Screening of Pioneer Metallophyte Plant Species with Phytoremediation Potential at a Severely Contaminated Hg and As Mining Site

: Phytoremediation of mine soils contaminated by potentially toxic elements (PTEs) requires the use of tolerant plants given the speciﬁc conditions of toxicity in the altered soil ecosystems. In this sense, a survey was conducted in an ancient Hg-mining area named “El Terronal” (Asturias, Spain) which is severely affected by PTE contamination (As, Hg, Pb) to obtain an inventory of the spontaneous natural vegetation. A detailed habitat classiﬁcation was performed and a speciﬁc index of coverage was applied after a one-year quadrat study in various sampling stations; seven species were ﬁnally selected (Agrostis tenuis , Betula celtiberica , Calluna vulgaris , Dactylis glomerata , Plantago lanceolata , Salix atrocinerea and Trifolium repens) . A total of 21 samples (3 per plant) of the soil–plant system were collected and analyzed for the available and total concentrations of contaminants in soil and plants (roots and aerial parts). Most of the studied plant species were classiﬁed as non-accumulating plants, with particular exceptions as Calluna vulgaris for Pb and Dactylis glomerata for As. Overall, the results revealed interest for phytoremediation treatments, especially phytostabilization, as most of the plants studied were classiﬁed as excluder metallophytes. Author Conceptualization, J.L.R.G. and N.M.; Methodology, J.L.R.G., E.A. and N.M.; Validation, J.L.R.G. and T.E.D.; Formal Analysis and Investigation, T.E.D., J.L.R.G. and N.M.; Resources, J.L.R.G., E.A. and N.M.; Data Curation, N.M.; Writing—Original Draft Preparation, N.M.; Writing—Review & T.E.D. J.L.R.G.; N.M.; T.E.D. J.L.R.G.;


Introduction
Contamination derived of mining activities has been increasing dramatically since the beginning of the industrial revolution. Mining operations have produced many environmental problems, including soil contamination and ecosystem degradation [1,2], specially wherever potentially toxic elements (PTEs) were included within the ores exploited [3,4]. Furthermore, uncontrolled tailing disposal implies an important environmental impact especially on soils (losses of biological activity, structure, and fertility) and other environmental compartments affected by wind dispersion, water erosion, leaching, etc. [5,6].
Studies have shown that PTEs are persistent and widely dispersed in the environment; they interact with different natural components and pose threats to human health and the environment [7,8]. PTEs from air emissions can deposit directly on soils, and may be accumulated through rainwater transport [9,10]. Within usual PTEs found in mining areas As, Hg, and Pb are well-known toxics in low concentrations usually mobilized and adsorbed by animals and plants, and are also toxic by ingestion or inhalation for humans. Specifically, As toxicity has caused environmental problems in relation to groundwater and human illnesses [11][12][13]. Pb is potentially toxic to earthworms, but also for predators and detritovores in terrestrial food webs [14]; in humans an excessive intake of Pb may damage neurologic, vascular, endocrine, and immune systems [15]. Hg participates in a number of complex environmental cycles and, once in the environment, can be converted into organomercury compounds which are highly toxic to most organisms [16], causing neurological 1.
To identify and describe species growing in a paradigmatic mining area affected by As, Hg, and Pb contamination.

2.
To determine PTEs contents (in soils, roots, and aerial parts) and behavior of most representative plant species.

3.
To assess the selection of the most suitable combination of plant species to design phytoremediation strategies.

Site Description
The study site is an abandoned Hg mine and metallurgy area called "El Terronal", located in a geographical radius of 10 km around of Mieres (20 km of Oviedo), in Asturias, NW Spain ( Figure 1). With abundant Hg ore deposits, Asturias, from the 1950's until the 1970's, was an important mercury producer on a world scale, with an average annual production of 15,000 flasks (1 flask = 34.5 kg) [28].
The study area was exploited from the Roman occupation of the Iberian Peninsula in the first and second century A.D. [29]. In the late 1960's-early 1970's, production peaked, nevertheless, in 1974, the mine and the metallurgy ceased activity. The legacy remained for decades in the form of abandoned industrial installations and, heaped on a hillside, a large volume of waste contained heavy metals and arsenic disposed along the San Tirso River valley.
Mineralogical studies of El Terronal spoil heaps showed that iron sulphides are very abundant (pyrite, marcasite and pyrrhotite), and, in general, weathering has altered them to hematite and iron hydroxides [30]. Mercury generally appears in the form of cinnabar, irregularly distributed in a brecciated conglomerate and disseminated in the matrix, and arsenic appears as realgar and As-rich pyrite [31]. The study area was exploited from the Roman occupation of the Iberian Peninsula i the first and second century A.D. [29]. In the late 1960's-early 1970's, production peaked nevertheless, in 1974, the mine and the metallurgy ceased activity. The legacy remaine for decades in the form of abandoned industrial installations and, heaped on a hillside, large volume of waste contained heavy metals and arsenic disposed along the San Tirs River valley.
Mineralogical studies of El Terronal spoil heaps showed that iron sulphides are ver abundant (pyrite, marcasite and pyrrhotite), and, in general, weathering has altered them to hematite and iron hydroxides [30]. Mercury generally appears in the form of cinnabar irregularly distributed in a brecciated conglomerate and disseminated in the matrix, an arsenic appears as realgar and As-rich pyrite [31].

