Arsenic transformations in the soil rhizosphere plant system fundamental and potential application to ph

Arsenic transformations in the soil rhizosphere plant system fundamental and potential application to ph

Arsenic transformations in the soilÁ/rhizosphereÁ/plant system: fundamentals and potential application to phytoremediation Walter J. Fitz, Walter W. Wenzel * Institute of Soil Science, University of Agricultural Sciences Vienna-BOKU, Gregor Mendel Strasse 33, A-1180 Vienna, Austria Received 3 September 2001; received in revised form 24 May 2002; accepted 27 May 2002 Abstract This paper reviews major processes that potentially affect the fate of arsenic in the rhizosphere of plants. Rhizosphere interactions are deemed to play a key role in controlling bioavailability to crop plants and for a better understanding and improvement of phytoremediation technologies. Substantial progress has been made towards an understanding of As transformation processes in soils. However, virtually no information is available that directly addresses the fate of As in the rhizosphere. We are proposing a conceptual model of the fate of As in the soilÁ /rhizosphereÁ/plant system by integrating the state-of-the art knowledge available in the contributing disciplines. Using this model and recent studies on hyperaccumulation of As, we discuss research needs and the potential application of rhizosphere processes to the development of phytoremediation technologies for As-polluted soils. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Arsenic; Bioavailability; Hyperaccumulation; Mycorrhiza; Phytoremediation; Rhizosphere 1. Introduction Arsenic is an ubiquitous trace metalloid and is found in virtually all environmental media. How- ever, concentrations of As in non-contaminated soils are typically well below 10 mg kg (1 . Its presence at elevated concentrations in soils is due to both anthropogenic and natural inputs. Anthro- pogenic sources include mining and smelting processes besides application of As-based insecti- cides, herbicides, fungicides, algicides, sheep dips, wood preservatives, dyestuffs, feed additives and compounds for the eradication of tapeworm in sheep and cattle (Adriano, 2001). Geochemical sources of As-contaminated soils include As-rich parent material as As easily substitutes for Si, Al or Fe in silicate minerals (Bhumbla and Keefer, 1994). Arsenic is also commonly associated with sulfides, e.g. in sulfidic ore deposits. Other natural sources of As include volcanic activities, wind- born soil particles, sea salt sprays and microbial volatilisation of As (Nriagu, 1990; Frankenberger and Arshad, 2002). It has been estimated that there are potentially 1.4 million contaminated sites within the European Community impacted to various extent by organic and/or trace metal/metalloid pollutants (European * Corresponding author. Tel.: '/43-1-47654-3119; fax: '/43- 1-4789-110 E-mail address: wwenzel@edv1.boku.ac.at (W.W. Wenzel). Journal of Biotechnology 99 (2002) 259 Á /278 www.elsevier.com/locate/jbiotec 0168-1656/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 1 6 5 6 ( 0 2 ) 0 0 2 1 8 - 3 Topic Centre Soil, 1998). Forty-one percent of the superfund sites in the USA for which US EPA has signed records of decision are contaminated with As (US EPA, 1997), more than 10 000 As-con- taminated sites have been reported for Australia (Smith et al., 2002). Though considerable progress has been made in reducing atmospheric inputs of As in Western Europe (Schulte and Gehrmann, 1996), pollution by As and other trace metals at a large scale can still occur as shown by the Don ˜ ana ecological disaster in southern Spain (Pain et al., 1998). Drinking of As-contaminated groundwater is perhaps the most common exposure pathway of humans to As toxicity. The biggest known As calamity occurred in the Bengal Delta (Bangla- desh/West Bengal) where millions of people de- pend on As-rich drinking water (Chakraborti et al., 2001). Natural contamination of groundwater by As has been also recorded for many other parts in the world. Berg et al. (2001) reported a recently discovered case of groundwater contamination in Hanoi (Vietnam) with contamination levels vary- ing from 1 to 3050 mgl (1 . Technologies currently available for the reme- diation of metal/metalloid contaminated soils are expensive, time consuming, can create risks to workers and produce secondary waste (Wenzel et al., 1999a; Lombi et al., 2000a,b). Recently phy- toremediation, the use of green plants to clean up contaminated soil, has attracted much attention (Baker et al., 1991; McGrath et al., 1993). Basi- cally, the strategy of phytoremediation can be divided into five fundamental processes that apply to As, including phytoextraction, stabilisation, immobilisation, volatilisation and rhizofiltration (Salt et al., 1998; Wenzel et al., 1999a,b). A major step towards the development of phytoremedia- tion of As-impacted soils is the recent discovery of the As-hyperaccumulating ferns Pteris vittata and Pityrogramma calomelanos. Both plants produce large biomass and are therefore promising candi- dates for phytoextraction purposes (Ma et al., 2001; Francesconi et al., 2002; Tu and Ma, 2002; Visoottiviseth et al., 2002). This paper is focused on the basic processes involved in As transformation in the soilÁ / rhizosphereÁ/plant system. Rhizosphere interac- tions are deemed to play a key role in controlling bioavailability to crop plants (Hinsinger, 2001) and for a better understanding and improvement of phytoremediation technologies (Wenzel et al., 1999b; Lombi et al., 2001). However, virtually no literature is available which refers particularly to the biogeochemistry of As in the rhizosphere. The bulk of available literature is related to more general aspects of soilÁ /plantÁ/As relationships. Several comprehensivereviews are available on As in the soil system (e.g. Bhumbla and Keefer 1994; O’Neill, 1995; Sadiq, 1997; Smith et al., 1998; Adriano, 2001). However, the fate of As in the rhizosphere has yet not been explored. This review addresses major processes potentially in- volved in the fate and transformation of As in the soilÁ /rhizosphereÁ/plant system in order to present conceptual models and hypotheses while high- lighting future research needs to enhance the scientific basis for further development of phytor- emediation technologies. 