Technosol Micromorphology Reveals the Early Pedogenesis of Abandoned Rare Earth Element Mining Sites Undergoing Reclamation in South China

Abstract

The process of anthropogenic pedogenesis has necessarily become an important aspect of the study of today’s soils. The sustainable reclamation or remediation of soils degraded by industrial or mining activities is currently of great interest worldwide. In this field, the study of thin soil sections can provide relevant answers, particularly to questions concerning the evolution of these soils under the impact of reclamation practices. Here, we describe an experiment to reclaim former rare earth element mining sites in China using organic soil amendments and plantations of a local fiber plant, Boehmeria nivea. Two years after the start of the experiment, a study of soil structure, considered as an indicator of soil biofunctioning, was carried out on the different plots, supplemented by monitoring of physico-chemical properties. Morphological (light microscopy) and analytical (SEM-EDX, µ-XRF) characterization of thin sections allowed us to pinpoint some pedological processes as aggregation with particular reference to the contribution of biological factors and mineral species, highlighting the impact of the practices implemented. Using a soil micromorphology approach enabled us to track the rapid evolution of the early stages of pedogenesis of these Technosols and to provide insight into the potential for reclamation of these mined sites in the future.

1. Introduction

Soils are affected by various anthropogenic influences across extensive areas [1], making it essential to study their pedogenesis under the influence of both natural and anthropogenic factors. Such monitoring allows us to understand the early stages of the development of soils altered by anthropogenic activities and thus to consider evolutionary pathways. One of the anthropogenic modifications of soils is the incorporation of artifacts, i.e., substances created or substantially modified or brought to the surface by human activities, such as garbage, mine spoil, or industrial waste [2]. In 2006, the group of Technosols was introduced into the World Reference Base for Soil Resources (WRB) to represent soils containing significant amounts of artifacts or that are sealed [3,4]. Studies have already shown that Technosols can be formed from parent materials such as wastes or mining residues [5,6,7,8,9,10,11]. As Technosol pedogenesis can be fast [7,12], it is essential to follow these processes from the earliest stages. In this way, analysis of soil micromorphology is a relevant methodology to highlight pedogenetic processes in Technosols, especially considering that they are not necessarily yet visible on a macromorphological scale.

The trajectory of these Technosols can be influenced by the reclamation practices [10,13] implemented to enhance the soil capacity to fulfill ecosystem functions such as (i) vegetation support for spontaneous or introduced vegetation, (ii) plant or animal diversity, (iii) nutrient reservoirs, and (iv) buffers against pollution and erosion. Phytoremediation, i.e., the use of plants and associated microorganisms, along with soil amendment and agronomic practices, to stabilize, degrade, or extract soil pollutants, is commonly implemented as a reclamation strategy in sites degraded by mining activities. This is an environmentally friendly remediation technology with broad application prospects [14,15]. The addition of organic and inorganic amendments, as well as the development of plants, is driving soil structure development and the accumulation of organic matter in mining waste [16,17]. Soil structure development can be seen as an objective for reclamation, as soil aggregation is one of the most important pedological processes and can be considered as an indicator of soil biofunctioning [18]. In this way, the analysis of soil micromorphology is a well-suited technique for monitoring structural dynamics in Technosols [19,20] and can be used to investigate the influence of reclamation practices on pedogenetic evolution of Technosols developing in mine tailings.

The use of heap leaching technologies to mine ion-adsorption rare earth element (REE) deposits in southern China for several decades has caused a number of environmental problems, including water pollution, soil erosion, and the destruction of vegetation [21]. Following the cessation of mining activities, large areas covered by tailings resulting from REE mining activities were reported [21], and no subsequent rehabilitation treatment efforts have been made. These tailings are characterized by poor physical structures, high erodibility, low pHs, and small amounts of organic matter and nutrients, as spontaneous plant colonization does not always establish itself easily. This physico-chemical infertility is accentuated by the high residual concentration of REEs in these mine tailings [22]. In addition to the risk of leaching of REEs into the environment, the high concentration of exchangeable Al³⁺ generally observed in these acidic tailings can also inhibit plant colonization. To overcome these limiting factors for greening, phytoremediation trials have been set up to facilitate the establishment of life in these ecosystems [23,24]. Phytoremediation practices (tillage, input of amendments, and planting of selected plants) induce an evolution in soil organization and the physico-chemical characteristics. This triggers a new phase of pedogenesis, leading to the formation of soil from the tailings, which can be classified as a Technosol. Understanding this new phase of pedogenesis is complicated by the complexity of the history that these deposits have undergone. Monitoring the pedogenetic trajectory would require information on each step of the evolution. Comparison of the properties and structure of the non-reclaimed and reclaimed tailings allows the changes induced by reclamation practices to be highlighted. However, it is difficult to unravel the various natural and anthropogenic factors of evolution and to date the successive processes occurring in the tailings.

Here, we studied the formation of Technosols from tailings resulting from the mining of ion-adsorption REE deposits in a subtropical climate. Reclamation practices, such as amendment input and the plantation of fiber plants, were performed in a demonstration base in Jiangxi Province (South China) [25]. Soil profiles were described two years after the experiment started in the reclaimed area and were compared with non-reclaimed tailings left to evolve naturally. Physico-chemical analyses and micromorphology studies were carried out. Thin soil sections were morphologically and analytically (MEB-EDS, µXRF) characterized in order to specify the aggregation process and the dynamics of major elements, as these are well-accepted markers of soil evolution.

2. Materials and Methods

2.1. Presentation and History of the Mining Site

The demonstration base is located in Jiangxi Province, China, in the Dingnan County of Ganzhou City (latitude: 24.99°; longitude: 115.05°). The study area is hilly, with altitudes ranging from 200 to 1200 m. The climate zone corresponds to that of a subtropical humid monsoon climate (18.9° mean annual temperature; 1620 mm mean annual rainfall). The typical vegetation type in this region is the middle subtropical evergreen broadleaf forest, including evergreen broadleaf forest and subtropical coniferous forest. Depending on the altitude and climatic conditions, natural soils in the region belong to red earths and yellow earths in the mountainous areas with high humidity in the Chinese soil classification. These highly weathered soils with low cationic-exchange capacities correspond to Cambisols, Nitosols, Acrisols, or Ferralsols in the World Reference Base for Soil Resources [3]. In the Dingnan area, the REE ion-adsorption deposits were derived from the weathering of a type of granite from the early Yanshanian period, corresponding to the Upper Jurassic period [26]. It was a fine-to-coarse-grained biotite monzonitic granite [27]. Weathering processes under this subtropical climate resulted in the formation of 10–20 m thick weathering crusts, showing a profile composed of A, B, and C horizons above the non-weathered bedrock. The dark brownish A horizon (<1 m thick) is rich in organic matter and aggregated. The yellow-to-red B horizon (4–10 m thick), corresponding to the completely weathered zone, contains almost no primary minerals from the granite and is enriched in clay minerals (up to 80%), mainly kaolinite and halloysite. Its structure results from the aggregation of clay minerals, Fe oxyhydroxides, and relict quartz grains, showing no trace of the granitic texture. Horizon C (2–3 m thick), corresponding to the semi-weathered zone, preserves the relict granitic texture. Its mineralogy is influenced by the weathering of primary granitic minerals (feldspars, micas) and resulting clay minerals (<30%). A transition zone between the saprolite and the bedrock, rich in rock fragments, is found below the C horizon. During weathering, numerous cations are released and leached, and REEs migrate and accumulate mainly in the lower part of the B horizon and the upper part of the C horizon, where they are adsorbed on clay minerals or Fe-Mn oxyhydroxides [27].

