Site description

The Pingle Manganese mine (long 110°46'E, lat 24°39'N) is located in Pingle county, about 108 km southeast from Guilin (Fig. 1), a well-known world-class karst scenic city in South China. This medium-sized mine lies in a hilly land (<180 m elevation), firstly extracted with opencast mining in 1958 and now left with a severe destruction of natural environment. The mine area is located in the middle subtropical monsoon climatic zone. The annual average temperature in this region is 19.9 °C with yearly average rainfall of 1636 mm and mean relative humidity of 79%. The regional vegetation is the typical subtropical evergreen broadleaf forest, and zonal soil is loess, usually 8-15 m deep covering the mine surface.

Figure 1.

Map of Guangxi, showing the study site (Pingle Mn mine) and a nearby Mn mine (Lipu) about 35 km west.

Sample Collection and preparation

Extensive ecological surveys and sampling were carried out in Pingle Mn mine from March 2005 to October 2006. According to the actual landform and vegetation the whole mineland was divided into four areas: the active mining area (MA), fresh tailings dam (FTD), reclaimed tailings dam (RTD) and restored area (RA). The active mining area is the highest among the first three covered with loose striped soil and overburden, but in the neighboring unexcavated area a good vegetation cover is seen with Imperata cylindrical var. major (a grass) dominating and some Pinus massoniana trees and Schima superba saplings standing out. Downward is a fresh tailings dam in use and some pioneer plants such as Phytolacca acinosa (herb) and Amaranthus tricolor (herb) colonizing. The reclaimed tailings dam (lowest in height) is a former tailings disposal pond but now reclaimed for use of orchard (mainly peach trees and Chinese chestnut). Underlain are natural herbaceous species, e.g. Erigeron acer, Panicum repens and Cynodon dactylon and the vegetation cover is about 80%. The restored area, a terraced hilly slope, is a large peach tree field (initiated by the local county government) and the overall vegetation coverage is around 30% with P. acinosa, I. cylindrical var. major, Hedyotis auricularia and Digitaris sanguinalis sparsely underlain. The peach was said (by growers) insipid, and the fruit size and annual yield smaller than usual.

Four soil sampling sites (MA, FTD, RTD and RA) were set up in the mineland, one site in each area. Samples with three replicates were collected from surface soil (0-20 cm) and usually 5 subsamples were merged into one single sample. Samples of major dominant plants (divided into leaf, shoot and root where appropriate) for each site were collected and 3-5 multi-point subsamples nearby were gathered and mixed into one sample. All soil and plant samples were sealed with polythene bags in the field and transported into the laboratory.

Soil samples were air-dried, homogenized and sieved through a 2-mm screen, then pulverized (with TAISITE FW100 stainless-steel pulverizer) and passed through a 100-mesh nylon sieve. Plant samples were gently washed with tap water, rinsed three times with deionized water, first dried at 105 °C for 30 min, and then at 70 °C to constant weight. Dried plant samples were ground in the same stainless steel pulverizer to fine powder.

Chemical analysis

Soil pH value (1:1 soil to distilled water, w/v) was measured using a pH meter. Soil organic matter (OM) was digested with H₂SO₄ + K₂Cr₂O₇ and determined with phenanthroline titration. Total nitrogen (TN) was determined by micro-Kjeldahl digestion followed by steam distillation. Total phosphate (TP) was determined colorimetrically by the molybdenum blue method. Soil samples were digested with concentrated HCl + concentrated HNO₃ HF + HClO₄ (10:5:5:3, v/v) and plant tissues digested with concentrated HNO₃ + HClO₄ (20:3-5, v/v) for metal determination (Li et al. 2007). The total metal concentrations (Cd, Cr, Cu, Pb, Zn and Mn) in digestates were determined by the atomic absorption spectrophotometer with flame atomizer (WFX-110, Beijing). The “plant—available” portion of metals in soil was extracted with 0.1 M HCl solution and determined with the same method. Quality assurance of metal analysis of soil and plant materials was performed using recovery rate (86%-115%) of the added certified reference materials (provided by Guangfu Fine Chemical Research Institute) into the digested solution.