2. Arsenic transformations in the soil Á /plant Á / microbe system 2.1. General Arsenic has been known to have a high affinity for oxide surfaces, which is affected by several biogeochemical factors such as soil texture, or- ganic matter, nature and constituents of minerals, pH, redox potential and competing ions (Adriano, 2001). The activity of As in soil solution is most commonly controlled by surface complexation reactions on oxides/hydroxides of Al, Mn and especially Fe (Inskeep et al., 2002). Smaller textural fractions contain larger sorbed and total amounts of As (Lombi et al., 2000a,b), as sorbing oxides/hydroxides are typically concentrated in the clay size fraction ( B /2 mm) due their small size. This explains also the lower toxicity of fine textured compared with coarse textured As-pol- luted soils (Jacobs et al., 1970). Analyses of drainage waters derived from mine tailings have shown that suspended material (! /0.45 mm) is the main carrier of arsenic and mainly responsible for W.J. Fitz, W.W. Wenzel / Journal of Biotechnology 99 (2002) 259 Á/278260 metal fluxes into ground and surface waters (Roussel et al., 2000). About 25 different As compounds havebeen identified in biological samples, mainly in marine ecosystems (Francesconi and Edmonds, 1993). However, usually only the organic species mono- methylarsonic acid (MMAA) and dimethylarsinic acid (DMAA) are found in detectable concentra- tions in soils besides abundant inorganic As V and As III species (Takamatsu et al., 1982, Tlustosˇ et al., 2002). Paddy soils typically show larger extracta- ble concentrations of MMAA and DMAA which suggests that methylated arsenicals are produced under anaerobic conditions (Takamatsu et al., 1982). In very few cases trimethylarsine oxide (TMAO) and arsenobetaine (AB) havebeende- tected as minor compounds in soil extracts (Geis- zinger et al., 2002). Toxicity and chemical behaviour of As com- pounds are largely influenced by the form and speciation of As. As III is more mobile and more toxic than As V . Gaseous arsines are most toxic whereas arsenobetaine and arsenocholine (mainly found in marine organisms) are nontoxic. As a rule, inorganic arsenicals are more toxic than organic arsenicals and the trivalent oxidation state is more toxic than the pentavalent oxidation state (Fowler, 1977; Adriano, 2001). Though most studies did not directly investigate the fate of As in the rhizosphere we highlight in the following the major processes taking place in the rhizosphere to assess the potential interactions with the fate of As in the soilÁ /plantÁ/microbe system. 2.2. Fate of arsenic as related to rhizosphere acidification/alkalinisation It is generally known that rhizosphere pH may considerably differ from that in the bulk soil. Depending on plant and soil factors pH differ- ences can be up to two units. Factors affecting rhizosphere pH are the source of nitrogen supply (NO 3 ( vs. NH 4 ' uptake), nutritional status of plants (e.g. Fe and P deficiency), excretion of organic acids, CO 2 production by roots and rhizosphere microorganisms, and the buffering capacity of the soil (Marschner, 1995). Several studies have been carried out on pH- dependent As sorption in soils and on pure mineral phases. Studies using soil and pure Fe hydroxides generally agree that As V solubility increases upon pH increase within pH-ranges commonly found in soil (pH 3Á /8), whereas As III tends to follow the opposite pattern (Manning and Goldberg, 1997; Smith et al., 1999; Tyler and Olsson, 2001; Raven et al., 1998; Jain and Loep- pert, 2000). Thermodynamic calculations suggest that H 2 AsO 4 ( dominates below and HAsO 4 2( above pH 6.97 (Sadiq, 1997). Furthermore, net surface charges of soil constituents become more negative as functional groups dissociate protons upon pH increase. Conversion of H 2 AsO 4 ( to HAsO 4 2( along with increasing negative surface charges of soil constituents lead to As V mobilisa- tion as electrostatic repulsion is enhanced particu- larly abovepH7(Â /pK 2 ). Moreover, as well the oxide concentration of soil has considerable influ- ence on the pH-dependent solubility of As. Smith et al. (1999) found that for soils low in oxidic minerals pH had little effect on the amount of adsorbed As V whereas highly oxidic soils showed a pronounced decrease of As V adsorption upon pH increase. In contrast to As V , solubility of As III decreases with decreasing pH in soil. The pK 1 of arsenous acid (H 3 AsO 3 0 ) is 9.22, which implies that below pH 9.22 the As III species is mainly uncharged (Sadiq, 1997). This may contribute to the generally larger solubility of As III in soil systems. Most soils exhibit oxic conditions, hence an increase of rhizosphere pH could favour mobilisa- tion of labile and exchangeable As V -fractions in the root vicinity and consequently enhance plant uptake. Nitrogen nutrition, as it is most respon- sible for