Specific Features of the Accumulation and Distribution of Heavy Metals in Soils of the Floodplain and Deltaic Landscapes of the Don River

Corresponding Author: Saglara S. Mandzhieva Southern Federal University, 344006, 105/42 st. Bol’shaya Sadovaya, Rostov-on-Don, Russia Email: msaglara@mail.ru Abstract: The purpose of this work was to study specific features of the accumulation and distribution of the total content and mobile forms of Heavy Metals (HM) in soils of the estuarine ecosystems of the floodplain and deltaic landscapes of the Don River. The results allowed us to detect regional specific features in the formation of HM compounds in floodplain soils of the Don River estuarine region and the changes that occur with soil pollution and to detect the influence of different factors on the transformation of metal compounds in the soils and estimate them ecologically. It was shown that the distribution of total HM content in the investigated soils is determined primarily by their content in pedogenic species of the Don river floodplain, as well as by soil factors: The organic matter content and particle-size fractions. The system of chemical element compounds forms a relationship of different compound groups, with strongly bound metal forms predominating. The soils can be ordered by their ability to strongly retain Cd, Pb, Mn, Zn, Cu, Ni: Alluvial-meadow heavy-loam > alluvial meadow light-loam > meadow alluvial-deposited > alluvial-meadow sandy and sandy-loam > alluvial-stratified sandy. This series completely corresponds to the decrease in their HM buffering capacity. Regional characteristics of HM behavior in Lower Don floodplain and delta soils include the weakly bound compounds Cd, Pb, Mn, Cu, Ni, which are represented mainly by specifically absorbed forms on carbonates and Fe-Mn (hydr) oxides.


Introduction
The floodplain of the estuarine region of the Don River is located in a semiarid climate. It is unique for the productivity of the territory, which is characterized by extremely favorable natural conditions. The high soil fertility and hydrological regimen of Don River have resulted in the formation of a highly productive meadow biocenosis of the floodplain and deltaic landscapes that serve as spawning areas for valuable fish breeds. At present, this territory is affected by active technogenic influences. The mostly important, in terms of its consequences, was the overregulation of the Don River runoff. The estuarine ecosystems serve as natural barriers, accumulating substances carried by the water, including contaminants. The cities of Rostov-on-Don, Novocherkassk, Azov and Bataisk (Russia) are located in the Don River estuarine region ( Fig. 1), as are large industrial establishments that emit large amounts of contaminant substances to the atmosphere. Emissions by the Novocherkassk Power Station (NPS) make up to 1% of the total emission of polluting substances into the Russian air, up to 35% and up to 50% of the aerosol emissions in Southern Federal district and Rostov region (ENPTS, 2001). As a result of the location of the main sources of air pollution and the dominant wind direction, a substantial portion of the fallout precipitates within the floodplain of river estuarine region (Fedorov et al., 2012). Research on the concentration and migration of chemical elements in the floodplain soils and their influence on crop capacity has been conducted in the past and is being conducted at present (Birch, 2011;Helz, 1976;Olawoyin et al., 2015;Ray and McCormick-Ray, 2013;Rovira et al., 2014;Sun et al., 2015;Zhang et al., 2010). A substantial number of works is dedicated to the geochemistry of Heavy Metals (HM) in the studied region (Alekseenko and Bofanova, 2002;Minkina et al., 2008a;2008b;2009;Fedorov et al., 2012). Despite the collected material, many questions on the accumulation, dispersion and transformation of HM in soil cover of the Don River estuarine region remain unsolved. Thus, this work is relevant.
The purpose of this work was to study specific features of the accumulation and distribution of the total content and mobile forms of HM in soils of the estuarine ecosystems of the floodplain and deltaic landscapes of the Don River.

Soil
From the viewpoint of geomorphology, the studied territory is confined to the Don River floodplain; as for hydrology, it is situated within its present estuarine region, the upper limit of which passes in range of Razdorskaya stanitsa (Gar'kusha and Fedorov, 2010). In turn, the delta is a part of the Don River estuarine region; the soil studies were mainly conducted in the floodplain of this region.
Monitor stations were installed to study specific features of the geochemical condition of the soil cover in landscapes of the estuarine region floodplain (Fig. 1).
The soil cover on the examined ground is represented by meadow, alluvial-meadow, saturated and alluvialstratified soils under laid with alluvial sediments (CDRS, 2004)-Fluvisols group (WRB, 2006. Soil samples were taken from a depth of 0-20 cm.

