Accumulation of As, Ag, Cd, Cu, Pb, and Zn by Native Plants Growing in Soils Contaminated by Mining Environmental Liabilities in the Peruvian Andes

The capability of native plant species grown in polluted post-mining soils to accumulate metals was evaluated in view of their possible suitability for phytoremediation. The study areas included two environmental liabilities in the Cajamarca region in the Peruvian Andes. The content of As, Ag, Cd, Cu, Pb, and Zn was determined in individual plant organs and correlated with soil characteristics. The degree of the pollution depended on the metal with results ranging from uncontaminated (Cd) to moderately (Zn), strongly (As, Cu), and extremely contaminated (Pb, Ag) soils. The metals were mainly present in the fractions with limited metal mobility. The bioaccumulation of the metals in plants as well the translocation into overground organs was determined. Out of the 21 plants evaluated, Pernettya prostrata and Gaultheria glomerate were suitable for Zn, and Gaultheria glomerata and Festuca sp. for Cd, phytostabilization. The native species applicable for Cd phytoremediation were Ageratina glechonophylla, Bejaria sp., whereas Pernettya prostrata Achyrocline alata, Ageratina fastigiate, Baccharis alnifolia, Calceolaria tetragona, Arenaria digyna, Hypericum laricifolium, Brachyotum radula, and Nicotiana thyrsiflora were suitable for both Cd and Zn. None of the studied plants appeared to be suitable for phytoremediation of Pb, Cu, As and Ag.


Introduction
Mining has been around since ancient times and has been essential for economic and industrial development of societies [1]. The extraction of natural resources has generated an economic surge in Latin America [2]. During the recent decades, the mining sector has been one of the main economic pillars of Peru [3], making it the main producer of gold, zinc, lead, and tin in Latin America and the second largest producer of copper, silver, and zinc worldwide [4].
Nevertheless, mining operations are also responsible for significant environmental damage by producing a large amount and diversity of residues, affecting the quality of water, soil, and air [5], of which soil pollution is one of the most crucial environmental problems [6]. Likewise, the abandoned mines without effective shut-down procedures are causing a nuisance in the mining districts due to metal transfer to the environment [7], being one principal reason for the emergence of conflicts in Peru [8]. Metals are persistent in the environment and, being nonbiodegradable [9], cause serious harm bioaccumulating in flora and fauna [6,10]. The installations, effluents, emissions, remains, or deposits of residues from abandoned or inactive mines constituting a permanent and potential risk for health, the surrounding ecosystem, and property are defined as mining environmental liabilities (MELs) [11,12]. The lack of adequate legislation has caused MELs to increase

Study Area and Sampling
The sampling sites were located in the Andes of northern Peru, in the Hualgayoc district, in the west of the province of Hualgayoc, between the districts of Chugur and Bambamarca in the department of Cajamarca.
Hualgayoc district is located in the west of the province of Hualgayoc in the department of Cajamarca in the Andes of Northern Peru. It is ca. 45 km away, as the bird flies, from Cajamarca city and 865 km by road from Lima city [25]. Since Spanish colonial times, Hualgayoc has been famous for Ag-rich polymetallic mineralization, in addition to Zn, Pb, and Cu [23,24,26]. The main mineralized structures in the Hualgayoc district are mantos and veins hosted largely in the upper Goyllarisquizga Group, the Inca Formation and lower Chulec Formation rocks [23,24,26]. The Hualgayoc district has 943 MELs [14], two of them belonging to the ex-mining unit Los Negros have been studied here; they are found on the left bank and downstream from the Hualgayoc river creek in the flak of the Llaucano river valley. The identity of the liable parties of the generation of these MELs have not been identified yet [14]. In order to evaluate the environmental impact of the abandoned mines the two chosen sampling sites included a leach pad (S1) located on top of a mine waste deposit and a mine dump (S2) made up of clay gravel with sand.
Sampling area #1 leach pad (a latitude 6 • 44 49 S, longitude 78 • 35 36 W, and altitude of 3249 m above sea level) was located on top of a mine waste deposit and had no drainage. Sampling area #2 mine dump (latitude 6 • 44 53 S, longitude 78 • 35 49 , and altitude of Sampling area #1 leach pad (a latitude 6°44′49″ S, longitude 78°35′36″ W, and altitude of 3249 m above sea level) was located on top of a mine waste deposit and had no drainage. Sampling area #2 mine dump (latitude 6°44′53″ S, longitude 78°35′49″, and altitude of 3399 m above sea level) had granular material made up of clay gravel with sand, no stabilization work was observed and there was no revegetation nor drainage ( Figure 1). Soil and plant samples were taken following the criteria based on the distance from the point of contamination, wind direction, slope and inclination, vegetation cover, and soil texture [27]. The location of the points for soil sampling in both sites was fixed at 0, 15, 30, 45 and 60 m away from the point of contamination, forwards downhill slide. For the determination of soil profile characteristics, samples were taken at two depths between 0-15 and 15-30 cm intervals. Following this pattern, a total of 20 soil samples were collected from the two sites. Plants and soil samples were collected in July 2018 (dry season). Sampling of flora specimens was carried out in the same locations as for soil sampling. To do so, an area of 4 m 2 around each soil sampling point was delimited. Each plant was stored in polyethylene bags before being carried to the laboratory. A total of 47 plant samples (from 1 to 5 replicates) were collected from the two sampling sites, including 29 samples from site #1, and 18 samples from #2. Five species were found in both areas so a total of 21 unique species of native flora have been identified. The taxonomic identification of the collected plants was carried out by Dr. Manuel Timaná (the Pontifical Catholic University of Peru, Peru) and Mg. Paul Gonzales Arce (the Laboratory of Floristics of the Herbarium of the Natural History Museum of the Universidad Nacional Mayor de San Marcos, Peru).

