Is rhizosphere remediation sufficient for sustainable revegetation of mine tailings?

Abstract

Background

Revegetation of mine tailings (fine-grained waste material) starts with the reconstruction of root zones, consisting of a rhizosphere horizon (mostly topsoil and/or amended tailings) and the support horizon beneath (i.e. equivalent to subsoil – mostly tailings), which must be physically and hydro-geochemically stable. This review aims to discuss key processes involved in the development of functional root zones within the context of direct revegetation of tailings and introduces a conceptual process of rehabilitating structure and function in the root zones based on a state transition model.

Scope

Field studies on the revegetation of tailings (from processing base metal ore and bauxite residues) are reviewed. Particular focus is given to tailings' properties that limit remediation effectiveness. Aspects of root zone reconstruction and vegetation responses are also discussed.

Conclusions

When reconstructing a root zone system, it is critical to restore physical structure and hydraulic functions across the whole root zone system. Only effective and holistically restored systems can control hydro-geochemical mobility of acutely and chronically toxic factors from the underlying horizon and maintain hydro-geochemical stability in the rhizosphere. Thereafter, soil biological capacity and ecological linkages (i.e. carbon and nutrient cycling) may be rehabilitated to integrate the root zones with revegetated plant communities into sustainable plant ecosystems. A conceptual framework of system transitions between the critical states of root zone development has been proposed. This will illustrate the rehabilitation process in root zone reconstruction and development for direct revegetation with sustainable plant communities. Sustainable phytostabilization of tailings requires the systematic consideration of hydro-geochemical interactions between the rhizosphere and the underlying supporting horizon. It further requires effective remediation strategies to develop hydro-geochemically stable and biologically functional root zones, which can facilitate the recovery of the microbial community and ecological linkages with revegetated plant communities.

INTRODUCTION

Revegetation of land destroyed by mining and mineral processing activities involves the full reconstruction of ecosystems, comparable to the earliest stages of primary succession following major land disturbances, but within a much shorter timeframe (Bradshaw, 1983, 1997; Walker and del Moral, 2003). Surface-mined lands such as open-cut mines of coal, bauxite, mineral sands and metals can be rehabilitated into desirable plant ecosystems using complex technical restoration approaches (Prach and Hobbs, 2008). These approaches range from native seed bank conservation, soil improvement and engineering (e.g. fertilizing, topsoil-overburden layering) to landscape design (e.g. slope, contour and catchment connectivity) specific to catchment requirements and local climatic conditions (Sutton and Dick, 1987; Barnhisel and Hower, 1997; Koch, 2007; Gravina et al., 2011). Successful rehabilitation with sustainable plant ecosystems has been reported at sand and bauxite mines (Koch, 2007; Gravina et al., 2011), but not on tailings (Manab and Maiti, 2007; Mendez and Maier, 2008). What makes phytostabilization of mine tailings more challenging are the intrinsic abiotic constraints present in the wastes, including a dysfunctional physical structure and hydraulic capacity, unstable geochemistry and toxic chemical conditions (Bell and Jones, 1987; Licsko et al., 1999; Osborne, 2000; Manab and Maiti, 2007; Mendez and Maier, 2008).

Tailings are residue wastes from processing ores (e.g. rock ores for metals such as Au, Cu, Ni, Pb, Zn, U) and industrial minerals (e.g. bauxite, coal), which contain unstable primary and secondary minerals rich in metals and metalloids (e.g. Al, As, Au, Cr, Cu, Ni, Pb, Zn, U), salts and unwanted gangue minerals (e.g. silicates, carbonates, oxides/hydroxides, sulphides, etc.) (Lottermoser, 2010). These residues of fine particle sizes are pumped in slurry form into purposely built facilities and deposited as sediments across hundreds to thousands of hectares of landscape at each mine, accumulatively occupying millions of hectares of land worldwide (Sutton and Dick, 1987; Gordon, 2002; Li, 2006; Mudd, 2007). From the exposed tailings surface, dispersion of fine particles rich in pollutants (e.g. metals, metalloids, radioactive elements) into the surrounding natural and populated environments can occur via aeolian pathways in arid/semi-arid regions and water-induced surface runoff and leaching in high rainfall regions (Azcue et al., 1995; Castro-Larragoitia et al., 1997; Mendez and Maier, 2008). A recent catastrophic spillage of bauxite residues (red mud) over 40 km² at Ajka (Hungary) has further highlighted long-term environmental risks of exposed and unstabilized tailings in the event of erratic weather patterns (Gelencsér et al., 2011; Ruyters et al., 2011). Phytostabilization using sustainable plant communities has been advocated as a long-term sustainable solution to the closure of tailings storage facilities (Mendez and Maier, 2008).

The lack of sufficient volumes of topsoil and/or inert spoils and overburden materials to create a deep soil profile, thus isolating the toxic tailings, has prompted increasing interest in the direct revegetation of tailings. Root zones have been directly reconstructed in tailings through remediating measures, for example deep tillage, the addition of lime or gypsum and organic matter (OM), and microbial inoculation (e.g. Bell and Jones, 1987; Mulligan et al., 2006; Wehr et al., 2006). However, cases of sustainable phytostabilization of tailings via a direct revegetation approach are not yet available in the literature. Past efforts have mostly concentrated on rhizosphere remediation in the top 20–30-cm layer, rather than examining remediation requirements of whole root zones, the development of the soil matrix and horizons, and the rates and trajectories of plant community development (Bell and Jones, 1987; Cleland, 1988; Roseby et al., 1998; Wehr et al., 2006; Mendez and Maier, 2008). In many cases, remediation benefits on growth conditions in the tailings are short-lived. They fail to effectively stimulate the development of soil structure and functions in the root zone, thus leading to poor plant survival and failure of newly established plant communities within a few years (Fig. 1) (e.g. Hunter, 1974; Bell and Jones, 1987; McNearny and Wheeler, 1995). In some cases, temporary (months to years) establishment of multiple plant species may be achieved, but these are far from sustainable, requiring continual inputs of remediating resources (e.g. OM, nutrients and water) to increase plant productivity. Furthermore, supplementary seeding/planting to improve species diversity of target plant communities is often required (Roseby et al., 1998; Gravina et al., 2004; Wehr et al., 2006).

Fig. 1.

A conceptual diagram illustrating three typical scenarios of reconstructed root zones in tailings: (1) rapid decline of the improved root zone conditions regardless of resource inputs, probably due to incorrect assessment of primary constraints and/or inappropriate treatments; (2) initially stable root zone conditions improved by resource inputs and treatments, but gradual decline of the conditions after the withdrawal of external resources (such as water and fertilizers); and (3) sustainable root zone improved by initial inputs of external resources, becoming infinitely self-sustainable even after the withdrawal of resource inputs.

The importance of soil formation in ecosystem reconstruction on mined land has long been recognized (Bradshaw, 1997; Walker and del Moral, 2003). Soil formation and development are the result of many physical, chemical and biological processes and interactions under local climatic conditions over long time periods (Matthews, 1992; Bradshaw, 1997). To achieve this within a much shorter time frame, effective and long-lasting remediation strategies must be adopted to reconstruct a stable and sustainable physical structure (e.g. macropores and water-stable aggregates). Key hydraulic and chemical processes (e.g. water infiltration, ion adsorption/desorption) and conditions (e.g. redox status, pH, soluble ions) need to be restored to foster and stimulate biological activities of functional microbes and roots in the root zone system.

Unfortunately, historical efforts in tailings revegetation have largely focused on the use of native and/or exotic plant species tolerant to well-known adverse factors (e.g. high salinity, acidity/alkalinity, high levels of metals and metalloids) in the tailings, including halophytes and metallophytes (Bell et al., 1993; Clemens et al., 2002; Whiting et al., 2004; Baker et al., 2010). Lacking has been an holistic and systematic approach to link short-term remediation measures with rates and trajectories of changes in root zone structure and functions in the long term. Databases on remediation requirements for long-term soil development and formation in directly revegetated tailings (rather than deeply covered by inert spoils and top soil) remain limited and/or inaccessible. This is due to the lack of long-term (i.e. decades) field trials to monitor and evaluate changes in the reconstructed root zones in relation to plant growth responses and community performance (Gravina et al., 2004; Manab and Maiti, 2007; Mendez and Maier, 2008), although some information is available (e.g. Roseby et al., 1998; Bendfeldt et al., 2001; Shukla et al., 2005).