Due to its abundant reserve of ion-adsorption REE deposits, Dingnan county has been mined since the 1970–1980s but more intensively in the 2000s. Since the 1990s, these deposits have been surface/mountaintop-mined using the heap leaching technique, which consists of removing the vegetation, excavating the REE-rich clayey soil layer (B/C horizon), placing it in a 1–5 m high heap on an impermeable layer, and leaching it with an electrolyte solution to recover the REEs adsorbed on the clays by ion exchange. Sodium chloride was used as a lixiviant in the 1970s, but this was replaced with ammonium sulfate in the 1980s because NH₄⁺ has a better desorption capacity than Na⁺ [28].

The studied REE mine tailings resulted from the exploitation of REE deposits by heap leaching technology, which continued until 2008. After abandonment, the tailings were terraced in 2011 in order to prevent erosion and landslides (Figure 1). Until 2015, the vegetation coverage was scarce and limited to spontaneous species (Miscanthus sinensis and Pinus massoniana). Since 2015, the site has served as a demonstration base to test different reclamation strategies using organic amendments and plants. Reclaimed zone A (around 2000 m²) of the demonstration base (Figure 1(1),(5)) was set up to test the feasibility of producing fiber plants on the tailings. A mixture of pig manure and sawdust (with a ratio of 2:1, v/v) was mixed by tillage with the first 10–12 cm of the tailings at a rate of 4–6 kg m⁻² in 2015 and of 1.5–3 kg m⁻² in 2016. First, Chinese hemp (Cannabis sativa L. var. indica), an annual fiber plant, was sown with a density of 3 g m⁻² in July 2015; then, one-year ramie (Boehmeria nivea L.) seedlings of about 35 cm in height (cultivar Zhongzhu No. 1) were planted in rows (3–5 plants per m²) in July 2016 after hemp removal. Ramie is a perennial fiber plant that colonizes tropical and subtropical regions. It has been cultivated in South China for thousands of years and is known for the high quality of its fibers for agricultural and industrial uses. Ramie shows strong adaptability to drought, low soil pH, and soil infertility [29], making it a good potential candidate for post-mined soil phytoremediation, while producing fibers of economic value [25,30]. No intervention was made in 2017. Other plants, especially Miscanthus sinensis, have grown between rows of ramie. Other parts of the demonstration site were left without reclamation for comparison, in particular parts of the tailings on steep slopes (>45°) (Figure 1(4), C profiles).

2.2. Soil Description and Sampling

The REE ion-adsorption mine tailings of the site were studied as parent materials for soil development. For that, bulk samples were collected from the surface of the mine tailings at two locations on the demonstration base. The area T1 was laid out in small basins connected by pipes for the heap leaching process the year before sampling (Figure 1(2)). Samples were collected in three basins. For each basin, one surface sample was taken from a depression, where finer particles accumulated at the surface, and one sample was taken from the upper part of the basin with coarser materials. In total, six samples were taken from the surface of site T1. The flat T2 area at the top of the demonstration base was terraced five years before sampling, and there was very little vegetation growing on these tailings (Figure 1(3)). Five surface samples of bare tailings were collected in the zone. Bulk soil samples were analyzed for physico-chemical properties, as explained in Section 2.3.

To study the effects of reclamation on soil development in the mine tailings, five soil profiles (0.8 m deep) were randomly dug in the reclaimed zone A (profiles A1 to A5) and three were excavated at the bottom of the non-reclaimed sloping tailings for comparison (profiles C1 to C3) in November 2017 (Figure 1(1)). Soil profiles were described to delimitate layers based on color, structure, and degree of compactness. Bulk soil samples were collected in each layer described along the profiles for physico-chemical analyses. One soil cylinder (100 cm³ in volume, 5 cm in height) was sampled in each layer vertically, and the samples were stored individually in plastic boxes to be transported to the laboratory for the characterization of bulk density and water retention. Vertical distribution of roots was investigated by using in situ mapping of root intersections with vertical planes. Root mapping was performed by placing a grid (40 × 60 cm) with a mesh of 2 × 2 cm on the vertical faces and marking the presence or absence of root contact in each 2 × 2 cm square. To study micromorphology, Kubiena boxes were collected in each identified layer along three profiles of the non-reclaimed (0–30 and 30–60 cm) and reclaimed (0–15, 15–30, and 30–60 cm) tailings, with two repetitions.

2.3. Physico-Chemical Analyses

Soil samples were air-dried and sieved through a 2 mm sieve. The mass of the coarse fraction (>2 mm) was determined. Soil texture was determined for most layers by sieving and using a densitometer (method LY/T 1225-1999) at the Guangdong Academy of Agricultural Sciences (Guangzhou, China). Concentrations of major elements were determined using air-dried soil samples (<100 µm) after lithium metaborate fusion by X-ray fluorescence at ALS Chemex Co., Ltd. (Guangzhou, China).

The soil cylinders were stored at ambient temperature until processing. When the cylinders were not completely filled with soil, the missing soil volume was estimated. Cylinders were saturated with water by capillary rise for 48 h, weighed, and dried at 105 °C until they reached a constant mass. The bulk density (ρ_b) was calculated as the dried mass of the soil in the cylinders divided by the soil volume. The volumetric water content at saturation (θ_sat) was obtained by calculating the difference in the masses of wet and dry soils divided by the soil volume. The total porosity (ε_tot) was calculated from the bulk density and ρ_s, the particle density, which is here considered to be equal to 2.65 g cm⁻³. The capillary porosity (ε_cap) was estimated by the water content following saturation by capillary rise for 48 h and mainly represented pores with diameters between 0.2 and 30 µm. The proportion of the capillary porosity over the total porosity (ε_cap/ε_tot) was calculated.

The total carbon and nitrogen concentrations were determined by dry combustion using a Vario EL cube elemental analyzer (Laboratory of Environmental Pollution Control and Remediation Technology, Guangzhou, China). The total organic carbon (TOC) concentration was considered equal to the total C concentration because inorganic C is insignificant in acidic tailings. The pH was determined in water at a ratio of 1:2.5 (w/v) after shaking for 1 h. The electrical conductivity was measured in water suspension at a ratio of 1:5 (w/v) and corrected from the temperature to be expressed at 25 °C. Cation-exchange capacity (CEC) and exchangeable cations (Ca²⁺, Mg²⁺, K⁺, Na⁺, and Al³⁺) were determined using a cobalt hexamine chloride (Co(NH₃)₆Cl₃) solution at 0.0166 mol/L at a ratio of 1:5 (w/v) for 1 h (NF ISO 23470). The CEC was determined by spectrophotometric analysis and the concentrations of exchangeable cations by dosing the extracts using ICP-OES (Perkin Elmer Optima S300 DV Laboratory of Environmental Pollution Control and Remediation Technology, Guangzhou, China). Selective extractions were carried out for the quantification of oxy(hydr)oxides (Fe, Mn, and Al). Oxalate extraction allows the solubilization of poorly crystalline Al, Fe, and Mn oxides, as well as allophane, gels, and Al organic complexes. The method has been adapted from [31], with a ratio of 1:10 after shaking for 4 h in the dark. The dithionite solution allows the solubilization of pedogenic oxides and hydroxides, including crystalline Fe oxides. The method was adapted from [32], with a solution of sodium citrate, acetate, and dithionite in a ratio of 1:20 which was shaken at 60 °C for 3.5 h. After filtration and dilution, the extracts were analyzed by ICP-OES (Perkin Elmer Optima S300 DV).

