Distribution of Heavy Metals (Cd, Cu, Co, Cr, Hg, Mn, Ni, Pb, Zn, and As) in Overbank Sediments of Ibar Tributaries (in Serbia)


 The Ibar River is an international river whose watercourse passes next to numerous technogenic landfills and chemical-industrial complexes. They can affect the quality of water in the river, as well as the quality of overbank sediments. The river flows into the West Morava River near Kraljevo and it represents its largest tributary. The analysis of the overbank sediments of Ibar tributaries showed high concentration of heavy metals at some points. The statistical analysis of the content determined the existence of two synthetic factors around which heavy metals were concentrated. The comparison of the chemical composition of overbank sediment with the composition of geological and pedological substratum on which they were developed showed that the increased concentrations of heavy metals were mostly a product of geological and pedological composition of the river’s substratum and basin, and had no connection with (possible) anthropogenic factors. However, due to the sensitivity and proximity of potential pollutants, the necessity to constantly monitor chemical composition of overbank sediments in the river basin was determined.


Introduction
The Ibar River is an international river. Its river source is located on the Hajla Mountain in Montenegro at an altitude of 2,400 m. It ows through Montenegro and Serbia. In central Serbia, near Kraljevo, it ows into the West Morava River as its largest tributary. The upper, middle, and lower course can be differentiated. The upper course is located mostly in Montenegro and only partly in Serbia (Kosovo and Metohija). The middle and lower course of the river were investigated in this paper. The river often ows through narrow and not easily accessible gorges. A 110 m high dam was built 24 km upstream from Kosovska Mitrovica on the Ibar, and because of this, the arti cial lake Gazivode was created. The total length of Ibar is 276 km, and the river's basin surface area is 8,059 km 2 (Fig. 1). Larger tributaries in Serbia are Sitnica River, Rudnjačka River, Raška River, Brvenica River, Jošanica River, Studenica River, and Ribnica River.
The Ibar River ows through areas with speci c geomorphological structure. The in ow of wastewater, which originates from mining-industrial complexes, tailings, intensive agriculture, illegal land ll sites, etc., into the river basin is high. Due to orography, the river often over ows. Numerous studies indicate that more than 12% of watercourses that belong to the Ibar River basin are susceptible to ooding (Dragićević et al. 2019). When the water that ooded the area retreats, it leaves pollutants behind. Numerous studies have con rmed this. There are few studies that investigate the direct origin and chemical composition of this material.
The middle course of Ibar has a high in ow of industrial wastewater that originates from the wider area of Kosovska Mitrovica, Zvečan, and Leposavić. Waters from the wider area of Kosovska Mitrovica come from several mining-metallurgical-chemical plants and land lls of Trepča Mines. The watercourse passes by the Pb-Zn ore otation tailings site in Leposavić and further downstream next to the otation and otation tailings of Suva Ruda (Rudnica and Kukanjica streams). The contouring showed that the Rudnica tailings covered the area of 18 hectares 79 acres 8 m 2 . The granulometric composition of the tailings content showed that it consisted of siltstone / sand. The content of chalcopyrite was three times higher than content of andradite in one part of the tailings site, while other minerals were less present. In other part of the site quartz and gypsum were equally present, and in the third part of the site quartz was predominant. The distribution of heavy metals in the tailings site did not exhibit any patterns. One part of the site had elevated U content compared to the natural background radiation. Kukanjica Stream tailings covers the area of 10 hectares 48 acres 9 m 2 . The granulometric composition showed the presence of clay/ siltstone. The X-ray analysis determined that quartz was predominant compared to clay, siderite, pyrite, gypsum, and ankerite, and feldspars and calcite occurred locally. Hydro-chemical analysis showed increased contents of Pb, Zn, and Fe compared to standards for class II water, while the concentrations of As and Sb were above the allowed for class IV waters. The soil samples from the area surrounding the tailings site contained the concentrations of Zn and As in the phytotoxic range, and Co, Cr, Sb, Sn, Cu, and Cd had increased content locally (Đokić 2011, Đokić et al. 2013).