the cation/anion uptake ratio, greatly affects rhizosphere pH (Marschner and Ro ¨ mheld, 1983). Hence, fertilisation of plants grown on As- contaminated soil with NO 3 ( as the N source, would potentially increase rhizosphere pH, and thus possibly enhance As accumulation in plant tissues. On the other hand there are distinct differences in rhizosphere acidification among plant species. For instance legumes and actinorhi- zal plants meet their N supply by symbiontic N 2 W.J. Fitz, W.W. Wenzel / Journal of Biotechnology 99 (2002) 259 Á/278 261 fixation. N 2 enters the root uncharged, thus the cation/anion uptake ratio of N 2 fixing plants is large and results in a net H ' release by the plant (Marschner, 1995). Rhizosphere acidification by N 2 -fixing symbionts would favour As V immobili- sation in soil under oxic conditions. The As hyperaccumulator P. vittata was re- ported to prefer calcareous soils of neutral to slightly alkaline pH (Jones, 1987; Ma et al., 2001). This implies that changes of rhizosphere pH would be no prerequisite for As-hyperaccumulation due to the high pH-buffer power of calcareous soils. However, P. vittata and P. calomelanos havebeen as well found on acidic soils and mine tailings in Thailand. An increase of the rhizosphere pH could potentially increase As V solubility and possibly plant uptake on such substrates. On the other hand very low pH values may dissolve As sorbents such as Fe oxides/hydroxides (see Sections 2.6 and 2.4). 2.3. Root exudation It has been reported that P-deficient plants show an enhanced exudation of carboxylic acids, such as citric and malic acid (Hoffland, 1992; Neumann and Ro ¨ mheld, 1999; Kirk et al., 1999). This response is thought to change soil pH, to displace P from sorption sites, to chelate metal cations that could immobilise P or to form soluble metal- chelate complexes with P, resulting in enhanced availability of P (Kirk et al., 1999). Cluster roots of P-deficient plant species such as Lupinus albus and members of the Protaceae exude particularly strong organic acids and phenolics (Dinkelaker et al., 1995). Arsenic and P belong to the same chemical group and have comparable dissociation constants for their acids and solubility products for their salts, resulting in similar geochemical behaviour of As and P in soil (Adriano, 2001). Hence, it is reasonable to assume that carboxylate exudation could play a role in the mobilisation of As in the rhizosphere and enhance As uptake by plants. Basically two strategies have been identified for acquisition of Fe by higher plants. Strategy I exists in monocotyledenous species, with the exception of graminaceae (grasses), and dicotyledenous spe- cies, and involves three processes: (1) enhanced net excretion of protons, (2) a plasma membrane- bound inducible reductase, and (3) enhanced release of reducing and chelating agents. Strategy II, confined to grasses, is characterised by release of phytosiderophores and a high-affinity transport system for Fe uptake (Marschner and Ro ¨ mheld, 1994). Fe-oxides/hydroxides typically dominate As sorption in soil (see Section 2.7). Laboratory studies of arsenate and arsenite adsorption on Fe-oxide surfaces indicate that both species are bound as mono and bidentate surface complexes (Waychunas et al., 1993; Sun and Donner, 1996). The excretion of protons and/or the release of reducing and chelating compounds by strategy I plants also could result in co-dissolution of As from Fe-oxides/hydroxides, rendering As more soluble and available to plants. Admittedly, virtually nothing is known about Fe nutritional aspects and related rhizosphere processes of fern plants. They are sensu strico neither strategy I nor strategy II plants as ferns belong to the Pteridophyta. However, ferns such the As-hyperaccumulator P. vittata and P. calo- melanos certainly acquire Fe. P. vittata is known to grow commonly on calcareous soil (Jones, 1987). It has been reported that root exudates (oxalic and citric acid) of acidifuge plants effec- tively mobilise P and Fe from lime stone (Stro ¨ met al., 1994). Porter and Peterson (1975) found a highly significant correlation (P B /0.001) between As and Fe in several As-tolerant plants from different mine sites in UK. No correletions were found between As and other elements (Pb, Cu, Zn), not even for P. In conclusion we suggest that P, Fe and As uptake by As hyperaccumulator species may be related to each other. Reductive dissolution of Fe III minerals inevitably dissolves Fe-bound As, root exudates enhancing P mobilisation are likely to desorb As as well. Besides rhizosphere processes As-hyperaccumulator most likely posses a particular As-uptake mechanism whereas sup- pression of the high-affinity phosphate uptake system is involved in adaptive tolerance of plants to As (Meharg and Macnair, 1992a see Section 2.7). W.J. Fitz, W.W. Wenzel / Journal of Biotechnology 99 (2002) 259 Á/278262 2.4. Redox potential The redox potential significantly influences speciation and solubility of As in soils (e.g. Deuel and Swoboda, 1971; Masschelyn et al., 1991; Marin et al., 1993; McGeehan and Naylor, 1994; Onken and Hossner, 1995, 1996). Generally, inorganic As is mainly present as As V in aerobic conditions (high redox potential) and as As III in flooded (low redox potential) soils. Arsenic is less toxic and less mobile in the ' /V than in the '/III oxidation state. It has been repeatedly observed that increased As solubility under reduced condi- tions is associated with dissolution of Fe and Mn oxides/hydroxides. Significant correlations have been found between dissolved Fe and As (Masschelyn et al., 1991; Marin