The present review has been triggered by this knowledge gap linking short-term remediation measures to the long-term rate and trajectory of root zone development in directly revegetated tailings, rather than surface-mined land. It aims to illustrate the importance of hydro-geochemical stabilization across the whole root zone, reconstructed from remediated tailings, by considering interactions between the hydro-dynamically inter-linked rhizosphere and the underlying horizons. The development process of soil structure and functions in the root zones has been conceptually divided into short-term hydro-geochemical stabilization and the rehabilitation of soil biological capacity and ecological linkages in the intermediate to long term (Fig. 2). Two critical transition phases have been proposed to set conceptual goals of remediation measures for manipulating the rate and trajectory of system changes in the root zones: (1) hydro-geochemical stabilization induced by effective remediation targeting mineral and geochemical characteristics in tailings and their dynamic interactions with local climatic conditions (e.g. temperature, evaporative water deficit), and (2) the consequent rehabilitation of soil biological capacity characterized by the succession of a microbial community structure from autotrophic to heterotrophic bacteria (including rhizobia, actinomycetes) and fungi, and biological functions from decomposing exogenous OM added in tailings to litters from plant communities in situ. Dynamic and responsive ecological linkages (i.e. carbon and nutrient cycling) connect the reconstructed root zones with target plant communities, thus sustaining rehabilitated plant ecosystems as a whole (Wardle et al., 2004). Finally, a conceptual model of root zone system transition has been developed for the holistic and systematic evaluation of the developmental process of reconstructed root zone systems towards a stable and sustainable state. From here, future studies may develop indicators and criteria to evaluate the development of the root zone system at each phase.

Fig. 2.

A conceptual diagram illustrating the interactive relationship between reconstructed root zone and the immediately adjacent supporting zone: (1) improved conditions in the root zone due to amendments and inputs of resources (such as irrigation and nutrients); and (2) deteriorating impacts of mobile constraint factors in the support layer on improved conditions in the rhizosphere layer, due to their mobility via capillary rise (soluble salts, alkalinity, acidity, etc.).

LIMITATION OF RHIZOSPHERE REMEDIATION AND IMPORTANCE OF REHABILITATING WHOLE ROOT ZONE SYSTEMS

Many past studies have shown that rhizosphere remediation of tailings for direct revegetation is not sufficient to support plant establishment and development of desired plant communities, due to the rapid deterioration of growth conditions initially improved by remediation (see selected examples in Table 1), such as increased compaction, pH decline or rise, salinization and nutrient depletion. In semi-arid regions, tailings remediation via surface tillage and amendments did not support the initial establishment of plant species without ongoing irrigation (Hunter, 1974; Ison, 1976). This was a result of high evaporative loss of water and enrichment of soluble salts in the rhizosphere from chemical weathering of minerals in the tailings beneath the rhizosphere (Wehr et al., 2006; Smuda et al., 2008). This ineffective remediation strategy to stabilize hydro-geochemical dynamics in the tailings resulted in a rapid failure of the revegetation efforts [Fig. 1(1)].

Table 1.

Field examples of tailings phytostabilization of base metal mine and gold mine tailings and bauxite residue mud in field trials

Revegetation mode	Type of tailings	Remediation methods	Climate/plant species used	Performance of plant communities	References
Direct revegetation	Mt Isa Cu/Pb/Zn mine tailings	Shallow ripping (about 17 cm). Surface incorporation of fly ash, N/P-fertilizer, mulch	Semi-arid subtropical climate (400 mm rain per year, mostly in December–March), mixed grass species (sesbania, vetch, cereal rye, buffel grass, Rhode grass, couch, sudan grass, etc.)	Predominantly buffel, couch and Rhode grasses, low biomass, only short lived	Hunter, 1974; Ison, 1976
	Kidston Gold mine tailings	No soil cover. Tailings fertilized with NPK fertilizers and irrigated with drip-irrigation	Tropical with distinct wet–dry season (about 719 mm rain per year, mostly between November and April), trees (seedlings): Acacia, Eucalyptus, Casuarina, Melaleuca species, grasses (seeding): Cenchrus, Chloris, Cynodon, Echinochloa species, etc.	(1) Low survival and poor growth in acidic tailings without amendments (2) By 2005, 300 ha tailings successfully revegetated, but not yet reached a self-sustaining native plant community	Roseby et al., 1998; Mulligan et al., 2006
Capped revegetation	Mt Isa Cu/Pb/Zn mine tailings	Tailings ripped to 1 m, covered with 50 cm gravelly hillside materials, amended with sewage sludge	Mixed tree and grass species: Eucalyptus, Acacia, Triodia, Ptilotus, Atalay, Stylo, Cenhrus, Atriplex species, etc.	(1) Prodcued best and most consistent vegetative growth; natural colonization of Kopok, Ptilotus	Hodge et al., 1997
				(2) Good establishment of Acacia species and Cenchrus
				(3) Predominantly buffel (Cenchrus) after 8 years
	Bauxite residue mud	Capping profile: topsoil or subsoil over a layer of capillary break (e.g. residue sand or low-grade bauxite)	Tropical monsoonal with wet–dry season, about 1330 mm annual rainfall, native Acacia and Eucalyptus, grasses (Chloris gayana, Cynodon dactylon, Sporobolus virginicus, etc.)	Varying success in different areas, with decreasing diversity of species and variable biomass cover	Wehr et al., 2005, 2006

The combination of topsoil placement (20–30 cm), deep tillage (1 m) and addition of sewage sludge led to initial establishment of diverse native and exotic plant species in Mt Isa Mine tailings in Queensland, Australia. However, species diversity declined rapidly within a couple of years, resulting in the dominance of buffel grass (Cenchrus ciliaris), which is highly competitive in water acquisition (Hodge et al., 1997). This decline was perhaps due to the short-lived effectiveness of initial remediation measures for reconstructing root zones, such as re-emergence of compaction over time, discontinuity between the rhizosphere layer (topsoil) and supporting zone (tilled tailings), and inadequate hydraulic capacity (water infiltration and storage in depth), which was further exacerbated by the aggressive competition of buffel grass for water [Fig. 1(2)]. In long-term field trials at Kidston gold mine, Queensland, Australia, tailings under tropical monsoonal climatic conditions (719 mm annual rain mostly in November–April), direct revegetation by seeding and planting tube-stocks of native plant species in the tailings has shown some success in the establishment phase, with the aid of fertilization and irrigation (Mulligan et al., 2006; Roseby et al., 1998). Here also, the established plant communities were far from sustainable after many years (Mulligan et al., 2006). This rehabiliation effort could be categorized as the temporary establishment of plant communities of target species [Fig. 1(2)]. The development of root zones and limiting factors hindering plant community development have yet to be investigated in these studies. To phytostabilize highly alkaline bauxite residue mud at the Gove refinery Northern Territory, Australia, soil horizons were engineered to include a capillary break layer consisting of residue sand (coarse fraction) and low-grade bauxite between the rhizosphere layer and the tailings. This minimized the movement and enrichment of alkalinity and salinity from the tailings layer into the rhizosphere, which resulted in stable pH (<8·2) and EC (<0·2 dS m⁻¹) conditions in the rhizosphere zone (Wehr and Menzies, 2005; Wehr et al., 2006). Nevertheless, these engineered soil horizons did not develop sufficient hydraulic capacity and processes to respond to the wet–dry seasonal climatic patterns, leading to an unsustainable plant community structure during the prolonged dry season. As a result, different remediation strategies may result in different rates and trajectories of root zone development in physical, hydraulic and chemical conditions over time, depending on tailings characteristics and local climatic conditions. Climatic conditions may intimately modulate weathering of unstable minerals and thus the intensity of toxic chemicals in the root zone.