For each measured parameter, the mean and standard deviation were calculated on the number of samples collected (3 to 6). When some concentrations were below the detection limit (DL), half the detection limit was used for the calculations.

2.4. Micromorphology

Thin soil sections were prepared at the Laboratoire Sols et Environnement (Vandoeuvre-lès-Nancy, France) according to a specific methodology [33]. Soil blocks from Kubiena boxes were dried with acetone, embedded in resin (Norsodine 83 V), and polymerized at 35 °C. Then, 30 µm thick soil sections (50 × 80 mm) were realized from each block and glued on a glass slide with no further recovery. Soil microscopic features were described according to [34], using a Leica MFLZIII stereomicroscope (Laboratoire Sols et Environnement, Nancy, France) with magnifications from ×8 to ×100. To specify the elemental composition of the soil constituents, some thin soil sections without any coating were further analyzed using two methods at the Service Commun de Microscopie Electronique et Microanalyse X (GeoRessources, Vandoeuvre-lès-Nancy, France):

(1)

Micro X-ray fluorescence. A thin section of the surface layer of profile A1 in the reclaimed tailings was completely mapped using a Bruker M4 Tornado µ-XRF (GeoRessources, Vandoeuvre-lès-Nancy, France) consisting of a Rh anode X-ray tube and a Bruker XFalsh double detector (SDD-type). The map was obtained at 50 keV/600 µA with a 20 µm spot, a spacing between two spots of 45 µm, and a counting time of 8 ms. Elemental maps of the most abundant elements detected (Al, Ca, Ce, Fe, Si, Mn, Mg, K, Ti, and Zr) were processed using the integrated software of quantitative mapping Q map by binning 3 pixels and deconvolution of the spectra. The intensity of the color on the maps is proportional to the concentration of the element. The map of Ca showed its presence in the resin used to make thin sections.

(2)

Scanning electron microscopy (SEM; GeoRessources, Vandoeuvre-lès-Nancy, France). Several thin sections collected along profiles A1 and A2 in the reclaimed mine tailings were analyzed punctually using an SEM Tescan coupled with an electron-dispersive X-ray spectroscopy (EDX) microanalyzer, running at 15 keV or 18 keV, respectively, for analysis or mapping, with a pressure of 20 Pa allowing a low vacuum.

3. Results

3.1. Characteristics of Mine Tailings

The tailing materials were mainly composed of silicon (Si), aluminum (Al), potassium (K), and iron (Fe) with very low concentrations of total organic carbon (TOC), total nitrogen (N), and total phosphorus (P) and relatively high residual concentrations of REEs (

Table 1. Main physico-chemical properties of the REE ion-adsorption tailing materials sampled from the surface of recently leached tailings (T1) and tailings abandoned and terraced for 5 years at the top of the demonstration base (T2). For each parameter, the mean ± standard deviation was calculated on n = 5 or 6 samples from the bulk surface layer of the mine tailings.

		Recently Leached Tailings (T1, n = 6)	Abandoned Terraced Tailings (T2, n = 5)
Elemental composition (total concentrations)
Ctot	% (w/w)	0.20 ± 0.04	0.17 ± 0.03
Ntot	% (w/w)	0.035 ± 0.004	0.015 ± 0.002
C:N	-	5.8 ± 0.9	12.0 ± 0.7
Altot	% (w/w)	10.4 ± 1.7	8.8 ± 0.6
Fetot	% (w/w)	3.0 ± 0.6	2.2 ± 0.5
Ktot	% (w/w)	3.9 ± 0.3	4.7 ± 0.2
Sitot	% (w/w)	30.7 ± 2.1	32.6 ± 1.2
ΣREEs 1	mg kg−1	352 ± 104	391 ± 111
Particle size distribution
Clay (<2 µm)	% of fine earth	8.7 ± 4.6	7.4 ± 1.1
Silt (2–50 µm)	% of fine earth	34.7 ± 8.6	16.4 ± 1.1
Sand (50–2000 µm)	% of fine earth	56.7 ± 12.4	76.2 ± 0.9
Coarse fragments (>2 mm)	% (w/w)	14.2 ± 7.8	17.4 ± 6.6
pH, CEC, and exchangeable cations
pHwater	-	5.5 ± 0.4	4.6 ± 0.1
EC 2	dS m−1	n.a.	0.019 ± 0.003
CEC 3	cmol kg−1	3.9 ± 1.1	2.8 ± 0.7
Al/T 4	%	56 ± 12	77 ± 11
S/T 5	%	5.8 ± 0.7	2.9 ± 0.3
Forms of Fe
Feo 6	mg kg−1 (% of total Fe content)	148 ± 56 (0.5 ± 0.1% of Fetot)	228 ± 82 (1.0 ± 0.3 of Fetot)
Fed 7	mg kg−1 (% of total Fe content)	11,091 ± 2003 (38 ± 4 of Fetot)	8615 ± 1341 (39 ± 6 of Fetot)
Feo/Fed 8	%	1.3 ± 0.3	2.6 ± 0. 6
(Fed-Feo)/Fetot 9	%	37 ± 4	38 ± 6

¹ ΣREEs: sum of the total concentrations of the 14 lanthanides (La to Lu); ² EC: electrical conductivity; ³ CEC: cationic-exchange capacity; ⁴ Al/T: aluminum saturation (exchangeable concentrations of Al³⁺ divided by the CEC); ⁵ S/T: non-acid saturation (sum of exchangeable concentrations of Ca²⁺, Mg²⁺, K⁺, and Na⁺ divided by the CEC); ⁶ Fe_o: Fe concentration extractable by oxalate (non-crystallized hydrated oxides and organic complexes); ⁷ Fe_d: Fe concentration extractable by citrate–dithionite (pedogenic more- or less-crystallized Fe (hydr)oxides, ‘free iron’); ⁸ Fe_o/Fe_d: degree of disorganization of free Fe (hydr)oxides; ⁹ (Fe_d-Fe_o)/Fe_tot: degree of crystallinity of Fe (hydr)oxides.

). These materials were acidic with a low salinity and a low cationic-exchange capacity mainly filled by Al³⁺. They contained 10% to 30% of coarse fragments and their texture was sandy loam to loamy sand, except in the depressions of the recently leached tailings, which were more silty with a loamy texture at the surface. The recently leached tailings also displayed a higher pH and slightly higher N concentrations, probably related to ammonium sulfate leaching.