Chemical Analysis of Soils
The soil samples were analyzed for pH, electrical conductivity, organic matter and texture at the Soil, Plant, Water and Fertilizer Analysis Laboratory of the Universidad Nacional Agraria La Molina in Lima (Peru). pH was measured using a suspension of soil Soil and plant samples were taken following the criteria based on the distance from the point of contamination, wind direction, slope and inclination, vegetation cover, and soil texture [27]. The location of the points for soil sampling in both sites was fixed at 0, 15, 30, 45 and 60 m away from the point of contamination, forwards downhill slide. For the determination of soil profile characteristics, samples were taken at two depths between 0-15 and 15-30 cm intervals. Following this pattern, a total of 20 soil samples were collected from the two sites. Plants and soil samples were collected in July 2018 (dry season). Sampling of flora specimens was carried out in the same locations as for soil sampling. To do so, an area of 4 m 2 around each soil sampling point was delimited. Each plant was stored in polyethylene bags before being carried to the laboratory. A total of 47 plant samples (from 1 to 5 replicates) were collected from the two sampling sites, including 29 samples from site #1, and 18 samples from #2. Five species were found in both areas so a total of 21 unique species of native flora have been identified. The taxonomic identification of the collected plants was carried out by Dr. Manuel Timaná (the Pontifical Catholic University of Peru, Peru) and Mg. Paul Gonzales Arce (the Laboratory of Floristics of the Herbarium of the Natural History Museum of the Universidad Nacional Mayor de San Marcos, Peru).

Chemical Analysis of Soils
The soil samples were analyzed for pH, electrical conductivity, organic matter and texture at the Soil, Plant, Water and Fertilizer Analysis Laboratory of the Universidad Nacional Agraria La Molina in Lima (Peru). pH was measured using a suspension of soil in deionized water at a ratio 1:1 using a Consort pH-meter (Turnhout, Belgium) and electrical conductivity was measured in aqueous extract of soil in water at a ratio 1:1. Organic matter content was determined using the Walkley and Black method [28] based on the oxidation of organic carbon with potassium dichromate, sulfuric acid, and concentrated phosphoric acid. Texture evaluation was conducted using the classical Bouyoucos Hydrometer method-the mechanical analysis of soils utilising the suspension of solids for the quantification of the content of sand, silt, and clay in percentage [29].
Metal content in soils was analyzed at the SGS Geochemical Assays Laboratory, Lima, Peru. Approximately 0.20 g of soil with particle size below 106 µm was digested in two steps in Hot Blocks (Environmental Express, Charleston, SC, USA). For the first digestion step (at 90 • C for 30 min), 1 mL of concentrated nitric acid was added; for the second one (at 90 • C for 1 h), 3 mL of concentrated hydrochloric acid was added. Once the digest cooled to room temperature, 2 mL of concentrated hydrochloric acid was added and the mixture was diluted with ultrapure water (18.2 MΩ·cm) up to 20 mL. The concentration of metals was determined by an inductively coupled plasma emission spectrometer (ICP-OES) using a Perkin Elmer Model Optima 8300 DV instrument. The certified soil reference materials OREAS 906, OREAS 907, and OREAS 522 (from ORE Research & Exploration Pty Ltd., Bayswater North, Victoria, Australia) were used for quality control.

