Artificial soils from alluvial tin mining wastes in Malaysia – A study of soil chemistry following experimental treatments and the impact of mycorrhizal treatment on growth and foliar chemistry

Abstract

For decades Malaysia was the world's largest producer of Sn, but now the vast open cast mining operations have left a legacy of some 100 000 ha of what is effectively wasteland, covered with a mosaic of tailings and lagoons. Few plants naturally recolonise these areas. The demand for such land for both urban expansion and agricultural use has presented an urgent need for better characterisation. This study reports on the formation of artificial soils from alluvial Sn mining waste with a focus on the effects of experimental treatments on soil chemistry. Soil organic matter, clay, and pH were manipulated in a controlled environment. Adding both clay tailings and peat enhanced the cation exchange capacity of sand tailings but also reduced the pH. The addition of peat reduced the extractable levels of some elements but increased the availability of Ca and Mg, thus proving beneficial. The use of clay tailings increased the levels of macro and micronutrients but also released Al, As, La, Pb and U. Additionally, the effects of soil mix and mycorrhizal treatments on growth and foliar chemistry were studied. Two plant species were selected: Panicum milicaeum and Pueraria phaseoloides. Different growth patterns were observed with respect to the additions of peat and clay. The results for mycorrhizal treatment (live inoculum or sterile carrier medium) are more complex, but both resulted in improved growth. The use of mycorrhizal fungi could greatly enhance rehabilitation efforts on sand tailings.

---

Environmental impact

Worldwide there is an environmental legacy from past mining activities. Malaysia was the world's largest producer of Sn, but now the vast open cast mines have left over 100 000 ha of wasteland, covered with a mosaic of tailings and lagoons. Such areas are difficult to reclaim, since the remaining material has poor water and nutrient retention and often contains toxic metals. Few plants naturally recolonise these areas. However, with rapid urban growth there is now pressure to consider the use of this land for agricultural purposes. This multidisciplinary study reports on the formation of artificial soils from alluvial Sn mining waste and the effects of mycorrhizal treatments on growth and foliar chemistry in order to better elucidate appropriate reclamation treatments.

Introduction

Malaysia's former position as the world's largest exporter of Sn has left it with a legacy of some 100‚000 ha of mined land comprising a complex mosaic of lagoons, overburden dumps and tailings.¹ These latter provide a spectrum of edaphic conditions ranging from almost pure clay to almost pure sand, and it is the sandy fractions (sand tailings) that dominate. The proximity of such mined land to urban centres has made them a target for agricultural use since the 1930s,² while loss of productivity from the natural forest estate has more recently made them a target for establishment of monoculture crops under the “Compensatory Forest Plantation Project”.³ However, while rehabilitation of the clay tailings is relatively straightforward, efforts to vegetate the much more extensive sandy tailings are usually hampered by the physical intractability of the substrate, which may comprise as little as 1% clay and silt-sized fractions.⁴ Natural rehabilitation of such mined land is immensely slow, due to a combination of excessive drainage and low nutrient status.^5,6 However, these constraints may be overcome by heavy amendment with bulky organic materials.,⁷ simultaneous irrigation and mineral fertilization or mixing with the clay fraction which was separated during mining.⁸

Although the agricultural use of tailings increased during the 1990s, it utilised less than 5% of the available land resource toward the end of that decade.⁹ It was also during this period that concerns were first raised over the possibility that food crops grown on Sn tailings could be contaminated by potentially toxic elements.

The geological genesis of commercial Sn deposits means that tailings produced during their exploitation are likely to retain residual quantities of a range of heavy minerals, including some with toxic potential.¹⁰ Shamshuddin et al.,¹¹ demonstrated high concentrations of As in both sand and clay tailings, while Ang et al.,⁹ found that Pb and Zn concentrations in some agronomic species cultivated on these tailings exceeded locally permitted levels. In our own studies, we have found that elements such as Pb, U and As were widespread in both sand and clay tailings, but were not responsible for influencing colonization of these tailings by the local pioneer flora.

A number of organic materials are commonly applied to sand tailings in Malaysia, including palm oil processing effluents, rice straw, peat, and chicken manures. Each of these will have different impacts upon trace metal availability in sand tailings, depending upon their individual chemical characteristics, in turn dependent upon their state of decomposition. Although polyphenolic compounds in newly cut plant materials will complex with trace elements, decomposition leading to carboxyl and phenolic groups will greatly enhance their complexation potential. Indeed, the greater effectiveness of leguminous residues in reduction of Al availability may be due in part to their more rapid mineralization.¹² Whether additions of organic matter to sand tailings would result in synergistic or antagonistic interactions with any clay amendments has not yet been examined.

Common rehabilitation practises, such as the addition of clay and organic matter, can have profound effects on the trace and nutrient element chemistry of sand tailings and other soils. Previous studies of clay tailings generated by the mining of alluvial tin deposits have shown that they are dominated by kaolinite.¹³ Kaolinite has a limited capacity for exchangeable binding of cations, and this capacity diminishes with pH due to saturation of exchange sites by hydrogen ions.¹⁴ In contrast, the products of kaolinite weathering (such as gibbsite) are known to act as strong adsorbents of anions, including phosphate and its analogue arsenate.¹⁵ Under acidic conditions, gibbsite itself may dissociate, releasing Al ions that further compete with cation binding sites in the soil. If kaolinite and its weathering products are found in clay tailings, then the effects of adding such a material to sand tailings could include: improved water retention, small increases in cation exchange capacity, but also a tendency to adsorb nutrient and trace element anions. An abundance of Al ions will also tend to increase acidity and compete for cation exchange sites with nutrient cations such as Ca and Mg, which may therefore be deficient for colonising or planted flora.

Adding organic amendments should also improve the water and cation exchange capacities of sand tailings, but will also impact upon trace metal behaviour. Lund & Fobian¹⁶ showed that As was retained in both inorganic and organic soil horizons, although adsorption by soil organic matter in the topmost horizon was thought to be more significant than adsorption by clay minerals. Cao & Ma¹⁷ demonstrated that composted plant material could adsorb arsenate but that adding phosphate to the amended soil would counteract this beneficial effect by displacing arsenate from the exchange sites. Many other effects of trace metal/organic matter interactions have been examined, and from these it is impossible to draw definite conclusions. Karaca¹⁸ studied Cu, Ni and Zn extractability from soils amended with various organic media. Two of these media significantly reduced Cu availability, one increased extractable concentrations of Ni, while all three increased extractable concentrations of Zn. He & Singh¹⁹ demonstrated an increase in cadmium availability (as measured in diethylene triamine pentaacetic acid (DTPA) extracts) as a result of adding sphagnum peat to various soils, and postulated that this was a result of peat-induced reduction in soil pH. Organic matter also plays a significant role in the detoxification of Al in some soil systems, although the mechanisms are diverse and may involve simultaneous chelation, complexation, adsorption and co-precipitation. When bound to soluble humic materials, Al is rendered non-phytotoxic,²⁰ while humic acids are also effective at reducing both the rhizotoxicity of La to maize²¹ and the phyto-availability of Pb.²²

Among the many attempts to rehabilitate former minesoils in Malaysia, few have examined the potential benefits of mycorrhizal fungi. Although primarily responsible for the improved P nutrition of their host plants,²³ the involvement of these symbiotic fungi in a number of other nutrient and trace metal effects has been extensively researched. Not only can their presence determine plant community structure during early successional stages,²⁴ but they are thought to improve plant water relations,²⁵ N nutrition²⁶ and micronutrient uptake.²⁷ Mining activity is known to disrupt or even eliminate mycorrhizal populations in topsoils,²⁸ and plants are often deliberately inoculated prior to use for rehabilitation of degraded soils.²⁹ Their use in reclamation of sand tailings in Malaysia has been limited,^30,31 but they are widely used elsewhere during the rehabilitation of similar substrates.^32,33 It is impossible to generalize on the role of mycorrhizal fungi in enhancing or reducing the uptake of potentially toxic elements from contaminated soils, since their influence varies depending on plant growth conditions, the symbiotic species and the element under consideration.^34,35 However, Chen et al.³⁶ suggested that mycorrhizal inoculation reduced Pb toxicity in plants by sequestering Pb within plant roots. Similar effects were demonstrated for As partitioning in the grass Holcus lanatus by Gonzalez-Chavez et al.³⁷ and for U in barley (Hordeum vulgare) by Chen et al.³⁸ Rufyikiri et al.³⁹ also demonstrated reduced U concentrations in the shoots of mycorrhizal subterranean clover (Trifolium subterraneum) when compared with non-mycorrhizal controls.

