Environ. Sci. Technol. 2008, 42, 7146–7151
Metal Solubility and Speciation in the Rhizosphere of Lupinus albus Cluster Roots ´ ,† J. DESSUREAULT-ROMPRE B . N O W A C K , * ,†,‡ R . S C H U L I N , † M.-L. TERCIER-WAEBER,§ AND J. LUSTER| Institute of Terrestrial Ecosystems (ITES), ETH Zurich, Switzerland, Empa - Swiss Federal Laboratories for Materials Testing and Research, St. Gallen, Switzerland, CABE, Department of Inorganic and Analytical Chemistry, Sciences II, University of Geneva, Switzerland, and Swiss Federal Institute for Forest, Snow, and Landscape Research (WSL), Birmensdorf, Switzerland
Received January 17, 2008. Revised manuscript received June 5, 2008. Accepted June 12, 2008.
The objective of this study was to investigate the influence of root exudation of organic acid anions on the speciation of major and trace metal cations in the rhizosphere of Lupinus albus cluster roots. Plants were grown in rhizoboxes containing repacked weakly acidic loam. Bulk soil solutions and, during the lifetime of cluster roots, rhizosphere solutions were collected using micro suction cups. During organic acid anion exudation bursts, metals in the rhizosphere of cluster roots were strongly mobilized. The concentrations of dissolved organic carbon derived from soil organic matter increased parallel to organic acid anions. Speciation calculations revealed that, during exudation, Al, Ca, Mn, and Zn in the cluster root rhizosphere were mainly bound with citrate, while Cu and Pb were always strongly bound to soil-derived dissolved organic matter. Our results indicate that cluster root exudation led on one hand to direct mobilization and complexation of metals like Al, Fe, and Zn by citrate and on the other hand to the mobilization of soil organic matter which complexes and solubilizes Cu and Pb.
Introduction Root exudates can alter the solubility of ions and molecules in the rhizosphere (1–3). The amount and composition of root exudates depends strongly on the plant’s nutritional status. Some species exude low-molecular-weight organic acid anions (OAAs) in response to a P deficiency or phytosiderophores in response to an Fe or Zn deficiency (2, 4). Exudates can increase the phytoavailability of nutrients (5) but also restrict the uptake of toxic metals by the formation of nontoxic metal-ligand complexes. Potential ligands in root exudates are OAAs, phosphate, or polysaccharides (6). Many OAAs, which occur as anions under a wide range of soil conditions, are able to dissolve manganese and iron oxides or aluminum, calcium, and iron phosphates and to liberate P for root uptake (1, 2, 5, 7). Furthermore, OAAs can * Corresponding author e-mail:
[email protected]. † ITES. ‡ Empa. § University of Geneva. | Swiss Federal Institute for Forest, Snow, and Landscape Research. 7146
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 19, 2008
displace phosphate and sulfate from mineral surfaces by anion exchange (1, 2, 7, 8). The reaction of OAAs with metals in soils depends not only on the complexation ability of the OAAs but also on their sorption/desorption reactions, and their microbial degradability. Therefore, the degree of metal mobilization depends strongly on the amount and type of exuded OAA and on the physicochemical and biological properties of the soil (2). The impact of root exudates on metal concentrations in the soil solution has been studied by soil extraction, adsorption experiments (2, 9–11), or pot experiments combined with soil extraction (12). Most of these studies have looked at the effects of OAAs on metal desorption from soil (2). Independent of the soil properties and the concentration of OAAs, they all observed an increased metal desorption in the presence of OAAs. Furthermore, changes in rhizosphere chemistry during an exudation event are not only related to one specific OAA but to the overall exudation effect including enhanced microbial growth, as well as changes in pH and dissolved organic carbon (DOC) concentration. The effect of root exudation on the DOC concentration in the rhizosphere solution has not received much attention. Because of its strong affinity to polyvalent cations, DOC plays an important role in the biogeochemistry of cationic nutrients and pollutants in soils (13). On one hand, the exudation of organic compounds increases the DOC concentration in the rhizosphere directly. Up to 9 mM citrate was measured in the rhizosphere solution of Lupinus albus, which corresponds to 650 mg of C L-1 (14). In a soil solution from the root zone of willow, an OAA (malate and oxalate) contribution to DOC of 45-51% was measured (15). On the other hand, OAAs can also increase DOC derived from soil organic matter (SOM), either indirectly via an increase of microbial activity and related SOM degradation or directly by anion exchange or complexation of Ca, which stabilizes SOM (16, 17). The objective of this study was to investigate in situ the impact of the root exudation of OAAs by Lupinus albus on the concentration and speciation of major nutrient and trace metal cations in the rhizosphere of an unpolluted soil. For this purpose, we combined plant growth in rhizoboxes with the spatially highly resolved collection of soil solution using micro suction cups (18). Lupinus albus was chosen as a model plant well-known to exude high amounts of citrate from specialized roots, the so-called cluster roots (19). The potential impact of OAA exudation on metal speciation in solution was estimated by model calculations.