Calculation of the Geo-Accumulation Index (I geo )
The geo-accumulation index (I geo ) was introduced by Muller [30] for the evaluation accumulation of metals in river sediments. Since then it has been adopted by many researchers in order to classify sediments and soils from the point of view of metal pollution level. The values necessary to calculate I geo are concentration of metal in given soil and the background level concentrations. In our study, baseline values obtained by Santos-Francés et al. [22] from La Zanja region at ca. 30 km from the sampling place are used for calculation purposes, except for silver, for which the natural abundance in the Earth's crust was used [31]. The values used for calculations are given in Supplementary Information (Table S1).
The I geo was calculated according to the equation: where C i is the measured concentration of the i metal examined in the soil and B i is the background level of this metal. The factor 1.5 is used to correct possible variations in the background values of a particular metal in the environment. Depending on the value of the I geo , the soils are classified into seven groups [32]: unpolluted (I geo < 0), unpolluted to moderately polluted (0 ≤ I geo < 1), moderately polluted (1 ≤ I geo < 2), moderately to strongly polluted (2 ≤ I geo < 3), strongly polluted (3 ≤ I geo < 4), strongly to very strongly polluted (4 ≤ I geo < 5), and very strongly polluted (I geo ≥ 5).

Sequential Extraction Procedure
In order to identify how the metals are distributed among the soil components, the soil samples were subjected to sequential extraction based on the Tessier approach [33], using the modified method [34] adapted to the soil parameters consisting of the four following steps: All samples were centrifuged at 400 rpm for 10 min and filtered (0.45 µm). Metals present in the supernatants were quantified by ICP mass spectrometry (ICP-MS) using the reaction cell mode pressurized with He and H 2 gas. The dilutions performed allowed work in the calibration range of 0.01-10 ppb; an 8-point calibration curve was used. The isotopes monitored were 107 Ag, 109 Ag; 75 As; 111 Cd, 112 Cd, 114 Cd; 63 Cu, 65 Cu; 206 Pb, 207 Pb, 208 Pb; 64 Zn, 66 Zn, and 68 Zn. Analytical blanks were run in parallel.

Chemical Analysis of Plants
The vegetal material was washed with tap water (for the disposal of soil remains) and then with distilled water. All the plant samples were separated into leaves, stems, and roots, which were dried at 30-40 • C. Foliar samples were crushed and ground, and 0.10 g subsamples were digested in the heat block (SCP Science, Courtaboeuf, France) in triplicate. The digestion was carried out in two steps: the first one using 2 mL of nitric acid (70%) at 85 • C for 4 h 30 min and the second one using 0.75 mL of hydrogen peroxide (30%) at 85 • C for 4 h 30 min. The digest was diluted to 50 mL with ultrapure water (18.2 MΩ·cm). The elemental composition was analyzed by ICP mass spectrometry (ICP-MS) using an Agilent 7500 (Agilent, Tokyo, Japan) instrument equipped with a reaction cell pressurized with He and H 2 gas. The dilutions performed allowed work in the calibration range of 0.  66 Zn, and 68 Zn. Analytical blanks were analyzed in parallel. The Standard Reference Material 1573a Tomato Leaves of the NIST (Gaithersburg, MD 20899, USA) was used for quality control.

Calculation of the Translocation Factor (TF) and the Bioconcentration Factor (BCF)
The translocation factor (TF) and the bioconcentration factor (BCF) were calculated to assess the phytoremediation potential of the studied plants. The BCF is defined as the ratio of the metal content accumulated in the plant to the content in the soil [35], as given below: BCF= C p /C so (2) where C p is the metal concentration in the plant (shoot) and C so is the total metal concentration in the soil. Plants with a BCF value higher than one are considered to be suitable candidates for phytoextraction [36,37]. The ability of the trace metal transfer from the roots to the aerial part is expressed using the translocation factor TF [38], calculated as follows: where C s is the metal concentration the aerial part (shoot) of the plant and C r is the concentration in roots. Plants with TF values higher than one can be potentially used for phytoextraction [38].

Statistics
The measurements were carried out in triplicates and the results with relative standard deviation higher than 10% were discarded and the measurements repeated. Values are reported as mean ± standard deviation (SD) of three replications.
In order to assess the differences in I geo values for the studied elements between the soil samples at each site, we first performed a normal test to determine whether the data was normally distributed. We found that the null hypothesis of normal distribution could not be rejected for any of the samples. Therefore, we decided to apply a t-test in order test for differences between each pair of elements. We applied the Bonferroni correction for multiple testing, which resulted in a significance threshold of alpha = 0.01/n, where n = 15 was the number of possible pairs of elements.