If potentially toxic elements are selectively adsorbed by organic or clay amendment of sand tailings, it is unknown whether mycorrhizal fungi (which would preferentially colonize the organic fractions of soil⁴⁰) would in turn liberate these elements, especially under low nutrient conditions. Unlike rhizobial associations, which fix N from the atmosphere, mycorrhizae are only able to access existing soil nutrient reserves – and these may be so low in sand tailings as to render mycorrhizal inoculation inappropriate.

The aim of this current study was primarily to examine the effects of using additions of clay tailings and peat on the nutrient and potentially toxic element contents of Malaysian sand tailings. The effects of inoculation with mycorrhizal fungi were also examined. Mycorrhizal fungi are widely used to improve plant establishment on degraded soils, including former mine sites.²⁹ Commercial mycorrhizal inocula do not generally contain pure isolates of mycorrhizal fungi or their infective propagules. Instead, mineral bulking agents such as expanded clays and zeolites are added to both buffer and bulk the infective material, protecting it during transport and storage, as well as facilitating its application in the field. The commercial TerraVital-D inoculum (Plantworks UK, Sittingbourne, Kent) used during this research contained (an unspecified) expanded clay, together with particles of slate-argillite and zeolite in addition to the active propagules: fungal spores, mycelia and colonized root fragments. Zeolite is used to reduce the leaching of heavy metals from contaminated land,^41,42 and has been shown to be particularly effective in reducing Cd mobility.⁴³ Since the mycorrhizal carrier medium may influence the nutrient and trace metal chemistry of soils over and above effects exerted by the mycorrhizal fungi themselves, we also sought to examine the influence of such a medium on the chemistry of amended sand tailings.

Materials and methods

Creation of artificial soils

Three different waste materials (sand tailings, clay tailings and peat) were collected from a working mine site near the town of Dengkil in the Malaysian State of Selangor.

Surface (0–15cm) samples of sand tailings were collected systematically from a half-hectare area of recently mined land. Clay tailings were collected in the same way from a former slime-retention lagoon, although extensive plant colonization first required the clearance of both plants and a thin organic horizon to expose the clay beneath. Peat was sourced from a further area of the site, cleared of forest in preparation for mining, where it formed a one metre layer on top of the alluvial overburden. Fine grade organic material was collected systematically across an exposed face of approximately one hundred metres.

A range of peat and clay additions were used to examine the individual and interactive effects of these materials. Since Lim et al.⁴⁴ indicated that the greatest improvement in yield of Brassica chinensis occurred with a mix of seventy percent sand and thirty percent clay, thirty percent was fixed upon as the highest added clay treatment. The organic C content of natural tropical soils is variable, with Kauffman et al.,⁴⁵ giving a range of 0.8 to 5.0% from arenosols and podzols respectively. Six percent was therefore chosen as an upper limit for peat treatments.

The fresh volume/dry weight ratios of each material were calculated to facilitate the soil creation process. This ratio was determined for field-dry sand tailings, field-dry peat and clay slip (a liquid mixture of clay tailings and rainwater). The slip was created as the simplest method for ensuring adequate homogenisation of the mineral components of the experimental soils. The required volumes of clay and sand were manually mixed, air-dried and disaggregated. The appropriate volume of peat was then incorporated and the resulting soils (Table 1) air-dried again before disaggregation and use. The commercial mycorrhizal carrier medium, an expanded clay (details not provided by the supplier), together with particles of slate-argillite and zeolite, was added in the ratio of approximately one to nine (200 ml of carrier medium to 1600 ml soil), and the resulting thirty-two treatments used in glasshouse trials. To differentiate between genuine mycorrhizal effects and the impact of the carrier medium, all active mycorrhizal treatments were complemented with treatments containing the sterile carrier, mixed at the same rate as the inoculum. Subsamples of all experimental soils were removed for laboratory characterization (after allowing them to age for about six weeks).

Table 1 Ratios (percentage by oven-dry (103 °C) weight) of the three waste materials used to create the sixteen artificial soils

Soil number	Sand	Clay	Peat
1	100	0	0
2	100	0	2
3	100	0	4
4	100	0	6
5	90	10	0
6	90	10	2
7	90	10	4
8	90	10	6
9	80	20	0
10	80	20	2
11	80	20	4
12	80	20	6
13	70	30	0
14	70	30	2
15	70	30	4
16	70	30	6

---

If existing forest communities growing on sandy soils are to provide a model for plant growth on sandy tailings, it is necessary to identify whether these substrates are chemically and physically comparable. Thus, two natural soils were also collected to serve as analogues for the sand and clay tailings. Rudua series soils are derived from monsoonal beach terraces along the eastern coast of Peninsular Malaysia. The upper horizons of the resulting podzols are freely draining with low inherent fertility.⁴⁶ Rengam series soils are granite-derived, kaolinitic and widespread across Peninsular Malaysia.⁴⁷ Both soils were systematically collected by taking surface (0–15cm) samples from across areas of approximately one hectare. These were bulked together and sub-sampled prior to analysis. Rudua soils were collected from a location beneath virgin heath forest near the village of Jambu Bongkok on Peninsular Malaysia's east coast. Rengam (kaolisol) soils were collected from beneath secondary forest in Kuala Lumpur.

Sample pre-preparation

All soils were air-dried at approximately 35 °C before sieving through 2mm square nylon mesh. For C and N determinations only, a subsample of the sieved material was ground in a stainless steel disc mill (Tema, UK) until capable of passing a 250μm nylon mesh sieve.

Analytical techniques

To approximate plant-available fractions of nutrient and potentially toxic elements, <2mm soils were subjected to Mehlich-3 extraction, with the resulting solutions tested using a combination of ICP-AES (Varian Liberty 200, Varian UK Ltd), ICP-MS (VG Elemental PlasmaQuad 3, Thermo Electron Corporation) and FAAS (Varian SpectrAA 400 Plus, Varian UK Ltd). Mehlich-3 is a universal multi-element soil extractant that is universally gaining popularity with soil testing bodies for its ease of use. It comprises: 0.2N CH₃COOH/0.25N NH₄NO₃/0.013N HNO₃/0.015N NH₄F/0.001M EDTA. Prior to adoption, the Mehlich-3 extractant was compared with a 0.01M CaCl₂ extraction using samples of sandy tailings and the protocols proposed by the Soil and Plant Analysis Council.⁴⁸ The results were low for most elements by both extractants, but since it was possible to determine Ca using the Mehlich-3 extract, this was adopted as the preferred method.

Soil pH was determined in a 1 : 2.5 (w/v) suspension of <2mm soil in distilled water, while total C and N contents were determined in <250μm material by autoanalysis (CE Instruments EA 1100 CHNS Autoanalyzer, Thermo Quest, Italy).

Effective cation exchange capacity was determined by summing Mehlich-3 extractable concentrations of Mg, Ca and K, together with exchangeable acidity (H⁺ + Al³⁺), determined in 1N KCl extracts.

Analytical quality was assured through the use of a certified reference materials: CANMET: Sandy soil SO2 (obtained from the Canadian Certified Reference Materials Program, Natural Resources Council, Canada), and WEPAL ISE954 clay soil and WEPAL ISE971 sandy soil (obtained from the Wageningen Evaluating Programme for Analytical Laboratories (WEPAL) the University of Wageningen, Netherlands), and internal standards and reagent blanks (where appropriate). All analyses were performed in at least triplicate. Although the SO2 soil is not certified for elements extracted by either 0.01M CaCl₂ or Mehlich 3, the WEPAL materials are suitable for comparison, although the values quoted are consensus rather than certified values. In general, the values obtained were within 10% of the consensus values for the 17 elements listed.

Analyte selection

The selection of analytes was based on available data.^9–11 The pH, total C, N, P and K were determined using the techniques described above. The following elements were also determined: Al, As, Ca, Cu, Fe, La, Mn, Ni, Pb, S and Zn, again using the techniques described above (see also Table 2).

Table 2 Mehlich-3 extract data for artificial soil ingredients and a natural kaolisol. All data in mg kg⁻¹ except pH, C (%) and N (%)

Parameter/Element	Sand	Clay	Peat	Kaolisol
pH	4.2	3.57	3.65	4.54
C	0.11	0.79	38.7	0.77
N	0	0	1.75	0
P	0.304	1.26	0.229	0.291
K	5.05	14.6	0	14.6
As	0.054	0.165	0.166	0.058
Al	127	1020	1700	538
Ca	13	74.1	1670	189
Co	0.019	0.544	0.104	0.011
Cu	0.047	2.72	0	0.179
Fe	18.8	169	170	41.6
La	0.058	1.55	0.021	0.213
Mg	1.18	58.2	177	9.04
Mn	0.077	13.3	3.37	1.68
Ni	0	0.503	0.1	0.061
Pb	0.956	20.9	0.621	0.659
S	25.4	329	113	47.4
U	0.165	0.243	0.008	0.005
Zn	0.235	4.01	1.09	0.602

---

Statistical analysis

To determine whether there were statistical differences between sets of data obtained for the soil studies, the one-way ANOVA procedure was used. Statistical analyses were performed using a General Linear Model within the Minitab Release 13.32 (Minitab Inc, USA). Where necessary, multiple comparisons were made using Tukey tests. However, for more sophisticated identification of the significance of each experimental factor (clay, peat and mycorrhizae) and their interaction on the weights of Panicum and Pueraria, three-way ANOVA was used (Genstat software, Rothamsted, UK).

