1 Introduction

Mining is economically important to producing regions and countries, but it is an activity that generates a large amount of wastes and poses a serious threat to the environment. The changes driven by mining on the landscape and local biodiversity are noticeable and persist for a long time when mined areas are not reclaimed. Soils impacted by mining activities pose serious risks to the health of exposed populations, either by ingestion of soil or contaminated food and water (Silva et al., 2019). In order to recover such areas, it is important to develop studies that promote the understanding of how degradation and contamination processes occurred (Yao et al., 2021), especially when comparing non-mining and mining areas in the same region and subject to the same environmental conditions. Understanding changes in soil characteristics and how these changes influence the spread of degradation facilitates the monitoring and decision-making for mined areas. Studies investigating the potential risk of soil contamination by heavy metals and their diffusion to aquatic systems and food crops are recurrent in the literature (Sun & Li, 2021; Xiang et al., 2019; Zheng et al., 2020). However, few studies address the change in soil physical and chemical characteristics after mining shut down.

The calcium tungstate mineral scheelite (CaWO4) is an important ore of tungsten. The largest scheelite deposits in Brazil are located in the country’s semi-arid region (Dantas, 2002). These mines process scheelite ores and generate large amounts of tailings and overburden accumulated in heaps exposed to wind and rainwater. A previous study has shown that domestic and agricultural water supplies in the region have sediments originating from these mining waste heaps that altered the water quality (Petta et al., 2014); however, no study has addressed the consequences of this scheelite mining on soil quality.

Several scheelite mines in the region studied were closed, but no recovery measures were implemented and the exploited sites were naturally covered by indigenous vegetation. The spontaneous vegetal cover can impair the contamination spread and ameliorate the chemical, physical, and microbiological characteristics of the soils (Xiang et al., 2019; Zheng et al., 2020). In this scenario, more studies are needed to investigate how vegetation cover in semi-arid settings can contribute to the reclamation of mined areas.

The socioeconomic importance of mining activity is evident, but environmental impacts should not be neglected. Therefore, investigations into mining impacts on soil quality, including heavy metals inputs in the soils, are crucial to guarantee that human health and environmental risk are taken into account. The risk assessment is particularly relevant to naturally fragile ecosystems, such as the Brazilian semi-arid region. In this scenario, the objective of this work was to evaluate the changes in soil physical and chemical characteristics caused by scheelite mining in a semi-arid setting in northeastern Brazil. The mining impacts on the soil background concentrations of the metals Cd, Cr, Cu, Ni, Pb, Zn, Fe, and Mn were also assessed and compared with regulatory levels adopted in the country. Our data provide useful information for the monitoring of soil quality not only in the study area but also in mining sites under similar tropical settings.

2 Materials and methods

2.1 Study area

The scheelite mines are located in Currais Novos (6°15′39″S, 36°31′04″W), Rio Grande do Norte State, northeast Brazil (Fig. 1). The mines lie in the Seridó Scheelite Province, which holds the largest reserve of calcium tungstate–scheelite (CaWO4) in Brazil (Dantas, 2002). Most of the deposits were discovered in the early 1940 and, since then, have been exploited for tungsten. The climate is hot semi-arid (BSh) (Alvares et al., 2013), characterized by scarcity and uneven rainfall distribution, with a mean annual precipitation of 610 mm and a rainy season from February to April. The vegetation is the hyperxerophilic Caatinga, and Leptsols are the predominant soils (FAO, 2015). This soil order is characterized by very shallow profile depth and large amounts of gravel.

Fig. 1
figure 1

Location of the study area and soil sampling sites. Post-mining (A and B) and mining sites (C and D)

Full size image

The soil characteristics were investigated along a temporal sequence of scheelite mining: non-mining, mining, and post-mining. The activities at the mines began in 1947, were interrupted in the late 1970s due to the price fall of tungsten in the ore’s international market, and returned in 2005 with the prospect of reusing the scheelite waste in old tailings. Mining waste dumps accumulated since the beginning of the mining operations have been exposed for the last 70 years. At least two of these dumps cover an area of 121,500 m2, at a volume of 1,943,200 m3, totaling 3,110,400 tons (Petta et al., 2014).

