Distribution of Heavy Metals in the Surrounding Mining Region of Kizhnica in Kosovo

Abstract

This study investigated the distribution of heavy metals in agricultural soils in the vicinity of three large mining landfills of the Kizhnica mine in the Republic of Kosovo. The mining sector is one of the most important sectors of Kosovo’s economic development, and the Kizhnica mine is one of the most important ore producers in Kosovo. Besides the positive aspects, the development of production also has some negative side effects, such as the generation of industrial waste and the possible contamination of surrounding areas, including agricultural land. Therefore, ten sampling sites were selected in the vicinity of the Kizhnica mine. These sites were characterized and assessed as the most important due to the anthropogenic impact of mineral processing and open-tailing waste deposits in Kizhnica. The concentration of Pb, Zn, Cu, As, Cd, Ni, Mn and Sb in the selected samples was determined using inductively coupled plasma–optical emission spectrometry. The data obtained were used to create geochemical maps and calculate the contamination factor, pollution load index and geoaccumulation index. Cluster analysis, Pearson correlation coefficient and air spatial distribution patterns using the air dispersion model were used to evaluate within the area. The results showed that heavy metal levels are influenced by the anthropogenic nature of pollution, confirming a current ecological threat from mining activities in the region. In order to improve waste management, reduce the hazardous impacts of mining and contribute to the sustainable development of the region, the potential reuse of the deposited waste material in the construction industry is proposed.

1. Introduction

Mining is the most important industrial activity in the Republic of Kosovo, which, in addition to its positive economic impact, also has various negative consequences, mainly related to environmental and health issues [1,2,3]. Environmental pollution by heavy metals has been studied for some time. Ferati et al. [4] investigated heavy metal contamination in the water and sediments of the Trepça and Sitnica rivers (northern part of Kosovo). They reported that the high concentrations of As, Cd, Pb and Zn in the water samples originated primarily from anthropogenic sources, such as the discharge of industrial water from mining flotation and mine waste eroded from the riverbanks. In addition, they reported high levels of contamination in the sediments examined, particularly in the sediments of the Trepça River. Borgna et al. [5] investigated the concentration of 12 metals (As, Cd, Co, Cr, Cu, Ni, Pb, Sb, Tl, Th, U and Zn) in the surface soil (0–5 cm) and slightly deeper soil layers (5–50 cm) in the river valleys of Iber and Sitnica (northern part of Kosovo, area 350 km²). They found an increasing trend in the concentrations of the analyzed metals when approaching the ore smelters in Zveçan. The pollution ranged from 15 to 22 km around Zveçan. The concentrations of Pb, Zn, Cu, As, Cd and Sb were particularly high. The main pollution was found in the surface layer. Pollution decreased with increasing depth, and at a depth of 2 m, no pollution was detected in most areas. An exception was the samples taken near the tailings, where low concentrations of pollutants were still detected at a depth of 2 m [6,7,8,9]. Sajn et al. [10] carried out a very similar study but focused on the Mitrovica region (301.5 km²), where the large Trepça mine is active. They analyzed the content of 36 elements in the soil (Ag, Al, As, Au, B, Ba, Bi, Ca, Cd, Co, Ga, Hg, K, Si, Ni, P, Pb, S, Sb, Sc, Se, Sr, Th, Ti, Tl, U, V, W and Zn) and statistically determined the anthropogenic origin for 11 of them (Ag, Pb, Sb, Bi, Zn, Cd, As, Cu, Hg, Au, Tl and Mo). In addition, the detected concentrations of Pb, Cd, Hg, As, Zn and Cu were well above the European average. Prathumratana et al. [11] also reported elevated Pb concentrations (according to the U.S. Environmental Protection Agency standard) in the topsoil in the Mitrovica region and their anthropogenic origin. The anthropogenic origin of the heavy metal contamination of soils in northern Kosovo has also been identified by other scientists [3,12,13,14].

In contrast to the number of studies conducted for the northern part of Kosovo, the studies on the negative impacts of mining conducted for other parts of the country were very rare. Markovic et al. [15] analyzed heavy metals in spring water in Kosovo. In addition to samples from two municipalities in the northern part of Kosovo (Zvecan and Leposlavic), they also analyzed spring water in the municipality of Novo Brdo, which is located in the southeastern part of Kosovo. They reported anthropogenic pollution by As, Pb, Fe and Ni. There are some reports of pollution of water by heavy metals in the river Drenica [16] (central part of Kosovo) and in the rivers Mirusha, Stanishor and Morava in the municipality of Gjilan in the southeastern part of Kosovo [17], but these instances of pollution seem to be due to previous industrial activities (during the time when Kosovo was part of ex-Yugoslavia) and not so much due to the current mining activities.