Mycorrhizal treatments

Mycorrhizal inoculum (TerraVital-D) was purchased from a commercial source (Plantworks UK, Sittingbourne, Kent). The active constituents of this product were fungal spores, mycelia and root fragments colonized by five mycorrhizal species: Glomus manihotis, G. mosseae, G. geosporum, G. microaggregatum and Gigaspora rosea. These were held in an inert carrier medium containing expanded clay (not specified by the supplier), together with particles of slate-argillite and zeolite, producing a medium with pH 7.6.

Zeolite is often used to reduce the leaching of heavy metals from contaminated land, where it is particularly effective in reducing cadmium mobility.⁴³ It might therefore be expected to exert an influence on experimental soils over and above that produced by the mycorrhizal fungi themselves, especially given the disparity in pH between the different media (sand tailings often have a pH of 4.0 or less⁶). To differentiate between genuine mycorrhizal effects and the impact of the zeolitic carrier medium, all active mycorrhizal treatments were complemented with treatments containing the sterile carrier, mixed at the same rate as the inoculum. All treatments were replicated four times.

Plant species

Panicum milicaeum (Proso Millet) is a rapidly-maturing graminaceous species, widely planted in temperate and subtropical regions on soils which are too shallow or too infertile for the successful cultivation of other cereals. Although the use of this species in the human diet is declining, it has been recommended as a short-duration catch crop for South-East Asia, being able to take advantage of residual soil moisture during dry periods.⁴⁹ Seed of this species was obtained locally. Pueraria phaseoloides (Tropical Kudzu) is a leguminous species widely grown as ground cover in palm oil and rubber plantations⁵⁰ and is also found as a pioneer on sandy tailings.⁵¹ It has “moderate to low” requirements for Ca and P, being well suited to acid tropical soils.⁵⁰ Seed of this species was purchased from B & T World Seeds (Olonzac, France).

Procedure

Seed of both species was surface sterilized by immersion in 5% sodium hypochlorite solution for 15 min, followed by repeated rinsing in distilled water.⁵² They were then placed in a sterile petri dish lined with filter paper, and allowed to air-dry before use. Five seeds of Panicum miliaceum were sown directly into each pot containing artificial or natural soils. Germination was extremely rapid, normally taking two or three days. Germination of Pueraria phaseoloides was more erratic, taking up to three weeks. To reduce variability within the experiment, all seed of this species was sown in plastic troughs filled with solar-sterilized mine sand,⁵³ and allowed to grow on for a week after germination. Similarly sized seedlings were selected from this source and transplanted into the experimental soils. Any plants which did not survive this process were replaced until all pots contained single established seedlings of similar size at a period of three weeks. At this point all were randomized within the greenhouse.

Soils were watered to near field capacity at least twice daily, using rainwater. Although it was possible that collected rainwater would contain wind-blown mycorrhizal spores, this would more closely resemble field conditions and it was not felt necessary to provide a sterile water source. However, to reduce the chances of airborne cross-infection within the greenhouse, the floor and wall surfaces were damped-down with tap water twice daily.

Panicum milicaeum plants began to flower after two months, and were harvested at that point. Pueraria phaseoloides were harvested after three months, when still growing vegetatively. All above-ground parts of the plant were removed and weighed to 1.0mg. The material was then transferred to paper envelopes and dried in a laboratory oven at 80 °C until they reached constant weight.⁵⁴ Leaves were carefully brushed to remove surface debris, but were not otherwise cleaned prior to analysis. Dry weights were recorded to 0.1mg.

Dissolution of plant material for analysis was achieved by digesting dried, ground sample in hot, concentrated nitric acid.⁵⁵ Concentrations of Al, As, Cu, La, Mn, Ni, Pb, U and Zn were determined in the resulting solutions using a combination of ICP-AES and ICP-MS. Very low growth under some treatments meant that it was impossible to perform replicate analyses for much of the foliar material. Data for foliar chemistry are therefore presented as n = 1, without statistical analysis.

Results and discussion

Soil studies

The data in Table 2 reveal that although the natural soil is acidic, the mine wastes are more so, the lowest pH being recorded for the clay fraction. This may be due to the presence of oxidised pyrites (which could also be responsible for the elevated S content of clay tailings), but also be a result of Al released from the weathering of clay minerals.⁵⁶ The presence of organic matter (as extrapolated from C and N data) is low in the natural soil, but lower in sand tailings and at an equivalent level in the clay tailings. Any impacts of organic matter upon trace element characteristics in the artificial soils are therefore most likely to derive from the addition of peat rather than clay.

Other macronutrient elements (K, Mg and S) were all present at lower concentrations in sand tailings than the kaolisol, while concentrations of P were rather similar – possibly due to immobilisation of phosphate by sesquioxides in the natural soil, and its presence in heavy minerals such as monazite and xenotime which are found in sand tailings.⁵⁷

Concentrations of potentially toxic elements were rather variable among the different substrates, although clay and sand tailings retained higher concentrations of Pb and U than either peat or the kaolisol – the Pb content of clay tailings was particularly elevated, and might impact upon the use of this substrate in agricultural rehabilitation efforts.

Table 3 shows the data for pH presented for Rudua (sandy), Rengam (kaolinitic) and artificial soils containing various proportions of sand and peat (% by weight) and equivalent soils dosed with mycorrhizal carrier medium. The full data set for C, N, P, K, and other metals studies may be found in the Electronic Supplementary Information (ESI) Tables S3a–3s.

Table 3 pH data for Rudua (sandy), Rengam (kaolinitic) and artificial soils containing various proportions of sand and peat (% by weight) are tabulated along with data for equivalent soils dosed with mycorrhizal carrier medium. Data within columns (for separate mycorrhizal treatments) followed by different alphabetic superscripts and data within rows (for separate mycorrhizal treatments) followed by different numerical superscripts are significantly different at p ≤ 0.05. Comparisons between non-mycorrhizal and mycorrhizal carrier treatments are also included (*denotes a significant difference at p ≤ 0.05)

Non-mycorrhizal
		Peat (%)
		0	2	4	6
Sand (%)	100	4.05 (0.02)^a,1	4.10 (0.01)^c,2	4.04 (0.02)^d,2	3.87 (0.05)^c,2
	90	3.81 (0.01)^d,1	3.70 (0.01)^b,2	3.63 (0.05)^c,3	3.51 (0.01)^b,4
	80	3.68 (0.01)^c,1	3.56 (0.01)^a,2	3.53 (0.04)^b,2,3	3.50 (0.01)^b,3
	70	3.73 (0.01)^b,1	3.56 (0.01)^a,2	3.38 (0.01)^a,3	3.39 (0.01)^a,3
Rudua		4.63 (0.00)^f
Rengam		4.61 (0.01)^e

---

Mycorrhizal carrier medium
		Peat (%)
		0	2	4	6
Sand (%)	100	6.19 (0.01)^d,1	6.38 (0.01)^c,2	6.00 (0.01)^d,2	5.68 (0.01)^d,4
	90	4.93 (0.01)^c,1	5.48 (0.02)^b,2	6.42 (0.01)^c,3	5.84 (0.01)^c,4
	80	6.12 (0.01)^b,1	5.73 (0.02)^a,2	6.17 (0.00)^b,3	6.11 (0.01)^b,1
	70	5.24 (0.00)^a,1,3	5.73 (0.00)^a,2	5.23 (0.01)^a,3	5.39 (0.01)^a,4

---

Comparisons
		Peat (%)
		0	2	4	6
Sand (%)	100	*	*	*	*
	90	*	*	*	*
	80	*	*	*	*
	70	*	*	*	*

---

Effects of clay and peat additions on the chemistry of sand tailings

pH, total carbon and macronutrients. Adding both peat and clay caused significant decreases in soil pH, although the effect was more pronounced by the additions of clay than peat. Although many tropical species are adapted to acidic soil conditions, and the natural soils were both acidic, none of the experimental media would be favourable for crop production without liming or other acid-mitigation intervention (such as the use of green plant residues.²⁰

Carbon contents increased with the addition of both peat and clay, with more significant increases resulting from peat additions. Levels of N in most of the soils fell below the limit of detection for this element (0.1 g kg⁻¹) and there was no consistent pattern relating to the addition of clay. However, peat added at levels of 4% or above did result in measurable increases in N.

