Research Electron Microprobe and Synchrotron X-ray Fluorescence Mapping of the Heterogeneous Distribution of Copper in High-Copper Vineyard Soils A S T R I D R . J A C O B S O N , * ,† SYLVIE DOUSSET,‡ FRANCIS ANDREUX,§ AND P H I L I P P E C . B A V E Y E †,| Laboratory of Geoenvironmental Science and Engineering, Bradfield Hall, Cornell University, Ithaca, New York 14853-1901, Centre des Sciences de la Terre et de l’Environnement, Universite´ de Bourgogne, 6 Boulevard Gabriel, 21000 Dijon, France, and SIMBIOS Centre, University of Abertay Dundee, Bell Street, Dundee DD1 1HG, UK
The response of microorganisms to metal contamination of soils varies significantly from one investigation to another. One explanation is that metals are heterogeneously distributed at spatial scales relevant to microbes and that microoorganisms are able to avoid zones of intense contamination. This article aims to assess the microscale distribution of Cu in a vineyard soil. The spatial distribution of Cu was measured at two resolutions (0.3 mm and 20 µm) in thin sections of the surface 4 cm of undisturbed soil by electron microprobe and synchrotron X-ray microfluorescence spectroscopy. Bulk physicochemical analyses of Cu, pH, organic matter, texture, and mineralogy were performed. The results indicate that the Cu distribution is strongly heterogeneous at both scales of observation. Entire regions of the thin sections are virtually devoid of Cu, whereas highly localized “hotspots” have Cu signal intensities thousands of times higher than background. The distribution of Rb, or Al and Si, indicators of clay minerals, or Fe (iron (hydr)oxides), show that Cu is not preferentially associated with these mineral phases. Instead, Cu hotspots are associated with particulate organic matter. These observations suggest modification of current sampling protocols, and design of ecotoxicological experiments involving microorganisms, for contaminated soils.
Introduction Heavy metals are toxic to all organisms, if present in high concentrations. Microorganisms are no exception, and exposure to heavy metals has been well documented, for * Corresponding author present address: Utah State University, Plants, Soils and Climate Dept., tel: (435) 797-2233; fax: (435) 7973376; e-mail:
[email protected]. † Cornell University. ‡ UMR 5561 Bioge ´ osciences, CSTE, Universite´ de Bourgogne. § UMR 1229 Microbiologie des Sols et Environnement, CSTE, Universite´ de Bourgogne. | University of Abertay Dundee. 10.1021/es070707m CCC: $37.00 Published on Web 08/17/2007
2007 American Chemical Society
over a century, to negatively affect the growth, metabolism, and survival of microbial cultures (1). In soils, research carried out in the last two decades on the effect of metal contamination on microbial activity, as well as community or population composition, has identified similarly deleterious effects in a number of cases (e.g., 2-5). Yet, experimental observations are not always clear. Large variations have been reported in the soil heavy metal concentrations constituting a threshold for the selection of tolerant microorganisms. For copper (Cu), this threshold ranges from as low as 5 mg kg-1 to upward of 1000 mg kg-1 (2, 6, 7). Microorganisms with a high sensitivity to heavy metals are routinely found in highly contaminated soils. In a number of short-term field experiments, addition of significant amounts of heavy metals resulted in no noticeable effect on microbial biomass (e.g., 6, 8), or in increases, rather than decreases, in microbial metabolism (1). Some of these contradictory observations are likely due to the fact that the soils used by various investigators have different chemical and physical properties, which affect active metal concentrations (1). Undoubtedly, methodological differences among studies also explain some of the conflicting picture. Nevertheless, in spite of efforts to account for specific soil properties and to standardize experimental protocols, the exact influence of heavy metals on soil microorganisms remains somewhat unclear, quantitatively as well as qualitatively. Yamamoto et al. (9) suggested that a possible explanation in that respect is the heterogeneous spatial distribution of metal contaminants in subsurface environments, with highly contaminated “hotspots” neighboring regions in the soils that are essentially devoid of metal contamination. By analogy with other organisms, for example earthworms (10, 