Slow Recovery of Major Soil Nutrient Pools during Reclamation in a Sub-Alpine Copper Mine Area, Southeastern Edge of the Tibetan Plateau, Sichuan Province, SW China

: Improvement of soil nutrients is crucial for the long-term development and stability of restored ecosystems in mine areas. However, knowledge about the variation in soil nutrients and their inﬂuencing factors during the reclamation of sub-alpine metal mine soil is still scarce. We assess the status of soil organic carbon (SOC), total nitrogen (TN), N fractions, total phosphorus (TP), and available P in reclaimed soil at a copper mine site (2702 m above sea level) in the southeastern edge of the Tibetan Plateau, southwest China. The mine area had been reclaimed by distributing stockpiled topsoil (~15 cm) in 2008, sowing seeds of ryegrass ( Lolium perenne L.) in 2009, and planting seedlings of A. ferdinandi-coburgii and Rosa omeiensis f. pteracantha in 2010. We found that, eight years after the reclamation, although the concentrations of SOC (24.3 g kg − 1 ) and TN (2.21 g kg − 1 ) in the reclaimed soil increased by 25% and 29% compared with the stockpiled topsoil, respectively, they only accounted for about half of the levels in the undisturbed topsoil. In contrast, the concentration of TP (498–570 mg kg − 1 ) did not signiﬁcantly change between the reclaimed and stockpiled topsoil. The concentrations of NH 4 -N (30.1 g kg − 1 ), NO 3 -N (17.2 g kg − 1 ), and available P (11.1 mg kg − 1 ) in the reclaimed soil were 2.2, 1.3, and 1.6 times the levels in stockpiled topsoil, respectively, but still lower than those in undisturbed soil. The concentrations of microbial C, N, and P in the reclaimed soil had a similar variation pattern to the available nutrients. The soil C:P and N:P ratios and microbial biomass C:P and N:P ratios in the reclaimed soil were signiﬁcantly lower than the levels in the undisturbed forest soil. The average accumulation rates of SOC and TN in the reclaimed soil were 85.3 and 11.4 g m − 2 year − 1 , respectively. The rates are much lower than those of reclaimed mine soils with similar reclaim duration but better climate conditions. Relatively slow development of vegetation and soil microorganisms and leaching due to a freeze–thaw cycle controlled by the sub-alpine climate are likely responsible for the slow recovery of soil SOC and TN.


Introduction
Mining exploitation has caused irreversible damage to many natural ecosystems around the world. Reclamation is an effective technology to recover a disturbed ecosystem and restore its biodiversity and function [1,2]. Mine soil quality is a key indicator to assess the success of reclamation in mine sites [3][4][5] because the long-term development and stability of recovered plant communities are based on the systematic development and improvement of soil properties, including microbial functions [6,7]. Organic carbon (C) and nitrogen (N) pools in mine soil are massively degraded

Site Description
This study was conducted at an experimental reclamation project in a copper mine (chalcopyrite) (28 • 26 49" E, 101 • 41 43" E, 2702 m above sea level) in the southeastern edge of the Tibetan Plateau, southwest China ( Figure S1). The protolith of this mine is a set of argillaceous-arenaceous marine sedimentary rocks in the passive continental margin, containing a small amount of volcanic rocks. Metallogenic materials mainly come from the ore-bearing rocks characterized by multiple phases and many geneses. The formation occured during the Yanshan period (133-143 Ma) by middle to high hydrothermal fluid mineralization [28]. The mean annual temperature is 7 • C ( Table 1). The diurnal temperature range is 21 • C with the maximum annual temperature of 21 • C in July and the minimum annual temperature of −15.6 • C in January ( Table 1). The mean annual precipitation of 802 mm is concentrated in the summer season (Table 1). Soil freeze-thaw cycles occur from November to March (Table 1), showing a cryic temperature regime to pedogenic processes in this region. This climate regime is typical and common in the sub-alpine region in the southeastern edge of the Tibetan Plateau. The undisturbed forest in the vicinity of the mining area is dominated by alder (Alnus ferdinandi-coburgii) with a height of 2-10 m. The coverage percent of the undisturbed forests is 90-100% (Table 1). The soil in the undisturbed forest is classified as leached brown soil according to the Genetic Soil Classification of China (GSCC) [29] or Alfisols according to soil taxonomy (ST) [30]. The slope of the reclamation area is 5-10 • facing southwest. The experimental reclamation action was to cover stockpiled topsoil 15 cm deep in 2008. The topsoil was taken from the surface mineral soil (0-30 cm) before this area was mined. In April 2009, seeds of ryegrass (Lolium perenne L.) were sown. In April 2010, seedlings of A. ferdinandi-coburgii and Rosa omeiensis f. pteracantha were planted with a distance of about 5 m to each other. The survival rate of the two species was recorded as between 30% and 40% in 2017. The coverage percent of the restoration vegetation was 30-50% (Table 1). This reclamation method is common in most mine areas in the southeastern edge of the Tibetan Plateau.

