Water, Air, and Soil Pollution

, Volume 200, Issue 1, pp 119–132

Tracing the Metal Pollution History of the Tisza River Through the Analysis of a Sediment Depth Profile


  • H. L. Nguyen
    • Department of Analytical and Environmental ChemistryVrije Universiteit Brussel
    • Faculty of Chemical TechnologyHanoi University of Technology
  • M. Braun
    • Department of Inorganic ChemistryUniversity of Debrecen
  • I. Szaloki
    • Department of Nuclear TechniquesBudapest University of Technology and Economics
  • W. Baeyens
    • Department of Analytical and Environmental ChemistryVrije Universiteit Brussel
  • R. Van Grieken
    • Department of ChemistryUniversity of Antwerp
    • Department of Analytical and Environmental ChemistryVrije Universiteit Brussel

DOI: 10.1007/s11270-008-9898-2

Cite this article as:
Nguyen, H.L., Braun, M., Szaloki, I. et al. Water Air Soil Pollut (2009) 200: 119. doi:10.1007/s11270-008-9898-2


The vertical profiles of 20 major and trace metals were investigated along a 180-cm-long sediment core, which was sampled at Kiss-Janosne-Holt Tisza, an oxbow lake located in the upper part of the Tisza River in Hungary. The vertical profiles showed sharp peaks at different depths, reflecting historical pollution events and unusual changes of river water characteristics. Five different groups of metals, containing metals which were strongly correlated and showing a similar behaviour, could be distinguished by factor analysis. Six areas, with variable degrees and types of contamination, were classified in the sediment core with cluster analysis. The most polluted sections were found in the upper 50-cm part (significantly contaminated by Cu, Zn, Pb, Cd and Hg) and the deeper 100–120-cm part (characterised by high concentrations of metals associated with mining activities, such as Fe and Mn, as well as Cu, Zn and Pb). In recent years, important pollution events, such as the one which took place in March of 2000, were the reason for pollution of the upper sediment layers, whereas mining activities during the last century were responsible for the pollution of the deeper core sections.


Heavy metalsSediment corePollutionTisza River

1 Introduction

The origin of metals which accumulate in sediments is partly from natural sources through the weathering of rocks and partly arising from a variety of human activities, including mining, smelting, electroplating and chemical manufacturing plants, as well as domestic discharges, shipping, boating activities…(Forstner and Wittmann 1983). Due to their particle reactivity, trace metals tend to accumulate in sediments and may persist in the environment long after their primary source has been removed (Park and Presley 1997). Consequently, sediments are not only important carriers but also potential sources of contaminants in aquatic environments (Salomons and Forstner 1984). Being formed by the deposition of fine particles with their associated contaminants, each layer of buried sediment represents a record of the environmental conditions, reflecting the water quality and possible effects of anthropogenic contamination at a certain period (Von Gunten et al. 1997). As a result, sediments are one of the most important tools to assess the contamination level of aquatic ecosystems. It has been reported that both natural background levels and man-induced accumulation of metals over an extended period of time can be traced by studying dated sediment cores (Forstner 1990). Using this technique, a historical record of the various influences on the aquatic system can be reconstructed. A lot of research on sediment depth profiles in lakes or rivers has been carried out and has provided valuable information about temporal variation of metals, the relationship with the evolution of adjacent terrestrial ecosystems and emission sources or the reconstruction of the metal pollution history of the aquatic area, for example (Iskandar and Keeney 1974; Park and Presley 1997; Dauvalter and Rognerud 2001; Ying et al. 2002; Ciszewski 2003).