et al., 1993; McGeehan and Naylor, 1994), confirming that Fe oxides/hydroxides represent the major sorbing agents for As in soils (see Section 2.6). Flooding had no influence on soluble Ca and Al (Massche- lyn et al., 1991). Masschelyn et al. (1991) investi- gated redoxÁ /pH relations of As V and As III stability using an apparatus which allowed pH and redox control of a stirred soil suspension. Under oxidised conditions, soluble As concentra- tions were three times larger at pH 8 than at pH 5, because of the decreased positive surface charge at high pH. Under reducing conditions As III became the major dissolved species with total soluble As being smaller at pH 8. Dissolved Fe concentrations did not significantly increase upon reduction at pH 8(Masschelyn et al., 1991). In contrast, Marin et al. (1993), using the same experimental set up, reported increased As solubility upon pH decrease (7.5Á /5.5) for both reduced and oxidised conditions without providing any explanation. As concentra- tions in rice (Oryza sativa L.) increased upon decreasing redox potential (Marin et al., 1993). The oxidation of the rhizosphere is a well known phenomenon for paddy rice as these plants are able to transport O 2 through aerenchyma to roots, which results in a leakage of O 2 into the rhizo- sphere (Flessa and Fischer, 1992). Rice roots grown in reduced suspensions were coated with Fe plaque containing As (Marin et al., 1993). Doyle and Otte (1997) found formation of Fe plaque also around roots of salt marsh plants which led to an effective fixation and consequently detoxification of As and Zn in the rhizosphere. 2.5. AsÁ /P interactions Similar to carboxylic acids released by plant roots, other organic and inorganic anions may compete with As for sorption sites. The phosphate ion plays a prominent role in anion Á /As interac- tions due to its physicochemical similarity to As (Adriano, 2001). Moreover, arsenate is thought to be taken up via the phosphate uptake system and may consequently interact with plant P nutrition (Asher and Reay, 1979; Meharg and Macnair, 1990). Though numerous studies on As Á /P inter- actions have been published, results have not been explored systematically and yet have not been applied to the rhizosphere. Table 1 gives a compilation of studies on AsÁ /P interactions with respect to mobilisation/extractability, plant uptake and phytotoxicity of As. Generally it is reasonable to distinguish between hydroponics (solution cul- ture), pot/column/batch and field experiments. Hydroponic experiments inevitably overestimate the importance of uptake kinetics of the plant in consideration (Meharg et al., 1994) and typical processes of soilÁ /plant relationships such as water flow, nutrient/pollutant mass flow to the root surface, diffusion, adsorption/desorption and ion exchange are not considered as soil is absent in such an experimental set up. Consequently, P additions in solution culture studies decrease As uptake and mitigate As-caused phytotoxicity symptoms (Hurd-Karrer, 1939; Asher and Reay, 1979; Tsutsumi, 1982; Meharg and Macnair, 1990; De Koe and Jaques, 1993). Summarising results of pot, laboratory and field experiments leads to a different conclusion. Phos- phorus additions at high rates enhance As leaching in laboratory column studies (Woolson et al., 1973; Peryea and Kammereck, 1995), increase extractable fractions of As in batch experiments (Carrow et al., 1975; Peryea, 1991) and reduced sorption of As V and As III onto soils (Smith et al., 2002). Plant uptake of As has been shown to increase upon P application in pot experiments (Creger and Peryea, 1994; Jiang and Singh, 1994; Woolson, W.J. Fitz, W.W. Wenzel / Journal of Biotechnology 99 (2002) 259 Á/278 263 1972, Woolson et al., 1973) and at field scale (Small and McCants, 1962). In contrast to solution culture studies, presence of P causes As Á /P compe- tition for sorption sites resulting in increased As bioavailability, and hence higher As concentra- tions in plants. Quaghebeur and Rengel (2001) studied As Á /P interactions in the rhizosphere and found that the presence of P significantly increased As concentrations in shoots and roots of both tolerant and non-tolerant clones of Holcus lanatus. Jacobs and Keeney (1970) compared As accumu- lation in corn from artificially contaminated soils (20 and 80 mg kg (1 ). Arsenic concentrations were larger in plants grown on sandy soil compared with a silty loam. Increasing the level of P in soil had little effect on plant uptake of As on the silt loam but showed a marked increase on the sandy soil when As was present at 80 mg kg (1 . Both reduced and increased phytotoxicity symp- toms had been found after P additions to soil grown plants. Differences in texture and mineral content affect also As Á /P relationships. Hurd- Karrer (1939) found in pot experiments improved growth of As-injured wheat on clay loam and sandy loam soils upon addition of P. In contrast, Woolson et al. (1973) reported reduced growth of corn after P fertilisation on a sandy loam and enhanced growth on silty clay loam. Jacobs and Keeney (1970) found enhanced As-toxicity to corn on a sandy soil but little effects on a silty loam. Benson (1953) found good response of barley growth to added P only on seven of 17 soils with toxic concentrations of As. Schweizer (1967) studied toxicity of disodium methanearsonate (DSMA) residue applied on two silty loam soils and found that P additions increased phytotoxicity symptoms. Benson (1953) tested P additions also at field scale and found no yield response. In this study superphosphate fertiliser was applied in dry form in 10 cm