Plant Classification
The diversity of species with a significant presence in the study site was identifie and the predominant plant species were characterized. The identification took place in tially "in situ" in zones of the site where the PTEs contents were high as determined i previous studies [4,32], and the conflicting specimens were herborized for identification according to the methodology and traditional methods of plant taxonomy.

Plant Classification
The diversity of species with a significant presence in the study site was identified and the predominant plant species were characterized. The identification took place initially "in situ" in zones of the site where the PTEs contents were high as determined in previous studies [4,32], and the conflicting specimens were herborized for identification, according to the methodology and traditional methods of plant taxonomy.
For the nomenclature and the general characteristics of the plants used, the criteria of Castroviejo (1986-2012) [33] was followed, except in the case of the botanical families of the Grasses (Poaceae) and Betula genus, in which Hubbard (1985) and Ashburner & Mc Allister (2013) criteria were respectively followed [36,37].

Soil and Plant Sampling
As indicated above, previous studies were useful to identify zones of the site with high contamination levels [31,32], whereas another general study of plants potentially useful for phytoremediation [2] was also taken into account.
All things considered, samples of the soil-plant system in the area "El Terronal" were selected according to the density, frequency, and surface covered in the most contaminated areas of the site. The importance and predominance of each plant was categorized from level 1 to 4 ( Table 1) by means of a coating index (CoI) [38][39][40][41]. This plant population frequency study was carried out monthly for one year by the quadrat method (1 m × 1 m) at 7 sampling areas selected after an initial screening of contaminated soils (i.e., areas selected presented the highest contamination levels). In these areas, 39 plant species were identified (see results); finally, 7 were selected and a total of 21 plant samples (7 different species with 3 replicas each) were taken. After plant sampling, individuals were sorted by hand to separate plant structures (aerial parts and root samples). These were then thoroughly washed with tap water several times followed by distilled water, and then cleaned using an ultrasonic bath to remove external contamination, and subsequently dried at room temperature for two weeks. Samples were ground in a universal rotor and variable speed Ultra Centrifugal Mill ZM 200 (Retsch, Haan, Germany) (from 6.000 rpm to 20.000 rpm). The milled samples were collected in stainless steel containers, homogenized, and screened to a size of less than 50 µM.
Representative soil samples (21) were taken in the tilled depth (0-25 cm) in the same sampling stations that the plants were sampled. Sampling was carried out using a soil auger and individual plastic bags were used for storage. All of the soil samples were dried to a constant weight at room temperature over a period of 20 days. Samples were then sieved through 2 mm, ground below 150 µM 400 rpm, (RS100 Retsch vibratory disc mill), homogenized, and quartered by means of an aluminum riffler (cleaned between samples using ethanol, distilled water, and compressed air) to provide a representative subsample of approximately 500 g.

Soil Analyses
A physicochemical characterization of the composite soil samples were carried out according to standard procedures (3 determinations per sample), pH was measured in a suspension of soil and distilled water (1:2.5) with a glass electrode [42], and electrical conductivity (EC) was determined in a 1:5 suspension of soil and wate, using a conductivity meter. Organic matter was measured by weight loss at 450 • C (loss-on-ignition method) [43]. Total N was determined by Kjeldahl digestion [44]. Mehlich 3 reagent [45] was used to colorimetrically determine available P. Exchangeable Al was extracted with 1 M KCl, and exchangeable cations (Ca, Mg, K and Na) with 1 M NH 4 Cl; both were then analyzed by atomic absorption spectrophotometry [46] in a AA200 Perkin Elmer system (Massachusetts, USA). The effective cation exchange capacity (ECEC) was estimated as the sum of exchangeable Al and exchangeable cations. Particle-size distribution was determined by the pipette method, after particle dispersion with sodium hexametaphosphate and sodium carbonate [47].
Representative soil subsamples were leached by means of an 'Aqua regia' digestion (HCl + HNO 3 , 1:1) (250 mg of sample for 8 mL of 1aqua regia) in an Anton Paar 3000 microwave (Graz, Austria) operated for 25 min at 1000 W. The samples were diluted and filtered. Major elements were quantified by an inductively coupled plasma mass spectrometer (ICP-MS 7700, Agilent Technologies, Santa Clara, CA, USA) using isotopic dilution analysis (IDA). Reference materials, ERM-CC141 and ERM-CC018, were used. Detection limits for As, Hg, and Pb were 0.1 µg.L −1 , with RSD > 5% (reproducibility), and the percent of recovery was above 95%.
PTEs soil phytoavailability was estimated by a sequential extraction procedure based on the first two fractions (exchangeable and carbonate-bound) of the Tessier method [48].
Both extracts were passed through 0.45-µM PTFE filters and diluted 1:10 prior to analysis by ICP-MS, as referred above.