The Tisza river is the longest tributary (977 km) of the Danube river, having the largest drainage basin (157,200 km2), which is shared by five countries: Ukraine, Romania, Slovakia, Hungary and Servia (Laszlo et al. 2000). The upper Tisza river, flowing through Hungary, has been subject to severe metal pollution due to mining activities and major industrial complexes in Romania, near the border of Romania and Hungary. Heavy floods, which occur regularly due to snow-melting and intensive precipitation, are a great potential risk of enhancing these metal pollutions by leaching of the mine tailing, which are stocked nearby the river banks, as well as spreading of the contamination outside the river bed and further downstream (Hamar and Sarkany-Kiss 1999, Osan et al. 2002, 2007). The river has about 70 oxbow lakes, with surface areas larger than 5 ha, that are only connected to the main river during flooding events (Braun et al. 2000). During these events, a large amount of suspended particles and associated metals are transported by the river and deposited onto the floodplains. As a result, sequential sediment layers have built up, containing important geochemical information about changes in chemical and hydrological conditions of the river at the moments of deposition. This relation is most evident in areas of extensive ore exploitation and processing when the pollution with heavy metals is large (Ciszewski 2003). Therefore, sediment depth profiles, especially in oxbow lakes of Tisza River, are considered as historical records that provide valuable information on environmental changes and industrial or/and mining activities in the river catchment area.

In this study, the vertical distribution of major and trace metals in a sediment core of an oxbow lake was investigated. Different groups of metals were identified based on their common source, similar behaviour and distribution within the core. Dating the scale of the core was estimated by linking contaminated sections with flooding events. From these analyses, a reconstruction of the metal pollution history of the Tisza River could be made.

2 Materials and Methods

Using a 5-cm-diameter Livingstone piston corer (Walker 1964), a long sediment core (180 cm in length) was collected at Kiss Janosne Holt Tisza (48°05′17.40N and 22°33′12.07E), one of the hundreds of oxbow lakes located in the Upper Tisza river (Fig. 1). The oxbow lake is located near the village of Tivadar, located at about 200 km from Baia Mare and the Baia Borsa mining area of Romania. The sampling was carried out in August 2001, 1 year after the severe cyanide and heavy metal spill from the Baia Maire and Baia Borsa mining fields into the rivers Szamos and Viseu, tributaries of the Tisza river. The sediment core was sectioned into 89 sections of 2-cm thickness, and then the samples were frozen, dried in a lyophiliser and ground into finer particles. For the determination of metals, sub-samples were digested with Aqua Regia HCl/HNO3 (3/1 v/v) in a microwave oven CEM MDS 2000 (Nguyen et al. 2004a, b). To verify the reproducibility of the digestion procedure, blanks and the standard reference material, IAEA 405, Trace Elements in Estuarine Sediment, were included in each digestion batch. Mercury was quantified by a Cold vapour Atomic Absorption Spectrometry (Thermo Separation Products). Other trace metals were determined by inductively coupled plasma–mass spectrometry using a VG-Elemental Plasma Quad II instrument. Major metals, phosphorus and sulphur were analysed using inductively coupled plasma–atomic emission spectrometry with a Thermo Jarrell Ash IRIS Advantage/1000 Radial Plasma Spectrometer. For methyl mercury analysis, samples were extracted with HNO3 (4M) in a Microwave Prolabo A300, followed by extractions with CH2Cl2 and back extraction into MilliQ water. Methylmercury is then analysed by aqueous phase ethylation and headspace gas chromatography with atomic fluorescence detection (Leermakers et al. 2003).
Fig. 1

Map of the sampling area

The analytical results were statistically processed with factor analysis to group metals with similar behaviour and with cluster analysis to identify sections with various levels of contamination in the sediment profile. Based on the obtained clusters and a vertical time scale using water levels and specific flooding and pollution events, the metal pollution history of the Tisza river sediment was reconstructed.

3 Results and Discussions

3.1 Vertical Variation of Heavy Metals in Sediment of Kiss Janosne Holt Tisza, Tisza

The average trace metal concentrations over the whole length of the core (180 cm) were 110 ± 13.5 μg g−1 for Cr, 21.2 ± 2.3 μg g−1 for Co, 70 ± 8 μg g−1 for Ni, 62 ± 14 μg g−1 for Cu, 216 ± 55 μg g−1 for Zn, 0.7 ± 0.4 μg g−1 for Cd and 48 ± 34 μg g−1 for Pb, respectively. These values are higher than natural background levels in Hunagrian soils (Hungarian Ministry of Environment and Water 2000). The concentrations range in between the values considered as polluted sediments and concentrations toxic to biota (Table 1).
Table 1

Hungarian authority standards for soil (mg·kg−1 dry weight)