deep trenches 10 cm away from the seeds. However, phosphorus is known to be very immobile in soils (Marschner, 1995), which likely resulted in spatially confined AsÁ /PÁ/root interac- tions. In case of As-hyperaccumulating plants it is very unlikely that P fertilisation may cause phytotoxi- city problems as it has been reported that P. vittata accumulates up to 22 630 mg As kg (1 in dry matter on soil spiked with 1500 mg As kg (1 (Ma et al., 2001). Therefore, P additions may in Table 1 Selection of literature on the response of mobilisation, plant uptake, phytotoxicity of As to P additions in hydroponics, pot/column/ batch and field experiments Effect Species Response Experimental set up Hydroponic Pot/column/batch Field Mobilisation/extractability As V Increased 1, 2, 3, 6, 13, 18 In soil As III Increased 18 As-plant uptake As V Increased 12, 14, 13, 17 15 Decreased 5, 10, 19 Phytotoxicity Root elongation As V Increased 9, 11, 19 As III Slightly increased 9 Yield As V Increased 4 3, 4, 6, 8 Decreased 3, 14 No response 8, 14 8 Plant height DSMA a Decreased 7 a Disodium methanearsonate. 1, Peryea and Kammereck (1995);2,Peryea (1991);3,Woolson et al. (1973);4,Hurd-Karrer (1939);5,Asher and Reay (1979);6, Carrow et al. (1975);7,Schweizer (1967);8,Benson (1953);9,Tsutsumi (1982); 10, Meharg and Macnair (1990); 11, De Koe and Jaques (1993); 12, Jiang and Singh (1994); 13, Jacobs and Keeney (1970); 14, Woolson (1972); 15, Small and McCants (1962); 16, Creger and Peryea (1994); 17, Quaghebeur and Rengel (2001); 18, Smith et al. (2002); 19, Sneller et al. (1999). W.J. Fitz, W.W. Wenzel / Journal of Biotechnology 99 (2002) 259 Á /278264 the first place enhance plant growth and secondly mobilise exchangeable As resulting in increased total As uptake. Most studies on AsÁ /P interactions were carried out using spiked soil and delivered valuable information. However, concentrations of labile As and P can be expected to be substantially different in soils that gradually received As during extended periods through various anthropogenic processes such as mining and smelter activities (Wenzel et al., 2002a). It has been demonstrated that As rapidly becomes recalcitrant in soil with time (Lombi et al., 1999; Onken and Adriano, 1997), resulting in reduced toxicity (Jiang and Singh, 1994). Large proportions of total soil P may be present in organic forms such as phytates (Dalal, 1977; Marschner, 1995). Similarly, up to 70% of dis- solved P in soil solution were found to be present as organic P (Helal and Sauerbeck, 1984). Phytic acid is expected to compete with As for sorption sites due to its anionic nature. However, interac- tions of As with organic P have not been studied yet. 2.6. Binding forms of As in soil Despite apparent similarities between the chem- istry of As and P, some important differences have to be considered. Unlike P, As is present also in oxidation state III, and besides oxygen other ligands may form stable species that are not found with P (O’Neill, 1995). The traditional Chang and Jackson (1957) procedure, developed for sequential extraction of P, has been adopted for fractionation of As in soils (e.g. Woolson et al., 1971, 1973; Akins and Lewis, 1976; Onken and Hossner, 1996; Onken and Adriano, 1997; Wasay et al., 2000). It has been assumed that this extraction scheme addresses, with respect to P, the so-called water-soluble plus adsorbed (NH 4 Cl-extractable) As and the Al- (NH 4 F-extractable), Fe- (NaOH-extractable), and Ca-bound (H 2 SO 4 -extractable) As fractions. Based on this extraction procedure (Woolson et al., 1971, 1973; Akins and Lewis, 1976; Wasay et al., 2000) and comparisons between extractable fractions of As and Fe/Al oxide/hydroxide content of soils (Wauchope, 1975; Johnston and Barnard, 1979; Polemio et al., 1982; Manning and Gold- berg, 1997; Chen et al., 2002) it has been suggested that Fe-oxides/hydroxides represent the major sink for As sorption in soils, whereas the importance of Al- and Ca-bound fractions are variable. In none of the studies using a modified Chang and Jackson (1957) procedure for As fractionation co-dissolved Al, Fe and Ca were analysed in the extracts. Studies presenting data on co-dissolved Al, Fe and Ca prove that the Ca-bound As plays a minor role in As sorption even in calcareous soils (Wenzel et al., 2001a; Shiowatana et al., 2001). Hence, only minor proportions of As were found in extracts (1 M sodium acetateÁ /acetic acid buffer) of soils addressing Ca-bound metal fractions (Wenzel et al., 2001a). These findings are in agreement with results of energy dispersive X-ray microanalysis (EDXMA), providing evidence for strong association of As with hydrous Fe oxides (Lombi et al., 2000b). Oxides/hydroxides of Al, Mn and Fe are also known to form coatings on other soil particles such as clays. Fordham and Norrish (1983) reported that adsorption of As V in a lateritic Podzol was mainly controlled by Fe oxide deposits on kaolin flakes, showing a high degree of substitution of Al for Fe within Fe oxide particles. Little research has been done on As adsorption by organic matter. Thanabalasingam and Picker- ing (1986) found sorption of As onto humic acids in batch experiments. However, this was primarily related to the ash content of the humic acids used. There is no evidence that soil organic matter (SOM) would contribute in significant quantities to As sorption in soils, especially in the presence of effective sorbents such as hydrous Fe oxides (Livesey and Huang, 1981; Wenzel et al., 2001a). Risk assessment of As-polluted sites of both anthropogenic and geogenic origin even revealed enhanced As solubility