Adding peat seemed to diminish the extractability of P, although there was little statistical confirmation of this, and it may be a consequence of dilution by peat during creation of the soils. In contrast, increasing clay contents actually raised the extractable P contents of soils. Phosphate binding is common in weathered tropical soils, which are rich in Fe and Al sesquioxides,⁵⁸ but it appears that the alluvial clays extracted during mining either do not contain excesses of these minerals, or that anion exchange sites are already occupied – possibly by arsenate. Another possibility is that the extractant is sufficiently aggressive to displace such phosphate pools, although this seems unlikely given the relatively lower quantities of phosphate extracted by Mehlich-3 when compared with other extractants such as Bray-2 (0.03M NH₄F + 0.1M HCl).⁵⁹

Potassium concentrations were not enhanced by the addition of peat but were significantly increased at each addition of clay, possibly reflecting an abundance of micaceous minerals¹¹ in this medium. As with P, concentrations of K in the Rudua (sandy) soil were extremely modest, even when compared with pure sand tailings, reflecting a residue of micaceous minerals in this latter substrate that is removed during marine weathering and subsequent development of Podzols.

Sulphur levels were not significantly altered by additions of peat, despite the relative abundance of this element in this substrate. In contrast, clay did have a significant impact on S levels, which increased with soil clay content. Clay tailings are very poorly drained, and under water-logged conditions, sulphide generation may occur. The clay tailings might also harbour greater quantities of heavy sulphide minerals than the sand tailings. Concentrations in the natural sandy soil were significantly lower than those of sandy tailings to which no clay or peat had been added, while those of the kaolinitic soil were exceeded as soon as sandy tailings had been amended with 10% clay tailings.

The response to peat addition is very clear, with increased percentages of organic matter resulting in increased concentrations of extractable Mg (with one exception). Adding clay also increased extractable Mg concentrations, but only up to soils with 20% clay content. Levels in the natural soils were very low.

The Ca response was similar to that for Mg, increasing with both peat and clay additions, up to soils with 20% clay content. The naturally sandy soil retained very little extractable Ca (even in comparison with the sand tailings). Levels in the kaolinitic soil were surprisingly high, given their weathered, acidic state. Further investigation is required to identify the source of this elevation.

Trace elements. Both clay and peat were shown to contain similar Mehlich-3 extractable concentrations of Fe, and adding both these amendments resulted in increases in this element. Baseline concentrations in sand tailings were higher than their natural analogue, and additions of either amendment rapidly exceeded the concentration found in the kaolinitic soil, suggesting an abundance of Fe in both amendments.

Overall, adding peat did not significantly influence extractable Mn levels in sand tailings, although there is limited evidence that it increased Mn availability in soils to which clay had not been added. In clay-amended tailings, levels of extractable Mn increased, the limited effects of peat were no longer significant. The response for Zn was similar, increasing stepwise with increases of soil clay content. Concentrations were greater in all artificial soils when compared with their natural analogues, although further investigation would be necessary to determine whether these concentrations were phytotoxic.

de Matos et al.⁶⁰ found that the organic component of tropical soils was significant in retarding Cu leaching from these soils, while Karaca¹⁸ found no consistent effect on DTPA-extractable Cu concentrations of adding mushroom compost to soil. Although there is some evidence of organic retention of Cu in the artificial soils, this only becomes significant for a minority of cases. However, concentrations increased significantly as clay contents were increased.

Nickel extractability was not significantly influenced by peat additions, despite the findings of Karaca¹⁸ that extractable Ni concentrations were greater in soils to which mushroom compost had been added. Adding clay resulted in increased Ni concentrations, although this was only significant in a stepwise fashion for those soils to which 6% peat had also been added. As with Mn, concentrations of Ni were greater in all artificial soils when compared with natural soils – indeed they were below the limit of detection (0.001mg kg⁻¹) for this element in the sandy soil.

Al contents of both peat and clay were extremely high (Table 2), and increases in Mehlich-3 contents of this element with additions of each material would be expected. These are observed in the available data, although when taken in combination, it is the increases of clay, rather than peat which result in significant increases in extractable Al. Haynes & Mokolabate²⁰ reported reduced Al availability following increases in soil organic content, but this is not seen here. Indeed, significant increases in extractable Al occurred in many experimental soils upon incorporation of peat, possibly due to the high loading of peat by this element, reducing the availability of adsorption sites for excess Al contributed by clay amendments. Tipping⁶¹ demonstrated that Al competed for binding sites on fulvic acid, liberating Zn and Pb at both low and neutral pH. Thus, although dissolved organic acids might reduce Al availability, there is a danger in contaminated soils of a concomitant increase in concentrations of potentially phytotoxic elements. Despite the elevations witnessed by adding clay, levels were generally greater in the natural kaolinitic soil. The very low occurrence of Al in the sandy soil (compared with sand tailings) might indicate difficulties in any attempts to establish ‘artificial’ heath communities on sand tailings, since much of the heath flora may not be pre-adapted to high-Al soils.

Arsenate is known to be strongly adsorbed by various soil constituents, including Fe, Al and Mn oxides, and soil organic matter.⁶² Adding peat at any level to sand tailings significantly reduced Mehlich-3 extractable As if no clay was added. Extractable concentrations in such soils were lower than those of the natural kaolisol, although still higher than those of the sandy soil. In contrast, adding clay produced not significant decreases, but significant increases in extractable As, suggesting that available adsorption sites were already occupied.

Dissolved organic acids are thought to increase the phytoavailability of micronutrients in soils, but as a secondary effect, may increase the phytoavailability of potentially toxic elements.⁶³ Indeed, low molecular weight organic acids are known to increase La uptake in barley.⁶⁴ Under the experimental conditions used here, addition of peat actually reduced La extractability, and for soils with 20 and 30% added clay, there was a reduction in concentration with peat, suggesting strong adsorption of this element. Clay tailings increased La concentrations in all non-mycorrhizal soils, again suggesting its occurrence in a residual heavy mineral fraction within this medium.

Soil organic matter is known to reduce the availability of Pb,⁶⁰ but the response for Mehlich-3 extractable concentrations of this element are not so conclusive, with peat having no overall significant reductive effect on Pb. As with La, step-wise additions of clay resulting in increases in Pb availability.

Shahandeh & Hossner⁶⁵ demonstrated reduced plant uptake of U when plants were grown in clayey soils with elevated contents of Fe, Mn and organic matter, thanks to strong adsorption of U by these fractions. Adding peat to sand tailings resulted in similar reductions, which were almost step-wise with increases in peat. Although clay tailings produced an increase in extractable U, these increases were greatly offset by the addition of peat. Generally, the use of the mycorrhizal carrier medium had no significant effect, but where it did (as with Pb), this effect was reductive. Concentrations of this element were much greater in the mine wastes than natural soils.

Effects of mycorrhizal carrier medium on the chemistry of amended sand tailings

pH, total carbon and macronutrients. The mycorrhizal carrier medium had a significant effect on all experimental soils, increasing pH to levels at which the availability of many potentially toxic elements would significantly decrease,⁵⁶ possibly leading to improved crop performance without any additional benefit incurred as a result of inoculation with mycorrhizal fungi. The carrier medium also produced a significant increase in C contents, although this was more notable in soils to which no or only 2% peat had been added. This might also result in significant changes to trace element availability.

No significant effects were apparent in total N, Mehlich-3 extractable P or S. Since the exploitation of limited soil phosphate pools is regarded as a significant mycorrhizal faculty, often resulting in improved growth in inoculated crops,⁶⁶ it is gratifying to note that such a response cannot be attributed to the carrier medium alone.

However, the carrier medium did produce significant increases in K and Ca concentrations in the artificial soils, although this is unlikely to improve crop nutrition (given the low N and P status of the soils). Magnesium concentrations were also significantly enhanced by addition of the carrier medium – but only in the treatments to which no clay had been added.

Trace elements. Responses to the carrier medium of trace element extractability were extremely variable. Adding it to sand tailings to which peat (only) had been added caused an increase in extractable Mn, while concentrations declined with added clay. Likewise, an increase in extractable La from unamended sand tailings became a decrease once clay had been added. Al and Cu extractability were not affected by the carrier medium, while Fe was apparently suppressed. Lead extractability was reduced in those treatments for which a significant effect could be determined.

Perhaps the most important trace element affect of the carrier medium was on As availability, which significantly increased as a result of its addition in almost all artificial soils. This was possibly due to the increased soil pH, which is known to lead to increased desorption of H₂AsO₄⁻.⁶⁷

Plant growth: Panicum miliaceum

Only fifty percent of non-mycorrhizal soil treatments produced significant growth enhancements when compared with the control (Table 4). It appears that adding both peat and clay improved the growth of this species, although the effect of peat was reduced in those soils containing the highest percentage of clay (30%). All inoculated treatments produced significant growth enhancements, as did all treatments using the sterile mycorrhizal carrier medium. Under these conditions, plant growth was improved by additions of clay, but such improvements were subsequently offset by addition of peat. Plants grown on both natural soils were significantly larger than those grown in the experimental media (without mycorrhizal treatment).