Sampling and Analyses
In the reclaimed area, six plots (5 × 5 m) with a distance of >30 m from each other were selected to collect soil samples. In each plot, ten soil cores were collected by a soil auger. The depth of each soil core was 30 cm. The cores were divided into two layers to collect samples: 0-15 cm-reclaimed soil (R S ) and 15-30 cm-reclaimed spoil (R SP ) (Figure. S2). The collected sub-samples were mixed thoroughly to a composite sample for each horizon at each plot. Thus, we collected two composite samples from each plot, i.e., data of 12 composite samples from the reclaimed area. A litter horizon was not found in the reclaimed plots, therefore we did not collect litter samples.
In addition to R S and R SP , soil samples were also collected from the adjacent undisturbed forests (U TOP : 0-15 cm, U SUB : 15-30 cm) as reference soil ( Figure S2). The initial stockpiled topsoil (CK S ) and mine spoil (CK SP ) were collected as control groups ( Figure S2). Soil samples were stored in a cooler with ice (1-4 • C) before moving to a laboratory. Samples for the analysis of soil properties were sieved (<2 mm) to remove coarse rocks and roots. Then, the soil samples were divided into two parts and were air-dried or stored at 4 • C (fresh soils), respectively.
Soil bulk density (BD) was determined by excavating holes with a certain volume. The volume of excavated soil could be measured by backfilling with a known volume of water [31]. Samples for BD were oven-dried at 105 • C for 48 h, and the weight and volume of coarse fragment (>2 mm) were subtracted from the total soil weight and volume to calculate BD. Air-dried soil samples were used for the rest of analyses except for determining NH 4 -N, NO 3 -N and microbial biomass C, N, and P (MBC, MBN, and MBP). Soil pH was determined in H 2 O suspensions (soil:water ratio of 1:2.5) using a pH meter (FE20K, Mettler Toledo, Germany). The particle size distribution was determined by a laser particle size analyzer (Mastersizer 2000) after samples were treated with 3%H 2 O 2 (clay: <2 µm; silt: 2-20 µm; sand: 20-2000 µm). SOC and total N (TN) concentrations were measured using an element analyzer (Elementar vario MACRO cube, Germany). Total P (TP) was measured by a profile DV (USA Teledyne Leeman Labs) inductively coupled plasma atomic emission spectrometer (ICP-AES) (detection limit: 0.02 mg P/L) after digestion by refluxing with the nitric acid (6 mL), hydrofluoric acid (2 mL), and perchloric acid (1 mL). Available P was determined using the method proposed by Olsen and Sommers (1982) [32]. The concentrations of NH 4 -N and NO 3 -N were determined using a flow autoanalyzer (SEAL analytical AutoAnalyzer 3, Germany) after extraction by 1 M KCl solution. The contents of MBC, MBN, and MBP were measured by the chloroform fumigation-extraction method [33,34]. Two replicates were used when determining the above soil properties.

Calculations
The pools of SOC and TN (Pool, kg m −2 ) in different soil horizons were calculated using the following formula: where Conc., BD and D are the concentration of SOC or TN (g kg −1 ), bulk density (g m −3 ), and depth (cm) in each horizon, respectively. The accumulation rate of SOC and TN (R, g m −2 year −1 ) was calculated using the following formula: where Age is the reclamation time (8 years).