11), it is conceivable that microorganisms, depending on their metal sensitivity, would either avoid heavy metal hotspots entirely or populate them preferentially. Recent experimental evidence of highly heterogeneous microbial distribution in soils lends credence to Yamamoto et al.’s (9) perspective. Franklin and Mills (12) used nested soil samples, collected at separation distances ranging from 2.5 cm to 11 m, to show that there is significant multiscale variation in the spatial distribution of microbial communities in soils. They attributed this pattern to the variable response of individual populations to the spatial heterogeneity of soil physical and chemical properties. More recently, Becker et al. (5) sampled a soil chronically contaminated with Cr and Pb, using a nested scheme with sampling points from less than 1 cm up to 18.9 m apart. They found significant heterogeneity in the spatial distribution of microorganisms. Of 68 different phylotypes found in the soil, 51% were present in less than 10% of the samples, and 82% in less than 20% of them. Microbial activity and metal content were not significantly correlated, which, as Becker et al. (5) suggest, could be because the metals were not bioavailable or because there was selection for resistant microbial communities in microsites with high metal concentrations. Another plausible explanation is that even the centimetric scale of their homogenized soil samples was too large to discern spatial heterogeneity in metal distribution susceptible to have an effect on microbial colonization. This possibility is supported by the work of Nunan et al. (13, 14), who in intact soil thin sections observed significant clustering of bacterial cells, and large differences in bacterial population density, over micrometer distances. Consequently, Nunan et al. (14) advocated that attempts to determine which chemical or physical VOL. 41, NO. 18, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6343
factors have an impact on bacterial communities should be carried out at a number of scales, extending upward from the micrometer scale into the centimetric and metric scales typical of past research on the distribution of metals in soils. Until just a few years ago, no available experimental technique allowed the mapping of elemental distribution in geological samples or soil thin sections at micron-scale resolution over even millimetric, let alone centrimetric, areas. Fortunately, the advent of microfocused synchrotron-based X-ray fluorescence spectroscopy (µ-SXRF) has since made this type of measurement feasible. Much of the research published to date using µ-SXRF, however, has focused on the distribution of specific elements (e.g., Fe, Zn, Pb, Se, or As) in micron-sized nodules or aggregates (15-17), or in a few isolated regions or transects, at most covering fractions of square millimeters in total area, in soil thin sections (e.g., 18-20). Only Jones and Doner (21), using an X-ray fluorescence microprobe, mapped the concentration of metals (Cr, Pb, and Ti) in areas larger than the square millimeter (0.05 cm × 2.9 cm). Their maps show significant heterogeneity in the spatial distribution of metals, with repeated patterns of 50- to 100-fold differences in metal concentrations occurring over distances smaller than a millimeter. However, the resolution of these maps, with a pixel size of 0.18 mm, is too coarse to be directly relevant to the pore-scale ecology of microorganisms. In this general context, the primary objective of the research described in this article was to assess whether microfocused synchrotron-based X-ray fluorescence spectroscopy, in conjunction with electron scanning microscopy and microprobe analysis, could provide information on the heterogeneity of the distribution of heavy metals in contaminated soils in a range of measurement scales from micrometric to centimetric. A second objective was to determine the extent to which these same techniques could shed light on the nature of the soil constituents with which metals seem to associate at spots where metal concentration is high. Vineyard soils with high levels of copper were selected for the study, in part because Cu is a common component of a variety of pollutants introduced into soils, such as wood preservatives, inorganic fungicides, or sewage sludge, and because a very significant body of literature has been devoted to the effects of Cu on soil microorganisms and on plants.