Three-way ANOVA indicated that all main factors and their interactions were significant at p ≤ 0.05 (data not shown). The separate mycorrhizal treatments were subsequently subjected to two-way ANOVA, indicating that all main factors and interactions were again significant within each mycorrhizal treatment, except that additions of peat had no significant effect on the dry weight of Panicum when inoculated (ESI Table S4a).

Under non-mycorrhizal conditions, adding clay reduced the growth of Panicum miliaceum, although these differences were only significant when peat had also been added at 2 or 4%. Adding peat significantly improved growth except in the 70% sand (30% clay) treatment. These improvements did not significantly continue with additional peat, suggesting that a 2% level was sufficient to obtain the best growth improvement of this species on sand tailings. Sliming (the practise of adding clay to sand tailings) could not be recommended based on these greenhouse trials. However, under field conditions (where water might be limiting) the effects of such treatment might prove beneficial (ESI Table S4b).

The significant overall improvement in plant growth following mycorrhizal inoculation has already been noted. However, it is also possible to note significant impacts of the other soil treatments on growth of the mycorrhizal plants. In only two instances did clay addition significantly enhance growth – when 4% peat was added to either 80 or 90% sand mixes. Otherwise, neither peat nor clay additions generally improved growth to a significant degree.

Significant enhancement of Panicumgrowth also resulted from soils treated with sterile mycorrhizal carrier medium, although these improvements were not so great as demonstrated for inoculated treatments. Examining the combined statistics shows that adding clay generally improved growth (but no further gains were made by adding more than 10% clay) while adding peat reduced it.

Plant growth: Pueraria phaseoloides

Growth of Pueraria phaseoloides was significantly enhanced (when compared with the control) under almost all experimental conditions (Table 5). The impact of inoculation on foliar weights is apparent, as is the impact of adding the sterile carrier medium. Plant growth was significantly greater on both natural soils when compared with those established on the sand tailings. In contrast with data for Panicum miliaceum (Table 2), adding clay appeared to improve growth, but these improvements were then offset by the addition of peat.

Table 4 Fresh weights (mean and standard deviation) of millet (Panicum miliaceum) in various artificial and natural soils (mg). All data are compared with the control (100% sand, no peat, no mycorrhizal treatment), with which they are significantly different (at p ≤ 0.05) if followed by different superscripts (n = 4)

Non-mycorrhizal
		Peat (%)
		0	2	4	6
Sand (%)	100	12.89 (3.47)^a	33.2 (12.7)^b	42.5 (6.46)^b	33.3 (7.63)^b
	90	33.2 (12.7)^b	42.5 (6.46)^b	33.3 (7.63)^b	11.5 (0.00)^a
	80	8.99 (4.64)^a	13.5 (9.83)^a	18.7 (3.89)^a	25.1 (5.28)^b
	70	22.7 (5.82)^b	18.4 (2.97)^b	15.9 (4.55)^a	21.1 (7.68)^a
Rudua		46.5 (4.58)^b
Rengam		144 (39.1)^b

---

Mycorrhizal
		Peat (%)
		0	2	4	6
Sand (%)	100	266 (45.2)^b	335 (41.5)^b	215 (72.6)^b	255 (74.8)^b
	90	302 (137)^b	320 (48.7)^b	417 (63.3)^b	283 (30.7)^b
	80	524 (75.0)^b	299 (96.5)^b	366 (50.2)^b	344 (87.0)^b
	70	535 (106)^b	526 (85.0)^b	270 (23.8)^b	232 (35.8)^b

---

Mycorrhizal carrier medium
		Peat (%)
		0	2	4	6
Sand (%)	100	161 (38.1)^b	254 (105)^b	199 (94.9)^b	91.9 (37.0)^b
	90	226 (22.1)^b	186 (13.9)^b	142 (24.3)^b	140 (15.8)^b
	80	248 (64.7)^b	144 (28.7)^b	111 (9.93)^b	96.2 (25.0)^b
	70	481 (52.5)^b	240 (113)^b	142 (5.26)^b	121 (15.2)^b

---

Table 5 Fresh weights (mean and standard deviation) of Kudzu (Pueraria phaseoloides) in various artificial and natural soils (mg). All data are compared with the control (100% sand, no peat, no mycorrhizal treatment), with which they are significantly different (at p ≤ 0.05) if followed by different superscripts (n = 4)

Non-mycorrhizal
		Peat (%)
		0	2	4	6
Sand (%)	100	107 (26.1)^a	191 (130)^a	148 (21.3)^a	110 (45.8)^b
	90	671 (221)^b	337 (126)^b	239 (86.8)^b	214 (42.4)^b
	80	657 (98.1)^b	314 (97.1)^b	184 (102)^a	153 (40.9)^a
	70	817 (238)^b	483 (62.8)^b	204 (61.8)^b	106 (38.3)^a
Rudua		608 (173)^b
Rengam		8180 (2370)^b

---

Mycorrhizal
		Peat (%)
		0	2	4	6
Sand (%)	100	6940 (475)^b	9190 (1130)^b	9970 (981)^b	8580 (2400)^b
	90	7860 (1090)^b	10600 (1230)^b	9390 (1500)^b	10800 (1110)^b
	80	7800 (1340)^b	10000 (2200)^b	9360 (705)^b	6010 (1680)^b
	70	9550 (764)^b	10600 (970)^b	9260 (2540)^b	6420 (1650)^b

---

Mycorrhizal carrier medium
		Peat (%)
		0	2	4	6
Sand (%)	100	120 (28.9)^a	1100 (322)^b	1050 (189)^b	1280 (162)^b
	90	3290 (880)^b	2620 (633)^b	2310 (591)^b	1808 (499)^b
	80	3340 (489)^b	3110 (504)^b	2300 (591)^b	2330 (562)^b
	70	4000 (824)^b	3620 (513)^b	2830 (962)^b	2270 (216)^b

---

Three-way ANOVA indicated that all main factors and their interactions were significant at p ≤ 0.05 (data not shown). The separate mycorrhizal treatments were subsequently subjected to two-way ANOVA, indicating that all main factors and interactions were again significant within each mycorrhizal treatment, except that additions of clay (reductions of sand) had no significant effect on the dry weight of Pueraria when inoculated (Table 6).

Table 6 Al concentrations in foliage of Panicum miliaceum and Pueraria phaseoloides (μgg⁻¹)

		Panicum miliaceum				Pueraria phaseoloides
Non-mycorrhizal
		Peat (%)				Peat (%)
		0	2	4	6	0	2	4	6
Sand (%)	100	834	355	137	84.5	4530	2230	859	372
	90	852	611	625	307	1310	2330	1920	1260
	80	2350	1660	1590	1240	1010	3670	1470	1680
	70	1600	1700	2310	3620	1460	1640	1630	2420
Rudua		142				467
Rengam		79.8				319

---

Mycorrhizal
		Peat (%)				Peat (%)
		0	2	4	6	0	2	4	6
Sand (%)	100	47.6	46.2	36.6	38.6	256	273	198	240
	90	87.0	112	70.6	53.2	319	388	311	269
	80	196	169	125	286	421	467	389	313
	70	228	57.8	194	193	311	314	303	285

---

Mycorrhizal carrier medium
		Peat (%)				Peat (%)
		0	2	4	6	0	2	4	6
Sand (%)	100	62.1	37.1	47.2	26.0	339	256	154	150
	90	151	167	36.4	129	402	492	267	276
	80	85.6	107	56.8	74.2	451	329	372	250
	70	84.9	67.4	120	181	536	584	235	238

---

In contrast to the findings for Panicum miliaceum, adding clay improved the growth of Pueraria phaseoloides under non-mycorrhizal conditions, although these improvements were only significant for two treatments (Table 6). Adding peat reduced growth, although this reduction was also only significant in two treatments, it seems to suggest that higher peat levels had an increasingly negative impact on Pueraria. Performance on the natural soils was only significantly greater in those plants established on the kaolinitic Rengam soil, further suggesting that this species favours clay conditions over sand, even when both are supplied with adequate moisture. When no mycorrhizal inoculum was added, Pueraria responded most strongly to the highest additions of clay (70%) and the lowest additions of peat (0%). This directly opposes the data for Panicum, indicating that species choice is absolutely crucial when considering the various rehabilitation regimes for mined land.

Improvements in plant growth as a result of mycorrhizal inoculation were immense (in some cases approaching a one hundred-fold increase in dry weight). These improvements generally masked any individual treatment effects, although additions of peat at 6% did reduce growth in both 70% and 80% sand mixes when compared with 0% peat treatments. Overall, neither peat nor clay additions improved growth to a significant degree.