Mineralogical analyses of these tailings showed that the main minerals detected in the tailings were quartz, potassic feldspars, and phyllosilicates such as chlorites, micas, and kaolinite and halloysite [35]. Despite the reddish color of the tailings, Fe oxides were not detected by X-ray diffraction on bulk tailing samples, probably because of the relatively low total concentrations of Fe in the tailings (2%–3%) and also the limited proportion of total Fe present as well-crystallized oxides. Indeed, selective extractions of Fe showed that free iron (Fe_d), i.e., pedogenic more- or less-well-crystallized Fe (hydr)oxides, represented 32% to 47% of the total Fe concentration, with the rest of the Fe being present in the lattices of silicates. From the extractions, the degree of crystallinity of the Fe (hydr)oxides ranged from 30% to 46% of the total Fe concentration and the degree of disorganization of the free Fe(hydr)oxides (Feo/Fed) was rather low, between 1% and 4% (Table 1).

3.2. Description of Soil Profiles

All the profiles were marked by a reddish color of the materials, even if more- or less-colored areas were heterogeneously distributed along the profiles (Figure 2). At the top, non-reclaimed tailings on the slope showed an accumulation of washed coarse particles from the surface, suggesting that fine particles have been lixiviated or were covered by dark biocrusts (Figure 2(2),(3)). The profiles of the non-reclaimed tailings showed no clear distinction between layers, except perhaps at 30/35 cm, where a more compacted layer in the subsoil (30–50 cm) is separated from the topsoil (0–30 cm). No or few roots were observed along these profiles. No or few fine structural units were observed. The materials were friable with rock fragments (<1 cm). The structure can be considered to be single-grain. In terms of these characteristics, the layers along the profiles were described as C_u horizons, as they consist of the more- or less-weathered parent materials (C horizon), i.e., the mine tailings, which were considered to be artifacts (u) [34].

The most obvious difference between the profiles of the non-reclaimed (Figure 2(1)) and reclaimed (Figure 2(4)) tailings was the presence of a brownish superficial layer (0–10 cm) in the reclaimed zone. At the top of the tailings, there was an accumulation of ramie leaves and pieces of stems and organic residues (Figure 2(6)). The surface layer (0–10 cm) was crisscrossed by many roots and presented residues of organic matter as well as fine aggregates, leading to a weak granular structure. Roots were still very present down to a depth of 30 cm; then, their density gradually decreased and they were practically absent at 60 cm (

Table 2. Profile description and physical properties in the layers sampled along the profiles of non-reclaimed tailings and reclaimed tailings (zone A). The quantitative results are presented as mean ± standard deviation calculated on 3 or 5 profiles.

		Non-Reclaimed Tailings (n = 3 Profiles, C1 to C3)		Reclaimed Tailings—Zone A (n = 5 Profiles, A1 to A5)
Layer Depth	cm	0–25 or 30	25 or 30–60	0–10	10–30 or 45	30 or 45–60
Profile description
Compactness 1		1–3	2–3	0	1–2	2–3
Structure		Single-grain	Single-grain	Granular	Angular blocky	Angular blocky
Root frequency	%	0	0	96.8 ± 4.9	72.0 ± 9.4	38.3 ± 22.8
Particle size distribution
Coarse fragments	% (w/w)	32.3 ± 1.8	29.2 ± 7.6 *	26.7 ± 6.4	34.7 ± 3.5	31.5 ± 5.4
Sand	% of fine earth	71.2 ± 6.5	69.3 ± 2.1 *	72.4 ± 7.4 *	70.2 ± 2.0 *	69.4 ± 0.0 *
Silt	% of fine earth	20.0 ± 4.6	24.9 ± 4.9 *	20.7 ± 9.4 *	22.3 ± 2.8 *	22.1 ± 0.6 *
Clay	% of fine earth	8.8 ± 2.0	5.8 ± 2.8 *	7.0 ± 2.0 *	7.6 ± 1.1 *	8.5 ± 0.6 *
Porosity
ρb 2	g cm−3	1.19 ± 0.14	1.42 ± 0.07 *	0.81 ± 0.07	1.29 ± 0.06	1.27 ± 0.08
θsat 3	% (v/v)	42.8 ± 2.5	39.7 ± 2.5 *	37 ± 5.2	38.2 ± 1.7	42.8 ± 1.7
εtot 4	% (v/v)	55.2 ± 5.2	46.6 ± 2.6 *	69.6 ± 2.7	51.2 ± 2.3	52.2 ± 3.0
εcap/εtot 5	% (v/v)	77.9 ± 7.5	85.3 ± 0.6 *	53.2 ± 7.4	74.9 ± 5.4	81.9 ± 2.4

¹ Compactness scale: 0, loose; 1, weakly compacted; 2, moderately compacted; 3, strongly compacted. ² ρ_b: bulk density; ³ θ_sat: field water content after 48 h of saturation by capillary rise; ⁴ ε_tot: total porosity calculated from the bulk density; ⁵ ε_cap/ε_tot: proportion of capillary porosity (taken as the water content at saturation) over the total porosity. * indicates that this layer was analyzed in only 2 profiles.

). The underlying layers were reddish and not well differentiated. Separation can be carried out at 30 or 45 cm depending on the degree of compactness and the size of rock fragments increasing with depth. Fine-to-coarse angular aggregates were observed, respectively, at a depth between 10 and 30 cm and between 30 and 60 cm. The structure can be considered to be angular blocky.

The surface layer was an organo-mineral horizon (A horizon) resulting from tillage, the incorporation of organic amendments, and the planting of tailings considered as artifacts (u), known as A_u. The underlying layers can be considered as C_u mineral horizons, as in the non-reclaimed tailings [36].

3.3. Physico-Chemical Analyses Along the Profiles

In terms of physical properties, the granulometric distribution of particles was relatively similar in the non-reclaimed and reclaimed tailings and constant along the profiles (Table 2). The content of coarse fragments (>2 mm) varied from 21% to 39% of the soil mass. All layers analyzed for texture showed a sandy loam texture class with sand content ranging from 64% to 78% of the fine soil and clay content ranging from 4% to 11% of the fine soil. This is in line with the texture of the tailings that had been abandoned for several years (Table 1). There was no evidence of textural variation along the profiles.

The bulk density of the non-reclaimed tailings varied from 1.08 to 1.46 g cm⁻³, corresponding to a total porosity of 45% to 59%, showing a slight compaction with depth. After reclamation, the surface A_u horizons tended to have a lower bulk density (0.69 to 0.88 g cm⁻³) and thus a higher total porosity (67% to 74%), considering the solid density as constant, than the deeper C_u layers (1.15 to 1.35 cm⁻³) (Table 2). The capillary porosity represented 71 to 86% of the total porosity in the non-reclaimed tailings. This proportion was similar in the deeper C_u layers of the reclaimed tailings (66% to 84%) but lower in the surface A_u horizons (47% to 66%). This indicates a higher proportion of macroporosity (pores > 30 µm in diameter) out of the total porosity in the surface-amended A_u horizons.

As shown in Table 1, the tailings were mainly composed of Al, Si, Fe, and K with no specific trend in relation to depth along the profiles or changes in reclamation practices (

Table 3. Elemental composition and chemical properties in the layers sampled along the profiles of non-reclaimed tailings and reclaimed tailings (zone A). The results are presented as mean ± standard deviation calculated on 3 or 5 profiles. * indicates that this layer was analyzed in only 2 profiles.