The tailings sites are located in Tadenje (Progorelica), Jarandol, Pobrdje, Piskanja, Bela Stena, and Korlaća. Wastewaters from these tailings ow directly into the Ibar. Jarandol and Progorelica tailings were created as a result of the exploitation of coal. The exploitation began at the end of the 19th century. The areas of initial exploitation and tailings were spontaneously recultivated with pioneer plants. The mean U and Th content at this site is higher compared to their mean content in the earth's crust. The excavated coal at the Piskanje location is re ned by separation process. The tailings site for boron minerals is located at Pobrdje. This site was formed after manual sorting of the excavated ore. Hydrochemical analysis showed increased content of Fe and Pb, and increased contents of Zn, Cu, Cd, Cr, Co, Ni, Sb, As, and Hg were high as well. Bela Stena tailings site was formed after the exploitation of magnesite during the last twenty years of the 20th century. The tailings site at Korlaća was formed after the exploitation and processing of asbestos (Đokić 2011).
The lower course of Ibar receives the in ow of industrial and communal waters from Raška and Baljevac (Nikolić et al. 2014). Ibar is polluted by industrial wastewaters that originate from dairy and meat industries, municipal wastewater, etc. The river passes by 47 settlements in its middle and lower course which endanger the quality of the water and its sediments (Nikolić et al. 2014). All in all, the river is polluted by industrial, agricultural, and municipal wastewaters.
The construction of several hydroelectric power plants is planned on the Ibar. Numerous analyses have been performed on how these hydroelectric power plants could affect the quality of water. The conclusion was that the planned accumulations would prevent the formation of the stable thermal strati cation during the warmer part of the year, which would further impede the full realization of eutrophication potential (Jaćimović et al. 2014).
The quality of the water in river is systematically and continuously controlled by measuring the parameters that in uence it the most, which are the concentrations of Pb, Cu, Ni, Co, Cd, Mn, Zn, Cr, and Hg. These measurements are performed at several locations along the Ibar watercourse (SEPA 2019).
The suspended material transported by the river during the ood, which is frequent, is deposited as the overbank sediment in the soil on the banks of the Ibar. This sediment, most of all, affects the chemical properties of the soil. However, the geological stratum on which the soil is developed has the greatest in uence on the structure and mineral and chemical composition of the soil (Dorfer et al. 2018a, Dorfer et al. 2018b). The climate also has considerable in uence on these properties (Gore 2009).
The overbank sediment is a ood sediment formed under the conditions of higher energy of relief from the material transported in the suspension. The overbank sediment is a complex, dynamic system susceptible to changes.
The tributaries of Ibar often have the characteristics of a torrent, which can cause the mixing of materials These tributaries also ow through numerous settlements and receive their wastewaters. This way they directly or indirectly affect the quality of the water in Ibar and the quality of the sediment as well. The sources of the pollution can be diffuse, point, and combined, and in the case of this river, they are combined. The regularity of the increased presence of heavy metals (Cd, Cu, Co, Cr, Hg, Mn, Ni, Pb, Zn, and As) in the overbank sediments of the tributaries of Ibar was analyzed in this paper.

Sampling
The collection of the samples was performed according to the recommendations for the drawing of the Geochemical Map of Serbia 1:500,000, stream, overbank, and oodplain sediments (Đokić 2016). The coordinates for the sampling were determined using a GARMIN-GPSmap60CSx. The depth of sample collection was from 0 to 10 cm. To avoid the in uence of organic matter, samples were collected under the layer of humus. The collection was performed by an Eijkelkamp probe with a nozzle for soil testing and a spade with a stainless steel blade. Each sample consisted of 10 subsamples which were homogenized in the eld. A total of 75 samples (750 subsamples) were collected. The samples collected for the realization of the project Geochemical Map of Serbia 1:500,000, stream, overbank, and oodplain sediments were used in this study (Đokić 2017, Đokić 2018).

Analytical procedures, methods, and concentrations
The samples were dried at room temperature under atmospheric conditions for 60 days, and then they were quartered. The reduced sample was dried in a porcelain dish, in a laboratory dryer at 40° C. The sample weighing 100 g was ground and then sieved to a grain size of 100µm. The sample was then chemically analyzed.