Physicochemical Analyses
According to the guidelines (AMSS, 1990) the main physico-chemical attributes of soils were estimated: Particle-size fractions, organic matter content, carbonates, exchangeable Ca 2+ and Mg 2+ , pH and dense water extract residue.
The condition of the examined soils was assessed by the value of the gross content of chemical elements in soils and by indices of the soil retention strength of metals and metalloids. The total content of Mn, Cr, Ni, Cu, Zn, Pb, Cd and As in soils was determined by X-rayfluorescent technique and expressed in mg/kg.
Weakly bound metal compounds include externalsphere metal complexes accompanied by a variety of functional groups of solid soil phases. Strongly bound metal compounds, which result from the formation of internal-sphere complexes with solid phases, are nonventilated, original minerals. We propose the use of the content of weakly bound HM compounds in soil as an integral indicator of metal mobility and the capability to participate in mass transfer processes (Mandzhieva et al., 2014).
Weakly bound HM compounds in soils include exchangeable, complex and specifically sorbed compounds, the content of which in soil is tightly associated with their content in plants (Minkina et al., 2009). HM compounds labeled as weakly bound are transferred into a solution by parallel extractions by reagents (Minkina et al., 2013).
About 1 N ammonium-acetate buffer (NH4Ac) with a pH of 4.8 (soil-to-solution ratio of 1: 10 and extraction time of 18 h), able to transfer exchangeable metal forms into a solution, which characterizes their "actual" mobility.
About 1% EDTA NH4Ac solution with a pH of 4.8 (soil-to-solution ratio of 1: 10 and extraction time of 18 h), which allegedly, along with exchangeable metal forms, transfers their complex compounds into a solution.
By the difference between the metal content in EDTA extract in NH4Ac and NH4Ac, we calculated the metal content found in the composition of complex compounds (Nosovskaya et al., 2001;Protasova and Gorbunova, 2006).
Acid-soluble metal forms derived with 1 N HCl solution (soil-to-solution ratio of 1: 10 and extraction time of 18 h), which characterizes the potential reserve of mobile compounds in soil. They are supposedly represented by exchange-capable metal ions and specifically retained compounds, including those retained by amorphous Fe and Mn oxides and carbonates.
We calculated the number of specifically retained metal compounds from the difference between the metal compounds in HCl and NH4Ac extracts. The additivity of extracts is experimentally proven (Minkina et al., 2009). The metal content in extracts is estimated by Atomic Absorption spectrophotometry (AAC).
The HM content in the composition of strongly bound compounds was estimated by the difference between the total metal content in the soil and the content of their weakly bound compounds.

Statistical Analysis
The results of chemical analysis correspond to the mean of three replicates. Descriptive data analysis, comprising minimum value, mean value, maximum value and confidence limits for mean, was carried out with STATISTICA 10.0. Statistical significance of the differences among means was determined by Fisher LSD test. Differences were considered not significant at values of p>0.05.

Results
Comparison of the results of chemical analyses of meadow, alluvial-meadow and alluvial-stratified saturated soils from the examined monitor stations (Table 1) demonstrates that the particle-size fractions of the soil is rather mixed and is presented mainly by sandy, sabulous and light-loamy species. Fractions of fine and medium sand are predominant. The examined soils have neutral, alkalescent, or strongly alkaline medium reactions (7.3-8.9) and a low humus content of 0.2-3.7, which was mainly determined by the qualitative content of the infused and redeposited material composing the inwashed and buried horizons. The overwhelming majority of the examined soils are not saline (the value of dense residue does not exceed 0.15%). The high carbonate content from soil covers in some stations is associated with the presence of biogenic calcite in deposited horizons. Calcium predominates in the composition of the absorbent complex (Table 1).
The gross soil analysis makes it possible to trace changes in the chemical composition of the most stable part of soils, the minerals. A macroelement analysis of soils showed that the SiO 2 content at stations 1, 4, 5, 7, 8a, which have a lite particle-size fractions, is increased relative to the amount of R 2 O 3 and other oxides, which contain primary nutritional elements: CaO, MgO and K 2 O (Fig. 2).
The accumulation and distribution of HM in soils is determined by external and internal factors. The external factors are conditioned by physico-chemical properties of soils, the geomorphological features of the examined territory and the association with sources of technogenic emission. The internal factors are related to the atomic properties of chemical elements and their compounds Sparks, 2003;Violante et al., 2007). A complex of physico-chemical parameters directly performs soil-protector functions. To reveal them, it is necessary to assess the soil protective abilities (buffering capacity) for HM.
The buffering capacity of the examined soils in the Don River floodplain was assessed according to technique of (Ilyin, 1995), which is based on an accounting of the inactivation capability of soil characteristics: Humus, physical clay (fractions <0.01 mm), carbonates, sesquioxides and pH.
Assessment of the soil buffering capacity has shown that it is somewhat different in the studied soils (Fig. 3). The major factors in the formation of the buffer value of the studied soils for HM are physical clay, carbonates and humus. The contribution of pH and sesquioxides is practically constant for the studied objects.
According to the buffering capacity gradation for HM, alluvial-stratified, sandy soil (station 4) and alluvial-meadow, sandy-loam soils (stations 7 and 8a) have a medium degree of this parameter, while the highest buffering capacity index characterizes meadow, light-loamy soil (station 2).  Table 1. Physicochemical soil properties of monitoring stations of the estuarine ecosystem of the floodplain and delta of the river Don (0-20 cm layer) Exchangeable cations, Size (mm) and mmol (+)/100 g contents (%) of particle-size fractions Light loam AL-alluvial-meadow soil, L-meadow soil, AS-alluvial-stratified soil The soils can be arranged by the buffering capacity value for HM in descending order: Meadow alluvialdeposited light-loamy ≥ alluvial-meadow light-loamy > alluvial-meadow heavy-loamy > alluvial-meadow sandy and sandy-loam > alluvial-stratified sandy.
The intensity of the accumulation and distribution of HM in soils is directly determined by the ecological conditions of their formation and buffering capacities. High values for the soil content of humus and clay particles facilitate active metal accumulation (stations 1, 3, 8, 9). In sandy-loam and sandy, alluvial-meadow, saturated and alluvial-stratified saturated soils (stations 4, 5, 7, 8a), hydrogenic accumulation processes take primary importance, but the HM content in such soils is low because of the low soil humus content and the low content of absorbed cations and clay particles (fractions <0.01 mm) (Table  1 and Fig. 3).
The distribution of total HM content in the soils of the monitor stations is presented in Fig. 4.

