e.g., As, Cd, and Pb) was higher in the subsurface horizons (40-60 cm depth), which suggests contaminants leaching to the lower layers or resulting from the infiltration of pollutants into the riverbanks by polluted waters.For the RIC-9 profile, the concentrations of As, Cd and Pb are comparable with the values obtained for the STO-6 profile and show a similar pattern for Cd and Pb (Figure 3).Maximum concentrations of lead, for example, are reached at a depth of 40 cm (34.4 mg kg −1 ), a similar value to that obtained for the STO-6 profile (35.7 mg kg −1 ) at the same depth (40 cm).Note that profile RIC-9 is situated 31 m away from the riverbanks and no trace of soil contaminated with hydrocarbons was detected in the profile.It is possible, however, that other pollutants (such as Cd and Pb) carried off by the river may have possibly reached this site (RIC-9) during major floods [35], or that these contaminants found in subsurface layers (20-40 cm) come from atmospheric fallout.It is known that the fine particulate matter of lead and of other pollutants can be transported over long distances [2,3] and be redeposited on the soil surface.Over time, these pollutants penetrate the lower horizons through migration or leaching [12,36]; cadmium and zinc are usually more easily remobilized than lead, however [15].Still, the concentrations of trace metals (As, Cd, Pb) detected on the surface of the profiles (RIC-9 and STO-6) are rather low.
The values of all the profiles (Figure 3) show many peaks that indicate a concentration level of lead and other elements (As, Cd) found at depths of 40, 60, and 80 cm and two peaks with stronger concentration levels of Cd at depths of 100 and 180 cm in the STE-1 profile, a relative high concentration of As at a depth of 80 cm in profile STO-6, and very high concentrations of As, Cd and Pb at a depth of 220 cm in profile MAS-13.These peaks in the soil profiles suggest several separate contamination events that might be associated with various pollutant discharges into the river.The peaks found at the bottom of the soil profiles (100, 180 or 220 cm) can only really be associated with the transport of pollutants by the river due to their position in the profile, while the pollutants found at lower depths (40 cm) most likely also come from the transport of pollutants by the river [32], but contamination through atmospheric fallout (e.g., anthropogenic deposition in the soil surface) cannot be ruled out.These two likely sources (river and atmosphere) could partly account for the variability of the isotopic values found in the different profiles (Figure 4), though such variability could be attributable to the type of products discharged into the river or transported by it (e.g., mixing of lead with the hydrocarbon layer or lead from mine waste).
the 208 Pb/ 206 Pb ratios obtained for the STE-1 and MAS-13 soil profiles ranged from 2.089 to 2.111 and 2.086 to 2.105, respectively.The highest variations were recorded in the STO-6 profile, but there were also significant differences in the MAS-13 and STE-1 profiles.It is interesting to note that the differences between the isotope ratios are more pronounced at the bottom of the profiles.