The above outcomes from the field trials are not surprising as plant roots require not only favourable growth conditions in the rhizosphere horizon, but also a stable and hydraulically functional horizon beneath the rhizosphere. In natural soil systems, the majority of biological activities and root growth occur in the A (+O) horizon and some roots are present in the B horizon and may penetrate into the C horizon (e.g. deep tap roots), depending on the profile characteristics, plant species and climate (Brady and Weil, 2002). Therefore, root zones are an integrated dynamic continuum, in which there is a continuous upward–downward flux of water and solutes across the A, B and, where possible, C horizons. As a result, growth conditions in the rhizosphere (i.e. equivalent to A horizon) are not only the result of remediation, but also of the ongoing hydro-geochemical dynamics in the supporting horizon (i.e. equivalent to B or B + C horizons) beneath the rhizosphere (Fig. 2). The hydro-geochemical dynamics and stability in reconstructed root zones in tailings are closely dependent on the characteristics of tailings mineralogy and geochemistry and local climatic conditions (e.g. rainfall intensity and distribution) (Dudka and Adriano, 1997; Manab and Maiti, 2007; Smuda et al., 2008; Hayes et al., 2009) (Fig. 3).

Fig. 3.

A conceptual model describing the process from understanding characteristics of tailings mineralogy and geochemistry, remediation interactions with tailings, to soil formation and development of functional processes, and the rehabilitation of ecological linkages with target plant communities. Reverse arrows indicate feedback interactions between processes.

As a result, hydraulic processes in reconstructed root zones should be carefully considered in relation to the mobility of adverse factors such as acidity, alkalinity and/or soluble salts in the subsoil layer underlying the rhizosphere, as this may render efforts of rhizosphere remediation futile (Fig. 2). Salt efflorescence at the tailings surface is a common phenomenon under semi-arid climatic conditions, due to high evaporative deficit and strong capillary suction (Dold and Fontbote, 2001). Large amounts of sulfates and cations were found to be enriched in the surface layer due to the fine particle sizes of tailings (silt–fine silt) and capillary rise of pore water in Cu–Pb–Zn tailings under semi-arid subtropical climatic conditions (Dold and Fontbote, 2001; Smuda et al., 2008). The movement of alkalinity and salinity from the tailings layer into the rhizosphere poses a great challenge to the long-term survival and development of plant communities established on bauxite residues (Wehr et al., 2005, 2006).

TAILINGS PROPERTIES, HYDRO-GEOCHEMCIAL DYNAMICS AND REMEDIATION REQUIREMENTS

Tailings, particularly those from processing metal ores (e.g. gold, base metals) and bauxite, are geochemically unstable when exposed to cool and oxygenated surfaces caused by water infiltration, due to the intrinsic properties of tailings particles, i.e. high surface area and sulfidic minerals (Dold and Fontbote, 2001; Smuda et al., 2008; Lottermoser, 2010; Jones and Haynes, 2011). As a result, soil development and formation processes are only possible after remediated tailings in root zones have gone through the first stage of hydro-geochemical stabilization. Specifically, following transition from the rapid reaction phase (e.g. oxidation, dissolution), which results in the development of high concentrations of secondary minerals and extreme pH conditions, into the slow reaction phase, in which impacts of products and conditions from chemical reactions has declined to a degree at which plant roots can tolerate and survive. The emphasis on hydro-geochemical dynamics arises from the necessity to rehabilitate the hydraulic capacity in root zones, while also understanding the critical roles of water and oxygen in the geochemcial reactions (or chemical weathering) of tailings minerals. The rate and trajectory of this initial hydro-geochemical stabilization are closely linked to several key factors: (1) the geochemistry of tailings minerals; (2) the physical and chemical interactions of tailings minerals with remediating materials (e.g. neutralizing agents) and practices (e.g. tillage, layering and mixing); and (3) local climatic conditions (e.g. temperature, evaporative intensity and rainfall patterns) (Fig. 3). In particular, impacts of local climatic conditions [e.g. rainfall amount and frequency (i.e. seasonal distribution)] on hydro-geochemical dynamics in root zones must be closely investigated when formulating remediation strategies relevant to long-term field performance. For example, rooting conditions could be substantially improved in Pb/Zn tailings to permit the natural colonization of diverse plant species within 3–16 years of decommissioning of tailings storage facilities under subtropical and monsoonal climate (1300–1500 mm rainfall per year) (Shu et al., 2005). However, under semi-arid climatic conditions, the time required for the natural improvement of root zone conditions is much longer (decades), allow some degree of freedom for natural colonization of tolerant native plant species (Conesa et al., 2007a; Mendez and Maier, 2008).

Influence of tailings mineralogy on hydro-geochemical dynamics and remediation requirements

Comprehensive reviews of the mineralogy, geochemistry and physical–chemical properties of tailings from various mineral processing are beyond the scope of the present discussion and have been provided by several authors (see Blowes and Jambor, 1990; Dold, 2003; Duker et al., 2005; Donato et al., 2007; Smuda et al., 2008; Lottermoser, 2010; Kossoff et al., 2011). Some typical physical and chemical constraints of deposited tailings are fine texture, high mechanical impedance, extreme pH conditions (e.g. pH < 4 in metal tailings, and pH > 10 in red mud), hyper-salinity (e.g. EC_1:5 > 15 mS cm⁻¹), and elevated concentrations of metals and metalloids (see examples in Table 2). Bulk analysis of geochemical and chemical properties may provide an initial assessment of likely chemical constraints to plant growth in tailings at the time of sampling, for example total and net acidification capacity, total and extractable pools of metals and metalloids by various acid leaching, and sequential fractionation with chemical agents (Dold, 2003; Conesa et al., 2008; Smuda et al., 2008; Álvarez-Valero et al., 2009). However, it is important to note that minerals in the tailings are never weathered uniformly and consistently in the fashion as artificially created in various chemical extractions and batch tests. It is fundamentally important to characterize the composition and assemblages of minerals, rates of oxidation, dissolution, precipitation and surface coating of mineral particles, in relation to hydraulic properties (e.g. infiltration and capillary suction), and to evaluate temporal changes of state of geochemical dynamics in the tailings, under seasonal events of local climatic conditions [Fig. 3(1)]. This information provides the basis for designing remediation measures to achieve long-term stability of root zone conditions for plant growth.

Table 2.

Some examples of primary constraint factors in mine tailings in relation to revegetation feasibility

Mine waste	Physical and hydraulic constraints	Chemical constraints	References
Base metal mine tailings	Fine particle size (e.g. mostly <100 µm), high bulk density and mechanical compaction, lack of aggregation and macropores, slow to very slow hydraulic conductivity and poor water infiltration	pH conditions: acidic (e.g. 2–3), near neutral (e.g. 6·5–7·5) or alkaline (>8) (depending on calcite/pyrite ratio)	Bell and Jones, 1987; Dold and Fontbote, 2001; Shu et al., 2005; Conesa et al., 2007a; Smuda et al., 2008; Huang et al., 2011b
		High levels of soluble salts (e.g. Mg, Na, sulfate, chloride)
		High levels of total concentrations of metals, metalloids, etc.
		High levels of soluble metals and metalloids depending on pH
		Low cation exchange capacity, low total and/or available nutrients (such as N, P, micronutrients at alkaline pH)
		Lack of organic matter
Gold mine – heap leach residue and tailings	Porous (heap leach residue, variable with tailings), dispersive (due to high sodicity)	High pH (>8·5)	Lottermoser, 2010
		High levels of cyanides (complexes with metals, such as Pb, Cu and Zn) in soluble and solid phases
		High levels of metals, metalloids
		High sodium levels and sodicity
		Deficiencies of micronutrients (Mn, Zn, Cu, etc.)
		Lack of organic matter
Bauxite residue mud	Small particle size (mainly silt and clay) and high bulk density, high compaction and low hydraulic conductivity, highly dispersive due to high sodicity.	High pH (10–12)	Wong and Ho, 1993; Courtney and Timpson, 2005b; Ippolito et al., 2005; Wehr and Menzies, 2005; Woodard et al., 2008; Harris, 2009
		High salinity (dominated by NaCl) and sodicity
		High levels of Na and low levels of exchangeable K
		Poor nitrogen retention capacity due to the high pH and low organic matter
		Low availability of micronutrients (e.g. Mn, Zn and B)

