Traçage isotopique

Traçage isotopique

Mercury (Hg) isotope composition was investigated in lichens over a territory of 900 km2 in the northeast of France over a period of nine years (2001-2009). The studied area was divided into four geographical areas: a rural area, a suburban area, an urban area, and an industrial area. In addition, lichens were sampled directly at the bottom of chimneys, within the industrial area. While mercury concentrations in lichens did not correlate with the sampling area, mercury isotope compositions revealed both mass dependent and mass independent fractionation globally characteristic of each geographical area. Odd isotope deficits measured in lichens were smallest in samples closeto industries, with∆199Hg of-0.15 ( 0.03 ‰, where Hg is thought to originate mainly from direct anthropogenic inputs. Samples from the rural area displayed the largest anomalies with ∆199Hg of -0.50 ( 0.03‰. Samples from the two other areas had intermediate ∆199Hg values. Mercury isotopic anomalies in lichens were interpreted to result from mixing between the atmospheric reservoir and direct anthropogenic sources. Furthermore, the combination of mass-dependent and mass independent fractionation was used to characterize the different geographical areas and discriminate the end-members (industrial, urban, and local/ regional atmospheric pool) involved in the mixing of mercury sources. Introduction Mercury (Hg) is a hazardous pollutant discovered and used since the Roman period (1). From the beginning of the industrial era, anthropogenic fluxes to the atmosphere increased systematically and now dominate the natural fluxes, according to estimates from global modeling (1). It is believed that Hg deposition to eco- and geosystems has increased 3-fold compared to the estimated natural background over this time period (2). Regulations in different countries led many industrial companies to produce goods with low Hg contents and to revise their removal systems and filters, but Hg emissions to the atmosphere from anthropogenic sources remain elevated (3, 4). Gaseous elemental mercury (GEM) is the dominant species in the atmosphere. Because of its long atmospheric residence time, GEM is considered a transboundary pollutant because it may be deposited far from its emission source (5, 6). For this reason, atmospheric depositions at the scale of a small territory will represent mixtures between longrange transported GEM and gaseous oxidizedmercury (GOM) from local sources with shorter lifetimes (5, 6). Deciphering the origin of pollution (global, regional, or local sources) or natural sources is crucial for developing policies for controlling mercury emission or deposition from human activities (2). However, tracing anthropogenic sources of heavy metals in a multiple source environment is currently a difficult task. Therefore, many such studies have been conducted around a single point source of pollution using multielement geochemistry (7, 8) or an isotopic approach. The latter mostly comprise Pb isotope tracing, but more recently other heavy metals such as Cd (9), Zn (10, 11), and Hg (12) were used as well. Identifying anthropogenic atmospheric sources using Pb isotopes measured in lichens, rain, and snow is possible because the compositions of potential sources are known (7, 13-16). Mercury possesses several physicochemical features that may be used to understand and investigate pollution issues. Mercury concentration combined with its speciation are used to document the presence and the potential toxicity of a contamination. A third dimension of investigation has recently become available with the precise measurement of Hg isotope ratios, which shows that mass-dependent fractionation (MDF) of Hg isotopes could be used to trace anthropogenic emissions (12, 17-19) as well as natural sources (20-26). A fourth aspect has been recently added with the discovery of mass-independent fractionation (MIF) of Hg isotopes during photochemical or non photochemical reduction (22, 27, 28) as well as physical (29) and chemical (30) processes. Such MIF anomalies may be used to identify fractionation processes or sources in natural environments (18, 31-36). Combining these different analytical levels and scientific concepts would yield a deeper view of the complex issue of the origin and fate of Hg pollutionin the environment. This study reports the mercury isotopic composition of atmospheric matter recorded in lichens, sampled over a territory of 900 km2 that contains multiple sources (anthropogenic and natural) and that was divided into four major geographical areas (rural, suburban, urban, and industrial). Sampling was performed over a period of nine years. The aim of the study is to show that MDF as well as MIF of Hg isotopes can discriminate multiple atmospheric sources and thus trace sources of environmental contamination.