Statistics
A test of homogeneity of variances was applied to pH, the contents of sand, silt, and clay and the concentrations of SOC, TN, TP, NH 4 -N, NO 3 -N MBC, MBN, and MBP (Table S1). If the variance of a variable was not homogeneous, the variable was transformed by the natural logarithm function. Variables were subjected to a one-way analysis of variance (ANOVA) to test differences between R S , R SP , CK S , CK SP , U TOP , and U SUB (Table S2), after which LSD's post-hoc tests were performed (Table S3). Before a correlation analysis was applied, variables were tested for normality (Q-Q plots, Figure S3). The correlations between these variables were determined by Pearson's correlation coefficients. A significance level of p < 0.05 was used in this study (except where noted).

Soil Properties
The bulk density in the reclaimed soil (R S ) was lower than that in the initial stockpiled topsoil (CK S ) but higher than that in the undisturbed topsoil (U TOP ) ( Table 1). The proportion of clay in the R S was lower than that in the CK S and U TOP . The reclaimed spoil (R SP ) had more clay than the initial mine spoil (CK SP ). We found no significant difference of pH between the R s and CK s . In contrast, pH of the R SP increased by 1.1 units compared with the CK SP (Table 1).

Variations in soil C, N and P
The concentrations of SOC (24.3 ± 1.7 g kg −1 , mean ± SD) and TN (2.21 ± 0.12 g kg −1 ) in the R S increased significantly compared with the CK S (SOC: 19.4 ± 1.6 g kg −1 , TN: 1.71 ± 0.19 g kg −1 ), but only accounted for 44% and 54% of those in the U TOP , respectively (Figure 1a,b). The concentrations of SOC (2.46 ± 0.21 g kg −1 ) and TN (0.57 ± 0.03 g kg −1 ) in the R SP increased slightly compared with the CK SP (SOC: 1.81 ± 0.24 g kg −1 , TN: 0.43 ± 0.06 g kg −1 ). The concentrations of NH 4 -N (30.1 ± 5.8 g kg −1 ) and NO 3 -N (17.2 ± 3.7 g kg −1 ) in the R S were 2.2 and 1.3 times of those in the CK S (Figure 2a,b). The concentrations of NH 4 -N and NO 3 -N increased from 3.2 and 0.8 mg kg −1 in the CK SP to 13.2 and 5.5 mg kg −1 in the R SP , respectively. The concentration of NH 4 -N was higher than that of NO 3 -N in both the R S and R SP . However, there was higher concentration of NO 3 -N than NH 4 -N in both the U TOP and undisturbed subsurface soil (U SUB ) (Figure 2a

Soil Properties
The bulk density in the reclaimed soil (RS) was lower than that in the initial stockpiled topsoil (CKS) but higher than that in the undisturbed topsoil (UTOP) ( Table 1). The proportion of clay in the RS was lower than that in the CKS and UTOP. The reclaimed spoil (RSP) had more clay than the initial mine spoil (CKSP). We found no significant difference of pH between the Rs and CKs. In contrast, pH of the RSP increased by 1.1 units compared with the CKSP (Table 1).
The C:N ratios (mass ratio) in the R S (11.0 ± 1.1) and R SP (4.3 ± 0.3) did not significantly change compared with the CK S (11.4 ± 0.8) and CK SP (4.2 ± 0.3). They also were lower than the U TOP (13.4 ± 0.6) and U SUB (14.4 ± 0.4) (Figure 3a). In contrast, the C:P (44.3 ± 3.7) and N:P (4.1 ± 0.4) ratios in the R S were significantly higher than those in the CK S (34.4 ± 1.5 and 3.0 ± 0.3), although they were still largely lower than those in the U TOP (105.8 ± 10.7 and 7.9 ± 0.8) (Figure 3b,c). The C:N ratios (mass ratio) in the RS (11.0 ± 1.1) and RSP (4.3 ± 0.3) did not significantly change compared with the CKS (11.4 ± 0.8) and CKSP (4.2 ± 0.3). They also were lower than the UTOP (13.4 ± 0.6) and USUB (14.4 ± 0.4) (Figure 3a). In contrast, the C:P (44.3 ± 3.7) and N:P (4.1 ± 0.4) ratios in the RS were significantly higher than those in the CKS (34.4 ± 1.5 and 3.0 ± 0.3), although they were still largely lower than those in the UTOP (105.8 ± 10.7 and 7.9 ± 0.8) (Figure 3b,c).  [35]. The orange dash line represents the value of soil C:N, C:P, and N:P ratios in China [36]. Data are shown as means ± SD of 6 replicates. Different letters indicate significantly different variables between different samples at the p < 0.05 level.  in the R S were still lower than those in the U TOP (602 ± 59, 83.1 ± 9.3 and 18.3 ± 3.4 mg kg −1 ), they were comparable to those in the U SUB (210 ± 35, 22.4 ± 4.9 and 5.5 ± 0.9 mg kg −1 ). MBC:N (9.8 ± 1.4), MBC:P (25.9 ± 4.2), and MBN:P (2.6 ± 0.1) ratios in the R S were higher than those in the CK S and R SP (Figure 5a-c). Compared with the U TOP , the R S had a comparable MBC:N ratio but lower MBC:P and MBN:P ratio. The concentration of MBC, MBN, and MBP in the CK SP was not detected.