Soil Physicochemical Characterization
Physicochemical properties of the soils are given in Table 1. The two sites had similar extremely acidic pH values in the range of 3.59-4.19 in site #1 and 2.77-4.02 in site #2. As a result, no carbonates and low organic matter contents were present. Soil was nonsaline and had low electrical conductivity (EC) with an average value of 0.08 dSm −1 in site #1 and of 0.15 dSm −1 in site #2. In the texture of soils, the sand fraction dominated at 42-66% in site #1 and at 52-62% in site #2, varying from sandy clay loam to clay loam textures in the different layers. The mineralogical composition of soils was dominated by illite/sericite, kaolinite, quartz, and jarosite [39]. The most important factor affecting trace metals availability is soil pH [1], which usually increases the mobility and bioavailability of the metal with lower pH values, with the content of organic matter and clay minerals [10,40]. On the other hand, soils rich in organic matter can bind metals efficiently in acidic conditions, however, bonding capacity depends on the metal [10]. The samples had fine granulometry due to the major presence of sandy fraction; the fine soil particles can be easily transported by wind and water erosion to the nearby environment [41]. The acidity of the soil and their unbalanced properties pointed to the poor soil management [42]. Santos-Francés et al. [22] at Colquirrumi mine, located 6 km NW of our study area, reported acid soils (pH = 4.6), low organic matter content (5.04%), and silt sand texture (29.5% sand, 41.5 silt, and 30.8% clay); also, soils Colquirrumi are on quartzite and dacite substrates, which are much less reactive [39]. These results show significant differences in comparison with the study reported by Bech et al. [43] for the Carolina Mine, located 4 km SW of the study area, which reported basic pH of soil (i.e., pH = 6.8-8.0), high calcium carbonates levels, slight electrical conductivity (0.1-0.8 dSm −1 ), and silt loam texture.

Soil Metal Composition
Following the initial screening of 34 elements [39], the study of Pb, Zn, As, Cu, Ag, and Cd was found of the highest interest because of their extremely high contents and potential toxicity. The results obtained for these elements for 15, 30, 45, and 60 m distances from the points of contamination at two different depths (0-15 and 15-30 cm) are presented in Figure 2a,b for sampling sites #1 and #2, respectively. In view of the findings, it can be concluded that the whole studied area was characterized by relatively uniform pollution. Moreover, the analysis results (not shown) of soil from two presumably uncontaminated sites in the vicinity of the sampling areas still showed high concentrations of the studied elements.
In general, neither significant differences nor a particular trend in the concentration levels for any of the studied elements for the soils from sampling site #1 (Figure 2a) was observed, except for Ag, whose concentrations in the upper soil layer (0-15 cm) were regularly higher than in the lower (15-30 cm) one. In one particular sample (the upper layer of the soil taken at 60 m from the leach pad, the presumed contamination source), significantly higher (twice as high as the next highest value) concentrations of Pb, Ag, and Cd suggested the presence of a grain of rock enriched in these three elements. Among the elements studied, the most abundant metal was Pb, found at the average level of 2050 mg kg −1 . Zinc concentration was in the range between 287 and 724 mg kg −1 , and was at the same level as that observed for Cu concentration in the range between 245 and 512 mg kg −1 , and As concentration was in the range between 252 and 401 mg kg −1 . Much lower was the concentration of Ag, in the range between 4.1 and 20.7 mg kg −1 , and of Cd, with an average concentration of 1.5 mg kg −1 .
Concerning the results obtained for sampling site #2 (Figure 2b), a point with unusually high contents of Pb, Ag, and Cd was recorded in the upper layer of the mine dump (the presumed contamination source). Apart from this observation, no other clear trend in the measured concentrations was observed. In addition, for this sampling area (#2), lead was the metal yielding the highest concentrations with values between 934 and 2367 mg kg −1 . The levels of Zn, Cu, and As remained comparable with the concentrations of Zn ranging from 120 to 305 mg kg −1 , Cu from 131 to 283 mg kg −1 , and As from 119 to 343 mg kg −1 . Nevertheless, two exceptionally high values of ca. 730 mg kg −1 were recorded for As for both layers of the soil sample located farthest from the point of contamination. The average values for these three elements were significantly lower than in the case of sampling area #1. The total Ag concentration was in a range of several mg kg −1 , and that of Cd was in general below 1 mg kg −1 . The Peruvian Environmental Quality Standards (PEQS) for soil set the maximum concentration permitted for Pb, As, and Cd in dried soils for agricultural uses at 70, 50, and 1.4 mg kg −1 , respectively [44], which coincides with the Pb and Cd values described in the Canadian Soil Quality Guidelines (CSQG), being stricter for As with 12 mg kg −1 [45]. In sampling sites #1 and #2, lead was the metal with the highest concentrations considerably exceeding the maximum values allowed by soil regulations; likewise, As content also showed great excess. However, most Cd concentrations recorded low values, except in site #1 at the furthest distances from the point of contamination (45 and 60 m), where Cd values slightly exceeded the permitted standards. The PEQS do not contemplate maximum values for Zn, Cu, and Ag, unlike the Canadian soil standards, which set permitted values at 250, 63 and 20 mg kg −1 , respectively [45,46]. According to Canadian standards, Zn exceeded the maximum allowed values in sampling site #1, whereas at sampling site #2 the concentration of Zn was above the legal threshold only at 0, 30, and 45 m away in the upper soil layer (0-15 cm). Copper concentrations significantly exceeded the maximum values allowed in both studied sites, and Ag was above the Canadian legal threshold only in samples from site #1 located furthest from the point of contamination (45 to 60 m). On the other hand, the Colquirrumi mine reported high content of heavy metals largely exceeding maximum allowed concentrations according to both PEQS and CSQG (Pb = 2069 mg kg −1 , Zn = 1893 mg kg −1 , Cu = 198 mg kg −1 , As = 428 mg kg −1 , and Cd = 13 mg kg −1 ) [22,39]. In addition, the Carolina mine, when it was an active mine in the same Hualgayoc district, reported high concentrations of Pb, Zn, Cu, and As (3992-16,060, 11,550-28,059, 256-2070, and 280-1030 mg kg −1 , respectively) [47]. In addition, Table 2 shows concentration levels of Pb, Zn, Cu, As, Ag, and Cd found reported recently in other abandoned mines in Latin America, the majority of which are with acid soils. Lead, As, Ag, Cd, as well as excess of Zn and Cu are toxic, nonbiodegradable and persist for long periods in the environment [6,21]. In strongly acidic conditions, as observed in our study case, these metals are released and cause potential hazards constituting a serious source of pollution to the surrounding areas [52], affecting directly or indirectly soil, water, flora, fauna, and human health [21]. The concentrations of Pb, Zn, As, Cu, Ag, and Cd studied in the mining environmental liabilities exceeded Peruvian and Canadian soil standards, suggesting a potential risk of the contamination. Thus, there is a vital need to implement sustainable techniques for the remediation of these mining sites [21].

