Iron mineralogy and uranium-binding environment in the rhizosphere of a wetland soil
Graphical abstract
Introduction
Wetland environments are effective at mitigating the migration of many groundwater contaminants because of their unique combination of geochemistry, microbiology, and hydrology (Grybos et al., 2007, Kennish, 2002). These ecosystems promote the decay of plant litter to form soil organic matter (OM). The resulting OM has several direct and indirect properties that enhance the wetland soil's capacity to bind contaminants, including increasing microbial activity, lowering the oxidation state, and creating more surface binding sites (Bowden, 1987, Johnston, 1991, Schlesinger, 2009, Sposito, 2011). These conditions often lead to groundwater metals (Frohne et al., 2014, O'Geen et al., 2010) and radionuclides (Kaplan et al., 2013, Wang et al., 2014, Wang et al., 2013) becoming immobilized in the wetlands.
Contaminant U has been reported in wetlands associated with U mining operations (Wang et al., 2014, Wang et al., 2013) and nuclear material production (Bertsch et al., 1994, Chang et al., 2014, Sowder et al., 2005). Uranium is more mobile in its oxidized state, U(VI), which commonly forms strong complexes with carbonates and hydroxides, and is relatively soluble (Langmuir, 1978). Uranium in far less mobile in the reduced state, U(IV), which is appreciably less soluble than U(VI) and binds strongly to mineral and organic matter. Uranium(VI) can be readily reduced to U(IV) by dissolved and solid forms of Fe(II), Mn(II), and sulfides, and by OM and several naturally occurring microbes(Lovley, 1997, Wall and Krumholz, 2006).
Roots have long been known to alter soil microbiology, hydrology, and chemistry (Bronick and Lal, 2005). This is especially pronounced in wetlands where plants not only create steep pH, plant nutrient, and organic C gradients, but they also create dissolved O2 gradients. Downward transport of oxygen through the plant into the soil is a physiological adaptation of wetland plants to enable them to grow in water-logged, reducing conditions (Bacha and Hossner, 1977, Chen et al., 1980, Mendelssohn and Postek, 1982). The oxygen introduced to the carbon-rich rhizosphere, the root-impacted soil zone, promotes the abiotic and biotic oxidation of dissolved Fe(II) to form Fe(III)-oxyhydroxide precipitates on the root surface, referred to as plaques (Gibberd et al., 2001, Thomas et al., 2005). Plaques can include several Fe-oxyhydroxides, including ferrihydrite, lepidocrocite, goethite, Fe-hydroxide, and siderite (Bacha and Hossner, 1977, Chen et al., 1980, Hansel et al., 2001, Taylor et al., 1984, Wang and Peverly, 1999). These mineral formations have been reported to be associated with several different plant species, including indigenous wetland plants, agricultural plants, and trees (Crowder and Macfie, 1986, Green and Etherington, 1977, Menhelssohn, 1993, Otte et al., 1989). The presence of plaques is largely controlled by plant type and the hydrological cycle and they can act as barriers to phytotoxic metals (Hansel et al., 2001, Otte et al., 1989, St-Cyr and Crowder, 1990). It is not clear how the plaque Fe-oxyhydroxides differ from those in the bulk soil. The high organic carbon and redox conditions in plaques are favorable for many Fe-reducing and Fe-oxidizing bacteria (Emerson et al., 1999, King and Garey, 1999). Biogenic Fe(II)/Fe(III) cycling may be an important process affecting other redox-active contaminants, such as U. Above pH 5.4, Fe(III)-U(IV) precipitates in bioreduced soils have been detected by X-ray absorption fine structure (XAFS) (Kelly et al., 2008). Furthermore, phosphate has been reported to enhance U sorption on Fe(III)-oxides (Singh et al., 2012, Singh et al., 2010). U(VI) species initially associated with carbon- and phosphorus-containing ligands were transformed to U(IV) associated with Fe and uraninite (Kelly et al., 2010).
Uranium migration has been significantly attenuated in the Tims Branch wetland on the Savannah River Site (SRS) in South Carolina, a nuclear processing facility (Evans et al., 1992). Of the 44 metric tons of depleted U introduced between 1954 and 1985 into the Tims Branch system, 70% remains in the wetland. Based on XAFS and sequential extraction characterization, most of the wetland U exists in association with OM (Bertsch et al., 1994, Li et al., 2015, Sowder et al., 2005). Furthermore, while this site is moderately acidic (pH ~ 5.5) and has microbes that have the potential to promote U reduction (including Geobacter spp. and sulfur reducing bacteria) (Turick et al., 2008) the U at the site exists almost exclusively as U(VI) (Bertsch et al., 1994, Li et al., 2015). Greenhouse mesocosm studies, simulating Tims Branch conditions, showed that U concentrations near the root were 3 × greater than these in root-free soils (Chang et al., 2014, Jaffé et al., 2014, Li et al., 2015). Furthermore, the rhizosphere had a distinct color difference compared to the bulk soil; it was brick-red, whereas the bulk soil was either brown or yellowish white. The red coloration originated from Fe(III)-mineral formation. It was also shown that the composition of OM differed greatly between the rhizosphere and non-rhizosphere; the rhizosphere OM molecules generally had greater overall molecular weights, less aromaticity and a greater hydrophilic character (Kaplan et al., 2016).