Hydro-geochemical dynamics in tailings are closely related to the presence of primary and secondary minerals from ores, gangue materials and the type of mineral processing. Their mineralogical and geochemical properties at the time of sampling are subject to further weathering events under local climatic conditions (Dudka and Adriano, 1997; Conesa et al., 2007a; Smuda et al., 2008; Lottermoser, 2010). For example, base metal ores, such as porphyry Cu–Mo, pyrite (FeS₂)/chalcopyrite (CuFeS₂) Cu, Fe oxide Cu–Au ores and carbonate-hosted Pb–Zn ores, commonly coexist with gangue minerals, such as feldspars (a group of silicates, KAlSi₃O₈–NaAlSi₃O₈–CaAl₂Si₂O₈), magnetite (Fe₃O₄), hematite (Fe₂O₃), galena (Zn–PbS) and carbonates (limestone, dolomite, etc.) (Dold, 2003; Hansen et al., 2005; Conesa et al., 2008; Smuda et al., 2008). In bauxite residues, the dissolution of sodalites (sodium aluminosilicates) via the Bayer reaction and subsequent neutralization is the major source of alkalinity and sodicity in reconstructed root zones (Zheng et al., 1998; Jones and Haynes, 2011). When exposed to cool and oxygenated surfaces under wet–dry cycles, these finely ground minerals and residual chemicals (cyanide, lime, caustic soda, etc.) from mineral processing trigger a series of complex reactions at variable rates. Such processes include oxidation (e.g. sulfides), acidification, neutralization, dissolution, sorption through precipitation and co-precipitation. All these processes are further affected by the co-presence of cations and anions (also from added chemicals and fertilizers) and local climatic conditions (e.g. temperature and rainfall) (Davis et al., 1993; O'Day et al., 1998; Pandey et al., 2007; Schippers et al., 2007; Schuwirth et al., 2007; Hayes et al., 2009). In general, silicate minerals are weathered into clay minerals and quartz, and sulfide minerals into hydrated oxides, sulfates, phosphates and carbonates, depending on the availability of anions and pH (Nabon, 1991). At the same time, the presence of acidophiles (autotrophic S- and Fe-oxidizing bacteria) in unsaturated zones of tailings enhances the weathering of sulfide minerals (pyrite, pyrrhotite, chalcopyrite, sphalerite, galena, etc.) (Tyson et al., 2004; Kock and Schippers, 2006; Schippers et al., 2010). As a result, the trajectory of hydro-geochemical dynamics in tailings determines the chemical conditions in the root zones (particularly, in pore water), such as acidity or alkalinity, which are fundamentally related to the mineralogy and geochemistry of minerals present in the tailings.

The pH condition in tailings is one of the major consequences of the complex and temporal hydro-geochemical dynamics of the minerals, and at the same time is one of dominant factors regulating the solubility of metals, metalloids and other ions in pore water, which are toxic to plant growth. Acidic and metallic drainage, resulting from oxidation (or weathering) of pyritic minerals (or sulfides), acidification and associated dissolution of metals and metalloids from primary minerals, is a common and priority problem in base metal mine tailings (O'Day et al., 1998; Dold, 2003; Willscher et al., 2007; Álvarez-Valero et al., 2009). This is dependent on the balance between pyrite oxidation rate and dissolution rate of carbonates and base cations (Dold and Fontbote, 2001, 2002). For example, the ratio of pyrite (or equivalent) to calcite (or equivalent) minerals in base metal mine tailings determines the long-term potential of acidification in the tailings (Dold and Fontbote, 2001, 2002). The relatively high content of sulfide (pyrite equivalent) to carbonate (calcite equivalent) in the oxidation zone of tailings was the primary cause for low pH (e.g. 2–3) (Dold and Fontbote, 2001). Acidification in tailings is unlikely to occur when the neutralization potential of carbonates greatly exceeds the acid generation potential of pyrite, for example as seen in the tailings of the Fe oxide Cu–Au mine Pedro A. Cerda in the Atacama desert of northern Chile where a ratio of 10 wt% calcite to <2·5 wt% pyrite was reported (Dold and Fontbote, 2002). As a result, stabilization and control of long-term pH changes in tailings are often one of the priority requirements in tailings remediation for root zone reconstruction.

The rate of hydro-geochemical stabilization is not only influenced by the co-presence of many minerals but also subject to the rate of hydro-geochemical reactions (or weathering) of various minerals in contact with water and oxygen and their spatial assemblages over time. Such examples include oxidation of sulfides, dissolution and precipitation of metals and metalloids, and sequestration of metals/metalloids by Fe/Mn oxides/hydrous oxides (Davis et al., 1993; Doye and Duchesne, 2005; Kossoff et al., 2011). An initial rapid oxidation of sulfides has been reported when fresh tailings were deposited in tailings storage facilities, resulting in a pH decline from 9–10 to 8·7–9·8 (Dold and Fontbote, 2002; Smuda et al., 2008). However, the oxidation rate may quickly slow due to the lack of oxygen in the pore space caused by sedimentation and compaction. The potential of rapid reactions may therefore be temporarily contained, but unleashed later when again exposed to water and oxygen. Contact between unstable minerals and water and oxygen can occur during natural seasonal wet–dry cycles and upon physical disturbance such as tillage and amendment. This may trigger a rapid oxidation of sulfide minerals, such as galena (PbS), sphalerite [(Cd,Zn)S] and chalcopyrite (CuFeS₂) into sulfates of different solubility. Metals released into the solution phase – in the presence of these anions – can then be further transformed into insoluble silicates, carbonates and/or phosphates and sequestrated by Fe/Mn oxides and hydrous oxides (Hudson-Edwards et al., 1995). These insoluble products may coat the sulfide mineral granules and create a barrier against the progression of oxidation towards their inner sphere (Weisener et al., 2011). This may further slow the oxidation of sulfide minerals, thus alleviating further pH reduction and the acute release of toxic metals into the solution phase. From the above evidence, it is important initially to carry out column leaching tests with remediation treatments, rather than just batch tests, to simulate hydro-geochemical processes and to characterize the rate of oxidation or chemical weathering of unstable minerals (e.g. rapid vs. slow weathering or oxidation) over time. This would provide useful information for designing temporal combinations of remediation measures and to stabilize the hydro-geochemical dynamics of tailings in root zones, for example through stimulated oxidation and leaching, induced precipitation, and coating and controlling oxidation.

The textural characteristics of tailings have profound impacts on the geochemical dynamics in the tailings profile, as they determine pore sizes, diffusion and infiltration of water and oxygen. The physical and hydraulic properties of tailings are less well described compared with chemical/geochemical characterization. Efficient grinding of ores and minerals is typical of modern mineral processing, leading to fine particle sizes (mostly very fine silt to silt) and highly consolidated sedimentation (Lottermoser, 2010; Jones and Haynes, 2011). In a survey of physical properties in recently deposited tailings from Mt Isa mines, it was found that the particle sizes were mostly distributed in the range 2–63 µm (fine silt to silt fraction), which were much finer than those from a decommissioned old tailings as a result of the implementation of technologies to increase processing efficiency (Forsyth et al., 2011; Huang et al., 2011a). Particle sizes of residue mud (or red mud) from a bauxite refinery are mostly <20 µm, with a small proportion (20–30 %) at <2 µm (Jones and Haynes, 2011). The consolidation of these fine particles causes high bulk density (e.g. base metal tailings 1·6–1·8 Mg m⁻³, red mud 2·8–3·3 Mg m⁻³) and compaction (>2·0 MPa), restricting root penetration (our unpublished data for Cu/Pb/Zn tailings at Mt Isa Mines; Woodard et al., 2008). Water infiltration into such compacted tailings can be slow or very slow, due to the uniformity of the particle size distribution and its maximum in the silt fraction (Wehr et al., 2005, 2006; Huang et al., 2011a). Only particle sizes associated with the clay fraction from weathered ore (e.g. bauxite) consist of clay minerals. Clay-sized particles originating from ground rock lack the structure of clay minerals and this may affect their swelling–shrinkage capacity, which may have a negative impact on the development of secondary pore systems.