Index of Geoaccumulation
The degree of soil pollution in the studied areas was assessed using the index of geo-accumulation (I geo ). I geo values for the studied metals are summarized in Figure 3. In general, the degree of pollution was similar in the two areas with results ranging from uncontaminated to extremely contaminated, depending on the metal. Considering Cd, the soils could be considered as unpolluted (I geo ≤ 0), whereas pollution with Zn was in the range from moderate to strong. Similarly, although higher pollution levels were observed for As and Cu. The highest I geo values were obtained for Pb (from strongly to very strongly contaminated) and for Ag (very strongly polluted). Finally, after the statistical analysis (detailed in Section 2.7) the calculated geo-accumulation index showed the following order of contamination levels for both examined areas: Cd < Zn < As ∼ = Cu < Pb < Ag.

Sequential Extraction Study from the Soil
The total content of metals in soils does not fully reflect their potential environmental risk; from the ecological point of view, it is important to know which part of the total concentration of a metal is available to determine its toxic effect [33,34]. The proportions of metal fractions that determine the availability and mobility of trace elements in the soil and within the ecosystem, vary depending on the mineralogical composition of soils and other factors such as pH, organic matter content, and charge characteristics [53].
In our study, Pb, Zn, Cu, As, Ag, and Cd present in soil samples were separated into four fractions, providing a closer insight into their potential bioavailability. The fractions included (i) exchangeable metals forms, followed by metals bound to (ii) hydrated iron and manganese oxides, (iii) organic matter and, finally, (iv) mineral metal fraction. The partitioning of metals from the sampling areas #1 and #2 are presented in Figure 4a,b,   Figure 3. Geo-accumulation index (I geo ) for Pb, Zn, Cu, As, Ag, and Cd at the sampling sites.