Materials and Methods Site Selection. A previous study (22) involved a number of vineyard soils from Odenas (Bj), in Beaujolais, and VosneRomane´e (VR), in Burgundy. Vineyards in these two regions have been in cultivation for over two millennia. Starting in 1885, “Bordeaux mixture”, a mix of copper sulfate, hydrated lime (calcium hydroxide), and water, has been extensively used as a fungicide, at application rates that until recently were not regulated and lead to significant copper accumulation in surface soil horizons. Of the soils investigated by Jacobson et al. (22), the two soils with the highest average copper content in the 0-20 cm layer were selected for the present research. These are Bj-150, an acidic soil from Odenas (Beaujolais) with a Cu concentration of 151 mg kg-1, and VR-450, a calcareous soil from Vosne-Romane´e (Burgundy) with a bulk soil total Cu concentration of 417 mg kg-1. The soils were physically and chemically characterized for texture, pH, carbonates, total organic carbon (TOC), total Cu, and free Cu2+ activity (Table S1) as described in the Supporting Information (SI). Results for and all further references to the granitic soil from Odenas can be found in SI. These analyses, as well as the earlier data obtained by Jacobson et al. (22) for the 0-20 cm surface layer (Table S1, SI) indicate that the soil from Vosne-Romane´e (VR-450) is calcareous with a total carbonate content around 22% and 6344
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 18, 2007
a pH around 8.3. It is gravelly (33%), but the 9.0. Analysis of the soil thin sections by polarizing light microscopy shows that the VR soil is rich in carbonate minerals (calcite, aragonite, dolomite, micrite) and phyllosilicates (glauconite, cerolite, serocite, mica, biotite) but does not contain Cu-rich minerals that might interfere with the interpretation of spectrometric results. Soil Collection. An intact soil sample was collected by pressing an aluminum box without its end caps (7 cm (l) × 8 cm (w) × 4 cm (h)) into the surface of the soil. When the top edges of the box were level with the soil surface, a shovel was used to lift the sample and surrounding soil from the ground. The excess soil was trimmed from the outside edges of the box, and the top cap was replaced. Soil was then sliced level with the bottom of the box and covered with the bottom cap. The box was labeled to indicate the soil surface. Preparation of Thin Sections. The structured soil sample was air-dried at 20 °C, without removing it from the box, until it reached a constant weight, and was then impregnated under vacuum with a polyester resin mixed with styrene and acetone. The resin was allowed to harden at room temperature for approximately 5 weeks. Four 1-mm thick and four 30-µm thick vertical thin sections (20 mm × 40 mm) were cut from each resinimpregnated soil block. The 1-mm thick thin sections were necessary to have sufficient sample volume to generate detectable element signals with the electron microprobe. The 30-µm thick thin sections were used for the µ-SXRF analyses to limit the sampling depth, which depends on the escape depth of the fluorescence X-rays of interest (typically 10-80 µm for 3d transition metals) (23), to visible soil components (e.g., sand grains, particulate organic matter, etc.). The thin sections were mounted on glass slides and the surfaces polished smooth with silicon carbide (7 µm) followed by cerium oxide (3 µm). Each slide was scanned directly on a flatbed scanner prior to any other analysis. Microfocused Synchrotron-Based X-ray Fluorescence (µ-SXRF). Elements in the 30-µm thick soil thin section were mapped by µ-SXRF on the F3 beam line of the Cornell High Energy Synchrotron Source (CHESS). The nominal beam energy was 18 KeV, which is sensitive to K lines of elements 20 (calcium) to 38 (strontium). These elements include Cu and the major elements we were interested in mapping in the soils (Ca, Ti, Mn, Fe, Zn, and Rb). Dwell times were generally 10 to 20 s per spot. Due to cyclic variations in incoming beam intensity, dwell times were adjusted to maintain a constant incident flux. The output count rate was between 500 000 cts and 1.7 × 106 cts at each measurement location. Two beam sizes were used. First, the entire thin section was mapped using a collimated 0.3-mm beam. Cu hotspots identified on the thin sections were then reanalyzed using a capillary focused beam with a 50-mm focal distance and full width half