In this study, we have focused on the Kizhnica mine (municipality of Gracanica) and the pollution of the surrounding agricultural land by heavy metals. The Kizhnica mine is located about 10 km southeast of the city of Prishtina and is rich in lead and zinc ores [18]. Ore production in the mine was increased by using the flotation process, the waste from which was disposed of in open landfills. It is known that the activities associated with lead and zinc production generally lead to significant pollution in the surrounding areas [19,20,21,22]. As heavy metals are not degradable, it is extremely difficult to restore the polluted environment to its former state [23,24].

The sulfide ore, mined at Trepca, is not exposed to an oxidizing environment when it is underground. However, as soon as it has been mined and processed and finally stored in a tailing, it is exposed to and made accessible by water and oxygen. Tailings are defined as composite suspensions of process water and finely ground residues remaining after ore beneficiation, containing secondary sediments and reagents for mining processing or extraction chemicals [25]. They often have an earthy character, physically and chemically similar to raw ore that enters the processing chain, except for changes in particle size resulting from the crushing process [26]. Many mined chalcophile metal deposits (Cu, Zn-, Pb-bearing ores) undergo oxidation processes prior to exploitation; however, sulfide minerals may remain abundant in the generated waste [27]. Mineralogically, sulphide minerals in the waste tailing, such as pyrite (FeS2), are typically the most abundant [28]. Other major phases include chalcopyrite (FeCuS2), covellite (CuS), sphalerite (ZnS), pyrrhotite (FeS), arsenopyrite (FeAsS) and galena (PbS) [28], as well as a range of minerals such as quartz, phyllosilicates, carbonates and some oxides [29]. Deposits of old polymetallic (Zn, Cu, Pb, Fe, Ag) sulphide deposits are investigated worldwide because of their environmental impact. The greater effect of porosity, as a result of coarser grain size, and the unsaturated conditions that occur in the tailing piles have enhanced sulphide weathering, compared with the fine-grained and nearly saturated flotation material present in the settling basins [30,31]. However, numerous physical and stability problems were also observed, mainly due to the particular hydrogeotechnical and mineralogical properties of the tailings [32], to propose new tailing management approaches. To prevent oxidizing, the tailings are deposited underwater in the dam, ensuring that the tailing material is not exposed to oxygen or other weathering processes. However, in Kosovo, the tailings are on the surface and exposed to weathering processes. This tailing from the Kizhnica mineral processing plant is very close to Graçanicë (with 7,700,000 tons of tailing waste) and is a source of dust in dry periods [33]. Due to wind, the dust from these piles might be moved into the surrounding area and into the river. The tailing material at Trepca can thus be spread to the environment by the wind or through erosion. When released into the air, it may travel long distances before settling on the ground [33].

Studies dealing with the negative impact of mining in Kizhnica on the environment are very rare. In fact, we could only find the study by Korça and Demaku [34], which analyzed the occurrence of Pb, Zn, Cu, Cd, Ni and Fe in three types of samples: water, soil and flotation sludge. Unfortunately, no spatial analysis was performed, and the number of samples (four samples per sample type) is insufficient to draw reliable conclusions. This lack of detailed research on the negative impacts of mining in Kizhnica confirms the novelty of our study. The aim of this study is to investigate the distribution of heavy metals in soils in the wider area surrounding the Kizhnica mine, which consists mainly of agricultural land. The main specific objectives are (1) to determine the concentrations of heavy metals in the soil around the flotation in Kizhnica and assess the associated ecological risk, (2) analyze the distribution of heavy metals, and (3) identify the sources of heavy metals using statistics in combination with an air dispersion model and an environmental assessment.

The agricultural soils investigated are located in relative proximity to the three large mining landfills. One of these landfills (10 ha) is active, while two others (70 ha) are inactive and only partially covered with soil. Depositing mining waste in open tailing and in the ground is not viable method due to future environmental costs. Tailings are subject to weathering, acid mine drainage and the mobilization of metals. Weathering and acid mine drainage release the metals in the soil and transport them further into the surface/groundwater and into the food web [7]. In this context, the mobilization of metals in acid mine drainage has become a significant environmental problem for many countries [35,36,37,38,39,40]. Various factors influence the mobilization, deposition, and concentration of metals from mining waste [6,7]. For example, the pH value plays an important role in the bioavailability and mobilization of heavy metals. Under highly acidic conditions, these metals are mobilized in the environment. This means that mining waste offers suitable conditions for the mobility and bioavailability of metals. The high concentration of heavy metals poses a potential problem for human health due to their hazard potential when directly exposed to the contaminated area. Metals of particular concern to human health are lead, arsenic, cadmium, chromium, copper, nickel, selenium and zinc. Maximum levels for toxic metals in soil, water and the food chain have been set by various organizations such as the United States Environmental Protection Agency (US EPA) and the World Health Organization (WHO). The limits for heavy metals in soil have been determined using various standard digestion methods based on the total concentrations of metals [7,41,42]. Various pollution indicators are used to assess metal contamination in soil, such as the pollution load index (PLI), geoaccumulation index (I_geo), and enrichment factor (EF) [42,43].