The current non-mining area has no previous mining and is covered by native vegetation (Fig. 1a); it is located in the surroundings of the mining areas and can hence serve as a reference area. The mining areas have been used for scheelite exploration since 1947, with a 15-year break from operations. The post-mining areas are sites where operations were shut down in 1976, i.e., roughly 44 years ago. The piled waste was left on the abandoned sites with no measures taken for ecological restoration. The scheelite processing includes fragmentation, gravimetric concentration, and roasting followed by electromagnetic separation. Coarse and fine tailings and metallic rejects are deposited together in the same tailings pile (Fig. 1).

2.2 Soil sampling and analysis

Five soil sub-samples (0–20 cm) were collected from different locations within each area to form a composite sample to analysis. Thirty-six composite soil samples were collected in total and distributed as follows: non-mining (locations 1–3), mining (locations 4–24), and post-mining (locations 25–36) areas. The samples were air-dried, homogenized, and passed through a 2-mm sieve, and then chemically and physically characterized using standard methods (Teixeira et al., 2017).

Samples were analyzed for bulk density by the beaker method, while particle size distribution was determined through the hydrometer method. The pH was measured in water (1:2.5 ratio). The K+ and Na+ exchangeable contents were obtained by flame photometry after extraction with Mehlich-1. The contents of Ca2+ and Mg2+ were determined by titration after extraction with 1 mol L−1 KCl solution. The cation exchangeable capacity (CEC) was calculated from the sum of K+, Na+, Ca2+, and Mg2+. Potential acidity (H+  + Al3+) through titration after extraction with calcium acetate buffered solution at pH 7.0. Available P was measured by colorimetry after extraction with Mehlich-1; total N was obtained by the Kjeldahl method (Malavolta et al., 1997). Soil organic carbon (SOC) was determined by the modified Walkley–Black method (Silva et al., 1999). Soil C stock within 0–20-cm depth was calculated by multiplying soil C content by the bulk density and the depth of the soil layer (Veldkamp, 1994).

For quantifying the heavy metals, 0.5 g of soil was digested in Teflon vessels with 9 mL of HNO3 and 3 mL of HCl USEPA 3051A (USEPA, 1998) in a microwave oven (MarsXpress) for 8 min and 40 s on the temperature ramp, the necessary time to reach 175 °C. This temperature was maintained for an additional 4 min and 30 s. After digestion, all extracts were transferred to 50-mL certified flasks (NBR ISO/IEC), filled with ultrapure water (Millipore Direct-Q System), and filtered in a slow filter paper (Macherey Nagel®). High purity acids were used in the analysis (Merck PA). Calibration curves for metal determination were prepared from standard 1000 mg L−1 (Titrisol®, Merck). Sample analysis was done only when the coefficient of determination (r2) of the calibration curve was higher than 0.999. The concentrations of Cd, Cr, Cu, Ni, Pb, Zn, Fe, and Mn were determined with an atomic absorption spectrophotometer (Perkin Elmer AA 800). The quality of the analysis was assessed using spikes and reference material (SRM 2709a San Joaquin Soil) with certified values for metals; recoveries ranged from 87 to 103%.

The concentrations of Cd, Cr, Cu, Ni, Pb, and Zn in the sites were compared with the prevention value (PV) and investigation value (IV) established by CONAMA Resolution No. 420/2009 (CONAMA, 2009) and also with the Quality Reference Value (QRV) established by Preston et al. (2014) for the soils of the Rio Grande do Norte state. The PV corresponds to the metal concentration in the soil that poses risks to agricultural activities and crops contamination. In turn, the IV indicates that metal concentration in the soil is above tolerable risks to human health in agricultural, residential, or industrial scenarios.

2.3 Statistical analysis

The soil physical and chemical characteristics were analyzed for descriptive statistics analysis using Statistica v.7. Pearson’s linear correlation was performed to study the relationship between soil characteristics and heavy metal concentrations in soils. A two-way cluster analysis was accomplished to assess similarities among samples and sites. Principal component analysis (PCA) was performed using PC-ORD® v.6 to reduce the data mass and facilitate the choice of quality indicators in future environmental reclamation and monitoring programs.

3 Results

3.1 Soil physical and chemical characterization

The soils analyzed were predominately sandy, with textural class varying from loam-sandy to sandy-loam (Table 1). Mining and post-mining sites were sandier than non-mining and hence had the lowest clay and silt contents. Sampling locations 31 to 36 (post-mining sites) had higher clay contents close to the mean contents in the non-mining site (Table 1). The post-mining sites were probably areas of fine tailings disposal. The mean bulk density values were highest in mined areas, while mean density did not differ in post-mining sites and non-mining sites. Eight out of the twenty-one mining sites sampled showed density values above 1.75 g cm−3. Given that soil bulk density is a function of soil texture, bulk density in mining and post-mining were positively correlated with sand content and negatively correlated with silt. The higher the sand content of the deposited tailings, the higher the bulk density and the lower the silt content.