The objective of this study was to build upon results from mesocosm studies (Chang et al., 2014, James and Rubin, 1986, Kim et al., 2003) and to conduct a field investigation to determine how rhizosphere and non-rhizosphere soil differ in terms of mineralogy and geochemistry that might influence uranium binding. Our hypothesis was that wetland plant roots contribute OM and release O2 within the rhizosphere that promote the formation of Fe(III)-(oxyhydr)oxides. In turn, these Fe(III)-(oxyhydr)oxides stabilize organic matter that together contribute to contaminant immobilization. The general approach was to collect soils containing roots from the Tims Branch wetland and characterize subsamples designated as near (rhizosphere) and far (non-rhizosphere) from the roots. The samples were characterized using wet chemistry, and various types of spectroscopy and microscopy. Additionally, soil porewater samples were collected from depth-discrete diffusion samplers to provide information about the aqueous chemical conditions at the study site.
Section snippets
Study site
Tims Branch is a second-order stream located on the SRS and it drains an area of approximately 16 km2 before entering the Savannah River (Fig. 1). Tims Branch passes through M-Area, a former target and fuel processing facility for nuclear materials (Pickett, 1990). As a result of operations between 1954 and 1985, 44 tons of U was released into the Tims Branch system, which accounts for 97% of the gross alpha activity released by SRS. As a remnant of former agriculture activities in the area, Tims
Results and discussion
The characterization techniques used in this study probed different scales. The diffusion samplers probed tens of grams of (liquid) sample. The bulk soil chemistry (pH, XRD, CEC, Total Fe and Mn, DCB Fe and Mn, and U) probed grams of sample. Mössbauer spectroscopy probed bulk Fe mineralogy on milligrams of sample and the various microscopy techniques probed fractions of a milligram of sample (individual or groups of particles). As such, it was important to keep track of these various scales
Conclusions
Uranium in a contaminated wetland soil existed primarily in a dithionite-extractable fraction (> 79 wt.%), likely in associating with soil coatings. A very small, but important fraction (10-3 wt.%) of the U existed in the porewater at sub-μM concentrations. Porewater U concentrations in the wetland generally varied in a systematic manner in the centimeter scale with other aqueous porewater constituents (e.g. pH, Eh, Mn, Fe, and DOC). The μ-XANES analysis of the bulk soil in this wetland system
Acknowledgements
This work was supported by the Subsurface Biogeochemistry Research Program within the Office of Biological and Environmental Research (OBER), Office of Science, U.S. Department of Energy, Grants DR-FG02-08ER64567 and ER65222-1038426-0017532. Mössbauer spectroscopy, XCT, TEM, He-IM, and SEM/EDS were conducted at EMSL, a national scientific user facility sponsored by DOE's OBER program. EMSL is located at the PNNL in Richland, WA, USA. Use of Advance Photon Source was supported by the U.S. DOE
References (60)
- et al.
The effect of pH and plaque on the uptake of Cu and Mn in Phragmites australis (Cav.) Trin ex. Steudel
Ann. Bot.
(2000) - et al.
Soil structure and management: a review
Geoderma
(2005) - et al.
Role of fungi in the biogeochemical fate of depleted uranium
Curr. Biol.
(2008) - et al.
Waterlogging tolerance among a diverse range of Tifolium accessions is related to root porosity, lateral root formation and aerotropic rooting
Ann. Bot.
(2001) - et al.
Is trace metal release in wetland soils controlled by organic matter mobility or Fe-oxyhydroxides reduction?
J. Colloid Interface Sci.
(2007) Uranium solution-mineral equilibria at low temperatures with applications to sedimentary ore deposits
Geochim. Cosmochim. Acta
(1978)Microbial Fe (III) reduction in subsurface environments
FEMS Microbiol. Rev.
(1997)- et al.
Simplified behaviors from increased heterogeneity: I. 2-D uranium transport experiments at the decimeter scale
J. Contam. Hydrol.
(2013) - et al.
Use of phosphate to avoid uranium toxicity in Arabidopsis thaliana leads to alterations of morphological and physiological responses regulated by phosphate availability
Environ. Exp. Bot.
(2009) - et al.
Mitigating nonpoint source pollution in agriculture with constructed and restored wetlands