A value (natural background)

B value (polluted)

C value (intervention to biological species)

































Vertical distributions of trace and major metals in the oxbow lake sediment are illustrated in Figs. 2 and 3, respectively. The depth profiles show that there is neither a gradually increasing nor decreasing trend in metal concentrations. Only in some parts of the core, sharp peaks are recognised probably as the result of different pollution events, rather than atmospheric depositions contaminating the sediments of the river catchment area (Yang and Rose 2003). Some of those metals follow similar variations along the depth profile, while others do not.
Fig. 2

Vertical profiles of trace metals and sulphur along the sediment core

Fig. 3

Vertical profiles of major metals along the sediment core

Cu, Zn and Pb show sharp peaks at 15 and 120 cm from the top of the core, suggesting their common origin or the same effects for these metals happening at the same moment. In other words, the pollution events occurring at these moments showed elevated levels of Cu, Zn and Pb. The concentrations are twice, triple (Cu, Zn) or even six times (Pb) larger than their average or local background levels (Table 1) implying a severe pollution of the river sediments by those events. Actually, two recent mining accidents which took place in Romania are considered the most suspect for the metal pollution in the top 15 cm of the core. In March 2000, a solid waste pond of Baia Borsa mining field, broken due to heavy rainfalls and melting of large snow layers, released about 20,000 tons of sludge contaminated with Pb, Zn and Cu to the Tisza river. Together with the cyanide and heavy metal spill from Baia Mare in January 2000, it was recently the most serious environmental disaster in Tisza river (Laszlo et al. 2000; WWF 2002). In other studies (Laszlo et al. 2000; Fleit and Lakatos 2003; Osan et al. 2002, 2007), high concentrations of these elements were also noticed in 10-cm surface sediment layers of Tivadar village, as well as the upper Tisza, as indications of these newly arrived pollution events. For the peaks occurring at a depth of about 120 cm of the core, the most possible explanation is heavy runoff from historical metal exploitation. Mining has been going on for centuries in the Baia Mare region, surrounding the streambeds of the Szamos river basin, the most important tributary of the Tisza. Important pollution from mining origin occurred about 40 years ago, together with the flourishing of the non-ferro industry (Baciu 2002), and left an imprint on the sediment of the Tisza river, as we observed at the depth of 120 cm. Several small peaks which appeared in the middle of the core may be the result of heavy floods, which occur regularly in Tisza catchment area (Hamar and Sarkany-Kiss 1999).

The profiles of Cr, Co and Ni are “broken” into two parts at the depth of 80–85 cm. At this depth, minimum concentrations of these metals are observed, implying a decrease of deposition of suspended metals from the catchment area due to a period with low rainfall amounts (Fig. 4) (VITUKI 2003) and a change in the sediment composition of the deposited matter. The major elements Al, Ba, K, Na and Sr also show a similar decrease at 80–85 cm. Apart from this “break”, frequent variations of these metal concentrations over the depth occur, but they are limited in intensity. A severe accident, which occurred in December 1999, at one of the Romanian industrial complexes near the border with Hungary, caused Cr level to rise to 20 times the permissible values (Judith 2001). Although the pollution only reached the lower part of the Tisza river, flooding events spread out the contamination over the whole catchment area, possibly resulting in the peaks occurring at about 20 cm depth of the oxbow sediment. Apart from this accident and effects resulting from frequent flood occurrences, no other link with specific environmental changes in these metal supplies to the river can be forwarded.
Fig. 4

Fluctuations of the water level of Tisza river during the last century

Cd and Hg show similar profiles and seem to be related to the S content of the sediment. Significantly higher levels of these elements (four to five times the mean levels) are recognised at a depth of 30–50 cm of the core. This may be due to contamination arising from discharged wastewater from the gold mining activities in the upper river. The peaks observed at the top of the core (5 cm) in Cd and Hg may also have originated from the metal pollution catastrophes which occurred in early 2000. Fleit and Lakatos (2003) also found increases of Cd in surface sediment layers sampled at Tivadar and the surroundings. MeHg concentrations are correlated to the total Hg concentrations. MeHg in the sediments is formed by in situ methylation of inorganic Hg by sulphate-reducing bacteria and can be decomposed to inorganic Hg by demethylating bacteria. Higher levels of MeHg are observed in the top sediment layers (0–30 cm) and sharp peaks at 35 and 45 cm coincide with the total Hg peaks, indicating that, during pollution events, Hg accumulated in the sediments in a form which was readily available for methylation.