Manganese
In the soils of the monitor stations, the manganese content changes from 469.8 to 1910.0 mg kg −1 (Fig.  4). It was established that the manganese content in pedogenic species and floodplain and delta soils varies within wider limits, from 100 to 2000 mg kg −1 (Cherkashina et al., 2001). Meanwhile, the maximum manganese content is associated with soils with predominance of a 0.005-0.05 mm fraction and high organic matter content. A distinct connection between the manganese concentration and landscape relief also appears-the relatively low contents (200-300 mg kg −1 ) are associated with "sandy" ridges, while the concentration increases to 600-800 mg kg −1 in inter ridge depressions. On average, the manganese contents in pedogenic species of the floodplain and delta is 300 and 400 mg kg −1 (Lukyanchenko at al., 2001). The highest manganese concentration is associated with the meadow soil of the floodplain (station 2). Moreover, an excess of manganese clarke (Vinogradov, 1959) and Maximum Permissible Concentration (MPC) (TsINAO, 1992), constituting 850 and 1500 mg kg −1 , respectively, is observed.

Chromium
The chromium content varies from 51.2 to 121.4 mg kg −1 of soil. In pedogenic species of the Don River floodplain and delta, the chromium content is 69 and 97 mg kg −1 , respectively. On many grounds, an excess is observed not just in background values but also in the MPC of this element (Fig. 4). In the soil, chromium is multivalent with a predominance of poorly soluble compounds. Most of the chromium is present in the form of Cr 3+ as a part of minerals or in the formation of different oxides, displaying an affinity with ironcontaining phases in soils. Siderite (38-76%) and ferric hydroxides (15-25%) account for a significant part of the heavy fraction content of the Don River floodplain and delta soils (Lukyanchenko et al., 2001).

Nickel
The amount of nickel in the soils of deltaic landscapes is below the clarke values of the lithosphere, 41.8 mg kg −1 and MPC. However, in floodplain landscape soils, the nickel content reaches 45.6-60.9 mg kg −1 , which exceeds the nickel concentration in pedogenic species of the floodplain (16 mg kg −1 ) and delta (37 mg kg −1 ). Nickel is mainly retained by hydroxides and ferric and manganese oxides (Minkina et al., 2013).

Copper
Copper is characterized by high organophilic qualities and is related to strong complexing agents, being fixated in the soil in the form of strong organic chelates. The amount of copper in the organogenic layer of the examined soils varies from 12.3 to 60.9 mg kg −1 (Fig. 4). Pedogenic species of Don River floodplain and delta contain 21 and 41 mg kg −1 of this element. The copper content in the Don River floodplain soils is notable for its quite high values, in some cases exceeding the MPC (Fig. 4).