Sequential Extraction Study from the Soil
The total content of metals in soils does not fully reflect their potential environmental risk; from the ecological point of view, it is important to know which part of the total concentration of a metal is available to determine its toxic effect [33,34]. The proportions of metal fractions that determine the availability and mobility of trace elements in the soil and within the ecosystem, vary depending on the mineralogical composition of soils and other factors such as pH, organic matter content, and charge characteristics [53].
In our study, Pb, Zn, Cu, As, Ag, and Cd present in soil samples were separated into four fractions, providing a closer insight into their potential bioavailability. The fractions included (i) exchangeable metals forms, followed by metals bound to (ii) hydrated iron and manganese oxides, (iii) organic matter and, finally, (iv) mineral metal fraction. The partitioning of metals from the sampling areas #1 and #2 are presented in Figure 4a,b, respectively. In general, no significant differences were found in the fractionation of Pb, Zn, Cu, As, Ag, and Cd between the upper (0-15 cm) and lower (15-  In general, no significant differences were found in the fractionation of Pb, Zn, Cu, As, Ag, and Cd between the upper (0-15 cm) and lower (15-30 cm) layers of soil. Out of 60 paired samples, only 6 (site #1: Zn-15 m, Ag-45 m and Cd-30 m and in site #2: Ag-0 m, Cd-0 m, Pb-0 m) displayed differences higher than +10% in the contributions of individual fractions into the total element content. The study of the distribution of elements in the soil showed the dominant presence of the fractions with limited metal mobility.
In sampling site #1, except for Cu, samples collected at 0 and 15 m from the leach pad presented notable differences from the three remaining ones collected at 30, 45, and 60 m. Copper was almost uniformly distributed among the mineral fraction and bound to organic matter and hydrated iron and manganese oxides; the remaining several percent of this metal was an exchangeable fraction. The two locations close to the presumed contamination source were characterized by the highest fraction of hydrated iron and manganese oxidesbound Pb, Ag, and Zn, whereas As was almost exclusively contained in the mineral fraction. Samples taken at 30, 45, and 60 m from the leach pad were characterized by a significant proportion of Pb, Ag, Zn, and As bound to organic matter, for which As reached more than 90% of its total content. The distribution of Cd was very variable with its total content distributed among all four fractions. The exchangeable fraction was absent for Ag, As, and Zn at 0 and 15 m from the leach pad.
The fractionation of elements from sampling site #2 is shown in Figure 4b. In general, the distribution of elements among the fractions did not vary with the distance from the contamination source. Cadmium was the only of the studied metals for which the exchangeable fraction was present at significant proportions (between 20 and 60%). Most of As (70-90%) was contained in the mineral fraction, whereas the majority of Ag was bound to hydrated iron and manganese oxides. Zinc was distributed among three main fractions including mineral, organic matter, and hydrated iron and manganese oxides, similar as in the case of copper where, additionally, 5-20% of the exchangeable fraction was present. In this area, the distribution of Pb was complex and completely different than in area #1, with a significant portion represented by mineral forms.
The first exchangeable metal fraction is considered to contain the most mobile metal species and usually those that bestow surge to toxicity problems, being available as ions [53]. Indeed, our study indicates low exchange capacity of the studied soils. The potentially high bioavailability can be also expected for metal forms contained in organic matter [54]. Metal bound with oxides Fe-Mn represents one of the dominant fractions; it indicates the role of Fe or Mn oxides in the immobilization of metals similar to that in the residual fraction, which implies the binding of metals in minerals [55]. According to the available literature on contaminated soils, the phase distribution of trace metals is the highest in Fe-Mn oxide or in the residual fractions [40], from which metals are mostly unavailable to plants [54]. Despite the highest metals content being found in the fractions with limited mobility, the acidic conditions of the soils could favor their mobility and availability [53].