The chemical analyses of the samples were performed in an accredited laboratory for chemical testing at the Institute of Mining and Metallurgy Bor. Cobalt (Co), cadmium (Cd), copper (Cu), lead (Pb), nickel (Ni), mercury (Hg), zinc (Zn), and arsenic (As) were examined by Inductively Coupled Plasma Mass Spectroscopy (ICP-MS), and chromium (Cr) and manganese (Mn) by X-Ray Fluorescence (XRF). The lower limit for determination of Zn and Mn was 0.001%; As <1 mg/l; Pb and Cu <0.5 mg/l; Ni, Hg, Co, and Exploratory factor analysis on Ln transformation of concentrations. Varimax rotation was performed for a more even distribution between the components. Based on the exploratory analysis by the extraction method, the factors around which the elements were grouped were singled out; Filling factor estimation matrix to determine the links between elements and factors; and Con rmatory factor analysis to con rm the percentage of variance grouped into wholes and determine (degree) the signi cance of variables for the factors.

Results
The concentrations of the tested heavy metals were grouped into concentration intervals. The length of the group interval was determined using the Sturge's rule: Based on the number of intervals, the width of the interval was calculated: Increased concentrations of phenol (organic pollutant) and other organochlorine compounds have been often recorded in Sitnica. These compounds were found in Ibar near the town of Kraljevo, and in the wells of the Kraljevo waterworks as well. The overbank sediments from the Sitnica River basin were not investigated in this paper.
RAŠKA RIVER BASIN. The source of the river is located on the slopes of Pester plateau. There is a tailings site in the basin of the river formed after the exploitation of andesite in Velika Bisina. A tailings site is also formed in Trnava due to the surface and underground exploitation of ultrabasite (speci cally dunite) (Đokić 2011). The city of Novi Pazar is supplied with drinking water from the spring of Raška (Beta 2019). The length of the river from its source to the mouth of the river near the town of Raska is 60 km, and the total area the basin covers 1,193 km 2 . The spring of the river was used for the needs of the Ras hydroelectric power plant in 1953, and a 4.5-km-long supply tunnel was built. The quality of the water is signi cantly lower downstream of the collectors in Novi Pazar. It is the most polluted in winter and autumn (Marinović and Rašljanin 2010). In certain parts the quality of the water corresponds to class V -sewage water (Korhner 2013). Smaller watercourses that form this basin are: Ljudska River I, II, III, IV, and V, Raška I, II, III, and IV, Dramićki Stream, Kuzmićka River, Đurićka River, Šaronjska River, Deževska River I and II, Nosoljinska River, Tušimska River, Kosurička River, Rečice, Trnavska River, Sebimiljska River, Jošanica I and II, Brezovačka River, Sebečevska River, Trnavska River I, II, and III, Izbička River, and Jovska River (Fig. 1). .. The construction of small hydro power plants of derivative type is planned at several locations (Tomić 2016). Smaller watercourses that form this basin are Rudnjačka and Lisinska rivers (Fig. 1).
RIBNICA RIVER BASIN. The length of this river is 26 km and it covers the surface of 115 km 2 . Its upper and middle courses are located in the mountain area and they have the characteristics of a torrent in some parts. The lower course is calm with the characteristic of a lake (Fig. 2). The quality of the river water is constantly controlled (SEPA 2019).
Geological and pedological structure of the investigated area Geological heterogeneity of the Ibar tributaries environment can have a great effect on the chemical composition of overbank sediments (Fig.2). The physical size of the basins does not have to affect the (increased) presence of the analyzed heavy metals (Cd, Cu, Co, Cr, Hg, Mn, Ni, Pb, Zn, and As) in overbank sediments. Table 1 shows the geological composition of the smaller basins of the tributaries.
The immediate sampling site may affect the chemical composition of overbank sediments. Table 2 shows the local geological structure of the immediate sampling sites for sediments in the investigated watercourses. Table 3 shows the occurrence of mineral raw materials that were found during geological mapping and drawing of metallogenic maps (Vanđel et al. 1981).