Plant Analyses
In order to determine the concentration of PTEs in the different plant organs, 0.2 g of powdered samples was digested with 8 mL of 50% nitric acid using a microwave at 800 W (Multiwave 3000, Anton Paar, Graz, Austria) for 15 min. The solutions were diluted to 50 mL with ultrapure water and passed through 0.45-µM PTFE filters before analysis. The elements of interest were measured using the same ICP-MS device, as described above. Standard reference material apple leaves NIST ® SRM ® 1515 were used with a percent of recovery above 95%.

Data Analysis
The correlations between different variables were evaluated using Pearson's coefficient. All statistical analysis (multiples regression) was performed using IBM (Armonk, NY, USA) SPSS Statistics software 22.0.

Accumulation Factors
To evaluate the Pb, As, and Hg accumulation efficiency in the plants, various factors were examined [49][50][51].
The biological concentration factor (BCF) was calculated as the metal concentration ratio of plant roots to soil (BCF = C root/C soil); values of BCF > 1 indicate the accumulation of a particular trace metal in the roots.
The translocation factor (TF) indicates the ratio of trace metals in the aboveground plant parts (shoot, branches, or leaves) to those in the plant root (TF = C above ground part/C root); TF > 1 indicates that plant translocate metals effectively from root to shoot.
Finally, mobility radio (MR), also known as the biological-accumulation coefficient (BAC), was calculated as the ratio of heavy metal in the aboveground plant parts (shoots, branches, or leaves) to those in the soil (MR = C above ground part/C soil). A mobility ratio of >1 indicates that the plant is enriched with metals (i.e., the accumulator, which can tolerate high tissue concentrations of trace metals), a mobility radio of =1 indicates an indifferent behavior of the plant toward metals (i.e., the indicator, which is characterized by metal uptake proportional to concentrations of trace metals in soil), and a mobility radio of <1 indicates that the plant excludes metals from the uptake (i.e., the excluder, which has a low rate of uptake or actively excludes trace metals).

Description of the Identified Plant Species
According to Díaz and Fernández (2007) [52], phytogeographically, the site is framed as the Eurosiberian Region, the European Atlantic Province, the Cantabrian-Atlantic Subprovince, the Ovetense Litoral District, the Galaico-Asturian Sector, the Ovetense Subsector. Bioclimatically, according to the cartography of Rivas-Martínez et al. (2004) [53], the site is included within the sub-Mediterranean oceanic temperate macroclimate.
The study habitat revealed the presence of many tolerant species in their mature state which spontaneously grew on the contaminated soil. Herbaceous species were predominant (81%, from which 46% corresponds to perennial herbaceous), 5% were arboreal, 12% shrubby, and 2% lichenic and muscinal (for details, see Supplementary Material, Table S1). In this sense, it is well-known that the herbaceous can tolerate areas with notable contamination [54,55]. In addition, herbaceous have a lot of advantages, as they are easy to cultivate and propagate. In general, they are self-sustainable, and many of them are perennial species, which have advantages for extracting or immobilizing the contaminants.
Once identified, the species that develop naturally in the site, and in order to select those species of greatest interest, the quadrat methodology and the subsequent calculation of the CoI revealed 7 plant species with the highest CoI (level 4, as shown in Table 2): Agrostis tenuis; Betula celtiberica; Calluna vulgaris; Dactylis glomerata; Plantago lanceolata; Salix atrocinerea, and Trifolium repens.

Physicochemical Characterization of Soil Samples
The main edaphological soil parameters of "El Terronal" samples that were taken in areas of maximum CoI are shown in Table 3. Overall, results revealed a moderate homogeneity, pH values generally had slightly low alkaline levels, and soils did not show salinity, whereas organic matter levels were high (above 6%). On the basis of the particle size distribution data, samples were classified as sandy soils with normal contents of total N, a very high C/N relation, and P and Mg deficiencies.