in organic surface horizons of forest soils (Brandstetter et al., 2000; Wenzel et al., 2002a). This can be explained by the anionic nature of many organic compounds in soil, result- ing in reduced As adsorption on Al and Fe oxides/ hydroxides (Fordham and Norrish, 1983; Xu et al., 1988). W.J. Fitz, W.W. Wenzel / Journal of Biotechnology 99 (2002) 259 Á/278 265 Turpeinen et al. (1999) used a modified metal sequential extraction procedure of Tessier et al. (1979) for As-fractionation. They reported that up to 14.4% of total As was organically bound, however, without testing the procedure for its applicability to As. Even though fractionation of soil-As by sequen- tial extraction procedures delivers only operation- ally defined As forms, its application in rhizosphere investigations may be useful in order to determine As pools of differential bioavailabil- ity, e.g. determination of As-pools potentially accessible to hyperaccumulator plants. Fig. 1 shows range and median values (%) of As dis- tribution among the five fractions of the Wenzel et al. (2001a) sequential extraction procedure. Twenty polluted soils of both anthropogenic and geogenic As origin were used. These results reveal that most As is associated with the Fe oxides/ hydroxides (fraction 3' /4). The amount of the non-specifically sorbed fraction (readily mobile) is small but most important for risk assessment with respect to potential ground water pollution (Brandstetter et al., 2000; Wenzel et al., 2002a). 2.7. Plant uptake of As There is no evidence that As is essential for plants, though growth is stimulated when supplied at low concentrations (Liebig et al., 1959; Lepp, 1981; Carbonell et al., 1998). From hydroponic experiments on plant uptake of As it is known that the chemical form of supplied As is more impor- tant than total As concentrations in solution. In solution culture experiments Marin et al. (1992) found that the phytoavailability for two rice cultivars followed the order DMAA B /As V B/ MMAAB/As III , while Carbonell-Barachina et al. (1998) obtained the order of DMAAB/MMAA$/ As V B/As III for two typical wetland plant species of the Lousiana salt marshes. However, both reports agree that upon absorption, inorganic species and MMAA were mainly accumulated in roots. In contrast DMAA was readily translocated to the shoots resulting in shoot/root As concentra- tion ratios of ! /1. Tlustosˇ et al. (2002) conducted a pot experiment on As uptake in radish grown on soil amended with As III ,As V and DMAA. As III was readily oxidised to As V , resulting in no differences in As accumulation and yield between these two treatments. Water extracts showed that DMAA was adsorbed to a much lesser extent than As V , causing a significant reduction of radish biomass production, although total As concentra- tions were similar to the other As treatments. Plants capable of accumulating exceptionally large concentrations of metals have been termed hyperaccumulators (Brooks et al., 1977). Recently, the first As-hyperaccumulating plants, the ferns P. vittata and P. calomelanos have been discovered. Both ferns produce large biomass, and are there- fore, promising candidates for phytoextraction purposes (Ma et al., 2001; Francesconi et al., 2002; Visoottiviseth et al., 2002). However, some confusion has entered the discussion on As-hyper- accumulator plants lately. Formerly reported As- tolerant plants grown on heavily As-polluted soils and mine tailings have been repeatedly termed as As hyperaccumulators (Francesconi et al., 2002; Francesconi and Kuehnelt, 2002; Visoottiviseth et al., 2002; Geiszinger et al., 2002). Table 2 provides an overview on reported shoot, root and substrate concentrations of As hyperaccumulators and As- tolerant plants. The biological absorption coeffi- cients (BAC, defined as the total element concen- tration in shoots with respect to total element concentration in soil, both in mg kg (1 ) and accumulation factors (AF, defined as the total element concentration in shoots with respect to total element concentration in roots, both in mg Fig. 1. Partitioning of As among the five fractions of a sequential extraction procedure in 20 test soils. Upper case symbols refer to: (a) non-specifically sorbed, (b) specifically- sorbed, (c) bound to amorphous and poorly-crystalline hydrous oxides of Fe and Al, (d) bound to well-crystallised hydrous oxides of Fe and Al. Arsenic pollution was caused by both natural and anthropogenic inputs. Extracted from Wenzel et al. (2001a). W.J. Fitz, W.W. Wenzel / Journal of Biotechnology 99 (2002) 259 Á /278266 kg (1 ) were calculated where possible. Compari- sons between hyperaccumulators and tolerant plants evidently show the difference in the As accumulation behaviour. Whereas tolerant plants tend to restrict soilÁ /plant and rootÁ/shoot transfer, hyperaccumulators actively take up and translo- cate As into above-ground tissues. Plants exhibit- ing AF and particularly BAC values B /1 do not represent candidates for phytoextraction. P. vittata grown on As-spiked soil resulted in dry matter concentrations of As as high as 22 630 mg kg (1 (Ma et al., 2001). Based on the accumu- latorÁ /excluder concept of Baker (1981) As toler- ant plants (Table 2) should be termed as excluders at AF ratios & /1, even they show elevated concentrations in above ground tissues. Hyperaccumulation of As seems to be rather constitutive than adaptive as populations from non-contaminated environments hyperaccumulate As as well (Table 2). Biomass production of P. vittata has been shown to increase upon As applications, suggesting the status of a beneficial element for this