Table 1 Soil physical and chemical characteristics of non-mining, mining and post-mining sites in scheelite mines of NE Brazil. Bulk density (BD); total cation exchange capacity (CEC); soil organic matter (SOM); available phosphorus (P); total nitrogen (N)
Full size table

The increase in clay contents in post-mining sites was positively correlated with the potential acidity (r = 0.88; p < 0.01), CEC (r = 0.82; p < 0.01), organic matter (r = 0.61; p < 0.05), total N (r = 0.62; p < 0.05), Fe (r = 0.90; p < 0.01), Cu (r = 0.92; p < 0.01), Ni (r = 0.96; p < 0.01), and Cr (r = 0.89; p < 0.01). On the other hand, soil density (r =  − 0.63; p < 0.05), pH (r = 0.66; p < 0.05), available P (r =  − 0.62; p < 0.05), sand fraction (r = 0.90; p < 0.01), and Zn (r = 0.92; p < 0.01) were reduced. In non-mining and mining sites, the clay content did not show significant correlation with any other soil variable evaluated.

In the post-mining site, the soil physical characteristics showed less spatial variability than the mining areas (Table 1); post-mining sites also showed a greater standard deviation in eight out of the 15 soil chemical characteristics than the mining sites. These results show that the soil formed after mining activity was more physically and chemically heterogeneous. The soils in the mining area showed a higher pH, but lower organic matter, potential acidity, CEC, and N and P contents (Table 1), which suggests that alkalization and plant nutrient losses resulted from mining.

The non-mining and post-mining areas have similar soils, which show the vegetation effect on soil reclamation. Therefore, the values of pH, soil organic matter, CEC, and N and P confirm that vegetation improved the soil quality in the post-mining sites over the four decades of natural regeneration (Table 1).

In the post-mining sites, the reduction in soil pH was accompanied by an increase in soil organic matter (r =  − 0.72; p < 0.01), total N (r =  − 0.69; p < 0.01), clay content (r =  − 0.66; p < 0.01), potential acidity (r =  − 0.80; p < 0.01), and reduction in bulk density (r =  − 0.67; p < 0.05). The available P and total N contents were also lower in the mining sites. No correlation was found between P and organic matter, which indicates that P is released from mineral phases, while organic matter is the primary source of soil nitrogen.

We found that soil carbon stock was reduced by mining. Even after 40 years of natural regeneration, the soil did not recover its natural capacity to store carbon (Tables 1 and 2). Also, post-mining sites showed positive correlations of carbon stock with clay, total N, and potential acidity, but negative correlations with Zn, Pb, and pH (Fig. 2).

Table 2 Soil C stock studies in areas undergoing restoration after mining
Full size table
Fig. 2
figure 2

Pearson’s correlation between carbon stocks in the post-mining site soils and clay (a); pH (b); potential acitidy (c); total nitrogen (d); Pb (e); Zn (f)

Full size image

3.2 Heavy metal concentrations in soils

In general, the mining activities increased the local background of metals in soils. Therefore, mining and post-mining areas had higher concentrations of Fe, Mn, Cd, Cr, Cu, Pb, and Zn than the reference site (Fig. 3). Heavy metal concentrations in the non-mining site were similar to the QRV for metals in the soils of the state of Rio Grande do Norte (Fig. 3). In contrast, the mining and post-mining sites had heavy metal concentrations above the QRV in at least one sampled location.

Fig. 3
figure 3

Heavy metal concentrations in the soils of non-mining, mining, and post-mining sites in scheelite mines in northeastern Brazil: Cd (a); Cr (b); Cu (c); Ni (d); Pb (e); Zn (f); Fe (g); Mn (h). QRV (blue line) quality reference values for heavy metals in soils of Rio Grande do Norte state (Preston et al., 2014), PV (red line) prevention value (CONAMA, 2009), IV (green line) investigation value for agricultural scenario (CONAMA, 2009), IV (magenta line) investigation value for residential scenario (CONAMA, 2009)

Full size image

The concentrations of Cd, Cu, Pb, and Mn were on average higher in mining areas than post-mining sites. On the other hand, concentrations of Cr, Ni, Zn, and Fe were higher in the post-mining area. Cadmium, Cr, and Cu concentrations in at least one location showed values that exceeded the IV values for agricultural and residential uses. Lead and Ni values exceeded the regulatory values for agricultural use. The soil disturbance caused by mining is reflected in the higher spatial variability of heavy metal concentrations compared to areas not mined (Table 1).