max (FWHM) spot size of 20-µm. Analysis of the X-ray spectra was performed with the in-house CHESS software package MCASPEC and program “sag2.m”, which runs under MATLAB, both written by Rong Huang (Pers. Communic.). Elemental concentrations could not be calculated with this software; however, relative photon intensities (counts per second) could be mapped. For a given element, relative intensities are proportional to concentrations. The proportionality constants vary from element to element and depend on the conditions under which measurements are performed. A color scale was applied to the maps ranging from deep blue (threshold count numbers measured for the resin or the resin-filled pore spaces) to
FIGURE 1. Scanned image of 30-µm thick soil thin section of the calcareous clay soil VR-450 from Burgundy (left), µ-SXRF map of Cu obtained with a 0.3-mm beam (center), and detailed maps obtained with a capillary-focused 20-µm beam. bright red for the highest count numbers for each element on a slide. The colors were automatically calculated and applied by the program “sag2.m”. Electron Microprobe Analyses. The surface of each 1 mmthick soil thin section was cleaned by sonication in ethanol and then coated with a layer of carbon (approximately 250 Å thick), using an Edwards Auto 306 vacuum coater. Elemental analyses were performed on a JEOL JXA-8900R EPMA Microprobe Microanalyzer outfitted with five automated wavelength dispersive X-ray spectrometers (WDS) (JEOL USA, Inc., Peabody, MA). Besides Cu, the microprobe allows mapping of a number of light elements (C, O, Al, Si, and S) that were not visible with the µ-SXRF setup. Under the conditions used, this Microprobe Microanalyzer produces pictures with a physical pixel size of 3 µm. The penetration depth of the beam is typically 4-5 µm, up to 10 µm in areas of high concentration in light elements. The microprobe was too slow to map an entire soil thin section but could be used effectively to map smaller subregions, e.g., about 1 mm2. Scanning Electron Microscopy (SEM) Images. The 1 mmthick thin section samples were mounted on aluminum stubs using double-sided carbon tape and then coated with four layers of carbon applied with a carbon thread sputter coater. Sample images were prepared using a Hitachi 4500 field scanning electron microscope at an acceleration voltage of 24 KeV.
Results and Discussion Cu Distribution Maps. In the 30-µm thick thin section of the VR-450 soil, µ-SXRF observations with the 0.3-mm diameter beam (Figure 1, center) show large, bright blue to yellowishgreen patches in the Cu distribution map consistent with the bulk chemical analysis of 492 mg kg-1 Cu in the surface 2.5 cm of the soil. Nevertheless, there is definite heterogeneity
in the distribution of Cu, with areas in the soil matrix where Cu is at or below detection limits, others where Cu content is significantly higher, and a few, very well delineated “hotspots” with a dark red color in the µ-SXRF map, indicating the highest Cu levels found in the thin section. When three different hotspots are reanalyzed with the 20-µm capillary beam, all three distribution maps (Figure 1, right) demonstrate significant fine-scale heterogeneity in the Cu spatial distribution. One map (top) shows very high levels of Cu (the intensity scale goes up to 20 000), distributed along intricate micropore patterns. The other two high-resolution maps display 1000-fold relative intensity differences among neighboring pixels in many places. High-resolution copper distribution maps obtained with the electron microprobe (Figure 2) demonstrate similar features. In the calcareous soil from Burgundy (Figure 2), mapping of a hotspot near two clearly identifiable wood shards reveals that Cu managed to diffuse somewhat within the un-decomposed organic matter and did not accumulate at its periphery. This observation is in line with Flores-Ve´lez et al.’s (24) scanning electron microscopy images of Cucontaminated wood shards in an acid sandy soil from Beaujolais and Adamo et al.’s (25) scanning electron microscopy and energy dispersive X-ray data showing that in Cu-contaminated soils from a mining and smelting region, numerous un-decomposed organic fragments contained high levels of copper, often in association with internal cellular tissues. Practical Significance of Heterogeneous Cu Distribution. The soil solution, rather than soil solids, is generally considered the transport medium of labile species to soil biota. However, at a given time, the soil solution contains only a small fraction of the total amount of an element contained in the solids (23). Moreover, in unsaturated, VOL. 41, NO. 18, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6345