The physical and hydraulic properties have profound implications in remediation requirements for rehabilitating hydraulic capacity (storage capacity) and functions (water infiltration and capillary rise), such as the creation and stabilization of macropores in the engineered soil matrix (i.e. remediated tailings) and horizons in the long term. The control of capillary rise of soluble salts requires the restoration/creation of aggregates and the formation of macropores in the root zone profile, which may be stimulated by effective amendment options, such as OM, gypsum and reactive Fe-oxides (Barral et al., 1998; Ippolito et al., 2005). Without addressing the long-term stability of the physical structure (e.g. the presence of an adequate proportion of stable macropores), it is unlikely to successfully rehabilitate the hydraulic capacity and processes required for water storage and gradual removal of toxic ions through leaching in reconstructed root zones. This is required not only for water supply to plants during a prolonged dry season, but also for preventing waterlogging in the root zone during torrential or extended rainfall (Woodard et al., 2008).

Physical properties can also be closely influenced by geochemical reactions, such as the formation of large amounts of Fe-oxides/hydroxides as result of oxidation in base metal tailings and high levels of sodium and hydroxysodalite in red mud (Forsyth et al., 2011; Jones and Haynes, 2011). Red mud particles are cemented together by hydroxysodalite at high pH, which can be broken up by lowering pH conditions to slightly acidic–neutral (Newson et al., 2006). Precipitation of iron oxides and oxyhydroxides has bonding effects in base metal mine tailings, creating massive barriers, which may restrict water infiltration without prior mechanical fracturing, but this effect can also be exploited to induce aggregation in remediated tailings (Duiker et al., 2003). Mineral dissolution and precipitation may change the sizes of water-filled pores and porosity in tailings, inducing marked changes of saturated and unsaturated hydraulic properties in remediated tailings (Wissmeier and Barry, 2009). As a result, short-term improvement in the physical structure can be modulated by ongoing geochemical reactions of minerals and interactions with remediation materials.

Influence of climatic factors on the distribution and mobility of soluble secondary minerals in the root zone profile

The stabilization of hydro-geochemical dynamics is closely influenced by local climatic conditions, particularly rainfall, in addition to mineralogy, geochemistry and texture [Fig. 3(1)]. In arid/semi-arid regions, secondary minerals containing metals and metalloids (sulfates, hydrous oxides, chlorides, etc) may accumulate in the tailings profile, due to the lack of frequent rainfall events and leaching effects (Blowes and Jambor, 1990; Jambor, 1994). When evaporative water deficit greatly exceeds the annual precipitation in semi-arid/arid regions, salt efflorescence may occur in tailings that are hydro-geochemically unstable, due to a net upward water flow via capillary forces, causing the enrichment of mineral salts in the surface layer of the tailings profile (Dold and Fontbote, 2001; our unpublished data). This up-/downward movement of secondary minerals through pore water flow may also occur during the prolonged dry season even when the total annual rainfall is high, highlighting the importance of rainfall distribution patterns on remediation and reconstruction of root zones in tailings (Wehr et al., 2005). In contrast, there is little accumulation, and even a decrease of soluble salts in the top layer of the tailings under extensive leaching effects from high annual rainfall (Davis et al., 1993; O'Day et al., 1998; Schuwirth et al., 2007). The stratification of a tailings chemical/geochemical profile may greatly alter and exacerbate physical/chemical conditions in the rhizosphere horizon, into which roots of revegetated plants are most likely to grow. This highlights the importance of understanding weather-induced heterogeneity in hydro-geochemical dynamics and distribution of chemical factors along the vertical profile. Great caution is required when extrapolating laboratory-based research results into field remediation strategies.

Remediation requirements for hydro-geochemical stabilization

Hydro-geochemical stabilization – the transition of the weathering process from the rapid into the slow reaction phase in hydrated and oxygenated environments – is the primary goal in tailings remediation when reconstructing root zones for direct revegetation (Fig. 3). This is critical to minimize the short-term failure of revegetated plant communities for tailings phytostabilization, due to acute toxicities of chemical factors in the root zones [Fig. 1(1)]. The pathways and timeframe to achieve this goal are closely dependent on the mineralogy, geochemistry and physical properties of tailings, the physical/chemical effects of remediation measures and local climatic conditions. When formulating remediation strategies, temporal combinations of stimulation and isolation measures may be adopted to first stimulate and exhaust the initial rapid weathering of geochemically unstable minerals (e.g. sulfides, sodalites) in the tailings. This avoids acute toxicities caused by rapid weathering of the minerals and thus avoids the failure of newly established plant communities. For example, the combination of deep tillage and incorporation of low-quality OM (e.g. wood chips, hay from native pasture) stimulates water infiltration and oxidation of sulfides, generating high concentrations of sulfates and chlorides and metals/metalloids in copper tailings (Fig. 4, data of metals/metalloids are not shown). The addition of liming (e.g. lime, dolomites) or acidifying materials (e.g. phosphor-gypsum, gypsum, sulfur-gypsum) in the tailings is determined by the potential and rate of acidity or alkalinity generated from hydro-geochemical reactions. With an increased understanding of the precipitation and coating phenomenon around particles of unstable minerals, isolation strategies may be formed by purposely incorporating materials rich in Ca, phosphorus and/or Fe (phosphate, Fe-oxides, limestone, etc.), to slow or contain the weathering intensity and associated toxicity to plants and soil microbes.

Fig. 4.

Responses of sulfate and chloride concentrations in pore water of Cu-tailings subject to organic matter remediation: 10 and 20 % hay (v/v). The hay was mixed thoroughly throughout the profile of about 100 cm in a large plastic container (1 × 1 × 1 m). The containers were situated on a tailings storage facility under field conditions (Huang et al., 2011a). The leachate solutions were collected by using ceramic suction cups placed horizontally at 15 and 80 cm below the surface under a maximum pressure of 400 Pa. The data were pooled from the results over a period of 12 months.

Neutralization of pH conditions is an important part of hydro-geochemical stabilization in reconstructed root zones as extreme pH conditions are common outcomes of the initial, rapid phase of hydro-geochemical dynamics in tailings. Very low or high pH is also a primary cause of the solubility and bioavailability of many metals, metalloids and salts in the pore water phase, such as in base metal tailings (Ye et al., 2000; Conesa et al., 2007b) and red mud (Fuller et al., 1982; Gräfe et al., 2011). Extensive liming and amendments with biosolids and woody debris raised soil pH from the original 5·1 to 11 (later settling to 7·7–8·4) and reduced heavy metal toxicity (Pb and Zn) to plants, leading to successful establishment of multiple native grass species in a field trial on metal mine tailings (Brown et al., 2007). The extremely high pH in red mud, a dominant factor directly causing physiological damage to roots, requires the incorporation of gypsum not only for lowering the pH, but also for countering the effects of high sodicity on particle depressiveness (Ippolito et al., 2005; Xenidis et al., 2005; Wehr et al., 2006; Woodard et al., 2008).