2. Materials and Methods

2.1. Sampling

Ten sampling sites close to Kizhnica mining complex were selected in Kizhnica, municipality of Gracanica, Kosovo (Figure 1). It was observed on the site that the dust was elevated and scattered over the settlements and agricultural land. The study revealed the need for assessment of the impact of the heavy metals bearing dust from the tailing to the soil contamination. Even though the contamination pathways were determined to also be by acid mine drainage from the tailings and erosion [33] into the surface water, in order to establish the relationship between the source (tailing waste deposit) [44] and soil contamination, the samples were taken in the area upstream of Gracanka river to avoid overlapping testing the pollution from different sources. Samples S1 and S2 were taken on the tailing surface, presenting the waste composition from different ages of deposit; S3 and S4 in the vicinity of the tailing waste deposit; S7, S8, and S9 a long distance from the tailing, from 1000 to 1500 m; and S5, S6, and S10 at elevated locations. The samples were collected using the soil monster (diameter, 4 cm) from the 0–20 cm soil layer at 10 random sample points. The soils taken from the points presented in Figure 1 were mixed into a composite sample and as one replicate sample. In total, 10 composite samples were established. Before sampling, the area was cleared of grass and plants, and the samples were taken directly from the soil. The samples were temporarily stored in cool boxes and transported to the University of Zagreb laboratory via air mail as quickly as possible.

2.2. Sample Preparations

After sampling, the samples were taken to the laboratory, air-dried at room temperature, sieved through a 2 mm sieve, mixed and homogenized using the coning and quartering method, and stored in polyethylene containers until analysis. A two-stage assisted microwave digestion method (HCl/HNO₃/HF/H₃BO₃) was applied for the determination of the total content of analyzed metals, using the MARSX XP1500 Microwave Digestion System, CEM, SAD. Approximately 0.30 g of sample was digested with the reagents (HCl, HNO₃, HF and H₃BO₃) as per the method presented by Kerolli Mustafa [7]:

	Two-stage method for microwave-assisted acid digestion [7]:
Stage	Reagents	Power, W Level (%)	Ramp Time, min	Max. Pressure, psi	Temperature, °C	Hold Time, min
1	3 mL HNO3 + 9 mL HCl + 3 mL HF	1200 (100)	15	800	210	15
2	25 mL 4% H3BO3	1200 (100)	15	800	170	10

After completion of digestion, the samples were cooled to room temperature, diluted with deionized water in 50 mL volumetric flasks, filtered through filter paper, and analyzed using inductively coupled plasma optical emission spectrometry (ICP-OES).

The reagents for microwave digestion were as follows: nitric acid (HNO₃, 67%) NORMATOM for trace element analysis was supplied by BDH PROLABO, England; hydrochloric acid (HCl, 30%) and boric acid (H₃BO₃, 99.9999%) were provided by Merck, Darmstadt, Germany; and hydrofluoric acid (HF, 48%) analytical reagent supplied by Fisher Scientific, Germany. All the plastic containers used for the analysis were cleaned by soaking them in 10% nitric acid for 24 h and then rinsed with distilled water before every use. Single-element standard metal ion solutions were used to control and prepare calibration standard solutions. All calibration standards were prepared by appropriate dilution of standard stock solutions (1 g L^–1) in a concentration range of 1 to 100 mg L^–1.

2.3. Chemical Analysis

Major and minor elements in soil samples were determined after treatment samples using the microwave-assisted digestion method (HCl/HF/HNO₃/H₃BO₃) and dissolution in aqua regia. The contents of Pb, Zn, Cu, As, Cd, Ni, Mn, and Sb in soil samples were determined using inductively coupled plasma-optical emission spectrometry (ICP–OES). A Teledyne Leeman Labs (Hudson, NH, USA) Prodigy High Dispersion ICP system was used. Assisted microwave digestion method 3015A (HCl/HNO₃/HF/H₃BO₃) was applied to determine the total content of major and minor elements using the MARSX XP1500 Microwave Digestion System, CEM, USA. Microwave digestion of 0.30 g soil samples was carried out in two stages using reagents HCl, HNO₃, HF and H₃BO₃. One drop of octanol was added to all prepared samples in order to avoid foaming. Additionally, three blanks were prepared. After the completion of digestion, the samples were cooled at room temperature, diluted with deionized water in 50 mL volumetric flasks, filtered through filter paper, and analyzed using ICP–OES. The carbon, hydrogen, and nitrogen contents were examined using the Croatian waste characterization standard [6,7]. The carbon, hydrogen, and nitrogen contents were measured using a CHN analyzer 1000 LECO following the ASTM D 5373 standard. The electrical conductivity and pH were measured using a conductivity meter (MA 5964 Iskra, Ljubljana, Slovenia) and pH meter (FE20, Mettler Toledo, Greifensee, Switzerland).