		Non-Reclaimed Tailings (n = 3 Profiles, C1 to C3)		Reclaimed Tailings—Zone A (n = 5 Profiles, A1 to A5)
Layer Depth	cm	0–25 or 30	25 or 30–60	0–10	10–30 or 45	30 or 45–60
Elemental composition (total concentrations)
Ctot	%	0.14 ± 0.02	0.16 ± 0.02 *	1.83 ± 0.45	0.22 ± 0.02	0.16 ± 0.01
Ntot	%	0.017 ± 0.006	0.021 ± 0.004 *	0.155 ± 0.035	0.021 ± 0.002	0.017 ± 0.006
C:N	-	8.8 ± 1.8	7.7 ± 0.2 *	11.7 ± 1.1	10.4 ± 1.0	10.7 ± 3.7
Altot	%	8.6 ± 0.9	9.0 ± 0.1 *	8.7 ± 0.8	9.2 ± 0.1	9.1 ± 0.2
Fetot	%	2.4 ± 1.1	3.1 ± 0.1 *	2.8 ± 0.1	3.1 ± 0.3	2.8 ± 0.2
Ktot	%	3.6 ± 0.5	3.4 ± 0.1 *	3.9 ± 0.1	4.2 ± 0.2	3.9 ± 0.3
Sitot	%	33.1 ± 1.9	32.2 ± 0.5 *	30.8 ± 1.4	31.6 ± 0.6	32.1 ± 0.2
Catot	g kg−1	0.14 ± 0.0003	0.09 ± 0.08 *	1.53 ± 0.44	0.45 ± 0.30	0.36 ± 0.18
Ptot	g kg−1	0.04 ± 0.01	0.03 ± 0.02 *	0.74 ± 0.21	0.06 ± 0.02	0.04 ± 0.01
Stot	g kg−1	0.11 ± 0.02	0.07 ± 0.07 *	0.34 ± 0.23	0.14 ± 0.08	0.13 ± 0.02
ΣREEs 1	mg kg−1	426 ± 178	562 ± 108 *	809 ± 643	334 ± 133	294 ± 60
Chemical properties
pHwater	-	4.44 ± 0.05	4.44 ± 0.04 *	5.25 ± 0.28	4.60 ± 0.19	4.54 ± 0.21
CEC 2	cmol+ kg−1	3.2 ± 0.9	4.0 ± 1.0 *	5.9 ± 1.3	4.4 ± 0.8	4.6 ± 0.4
Alexch 3	cmol+ kg−1	2.3 ± 0.6	2.2 ± 0.1 *	0.1 ± 0.1	2.2 ± 1.3	2.9 ± 0.7
Caexch 3	cmol+ kg−1	0.1 ± 0.1	0.5 ± 0.4 *	5.2 ± 1.4	1.6 ± 1.4	1.3 ± 0.9
Kexch 3	cmol+ kg−1	0.06 ± 0.02	0.10 ± 0.04 *	0.34 ± 0.12	0.11 ± 0.05	0.09 ± 0.01
Mgexch 3	cmol+ kg−1	0.01 ± 0.01	0.03 ± 0.02 *	1.01 ± 0.15	0.27 ± 0.19	0.09 ± 0.04
Naexch 3	cmol+ kg−1	<DL	<DL *	0.03 ± 0.01	<DL	<DL
Al/T 4	%	72 ± 3	57 ± 13 *	2.3 ± 2.6	52 ± 33	62 ± 16
S/T 5	%	5.7 ± 1.4	16 ± 8 *	113 ± 7	44 ± 29	31 ± 20

¹ ΣREEs: sum of the total concentrations of the 14 lanthanides (La to Lu); ² CEC: cationic-exchange capacity; ³ X_exch: exchangeable concentrations of cations; ⁴ Al/T: aluminum saturation (exchangeable concentrations of Al³⁺ divided by the CEC); ⁵ S/T: base saturation (sum of exchangeable concentrations of Ca²⁺, Mg²⁺, K⁺, and Na⁺ divided by the CEC).

). However, the reclamation induced an enrichment in C, N, P, and S in the amended surface A_u horizons. Concentrations of TOC were very low in the non-reclaimed tailings (0.12%–0.18%). They reached 1.1% to 2.4% in the surface A_u horizons of the reclaimed tailings, whereas the underlying C_u layers still showed similar concentrations to the non-reclaimed tailings (0.15%–0.24%). Similar trends were observed for N, P, and S (Table 3). The C/N ratio varied from 6 to 16 and tended to be slightly higher in reclaimed tailings than in the non-reclaimed tailings. Total Ca concentrations were significantly higher in the A_u horizons (0.12%–0.23%) but also in the underlying C_u layers (0.01%–0.09%) of the reclaimed tailings compared to the layers of the non-reclaimed tailings (<DL–0.01%) (Table 3).

In terms of chemical properties, the non-reclaimed mine tailings showed pH values of 4.4–4.5, as measured in the tailings abandoned for several years (Table 1). In the reclaimed tailings, the pH was higher in the surface A_u layers (pH 4.9 to 5.6) than in the deeper C_u layers (4.4 to 4.8). The cation-exchange capacity (CEC) of the mine tailings was low. It showed slightly higher values in some A_u layers (with a maximum value of 7.7 cmol⁺ kg⁻¹) compared to the C_u layers in the reclaimed tailings and in the non-reclaimed tailings. However, the concentrations of exchangeable cations changed significantly with reclamation. Exchangeable Al concentrations were much lower in the surface A_u layers (<6% of the CEC) than in the deeper C_u layers of the reclaimed tailings (32 to 82% of the CEC) and along the non-reclaimed tailings (48% to 75% of the CEC). The exchangeable concentrations of the base cations (Ca²⁺, Mg²⁺, K⁺, and Na⁺) showed an inverse trend to Al³⁺. The exchange complex was completely saturated (S/T > 100%) in the surface A_u layers, and the deeper C_u layers showed a higher base saturation (5% to 80% of the CEC) than in the non-reclaimed tailings (5% to 22% of the CEC). The most abundant base cation on the exchange complex was Ca²⁺ followed by Mg²⁺ and, to a lesser extent, K⁺. All the base cations showed the highest concentrations in the A_u layers. In addition, exchangeable Ca²⁺ and Mg⁺ concentrations were also higher in the deeper layers of the reclaimed tailings than in the non-reclaimed tailings.

3.4. Micromorphology Characterization

3.4.1. Structure Description

As a reminder, soil blocks in Kubiena boxes were collected from A_u and C_u horizons of the different profiles at one or two depths in the non-reclaimed tailings (0–30 and 30–60 cm) and two or three depths in the reclaimed profiles (0–15, 15–30, and 30–60 cm). Overall views of the thin tailing sections presented a distribution of the groundmass (i.e., coarse components and the micromass) and pores constituting the soil microstructure (Figure 3). The distribution of micromass versus individualized minerals was very heterogeneous, whatever the depth or soil treatment (reclaimed/non-reclaimed), even if micromass patches were less observed in the superficial layers of reclaimed soils. More pores were visible in the superficial layers (0–30 cm) of all profiles (Figure 3(1)–(3)) but particularly in the A_u horizons of the reclaimed soils (Figure 3(1)). Then, deeper down, the soil substrate appeared more compact (Figure 3(4)–(6)), sometimes with voids at 30–60 cm depth (Figure 3(6)). Considering the heterogeneity level of the thin section constitution, no other differences were identified at this scale.