Pollution assessment

The pollution index (Pi) and geo-accumulation index (I-geo) were employed to assess the pollution of individual metal in the Mn minesoils. Pollution index was calculated using the formula:

China Environmental Quality Standard for Soils (GB 15618-1995 Class II, pH < 6.5) was used as the assessment criteria (the standard values are: Pb, 250 mg/kg; Zn, 200; Cu, 150; Cr, 150; and Cd, 0.3) in this study since these figures indicate threshold levels of pollution warning. No value was given for Mn in this standard, thus Pi (Mn) was not calculated.

Based on Pi, soil contamination levels were classified into four grades: Pi < 1 (grade 1), unpolluted; 1 ≤ Pi < 2 (grade 2), slight pollution; 2 ≤ Pi < 3 (grade 3), medium pollution; Pi ≥ 3 (grade 4), heavy pollution. I-geo was classified into seven grades: I-geo ≤ 0 (grade 0), uncontaminated; 0 < I-geo ≤ 1 (grade 1), slightly moderately contaminated; 1 < I-geo ≤ 2 (grade 2), moderately contaminated; 2 < I-geo ≤ 3 (grade 3), moderately severely contaminated; 3 < I-geo ≤ 4 (grade 4), severely contaminated; 4 < I-geo ≤ 5 (grade 5), severely extremely contaminated; I-geo > 5 (grade 6), extremely contaminated (Müller, 1969).

Bioaccumulation Factor and Biological Transfer factor

For each metal, Bioaccumulation Factor (BAF) of a plant is the heavy metal concentration in aboveground part (mainly leaf or leaf plus shoot where appropriate) divided by the metal concentration in soil, and this is used to evaluate the plant's uptake capacity of the metal from soil. Biological Transfer Factor (BTF) is the heavy metal content of aboveground tissues divided by the metal content in roots, which indicates the transfer efficiency of a metal to the aerial parts.

Statistical analysis

All statistical analyses were performed using SPSS 12 for Windows. One-way ANOVA was carried out to compare the difference of means from various sampling sites, followed by multiple comparisons using the least significant difference (LSD) test. The level of significance was set at P < 0.05 (two-tailed).

Nutrient Status and Heavy Metals in minesoil

The soil nutrients, total heavy metal concentrations and 0.1 M HCl-extractable fractions were presented in Table 1. Soil pH ranged from 4.65 to 5.59 and averaged 5.03, indicating an acid nature. Significant variations existed among the four sites for the macronutrients. The active mining area had the highest N and organic matter (OM) contents probably because this area was freshly excavated from hilly slope with a good vegetation cover. The mean soil TN and OM barely fell into “moderate” class (0.8-1.0 g/kg for TN and 10-15 g/kg for OM) according to China Soil Fertility Standard for Green Food Production (NY/T 391-2000). However, when TP was concerned, this minesoil was considered very low (averaged 0.41 g/kg) since soil TP less than 0.8-1.0 g/kg can limit plant growth (Zhejiang Agricultural University 1991).

The minesoils were characterized by elevated levels of all the studied metals compared with the local (Guilin) and China soil background concentrations. The highest total metal levels were observed in the fresh tailings dam for all metals, especially for Mn, Pb and Cd for which concentrations were 80.9, 11.8 and 96.1 times higher than their local soil background values respectively. Basically heavy metal concentrations followed the order of Mn > Cr > Pb > Zn > Cu > Cd in the minesoil whereas in natural soil, the order was Mn > Cr ∼ Zn > Cu ∼ Pb > Cd. All the extractable metal fractions were below 5% of their total concentrations, indicating a pretty low availability (exchangeable + potentially available) to higher plants. Among the metals, Cd, on average, had the highest percentage (3.55%) of extracted portions while Cr the lowest (0.19%). A correlation analysis between the total and extractable fractions revealed no significant correlation existed for Pb, Cr and Cd, but positive correlations (p < 0.05) found for Mn (r = 0.81), Zn (r = 0.90) and Cu (r = 0.96), respectively.