2.4. Certified Reference Materials

Certified reference materials (S JR-3 and S Jsy-1) were analyzed to test the accuracy of the applied method for determining total metal concentrations using the microwave digestion method in investigated soil samples. Measured concentrations of elements showed very good agreement with the certified values of reference samples. A statistical comparison of the obtained data was performed using a two-paired t-test with a significance level of p = 0.05. Calculated t-values of 0.811 (for S JR-3) and 1.117 (for S Jsy-1) were lower than the critical value of 2.204, implying that the two sets of values for both certified reference standards were not significantly different.

2.5. Mineralogical Analysis

The mineralogical composition of the soil samples was determined using powder X-ray diffraction (XRD-6000, Shimadzu, Japan). The diffractometer operated at 40 kV with 30 mA current at room temperature to generate Cu-Kα radiation. The samples were scanned from 5 to 80° 2θ angles. A step size value of 0.02° was used with a collection time of 0.6 s per step. The furnace temperature was set to different values ranging from 100 to 1000 °C.

2.6. Statistical Analysis

The results were analyzed using Statistica^® data analysis software, version 14.1.0 (TIBCO Software, Palo Alto, CA, USA), MS Office package (Microsoft Corporation, WA, USA), and AERMOD View (Lakes Software, Waterloo, ON, Canada). Analyses of the experimental data was carried out using the Pearson correlation coefficient (r) and cluster analysis. The relationships between variables were tested using the Pearson coefficient as a non-parametric measure with a statistical significance of p < 0.05, while cluster analysis was used to cluster the heavy metals based on the detected patterns (similarities/dissimilarities) within the analyzed data. For modelling with the AERMOD View-Gaussian air dispersion model, the meteorological data for wind velocity, direction and frequency from the records of the Hydrometeorological Institute of Kosovo were used (Figure 2). The July, August and September 2023 data were used for wind rose plotting using WRPLOT View software, version 7.0 (Lakes Software, ON, Canada).

A specialized software package, SCREEN View, version 3.0 (Lakes Software, Ontario, Canada), was used to calculate the range of pollution and pollutant concentrations at discrete distances from the source. This software was developed to obtain pollutant concentration estimates for the single-source short-term calculations, calculating the maximum ground-level concentrations and distance to maximum. In this study, the terrain was defined as elevated, and the effects of simple area sources were modeled using a virtual point source method based on the wind direction for the samples taken during summer dry months [45]. The calculation included the following initial data: emission rate, source release height, surface of the area of the deposit, and wind direction.

2.7. Determination of Contamination Factor

The contamination factor ($CF$) for each studied heavy metal was calculated using Equation (1) to evaluate the pollution of soil samples.

$$CF=C_{sample}^i/B_n$$

(1)

The calculation is based on the methodology suggested by Min et al. [46] and Kerolli-Mustafa et al. [6]. $C_{sample}^i$ is the concentration of heavy metal i in the sample and B_n is the background or limited value for the same heavy metal in the soil as per Administrative Instruction No. 11/2018 [41]. The classes of soils based on contamination factor ($CF$) values are listed in

Table 1. Classification of soils based on contamination factor (CF) values [6,46,47].

CF Value	Class	Soil Description
CF < 0.10	I	very slight contamination
0.10 ≤ CF < 0.25	II	slight contamination
0.26 ≤ CF < 0.50	III	moderate contamination
0.51 ≤ CF < 0.75	IV	severe contamination
0.76 ≤ CF < 1.00	V	very severe contamination
1.10 ≤ CF < 2.00	VI	slight pollution
2.10 ≤ CF < 4.00	VII	moderate pollution
4.10 ≤ CF < 8.00	VIII	severe pollution
8.10 ≤ CF < 16.0	IX	very severe pollution
CF < 16.0	X	excessive pollution

.

2.8. Determination of Pollution Load Index

The pollution load index (PLI) is used to determine the extent of soil pollution by the analyzed metals, as proposed by Tomlinson et al. [47]. The PLI value was calculated according to Equation (2), where CF is the contamination factor, and n is the number of metals. The classes of soil pollution according to the PLI value are listed in

Table 2. Six classes of soil pollution according to the pollution load index (PLI) [47].

PLI Values	Class	Designation of Soil Quality
≤0	class I	no pollution
0–1	class II	low degree of pollution
1–2	class III	moderate degree of pollution
2–4	class IV	high degree of pollution
4–8	class V	very high degree of pollution
8–16	class VI	extremely high degree of pollution

.

PLI = (CF₁ · CF₂ · CF₃ · … · CF_n)^1/n

(2)

2.9. Determination of Geoaccumulation Index

The contamination of the soil samples with heavy metals was determined using the geoaccumulation index (I_geo), as proposed by Müller [43]. The assessment focused on the extent of metal contamination in seven classes, as shown in

Table 3. The classification of soil contamination using the geoaccumulation index (I_geo) [43].

Igeo Values	Class	Designation of Soil Quality
≤0	I	uncontaminated
0–1	II	uncontaminated to moderately contaminated
1–2	III	moderately contaminated
2–3	IV	moderately to strongly contaminate
3–4	V	strongly contaminated
4–5	VI	strongly to extremely contaminate
>5	VII	extremely contaminated

. This I_geo index is calculated according to Equation (3), where C_n is the concentration of the nth metal in the sample and B_n is the background or limited value of the same metal.