Materials and Methods Rhizobox System. The rhizobox used in this study was described in detail previously (18). A carbonate-free soil [pH 6.4 (0.01 M CaCl2), 15.1 g/kg Corg, 1.5 g/kg Ntot, 49 mg/kg Pavailable (0.5 M NaHCO3, pH 8.5), 862 mg/kg Porg, 36% sand, 49% silt, 15% clay] was air-dried, sieved (2 mm), and filled into the rhizoboxes at a bulk density of about 1.2 g/cm3. In order to equilibrate the soil, the rhizoboxes were flushed with 1 L of synthetic rainwater (for composition, see below) every week for 6 weeks. Seeds of Lupinus albus (“Weissblu ¨ hende Tellerlupine” cultivar, Ufa AG, Switzerland) were pretreated with 10% hydrogen peroxide (20) and then germinated in black garden soil for one week. Healthy plantlets were gently washed with deionized water to remove the soil and then transplanted into the rhizoboxes. Three rhizoboxes were planted each with a single plant. The experiment was conducted under controlled conditions in a climate chamber (light 16 h per day with an intensity 10.1021/es800167g CCC: $40.75
2008 American Chemical Society
Published on Web 09/03/2008
at canopy height of 150 µmoles m-2 s-1, 80% humidity, temperature day/night: 20/16 °C). The boxes were irrigated with synthetic rainwater (ionic composition in µmoles L-1: 70 NH4, 70 NO3, 3.2 PO4, 17 Cl, 3.1 SO4, 4.3 Na, 7.7 K, 5 Ca, 1.3 Mg, 0.15 Zn; pH ) 5.5) using wicks made of polyether sulfone (Rhizon irrigators, Rhizosphere Research Products, Netherlands) and installed at 5, 30, and 55 cm soil depth. A hanging water column of 40 cm was maintained between each wick and a corresponding reservoir in order to establish an approximately constant matric potential of -40 hPa throughout the rhizobox. Soil Solution Sampling. Samples were collected through the transparent front plate of the rhizoboxes as described previously (18). The rhizosphere solution sampling around cluster roots began with the emergence of rootlets between 4 and 7 weeks after sowing and was continued for a period of 7-10 days. Samples were collected each day during three periods (6:00-14:00, 14:00-22:00, and 22:00-6:00) from different positions around the cluster roots (one suction cup for each period; distance < 1 mm to tips of rootlets), which allowed for a sampling-free time at each position of 16 h to reequilibrate the soil (21). For details on the sampling devices and procedures, we refer to refs 14 and 18. In total, 11 cluster roots were sampled. In addition, the soil solution was sampled at 15 bulk soil locations (>2 cm from the nearest root). The average sample volume in 8 h was about 300 µL for rhizosphere and 600 µL for bulk soil solution samples. The influence of the rather weakly defined sampling zone on the results is discussed extensively in ref 14. In summary, measured concentrations should not be considered to quantitatively picture the situation at the place of sampling. They rather represent average conditions within the zone of influence, which is affected by sampling volume, applied vacuum, soil pore size distribution, and directional preferences due to gravitational and transpirational fluxes. All samples were analyzed for OAAs. Some bulk soil solution and the rhizosphere solution samples from four selected cluster roots (14:00-22:00 sampling period, during which the daily maxima of OAA exudation occurred (14)), were, in addition, analyzed for total metal concentrations. After OAA analysis, the samples from the rhizosphere solutions of the other cluster roots were pooled to get samples with different citrate concentrations. The criterion for pooling the samples was to obtain samples that represented the whole range of citrate concentrations found in the rhizosphere solutions from high concentrations during the exudative burst to low concentrations before or after. Five pools were made with citrate concentrations of 0.1, 569, 1147, 2378, and 9009 µmoles L-1. These pools and a pool of bulk soil solution samples were analyzed for total and dynamic metal concentrations as well as for UV absorption and total soluble phenolics. In addition, speciation calculations were performed for these pooled samples. Soil Solution Analysis. OAAs (acetate, citrate, formate, lactate, malate, oxalate, and propionate) and inorganic anions (nitrate, sulfate, and phosphate) were analyzed by ion chromatography (Dionex autosampler system; AS 11 column; eluent generator, potassium hydroxide (1 to 60 mM); flow, 1.5 mL min-1) with 200 