Details of non-reclaimed thin sections specified the distribution of coarse components (i.e., mineral grains) versus micromass (Figure 4). Despite the high heterogeneity of this distribution, the observation of the twelve slides confirmed the greater porosity of the 0–30 cm layer in the non-reclaimed tailings (Figure 4(1)) compared to the 30–60 cm layer (Figure 4(2)). Most minerals were small in size. Mineral sizes were distributed between 100 µm and 800 µm on the one hand and around 2 mm on the other, with a small proportion of larger minerals not exceeding 4 mm in size. Minerals, i.e., quartz, feldspars, and clays, were usually surrounded by a fine-grained reddish mineral micromass (Figure 4(3)–(5)). This micromass was present either as patches of various sizes (from 1 to 5 cm; Figure 3) or as a fine layer (2 µm) surrounding other minerals (Figure 4(3)–(5)). As micromass was associated to minerals (<50 µm), there could also be some stratification (Figure 4(6)) at this scale in the form of a layered coherent structure.

3.4.2. µ-XRF and SEM-EDX Analyses of Mineral Features

Similar mineralogy and mineral features were observed whatever the depth or type of profile (non-reclaimed/reclaimed) in the tailings. Microanalyses using µ-XRF and SEM-EDX were performed on some thin sections in the reclaimed tailings to specify the constitution of mineral phases (Figure 5 and Figure 6 as well as Figures S1–S6 for the EDS spectra and associated elemental composition). Elemental mapping of the thin section by µ-XRF showed the distribution of the mineral phases composed of the most abundant elements (Al, Si, K, and Fe) in the tailings. The principal minerals composed of Si were quartz phases (Figure 5(2)), which were either isolated, embedded in the micromass, or associated with other primary minerals such as feldspars in intergrowth or phyllosilicates and their iron-bearing weathering products (Figure 5(12)). Quartz phases were often cracked and some were fractured and filled with fine materials (Figure 6(1)). The main K-bearing phases were potassic feldspars (Figure 5(4)), which showed signs of weathering such as holes or reddening (Figure 5(8)). Some aluminosilicate phases were several millimeters in size with possible reddish, brownish, or greenish coloration (Figure 5(9)) and with a Si:Al atomic ratio close to one (Figure 6(8),(9) and Figure S6), which could correspond to kaolinitized feldspars associated with oxides. Aluminum was distributed in various aluminosilicates (Figure 5(3)) but was particularly concentrated in a few yellowish phases (Figure 5(10)). Based on EDS analysis of one of these phases, it contained phosphorus (P)- and P-enriched particle-bearing rare earth elements (REEs), which were observed along the crack of this mineral (Figure S2). This suggested the possible presence of aluminum phosphate minerals, some of which can host REEs, such as florencite [37]. Iron was present in several types of minerals (Figure 5(5)). One of the iron-bearing phases contained dark particles assigned to Fe (hydr)oxides (Figure 5(10)). Other Fe-bearing phases included brownish ferro-magnesian phyllosilicates, corresponding probably to biotites or chlorites. Most of them were weathered. Some were exfoliated (Figure 5(10)) and others showed rust-colored material along the cleavage planes (Figure 6(4)). Weathering of these phyllosilicates led to the formation of dark red patches, composed of veins of Fe-bearing phases as well as dark Fe (hydr)oxides containing titanium (Figure 6(4),(5)). Magnification of these dark red patches revealed networks of Fe-rich particles of one micrometer or less in size, probably (hydr)oxides (Figure 5(12),(13) and associated Figure S4 and Figure 6(1)–(3) and associated Figure S4). In addition, there were some black deposits resulting from the precipitation of Fe and Mn (hydr)oxides, sometimes associated with Ce oxides (Figure 6(6),(7) and associated Figure S5).

The darker or lighter reddish color of the micromass suggested the presence of iron in varying contents and/or speciations (Figure 5(6) and Figure 6(1)). The observations and analyses by SEM-EDS showed that it was composed of a mixture of mineral grains (5 to 100 µm in size) including lamellar phases and Ce- and Zr-bearing particles embedded in finer material (Figure 5(6) and Figure 7 and associated Figure S1). Based on several EDS analyses, the micromass was composed mainly of Si (18% ± 2% atomic, n = 9), Al (14% ± 2%), Fe (2.9% ± 4.4%), and K (1.6% ± 1.2%), suggesting a mixture of fine (alumino)silicates, phyllosilicates, and Fe (hydr)oxides.

3.4.3. Organic Matter and Biogenic Features

The development of an A_u horizon in the reclaimed tailings was the biggest difference in comparison with non-reclaimed profiles. Details of thin sections made in the superficial A_u horizon showed various organic matter residues (Figure 7). Still-intact leaf fragments appeared as individualized residues in this porous layer (Figure 7(1)). Some others were less easily recognizable because they were more fragmented. Numerous fine roots were also observed in the different voids (Figure 7(1),(2)). Many roots were also clearly associated with minerals (Figure 7(3) and Figure 8(3),(4) and associated Figure S7), underlining their contribution to aggregation. Fecal pellets were also dispersed in this layer (Figure 7(2)). These pellets were very different in size (50 µm, 100 µm, 250 µm, or 1500 µm or more), compaction, and degradation state (Figure S7). As the size and constitution of pellets are an indicator of the animal species, it is clear that there is faunal activity in the form of mites, collembola, enchytraeids, and millipeds in this soil. Some fecal pellets were composed of organo-mineral associations, as shown by the presence of minerals such as quartz (Figure 7(4)) or Fe oxides (Figure 8(1),(2) and associated Figure S8) mixed with organic residues or roots. These pellets can be considered, as newly formed aggregates, that are likely to be dispersed in the horizon (Figure 7(4)). Microflora was also observed as black fungal spores and hypha decaying organic matter (Figure 7(5)), present even within fecal pellets (Figure S7). Some biodegraded residues were also still recognizable in the 50–60 cm layer of the reclaimed profiles (Figure 7(6)). On the contrary, such organic features were nearly not observed in the non-reclaimed tailings, insofar as organic matter was virtually absent. All these observations underlined how much this A horizon is impacted by biological activity, evolving in the topsoil of the studied Technosols.

3.4.4. Structure Evolution

The main structure differences between the initial tailings and non-reclaimed (single-grain) profiles and the reclaimed (granular) profiles were due to biological activity. Roots contributed to higher total soil porosity in reclaimed profiles compared to non-reclaimed ones. Organic matter and biofeatures contributed to aggregate formation in the reclaimed profiles.

4. Discussion

4.1. Mine Tailings as Parent Materials

The REE ion-adsorption mine tailings are the parent materials of the soils and ecosystems that develop on these abandoned mine sites. These materials have undergone several stages of evolution: (i) weathering of the granite bedrock and development of thick leached soils in several phases since the Mesozoic period, (ii) excavation, transport, and leaching associated with the mining process (from 1980s to 2008), and (iii) leveling and weathering under natural climatic conditions (from 2011 to 2017). The characteristics of the tailings resulting from this complex history will determine the potential trajectory of the evolution of these soils and ecosystems.