Rate of Accumulation of Soil C and N
After eight years of reclamation, the SOC and TN pools increased from 3.85 and 0.34 kg m -2 in the CKS to 4.37 and 0.40 kg m -2 in the RS, respectively ( Figure 6). The SOC and TN pools in the RSP increased by 0.16 and 0.03 kg m -2 compared with those in the CKSP, respectively. The SOC and TN pools in the RS were 46% and 56% of those in the UTOP. For the RS, the accumulation rates of SOC and the RS were still lower than those in the UTOP (602 ± 59, 83.1 ± 9.3 and 18.3 ± 3.4 mg kg ), they were comparable to those in the USUB (210 ± 35, 22.4 ± 4.9 and 5.5 ± 0.9 mg kg −1 ). MBC:N (9.8 ± 1.4), MBC:P (25.9 ± 4.2), and MBN:P (2.6 ± 0.1) ratios in the RS were higher than those in the CKS and RSP (Figure  5a-c). Compared with the UTOP, the RS had a comparable MBC:N ratio but lower MBC:P and MBN:P ratio. The concentration of MBC, MBN, and MBP in the CKSP was not detected.

Rate of Accumulation of Soil C and N
After eight years of reclamation, the SOC and TN pools increased from 3.85 and 0.34 kg m -2 in the CKS to 4.37 and 0.40 kg m -2 in the RS, respectively ( Figure 6). The SOC and TN pools in the RSP increased by 0.16 and 0.03 kg m -2 compared with those in the CKSP, respectively. The SOC and TN pools in the RS were 46% and 56% of those in the UTOP. For the RS, the accumulation rates of SOC and Figure 5. Comparisons of (a) microbial biomass carbon to nitrogen (MBC:N), (b) microbial biomass carbon to phsophorus (MBC:P), and (c) microbial biomass nitrogen to phosphorus (MBN:P) ratios (mass ratio) between the reclaimed soil (R s ), reclaimed spoil (R sp ), initial topsoil (CK s ), and adjacent undisturbed forest soils (0-15 cm, U top ; 15-30 cm, U sub ) in a sub-alpine copper mine site, southeastern edge of the Tibetan Plateau. The blue dash line represents the value of global soil MBC:N, MBC:P, and MBN:P ratios [35]. Data are shown as means ± SD of 6 replicates. Different letters indicate significantly different variables between different samples at the p < 0.05 level.

Rate of Accumulation of Soil C and N
After eight years of reclamation, the SOC and TN pools increased from 3.85 and 0.34 kg m −2 in the CK S to 4.37 and 0.40 kg m −2 in the R S , respectively ( Figure 6). The SOC and TN pools in the R SP increased by 0.16 and 0.03 kg m −2 compared with those in the CK SP , respectively. The SOC and TN pools in the R S were 46% and 56% of those in the U TOP . For the R S , the accumulation rates of SOC and TN were 65.0 and 7.3 g m −2 year −1 , respectively ( Table 2). For the R SP , the rates were 20.3 (SOC) and 4.1 (TN) g m −2 year −1 , respectively.  (Table 2). For the RSP, the rates were 20.3 (SOC) and 4.1 (TN) g m -2 year -1 , respectively.

Variation in Soil Properties
The decrease in the bulk density of the RS compared with the CKS is likely a result of the accumulation of SOC with revegetation (Table 1, Figure 1). The decline of bulk density caused by the input of organic matter occurs in many young natural and recovered ecosystems. For example, soil bulk density decreased with increasing SOC along natural chronosequences in the Tibetan Plateau [37] and in the European Alps [38]. The decline of bulk density with the restoration period has also been reported in many recovered ecosystems [8,14,39,40]. The increase in pH of the RSP compared with the CKSP (Table 1) is probably caused by the deposition of base cations leached from the upper RS.