Similar to what was observed in the Cu, Zn and Pb patterns, sharp peaks appeared in the profiles of Fe and Mn (100–120 cm) and P (60–80 cm) which may be attributed to the chronically and accidentally polluted discharges from mining areas in the upper Tisza river during the last century. These peaks are as high as three to five times above the average background values of Fe (54 ± 6.5 mg g−1), Mn (630 ± 179 μg g−1) and P (842 ± 152 μg g−1). Consequences of these mining activities were also recognised, with high concentrations of Fe and Mn (at similar levels), in another sediment core collected at the oxbow lake Marot-zugi-Holt-Tisza situated in Upper Tisza river (Braun et al. 2000).

For the other major elements (Al, Ca, Mg, Na and Sr), apart from a decline at about 80-cm depth of the core, no clear pattern is determined along the sediment depth. Similar small peaks noticed at the same sediment layers reflect the impact of flooding frequency and intensity perturbing the normal sediment depth profile of an oxbow lake. Our observations are supported by the results of Braun et al. (2000), which showed similar consistent variations of these elements over an about 130-year time span.

3.2 Similar Behaving Metal Groups in Sediment of Kiss Janosne Holt Tisza Oxbow

To determine which of the 20 major and trace elements analysed in the Tisza sediment shows a similar behaviour or a common source, factor analysis was performed on all data. In such a way, the number of variables (metals) was reduced by reorganising them into groups, which included similar behaving metals. Five factors with Eigen values larger than 1 were extracted using a principal component extraction method, followed by a varimax normalised rotation to facilitate interpretation. Factor loadings of new variables are shown in Table 2. The composition of the five factors is the following: (1) major elements in sediments (Al, Ba, K, Mg, Na and Sr), (2) metals specifically originating from recent pollution events (Cu, Zn and Pb), (3) metals and species associated to sulphur (Cd, Hg, MeHg, S and Ca), (4) major metals of mining industry (Fe, Mn and P) and (5) the remaining metals (Cr, Co and Ni). Metals belonging to the same group reveal a strong correlation among them and suggest their common origin, similar behaviour and accumulation in the oxbow sediment of the Tisza area. For example, Cu, Zn and Pb are correlated in sediments of this area because their concentrations are regulated by a common source: discharges from intense ore mining and metal processing in the upstream of the Tisza river. Factor 3 also groups a number of elements and compounds, such as Hg, methylmercury and sulphides, that are interrelated by natural processes occurring in reduced environments. Methylation of Hg is controlled by the activity of sulphate-reducing bacteria also producing sulphides, as already reported for different areas (Baeyens et al. 1998; Regnell et al. 2001). Besides the natural formation of sulphides in reducing environments, sulphide-bearing minerals from mining activities are also transported as suspended matter in the Tisza River (Osan et al. 2002, 2007).
Table 2

Loadings (>0.6) of five extracted factors from principal component analysis


Factor 1

Factor 2

Factor 3

Factor 4

Factor 5





































































































































Scores of extracted factors are plotted vs the depth profile of the core, as shown in Fig. 5. It can be noted that random fluctuations around the zero level indicate no specific change of these metal groups with depth (factor 1 and factor 5). Conversely, sharp and relatively high peaks are noticeable for factor 2, factor 3 (in the upper and deeper part) and factor 4 (in the middle of the core). These peaks are totally in agreement with observations of individual metal concentration profiles along the depth of the sediment core described above. The vertical variation patterns of the metals are again confirmed by the distribution of the five factor scores along the sediment depth.
Fig. 5

Distribution of factor scores along the sediment depth

3.3 Cluster Analysis of the Metal Pollution along the Depth Profile

The metal contamination of Kiss Janosne Holt Tisza oxbow sediment can be classified into six categories as a result of the cluster analysis. Each category is characterised by a group of metals, which shows the most significant one with respect to a given cluster. The tree diagram for the classification of the sediment layers by cluster analysis is presented in Fig. 6.
Fig. 6