Physical manipulation of tailings through deep tillage and/or incorporation of coarse materials are integral steps in a tailings remediation and root zone reconstruction strategy. This amelioration of compaction triggers a whole suite of changes in root zone properties and processes, such as water infiltration, oxidation reactions, formation and dissolution of salts, and leaching of soluble salts (Heneghan et al., 2008). In our own field trial on a tailings storage facility, we found that deep tillage and incorporation of hay into Cu-tailings containing chalcopyrite significantly increased the concentrations of sulfate and chloride in leachate solutions collected over 12 months, under semi-arid climatic conditions in north-west Queensland (Fig. 4). Metal solubility and mobility in tailings can be enhanced by the incorporation of highly decomposed OM, such as biosolids and animal manure composts (Mendez et al., 2007; Santibáñez et al., 2008). Column leaching tests under simulated conditions in the laboratory are useful methods for evaluating likely rates and trajectories of geochemical weathering in relation to various remediation treatments (Kossoff et al., 2011). However, care needs to be taken when interpreting the results, in reference to field processes, by taking into consideration closely related influences of climate (radiation and rainfall) and hydro-geochemical dynamics in tailings.

Improving nutrient supply in root zones is also necessary for newly established plant species, but the effectiveness of added fertilizers can be significantly diminished without the above stabilization of hydro-geochemical dynamics, due to volatile loss of nitrogen at high pH and strong fixation of phosphate by the solid phase, such as in base metal mine tailings and red mud (H. Chen et al., 2005; C. Chen et al., 2010; Penn et al., 2005). As a result, remediation for improving nutrient availability should be of secondary priority to hydro-geochemical stabilization.

IMPORTANCE OF HYDRAULIC PROCESSES IN ROOT ZONE DEVELOPMENT

A functional root zone system is a continuum of soil horizons (e.g. A, B and C) with dynamic hydraulic processes in response to seasonal changes of water availability (Brady and Weil, 2002). The rhizosphere layer may be reconstructed with topsoil, subsoil, overburden, remediated tailings and/or their combinations, while the subsoil layer is usually tailings with or without amendments, depending on the nature of constraint factors in the tailings concerned (see Table 2 for examples). Key hydraulic processes may include water infiltration through the depth profile, capillary rise of pore water from depth and water storage capacity, which determines the capacity of reconstructed root zones to respond to local climatic conditions and water-use requirements of revegetated plant communities across wet and dry seasons. The failure to rehabilitate hydraulic functions in reconstructed root zones is one of the main problems causing the failure of newly established plant communities due to drought during the prolonged dry season and/or waterlogging in the wet season. Hydraulic processes in the root zone can further consolidate or deteriorate the state of hydro-geochemical dynamics through further oxidation, transformation of chemical forms of toxic elements, and vertical enrichment in the rhizosphere layer. The redevelopment of soil hydraulic processes in reconstructed root zones underpins the development and sustainability of newly established plant communities, particularly in regions with prolonged dry seasons (Wehr et al., 2005, 2006). In addition, soil hydraulic functions are important to soil development over long periods of time (Walker and del Moral, 2003).

The rehabilitation of hydraulic processes and functions requires the formation of aggregates and macropores in amended tailings, which may be stimulated by effective remediation options, such as OM, gypsum and reactive Fe-oxides (Barral et al., 1998; Bendfeldt et al., 2001; Ippolito et al., 2005). Emerging evidence has suggested that bulk density and stable aggregation are indicative of long-term soil quality improvement in reclaimed mine soils (Shukla et al., 2004). OM, apart from its benefits to soil biology and nutrient supply, can help improve soil physical and hydraulic properties in tailings due to its effects on aggregation and aggregate stability. Amending red mud with gypsum and green manure increased aggregation and particle size to 30 µm, compared with 5 µm in the original red mud, resulting in increased hydraulic conductance (Harris and Rengasamy, 2004; Harris, 2009). We may also draw lessons from studies on the role of OM in structural improvements of degraded soils. For example, in sandy soils, wheat residue amendments decreased mechanical compaction, and increased aggregation and water retention (Busscher et al., 2008). Grass with 1·9 % N and 24 : 1 C/N ratio resulted in better aggregation of fine clay than straw (0·31 % N, 145 : 1 C/N), while inert charcoal did not generate any micro-aggregation of clays (Watts et al., 2005).

However, correlative changes between OM quality, aggregation and physical structure in amended tailings has rarely been investigated in the development process of the root zone and plant communities in long-term field trials (i.e. over decades). OM added in tailings may be mineralized within a short period of time and the physical improvement in engineered soil matrix of root zones would eventually depend on the decomposition of in situ litter from the revegetated plant communities (Bendfeldt et al., 2001). This would require the evolution of microbial community structure and functions in remediated tailings from a state where it is in equilibrium with the added OM to a state where it has to adapt to local litter from the above-ground plant communities (refer to later discussion). Soil heterotrophic bacteria and mycorrhizal fungi can act as particle aggregators for microaggregates (0·25–0·05 mm) and macroaggregates (2·0–0·5 mm), respectively, through the binding effects of organic exudates, metabolites and hyphae (Preston et al., 1999; Caesar-TonThat et al., 2008; DeJong et al., 2010). Moreover, water-stable aggregation, bulk density and porosity have been tentatively suggested as quality indicators of soil development in revegetated mined land (Shukla et al., 2004).

In directly revegetated tailings, the impacts of OM quality on the aggregation process in reconstructed root zones has yet to be investigated in relation to the physical and chemical characteristics of tailings and amendment options. Beneficial effects of wood chips in tailings amendment have been reported in platinum tailings on the early establishment of grass communities (van Rensburg and Morgenthal, 2004). Woodchip amendment in bauxite residue improved physical structure and generated sufficient leaching in the root zone for vegetation establishment (Ippolito et al., 2005). High-quality OM is easily degradable (Baldock et al., 2004) and thus may not have long-lasting effects on bulk density and aggregation in remediated tailings and thus on soil development. Low- to moderate-quality OM, such as woodchip, is more biologically stable and decomposes at much lower rates (Baldock et al., 2004; Wardle et al., 2004). Stable carbon such as charcoal has low N content, but inert carbon and low-quality OM also have adsorption capacity to remove toxic ions (e.g. heavy metals) from pore water and increase cation exchange capacity for regulating nutrient buffering and supply to plants (Baldock et al., 2004). Knowledge of OM quality on physical and hydraulic improvements in tailings will help better select suitable types or combinations of different types of OM for tailings amendments. Apart from physical, chemical and nutritional benefits in root zones, OM of various types provides important substrates for fostering microbial colonization and the recovery of microbial community structure and functions.

MICROBIAL COMMUNITY STRUCTURE AND FUNCTIONS: BIOINDICATORS OF HYDRO-GEOCHEMICAL DYNAMICS AND ROOT ZONE DEVELOPMENT

After reaching hydro-geochemical stabilization, the second critical stage in the development of reconstructed root zones in tailings is the rehabilitation of soil biological capacity, particularly in the rhizosphere horizon. It includes the colonization of soil microbial communities dominated by heterotrophic bacteria, arbuscular mycorrhiza (AM fungi) and ectomycorrhiza, the establishment of mineralization processes capable of decomposing added (or exogenous) OM in tailings, and nutrient cycling (Fig. 3). Both community size and species richness of soil heterotrophic microbes increase significantly with pH neutralization in mine tailings (Mendez et al., 2008; DeJong et al., 2010). The rehabilitation of heterotrophic microbial communities in remediated tailings may be realized through OM addition and/or artificial inoculation of local or commercial isolates of bacteria and mycorrhizal fungi, such as growth-promoting bacteria, rhizobia, AM fungi and ectomycorhiza (Petrisor et al., 2004; Grandlic et al., 2008; de-Bashan et al., 2012). Different remediation measures can also affect microbial community structure and mineralization capacity in acidic tailings, with the best outcomes in plots that have been neutralized and covered with a topsoil layer (Moynahan et al., 2002).