Pollution Assessment of minesoil

The pollution assessment results with Pi and I-geo were shown in Table 2. Pi indicated that the mine-soil was not polluted by Pb, Cu and Zn, but was slightly polluted by Cr and seriously polluted by Cd. Fresh tailings dam had the greatest Pi and I-geo values, demonstrating this site was more severely contaminated than others.

For the five metals, all I-geos showed severer pollution than Pi, but the pollution trend of different metals remained. For instance, slightly moderate to moderately severe contamination were identified with I-geo for Cu, Pb and Zn, while no pollution were assessed as with Pi, and conversely, a moderate contamination for Cr with I-geo while a slight pollution was indicated by Pi. For Cd, both assessments identified the heaviest grade of contamination (Table 2).

Table 1.

The nutrients and heavy metal concentrations of minesoils (mean ± SE, n = 3) in Pingle Mn mineland. Data in parenthesis indicate the percentage of extracted fractions to the total metal concentrations.

Table 2.

The pollution (Pi) and geo-accumulation (I-geo) indices of minesoils and their pollution grading.

As for Mn, there was no stipulated value in China soil quality standard, but the calculated I-geos were 4.98, 5.75, 3.81 and 2.66 respectively for the mining area, fresh tailings dam, reclaimed tailings dam and restored area. These results reflected moderately severe to extreme contamination existed. Also, a reported appropriate Mn range in soil was 170-1,200 mg/kg (Wang, 1995), but Mn levels in this mineland were about 1.8-14.9 times of the upper limit, supporting that the minesoil was contaminated with Mn, especially in the active mining area and fresh tailings dam.

Phytoaccumulation and Transfer of Heavy metals

Heavy metal contents in the dominant plants were showed in Table 3. Different plants and different tissues of the same species varied greatly in heavy metal concentrations with Mn ranging from 1.3-9975.6 (average at 1341.5) mg/kg, Pb < limit of detection-51.5 (26.4) mg/kg, Zn 1.6-121.4 (46.3) mg/kg, Cd 0.01-11.39 (3.82) mg/kg, Cr 1.0-100.3 (41.1) mg/kg, and Cu 0.3-55.6 (11.8) mg/kg. In general, the normal heavy metal contents of terrestrial plants growing in uncontaminated soil were in the range of 0.1-41.7 mg/kg for Pb, 1-160 mg/kg for Zn, 0.2-0.8 mg/kg for Cd, 0.2-8.4 mg/kg for Cr, 0.4-45.8 mg/kg for Cu (Cao and Chi, 2001) and 1-700 mg/kg for Mn (Gerber et al. 2002). Thus, for these mineland plants, all Cu and Zn contents fell within the normal range, and some Mn and a few Pb levels above the range while all Cd and almost all Cr levels greatly exceeded the upper limit of the normal range, showing they accumulated higher Cd and Cr from substrate even though the bioavailability is low in relation to the total Cr.

Table 3.

Heavy metal concentrations of the plant tissues for the dominant species in Pingle mineland.

Table 4 presented the Bioaccumulation Factor (BAF) and Biological Transfer Factor (BTF) of the dominant plants. Almost all BAFs were less than 1.0 except the Mn hyperaccumulator P. acinosa in the restored area for Mn and E. acer, P. acinosa and H. auricularia for Cd. On average, plants tended to have a relatively higher BAF for Cd (0.66) but lower BAF for Pb (0.13). However, all plants showed high Mn transfer rate (BTF = 5.6) while Pb and Cu were few transported to the above-ground parts. In addition, S. superba and P. acinosa demonstrated unusually high Mn relocation ability and high Cd transfer feature as well.