Variation in Major soil nutrients
Comparing the RS with CKS, the significant accumulation of both SOC and TN (Figures 1a,b, and  6) and the slight decrease in TP (Figure 1c) showed a decoupled variation with the process of reclamation. The insignificant difference between the C:N ratios (Figure 3a) and the distinct changes in the C:P and N:P ratios (Figure 3b,c) between the RS and CKS also indicate this decoupled variation. In addition, the strong correlations between SOC, TN, MBC, and MBN and no correlation between TP and SOC, TN, MBC, MBN, and MBP (Table S3)  This difference is presumably because of the different ultimate sources and biogeochemical cycling of C, N, and P. Indeed, C, N, and P all can be returned to the soil by the litter of Lolium perenne L. and Astragalus membranaceus in our reclaimed site. However, the ultimate major sources of organic C and N in the soil are atmospheric through photosynthesis and fixation by plants and microorganisms, while the ultimate source of P is the weathering of primary mineral P [41,42]. As a result, these differences will gradually increase the C and N pools in soil but only "cycle" P from bottom to top horizons. More than that, the P content in soil tends to decrease with long-term pedogenesis [43]. Consequently, similar to our results, increasing C and N and decreasing soil P with the pedogenic process have been observed in many natural and reclaimed ecosystems in various climate types [7,[14][15][16]37,44,45].

Variation in Soil Properties
The decrease in the bulk density of the R S compared with the CK S is likely a result of the accumulation of SOC with revegetation (Table 1, Figure 1). The decline of bulk density caused by the input of organic matter occurs in many young natural and recovered ecosystems. For example, soil bulk density decreased with increasing SOC along natural chronosequences in the Tibetan Plateau [37] and in the European Alps [38]. The decline of bulk density with the restoration period has also been reported in many recovered ecosystems [8,14,39,40]. The increase in pH of the R SP compared with the CK SP (Table 1) is probably caused by the deposition of base cations leached from the upper R S .

Variation in Major soil nutrients
Comparing the R S with CK S , the significant accumulation of both SOC and TN (Figure 1a,b, and Figure 6) and the slight decrease in TP (Figure 1c) showed a decoupled variation with the process of reclamation. The insignificant difference between the C:N ratios (Figure 3a) and the distinct changes in the C:P and N:P ratios (Figure 3b,c) between the R S and CK S also indicate this decoupled variation. In addition, the strong correlations between SOC, TN, MBC, and MBN and no correlation between TP and SOC, TN, MBC, MBN, and MBP (Table S3)  This difference is presumably because of the different ultimate sources and biogeochemical cycling of C, N, and P. Indeed, C, N, and P all can be returned to the soil by the litter of Lolium perenne L. and Astragalus membranaceus in our reclaimed site. However, the ultimate major sources of organic C and N in the soil are atmospheric through photosynthesis and fixation by plants and microorganisms, while the ultimate source of P is the weathering of primary mineral P [41,42]. As a result, these differences will gradually increase the C and N pools in soil but only "cycle" P from bottom to top horizons. More than that, the P content in soil tends to decrease with long-term pedogenesis [43]. Consequently, similar to our results, increasing C and N and decreasing soil P with the pedogenic process have been observed in many natural and reclaimed ecosystems in various climate types [7,[14][15][16]37,44,45]. Table 2. Pools and accumulation rates of SOC and TN in reclaimed soils in mine areas with similar reclamation duration in the world.   The increase in the concentrations of NH 4 -N and available P in the R S (Figure 2a-c) reflect the improvement of soil available nutrient pools after revegetation. This is likely related to the increase in soil microbial activity according to their strong correlations to MBC, MBN, and MBP (Table S4). The ratio of MBC to SOC (MBC/SOC) increased from 0.63% in the CK S to 0.91% in the R S demonstrates the enhanced microbial activity due to revegetation. The high ratios of MBN to NH 4 -N (0.76) and NO 3 -N (0.51) also show the important role of microorganisms in transforming N [49]. In addition, the positive correlations of available P to SOC, TN, MBC, MBN, and MBP but not to TP (Table S4) and the high ratio of MBP to available P (0.78) further indicate that microbial transformation is a major mechanism of the increase in available P. Previous studies also related variations in soil nutrient status to microbial activities [19].