Zinc
In pedogenic species of the floodplain and delta, the zinc content is 49 and 84 mg kg −1 , which corresponds to the lithosphere clarke (according to Vinogradov (1959), the Zn clarke equals 50.0 mg kg −1 of species), background (72 mg kg −1 ) and MPC (100 mg kg −1 ). The high zinc content characterizes the soils with the most humus, those of stations 6, 1 and 3, which represent the biogenic accumulation of metal in the humus horizons.

Lead
In pedogenic species of the floodplain, lead is 14.0 mg kg −1 on average and 35.0 mg kg −1 in deltaic species, which is higher than lead clarke (10 mg kg −1 ). In the soils of the examined objects, the lead contribution varies widely, ranging from 2.5 to 33.0 mg kg −1 . The maximum metal concentrations are registered in monitor stations 1, 3, 5 and 5a.

Cadmium
In nature, cadmium appears in the form of small particles near smelting plants and from there they travel into the atmosphere, soil and water. Cadmium is not found in pedogenic species (Lukyanchenko et al., 2001). A small amount (0.4 mg kg −1 on average) is registered in the upper soil layer. Apparently, this can be explained by aerogenic pollution of the floodplain and deltaic landscapes by cadmium.

Arsenic
The main contribution of arsenic in the Don River floodplain and deltaic landscape ecosystems is associated with coal combustion products, waste products of the metallurgy industry and the usage of arsenic-containing pesticides. The arsenic contribution in the soil ranges from 6.2-11.6 mg kg −1 , which is several times that of the MPC and back-ground values (Fig. 4). The degree of arsenic pollution in the soils fluctuates from weak to strong. Meanwhile, arsenic accumulation, in addition to external factors (the presence of pollution sources), can be caused by its chemical properties and the ability to change its allotropic form in fluctuations of redox conditions. The background content (4.8 mg kg −1 ) and arsenic clarke (5.0 mg kg −1 ) in soil is 2.5 times higher than the MPC.

Discussion
The studied HM can be represented in sequentially descending order by their gross content in the soils of the monitoring stations in the floodplain and deltaic landscapes of Don River estuarine ecosystems: Mn > Cr > Zn > Ni > Cu > Bb > As > Cd.
The assessment of soil pollution by the gross metal content does not allow a determination of mobility or the ability to transfer to adjacent media, primarily, into plants and natural waters. The content of weakly bound HM compounds in the soil is more informative (Minkina et al., 2013). The amount of weakly bound HM forms varies from 18% in meadow soil to 81% in alluvial-stratified soil, depending on the metal (Fig. 5).
In floodplain landscape soils, the content of weakly bound compounds is 30-83% of the total content, on average, which strongly distinguishes them from zonal common chernozem soils, in which weakly bound compounds constitute only 10-20% ( Table 2). The contribution of weakly bound HM compounds is highest on the sandy soils of ground 4, which is linked with a low content of humus and clay particles, which are capable of retaining HM (Fig. 5 and Table 1). Thus, with a decrease in the soil buffering capacity, the contribution of weakly bound compounds increases, i.e., their mobility grows.
The contribution of strongly bound HM compounds varies greatly (from 18 to 82%) depending on examined soils of monitor stations (Fig. 5) and is mainly limited by the content of the fine-dispersed fraction. Of the total reserve of strongly bound HM compounds in soils, 69-74% are metals in silicate structures. The pollution level increases the amount weakly bound metal compounds, because the formation of extra amounts of mobile metal forms is observed in polluted soils (Mandzhieva et al., 2014).
Metals form the following ascending series by the soil content of weakly bound (% of total content): Cd > Pb > Mn > Zn > Cu > Ni. The contribution of weakly bound HM compounds is highest at monitor stations 1, 4, 5 and 6. With large volumes of emissions, the degree of HM mobility can reach 73-83% (Mazhaiskii et al., 2000). Regional characteristics of HM behavior in Lower Don floodplain and delta soils include the weakly bound compounds Cd, Pb, Mn, Cu, Ni, which are represented mainly by specifically absorbed forms on carbonates and Fe-Mn (hydr)oxides.
The content of the most mobile exchangeable forms is practically heterogeneous, from 1 to 63%, in the studied soils and directly depends on their buffering capacities and properties of chemical elements (Table 2 and Fig. 5).
A wide range in the content of exchangeable zinc forms is registered in the examined station soils (up to 30-fold). Mobile (exchangeable) Zn compounds slightly exceed the MPC in the soils of monitor stations 1 and 6 (Table 2 and Fig. 5). and Pb in soils of the floodplain and delta of the river Don, % of the total: I-meadow soil (station no. 2); II-alluvial-stratified soil (station no. 4); III-alluvial meadow soil (station no. 6) Table 2. Content of exchangeable, complex and, specifically sorbed forms soil of metals in the 0-20 cm layer of the monitoring stations of the floodplain and delta of the river Don, mg/kg The amount of exchangeable Cu forms is also insignificant and does not exceed the MPC, except for the most polluted monitor station (1), which is located in the floodplain in the general direction of the NPS and is affected by waste products.
The amount of exchangeable Pb forms in station soils is 5-33% of the total content. The higher relative content of Pb 2+ is linked with its ion radius. It is close to the ion radius of Са 2+ , which plays an important role in ionexchange interactions (Minkina, 2008a).
Exchangeable Cd forms, as well as their total content, exceed the MPC in the soils of many monitor stations ( Table 2). The highest extractability of Cd by ammoniumacetate buffer has been confirmed in works of researchers (Plekhanova et al., 2001;Fateev et al., 2001). The authors provide the share of Cd that transfers to this extract: It ranges from 22 to 60% of its total content.
By the content of exchangeable Mn forms, all of the studied grounds are not polluted (Table 2 and Fig.  5), however, a 2.5-to 10-fold variation of the content of this metal is observed depending on soil properties (stations 1 and 2).
HM in the studied soils form a series by the relative content of exchangeable forms (% of total amount): Cd > Pb ≥ Mn > Zn ≥ Ni > Cu.
The content of the studied HM in complex forms is in most cases higher than in exchangeable ones (Table 2 and Fig 5). HM form a series by the content of complex forms in soils (% of total amount): Pb ≥ Cu > Mn > Zn > Ni > Cu.
The relative content of complex forms of Zn and Ni in the studied soils is only 1-12% of the total content, which is explained by the weak complex-formation ability of Zn and Nu with organic matter (Piccolo and Stevenson, 1982). However, the contribution of complex forms of Pb can reach one-third of the total metal content in the soil.
The highest content of mobile metal compounds is represented by their specifically sorbed forms, which makes a potential reserve of elements. In alluvialmeadow saturated soils, which form on carbonate sediments and have an alkalescent reaction in HM immobilization, the role of ferric, and manganese hydroxides increases, as well as that for carbonates in the specific sorption of metals.
In the studied soils, the contents of specifically sorbed metal forms are ordered in the following sequence (% of total content): Mn ≥ Cu > Pb > Zn > Ni > Cd.