Tree diagram for the classification of the sediment layers by cluster analysis

The first cluster includes sediment layers which are highly polluted by Cu, Zn and Pb in the top of the core (10–15 cm) as a result of the recent pollution events. A rapid decline of these metal concentrations before and after the pollution events is another characteristic of cluster 1. The second cluster shows maximum concentrations of Hg and Cd at 30–35-cm depth. The third one is related to the low concentrations of major metals Al, Ba, K, Mg, Na and Sr and trace metals at a depth of 80 cm, probably resulting from dry periods. Significantly higher concentrations of Fe and Mn at a depth of 115–120 cm are characteristic for cluster 4, which are the result of historical mining activities. Cluster 5 groups sediment sections, which are associated with minor variations in Cr, Co and Ni concentrations (90–110 and 125–130 cm). The last cluster, the largest one, reveals no information on metal pollution on important flood events and covers the section from 130 cm to the end of the core. This part reflects background levels of heavy metals in sediments of the Tisza catchment area (Table 1).

3.4 Reconstructing Pollution History of the Tisza River

To identify unusual increases of metal concentrations in different sections of the core as a result of pollution events or severe floods, it is necessary to determine the age profile of the core. Several methods exist to assemble an age profile of the sediment core. In areas polluted by mining or industrial discharges, production changes recorded in the economic history of those catchment areas make it possible to determine the age profile of deposited sediments (Knox 1987). Isotopic dating with 14C, 234U, 230Th (Schwarcz 2002), 210Pb (Appleby and Oldfield 1992; Godoy et al. 1998; Lee and Cundy 2001) and 137Cs (Delaune et al. 1978; Catallo et al. 1995; Gallagher et al. 1996; Gambrell et al. 2001) are also techniques widely applied for this purpose (Macklin et al. 1994; Kudo et al. 2000; Gambrell et al. 2001). Analyses of historical maps and/or artefacts are further possibilities.

In this study, the age of the core is estimated by a time model, which is established using the maximum water levels of the Tisza river during the last century, as this determines the inundation of the oxbow. The Tisza River is able to run into the oxbow whenever the water levels exceed 650 cm. The principle of establishing this time model is based on the fact that drops of metal levels in the sediment core should match the water minima, and peaks appearing at a given sediment depth are correlated with flooding or pollution events. Scores of factor 2 are used as a representative for variation of metals, which are significantly affected by pollution associated with flooding events. The variations of the maximum river water level [in function of time (years)] and factor 2 scores (in function of depth) are shown in Fig. 7. Identification and fitting of matching parts of factor scores and water levels was performed using PC-SLOT, a programme for optimal sequence-slotting of two sequences subject to stratigraphic constraints (Clark 1995, 1996). The principle and algorithm of sequence-slotting were described in several publications (Gordon 1980; Clark 1985). Sequence slotting is a technique used for combining two sequences of data into a single consolidated sequence. This method is reported as the best to use for relating the depth scale of one sequence to another or an unknown timescale to a known one (Gardner 2000). PC-SLOT automatically does such matching of depth (or time) scales, using a technique known as “partial path length interpolation”. The partial path length, defined as the total “distance” along the optimal combined sequence from the start to the current element, was used for time calibrating (Fig. 8). The dating of the core could be estimated based on this time model.
Fig. 7