FIGURE 2. Scanned image of a 1-mm thick thin section of the VR-450 soil beside electron microprobe elemental maps of a typical Cu hotspot (white square in scanned picture, containing small wood shards). Small amounts of Cu are co-located with Al, which suggests an association with clay minerals. The highest Cu signal, however, is located on the surface of the wood shards penetrating the fragments to varying degrees and appears co-located with sulfur. structured soils the soil solution may be quite heterogeneous with elemental concentrations that depend not only on the amount of an element or species of interest on the solid but also the strength of sinks driving its partitioning between the solution and solids in soil micropores (26). In this context, the heterogeneous distribution of elements on soil solids is relevant to soil biota. Furthermore, even if soil microorganisms and plants exhibit only a limited ability to avoid the Cu hotspots identified in the previous section, and a fortiori if they are able to avoid these hotspots entirely, the observations presented in Figures 1 and 2 suggest that bulk Cu concentration measurements like those routinely performed by soil testing laboratories are not likely to be very meaningful in terms of biological responses to the presence of Cu. To predict the growth of plants, for example, in Cu-contaminated soils, data about the average bulk Cu concentration in the part of the soil typically explored by plant root systems would need to be complemented by some statistical measure of the heterogeneity of Cu distribution within the soil. In this context, and to deal adequately with the ecology of other biological populations in soils, research is needed to develop new sampling protocols that account for plant root architecture and its plasticity, as well as new methods for sample 6346
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 18, 2007
analysis that provide information on nutrient or contaminant distribution heterogeneity (27). Causes of Heterogeneous Cu Distribution. The causes of the large differences in Cu concentration at different points within the soil thin section may be elucidated by microscopic observation and through comparison of the different elemental maps obtained by µ-SXRF and microprobe analyses. A preliminary, qualitative observation concerning the scanned images of the thin sections is that even though the dark brown areas in the scanned images of the Burgundy soil (Figure 1) might contain significant amounts of organic matter or oxides, they are not particularly rich in Cu. Indeed, no correlation seems to exist between the soil color in the image and the copper distribution in the elemental map. Some of the dark areas are Cu rich and others are not. Indirect evidence of the presence of specific minerals is provided by µ-SXRF spectroscopy. Since spectral K line data were available for elements between Ca and Sr, elemental maps were prepared for all elements with discernible peaks in the spectra including Ca, Ti, Mn, Fe, Ni, Cu, Zn, As, and Rb. Not all the maps contained relevant information. The Ca distribution map, for example, was too color saturated. Otherwise, all the elements mapped were heterogeneously distributed in the soil and at both scales of observation (Figure
FIGURE 3. µ-SXRF maps of Cu, Rb (possible indicator of the presence of clay minerals), and Fe (which could indicate the presence of iron-bearing minerals) in the calcareous soil VR-450. The large maps (2 cm × 4 cm) were obtained with a 0.3-mm spot size. The small maps (1.5 mm × 1 mm) were obtained with a 20-µm spot size. 3). Of the resulting maps, those associated with Rb and Fe are of particular interest. Because Rb is similar in size and charge to potassium, it is often found at trace levels in clay minerals and therefore has been used in X-ray fluorescence analyses for the indirect determination of clay content in soils (28). Fe maps can help detect the presence of ironbearing minerals such as amorphous iron oxides that have been associated with Cu retention in soils (24, 29, 30). The calcareous soil (Figure 3) contains about 40% clay at the surface. At the coarser resolution (0.3 mm pixels), Rb is widely distributed in the