I_geo = log₂(C_n/1.5B_n)

(3)

3. Results and Discussion

3.1. Chemical Analyses

The pH values of soil samples are shown in

Table 4. pH values, electrical conductivity and elemental composition of carbon, hydrogen and nitrogen.

Samples	pH	Conductivity, µS cm–1	C, wt.%	H, wt.%	N, wt.%
S1	6.73	361	1.29	7.39	0.15
S2	3.22	2280	1.25	3.58	0.14
S3	2.35	3160	2.44	3.24	0.21
S4	2.68	2620	1.68	2.04	0.18
S5	6.88	13.5	1.89	3.51	0.13
S6	6.78	530	1.05	1.68	0.11
S7	7.68	72.4	1.98	1.06	0.23
S8	6.49	2060	2.71	1.98	0.18
S9	6.5	125	1.67	1.67	0.22
S10	6.82	25.1	0.94	0.81	0.16

. The initial pH values of samples close to the tailing damp were 2.35 and 2.68, while the pH values for other soil samples in the Kizhnica region ranged from 3.22 to 7.68. The pH values of the samples ranged from 2.35 to 3.22, indicating that the samples close to the mining waste had high acidic buffering capacity. This complies with the conclusion of other studies [3,6,7] that under these conditions, the mobility of heavy metals along the depth column from waste was a process that affected the environment on a larger scale. According to the results, the conductivity ranged from 13.5 for Sample 5 to 3160 μS cm⁻¹ for sample 3. Higher values were recorded between sample 2 (2280 μS cm⁻¹), sample 4 (2620 μS cm⁻¹) and sample 8 (2060 μS cm⁻¹) and other samples of soil, where the value of conductivity decreased in sample 5 (13.5 μS cm⁻¹), sample 10 (25.1 μS cm⁻¹), sample 7 (72.4 μS cm⁻¹), sample 9 (125 μS cm⁻¹), sample 1 (361 μS cm⁻¹) and sample 6 (530 μS cm⁻¹). Determining the elemental composition of soil samples showed that the carbon content was generally between 0.94 and 2.71 wt.%, while the hydrogen content differed. The content of hydrogen was higher in sample 1 at 7.39 wt.%. The results showed that the nitrogen content in soil samples ranged between 0.11 and 0.23 wt.%.

As mentioned in Section 2.3, the assisted microwave digestion method was used to determine the total metal content in soil samples. In

Table 5. Heavy metal concentrations in soil samples and limited values (concentration unit is in mg kg^–1 dry weight).

Samples	Concentration (mg kg–1)
	Pb	Zn	Cu	As	Cd	Ni	Mn	Sb
S1	2304.2	413.1	127.3	4260.1	0.1	37.9	317.0	125.9
S2	1623.4	687.3	86.7	646.8	2.3	216.3	5171.1	27.0
S3	1253.1	405.6	72.9	600.7	0.2	239.2	3397.0	23.7
S4	1714.1	1771.9	128.1	361.3	11.3	240.4	4173.2	18.8
S5	67.2	95.3	82.7	56.8	0.1	82.7	1905.9	2.9
S6	565.3	244.8	126.0	169.2	1.0	288.4	1551.7	4.4
S7	1129.7	1528.3	55.7	372.6	7.5	631.4	2015.5	12.5
S8	1051.6	608.3	54.0	163.9	2.4	590.2	2984.2	11.6
S9	1361.9	2401.1	78.9	310.6	10.5	349.6	2237.4	13.8
S10	354.9	364.5	64.1	127.6	1.5	155.0	1531.8	2.9
Limited values *	200	300	200	30	3	300	-	-

* Administrative Instruction of GRK No. 11/2018 on Limited Values of Emissions of Polluted Materials into Soil [41].

, the results of metal concentration are presented.

The concentrations of eight metals in all ten soil samples were found in the ranges of 67.2–2304.2 mg kg^–1 (Pb), 95.3–2401.1 mg kg^–1 (Zn), 54.0–128.1 mg kg^–1 (Cu), 127.6–4260.1 mg kg^–1 (As), 0.1–10.5 mg kg^–1 (Cd), 37.9–631.4 mg kg^–1 (Ni), 317.0–5171.1 mg kg^–1 (Mn), and 2.9–125.9 mg kg^–1 (Sb). As can be seen from Table 5, the concentrations of metals in sample 1 were higher for Pb and As than the concentrations of metals in other samples. As can be seen by the results, the concentrations of heavy metals in soil samples are higher in the area of tailing deposited waste. Heavy metal concentration in mining and mineral processing tailing waste, as well as other sources, such as weathering, other industrial activities, and traffic, is having a great impact on the environment due to the historical changes in industrial activities in the region [3,5]. Therefore, the open tailing dump of mining waste influenced by winds has a significant effect on the increase in heavy metal concentration on the surface of an area of Gracanica municipality. The contributions of these different sources are various, and their pollution characteristics might change. In fact, heavy metals are always heterogeneously distributed, and this might be the case for the higher concentration of the metals (including lead, arsenic and zinc) in the surface soil of the analyzed region. Overall, according to the results, samples contained a certain pollutant, which is a heavy burden on the environment. The obtained values for Pb, Zn, As, Cd, and Ni exceed the regulatory limit for heavy metals in soil [6,7]. The contamination risk for these metals is still very high due to the low pH and wind effect. The concentration of metal in soil samples was visualized in Figure 3. This graph shows the level curves of a function of major and minor element variables from two of the two-dimensional graphs as a function of concentration and location.