Conclusion
The in formativeness of the approach used in this study, which was based on a ranking of metal compounds that are strongly or weakly bounded in the soil, was proven. The results allowed us to detect regional specific features in the formation of HM compounds in floodplain soils of the Don River estuarine region and the changes that occur with soil pollution and to detect the influence of different factors on the transformation of metal compounds in the soils and estimate them ecologically.
It was shown that the distribution of total HM content in the investigated soils is determined primarily by their content in pedogenic species of the Don River floodplain, as well as by soil factors: The organic matter content and particle-size fractions. The system of chemical element compounds forms a relationship of different compound groups, with strongly bound metal forms predominating. The soils can be ordered by their ability to strongly retain Cd, Pb, Mn, Zn, Cu, Ni: Alluvial-meadow heavy-loam > alluvial meadow lightloam > meadow alluvial-deposited > alluvial-meadow sandy and sandy-loam > alluvial-stratified sandy. This series completely corresponds to the decrease in their HM buffering capacity.
Among weakly bound metal compounds, specifically sorbed forms predominate. At some monitor stations, pollution by exchangeable forms of copper, zinc and cadmium was found, which suggests technogenic HM accumulation. On the whole, the regularities in the distribution of mobile HM forms in the soil reproduce the regularities established for the distribution of their total content.

Acknowledgement
The work was supported by the Ministry of Education and Science (grant no. 5.885.2014/K) and the Russian Foundation for Basic Research (project no. 14-05-00586, 15-35-21134).

Author's Contributions
Tatiana M. Minkina: The head of the investigation, designed the research plan and organized the study, coordinated the data-analysis and contributed to the writing of the manuscript.
Dina G. Nevidomskaya: Participated in all experiments, coordinated the data-analysis and contributed to the writing of the manuscript.
Yuriy A. Fedorov: Contributed in drafting the manuscript and reviewing it critically for significant intellectual content. Saglara S. Mandzhieva: Coordinated the dataanalysis and contributed to the writing of the manuscript.
Tatiana V. Bauer: Collected the field data and contributed to the writing of the manuscript.
Victor A. Chaplygin: Collected the field data, performed the statistical analysis.
Aleksey K. Sherstnev: Collected the field data, determined the heavy metals.
Inna V. Zamulina: Analyzed the soil, performed the statistical analysis.