Use of the sterile carrier medium also enhanced Puerariagrowth to a significant degree when compared with plants grown without mycorrhizal amendment. However, these improvements were not so great as those noted for inoculated specimens, indicating an extremely positive benefit obtained from mycorrhizal inoculation.

Plant chemistry

Data for the metal concentrations found in Panicum miliaceum and Pueraria phaseolodes are presented in ESI Tables S6a–6i using the format shown in Table 6.

Foliar concentrations of Al were higher in Pueraria phaseoloides than Panicum milicaeum, suggesting a greater tolerance for this element in the former plant species (Table 6). Further work would be necessary to determine whether Al partitioning between roots and shoots was different for the two species, but it is likely that Al sequestration within Panicum milicaeum roots is responsible for lower transfer into the aerial parts of this species. Elevated root concentrations of Al might also be responsible for the generally very poor growth of this species in the artificial soils. Adding peat resulted in significant reductions in foliar Al concentration, suggesting its adsorption.²⁰ The responses to clay additions are more problematic, and suggest increases in Al availability with each addition, which were somewhat offset by added peat. However, the beneficial (reductive) effects of peat diminish as clay contents increase, suggesting saturation of available adsorption sites within the organic fraction. Mycorrhizal plants had much lower foliar concentrations of this element, indeed, concentrations were buffered across all experimental soils. Whether this was due to mycorrhizal binding of Al within the rhizosphere⁶⁸ is unlikely, given the equivalent concentrations in plants grown in soils to which the sterile mycorrhizal carrier medium had been added. This suggests that it is the chemical characteristics of the carrier that are responsible for any apparent mycorrhizal reductions in Al uptake.

This pattern was largely repeated with Mn, with concentrations in Panicum milicaeum reducing with peat additions, and increasing with clay. However, while clay resulted in increased uptake by Pueraria phaseoloides, peat had little effect on uptake by this species, suggesting that conditions within the rhizosphere enabled plant access to Mn adsorbed by the organic matter – a pool that was not available to Panicum milicaeum. Concentrations in both species were generally higher in response to the artificial soils than plants grown on natural soils. Both mycorrhizal treatments resulted in reduced foliar uptake of Mn in both species, but concentrations were often lower in non-inoculated plants, suggesting that although the available Mn pool was reduced by the carrier medium, mycorrhizal fungi partially compensated for this. Lambert, Baker & Cole Jr⁶⁹ made a similar observation with corn (Zea mays) and soybean (Glycine max) plants, in which foliar Mn concentrations increased when plants were inoculated with arbuscular mycorrhizal fungi.

Crop response to Ni uptake was similar, with peat causing rapid reduction in foliar concentrations. Response to clay was not so straightforward, but its addition resulted in increased uptake by both species. These increases were greatly offset by peat and it appears that uptake reached a maximum after the first addition of clay, which was not subsequently exceeded. There was little difference between inoculated plants and those grown with added carrier medium and it is therefore impossible to divine any mycorrhizal effect (enhancing or retarding) on Ni uptake by the two species. Foliar uptake was much higher in plants grown on the artificial soils, when compared with those grown on natural soils.

Substantial reductions in Cu uptake in response to added organic matter were restricted to Panicum milicaeum, in which foliar concentrations were reduced when 2% peat was added. Further reductions with increased soil organic contents cannot be determined from available data. Lack of response in Pueraria phaseoloides again suggests that rhizosphere conditions in this species enable crop access to Cu/organic matter pools. Plants grown in the artificial soils exhibited excessive foliar Cu concentration when compared with plants grown on natural soils, and this was exacerbated by addition of clay. Mycorrhizal inoculation generally reduced foliar Cu concentrations, in both crop species. Exceptions occurred with Panicum milicaeum in soils to which 4% or more peat had been added. These plants showed increased foliar Cu concentrations as a result of inoculation, although limited data prevents statistical analysis of these responses. As with Mn, the reduction in foliar concentrations of Cu seen in the carrier medium treatment were not repeated in the inoculated treatment, suggesting mycorrhizal-mediated improvements in Cu uptake.

Uptake of Zn was reduced in both plant species as a result of adding peat to the experimental soils. However, as with Cu, adding clay resulted in increased uptake. These increases were somewhat mitigated by the organic matter, but with clay levels of 30%, the reductive effects of peat were eliminated. Responses due to inoculation were different for both species. Mycorrhizal inoculation increased foliar Zn concentrations in Panicum milicaeum, but reduced them in Pueraria phaseoloides.

Adding organic matter to soils has been noted to increase adsorption (and reduce availability) of As.¹⁷ This should result in reduced foliar concentrations in plants grown in soils amended with organic matter. From available data there is little evidence of this, with some increases and some decreases in baseline levels in both experimental species. Adsorption of arsenate on clay minerals (and their derivatives) should also reduce its availability, but there is little evidence for this here. What is apparent is an increase in concentrations among inoculated plants, when compared with those grown in soils amended with the carrier medium. This effect is more marked as soil organic contents increase, suggesting mycorrhizal accessing of the organic matter/As pool. Mycorrhizal fungi are known to proliferate in organic matter,⁷⁰ and the chemical similarities between arsenate and phosphate may result in increased uptake of the former from organic sources when the latter is in limiting supply.

Organic acids have been noted to increase mobility of rare earth elements in the rhizosphere, increasing their phytoavailability.⁶³ Indeed, Han et al.⁶⁴ demonstrated increases in La uptake by barley (Hordeum vulgare) as a result of adding low molecular weight organic acids. These findings are not repeated here, as added peat reduced foliar concentrations of this element in most experimental soils. Adding clay had a very limited effect on La uptake. Inoculation with mycorrhizal fungi generally resulted in increased foliar concentrations of La, although this was most apparent in soils to which 20% (or less) clay had been added.

The reductive effects of organic amendment are again demonstrated with foliar Pb and U concentrations. Soil organic matter is known to significantly reduce Pb availability,⁶⁰ and this is repeated here. In contrast, the presence of organic acids has previously been shown to increase U uptake⁷¹ but this is not seen here in either test species. The reductive responses in foliar concentrations of U and P to organic matter are particularly striking.

Increased uptake as a result of clay addition has been noted for other trace elements (Al, La, Mn and Zn). This suggests increased concentrations of available forms of these elements in the clay tailings. Elevated foliar concentrations of Pb (and U) are of particular concern, and the use of clay tailings as an amendment for sand tailings cannot be recommended without further field trials with edible crop species. If suitably amended with organic matter, then the availability of these elements is reduced, but whether this is sufficient to render any crops safe for consumption would need to be verified on a case-by-case basis.

Mycorrhizal inoculation appears to have little effect on foliar concentrations of Pb in Panicum milicaeum, while it may increase these concentrations in Pueraria phaseoloides. General effects for U uptake are impossible to determine from available data.

Conclusions

Interactions of nutrient and trace elements with soil organic matter and mineral fractions are complex. However, the most significant characteristics in determining trace element availability are thought to be soil organic matter, clay, and pH.⁷² All of these were manipulated in the experimental soils created during this study. Iron and Mn oxides may have also been indirectly manipulated through the amendment of sandy tailings with clay and peat waste.

Adding both clay tailings and peat improved the cation exchange capacity of sand tailings, but they also reduced pH, potentially leading to increased trace element release.⁵⁶ Although organic material can reduce the toxicity of Al in solution,²¹ there was no evidence that it was strongly bound by peat when added to sand tailings. Despite this, concentrations of Al were generally lower in the artificial soils than those found in a natural kaolisol. Adding peat to sand tailings did reduce Mehlich-3 extractable levels of As, La and U, all of which could reduce potential negative health impacts from agronomic species established on sand tailings. Further work would be needed to establish the beneficial trace element binding properties of different organic amendments used when sand tailings are adopted for agriculture, and to establish whether the different rhizosphere conditions of different crop species resulted in different mineral/organic matter interaction and trace metal uptake. Peat also increased available concentrations of Ca and Mg, the latter being particularly significant in comparison with natural soils. It's use for improvement of sand tailings would be recommended but for the general threat to local peat-dominated ecosystems, that may render its use unsustainable. Use of other bulky organic materials (especially by-products from the palm oil industry) would be preferable, and these are known to bring nutritional benefits to degraded soils.⁷

In contrast, use of clay tailings cannot be recommended. Although its addition improved extractable concentrations of most macro and micronutrient elements, it also increased Al, As, La, Pb and U levels, which exceeded background concentrations found in natural sandy and kaolinitic soils, the increases in Pb being particularly notable. Further work will be necessary to isolate the agronomic sinks of Pb, As and U before clay could be deemed a suitable amendment for sand tailings.