plant (Tu and Ma, 2002). Root- induced rhizosphere processes can be anticipated to facilitate As uptake by hyperaccumulator plants. Arsenate is thought to be taken up via the phosphate uptake system (Asher and Reay, 1979; Meharg and Macnair, 1990). Solution culture studies revealed that tolerant populations of Agrostis capillaris (Porter and Peterson, 1975) and H. lanatus (Meharg and Macnair, 1991a) take up less As than non-tolerant plants. Meharg and Macnair (1992a) showed that arsenate uptake in solution culture of a As-tolerant H. lanatus population was caused by the suppression of the high affinity P uptake system (the high affinity uptake system for P is dominant at concentrations B /0.1 mmol l (1 ; Clarkson and Lu ¨ ttge, 1991), resulting in smaller As influx and accumulation in tolerant populations. Similar arsenate tolerance mechanisms were observed for As-tolerant popu- lations of Deschampsia cespitosa and to a lesser extent as well for A. capillaris (Meharg and Macnair, 1991b). In contrast, no down regulation of arsenate/phosphate transporters was found for As-tolerant Calluna vulgaris (Sharples et al., 2000a). Recent findings of Hartley-Whitaker et al. (2001) suggest that arsenate tolerance in H. lanatus requires both adaptive suppression of the high-affinity phosphate uptake system and consti- tutive phytochelatin production. Phytochelatins Table 2 Arsenic accumulation in hyperaccumulator and As tolerant plants Plant species As in plants (mg kg (1 ) As in soil (mg kg (1 ) BAC a AF b References Frond/shoot Root Hyperaccumulators P. vittata 22 630 1500 c 15 1 7234 303 97 74 23.8 1 755 6 126 1 P. calomelanos 8000 88 135 59 91 2 Tolerant plants (non-accumulators) A. capillaris 3470 26 500 0.13 3 Agrostis catellana 170 1000 17 000 0.01 0.17 4 Agrostis delicatula 300 1800 17 000 0.018 0.17 4 Cynodon dactylon 1600 10 850 9530 0.17 0.15 5 Paspalum tuberosum 1130 7670 0.147 6 Spergularia grandis 1175 7670 0.15 6 Only data of live plant material was collected. 1, Ma et al. (2001);2,Francesconi et al. (2002);3,Porter and Peterson, (1975);4,De Koe, (1994);5,Jonnalagadda and Nenzou (1996, 1997);6,Bech et al. (1997). a Biological absorption coefficient (shoot/soil concentration ratio). b Accumulation factor (shoot/root concentration ratio). c Spiked soil. W.J. Fitz, W.W. Wenzel / Journal of Biotechnology 99 (2002) 259 Á /278 267 seem to be also involved in detoxification of As in Silene vulgaris (Sneller et al., 1999)andNicotiana tabacum (Nakazawa et al., 2000). Most As in fronds of P. vittata and P. calomelanos is present as As III , whereas As V dominates in roots (Ma et al., 2001; Francesconi et al., 2002). As metabolism and complexation in plants havebeenreviewed by Meharg and Hartley-Whitaker (2002) and are therefore not discussed here. From hydroponic experiments one may con- clude that tolerant populations having a sup- pressed phosphate/arsenate high affinity uptake system are less efficient in absorbing P and hence would produce less biomass. Meharg et al. (1994) compared P uptake of tolerant and non-tolerant H. lanatus populations in both solution culture and in pot experiments with sterile potting com- post. At low P concentrations (0.5 and 5 mM) the tolerant clones showed reduced plant P concentra- tion and shoot biomass, but a higher percentage of root biomass. These differences were not found at high P concentrations in solution (50 mM). In contrast, tolerant plants grown in pots had smaller growth rates but higher P concentrations in their tissues. As discussed by the authors, hydroponic experiments tend to overestimate the importance of uptake kinetics, as influx rather than P diffusion (Nye, 1977), root morphology and soil parameters (Silberbush and Barber, 1983) appear to be the rate limiting steps. 2.8. Mycorrhizal associations and other microbial interactions in the rhizosphere Mycorrhizas are the most widespread mutualis- tic symbiotic association between microorganisms and higher plants and can be important for the mineral nutrition of the host plant (Wilcox, 1991). Apart from these well known beneficial effects on plant nutrition mycorrhizal associations may fulfil other functions for host plants growing on con- taminated land. Mycorrhizal fungi may alleviate metal toxicity to the host plant by acting as a barrier for metal uptake (Leyval et al., 1997). Recently, Sharples et al. (2000b) compared the short-term uptake kinetics of the ericoid mycor- rhizal fungus Hymenoscyphus ericae from an As/ Cu-contaminated mine site (As-resistant popula- tion) and from an uncontaminated natural heath- land (non-As-resistant population) in solution culture. Uptake kinetics of As V ,As III and phos- phate did not differ for resistant and non-resistant isolates. However, the mine-site fungi showed an approximately 90% enhanced efflux of As in the form of As III . Twenty-four-hours uptake of As V by hydroponically-grown mycorrhizal and non- mycorrhizal host C. vulgaris did not differ for mine-site plants. In contrast, inoculated heathland C. vulgaris accumulated 100% more As than non- inoculated individuals (Sharples et al., 2000a). The authors suggested that the mine site fungus acts as an As filter to maintain low As concentrations in plant tissues, while improving P nutrition of the host plant. Therefore, mycorrhizal fungi may be important for the revegetation/phytostabilisation of As-polluted sites. Arsenic tolerance of H. lanatus populations from non-contaminated sites was found to be polymorphic (Meharg and Macnair, 1992a,b). Meharg et al. (1994) investigated 50 tussocks from the same population of which 40% showed tolerance to As. The As-tolerant