thin section and as such appears generally co-located with Cu. At the higher resolution (20 µm pixels), it is evident that although some Cu is sparsely distributed over the same areas as Rb, the most intense Cu signals occur in portions of the soil where there is little, if any, Rb. For Fe, the low-resolution map is too color-saturated (i.e., there is far too much Fe) for a visual assessment of co-occurrence. The high-resolution map shows clearly one area (lower left corner) where there is a lot of Cu and virtually no Fe. However, at other locations, there seems to be some co-occurrence of the two metals, suggesting association of Cu with iron-bearing minerals. The preceding qualitative, visual conclusions can be made more quantitative by plotting pixel by pixel the relative intensities of Cu and Rb, and Cu and Fe, respectively (20, 23). The plots for the low-resolution maps (i.e., 0.3 mm pixel size) (Figure S4, SI) confirm the lack of correlation between Cu and either Rb or Fe. The correlation between Cu and Fe in the calcareous soil (Figure S4D, SI) is the least marginal, with an R2 of 0.4. Parat et al. (29) found a high correlation between Cu and Fe in calcareous soils from Burgundy, but only in the clay-sized fraction of the soil. In the present study the pixel by pixel correlation encompasses the whole soil (i.e., the soil was not size fractionated), which may explain in part the much lower correlation. Nevertheless, the plots (Figure S4C,D, SI) clearly show that the scatter in the data
is large. The data are scattered such that the diagonal quadrant, with high Cu and either high Rb or Fe values, is systematically empty. High Cu values are always associated with low Rb or Fe values, and vice versa. In the higher resolution (i.e., a 20 µm pixel size) correlation plots of the Cu hotspots in the calcareous soil (Figure S5C,D, SI), the data are distributed differently than they were in the low-resolution correlation plots. The largest Cu relative intensities are found at intermediate Fe relative intensity values, not at the lower end of the range. Nevertheless, and in spite of some differences between the two scales of observation, the correlation coefficients are all low (0.5 µm) organic debris, often referred to as particulate organic matter (POM), was Cu-enriched relative to other fractions (24, 25, 35, 36). Therefore, in agreement with earlier investigations, experimental results clearly point to minimally decomposed plant fragments as the strongest sorbents of Cu in the soil considered, an association that in and of itself would account to a large extent for the extremely heterogeneous distribution of Cu in the soil. Whether the plant fragments accumulated Cu after they were incorporated in the soil or, far more likely, they accumulated it during the (foliar) application of the Bordeaux mixture and retained Cu after incorporation, scanning electron micrographs (Figure S6, SI) indicate that their morphology is virtually intact and therefore that they have undergone very limited decomposition. Since the soils were sampled in early spring, almost 6 months after the fields had been harvested, the plant fragments were at least 6 months old. However, they could have been significantly older as well. As observed in other studies (e.g., ref 1), the high Cu content of these plant fragments may have delayed their degradation relative to that of other organic matter in the soil. Thus using electron microprobe and µ-SXRF spectroscopy, this article presents conclusive observations that Cu is very heterogeneously distributed in a Burgundy vineyard soil. Elemental maps obtained via µ-SXRF at 0.3-mm and 20-µm resolutions show that entire regions in the soil are virtually devoid of Cu and would therefore presumably provide a suitable environment for a wide variety of soil microorganisms. Most of the Cu appears to be present in very localized “hotspots” associated with minimally degraded organic matter, which may be protected from further degradation by its elevated copper content. This finding underscores the importance of investigating the microscale spatial distribution of contaminants to ascertain their possible impact on soil microbial communities and processes. It also suggests that current soil sampling protocols, typically leading to bulk contaminant concentrations in the whole root zone, may be largely misleading in terms of the ecotoxicology of metal contaminants in soils and would need to be modified substantially, to provide not only bulk data but also an estimate of the heterogeneity of contaminant distribution.