Trace Elements in Native Plants
Metals transport and distribution in plant tissues are impacted by the level of metal available and by the plant species [21]. In order to evaluate whether the native plants that grow in the two mining environmental liabilities could be indicators of potential metal mobility, metal concentration was evaluated in individual plant organs (leaves, stem, and roots). The results of the trace elements determination in organs of plants studied are given in Tables S1 and S2. Different plants have different abilities in the absorption of minerals, and the species that survive naturally in metalliferous soils are often only restricted to this type of area [56]. Metal availability can also be modified by plant roots through pH regulation affecting plant metal uptake [57,58]. Among the 21 plants collected, half of them were single specimens growing in one particular location, whereas for the other half, two to five individual specimens were found at different locations.
Copper: The concentrations of Cu in the majority of the native plants were the highest in the roots, of which, Ageratina glechonophylla (40 mg kg −1 -S1), Calamagrostis recta (55 mg kg −1 -S1), and Arenaria digyna (72 mg kg −1 -S1) presented the most elevated values. On the other hand, species having a higher concentration of metal in the leaves than in the roots were Achyrocline alata (16 mg kg −1 -S1 and 10 mg kg −1 -S2) and Hypericum laricifolium (11 mg kg −1 -S2). The only plant where the concentration in the aerial parts was higher than the normal range of 5-30 mg kg −1 [40] was Nicotiana thyrsiflora (49 mg kg −1 -S1). Another study carried out close to our sampling area, at the Carolina mine, also reported high values in different plant species studied (38-542 mg kg −1 in shoots and 43-396 mg kg −1 in roots); moreover, the values reported for the shoots of Achyrocline alata (38-130 mg kg −1 ) [47] were higher, exceeding the range for aerial parts (30 mg kg −1 ). Studies in countries bordering Peru, such as Ecuador, Chile, and Brazil, showed Cu concentrations in shoots at 1-85, 77-988, and 32-137 mg kg −1 , respectively [47,56]; likewise, other concentrations found were 40-243 mg kg −1 in Armenia (elevated) [38], and 20-29 mg kg −1 [60] in China. The results showed low Cu concentration in leaves, which may be due to low Cu translocation from roots to aerial parts, where it could interfere with photosynthesis and other essential processes [47].
Silver: Silver concentrations in all the studied native flora species were low, with roots showing higher contents than aerial parts. Nonetheless, some species exceeded the normal threshold of 0.5 mg kg −1 in their leaves [40], including Ageratina glechonophylla (0.75 mg kg −1 -S1), Brachyotum radula (1.11 mg kg −1 -S1), Arenaria digyna (2.28 mg kg −1 -S1), and Nicotiana thyrsiflora (2.97 mg kg −1 -S1). Previous studies showed Ag contents between 0.22-80 mg kg −1 [6]. The amount of Ag absorbed by plants is related to the concentration of soils; in many cases, Ag can be concentrated by plants to attain toxic levels [40]. Our study reported low Ag concentration in soil, and low percentage distribution of soil in the exchangeable and carbonate-bound fractions, thus the low concentration in the studied plants.
Trace elements concentrations in the investigated plants were very variable. In order to resist the toxic effects of metals, many plants developed a specific tolerance mechanism, such as restriction of metal translocation from roots into shoots [21]. High concentrations of As, Cd, Cu, Zn, and Pb were observed in shoots of plants from two botanical families: Poaceae (Cortaderia Hapalotricha and Cortaderia nitida), and Asteraceae (Ageratina sp, Baccharis latifolia, Baccharis rhomboidalis, Baccharis amdatensis) in Peru (the Carolina and Turmalina mines), Ecuador, and Chile by Bech et al. [47] where, however, metal concentrations were higher than in this study. It is interesting to mention that some studied plants, such as, e.g., Hypericum laricifolium, are used in traditional Peruvian medicine [62].
Some of the plant species were found at different sampling locations and, if the differences in soil metal content are not very high, can be considered as biological (quasi-) replicates. Indeed, these plants presented values reasonably close to each other (The relative standard deviation (RSD) for the large majority lower than 70%). The results are given in the Supplementary Information Table S3.

Bioconcentration (BCF) and Translocation Factor (TF) of the Native Plants
Native plants of this study grow naturally around the mining environmental liabilities polluted with trace elements, demonstrating their good adaptation and tolerance to contam-inated soil. The BCF and TF values reflected their metal accumulation and translocation. The BCF and TF values for the native Peruvian plants analyzed in this study are presented in Figures 5a-f and 6a-f, respectively. The capacity of a plant to accumulate trace metals indicates its potential applicability for phytoextraction or phytostabilization process [38]. Phytoextraction requires the translocation of metals from soil to plant roots, whereas phytostabilization is the capacity to reduce metal translocation from roots to shoots [63]. A plant's potential for phytoremediation can be estimated by the bioconcentration (BCF) and the translocation factor (TF) [37]; plants with these two indicators with values higher than one have potential to be used in phytoremediation. TF higher than one indicates a capacity to transport metal from roots to shoots, probably due to efficient metal transport systems and to the retention of metals in leaf vacuoles; low TF values show that more metals remain in the roots after plant uptake [20].