PTEs in Soil-Plant System
Total As, Hg, and Pb concentrations in soils and plants (aerial parts and roots) corresponding with the areas with maximum CoI (Figure 2) are shown in Table 4.  Table 4. Average content of metal(loid)s in soil and plant samples for the seven selected species analyzed (n = 3, uncertainties below 10% for all determinations). Data for elements with soil contents below 100 mg·kg −1 (= ppm) are not indicated and were not considered for the calculation of accumulation factors (see Table 5).

Specie Element
Concentration (mg·kg −1 = ppm) Soil Aerial Parts Roots As  197  3  23  Hg  131  5  18  As  107  9 15  Table 4. Average content of metal(loid)s in soil and plant samples for the seven selected species analyzed (n = 3, uncertainties below 10% for all determinations). Data for elements with soil contents below 100 mg·kg −1 (= ppm) are not indicated and were not considered for the calculation of accumulation factors (see Table 5).

Salix atrocinerea
As 112 6 12 Hg 229 2 5 With regards to soil, when compared with the soil screening levels established for PTEs in the Asturias region [56], the average values are notably above the levels in force, especially for As and Hg. Although PTEs contents are therefore very high, it should be noted that average phytoavailability levels found using sequential extraction were low (2.17% for As, 1.5% for Hg, and 0.2% for Pb).
The plant species studied notably exceed As concentrations usual for plants growing on uncontaminated soils which range between 0.009 and 1.5 mg/kg [57]. In fact, all of the species studied (except A. tenuis) revealed concentrations of As in the aerial parts surpassing 5-10 mg·kg −1 , i.e., levels that are usually considered to be toxic [58]. Similar effects are observed with Hg contents, as sub-lethal damage to vital functions occur at 1 mg·kg −1 ; depending upon the species, values that are exceeded in all the species with the exception C. vulgaris that on the contrary revealed high levels of Pb in the leaves (the normal range of metal concentrations in plants for Pb are 0.1-10 mg kg −1 ). Table 5 indicate that most of the plant species studied showed low values (below 1) for the three parameters. This may suggest that uptake of PTEs was very low even considering the high PTE contents in soil, thus plant species under study were classified as non-accumulating plants, with two particular exceptions: C. vulgaris for Pb translocation factor and D. glomerata for As translocation factor. In accordance, only these two plant species could be of interest for phtoextraction treatments whereas the other five could be useful in phytostabilization approaches. The distribution of trace elements in selected species at the mature stages is different for all of them.

Accumulation factors presented in
Previous studies support the different behaviors observed in our study. In fact, C. vulgaris has been identified as an As-tolerant species [59][60][61] and the same occurred with D. glomerata [62]. In turn, Plantago lanceolata is considered as a good As bioindicator, although, in contaminated soils, it recovered much less As than hyperaccumulator plants [63]. Other studies have reported low efficiency As extraction and translocation of T. repens [64] and S. atrocinerea [65]. Nevertheless, and disagreeing with our study, Agrostis tenuis was defined as hyperaccumulator of As [66,67] and T. repens showed strong Hg enrichment ability [68]. At any case, the efficiency and potential use of species in phytoremediation strategies is limited by factors such as the rate of growth, the development of roots, and the production of biomass [69].
The correlation between PTEs and accumulation factors, expressed as Pearson correlation coefficients (p < 0.05), did not reveal significant values. This can be explained as a result of the different mechanisms of assimilation and translocation factors for different elements in these plant species. However, significant correlations were found between contents of, for instance, As and Pb in soil and aerial parts (r = 0.972 and r = 0.905, respectively), and between soil contents and roots (0.940 for As and 0.896 for Hg). These facts show that plants contain information about the quality of the soil.
In a somehow different approach, we also found estimations of accumulation factors by obtaining a multiple linear regression equations as follows: BCF-As = −0.055 + 0.721 BCF-Pb + 1.

Conclusions
The examination of plant species growing in a former abandoned Hg-mining area revealed the presence of a diverse flora tolerant to the main toxic PTEs in the site (mainly As, Hg, and secondarily Pb). Remarkably, herbaceous species were predominant in the most contaminated areas of the site, as demonstrated in a one-year study of plant coverage.
The predominant species identified (Agrostis tenuis; Betula celtiberica; Calluna vulgaris; Dactylis glomerata; Plantago lanceolata; Salix atrocinerea, and Trifolium repens) revealed a methallophyte behavior consistent with a potential forthcoming use for phytostabilization. Therefore, regarding site remediation, future studies should focus on the application of phytostabilization as a first option. However, some of the species identified (Calluna vulgaris and Dactylis glomerata) are also of specific interest because of their ability to translocate Pb and As, respectively, thus pointing out to their potential as bio-indicators or even in phytoextraction.