The geological stratum has signi cant in uence on the formation of the soil. Due to the heterogeneous geological structure, overbank sediments were collected from different soil types (Table 4 (Goldschmidt 1954, Mason 1966, Jović and Jovanović 2004. The range of chromium concentrations is 7-1,500 mg/kg, and the mean content is 50 mg/kg. It is mostly found in ultrabasic rocks and serpentinite (1,600-3,400 mg/kg). Phytotoxic concentrations of chromium in soil on the surface amount to 75-100 mg/kg (Goldschmidt 1954, Mason 1966, Jović and Jovanović 2004. Mercury in unpolluted soil can be found in the concentration range from 0.01-0.5 ug/kg, and in the polluted soil it ranges from 0.5-50 ug/kg. The mean content value in clastic rocks is 300 ug/kg, in carbonate rocks 200 ug/kg, and in clay sediments 400 mg/kg (Mason 1966). The sources of the mercury in the environment are industrial plants that use it in the technological process (Jović and Jovanović 2004). The content of manganese found in the Earth's crust amounts to 950 mg/kg (Mason 1966). It can occur with oxides and hydroxides of iron and clay minerals with which it often forms a bond with microelements by adsorption (Jović and Jovanović 2004). In normal unpolluted soil, nickel concentrations measure 50 mg/kg, in soil developed on serpentinite 0.1-2%, and in mining areas 100-500 mg/kg (Cox 1995). The average lead content in the earth's crust is 14 mg/kg, and in unpolluted soil 2-200 ug/kg. Phytotoxic concentrations of lead in soils on surface are 100-400 mg/kg (Goldschmidt 1954, Mason 1966. The average zinc content in the earth's crust is 75 mg/kg, while in the soil it is 30-1,000 mg/kg. Phytotoxic concentrations of zinc in soils on surface are 70-400 mg/kg (Goldschmidt 1954, Mason 1966, Jović and Jovanović 2004. The average content of arsenic in the earth's crust is 1.5 mg/kg, and the concentrations in unpolluted soil range from 1-10 mg/kg. Phytotoxic total concentrations of arsenic in soils on surface amount to 15-50 mg/kg (Goldschmidt 1954, Mason 1966, Jović and Jovanović 2004.

Results And Discussion
Heavy metals in overbank sediments CADMIUM. This heavy metal occurs in contents higher than the maximum permissible value allows (6.4 mg / kg, (O cial Gazette of the RS 2012) in sediments found in the Rudnjačka river basin (Lisinska River -7.5 mg / kg). The contents higher than target value allows (0.8 mg/kg, (O cial Gazette of the RS 2012)) can be found in sediments from Gajovska and Kaznovićka rivers, in Jošanica river basin (maximum value obtained for Samokovka river), Maglašnica, Planska River II, and in Rudnjačka river basin (Rudnica). Somewhat lower contents, but still higher than the target value, were found in Raška river basin (Dramićki Stream), and in the sediments from Ibar I and II and Pavlićka River. The contents below the detection limit of the used method were found in sediments from the Raška river basin (Ljudska River I, II, and II), Raška (I and II), Đurićka, Šaronjska and Deževska River I and II, and Nosoljinska River. The content of Cd in half of the sediment samples was between 0.29 mg/kg and 0.53 mg/kg (Fig.4).

MANGANESE. This heavy metal was not included in the regulation on limit values for pollutants (O cial
Gazette of the RS 2012). The literature suggested that the content from 1,500 to 3,000 mg/kg can be toxic (Kabata-Pendias and Pendias 1984). It can be mobilized into the soil from the sediment. These high contents were found in the Raška river basin (Dramićki Stream, Tušminska River, and Jovski Stream), Brvenica river basin (Vrani Stream and Borovića River), Studenica river basin (Brevina), Jošanica river basin (Jošanica II and Lužnjaski Stream), Rudnjačka river basin (Lisinska River) and in Kaznovića River, Krivačka River, Planska River I and II, and Maglašnica. The content of this metal in half of the sediment samples was in the range from 929.90 mg / kg to 1,312.40 mg / kg (Fig. 4).