The colonization of heterotrophic bacteria in remediated tailings can suppress the presence and activity of S- and Fe-oxidizing bacteria and generate biofilms around mineral particles. These biofilms further retard oxidation of sulfide minerals and consolidate hydro-geochemical stability in root zones (DeJong et al., 2010). For example, in unsaturated zones of base metal mine tailings without plant cover, microbial communities are dominated by acidophiles, autotrophic S- and Fe-oxidizers (Schippers et al., 2007; Mendez et al., 2008), but the number of acidophiles significantly declined in covered pyrite-containing tailings (Kock and Schippers, 2006). Some heterotrophic bacteria have growth-promoting effects on roots through exudation of metabolites such as hormones (e.g. indole acetic acid) and metabolites (e.g. siderophore), thus improving root growth and nutrient acquisition (Grandlic et al., 2008; Raes and Bork, 2008; DeJong et al., 2010), and the initial establishment of plant species in the tailings. However, the secretion of chelating compounds by some species of heterotrophic bacteria tolerant of low pH may enhance the solubility of metals (e.g. Cd, Cr, Cu, Fe, Mn, Pb, Zn) released by autotrophic oxidation processes in sulfidic tailings (Moynahan et al., 2002). The presence of soil microbial communities dominated by heterotrophic bacteria and fungi may be used as an indicator of the hydro-geochemical stability in the root zones, due to their correlation with improved chemical conditions and root activities in reconstructed root zones. With the rapid progress of DNA sequencing technology, the community structure profile of soil microbes can be rapidly identified by means of modern meta-genomics tools (Raes and Bork, 2008).

The structure and functions of microbial communities (including heterotrophic bacteria and fungi) in amended tailings are bioindicators of a transition in root zone development from a hydro-geochemical dynamic state to a relatively stable state. These bioindicators are also the critical indicators of a transition to a self-regulating and sustainable state of root zones developed under revegetated plant communities [Fig. 3(2)]. Microbial community structure and functions may be recovered relatively quickly (years), with the stabilizing hydro-geochemical conditions resulting from effective root zone amendments (tailings), but are far from reaching the state observed in an unmined ecosystem (Banning et al., 2011; Mummey et al., 2002a, b). In 0·5- to 3-year-old revegetated sites of residue sand generated from an alumina refinery, microbial communities (both bacteria and fungi) emerged quickly with improving microbial functions, as indicated by declining microbial metabolic quotient. However, the structure of the microbial community was altered by changing physico-chemcial conditions in the residue sand and microbial biomass limited by low OM content (Banning et al., 2011). It remains to be investigated whether the community structure and function can evolve to generate adequate nutrient cycling from decomposition of in situ litter from developing plant communities towards sustainable coastal dune plant ecosystems in the future. In a surface mined site (not tailings), Mummey et al. (2002a, b) found that soil microbial community abundance and structure were different with the age of the rehabilitated plant ecosystem, with increasing ratios of fungi to bacteria biomass in soils under plant cover from a 20-year-old site. As a result, the presence of plant cover and OM in the root zone can significantly stimulate the recovery of microbial community structure and abundance (Mummey et al., 2002a; Banning et al., 2011), suggesting the importance of physical and chemical improvement for initial plant establishment and survival and OM amendments in reconstructed root zones.

As tailings are mostly devoid of OM, the addition of various OM is a common practice in tailings remediation and root zone reconstruction to improve physical, chemical and nutrient characteristics, and to improve and stimulate soil formation in the long term (Wong and Ho, 1994; van Rensburg and Morgenthal, 2004; Courtney and Timpson, 2005; Ippolito et al., 2005; Santibáñez et al., 2008). OM amendments can foster the recovery of microbial community structure and relative abundance (Cordovil et al., 2011). The added OM provides carbon substrates for microbial proliferation, generating coupled decomposition and nutrient cycling processes in the root zone (Baldock et al., 2004). However, OM quality (e.g. chemical resistance and C/N ratios) closely influences microbial community structure in soil (Merilä et al., 2010). In general, plant litter from Australian native tree species tend to have high C/N ratios, which would require increased abundance of fungi communities to break them down and recycle much needed nutrients such as phosphorus in the rhizosphere.

As revegetated plant communities develop, the gradual replacement of initially added OM and associated products with plant litter may induce shifts in the structure and abundance of the microbial community (Wardle et al., 2004). The build up of in situ OM from revegetated plant communities and the recovery of its recycling patterns are consequences of the diversity and productivity of revegetated plant communities and functions of recovered soil microbial communities in reconstructed root zones. In rehabilitated Jarrah forests at bauxite mining sites (surface mined land), there was an increasing trend of total carbon and nitrogen in the top 10 cm of soil along the pseudo-chronosequence of 1–17 years after rehabilitation, with a closing gap from those in soils from adjacent native Jarrah forest (Lin et al., 2011). In directly revegetated tailings without deep topsoil layers, the build up and recovery of decomposition processes of in situ OM from the revegetated plant communities would take longer than decades, because the diversity and productivity of revegetated plant communities would be more limited by the presence of hydro-geochemical factors and associated toxicities in the root zones directly developed from amended tailings. As a result, it is paramount to apply effective remediation options for achieving relatively stable hydro-geochemical conditions in reconstructed root zones to generate adequate plant community productivity and litter return to soil.

Numerous studies have reported short-term beneficial effects of OM amendments (such as green manure, sewage sludge, biosolids and wood chips) in tailings (e.g. gold, bauxite residue and base metal mines) on plant growth (Wong and Ho, 1994; van Rensburg and Morgenthal, 2004; Courtney and Timpson, 2005; Ippolito et al., 2005; Santibáñez et al., 2008). After break-down, OM particles and molecules may form film layers to bind particles together and form aggregates (Garcia-Pausas and Paterson, 2011). The types of OM commonly used in tailings amendments may be generalized into (1) municipal and animal wastes (e.g. biosolids, sewage sludge, manure), (2) composted biomass (e.g. mulch compost) and (3) raw biomass (e.g. hay, wood chips). However, information is not yet available on the relationship between different OM (e.g. with different C/N ratios) and soil biological capacity, and the transition of microbial community structure specific to exogenous OM to litter specific to local plant communities. This transition is important to the establishment of dynamic ecological linkages between the soil subsystem (reconstructed root zone system) and the plant subsystem (target plant communities) which controls the success of sustainable phytostabilization of tailings.

INTEGRATION OF ROOT ZONE FUNCTIONS WITH PLANT COMMUNITY REQUIREMENTS VIA ECOLOGICAL LINKAGES

There is increasing evidence of the importance of tight ecological linkages (e.g. carbon and nutrient cycling processes) in sustaining the succession of soil subsystems and plant communities within plant ecosystems (Walker and del Moral, 2003; Wardle et al., 2004; Merilä et al., 2010). The evolution of ecological linkages between the reconstructed root zones (soil subsystem) and plant communities (plant subsystem) would be critical to the transition of the rehabilitated plant communities into a sustainable state [Fig. 3(3)]. The quantity and quality of OM in reconstructed root zones are expected to shift from the dominance of exogenous (initially added) to native OM derived from litter of local vegetation, as reconstructed root zones develop and local vegetation productivity increases (Bendfeldt et al., 2001). This transition may depend on whether soil microbial community structure and functions in hydro-geochemically stabilized root zones could evolve together with the changes from exogenous OM initially added for tailings amendment, to native OM derived from litter of in situ plant communities. However, there are few long-term data from tailings revegetation on the relationships between improvement of physico-chemical conditions, changes in characteristics of OM composition, and shifts of microbial community structure, functions and biomass abundance. Nevertheless, findings on the recovery of soil OM and microbial communities reflecting target plant ecosystems from reclaimed, surface-mined land may provide some reference expectation for the root zone functional requirements in direct tailings revegetation (Mummey et al., 2002b; Lin et al., 2011).

The rate and trajectories of the decomposition of native soil OM from local vegetation are determined by soil microbial community composition, such as the composition and activities of heterotrophic bacteria and fungal and actinomycete populations (Garcia-Pausas and Paterson, 2011). Studies of long-term changes in soil microbial community structure and its relationship to soil OM quality in revegetated tailings/mine sites may provide useful data to proactively assess the state of reconstructed root zones in relation to the performance of plant communities. The shift of soil OM from exogenous to native sources may be characterized by soil δ¹³C and δ¹⁵N values, which are closely linked with litter composition (Dawson et al., 2002; Wang et al., 2011).