Acknowledgments We thank Edith Perrier and Thierry Pilorge (IRD, Bondy), Jacques Bonvalot and Pascal Taubaty (Universite´ de Bourgogne), John Hunt and Carole Daugherty (Cornell University), and Darren Dale (CHESS). This work is based upon research 6348
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 18, 2007
conducted at the Cornell High Energy Synchrotron Source (CHESS), which is supported by the National Science Foundation and the National Institutes of Health/ National Institute of General Medical Sciences under award DMR0225180.
Supporting Information Available Soil characterization methods and data (SI), the element distribution maps for thin sections of a sandy, granitic soil (pH 5.6) from a vineyard in Odenas, Beaujolais, France (SII), correlation plots (SIII), and SEM micrographs (SIV) referred to in the discussion are available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Bååth, E. Effects of heavy metals in soil on microbial processes and populations (a review). Water Air Soil Pollut. 1989, 47, 335379. (2) Dumestre, A.; Sauve´, S.; McBride, M.; Baveye, P.; Berthelin, J. Copper speciation and microbial activity in long-term contaminated soils. Arch. Environ. Contam. Tox. 1999, 136, 124131. (3) Courde L.; Reuter S.; Chaussod R.; Andreux F. La de´gradation des herbicides est-elle influence´e par le cuivre accumule´ dans les sols? In Produits phytosanitaires: analyse, re´sidus, me´tabolites, e´cotoxicologie, modes d’action, transfert, Actes du 30e`me Congre`s du Groupe Franc¸ ais des Pesticides, May 29-31, 2000, Reims, France; Couderchet, M.; Eullafroy, P.; Vernet, G., Eds.; Presses Universitaires de Reims: Reims, 2000; pp 69-77. (4) Merrington, G.; Rogers, S. L.; Van Zwieten, L. The potential impact of long-term copper fungicide usage on soil microbial activity in an avocado orchard. Aust. J. Soil Res. 2002, 40, 749759. (5) Becker, J. M.; Parkin, T.; Nakatsu, C. H.; Wilbur, J. D.; Konopka, A. Bacterial activity, community structure, and centimeter-scale spatial heterogeneity in contaminated soil. Microb. Ecol. 2006, 51, 220-231. (6) Huysman, F.; Verstraete, W.; Brookes, P. C. Effect of manuring practices and increased copper concentrations on soil microbial populations. Soil Biol. Biochem. 1994, 26, 103-110. (7) Arnebrant, K.; Bååth, E.; and Nordgren, A. Copper tolerance of microfungi isolated from polluted and unpolluted forest soil. Mycologia 1987, 79, 890-895. (8) Ranjard, L.; Echairi, A.; Nowak, V.; Lejon, D. P. H.; Nouaı¨m, R.; Chaussod, R. Field and microcosm experiments to evaluate the effects of agricultural Cu treatment on the density and genetic structure of microbial communities in two different soils. FEMS Microbiol. Ecol. 2006, 58, 303-315. (9) Yamamoto, H.; Tatsuyama, K.; Uchiwa, T. Fungal flora of soil polluted with copper. Soil Biol. Biochem. 1985, 17, 785-790. (10) Eijsackers, H.; Beneke, P.; Maboeta, M.; Louw, J. P. E.; Reinecke, A. J. The implications of copper fungicide usage in vineyards for earthworm activity and resulting sustainable soil quality. Ecotoxicol. Environ. Saf. 2005, 62, 99-111. (11) Lukkari, T.; Haimi, J. Avoidance of Cu- and Zn-contaminated soil by three ecologically different earthworm species. Ecotox. Environ. Safety 2005, 62, 35-41. (12) Franklin, R. B.; Mills, A. L. Multi-scale variation in spatial heterogeneity for microbial community structure in an eastern Virginia agricultural field. FEMS Microbiol. Ecol. 2003, 44, 335346. (13) Nunan, N.; Ritz, K.; Crabb, D.; Harris, K.; Wu, K.; Crawford, J. W.; Young, I. M. Quantification of the in situ distribution of soil bacteria by large-scale imaging of thin sections of undisturbed soil. FEMS Microb. Ecol. 2001, 36, 67-77. (14) Nunan, N.; Wu, K.; Young, I. M.; Crawford, J. W.; Ritz, K. In situ spatial patterns of soil bacterial populations, mapped at multiple scales, in an arable soil. Microb. Ecol. 2002, 44, 296-305. (15) Strawn, D.; Doner, H.; Zavarin, M.; McHugo, S. Microscale investigation into the geochemistry of arsenic, selenium, and iron in soil developed in pyretic shale materials. Geoderma 2002, 237-257. (16) Negra, C.; Ross, D. S.; Lanzirotti, A. Soil manganese oxides and trace metals: Competitive sorption and microfocused synchrotron X-ray fluorescence mapping. Soil Sci. Soc. Am. J. 2005, 69, 353-361. (17) Vantelon, D.; Lanzirotti, A.; Scheinost, A.; Kretzschmar, R. Spatial distribution and speciation of lead around corroding bullets in