Trends of Lead versus Isotope Ratios and Contamination
Events.The concentrations of lead in two soil profiles (STE-1 and MAS-13) are higher at the bottom of the profiles with peaks at 160, 180 and 220 cm (Figure 3).The concentrations of As at different depths in profiles STE-1 and MAS-13 also exhibit a similar pattern with maximum peaks at 160, 180 and 220 cm and another peak for the Cd concentration at 220 cm for profile MAS-13.For profile STO-6, the two maximum peaks for the As and Cd concentrations are reached at depths of 80 and 40 cm, and similar depths for the Pb concentration.By examining only lead in profiles STE-1, STO-6 and MAS-13, it was found that the concentration of this element gradually increases according to the depth.This was also the case for the concentration of As, except for the STO-6 profile.The similarity of soil profile curves suggest at Applied and Environmental Soil Science least two to three major contamination events that might be associated with spills or releases of pollutants into the waters of the Massawippi and Saint-Franc ¸ois rivers.Moreover, lead concentration levels are in fact highest in the hydrocarbon layers, that is, 50.82 mg kg −1 (180 cm) for the STE-1 profile and 149.13 mg kg −1 (220 cm) for the MAS-13 profile.The peaks recorded on the bottom of the profiles (160, 180, 220 cm) could be explained by oil spills or other petroleumbased products spills that occurred when water levels were at their lowest.It is very unlikely that the lead found at the bottom of the profiles is due to the leaching of this element to this depth.Lead fractions are not very mobile in soils and tend to be concentrated in surface horizons [36,37].Furthermore, the highest concentrations of lead analyzed in the profiles are associated with the layer of hydrocarbons most often found at the bottom of the profiles, especially in the Massawippi area where lead levels are the highest.Lastly, note that hydrocarbon compounds (e.g., gasoline, motor or fuel oils) are water immiscible and their migration into the groundwater is scant [38,39].It is nonetheless difficult to confirm the precise origin of the hydrocarbon (C 10 -C 50 ) contamination observed in the soil profiles.Furthermore, no study on contaminated soils or on the level of chemicals and isotopes in the natural soil in this area has been done.The only study that can be used for comparison comes from the Ministère des Ressources naturelles [40], which provides the concentration of various heavy metals obtained from the soil and sediment samples for sector A4 of the Appalachian geological region where our study area is located.Based on this study, the values obtained for the average concentration of As, Cu, Pb and Zn in the soil and sediments (natural background values) are 11.1 mg kg −1 (SD 10.1), 11.5 mg kg −1 (SD 66.5), 15.3 mg kg −1 (SD 17.5) and 67.3 mg kg −1 (SD 82.3), respectively.If these values are compared to those obtained in our profiles, the concentration exceeds the background rates by factors of 1.5 to 10 for the MAS-13 profile (excluding As), and is two to three times higher for the three other profiles (STE-1, STO-6, RIC-9), excluding arsenic.Records show relatively high concentrations of copper (44 mg kg −1 ) and Pb and Zn (50.82 and 144 mg kg −1 ) in the STE-1 profile (Table 3) at the base of the profile (80-100 and 160-180 cm) that may have originated from the Eustis mine.Part of these elements that may originate from former mining sites might have been carried downstream to the Windsor (STE-1) and Richmond (STO-6) sites.It is equally possible that the metallic content in the atmospheric dust produced by the mine accumulated on the subsurface horizons, though the Cu values are not overly high (between 29 and 34 mg kg −1 ) at the STE-1 profile, for example (see [33]).The ore extracted from the Eustis site consisted of copper and pyrite (chalcopyrite mineral), and about 39,000 tons of ore (copper mainly) were extracted annually from 1865 to 1939 [41].
Stable isotope ratios ( 207 Pb/ 206 Pb and 208 Pb/ 206 Pb) obtained for lead in the MAS-13 profile indicate a close similarity with the lead concentration profile (Figure 4).The correlation between 207 Pb/ 206 Pb and 208 Pb/ 206 Pb lead ratios in the same profile is exetremely high (r 2 = 0.96, P < .05,n = 12), and high between lead concentrations and 207 Pb/ 206 Pb and 208 Pb/ 206 Pb ratios (r 2 = 0.66 and r 2 = 0.78).The various peaks (at 160, 200, 220 cm) in the MAS-13 profile suggest a series of contamination events that include at least one that can be related to an oil spill (hydrocarbons layer at 200-220 cm).The isotope values at the base of the profile are different from those obtained on the subsurface horizons (20-40 cm).The values obtained from the soil sample contaminated with hydrocarbons for the STE-1 profile (180 cm in depth) indicate isotopic values ranging from 0.869 ( 207 Pb/ 206 Pb) to 2.109 ( 208 Pb/ 206 Pb), respectively.These represent values significantly different from the ones estimated for the surface profile (0-20 cm): 0.855 ( 207 Pb/ 206 Pb) and 2.089 ( 208 Pb/ 206 Pb).Also, in profile STE-1, concentration levels of lead are reached at a depth of 180 cm, with a value of 50.82 mg kg −1 (Figure 4).By comparing the isotope ratios obtained for the MAS-13 (0.869 for 207 Pb/ 206 Pb) and STE-1 profiles (0.855 for 207 Pb/ 206 Pb) are slightly different for each profile.This is an indication that the source of pollution may be different, either associated with petroleum product spills (e.g., gasoline, fuel oils) or other pollutant discharge.Considering the position of the peaks (160, 180, 200 and 220 cm) of the lead isotope ratios recorded in the soil profiles sampled from the Massawippi (MAS-13) and Saint-Franc ¸ois riverbanks (STE-1), it is possible, however, that at least two separate contamination events took place at different time intervals.For the STO-6 profile, there are slightly higher lead concentration levels at depths of 40 and 60 cm and where the associated isotope values are 0.865-0.868( 207 Pb/ 206 Pb) and 2.097-2.102( 208 Pb/ 206 Pb), respectively, which are values that are relatively different from those observed for the other profiles (MAS-13, STE-1).
In order to estimate the relative period of the deposits found in the alluvial soils of the STE-1 profile, the 210lead method (CRS Model-Constant Rate of Supply) was used to get a chronological series of sediment deposits and to assess the average annual rates of sedimentation for the site.Figure 5 presents the data that compare the different concentrations of metallic trace elements (e.g., As, Cu, Zn) for the STE-1 profile (Figure 5(a)), the concentration levels of the lead and isotope ratios (Figure 5(b)), in addition to the rates of sedimentation and calendar dates collected by the CRS Model (Figure 5(c)).The strongest concentrations of trace elements are recorded at the base of the profile (160-180 cm), precisely in the layer with the highest hydrocarbon contamination.The evidence shows that the hydrocarbon layer has permeated the fluvial deposits dated before or during 1700-1800 (in pre-industrial times) (Figure 5(c)).Remember that mining operations began in the region in the 1850's with the opening of copper mines in the Ascot Township, including the Ascot (1851-1864), Belvedere (1864-1865) and Eustis-Capelton mines (1863-1865) [41].It was only later in the 1900s and 1910s that extensive industrial activities such as wood processing, pulp and paper, heavy industries, chemical procedures took place although as early as 1850 various industries came to the region (textile and pulp and paper) [42].
Since contaminants were found at the base of the profiles, in particular in the alluvium which was deposited prior to the pre-industrial era, it can be shown that the rivers had reached very low levels that correspond perhaps to a prolonged low flow period (a very dry summer-fall season, for example).Finally, a second contamination peak is noted (As, Cu, Pb, Zn) at a depth of 100 cm in the profile that could be attributed to another contamination event associated with the discharge of pollutants in the river that may have infiltrated more recent deposits (between 1800 and 1850).In this respect, it is interesting to note in Figure 5(c) how the sedimentation rates have sharply increased during recent years, in particular in the 1970's and 1980's.The mean annual sedimentation rate is evaluated at 0.79 cm/y (1973)(1974)(1975)(1976)(1977)(1978)(1979)(1980)(1981)(1982)(1983)(1984)(1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992).This increase in sedimentation rates actually corresponds to an increase in flood events during the same period [17,43].