These parent materials are characterized by a sandy loam texture with less than 10% of clay-size particles, low organic matter, low CEC (<10 cmol⁺ kg⁻¹), and low base saturation. They contain primary minerals from the granite parent materials, such as quartz, potassic feldspars, and micas (biotite), as well as secondary minerals, such as clays and Fe-Mn (hydr)oxides. Analyses of the clay fraction in XRD indicated the presence of chlorite-type (14 Å), mica-type (10 Å), and kaolinite-type (7 Å) phyllosilicates. Observations in transmission electronic microscopy evidenced that halloysite is the predominant kaolinite-type clay [38]. Other minor mineral phases observed include aluminum phosphate minerals, which can host some REEs [38], Ce oxides, and zircons.

According to local geological data, the main minerals in the parent granitic rock from the upper Jurassic (early Yanshanian) period are potassic feldspars, plagioclase, quartz, and biotite with a minor amount of muscovite and hornblende and REE-bearing accessory minerals including zircon, apatite, and fluorocarbonates (synchisite) [27]. Weathering of granites under warm and humid conditions began in the Mesozoic in southern China, resulting in the formation of saprolites, which were partially eroded due to climatic variations and tectonic uplift in the Quaternary period. Soils formed on the remaining saprolite during periods of geomorphological stability were influenced by climatic changes and tectonic events in the Upper Pleistocene and Holocene [35]. During the granite weathering and then during soil development, the primary minerals more susceptible to weathering (plagioclase and biotite) were decomposed and secondary minerals were formed following different sequences, leading to the formation of kaolinite, halloysite, illite, vermiculite, smectite, gibbsite, chlorites, or Fe (hydr)oxides depending on the conditions (climate, hydrogeology, geomorphology, and parent materials) [26,35]. As the result, the exploited ore (weathered layer) is mainly composed of quartz, potassic feldspars, and kaolinite while biotite and plagioclase are mostly weathered.

The observations of thin sections in the mine tailings revealed various micromorphological features associated with mineral weathering and neoformation, such as possible kaolinitization of feldspars or alteration of ferro-magnesian phyllosilicates to Fe (hydr)oxides. These features correspond to typical alterations occurring during the weathering of granite and subsequent formation of red soils in a subtropical climate [26,35]. Other features, such as black precipitates of Fe, Mn, and Ce oxides, suggested the dissolution, migration, and precipitation of elements such as (hydr)oxides. Cerium is the only REE that can be oxidized from Ce³⁺ to Ce⁴⁺ and precipitate as an oxide (CeO₂) through oxidative weathering [26]. It has been detected under both oxidation states in the tailings [38]. The formation of Ce oxide occurs during the weathering of granite depending on local pH and redox conditions, possibly under the influence of the rhizosphere and associated mineralization of Fe and Mn oxides [39]; Ce oxides seem to not be affected by leaching during mining exploitation [38]. All these features are likely to be inherited from the weathering of granite and the subsequent formation of red soils.

In terms of structure formation, the results obtained on the non-reclaimed profiles, i.e., the mine tailings that have evolved for at least six years under local climatic conditions, indicated a rapid decrease in porosity with depth and the profiles are not well structured. The pseudo-aggregation observed on thin sections, resulting from the embedding of primary minerals by the fine-grained micromass, can be compared to pseudoplasmation. This process, observed in some tropical soils, is defined as the “in situ development of a structured loose material in which the original fabric of the saprolite is no longer visible” [7]. This concept is not entirely appropriate for Technosols, as technogenic parent materials can be fragmented, weathered, and transported by human activities. However, the formation of fine materials that turn red with time has also been described in Technosols formed on mine spoils made of rock fragments of dolomitic limestone in Brazil [7]. The fine-grained reddish matrix is made of a mixture of submicrometric phyllosilicates and Fe (hydr)oxides. Its reddening could be explained by the release of Fe during weathering of primary minerals such as biotite or chlorite and its precipitation as oxides. Additional microscopic observations of the fine fraction (<50 µm) of the tailings showed the presence of halloysite as a predominant clay mineral and of submicrometric amorphous Fe (hydr)oxide phases. Fe-XANES analyses on thin sections suggested the presence of iron (hydr)oxydes, goethite, and ferrihydrite in the matrix. In addition, this fine fraction was enriched in organic carbon compared to the bulk tailings [38]. This organo-mineral matrix probably results from the weathering of the minerals making up the parent rock. However, the determination of and the drivers of the formation of the micromass and pseudo-aggregates would have required a comparative micromorphological study of sequential samples of undisturbed regolith at different stages of its evolution. Iron presence in the micromass attests to its contribution to aggregation, indicating the importance of iron dynamics in the evolution of these Technosols. This process is well known for all soils including Technosols [40]. However, its dynamics still need to be further investigated in such mine tailings, where a subsequent portion of the primary Fe-bearing minerals have undergone weathering.

From the results obtained, it is difficult to specify the temporality of the observed processes of weathering and the arrangement of mineral associations and to attribute them solely to the geological origin of the materials, their pedological evolution prior to mining treatment, their even more extensive weathering and fragmentation during mining treatment, or even to the short evolution of the mine tailings under climatic conditions over several years. This would have required us to compare the materials at these different steps of evolution. As discussed previously, mineral (trans)formations, such as the weathering of relict primary minerals (feldspars or micas) observed in the tailings, are relatively common in the formation of red soils from granite in areas with subtropical climates [26,35]. Therefore, these transformations could have occurred during the formation of the ore and been preserved. Regarding the influence of the mining process, it has been shown experimentally that the lixiviation of ores (weathered soil horizons) with ammonium sulfate at acidic pH (4–4.5) induces mineralogical and microstructural changes in the material, such as the weathering of micas or the partial dissolution and migration of kaolinite and Fe hydroxides [40,41]. This leads to the loss of Al, Fe, and organic N content, as well as partial destruction of the cemented structure of the ore. This increases the proportion of quartz and potassic feldspars and create more voids, thereby changing the physical properties of the materials [40,41]. This suggests that the mining process has contributed to the fragmentation, weathering, and loss of fine reactive particles in the tailings. Based on our results, the comparison of the composition and properties of the recently leached tailings and those that have been abandoned for several years shows a decrease in pH and N content, which could indicate the leaching of some cations including NH₄⁺ adsorbed onto clays during the mining process after abandonment under the subtropical climate.

These materials, which are highly weathered and fragmented, make them specific in terms of their potential development, which is necessarily different from the development of other Technosols developed on more weatherable materials. The studied non-reclaimed profiles were not colonized by plants; however, a few scattered plants were observed, and in places, a biological crust was also present. Even if some leaching, mineral transformation, and pseudoplasmation processes exist in the tailings, profiles showed no visible soil development under the climatic conditions and restricted biological activity. Tailings are not well structured and are susceptible to rapid erosion. As these mine tailings can be considered as artifacts due to their modification by the mining process, incipient soils developing on them can be considered to be Spolic Technosols (Arenic, Hyperartefactic, and Protic) [2]. We might also wonder whether, in the long term, pedoplasmation, as an initial aggregation, might lead to the formation of a B horizon in these continually developing Technosols on mine tailings. Through iron dynamics, pseudo-pedoplasmation, and the presence of clays, the tailings still presented carbon-storage potentialities due to the existence of fine particles with some adsorption capacity when colonized by spontaneous vegetation. Taking into account the rapid mineralization of organic matter in these environments, micromorphological monitoring of these soils would thus make it possible to verify the association of carbon with fine mineral elements and determine the potential for medium-term carbon storage.