Variation trends of water level of Tisza river and Cu, Zn and Pb concentrations

Fig. 8

Fitting calendar of the water level with partial path length

By plotting the metal concentration variations on a time scale, the pollution history caused by accidental metal spill events or discharges from mining industries accumulating in oxbow lakes during flooding periods of the Tisza river was reconstructed. Figure 9 shows that sharp peaks of Cu, Zn and Pb found in the upper part of the core (15 cm) are the result of recent metal spills from Baia Mare and Baia Borsa old mine tailings occurring in January and March 2000. The rapid metal concentration rises in the deeper part (120 cm) are related to the runoff storm from mines occurring during 1955–1960. This also resulted in high elevations of Fe, Mn and P in the deeper part of the core (100–120 cm), corresponding to the period of the1950s. These are all the consequences of rapid industrialisation in Romania after the Second World War. Numerous mining and enrichment centres were built or continuously developed for the exploration of iron ores (during the 1950s) and non-ferrous metal deposits such as Cr, Ni, Al, Zn, Cu, Sn, Pb, Au, Ag… (throughout 1970s and 1980s). Together, non-ferrous metallurgy became increasingly important during those times (US-Congress, 2003). The appearance of Cd and Hg peaks before the cyanide and heavy metal spill accidents suggests uncontrolled or improper releases of these metals, by-products of the gold and silver mining processes from Baia Mare during 1975–1985. Sulphur, rarely exploited before World War II, was also obtained as a metallurgical by-product or refined from gypsum, an abundant mineral in Romania, with considerable amounts over these periods (US-Congress, 2003). A slight increase in background levels of Cr, Co and Ni starting from 120 cm to the surface of the core might be explained by a development of the metal industry from 1950 onwards. Small peaks observed in between clusters, from the middle of the core up to the surface, may be attributed to the flooding events occurring regularly in the Tisza catchment area.
Fig. 9

Reconstruction of metal pollution history of the Tisza river

4 Conclusions

Heavy metals which have accumulated in the sediment of the oxbow lake were correlated with each other based on either their common origin (Cu, Zn and Pb from recent severe mining accidents, Fe and Mn, as well as Cu, Zn and Pb, from mining runoff during the period of industrial development), their similar behaviour (Cd and Hg, associated to S) or their consistence distribution in the investigated area (Al, Ba, K, Mg, Na and Sr unaffected or Cr, Co and Ni moderately affected by the pollution events).

Based on the metal contamination levels and the metal groups involved, the depth profile of the Kiss Janosne Holt Tisza oxbow sediment can be separated into six parts. The most polluted sections were found in the upper 50-cm part (significantly contaminated by Cu, Zn and Pb with a peak at 15 cm and by Cd and Hg with a peak at 30 cm) and the deeper 100–120-cm part (characterised by high concentrations of metals associated with mining activities, such as Fe and Mn, as well as Cu, Zn and Pb). In recent years, important pollution events, such as the one which took place in March of 2000, were the reason for pollution of the upper sediment layers, whereas mining activities during the last century were responsible for the pollution of the deeper core sections. A decrease in metal concentrations, as well as major element concentrations, was observed at 80 cm, reflecting a change in sediment composition during an extended dry period. The bottom 45-cm section is more uniform, reflecting background levels, unaffected by severe pollution events.

The reconstruction of the metal pollution history of the Tisza river confirmed that the upper 15 cm of the core highly polluted by Cu, Zn and Pb was the result of metal spills from Baia Mare and Baia Borsa industrial complexes occurring in March 2000. Other mining or related metallurgical processing pollution events occurring during the 1950s and 1960s were the reason for high levels of these metals in the deeper part. Hg and Cd contaminations from the depth of 50 cm toward the surface originated from the Baia Mare gold mining area, where cyanide was used for gold extraction from ores. The time scale indicated that the pollution of these metals became severe in the period of 1975–1990. As a consequence of uncontrolled or improper mining discharges during the period of rapid industrialisation in the 1950s and 1960s, significantly large amounts of Fe, Mn and P, as well as Cu, Zn and Pb, accumulated in oxbow sediment. The extremely low water level of Tisza river observed during 1959–1961 was responsible for a sudden decline of Cr, Co, Ni and major element concentrations.

From all analyses, it can be concluded that the sediment core of the oxbow lake, containing valuable time-dependent information of the chemical and hydrological characteristics of the river, can be effectively used for reconstructing the metal pollution history of the river.


We are very grateful to Bela Csapo, teacher of Tarpa Primary School, for his help during the sampling expedition. This research was funded by the Flemish Government and Hungarian Education Ministry through the Flemish–Hungarian Bilateral Scientific and Technological Co-operation under contract project Nr. B-00/76 and by the Flemish government through a grant for H. L. Nguyen.

Copyright information

© Springer Science+Business Media B.V. 2008