phenotype had a 11% higher P-status and a 34% higher arbuscular mycorrhizal (AM)-infection rate of roots. Wright et al. (2000) conducted a field experiment using clones of tolerant and non-tolerant H. lanatus populations. Though no difference in AM mycor- rhization could be observed, tolerant plants accu- mulated more P in shoots, had largerer shoot and root biomass and produced considerably more flower panicles. Results of Meharg et al. (1994), Wright et al. (2000) show that conclusions drawn from studies on uptake kinetics in solution culture may have limited validity in more complex field conditions. Most ferns normally exhibit mycorrhizal asso- ciations (Jones, 1987). The role of mycorrhiza in As hyperaccumulation is not known yet. We found that P. vittata individuals grown in pots were colonised by MA fungi. Most well-studied hyper- accumulators belong to the Brassicaceae (Brooks, 1998; Baker et al., 2000), which generally do not form mycorrhizal associations (Marschner, 1995). To our knowledge no studies on the role of any type of mycorrhizal symbiosis in hyperaccumula- tion has been carried out so far. From present W.J. Fitz, W.W. Wenzel / Journal of Biotechnology 99 (2002) 259 Á/278268 [...]... mineral phases in the rhizosphere has been reported yet 2.10 Conceptual model of the fate of As in the soil Á /rhizosphere Á /plant system Based on the review of available literature and theoretical considerations presented in the previous sections, we propose a conceptual model of the fate of As in the soil Á /rhizosphere Á /plant system including As uptake, chemical speciation in soil solution and interactions... fate of As in the soil Á /rhizosphere Á /plant system This model highlights key processes and their interactions and may provide some guidance in future research on the fate of As in the rhizosphere of terrestrial plants Arsenic in soils is largely associated with Fe oxides/hydroxides The pH (H') in the rhizosphere may differ up to two units from that in bulk soil Under aerobic conditions As is mainly present... as AsV in soil solution which is desorbed from adsorption sites upon pH increase Both, plantinduced reductions of the redox potential (e () and drastic pH decreases in the rhizosphere may dissolve Fe oxides/hydroxides, resulting in concomitant release of As and P into the soil solution Redox potential and pH control the prevailing redox and hydrolysis species of As in solution Reduction of AsV to AsIII... translocate pollutants to the harvestable parts Phytoextraction can be divided in continuous phytoextraction (using hyperaccumulator plants) and induced phytoextraction (chemically induced accumulation of metals to crop plants) Induced phytoextraction has not yet been applied to As This technique potentially threatens deeper soil layers and ground waters by the artificially induced mobilisation of... examine rates and quality of root exudates potentially involved in As hyperaccumulation and/ or As tolerance; investigate the role of mycorrhizas in As tolerance and Á/hyperaccumulation; investigate the role of soil microbes on AsV/ AsIII transformations in the rhizosphere; develop rhizosphere management technologies (e.g based on knowledge obtained in rhizobox studies) in order to facilitate and improve... with the soil solid phase and plant nutrition of P and Fe (Fig 2) The Fig 2 Conceptual model of As in the soil Á /rhizosphere Á /plant system 270 W.J Fitz, W.W Wenzel / Journal of Biotechnology 99 (2002) 259 Á/278 individual processes depicted in the model as fluxes /transformations (solid lines) and influences/interactions (dotted lines) are well established, but yet have not been applied to describe the. .. heavy metal contamination of soil and vegetation around a copper mine in Northern Peru Sci Total Environ 203, 83 Á/91 Benson, N.R., 1953 Effect of season, phosphate and acidity on plant growth in arsenic- toxic soils Soil Sci 76, 215 Á/224 Benson, L.M., Porter, E.K., Peterson, P.J., 1981 Arsenic accumulation and genotypic variation in plants on arsenical mine wastes in SW England J Plant Nutr 3, 655... carboxylic ¨ acids and protons in phosphorous-deficient plants Plant Soil 211, 121 Á/130 Nye, P.H., 1977 The rate limiting step in plant- nutrient absorption from soil Soil Sci 123, 292 Á/297 O’Neill, P., 1995 Arsenic In: Alloway, B.J (Ed.), Heavy Metals in Soils Blackie Academic and Professionals, London, UK Onken, B.M., Hossner, L.R., 1995 Plant uptake and determination of arsenic species in soil solution... multiple tolerance to various metals and As include A capillaris (Porter and Peterson, 1975; Benson et al., 1981; Symeonidis et al., 1985), D cespitosa (Cox and Hutchinson, 1980, 1981) and S vulgaris (Paliouris and Hutchinson, 1991) In contrast to phytoextraction, plants are required that take up only small amounts of As and other metals in order to prevent transfer into the wild-life food chain Sharples... may therefore affect the fate of As in the rhizosphere which has not yet been explored 269 2.9 Precipitation phenomena in the rhizosphere Water flow, ion transport by convection and diffusion processes, plant uptake, changes of pH and redox potential, root exudation, etc alter the chemical composition of the soil Á/root interface and may result in precipitation phenomena, favouring pollutant immobilisation . directly investigate the fate of As in the rhizosphere we highlight in the following the major processes taking place in the rhizosphere to assess the potential. As in the soil Á /rhizosphere /plant system including As uptake, chemical speciation in soil solution and interactions with the soil solid phase and plant