Soil Pollution and phytotoxicity

In general, minesoil is characterized by the elevated heavy metal concentrations, deficiency of major nutrients, poor physical structure, extreme acidity and/or salinity, drought, and other stressful factors (Li, 2006; Shu et al. 2005; Wong and Luo, 2003; Yang et al. 2003). Heavy metal pollution of minesoil is a widely documented problem (Liu et al. 2004; Lei et al. 2005; Xie et al. 2005) but has been largely overlooked in China because they are usually remote to the densely populated urban areas. For example, in the five Pb/Zn mineland in south China, very high Pb, Zn, Cu and Cd concentrations were found in the substrates with plant communities (Shu et al. 2005); Investigations on the six tailings dams (mainly Sn and Sb mines) in northwest Guangxi showed very high levels of Pb, Zn, As, Cd, Cu, Sb and Sn in tailings substrate (Zhou et al. 2003). In a slag wasteland of Xiangtan Mn mine in the neighboring Hunan Province, minesoil was contaminated with Mn (7990.2 mg/kg), Pb (401.15 mg/kg), Zn (640.32 mg/kg), Cd (13.15 mg/kg) and Ni (91.33 mg/kg), especially most severely with Cd and Mn (Fang et al. 2006).

The pollution index (Pi) was extensively used for soil pollution assessment in China (Chen et al. 2005; Fang et al. 2006; Li et al. 2005; Zheng et al. 2006) and proved to be simple and effective because it related directly to a certain quality criterion. I-geo was also employed for evaluation of heavy metal contamination (Yang et al. 2004a; Li et al. 2006; Chen et al. 2006) of sediment when its natural background value was available. In this study, I-geo assessment indicated a severer pollution of minesoil than Pi, the similar scenario also reported by Hu (2003) who used a modified correction factor (1.0 instead of 1.5). The difference of the two assessments lies in that I-geo linked theoretically all “pollution contribution” to the single soil background value only, not discriminating the natural lithogenic effect from anthropogenic contribution within the mineland context. The high metal levels of minesoil may be caused by a combination of naturally elevated background value owing to the parent rock weathering with the extra released metals through mining (pollution). The index of geoaccumulation considered them all as pollution contribution, thus overestimated the real pollution effect. To make the I-geo assessment results comparable for minelands, the real background value of minesoil (usually higher than the local natural soil) must be reinvestigated or the correction factor be properly modified.

Table 4.

Bioaccumulation Factor (BAF) and Biological Transfer Factor (BTF) of the dominant plants in Pingle mineland.

Phytotoxicities have generally been associated with 3-8 mg/kg total Cd in soil, 60-125 mg/kg Cu, 1500-3000 mg/kg Mn (Ross, 1994), 500-1000 mg/kg Pb (Chaney, 1993), or 300 mg/kg Zn (Lepp, 1981). In this mineland, soil Mn, especially in the tailings dam and mining zone, is potentially phytotoxic and may constrain initial colonization by pioneers (Ye et al. 2002). However, it is widely accepted that the total metal concentration in soil overestimates the risk of phytotoxity, but in the present study, despite the low relative extractability of the metals, the bioavailable concentration of some metal is marginally high and surpass the thresholds of phytotoxity. In addition, several studies have shown a poor correlation between plant uptake and total metal levels in soil (Kabata-Pendias, 1993; Remon et al. 2005) mainly due to the various chemical association forms with different bioavailability (Maiz et al. 1997, 2000). Moreover, even the bioavailable metals are poorly correlated to the metals absorbed by plants (Wang et al. 2003), as the latter can modify the soil conditions in the rhizosphere, and may develop specific physiological strategies for excluding and accumulating metals. Using the same soil extraction scheme (0.1 M HCl) for Lipu Mn mineland (about 35 km away from Pingle), the mean mobilizable fractions were 1.2% for Pb, 2.8% for Cu, 2.4% for Cd, 1.6% for Zn and 3.1% for Mn (Li et al. 2007), the results agreeing well with those from this study.