NICKEL. The samples from Raška river Basin (Raška IV) and Vrški Stream had contents of nickel lower than the target value (35 mg/kg, (O cial Gazette of the RS 2012)). Samples obtained from Raška river basin (Ljudska River I, II, and III, Raška I, II, and IV, Kuzmićka River, Sebečevska River, and Trnavska River) and Rudnjačka river basin (Lisinska River) had nickel concentrations in the range between the target values and maximum permissible values (44 mg/kg, (O cial Gazette of the RS 2012)). Extreme values for nickel were recorded in sediments from the Brvenica river basin (Klisurski Stream), Pivnica, Raška river basin (Rečica), and Gokčanica river basin (Rudnjačka River). The concentrations of the rest of the samples were higher than the maximum permissible value. The content of nickel in half of the sediment samples was signi cantly higher than the maximum permissible value and it was in the range from 63.3 mg/kg to 682.6 mg/kg (Fig. 4).
LEAD. The contents higher than maximum permissible value (310 mg / kg, (O cial Gazette of the RS 2012)) was found in sediments from Jošanica river basin (Samokovka), Gokčanica (Zastupski Stream and Gokčanica), and in Gajovska and Kaznovićka Rivers. The contents between the target value and maximum permissible value was discovered in sediments from Raška river basin (Kuzmićka River and Deževska River I), Jošanica river basin (Jošanica III and Veleštica), Rudnjačka river basin (Rudnica), and in Krivačka River, Planska River II, Maglašnica, Vrški Stream, Pavlićka River, and Ibar I and II. The rest of the sediment samples had the lead content lower than the target value. The content of lead in half of the samples was lower than the target value and it was in the range from 36.80 mg/kg to 79.00 mg/kg ( g.4).
ZINC. The contents of zinc higher than maximum permissible value (430 mg/kg, (O cial Gazette of the RS 2012)) were found in the sediments from Kaznovićka River, Jošanica river basin (Samokovka), Gajovska River, and Maglašnica. The concentration between the target value (140 mg/kg, (O cial Gazette of the RS 2012)) and maximum permissible value was discovered in the sediments from Raška river basin (Dramićki Stream, Kuzmićka River, and Kosurička River), Rudnjačka river basin (Rudnica and Lisinska River), Gokčanica river basin (Zastupski Stream and Gokčanica), Planska River II, Lopatnica, Pavlićka River, and Ibar I and II. In other sediment samples the concentrations were lower than the target value. The content of zinc in half of the samples was below the target value and it ranged from 81.5 mg/kg to 127.5 mg/kg (Fig. 4).
ARSENIC. The extremely high concentrations of arsenic were discovered in the sediments from Gajovska River, Rudnjačka river basin (Samokovka), Kaznovićka River, Maglašnica, Pavlićka River, Gokčanica river basin (Dubovski Stream and Zastupski Stream), and Planska River II. The contents of As between target value (29 mg/kg, (O cial Gazette of the RS 2012)) and maximum permissible value (42 mg / kg, (O cial Gazette of the RS 2012)) was found in sediments from the Raška river basin (Dramićki Stream, Deževska River I, and Jarandolski Stream), Jošanica river basin (Jošanica III and Planska River I), Rudnjačka river basin (Rudnica), and Gokčanica river basin (Zastupski Stream and Gokčanica). In other sediments the content of arsenic was lower than the target value. The content of arsenic in half of the sediment samples ranged from 15.3 mg/kg to 38.4 mg/kg (Fig.4).
Cadmium and mercury were below detection limit of the used method; cadmium in the overbank sediments from Ljudska River I, II, III, and IV, Raška I, Đurićka River, Šaronjska River, Deževska River I and II, Nosoljinska River, and Pivnica; and mercury in the sediments from Ibar and Ribnica, Šaronjska River,

Factor analysis
Based on the obtained level of signi cance and visual review of the histogram for all 10 elements, it was determined that the assumption of normality of distribution could not accepted.