Bioconcentration (BCF) and Translocation Factor (TF) of the Native Plants
Native plants of this study grow naturally around the mining environmen ties polluted with trace elements, demonstrating their good adaptation and to contaminated soil. The BCF and TF values reflected their metal accumulation a location. The BCF and TF values for the native Peruvian plants analyzed in this presented in Figure 5a-f and Figure 6a-f, respectively. The capacity of a plant to late trace metals indicates its potential applicability for phytoextraction or phyto tion process [38]. Phytoextraction requires the translocation of metals from so roots, whereas phytostabilization is the capacity to reduce metal translocation f to shoots [63]. A plant's potential for phytoremediation can be estimated by th centration (BCF) and the translocation factor (TF) [37]; plants with these two with values higher than one have potential to be used in phytoremediation. than one indicates a capacity to transport metal from roots to shoots, probably ficient metal transport systems and to the retention of metals in leaf vacuoles; lo ues show that more metals remain in the roots after plant uptake [20]. The BCF values were calculated with respect to the total metal concentr (blue bars) and the sum of soils metal fractions with potential bioavailabi changeable and bound to carbonates, iron-manganese oxides, and organic bars). The part contained in the mineral fraction was considered as unavail such, not harmful from the point of view of environmental pollution. For As, Pb for all the studied plants, the BCF values were <1 with the only exception beckia tamnifolia, which was able to accumulate almost twice (1.76) the amoun able Cu. The factors influencing Cu uptake from contaminated soils and di roots have been discussed in detail by Cui et al. [60]. Results of Cu transloca for plants belonging to the same families: Asteraceae, Ericaceae, Hypericaceae, and Poaceae, for which several samples were available. This fact, however, can be due to the decisive role of individual variability of the plant specimens. Indeed, the role of the plant species seems to be important, which is demonstrated by significant variability in bioaccumulation capacities of plants growing in the same locations. For Ag, TF values ( Figure 6b) higher than one were obtained for Brachyotum radula (S1), Arenaria digyna (S1), Calceolaria tetragona (S2), and Achyrocline alata (S2). For the other plants, TF was lower than one. However, in all tested samples the BCF was <1, thus, no species was considered viable for Ag phytoremediation.
The translocation factor values for Pb were >1 (Figure 6f) in Brachyotum radula (S2), Calceolaria tetragona (S2), Chusquea scandens (S1), and Nicotiana thyrsiflora (S1), indicating that these four species can translocate metals from roots to the aerial part. These results are similar with previous studies, in which it was observed that the roots can absorb Pb from soils in a passive mode, where the rate of uptake is reduced by liming and low temperature, so that Pb translocation from roots to shoots is greatly limited [40] and, in consequence, Pb is not generally bioavailable in contaminated soil [63].
The BCF values for Zn were higher than one for the following plants species: Muehlenbeckia tamnifolia (S2), Buddleja interrupta (S2), Nicotiana thyrsiflora (S1), Calceolaria tetragona (S2), Arenaria digyna (S1), Achyrocline alata (S1 and S2), Hypericum laricifolium (S2). Zn TF >1 was also observed for Calceolaria tetragona (S2), Hypericum laricifolium (S1 and S2), The BCF values were calculated with respect to the total metal concentration in soils (blue bars) and the sum of soils metal fractions with potential bioavailability (i.e., exchangeable and bound to carbonates, iron-manganese oxides, and organic matter; red bars). The part contained in the mineral fraction was considered as unavailable and, as such, not harmful from the point of view of environmental pollution. For As, Ag, Cu, and Pb for all the studied plants, the BCF values were <1 with the only exception of Muehlenbeckia tamnifolia, which was able to accumulate almost twice (1.76) the amount of bioavailable Cu. The factors influencing Cu uptake from contaminated soils and distribution in roots have been discussed in detail by Cui et al. [60]. Results of Cu translocation (Figure 5d) showed that in roughly half of the studied species, including Hypericum laricifolium (S1 and S2), Ageratina glechonophylla (S1), Baccharis alnifolia (S1), Ageratina fastigiate (S1), Chusquea scandens (S1), Brachyotum radula (S1 and S2), Bejaria sp. (S1), Arenaria digyna (S1), Calceolaria tetragona (S2), Achyrocline alata (S1 and S2), and Nicotiana thyrsiflora (S1), the TF index values were higher than one. No notable similarities in BCF values were observed for plants belonging to the same families: Asteraceae, Ericaceae, Hypericaceae, and Poaceae, for which several samples were available. This fact, however, can be due to the decisive role of individual variability of the plant specimens. Indeed, the role of the plant species seems to be important, which is demonstrated by significant variability in bioaccumulation capacities of plants growing in the same locations.
The translocation factor values for Pb were >1 (Figure 6f) in Brachyotum radula (S2), Calceolaria tetragona (S2), Chusquea scandens (S1), and Nicotiana thyrsiflora (S1), indicating that these four species can translocate metals from roots to the aerial part. These results are similar with previous studies, in which it was observed that the roots can absorb Pb from soils in a passive mode, where the rate of uptake is reduced by liming and low temperature, so that Pb translocation from roots to shoots is greatly limited [40] and, in consequence, Pb is not generally bioavailable in contaminated soil [63].