3.2. Mineralogical Characterization

In order to evaluate the form of the heavy metals in the tailing and surrounding soil, X-ray diffraction was performed. The mineralogical composition of the samples was important for assessment due to the different behavior of minerals during weathering, as well as their mobility along the depth column. The samples are grouped as per their distance from the source (tailing). Samples 1 and 2 are from the tailing surface. Two samples were taken from the site as it was observed that there were different appearances of tailing, due to the sequence of deposition and advanced aging process. Samples 3 and 4 were not taken into consideration as a large excavation project was started covering their location. samples S7, S8, and S9 were at asimilar distance from the deposit in the wind direction blowing along the long side of the deposit. Finally, samples S5, S6, and S10 were located on the slopes, thus being protected from the wind. The X-ray diffraction (XRD) patterns for soil samples ((a) composite 1: S1, S2, (b) composite 2: S7, S8, S9, (c) composite 3: S5, S6, S10) were taken in order to confirm the phase composition of samples. A combined study using different characterization techniques was applied in order to better understand the structure and composition of the investigated samples. XRD patterns of original samples indicated that the major components of soil waste were quartz and other dominant fazes, including minerals of Al, Si, and K, as shown in Figure 4. There were strong peaks assigned to the C, H and N groups in all samples. In other samples, the presence of Fe, As and Mo, P, Li, Pd, Na, Al, NO groups, Cu, and Zn could be confirmed, but the presence of organic compounds was also shown in different places.

Silica, betaine phosphate, azobenzenes, ammonium triphosphate, and ash are mostly detected in the XRD graphs of the samples located at the tailing waste deposit (Figure 4a), with Chalcedonite and Jarosite included in the XRD graph of the samples located on the northeast and southeast of the deposit. On the opposite side, diffractometry for the samples taken from the elevated locations (Figure 4c) showed quartz, mica (muscovite), Cavansite, and phyllosilicate and organic compounds. It was interesting to observe that no galena was recorded on the diffractogram, nor pyrite or ankerite, as was the case with the other Trepca tailings [48]. This can be interpreted as indicating that the oxidation process is advanced and all sulfide minerals are oxidated. The analyses also show that the XRD method has constraints in detecting lower concentrations of elements; however, it confirms the source of the heavy metals in soil.

3.3. Statistical Analyses

The Pareto charts are used to verify the effects of each operational variable presented in Figure 5. In the figure, the bars extending to the vertical line correspond to the effects of a 95% confidence level. As can be seen in the figure, in all the evaluated responses with element concentrations in soil samples, the concentration effects and values of the effects are displayed, and a reference line is drawn on the chart. Any effect that extends over this reference line is potentially important. Analysis of the individual factors on the Pareto chart showed that the concentration was statistically significant since it overshot the critical value line (reference line). The interactions for the majority of elements do not differ much from the normal distribution.

In this study, explanatory data such as correlation matrix and cluster analysis were employed in order to obtain relevant data regarding the main variables (concentration of metals) of ten soil samples collected in Kizhnica. Thirteen variables (were compared to find similarities between metal distribution and pH values, electrical conductivity and elemental composition of carbon, hydrogen and nitrogen in soil samples (

Table 6. Pearson correlation matrix of heavy metals and pH values, electrical conductivity and elemental composition of carbon, hydrogen and nitrogen in soil samples.

	Pb	Zn	Cu	As	Cd	Ni	Mn	Sb	pH	Conductivity, µS cm–1	C, wt.%	H, wt.%	N, wt.%
Pb	1.000000
Zn	0.382236	1.000000
Cu	0.396471	−0.041432	1.000000
As	0.694510	−0.162572	0.471900	1.000000
Cd	0.293579	0.955152	0.037837	−0.271293	1.000000
Ni	−0.042726	0.421713	−0.553777	−0.429200	0.419457	1.000000
Mn	0.192090	0.232873	−0.116577	−0.440606	0.301268	0.178073	1.000000
Sb	0.748591	−0.127643	0.471808	0.995539	−0.236774	−0.413011	−0.362021	1.000000
pH	−0.392258	−0.074007	−0.221167	0.078777	−0.125394	0.225657	−0.795204	0.006272	1.000000
Conductivity, µS cm–1	0.377315	−0.015937	0.056684	−0.112253	0.041967	0.077827	0.786382	−0.033582	−0.896577	1.000000
C,wt.%	0.012788	0.104689	−0.509252	−0.221212	0.076021	0.546177	0.258306	−0.176810	−0.164339	0.451614	1.000000
H,wt.%	0.569875	−0.349644	0.470351	0.883341	−0.450701	−0.589655	−0.223098	0.894385	−0.117350	0.065569	−0.080523	1.000000
N,wt.%	0.276542	0.689743	−0.521162	−0.127564	0.591499	0.579868	0.133963	−0.098062	−0.090824	0.142459	0.549599	−0.310118	1.000000