The mycorrhizal carrier medium had a number of significant effects on soil chemistry, including significant increases in pH, Ca and K. The former would impact greatly upon trace and nutrient element availability, while the latter might competitively reduce the rhizotoxic effects of Al in the rhizosphere. The medium also increased extractability of As, possibly as a result of the increase in soil pH. It is therefore clear that the commercial medium in which mycorrhizal fungi are distributed can have significant effects on the chemistry of soils to which it is applied. Great care must therefore be taken when undertaking experiments to examine mycorrhizal response to ensure that the carrier medium is taken into account.

Growth patterns for the two test species were often contradictory. Panicum miliaceum responded well to additions of peat, while growth was suppressed when clay was added. Significant effects for Pueraria phaseoloides were fewer, although generally this species responded favourably to additions of clay and negatively to additions of peat. This highlights the need for post-mining rehabilitation schemes to consider not only soil conditions, but the possible effects that these might have on the desired vegetation. Applying peat at a rate of 4% resulted in the most significant increase in dry weight of Panicum under non-mycorrhizal conditions, while additions of clay at 30% produced the best enhancement in Pueraria (although a large standard deviation meant that this improvement was not statistically significant). Halim⁵⁰ noted that Pueraria performed best on heavy soils, and this was borne out by its performance on the kaolinitic Rengam soil. Whether clay additions in excess of 30% would produce further improvements in the species' performance on sand tailings is unknown, but logistical and economic constraints would be likely to limit such excessive additions of clay to sand tailings regardless of any potential improvements in growth.

The results for both mycorrhizal treatments (live inoculum or sterile carrier medium) are less straightforward to interpret. Certainly, both resulted in greatly improved growth in both test species, with inoculated plants generally much larger than those grown with the carrier medium, which were (in turn) much larger than those grown in the artificial soils to which no mycorrhizal treatment was applied. Use of inoculum damped out the treatment effects in both plant species, with few significant differences in plant weights across the sixteen different soils. However, underlying responses to the different soil treatments were still apparent when the carrier medium was used. In Panicum, this translated into a reversal of the effects seen in the non-mycorrhizal soils, with improved growth through addition of clay (although not significant) and reduced growth through addition of peat. This was not observed in Pueraria, where the underlying tendency for suppression of growth with added peat, and enhanced growth with added clay was still seen (only the latter significantly so).

Clearly, the mycorrhizal carrier medium is exerting significant influence on the performance of the two test species on the artificial soils. Zeolites are known to influence trace metal availability⁴³ and might also contribute to the available nutrient pool, enhancing growth. Despite this, if the effects of the carrier medium were subtracted from the inoculated plants, then the growth of the latter would still have been significantly improved by the presence of the mycorrhizal fungi. The use of mycorrhizal fungi could therefore greatly enhance rehabilitation efforts on Malaysian sand tailings. Without inoculation, it was possible to improve plant performance by either adding clay or peat (depending on species). In nutrient-deficient mine soils, Pueraria would offer significant advantages over Panicum, since it would not only offer improved ground cover (enhancing atmospheric nutrient interception) but its leguminous status would also result in the accumulation of nitrogenous plant litter, improving conditions for subsequent invasion by other species. The performance of this species under mycorrhizal conditions may be a combined result of mycorrhizal and rhizobial activity. Although the commercial inoculum was intended to provide fungal material, its use of root fragments as a component of the inoculum might also have provided appropriate rhizobial inoculant. Field tests with this species would be necessary to determine whether rhizobial or mycorrhizal (or both) inoculants were most effective at improving its growth on Malaysian sand tailings.

The effects of different soil treatments were extremely variable. However, adding peat generally reduced trace element uptake while adding clay increased it. This suggests that clay tailings retain elevated concentrations of many trace elements and that these are readily exchanged into soil solution for subsequent plant uptake. Adding peat often mitigated these effects, but the resulting foliar concentrations still exceeded those of plants grown in uncontaminated soils, and the inhibitory effects of peat were often overcome in soils to which higher percentages of clay had been added.

Since Panicum milicaeum and Pueraria phaseoloides exhibited such contradictory responses to peat and clay makes it impossible to state that foliar uptake of trace elements is a determining factor in their growth. Pueraria phaseoloides generally had higher foliar concentrations of trace elements, but still produced adequate growth, which was increased with additional clay (and therefore, soil trace element loadings). Further work will be necessary to examine the root/shoot partitioning of these trace elements, since it is possible that Al and trace element concentration in Pueraria phaseoloides foliage produces a trace element depletion zone within the rhizosphere, thereby avoiding any potential phytotoxic effects. Such mechanisms have been demonstrated elsewhere for the avoidance of Al stress,⁷³ and may well create trace element depletion zones as a by-product. In contrast, the reduced growth of Panicum milicaeum in soils to which clay had been added (and the increased foliar concentrations of Al and trace element which result) may indicate general intolerance for chemical stress within this species.

Inoculation with mycorrhizal fungi produced fewer net effects than might be expected from the literature. Copper, La and Mn uptake were improved by inoculation in both species, while increasing Zn uptake in Panicum milicaeum, but reducing it in Pueraria phaseoloides. More significantly, inoculation increased foliar As concentrations in both species, regardless of other soil treatments. It also increased Pb uptake by Pueraria phaseoloides. It was not otherwise possible to separate the effects of inoculation from the effects of adding the sterile carrier medium. Further work is necessary to provide evidence that it is the symbiotic fungi themselves which are altering trace (and nutrient) element uptake by plants, and whether there is any difference in fungi/root/shoot partitioning of these elements between the different plant species.

Combining responses for growth and foliar chemistry, it seems that Panicum miliaceum cannot tolerate the additional Al and trace element burden provided when clay tailings are mixed with sand, while Pueraria phaseoloides (adapted to acidic, high Al soils) responds favourably to clay additions, but harbours greater foliar concentrations of Al and trace elements as a result. Depending upon the required end use of sand tailings, great care should be exercised when selecting both soil ameliorant and potential crop species. Given the pressures on tropical peat resources, it is not appropriate to recommend their use in rehabilitation work unless they are produced as a by-product of the mining process and would otherwise be discarded. Since organic matter generally reduces plant uptake of Al and trace elements, its use is recommended over and above that of clay tailings, although in combination, the chemical influence of the clay may be overcome by the added organic material. Of most concern is As, which was not influenced by peat or clay additions. It had been expected that secondary clay minerals within clay tailings might strongly adsorb arsenate, but this did not occur – suggesting the saturation of appropriate binding sites. Increases in As uptake as a result of mycorrhizal inoculation must be closely monitored in field crops, especially those to which phosphate fertilisers have been added.

Acknowledgements

We are grateful to the Economic Planning Unit of Malaysia for permitting the fieldwork and CEO Timah Langat Holdings Sdn Bhd and SEK Mining