3.3 Multivariate statistical analyses

The first two components of the PCA analysis explained 68% of the data variability (PC1 = 44%, p = 0.001; PC2 = 24%, p = 0.001), most important variables for ordering the first component were pH (0.86), potential acidity (− 0.83), bulk density (0.81), clay (− 0.81), N (− 0.81), SOM (− 0.78), P (− 0.78), silt (− 0.58), sand (0.52), Ni (− 0.73), Cd (0.70), and Cr (− 0.68). For the second component, the most important variables were Fe (− 0.92), Mn (0.83), Zn (− 0.80), Pb (− 0.67), and Cu (− 0.55). Axis 1 segregated the sample plots of the mining sites from the non-mining and post-mining plots (Fig. 4), mainly due to the higher concentrations of nutrients and organic matter, potential acidity, and clay, along with lower density and sand content. Axis 2 segregated the non-mining from the post-mining area because of the higher heavy metal concentrations.

Fig. 4
figure 4

Principal component analysis of the soil physical and chemical characteristics from scheelite mines in Brazilian semi-arid. Non-mining (circle); mining (triangle); post-mining (square)

Full size image

Two-way cluster analysis demonstrates similarities between soil characteristics and sampling sites (Fig. 5). There were two main groups with different soil physical and chemical characteristics and heavy metal concentrations. The first group is characterized by lower pH values and greater soil fertility, while the second group has higher pH values, soil contamination, and decreased concentration of plant nutrients. These variables were also considered to distinguish mining and after mining soils. The post-mining group integrated the sample plots of the non-mined areas, indicating that the spontaneous vegetation cover in this area led to soil changes that resemble non-mined areas.

Fig. 5
figure 5

Two-way cluster analysis diagram of soil physical and chemical characteristics and sampling sites from scheelite mines. On the matrix, dark squares represent maximum values while white ones represent minimum values for each property (columns) in sampling sites (lines). SOM; total CEC

Full size image

4 Discussion

4.1 Physical and chemical characteristics of soils as affected by mining activities

The increase in soil bulk density is expected in mined areas due to heavy machinery traffic during digging (Shrestha & Lal, 2011), removal of vegetation, and compaction caused by the overload of tailings and overburden deposited in the soil. Vegetation cover has a close relation with soil density, as roots promote soil structuring, making the soil less dense and more permeable (Zeng et al., 2014). The increased density may affect some major soil functions, such as water infiltration and retention capacity, aeration, resistance to root penetration (Mora & Lázaro, 2014), and nutrient availability (Wang et al., 2011). Given the restriction to plant development, the increase in soil density is the most significant physical factor limiting the growth of vegetation in mined areas (Schroeder et al., 2010). In six out of the twelve post-mining sites sampled, soil bulk density exceeded the critical value for root growth, corresponding to 1.75 g cm−3 in sandy soils (Jones, 1983). In addition to soil compaction due to heavy traffic, mining waste is formed by mixing particles with different sizes so that smaller particles can fill the spaces between larger particles, resulting in a decrease in total soil porosity (Shrestha & Lal, 2011). The compaction caused by heavy machinery during ore extraction processes can be minimized by using lighter equipment, reducing heavy vehicle traffic, practicing deep ripping, and applying organic materials (Shrestha & Lal, 2011).

In general, Leptsols are predominantly shallow, sandy soils owing to the low weathering index in semi-arid conditions. Thus, the higher clay contents in the post-mining site may indicate that pit opening resulted in the burial of more weathered top horizons into the soil profile. Moreover, waste was deposited in the area and exposed to chemical weathering for at least four decades, favoring the increase in clay contents. The soils from the non-mining and mining areas soil have similar textural characteristics, which suggest the potential for reuse of the waste as a substrate in degraded areas after mining activity. Pedogenesis in post-mining sites in semi-arid regions occurs very slowly, which increases the time required to reclaim these sites. However, the monitoring of soil physical and chemical characteristics can demonstrate the evolution of pedogenesis after the closing of mining activities (Ciarkowska et al., 2016).