Material and Methods

 This study presents the Hg isotopic composition of lichens located in the northeast of France and covering an area of around 900 km2 (≈400 000 inhabitants). The main urban area is the city of Metz with 130 000 inhabitants (42 km2 ). Flatlands border this city to the south, east, and west, whereas an important industrial valley starts to the north of the city and continues to the Luxembourg border. Main wind directions are SW and NE (16). The territory was divided into four major geographical areas according mainly to the presence or absence of industrial/human activities (see Supporting Information (SI) Figure SI-S1). The four areas  are (1) Rural Area (RA, four sampling sites), (2) Suburban Area (SA, nine sampling sites), (3) Urban Area (UA, five sampling sites), (4) Industrial Valley (IV, five sampling sites). Within the IV, two lichens were sampled at two different industrial sites (I, two sampling sites). Eight species of epiphytic lichens were used for this study and sampled over a period of nine years (2001, 2003, 2006, 2008, and 2009). The Hg isotopic compositions of some of the lichens presented in this study were previously presented in Carignan et al. (33) (see SI Table SI-S1). Additional descriptions of all of these areas, types of lichens and sampling procedures are given in the SI. Furthermore, particular attention was focused on a sampling site located within the suburban area (see SI Figure SI-S1). This point (CP-27) was identified as anomalous because ofits high Hg concentration throughout the sampling years, with Hg contents three to four times higher than the average reported in all other geographical areas (see SI Table SI-S1 and results and discussion). Specific sampling of five lichens around this point was conducted in order to delimit this specific area. Additional descriptions concerning this contaminated point and the lichens sampled around it are given in the SI. Lichens were freeze-dried according to the method described in Carignan and Gariepy (13) or dried at 105 °C and digested according to the method described in Estrade et al. 2010 (37). Hg concentrations weremeasured using either gas-chromatography coupled to ICP/MS or a solid Hg analyzer, as described in the SI. Isotopic analyses were performed using the MC-ICP-MS Nu Plasma HR (Nu Instruments) at the IPREM/LCABIE, Pau. 

Tracing a Mercury 

Point source. Lichen sample CP-27 (SI Figures SI-S1 and SI-S5 and Table SI-S1) displayed elevated Hg concentrations throughout the years (average [Hg] of ca. 400 ng · g-1 ). Furthermore, a distinct positive δ202Hg averaging +1.38‰ (2SE, n ) 4) was measured. These results contrast with all other Hg concentrations and isotopic compositions measured in the studied area and CP-27 was thus considered as a contaminated point. Samples from surrounding localities have Hg concentrations and isotopic compositions typical of the suburban area (SA/ACP 38-39-40-41). One sample (SA/ACP 42) located within 200 m of the contaminated point presented intermediate Hg concentration and isotopic composition ([Hg] ) 196 ng.g-1 and δ202Hg ) +0.46‰, see SI Table SI-S1 and Figure SI-S6). Details on these particular sampling sites are given in the SI. The relationship between the contaminated point (CP-27), the intermediate point (SA/ ACP 42) and the suburban average (mean of SA/ACP 38-39-40-41) is linear in the δ202Hg vs 1/[Hg] diagram presented in Figure 4, strongly suggesting physical mixing between two mercury sources having distinct isotopic compositions and concentrations. The Hg-rich point source would have a high δ202Hg g 1.5‰ (Figure 4). A surprising observation is that these three points (one corresponds to the average suburban composition measured on five lichens) have negative ∆199Hg anomalies of ca. -0.45‰ suggesting that MIF was already developed in the contaminant source/ emission (SI Figure SI-S7). The exact nature of such a contamination is not known for now but our results strongly support the effectiveness of Hgisotopes foridentifying various sources and types of emissions of atmospheric mercury.