References

1. HSSI, Kinta River Flood Mitigation and Catchment Area Rehabilitation Project, 1st Edition, Ministry of Agriculture, Kuala Lumpur. 1994.
2. F. Birkinshaw, Malay J. Agric., 1931, 19, 470–475.
3. S. Appanah, & G. Weinland, Planting Quality Timber Trees in Peninsular Malaysia, 1st Edition, FRIM, Malaysia, 1993.
4. L. H. Ang, H. T. Chan, & H. A. Darius, Proc. Conf. on Forestry and Forest Products Res., Kuala Lumpur, Malaysia, 1–2 November, 1993. pp 270–278.
5. B. A. Mitchell, Malayan Forester, 1957, 20, 104–107.
6. W. Y. Wan Abdullah, M. H. Ghulam, & A. B. Othman, Physical and chemical properties of tin tailings in MARDI Station, Kundang, Selangor, 1st Edition, MARDI, Kuala Lumpur. 1992.
7. P. Vimala, A. Zaharah, C. S. Lee., & A. B. Othman, Proc. of the National Seminar on Ex-Mining Land and BRIS Soil: Prospects and Profit, Oct 1990, Serdang, Malaysia.
8. K. H. Lim, & G. Maesschalck, Proc.of the Conf. on Soil Science and Agricultural Development in Malaysia, Kuala Lumpur, Malaysia. May 1980. pp 93–105.
9. L. H. Ang, T. B. Ang, & L. T. Ng, Proc. of the Malaysian Sci. and Techno. Cong, Kota Kinabalu, Malaysia. Nov. 1998. pp 116–123.
10. M. O. Schwartz, S. S. Rajah, A. K. Askury, P. Putthapiban and S. Djaswadi, Earth-Sci. Rev., 1995, 38, 95–293.
11. J. Shamshuddin, M. Nik and S. Paramananthan, Pertanika, 1986, 9, 89–97.
12. M. T. F. Wong, P. Gibbs, S. Nortcliff and R. S. Swift, J. Agric. Sci, 2000, 134, 269–276.
13. I. M. J. Maene, C. K. Mok, & K. M. Lim, Proc. Conf. on Chemistry and Fertility of Tropical Soils. Kuala Lumpur, Malaysia. Nov 1973. pp 220–233.
14. V. Chantawong, N. W. Harvey and V. N. Bashkin, Water, Air, Soil Pollut., 2003, 148, 111–125.
15. F. J. Hingston, A. M. Posner and J. P. Quirk, Discuss. Faraday Soc., 1971, 52, 334–342.
16. U. Lund and A. Fobian, Geoderma, 1991, 49, 69–72.
17. X. Cao and L. Q. Ma, Environ. Pollut., 2004, 132, 435–442.
18. A. Karaca, Geoderma, 2004, 122, 297–303.
19. Q. B. He and B. R. Singh, J. Soil Sci., 1993, 44, 641–650.
20. R. J. Haynes and M. S. Mokolobate, Nutr. Cycling Agroecosyst., 2001, 59, 47–63.
21. E. Diatloff, S. M. Harper, C. J. Asher and F. W. Smith, Aust. J. Soil Res., 2001, 36, 913–919.
22. R. L. Zimdahl, JAPCA J. Air Waste Manage. Assoc., 1976, 26, 655–660.
23. K. Jayachandran, A. P. Schwab and B. A. D. Hetrick, Soil Biol. Biochem., 1992, 24, 897–903.
24. A. C. Gange, V. K. Brown and G. S. Sinclair, Funct. Ecol., 1993, 7, 616–622.
25. E. George, K. Häussler, S. K. Kothari, X. L. Li, & H. Marschner, In: D. J. Read, D. H. Lewis, A. H. Fitter & I. J. Alexander (Eds). Mycorrhizas in Ecosystems, 1st Edition, CAB International, Wallingford. pp 42–47. 1994.
26. T. Paré, E. G. Gregorich and S. D. Nelson, Plant Soil, 2000, 218, 11–20.
27. R. M. N. Kucey and H. H. Janzen, Plant Soil, 1987, 104, 71–78.
28. D. A. Jasper, A. D. Robson and L. K. Abbott, Aust. J. Bot., 1987, 35, 641–652.
29. G. Cuenca, Z. De Andrade and G. Escalante, Biol. Fertil. Soils, 1997, 26, 107–111.
30. M. N. Shamshuddin, In: P. Mikola (Ed). Tropical Mycorrhiza Research. 1980, 1st Edition, Clarendon Press, Oxford. pp 242–244.
31. P. Nadarajah, & A. Nawawi, In: Proc. Asian Seminar on Trees and Mycorrhiza. Kepong, Malaysia, April, 1988. pp 117–125.
32. P. Reddell and A. R. Milnes, Aust. J. Bot., 1992, 40, 223–242.
33. Y. Setiadi, P. Jeffries and J. C. Dodd, Indones. J. Biotechnol., 2000, 359–67.
34. I. Weissenhorn, C. Leyval, G. Belgy and J. Berthelin, Mycorrhiza, 1995, 5, 245–251.
35. C. Leyval, K. Turnau and K. Haselwandter, Mycorrhiza, 1997, 7, 139–153.
36. X. Chen, C. Wu, J. Tang and S. Hu, Chemosphere, 2005, 60, 665–671.
37. C. Gonzalez-Chavez, P. J. Harris, J. Dodd and A. A. Meharg, New Phytol., 2002, 155, 163–171.
38. B. Chen, P. Roos, O. K. Borggaard, Y. G. Zhu and I. Jakobsen, New Phytol., 2004, 165, 591–598.
39. G. Rufyikiri, L. Huysmans, J. Wannijin, M. Van Hees, C. Leyval and I. Jakobsen, Environ. Pollut., 2004, 130, 427–436.
40. T. V. St. John, D. C. Coleman and C. P. P. Reid, Ecology, 1983, 64, 957–959.
41. M. Puschenreiter, O. Horak, W. Friesl and W. Hartl, Plant Soil Environ., 2005, 51, 1–11.
42. X. Querol, A. Alastuey, N. Moreno, E. Alvarez-Ayuso, A. García-Sánchez, J. Cama, C. Ayora and M. Simón, Chemosphere, 2006, 62, 171–180.
43. P. Castaldi, L. Santona and P. Melis, Chemosphere, 2005, 60, 365–371.
44. K. H. Lim, G. Maesschalck, L. Maene, & W. H. Wan Sulaiman, Reclamation of Sn Tailings for Agriculture in Malaysia, 1st Edition, Universiti Pertanian Malaysia, Kuala Lumpur. 1981.
45. S. Kauffman, W. Sombroek, & S. Mantel, In: A. Schulte & D. Ruhiyat (Eds). Soils of tropical forest ecosystems – characteristics, ecology and management, 1998. 1st Edition, Springer-Verlag, Berlin. pp 9–20.
46. W. P. Panton, Reconnaissance Soil Survey of Trengganu. 1st Edition, Ministry of Agriculture, Kuala Lumpur. 1958.
47. IUSS Working Group WRB. World reference base for soil resources. 2nd edition. World Soil Resources Reports No. 103. FAO, Rome. ISBN 92-5-105511-4. 2006.
48. Soil and Plant Analysis Council. Soil Analysis Handbook of Reference Maethods. 1st Edition, CRC Press Inc, Boca Raton, 2000.
49. H. N. van der Hoek, & P. C. M. Jansen, In: G. J. H. Grubben & S. Partohardjono (Eds). Plant Resources of South-East Asia No. 10: Cereals. 1996, 1st Edition. Backhuys Publisher, Leiden, The Netherlands. pp 115–119.
50. R. A. Halim, In: L. 't Mannetje & R. M. Jones (Eds). Plant Resources of South-East Asia No. 4: Forages. 1992. 1st Edition. Pudoc, Wageningen, The Netherlands. pp 192–195.
51. V. M. Palaniappan, J. Appl. Ecol., 1974, 11, 133–150.
52. C. L. Boddington and J. C. Dodd, Plant Soil, 2000, 218, 137–144.
53. M. Brundrett, N. Bougher, B. Dell, T. Grove, & N. Malajczuk, Working with Mycorrhizas in Forestry and Agriculture. 1996, 1st Edition, ACIAR, Canberra.
54. J. B. Jones Jr, Commun. Soil Sci. Plant Anal., 1998, 29, 1543–1552.
55. B. A. Zarcinas, B. Cartwright and L. R. Spouncer, Commun. Soil Sci. Plant Anal., 1987, 18, 131–146.
56. D. L. Sparks, Environmental Soil Chemistry, 2nd Edition, Academic Press, San Diego, USA. 2003.
57. M. Y. Sulaiman, Geoderma, 1991, 127, 293–304.
58. L. C. Baillie, In: P. W. Richards. The Tropical Rain Forest, 2nd Edition, Cambridge University Press, Cambridge. 1996, pp 256–286.
59. J. J. Wang, D. L. Harrell, R. E. Henderson and P. E. Bell, Commun. Soil Sci. Plant Anal., 2004, 35, 145–160.
60. A. T. de Matos, M. P. F. Fontes, L. M. da Costa and M. A. Martinez, Environ. Pollut., 2001, 111, 429–435.
61. E. Tipping, Geoderma, 2005, 127, 293–304.
62. U. Lund and A. Fobian, Geoderma, 1991, 49, 83–103.
63. X. Q. Shan, J. Lian and B. Wen, Chemosphere, 2002, 47, 701–710.
64. F. Han, X. Q. Shan, J. Zhang, Y. N. Xie, Z. G. Pei, S. Z. Zhange, Y. G. Zhu and B. Wen, New Phytol., 2004, 165, 481–492.
65. H. Shahnandeh and L. R. Hossner, Water, Air, Soil Pollut., 2002, 141, 165–180.
66. J. G. Salinas, J. I. Sanz and E. Sieverding, Plant Soil, 1985, 84, 347–360.
67. W. J. Fitz and W. W. Wenzel, J. Biotechnol., 2002, 99, 259–278.
68. S. D. Koslowsky and R. E. J. Boerner, Environ. Pollut., 1989, 61, 107–125.
69. D. H. Lambert, D. E. Baker and H. Cole Jr, Soil Sci. Soc. Am. J., 1979, 43, 976–980.
70. H. Marschner and B. Dell, Plant Soil, 1994, 159, 89–102.
71. J. W. Huang, M. J. Blaylock, Y. Kapulnik and B. D. Ensley, Environ. Sci. Technol., 1998, 32, 2004–2008.
72. S. M. Ross, Toxic Metals in Soil-Plant Systems. 1994, 1st Edition, John Wiley & Sons, Chichester.
73. H. Matsumoto, In: M. N. V. Prasad & K. Strzałka. Physiology and Biochemistry of Metal Toxicity and Tolerance in Plants. 2002, 1st Edition, Kluwer Academic Publishers, Dortrecht. pp 95–109.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c1em10578g

---