Accumulation Rates of C and N
The accumulation rate of SOC at our site is higher than that at a lignite mine site in Western North Dakota (No. 1 in Table 2, [46]) and a coal mine site in India planted with Tectona grandis L.f (No. 2A in Table 2, [18]), slightly higher than that at a sand mine site in Southern Poland (No. 3 in Table 2, [47]), comparable to that at a coal mine site recovered by forest in Southeastern Ohio, USA (No. 4C in Table 2, 10 years, [8]), and lower than that at other sites with a similar reclamation duration but higher MAT and MAP (No. 2B-D, 4A, 4B, 5A, 5B, 6, and 7 in Table 2, [8,14,15,17,18]. Our accumulation rate of TN is in the middle of the range (1.5-29.7 g m −2 ·year −1 ) of the previous studies with various climate and vegetation types ( Table 2).
The differences of the rates of SOC and TN accumulation between our site and other sites can be jointly attributed to the effects of climate and vegetation type. First, relatively higher precipitation and temperature favor the development of vegetation and microbial activities, which further facilitate the fixation of N and the return of C and N to soil. This can be reflected by the observation that higher accumulation rates of SOC and TN generally occurred in sites with higher MAT and MAP (Table 2). Second, vegetation type and its interaction with climate can also influence the accumulation of SOC and TN. This can be shown by that the accumulation rates of SOC and TN in our site are even higher than those in several sites with higher MAT and MAP (No. 2A, 3, and 6 in Table 2). The large differences (11-fold) of the accumulation rate of TN between the Sites 6 and 8 further show the influence of vegetation type. The MAT and MAP are comparable at the two sites but Site 8 was dominated by the locust (leguminous tree) while Site 6 was dominated by non-leguminous species (Table 2). Singh et al. (2004) also found that leguminous trees cause a larger increase in soil N pool than non-leguminous trees [50]. In addition, the relatively high content of total Cu in the R SP (~368 mg/kg, personal communication) maybe prevent the growth of plants and activity of microbes, inhibiting the accumulation of SOC and TN.
The accumulation rates of SOC and TN at our site were largely higher than those in natural soil during primary succession, similar to some previous reports [20,47]. For example, our accumulation rate of SOC is 2.6-12.0 times of that at three "young" chronosequences (120-150 years) in the Swiss Alps [51][52][53]. Our accumulation rate of TN is 2.7-5.7 times of that at a lava chronosequence (142 years) in tropical Hawaii [44]. This is no surprise because the parent materials of natural successional soils are often lacking in C and N [52], while the reclaimed soil often has a certain amount of legacy C and N (Figure 1a,b), which can facilitate the rapid establishment of microorganisms and vegetation. As a consequence, considerable amounts of SOC and N can be rapidly fixed or returned to soil by litters in reclaimed sites.