Conclusions
Focusing on the contamination levels of the alluvial soils of the Saint-Franc ¸ois and Massawippi rivers (in southern Québec), the results of this study showed that pollutants (hydrocarbons and trace elements As, Cd, Pb and Zn) carried by the rivers can travel very long distances (over 100 km) and that the infiltration of these pollutants into the riverbanks is conditioned by factors such as hydrodynamic conditions (low levels of water, floods, etc.).In addition, the fluvial transport of pollutants turned out to be a relatively complex phenomenon because of the effect of recurring floods on the stirring up of fine sediments.By comparing the values obtained between soil profiles, the maximum and minimum values detected for arsenic, cadmium and lead vary from 3.01 to 37.88 mg kg −1 (As), 0.11 to 0.81 mg kg −1 (Cd), and 12.32 to 149.13 mg kg −1 (Pb), respectively, while the 207 Pb/ 206 Pb isotopic ratio values vary from 0.8545 to 0.8724 for all the profiles.The highest values of the trace elements (As, Pb and Zn) were detected in the hydrocarbon layer (C 10 -C 50 ), most often located at the bottom of the profiles (160, 200 and 220 cm in depth).The various peaks recorded in the soils and their position in the profiles suggest that various contaminants were transported by the river on several occasions and infiltrated the soil matrix or were deposited on floodplains during successive floods.Atmospheric particles which enter the river or deposited on riverbanks must also be considered as another source of pollution recorded in alluvial soils.In this study, it was found that alluvial soils are the efficient "environmental archives" that allow past contamination events to be traced and, as a result, can serve as an indication of the way contaminants can persist in the environment for a long period of time.
Attention is drawn to the importance of combining analytical methods of contaminant concentrations (including lead), which include isotope ratios of lead methods.Using only the concentration of trace metals in alluvial soils does not seem sufficient for detecting different levels of contamination in riverbank profiles.Dating methods (e.g., lead-210) used in combination with contaminant analysis can also turn out to be an additional analytical tool for interpreting data.Finally, the study shows that sediments found on the surface (0-20 cm) of alluvial soils are less contaminated than at deeper layers, pointing to evidence that water quality has in fact improved over time, owing to better management of pollutant discharge by municipal and government authorities over the last 10 to 20 years.
The next phases of this study will increase the number of soil samples in areas outside of floodplain zones (e.g., forested, upland, with little or no disruption by human activities) to discriminate between pollutant inputs associated with fluvial transport or atmospheric deposition.
Also, the use of comparative methods between Pb-Sc and Pb-Zr (Sc and Zr are conservative metals [44] with no significant anthropogenic source) will be evaluated.These will permit to better discriminate between natural background Pb and anthropogenic Pb.
Hydrocarbons layer at the bottom of soil profile (Windsor area) C 10 -C 50 : 470 mg kg −1

Figure 2 :
Figure 2: Soil profile with the hydrocarbon layer (C 10 -C 50 ) at a depth of 160-180 cm.The contaminated layer can be more than 30 cm thick in the riverbank in the Windsor area (left bank).

Figure 3 :
Figure 3: As, Cd and Pb concentrations in selected soil profiles (STE-1, STO-6, RIC-9 and MAS-13).The dotted line delimits classes A and B of soil contamination based on the MDDEP criterion [19].

Table 1 :
Description of sampling sites and some characteristics of soil profiles.
2.3.Age Dating and SedimentationRates with Pb.The isotopic 210 Pb dating method was used for soil profile at STE 1 (Windsor site) to determine sedimentation rates based on radionuclide activity in sediments.For this, the lead 210 CRS model (Constant Rate of Supply) was used to a close correlation between depth and time

Table 3 :
Lead and heavy metal concentrations and Pb isotopic compositions in soil profile at Windsor site (STE-1).