µL insert glass vials to reduce the sample volume. Because the samples contained formaldehyde to prevent microbial degradation, the DOC concentration was estimated from the UV absorbance at 254 nm (Varian Cary 50 spectrometer). A calibration curve of DOC (total organic carbon analyzer, Shimadzu TOC-V) versus UV was established using an extract of the same soil with synthetic rainwater (for composition, see above). This soil extract contained negligible amounts of OAAs. The UV absorbance of Fe-citrate solutions (0.2-1.0 mmol L-1 Fe) was used to correct for the contribution of Fe-citrate complexes to the UV absorbance. In this paper, the reported DOC is referred
to as DOCUV. This DOCUV is considered mainly soil-derived; that is, it does not include the contribution of root exudation. The reasoning for this is given in the discussion section, and we considered additionally results from (i) the UV absorbance of the supernatants of a batch experiment in which the soil, in the absence of plants, was treated with increasing concentrations of citrate, and (ii) the analysis of total soluble phenolics in the same samples and in the pooled rhizosphere solutions using the method described in ref 22. Total metal concentrations were analyzed in 10-times diluted samples acidified to pH 2 with suprapur nitric acid by using micro-inductively coupled plasma mass spectrometry (ICP-MS; Mg, Al, K, Ca, Mn, Cu, Zn, Cd, and Pb) and inductively coupled plasma-optical emission spectroscopy (ICP-OES; Ca and Fe). The laboratory precision for the ICP measurements was “< 10 %” or “< detection limit” (for values “< 10 × detection limit”). The detection limits of ICP-MS were 210, 0.7, 30, 620, 0.04, 0.08, 0.5, 0.002, and 0.002 µmol L-1 for Mg, Al, K, Ca, Mn, Cu, Zn, Cd, and Pb, respectively, in the undiluted samples. The reason for the rather high detection limit for Ca is that, because of Ar interference on mass 40, an isotopic line with a low abundance has to be used for this element. Detection limits of ICP-OES were 2.5 and 0.4 µmol L-1 for Ca and Fe, respectively. pH was measured using an ion-sensitive field effect transistor electrode (ISFET sensor, Sentron, The Netherlands). Square wave anodic stripping voltammetry (SWASV) using gel-integrated microelectrode (GIME) arrays was used for simultaneous measurements of Cu, Zn, Cd, and Pb in pooled samples. The GIME arrays consist of 5 × 20 interconnected Hg-plated Ir-based microdiscs (average radius of each Hghemisphere typically 5.6 µm), covered by a 300-µm-thick LGL Agarose gel (23). Thanks to the characteristics of microelectrodes and of the gel membrane (24), GIME-SWASV measurements in environmental samples at their natural pH allow sensitive (ppt level) and reliable detection of the dynamic fraction of Cu(II), Zn(II), Cd(II), and Pb(II), which is defined as the sum of the free metal ions and small labile and mobile complexes with a size of a few nanometers (for details, see ref 25), without chemical and physical interference from the media (24). The dynamic fraction corresponds to the maximum fraction of metal that is potentially bioavailable (26). Electrochemical measurements were performed using a computer-controlled Amel 433 potentiostat (Milan, Italy), coupled to a preamplifier, and a 5 mL Metrohm cell in a three-electrode configuration: a Metrohm Ag/AgCl/KCl(sat) reference electrode incorporated in an additional bridge of 0.1 M NaNO3 suprapur to avoid contamination of the test solution by metals present in the reference KCl(sat) electrolyte; a Metrohm platinum rod counter electrode; and the GIME working sensor described above. Before each set of analyses, Hg deposition on the Ir-microdisks was performed by applying a constant potential of -400 mV in a deoxygenated solution of 5 mM Hg(CH3COO)2 and 0.1 M HClO4 for 8 min (25). At the end of each day, Hg was reoxidized by scanning the potential linearly from -300 to +300 mV at 5mV/s in a deoxygenated 1 M KSCN solution. The SWASV conditions used for trace metal measurements in soil solution samples and for calibration were as follows: cleaning time ) 60 s; cleaning potential ) -100 mV; deposition potential: -1000 to -1200 mV; deposition time ) 0.5-3 min; frequency ) 50 Hz; wave amplitude ) 25 mV; step amplitude ) 8 mV. Calibrations, by successive standard additions of the target metal, were performed in a 0.1 M NaNO3 suprapur solution. The laboratory precision for the voltammetry measurements was