The neutral to alkaline pH in the study area is because scheelite is a calcium tungstate mineral containing CaO (19.4%) (Godeiro et al., 2011). The calcite influence extends to the water bodies in the region. The predominance of calcite in sediments maintains the water pH > 7 in this region (Petta et al., 2014). Soil pH has a strong influence on nutrient availability. Under conditions of pH above five, the mobility of heavy metals in soil is reduced; on the other hand, the plant growth is impaired by micronutrients (mainly Fe and Mn) deficiencies (Barros et al., 2010). Also, alkaline soil simply severe restrictions on plant growth (Negrão et al., 2017). Therefore, indigenous plant species are recommended to revegetate the studied sites, once these species have mechanisms to strive in harsh semi-arid conditions.

The reduction in SOM was expected in the mining areas due to tailings deposition and the absence of vegetation cover (Ussiri et al., 2006). The low SOM concentration results in serious consequences to soil quality, such as decreased water and nutrient retention and loss of soil structure, reducing the erosion resistance ability (Dou et al., 2020). Soil structure contributes to carbon sequestration since soil aggregates protect the carbon from physical and biochemical degradation (Vicente et al., 2019). Soil disturbance emits huge amounts of CO2 owing to the mineralization of the SOM (Mukhopadhyay & Maiti, 2014). In our study, soil C stock was not statistically different between mining and post-mining sites. However, C stock displayed increasing trends in post-mining sites, attributed mainly to the litter accumulation and fine root turnover along spontaneous vegetation development post-mining. Previous studies showed that soil C stocks increase with the restoration of post-mining areas (Table 2).

The soil C stock in our study was lower than many post-mining sites, which can be mainly attributed to the differences in soil type and plant species of a semi-arid setting (Table 2). The climatic conditions, weathering processes, nutrient cycling, and soil characteristics in semi-arid tropical regions can differ significantly from temperate and more humid regions (Galhardi et al., 2020). Soil carbon sequestration can offset CO2 emissions from mining activities. For example, Shrestha and Lal (2006) estimated that the carbon assimilated by restored vegetation was equivalent to that emitted by coal consumption over 50 years in the USA. Therefore, vegetation restoration is essential for rebuilding soil carbon pools in mining degraded lands (Yan et al., 2020).

In a study of post-mining scenarios lasting 10 to 20 years, revegetation with native species promoted greater carbon stocks than exotic species (Yan et al., 2020), although the latter is widely used due to their higher growth rate. Exotic herbaceous and shrub plants have negatively affected the soil C stock in post-mining sites (Cusack et al., 2015; Guillemot et al., 2018). In the present study, the plant species that spontaneously colonized the post-mining area were the indigenous Mimosa tenuiflora (Wild), Croton blanchetianus (Baill), Cereus jamacaru, and Cnidoscolus phyllacanthus; only P. juliflora is an exotic species, but adapted to the Brazilian semi-arid.

Batjes (2016) estimated that up to 2060 ± 215 Pg of C is found in a soil profile 2-m deep. The C stock in the soil is higher than the sum of the total carbon in the atmosphere (~ 800 Pg C) and terrestrial vegetation (~ 500 Pg C) (FAO & ITPS, 2015). This means that changes in soil carbon stocks in mined areas can significantly affect the local C cycle. It highlights the importance of restoration in post-mining areas, a process that also increases the soil capacity to store carbon.

The low P availability in mining soils (Table 1) may impair vegetation development and hamper soil reclamation using plants. Phosphorus deficiency is a limiting factor for primary production, mainly in the tropics and subtropics (Ramaekers et al., 2010). Our results showed that vegetation played a role in increasing P available contents in the post-mining site compared to the mining area. The total N concentrations were also increased in the post-mining site (Table 1). The positive correlations between N and organic matter in post-mining sites suggest that SOM is the primary N source. Nitrogen plays an essential role in the building of soil organic carbon and labile organic carbon fractions in post-mining sites since it is one of the major limiting factors to plant development (Yuan et al., 2018).

The absence of control measures in mining waste deposition and wastewater discharge can favor soil and nutrient losses and surface water contamination, accelerate environmental degradation, and expand the mining influence area. The environmental degradation in the mined areas studied here points to interventions that focus mainly on soil structuring, decompaction, and organic matter inputs, for instance, through revegetation. Revegetation is the last stage in the reclamation process. Long-term revegetation enhances the soil fertility and stability, controlling erosion rates in post-mining areas; plant and microbial biomass provide major contributions to rehabilitation of mined soils (Singh et al., 2012). Besides, plant nutrient availability is critical for mined soil reclamation (Feng et al., 2019) and post-mining ecosystem sustainability.