4.2. Effects of Reclamation on the Pedogenesis and Restoration of Soil Functions

Observation of the soil profiles of the reclaimed tailings clearly shows the development of an organo-mineral A_u horizon (surface soil layer—0–10 cm), thanks to soil amendment incorporation by tillage and planting. This A_u horizon also has much greater porosity. Deeper below the surface (15–60 cm), the amount of root organic matter continually decreased and porosity was also greater compared to in non-reclaimed soils. This A_u horizon also has higher total porosity with a higher proportion of macropores (>30 µm in diameter) and a lower proportion of capillary porosity involved in water retention, probably due to the effects of tillage and biological activity. The addition of amendments also induces an enrichment of carbon and nutrients (N, P, and S) and a saturation of the exchange complex with base cations, reducing the amount of exchangeable Al³⁺, which can be toxic to plant roots [42]. In addition, roots contribute to the accumulation and solubilization of carbon and nutriments, as shown by their enrichment in the rhizospheric soils around ramie roots developing in the tailings [25]. Most of the changes in chemical and physical properties are confined to the A_u horizon, with the deeper C_u horizons showing properties similar to those of the non-reclaimed tailings. However, the enrichment of Ca and exchangeable Ca²⁺ and Mg⁺ in the deeper layers of the reclaimed tailings suggests the leaching of some base cations in the profile [43]. In the deeper C_u layers (10–60 cm), the development of roots may contribute to the introduction of organic matter and to increased porosity, but these effects remain limited after 2 years.

Micromorphology confirms the impact of biological activity on the evolution of these materials by specifying the nature and transformation of organic matter along the profile. Leaves appeared to be rapidly mineralized, and fine roots also degraded fairly quickly, while larger roots were frequently observed. Ramie does indeed have a root system made up of large ligneous storage roots and fine roots involved in the absorption of nutrients [25]. Its roots contributed to soil porosity and external root tissues were associated with fine mineral particles, suggesting the release of aggregates in the profile. Biological activity can also be seen in the presence of fecal pellets of various origins, attesting to the biodiversity (i.e., enchytraeides, isopodes, millipeds) that has developed on these plots. The constitution and size of these excrements vary according to the species that produce them, giving them different turnovers, especially as some may already be associated with minerals. Thus, these aggregates are fully involved in the horizon structure formation. Evidence of the presence of fungi associated with organic matter attests to its biodegradation and thus the supply of nutrients to the whole ecosystem. The treatments that were carried out included planting, soil amendment, and also plowing (fragmentation and mixing effects). All these practices very quickly impacted the stored mine tailings, initiating the formation of Technosols. Compared to the non-reclaimed profiles, the new reclaimed soil profiles demonstrated an increase in soil porosity and carbon content in the amended and planted horizon, a significant degree of biological activity, and a soil structure with organo-mineral aggregates. These processes are essentially located in the most superficial horizons, and the stability of this new structure remains to be determined and certainly consolidated. Studying the structure of these Technosols over a longer period of time will enable us to determine whether this evolution is continuing or, because of the subtropical climate, the mineralization of carbon over time is greater than its storage capacity. The pedogenetic processes driven by living organisms, in addition to the reclamation practices, seem dominant at the early stages of formation of these soils. It remains to be seen whether the same applies thereafter and whether equilibrium is established with the physico-chemical pedogenetic processes. The impact of biological activity on mineral weathering and neoformation and its contribution to aggregation remains to be determined, given that, as in the case of pedoplasmation mentioned above, bioturbation plays an important role in the widespread turnover of soil materials and the structural changes observed at a microscopic scale in the transition zone between saprolite and soil [7]. Nevertheless, a period of only two years was enough to initiate soil development. In comparison, it was previously suggested that three years are required to develop an A horizon (surface soil layer) in rehabilitated soils that are initially less weathered [44]. The accumulation of pedogenic organic C is therefore considered to be an indicator of the success of soil development. In this way, the functions of carbon stock, vegetation support, and nutrient reserves can be fulfilled, depending on how these parameters evolve in the long term.

As a result, in just two years, the reclamation practices induced soil development (A_u/C_u profile), leading to a proposition for the soils to be classified as Spolic Technosols (Arenic, Hyperartefactic, and Ochric) [2]. This proves the effectiveness of the reclamation treatments in that, despite fairly low fertility and the presence of rare earth elements, vegetation has been able to establish itself and the soil has evolved. Pedogenesis rates in reclaimed soils are not higher than in natural soils; however, soil amendments used during the reclamation process likely create parent material conditions that promote vegetative growth (plantation or spontaneous colonization) and start a new stage of pedogenesis [45]. In addition, reclamation practices also contributed to reduce the availability of potentially toxic elements, such as rare earth elements and aluminum [25]. Recent work on these same materials highlights the value of morphological studies of thin sections coupled with synchrotron-based X-ray fluorescence spectroscopy for the speciation of elements, in particular rare earth elements (REEs), making it possible to specify the bearing phases of these elements and, therefore, the potential risk of dissemination of potentially toxic elements into the environment from these abandoned mine sites [38]. In particular, REEs presented high affinity for organic phases, suggesting a positive impact of phytoremediation on REE fate.

5. Conclusions

The results displayed here focused on the influence of reclamation practices on the early evolution of soils forming on tailings derived from the mining of rare earth elements in South China, based mainly on a micromorphological study. The mine tailings showed mainly micromorphological characteristics resulting from the weathering of granite in the subtropical climate and the subsequent formation of thick leached soil from which the REE ore was extracted. The evolution of the tailings after the cessation of mining activities appears to be relatively limited, with restricted biological activity. The reclamation practices, i.e., the addition of organic soil improvers and the planting of fiber plants, induced the development of an organo-mineral horizon with the accumulation of carbon and nutrients and the formation of organo-mineral associations driven by plant roots and soil organisms. This demonstrates the feasibility of growing fiber crops on low-fertility tailings with the use of amendments. However, assessing the sustainability of new ecosystems such as these would require long-term monitoring of soil function and development.

Micromorphology analysis enables us to identify some pedogenetic processes occurring in Technosols. Complementing the overall physico-chemical analyses, various existing microanalysis options on thin sections (e.g., µXRF, SEM-EDX, and X-ray fluorescence spectroscopy), as well as future methods (such as laser-induced breakdown spectroscopy), offer a wide range of techniques for the characterization of pedofeature constitution while identifying the in situ organo-mineral associations. Such advantages of the micromorphology approach counterbalance the reluctance to use this technique due to the potentially tedious conditioning of samples. This approach appears to be very useful particularly in (i) the characterization of elements of interest or potentially pollutants, (ii) the monitoring of structure development due to biological activity, and (iii) the study of early pedogenesis on anthropogenic materials. Taking into account the growing importance of the coverage of anthropogenic soils, micromorphology analysis appears to be relevant in this field of research, i.e., in the understanding of the evolution of these materials and their propensity to fulfill ecosystem functions. In terms of reclamation strategy, micromorphology analysis helps to identify the impacts of a given treatment and to adjust the reclamation/remediation practices with a view to sustainability.