Slow Increase in Soil Nutrient Pools in Sub-Alpine Mine Site
For our reclaimed soil, it would take around 124 years to reach SOC equivalence with the undisturbed forest soil, if keeping the current rate (Table 2). Because many previous studies have reported that the accumulation rate of SOC decreases with soil development [45,51], the actual recovery duration of SOC would be longer than this estimation. This recovery duration is much longer than that found in reclaimed sites with better climate conditions [4,14,40]. For example, the reclaimed site would need less than 20 years to achieve the SOC content of the reference forest in tropical Ranchi district, India [14]. Mukhopadhyay et al. (2014) found that soil quality after 17 years reclamation was comparable to that in a nearby undisturbed forest in a tropical coal mine site in India [4]. For a coal mine in a temperate zone, northern China, Yuan et al. (2018) found that the duration required to reach the nutrient level of undisturbed site was about 10 years [40].
The slow increase in soil SOC and TN at our site is likely related to the relatively slow development of vegetation (Table 1) and microbial activities controlled by the relatively cold and dry climate. This point can be supported by the variation in available nutrients, microbial biomass, and C:N:P ratios. First, although the concentrations of NH 4 -N and NO 3 -N at the R S increased compared with the CK S , they only accounted for 67.5% and 26.6% of those in the U top . The lower content of NO 3 -N content than NH 4 -N in the R S further illustrates that it will need a long time to reach the level of nutrients at the undisturbed site where the content of NO 3 -N is higher than NH 4 -N (Figure 2a,b). The low temperature and the long frost period (Table 1) is probably an important factor causing weak nitrification and, thus, the low content of NO 3 -N in the R S . In addition, the leaching of soluble NO 3 -N is also likely responsible for the low content of NO 3 -N in R S . This can be shown by the existence of NO 3 -N in the subsurface R SP (Figure 2b). The undetectable MBC and MBN in the R SP indicate that the NO 3 -N in the R SP came from the upper R S horizon by leaching. The occurrence of leaching at our reclaimed site is also demonstrated by the increase in clay content of the R SP and the decrease in clay content of the R S (Table 1).
Second, the concentrations of MBC, MBN, and MBP in the R S were 36.5%, 27.5%, and 47.3% of those in the U TOP (Figure 4a-c), showing the relatively low rate of microbial nutrient transformation controlled by the low temperature. This is also reflected by the observation that the MBC/SOC (0.9%), MBN/TN (1%), and MBP/TP (1.6%) in the R S were lower than those in the U TOP , and much lower than those in various types of soils around the world [54][55][56][57]. In addition, even in low altitudinal areas, it will take several decades or more for microbial biomass and diversity in disturbed soils to reach the levels of undisturbed soils [58,59]. Therefore, the slow recovery of soil microbial function is likely a reason for the slow recovery of soil nutrient pools at our site.
Finally, although the C:N ratio in the R S approached the ratios in China and global soils (Figure 3a) [35,36], the much lower C:P and N:P ratios than the global values (Figure 3b,c) indicate the insufficient contents of SOC and N in the R S . Similar to the soil C:N:P ratios, the MBC:N ratio in the R S increased to the same level of the U TOP ( Figure 5a); however, the MBC:P ratiois slightly lower than the global value and the MBN:P ratio much lower than that in the U TOP , U SUB , and global level (Figure 5a) [35]. These results illustrate not only the lack of C and N in the R S , but also the likely N limitation to the microbial community.

Conclusions
The SOC and TN pools in the reclaimed soil after eight years of reclamation increased significantly, compared with the initial stockpiled topsoil in the sub-alpine copper mine area, southeastern edge of the Tibetan Plateau, southwest China. In contrast, a significant difference in the TP pool was not observed. The SOC and TN pools in the reclaimed soil accounted for 46% and 56% of those in the adjacent undisturbed forest soil. The concentrations of MBC, MBN, and MBP in the reclaimed soil were only 37%, 27%, and 48% of the levels in the undisturbed soil, respectively. The relatively slow increase in soil SOC and TN pools are likely related to the relatively cold and dry sub-alpine climate, which does not favor the fast development of vegetation and microbial activities.
Supplementary Materials: The following are available online at http://www.mdpi.com/1999-4907/10/12/1069/s1, Figure S1: The location of the reclaimed area and the undisturbed forest in a sub-alpine copper mine site, southeastern edge of the Tibetan Plateau, Figure S2: Six classes of soil and spoil samples in this study, Figure S3: Q-Q plots of 11 variables in a sub-alpine copper mine site, southeastern edge of the Tibetan Plateau, Table S1: Test  of Homogeneity of Variances, Table S2: Results of ANOVA analysis of 11 variables, Table S3: Pearson's correlation coefficients between the concentrations of soil C, N, P, microbial biomass C, N and P (MBC, MBN and MBP) in a sub-alpine copper mine site, southern edge of the Tibetan Plateau (N = 30), Table S4: Pearson's correlation coefficients between the concentrations of available P, NH4-N and NH4-N and the soil C, N, P, microbial biomass C, N and P (MBC, MBN and MBP) in a sub-alpine copper mine site, southern edge of the Tibetan Plateau (N = 30).

Author Contributions:
The authors contributed equally to the conceptualization of the research design and the writing of manuscripts for this paper. The methodology, the preparation of empirical investigations, the collection of primary data in the field, data curation and the formal analysis was done by Y.C. The final review and editing of the manuscript was conducted by Y.C. and J.Z. The project administration as well as the funding acquisition was done by J.Z.