). This fact is supported by the use of Pearson’s correlation coefficient analysis, which can be used to measure the degree of correlation between the concentration of the major and minor elements in the investigated samples. Statistical analysis of the soil samples revealed several links between elements of interest. The similarity shown between elements refers to the same source of origin and the same mobility entrance in the environment based on pH and conductivity effects. The correlation matrix (Figure 6) between the elements of interest indicates a strong positive correlation of Zn with Pb, As and Pb. Cd showed a strong positive correlation with Zn, while Ni showed a strong correlation with Cd. Sb showed a strong correlation with Pb and As.

In general, the Pearson correlation coefficient used to analyze the correlation between elements in the investigated samples showed a very positive correlation between major and minor elements in soil samples coming from a common source of pollution, which were very much influenced by pH and conductivity. A matrix plot for the investigated elements is presented in Figure 6.

Severe air pollution occurs in the Kizhnica region, covering the whole area of Gracanica. The industrial waste deposits from mining and mineral processing production are being blown with the wind and scattered in an area covering more than 10 km. This estimation was calculated using SCREEN View, Screening Air Dispersion Model for the Wind class IV, with a wind speed of 8.8–11 m s⁻¹ in the east–northeast direction, calculated to be a resultant vector for the given location for the summer season (July, August and September), according to the data collected from the Hydrometeorological Institute of Kosovo, and presented in Figure 7.

The pollutants settle in a large area, polluting the soil and spreading throughout the region. There is a significant deposit of mining waste containing heavy metals in Kizhnica (Figure 8).

The study revealed that the average wind for the area is class IV with wind blowing from 5.7 to 8.8 m/s in the east and northeast direction. In addition, samples 1 and 2 were on the tailing surface, and the wind speed did not determine the dusting. Samples 3 and 4 were highly contaminated as they were very close to the waste, having slightly higher wind speeds of 2.10–3.6 and showing similar levels of contamination to the closest ones. On the other hand, samples S7, S8, and S9 were longer distances away but exposed to the stronger class III wind, 3.6 to 5.7 m/s. Samples 5, 6, and 9 were behind the hills and, on the higher ground, not exposed to the scattering of the heavy metals bearing dust particles.

Cluster analyses were carried out to identify the relatively homogeneous groups of elements in the study samples. The metal concentration was used to analyze the grouping of elements on soil waste. The following dendrograms of the investigated samples were obtained with the help of Ward’s method in order to define the similarities and dissimilarities of elements. Figure 9 shows the dendrogram analysis mean of elements in soil waste samples. When dendrograms of soil samples were examined, significant variations in pollution levels were observed. The dendrogram presents three major distinct clusters with four groups formed. Pb and Zn have good similarity and are clustered in one group, whereas Cu, Cd and Sb are clustered into another group. These two groups showed a close relationship with Ni and As in one of the major clusters. The third group showed close similarities with the last group of Mn. A similar relationship can be seen in elements of the other three groups as well. The major clusters indicated a strong anthropogenic influence. The results obtained by cluster analysis are consistent with those obtained by correlation analysis of elements in soil samples in Kizhnica.

3.4. Environmental Implication

The environmental implications of heavy metals are presented in

Table 7. The values of the contamination factor (CF), pollution load index (PLI) and geoaccumulation index (I_geo) of soil samples in Kizhnica.

CF
Metals	S1	S2	S3	S4	S5	S6	S7	S8	S9	S10
Pb	11.52	8.12	6.27	8.57	0.34	2.83	5.65	5.26	6.81	1.77
Zn	2.29	2.29	1.35	5.91	0.32	0.82	5.09	2.03	8	1.21
Cu	0.64	0.43	0.36	0.64	0.41	0.63	0.28	0.27	0.39	0.32
As	142.00	21.56	20.02	12.04	1.89	5.64	12.42	5.47	10.35	4.25
Cd	0.00	0.76	0.08	3.76	0	0.33	2.5	0.81	3.35	0.52
Ni	0.13	0.72	0.8	0.8	0.28	0.96	2.1	1.97	1.17	0.52
PLI
	0	2.13	1.25	3.24	0	1.2	2.84	1.7	3.08	1
metals					Igeo
Pb	3	3	2	3	0	1	2	2	3	1
Zn	0	1	0	2	0	0	2	1	3	0
Cu	0	0	0	0	0	0	0	0	0	0
As	6	2	4	2	1	2	3	2	3	2
Cd	0	0	0	2	0	0	1	0	2	0
Ni	0	0	0	0	0	0	1	1	0	0

. The contamination factor (CF) was calculated as the total metal concentration with reference to the national limited standards for safe limits of heavy metals in soil [41]. The analyzed elements are selected based on the toxicity index.