Phytoremediation and Restoration of mineland

Phytoremediation is defined as the use of plants to remove pollutants from soils or to render them harmless (Salt et al. 1998). Within the mineland context, phytoremediation technologies basically include phytoextraction in which hyperaccumulator plants remove metals from the abandoned substrates and concentrate them in the harvestable parts of plants (Kumar et al. 1995), and phytostabilisation where plants reduce the mobility and bioavailability of metals in the sediments either by immobilization or by prevention of migration (Smith and Bradshaw, 1972; Vangronsveld et al. 1995). Among the studied plants in this mineland, P. acinosa (a perennial herb) is a reported Mn hyperaccumulator and its maximum Mn content in leaves reached 19,300 mg/kg (Xue et al. 2004). However, the leaf Mn content was only 7122.4 mg/kg in this mineland, the difference may be caused by the great variation of soil Mn concentrations (52,359-114,013 mg/kg in Xiangtan vs. 2,095.6-17,839.7 in Pingle). Another plant of interest is S. superba, a wide distributed tree species commonly used for forest firebreak, contained 9975.6 mg/kg Mn in leaves. This result was very close to the suggested hyperaccumulator threshold by Baker and Brooks (1989). Furthermore, its BAF and BTF for Mn were 0.96 and 13.5, showing this woody plant could be a prospective Mn hyperaccumulator. A further pot growth experiment confirmed this species is a Mn hyperaccumulator (Yang et al. 2008).

While phytoextraction seems only promising for cleanup of slight to moderate polluted soil (Wong, 2003), phytostabilization is more realistic for remediation of minesites usually with higher heavy metal levels. A desired species for revegetating mine wasteland should best grow well on poor and harsh soils and develop vegetation cover in a relatively short time and accumulate biomass rapidly. In addition, native species with metals stored in roots (low BTF) are most beneficial to minimize dispersal. Shen et al. (2004) listed a wide range of plant species used for phytostabilization of minesites and evaluated their growth performances in different parts of China. In an almost barren slag wasteland of Xiangtan Mn mine, the initial natural herbaceous colonizers were Imperata cylindrical, Erigeron annuns, Cynodon dactylon, Paspalum ethunbergii, Arthraxon hispidus, Juncus effuses, and P acinosa (Yan et al. 2006). In the less harsh minesoil of Lipu Mn mine, Guangxi, 36 species (26 herbs and 10 woody) were found colonizing (Li et al. 2007), of which Digitaria sanguinalis, Erigeron canadensis, Urena lobara, Imperata cylindrica var. major, P. acinosa, Pteridium aquilinum var. latiusculum and Pteris multifida were dominant. These floras give some insights to the restoration design of Mn mine wasteland in Guangxi: firstly, the local pioneer species D. sanguinalis, C. dactylon, I. cyclindrica var. major, and E. canadensis could be employed to colonize and ameliorate the wasteland, and then an association with fast growing species C. henryi, P. massoniana and Eucalyptus spp. might restore some functional uses.

The aim of restoration of mine wastelands is to remediate ecological destruction and reduce pollution dispersion. If rehabilitation of mine spoils causes another serious pollution to human being, the loss outweighs the gain. A general tendency of mineland rehabilitation in China is the utilitarian reclamation for agriculture. From our investigation on the six major Mn minelands in Guangxi, more than 50 edible plants were found planting on the mine wasteland including Chinese chestnut, sugarcane, peanut, beans, potato and tubes, mandarin, orange, peach, Taiwan green Jujube, and tea trees. This conduct is more common in Guangxi than other areas of China probably owing to the poverty and severer shortage of cultivated land and to some extent is encouraged by the local government (Li et al. 2007). Worries arose because there were usually no protective treatments in place before planting and no monitoring of residual toxic metals was exercised before they enter the food chain. In the present study, both the peach fruit (P. persica) and Chinese chestnut (C. mollissima) greatly exceeded the Food Safety Standards of China in terms of the maximum allowable Cd, Cr and Pb levels (Table 3). Thus simple reclamation for agriculture use is of high risk unless sufficient treatments, e.g. separation layer, cover of guest soil over 50-cm deep (Wong and Luo, 2003), or substantial substrate amendments (Li, 2006), are in place before planting.

Other diverse restoration goals can also be considered. In Bayi Mn mineland of Guangxi, large areas were reclaimed with the fast-growing Eucalyptus forest for pulpwood while in Lipu Mine (Guangxi), some wastelands were reused as osmanthus (a local favorite gardening tree) nursery. This restoration mode has lower environmental risk and is more adaptive to different minelands. Furthermore, to leap out of the utilitarianism-oriented reclamation, biodiversity conservation, providing habitat to wildlife, checking soil and wind erosion and mitigating pollution dispersal, recreation and tourism, or beautifying the damaged landscape alone can become the aims of ecological restoration.