Soil microbial community structure can be altered by OM quality and N-availability, which in turn is linked to soil carbon and nitrogen mineralization and cycling (Cookson et al., 2005; Cordovil et al., 2011). For example, the availability of labile carbon stimulates the activity of specialist microbes, actinomycetes and fungi and the mineralization of recalcitrant OM (Garcia-Pausas and Paterson, 2011). In turn, changes in the quantity and quality of OM caused by changes in plant community composition and productivity can induce changes of microbial community structure and functions in soil (Merilä et al., 2010). As a result, OM of different characteristics in tailings amendment may have profound implications in the initial establishment of microbial community structure and functions, and associated characteristics of OM mineralization and nutrient cycling. This in turn may affect the interspecies competition and plant community structure (Fig. 5) (Wardle et al., 2004).

Fig. 5.

A conceptual diagram illustrating the importance of organic matter inputs and fertility remediation in the rehabilitation of ecological linkages between reconstructed soil subsystem and plant subsystem, for sustainable phytostabilization with target plant communities (adapted from Wardle et al., 2004).

OM with high N and low lignin contents (i.e. high-quality OM) would supply high levels of nutrients at a high rate and generate low levels of stable carbon in soil, which is more compatible with fast-growing and short-lived plant communities such as grasses (e.g. pasture), while OM containing low N and high phenolics and lignin (i.e. low-quality OM) is more compatible with plant communities (e.g. native plants) of low productivity through slow mineralization and nutrient supply and high soil carbon sequestration (Wardle et al., 2004). In many mine sites of arid/semi-arid regions, native plant communities are often preferred outcomes of tailings revegetation on, in particular, gold, bauxite and base metal mines, due to native plant species being highly adapted to local climatic conditions. As a result, high application rates of OM containing high N and labile C may be less desirable for tailings remediation in root zone reconstruction to support a native plant community, such as Spinifex–Acacia shrubland community under semi-arid climatic conditions (Hodge et al., 1997; Gravina et al., 2004; Huang et al., 2011b). High-quality OM stimulates the dominance of fast-growing grass species, which suppresses the growth of other native and slow-growing woody species and thus retards the succession of plant community (Wardle et al., 2004). This may be one of the causes for the dominance of buffel grass (Cenchrus ciliaris) and the absence of native plant species in a past tailings trial at Mt Isa Mine, where large volumes of sewage sludge were applied to Cu–Pb–Zn tailings. As a result, different combinations in timing and OM qualities may be necessary to balance the quick vegetation establishment and the competition between fast- and slow-growing plant species over time.

CRITICAL TRANSITION OF ROOT ZONE DEVELOPMENT: A HOLISTIC VIEW OF THE COMPLEX PROCESS

The development process of reconstructed root zones in tailings for direct revegetation is a complex and dynamic one, with many unpredictable trajectories of changes in physical, chemical and biological properties and functions over many years to decades. It depends on tailings properties, remediation strategy, local climatic conditions and plant community development. In general, the reconstructed root zones are required to develop from a state which is hydro-geochemically and biologically unstable and hostile, towards one which is biologically functional and ecologically sustainable (Whisenant, 2002) (Fig. 6). The reconstructed root zone in tailings needs to develop through two primary stages: (1) hydro-geochemical stabilization, via remediation strategies that are effective in rehabilitating primary structure and functions to alleviate physical and chemical constraints; and (2) the recolonization of soil microbial communities and functions (e.g. decomposition of added OM for supplying nutrients) and the establishment of ecological linkages between the root zone and the above-ground plant community, through adaptive changes of soil microbial communities for OM (plant litter) and nutrient recycling (Fig. 6). The relative depth of the rhizosphere horizon would depend on water storage capacity of the whole root zone profile and water requirements of the most drought-sensitive species in a target plant community under the in situ climatic conditions. A successfully reconstructed root zone in mined land should possess the characteristics of structural stability and functional sustainability.

Fig. 6.

A conceptual diagram of the development process of reconstructed root zones in tailings (adapted from Whisenant, 2002). The transition threshold (1) is dominantly governed by abiotic factors (physical and hydro-geochemical), and is represented by the state of hydro-geochemical stabilization. The transition threshold (2) is dominated by biological factors and is represented by the rehabilitation of soil biological capacity in remediated tailings and ecological linkages with the above-ground plant communities.

Given the time and financial investments, systematic monitoring of root zone development is necessary to provide some assurance and confidence of likely success of tailings revegetation programmes, rather than passively waiting for the responses of established plant communities in years/decades to come. Many soil quality indicators have been proposed for crop and forest soils sustainability (Abbott and Murphy, 2003; Tongway and Hindley, 2004). However, indicator systems are yet to be developed for specifically evaluating the trajectories of root zone development, particularly during the early development stage – the transition from hydro-geochemical stabilization to rehabilitation of soil biological capacity (Fig. 3). Recent studies have suggested that bulk density, stable organic C and aggregation stability are useful quality indicators of soil development in root zones of rehabilitated mined land (Six et al., 2000; Bendfeldt et al., 2001; Shukla et al., 2004, 2005). Further research on long-term soil development in remediated tailings should be carried out to develop indicators for assessing the progressive development of the reconstructed root zones through the critical thresholds, thus providing guidance for system intervention and recovery back to the positive trajectory of system development.

FUTURE RESEARCH

The present review has emphasized the fundamental importance of reconstructing stable and sustainable root zones in tailings to support target plant communities and achieve sustainable phytostabilization. It is doubtful if sustainable phytostabilization can be achieved simply based on tolerant plant species such as metallophytes (Clemens et al., 2002), without adequate knowledge of remediation requirements, which can foster soil formation and development within the reconstructed root zone system in the long term (Mendez and Maier, 2008). It is difficult to judge when the reconstructed root zones have reached the critical transition state with minimal risks of reversal, due to the lack of long-term data monitoring the correlative changes in key hydro-geochemical changes, soil biological processes and functions, and plant community development. This is because monitoring of mined land revegetation has historically focused on the changes and development of plant communities, rather than the changes in root zones, although the latter has started to emerge in recent literature. Our knowledge deficiencies on remediation roles in long-term soil development from amended tailings directly revegetated have been highlighted in several key aspects, including hydro-geochemical stabilization and the development of soil biological capacity and ecological linkages. The relationship between tailings properties and hydro-geochemical dynamics may be simulated in column leaching tests with remediation treatments, but have to be calibrated by field monitoring of changes in physical, hydraulic and geochemical properties and processes, as local climatic conditions (e.g. temperature and rainfall) can significantly modulate the effectiveness of remediation measures and change the intensity and distribution of abiotic factors in the root zone. Field lysimeters (or a simplified version) are an ideal tool for long-term monitoring of the hydro-geochemical dynamics, in response to remediation strategies under field conditions.

The rehabilitation of soil biological capacity and ecological linkages in root zones is critical to long-term soil development, biogeochemical cycling of mineral nutrients and the sustainability of established plant communities. Soil ecosystem processes are closely coupled to the development of plant communities in disturbed lands (Wardle et al., 2004). Research on this topic is limited, but could be facilitated with the aid of rapid and cost-effective DNA sequencing tools, together with other established soil science methodologies. Some lessons may be learnt from the investigation of natural vegetation colonization in abandoned tailings of legacy mines (Bradshaw, 1983; Winterhalder, 1996; Good, 1999; Moynahan et al., 2002; Shu et al., 2005; Hayes et al., 2009). More data are required from long-term (i.e. decades) field studies of tailings revegetation in different climate regions to develop integrative indicators to assess if the reconstructed root zones have developed beyond the critical thresholds with low risks of degradation. This will also provide a proactive system of intervention by designing effective remediation options at different stages of the development process.