(18)
(19)
(20)
(21)
(22) (23)
(24) (25)
(26)
a shooting range soil studied by micro-X-ray fluorescence and absorption spectroscopy. Environ. Sci. Technol. 2005, 39, 48084815. Manceau, A.; Marcus, M. A.; Tamura, N.; Proux, O.; Geoffroy, N.; Lanson, B. Natural speciation of Zn at the micrometer scale in a clayey soil using X-ray fluorescence, absorption, and diffraction. Geochim. Cosmochim. Acta 2004, 68, 2467-2483. Cousin, I.; Cornu, S.; Rose, J.; Deschatrettes, V.; Montagne, D.; Clozel, B.; Chevallier, P. Distribution of major and trace elements at the aggregate scale in a soil naturally rich in trace elements: A spatial approach using electron microprobe and X-ray microfluorescence analyses. Soil Sci. 2005, 170, 516-529. van Oort, F.; Jongmans, A. G.; Citeau, L.; Lamy, I.; Chevallier, P. Microscale Zn and Pb distribution patterns in subsurface soil horizons; an indication for metal transport dynamics. Eur. J. Soil Sci. 2006, 57, 154-166. Jones, S. D.; Doner, H. E. X-ray fluorescence microprobe analysis of a contaminated soil. Advanced Light Source (ALS) Compendium of User Abstracts, 1999. http://www-als.lbl.gov/als/ compendium/ AbstractManager/uploads/99057.pdf. Jacobson, A. R.; Dousset, S.; Guichard, N.; Baveye, P.; Andreux, F. Diuron mobility through vineyard soils contaminated with copper. Environ. Pollut. 2005, 138, 250-259. Manceau, A.; Marcus, M. A.; Tamura, N.; Quantitative speciation of heavy metals in soils and sediments by synchrotron X-ray techniques. In Applications of synchrotron radiation in lowtemperature geochemistry and environmental science; Fenter, P. A.; Rivers, M. L.; Sturchio, N. C.; Sutton, S. R., Eds.; Mineralogical Society of America: Washington D.C., 2002; pp 341-428. Flores-Ve´lez, L. M.; Ducaroir, J.; Jaunet, A. M.; Robert, M. Study of the distribution of copper in an acid sandy soil by three different methods. Eur. J. Soil Sci. 1996, 47, 523-532. Adamo, P.; Dudka, S.; Wilson, M. J.; McHardy, W. J. Chemical and mineralogical forms of Cu and Ni in contaminated soils from the Sudbury mining and smelting region, Canada. Environ. Pollut. 1996, 91, 11-19. Jacobson, A. R.; Baveye, P. Comment on “Critical evaluation of desorption phenomena of heavy metals from natural sediments”. Environ. Sci. Technol. 2004, 38, 4701-4702.
(27) Baveye, P. Comment on “Modeling soil variation: past, present and future”. Geoderma 2002, 109, 289-293. (28) Schnug, E.; Haneklaus, S. A rapid method for the indirect determination of the clay content by X-ray fluorescence spectroscopic analysis of rubidium in soils. Commun. Soil Sci. Plant Anal. 1996, 27, 1707-1719. (29) Parat, C.; Chaussod, R.; Le´veˆque, J.; Dousset, S.; Andreux, F. The relationship between copper cumulated in vineyard calcareous soils and soil organic matter and iron. Eur. J. Soil Sci. 2002, 53, 663-669. (30) Arias, M.; Lo´pez, E.; Ferna´ndez, D.; Soto, B. Copper distribution and dynamics in acid vineyard soils treated with copper-based fungicides. Soil Sci. 2004, 169, 796-805. (31) Ribeiro, A. B.; Nielsen, A. A. An application of discriminant analysis to pattern recognition of selected contaminated soil features in thin sections. Geoderma 1997, 76, 253-262. (32) Lamy, I.; Sterckeman, T.; Cambier, P.; Jaffrezic, A.; van Oort, F.; Baize, D.; Chaussod, R.; Denaix, L.; Cornu, S. Pre´sence et impact des e´le´ments en traces dans les sols. In Sols et environnement; Girard, M.-C.; Walter, C.; Re´my, J.-C.; Berthelin, J.; Morel, J.-L., Eds.; Dunod: Paris, 2005; pp 470-490. (33) Dijkstra, J. J.; Meeussen, J. C. L.; Comans, R. J. J. Leaching of heavy metals from contaminated soils: An experimental and modeling study. Environ. Sci. Technol. 2004, 38, 4390-4395. (34) Kim, B.; McBride, M. B. A test of sequential extractions for determining metal speciation in sewage sludge-amended soils. Environ. Pollut. 2006, 475-482. (35) Ducaroir, J.; Lamy, I. Evidence of trace metal association with soil organic matter using particle size fractionation after physical dispersion treatment. Analyst 1995, 120, 741-745. (36) Besnard, E.; Chenu, C.; Robert, M. Influence of organic amendments on copper distribution among particle-size and density fractions in Champagne vineyard soils. Environ. Pollut. 2001, 112, 329-337.
Received for review March 22, 2007. Revised manuscript received June 14, 2007. Accepted June 27, 2007. ES070707M
VOL. 41, NO. 18, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6349