Based on the obtained level of signi cance and visual review of histogram, the Ln variables for transformation values showed that the assumption of the normality of distribution could not be accepted for any of the tested heavy metals. Table 5 shows the values of Spearman's rank correlation for the analyzed heavy metals.
EXPLORATORY factor analysis on Ln transformation values by the extraction method determined that the examined heavy metals in overbank sediments were grouped around two synthetic factors and that they de ned 63.91% of the variance.
Based on the lling factor estimation matrix, it was determined that Cd, Zn, As, Pb, and Cu were concentrated around the rst synthetic factor, and Ni, Cr, and Co around the second factor.
CONFIRMATORY factor analysis con rmed that 68.96% of the variance could be explained by factor 1, and 95.16% by factor 2.
FACTOR 1. Table 6 shows the signi cance of correlation between the elements (Cd, Zn, As, Pb, and Cu) concentrated around FACTOR 1.
FACTOR 2. Table 8 shows the signi cance of correlation between the elements (Ni, Cr, and Co) concentrated around FACTOR 2.
The obtained numerical values of Spearman's rank correlation coe cient were in the excellent correlation range; Ni/Cr 0.952, Ni/Co 0.946, Co/Cr 0.910 (Table 5). Table 9 shows the surface distribution of element concentrations grouped around factor 2.
The quality of water in the Ibar River has improved from Raška to Kraljevo. Increased concentrations of heavy metals in the sediments of other watercourses have occurred due to the geological and pedological stratum underneath them.
Factor analysis identi ed two synthetic factors around which the analyzed heavy metals were grouped and further formed geochemical associations. Factor 1 explained 68.96% of the variance. Cd-Zn-As-Pb-Cu were concentrated around factor 1. The increased content of these elements occurred in northern and northeastern tributaries of Ibar. Extremely high contents were found in watercourses where occurrence and deposits of these heavy metals were already con rmed. The contents of Cd, As, and Cu were extremely high in the Rudnjačka river basin. The contents of Zn and Pb were high in the Jošanica river basin. The Gokčanica river basin had high contents of As and Pb, and Gajovska and Kaznovićka rivers had Zn, As, and Pb. The concentrations of Zn and As were increased in the Maglašnica river basin, and As content was increased in Pavlićka and Planska River II. The increased concentrations of Cu were only found in the Raška and Brvenica river basins.
Factor 2 was substantive and it de ned 95.16% of the variance. Ni-Cr-Co were grouped around factor 2. The high concentrations of these elements were present in almost all analyzed sediments, they were only slightly less present in southwestern tributaries of Ibar. This geochemical association was typical for this geological environment, i.e. ophiolites.
The analysis determined that heavy metals were grouped around these two factors based on similar geochemical behavior and that they actually represented geochemical associations. These geochemical associations were directly dependent on the environment lithology in which watercourses (or smaller basins) were formed and, to a lesser extent, on the immediate sampling environment. This was mostly minimized by the anthropogenic factor as a condition for the formation of these geochemical associations.

Conclusion
The Ibar is an international river. There are numerous objects along its watercourse or the watercourses that make up its basin that can affect the quality of sediments, and therefore the quality of water. The origin of the heavy metals in sediments from watercourses that form Ibar river basin and Ibar River were analyzed in this paper.
The quality of water in Ibar was mostly affected by the overbank sediments from Sitnica (Kosovo and Metohija) and Raška river basins. The anthropogenic factor had the greatest in uence on the quality of sediments from these two basins: the mining-industrial complex in Kosovo and Metohija on Sitnica river basin and municipal solid waste on the Raška river basin. However, due to the speci city of anthropogenic sediments (numerous tailing sites and mining-chemical compounds) in the Ibar river basin, especially in its upper and middle course, periodic control of the chemical composition of overbank sediments from its tributaries has been necessary.
The area through which the Ibar ows is ideal for the development of tourism. It has a potential for cattlebreeding and fruit growing that could be achieved with the right incentive. The Ibar river basin is covered in forest and has many geothermal water sources whose enormous potential is not fully exploited.