Different plants have been used to contain, destroy, or extract heavy metals from contaminated soils and waters, known as phytoremediation (Moosavi & Seghatoleslami, 2013). Studies have demonstrated that the highest heavy metal accumulation potential lies with P. juliflora (Prasad & Tewari, 2016; Usha et al., 2009). Varun et al. (2011) concluded that P. juliflora was an effective heavy metal phytoremediator in mining areas after observing a strong correlation between the degree of contamination and concentrations of Pb and Cd in plant samples. However, more studies on the ability of autochthonous plant species to accumulate potentially toxic elements are needed to improve the phytoremediation or phytostabilization of metals in soils from semi-arid settings.

4.2 Concentration of heavy metals in soils

Besides improving the soil physical and chemical characteristics of mined soils, soil quality monitoring regarding soil contamination is crucial to guarantee low risks to animals, humans, and the environment. Although post-mining soils’ physical characteristics and fertility demonstrate that the area approaches the condition of non-mined soils, the Pb and Ni concentrations in the post-mining soil exceed the permissible concentrations for agricultural soils (CONAMA, 2009). The study area lies in rural settings with extensive livestock; hence, there is a risk of food chain contamination due to grazing. Soils in mining areas are often polluted with multiple metals and metalloids, such as Pb, Cd, Cu, Ni, and Zn (Galhardi et al., 2020).

The QRVs for Zn, Cu, Ni, Cd, Pb, and Cr were exceeded in at least one of the mining or post-mining sites, which show that the scheelite mining contaminated soils. It should be noted that QRVs for Fe and Mn are not included in the regulation since they constitute oxides that are expected to be at naturally high concentrations in soil (Davies & Mundalamo, 2010). The Fe and Mn soil concentrations in non-mining areas were lower than those determined for the state of the Rio Grande do Norte (Preston et al., 2014). However, increases in the Fe and Mn concentrations in the mined areas demonstrate that mining contributed to the Fe and Mn inputs in soil.

Due to the sandy soils found in the studied sites, mining wastes are highly erodible. Therefore, heavy metals adsorbed to soil particles can be transported to adjacent areas and water bodies. Most of these particles are lost from soils by surface processes, including erosion and overland flow. Besides, water flowing across the soil surface can dissolve and transport soluble and easily desorbable metals (Nascimento et al., 2019; Silva et al., 2019). The concentration of rainfalls in few days increases the potential for contamination owing to the erosion process acceleration. In the semi-arid region of Spain, concentrated rain during the short rainy season contributed to the dispersion of the heavy metals Pb, Zn, Cu, and Cd from wastes of abandoned mines (Navarro et al., 2008).

In the mining district of Currais Novos, Cu and Zn were emitted from uncovered tailings piles, mainly by wind processes and runoff, enriching stream sediments that form the drainage system near the mining district and leading to Cu contamination (Petta et al., 2014). Considering that the water from these water bodies supplies the population of nearby cities (Cavalcante et al., 2018) and that heavy metals accumulate in the human body over time, the situation has a significant impact not only on the environment but also on the public health.

The heavy metal soil contamination depended on the type of metal, and the mining phase: Cr, Ni, Zn, and Fe concentrations increased in post-mining sites, while the largest concentrations of Cd, Cu, Pb, and Mn were found in mining sites. In the two mining phases, metals released from scheelite mining reached toxic levels in the soil and require further studies and planning of remediation strategies in these multi-contaminated soils. Besides, studies on the accumulation of metals by the local spontaneous vegetation and the metal transference from plants to grazing animals are warranted.

5 Conclusions

We assessed the effects of mining on the soil quality and heavy metals contamination of the largest deposit of scheelite in Brazil. We found that mining drastically affected the soil physical and chemical characteristics by compacting the soils and depleting plant nutrients (N and P). Heavy metal concentrations in mined and post-mined soils are above regulatory levels and pose a risk to the environment and human health. Our results showed that soil bulk density, pH, available P, total N, organic matter, and heavy metal concentrations were sensitive soil quality indicators and separate natural and mined areas. Therefore, these variables can be part of environmental monitoring and reclamation programs in the areas degraded by scheelite mining in the semi-arid region studied here and in similar climate settings in the world. The natural revegetation of the old tailings piles enabled the post-mining sites to be grouped with non-mining areas due to improvements in soil physical and soil fertility driven mainly by the organic matter accumulation in the soil. These findings highlight the importance of revegetation in the ecological recovery of mined sites.