The results in Table 7 show that Pb ranges from a very low degree of contamination to a very severe degree of pollution in sample 1. The content of Zn ranges from a low degree of contamination to a severe level of pollution in sample 9. Cd showed a moderate degree of contamination in samples 4, 7, and 9, while in other samples, it can be considered a low-risk element (Table 1).

The CF data indicate significant anthropogenic impacts at all sites. Using the classification system proposed by other authors [44,49,50,51], the overall range of CF values indicates a high degree of contamination of soil samples to excessive pollution with As in sample 1. The scope of PLI in the analyzed samples confirmed the degree of contaminants present, while in the majority of the samples, results show that the soil quality is deteriorating (Table 2).

Results of the geoaccumulation index are presented in Table 7 based on Muller’s classification [43] (Table 3). Cd and As belong to class VI (extremely contaminated) and Pb belongs to class III (moderately to strongly contaminated in sites S1, S2, S4 and S9). Therefore, the minerals and the sites S3, S7, and S8 are considered to be moderately contaminated (class II), while site S5, S6 and S10 are classified between uncontaminated to moderately contaminated (class I). Zn belongs to class III in S4, S7 and S9 (moderately to strongly contaminated) and to class I, which is uncontaminated to moderately contaminated, in all other sites. Cd belongs to class II (moderately contaminated) in sites S4 and S9, while in other sites, the result indicated that they are uncontaminated. Cu and Ni belong to class 0 (uncontaminated). While As results indicated extreme contamination in sample 1, all other analyzed samples show moderate to strong contamination. Heavy metals are critical contaminants due to their accumulation in the soil and easy uptake by plants, thus entering the food chain. Table 7 shows the range and average values of PLI values for each metal in the investigated sites. Based on the average values of PLI (Table 2), the ranking of the intensity of heavy metal pollution of the soil samples in Kizhnica is from uncontaminated to a high degree of pollution.

3.5. Impact on Sustainability

During the second half of the twentieth century, the whole region was focused on mining and metallurgy as the primary source of income and its development strategy. The environmental impacts of this industry were neglected, and environmental protection has only recently become important to society. With the mandatory recycling of lead acid batteries, replacing lead pipelines in water supply systems, new lead-free materials and other green transition technologies, mining of lead and zinc ores became less attractive and profitable. The technology used in Kizhnica mines is out of date and cannot fulfil the requirements of the environmental regulations imposed by the laws.

In order to achieve sustainability, mining, and minerals processing for heavy metals production should fulfill economic, environmental, and social equity requirements. If the production meets environmental requirements, the investment rate in passive tailing landfill remediation, green technology instalments, dust emission prevention, and acid mine drainage prevention will be very high, and the profit will be heavily impacted. On the other hand, the communities living around the mining and mineral processing facilities are employed by the mine, but at the same time, they are facing health issues. According to the Kosovo Agency of Statistics, the significant cause of death in Kosovo is diseases of the respiratory system [50], and according to Eurostat’s regional yearbook from 2023 [51], the incidence of respiratory diseases in Kosovo was much higher than in the Western Balkans or Turkey.

In this situation, the population has turned to the original source of income–agriculture production, having favorable climatic conditions, and natural resources. They faced obstacles in land use, as presented in this study, as the areas surrounding the mineral processing plant and tailing waste deposits are heavily loaded with heavy metals. In order to develop sustainable agriculture, massive land remediation should be performed, as well as detailed mapping of the soil pollution, based on the presented statistical model.

4. Conclusions

The environmental pollution in Kizhnica, municipality of Gracanica, Republic of Kosovo, is primarily due to mining activities. The measurements of heavy metal concentrations in these regions show high levels of metal concentration and bioaccumulation. The application of statistical analysis as a correlation matrix and cluster supports the fact that heavy metals are more concentrated close to the mining smelting area, but it has an impact on the surrounding area as well. The air dispersion modeling is in line with other chemical analysis and environmental implication indexes. The establishment of regular monitoring programs is an urgent need for Kosovo. The most important and cheap procedure is the application of phytoremediation as a method of great interest. The results also indicate that there is an increasing need for further research, mainly focused on the mechanisms of the remediation process and the potential use of the deposed waste material in different combinations in construction. This process will support the process of waste management and reduce the continued risk to the environment in the study area. In order to manage complex environmental problems but also the possible resource of valuable components, further work that deals with tailings and its impact on the environment will be published in a future paper. The work will include depth analysis of the metal formation and behavior in the oxidating process at surface tailings, and hydrogeochemical analysis of the tailing but also surrounding soil will be performed to understand and simulate heavy metal penetration and possible erosion of the tailings to the surface and ground waters.