Assessment of the Impact of a Motorway on Content andSpatial Distribution of Mercury in Adjacent Agricultural Soils

Abstract

The distribution of Hg in the vicinity of roads is probably not exclusively dependent on car emissions, but also on the presence of other point or diffuse sources of Hg emissions located from metres to several km away. The source of mercury in urbanised areas is pollution derived from the burning of fuels and industrial and transport waste, while in agricultural areas, it is constituent in mineral fertilisers and crop protection products. The research objective was to evaluate the content and spatial distribution of mercury in arable soils adjacent to the A1 motorway in Poland. The research material consisted of 40 soil samples taken from 20 test points on four transects at distances of 5, 10, 25 and 50 m from a noise barrier and in the direction of an arable field, and 10 m from the noise barrier in the direction of the motorway. Total mercury content was determined by atomic absorption spectrometry using an AMA 254 analyser. The spatial relationship between adjacent observations of variables was assessed using Moran’s I overall autocorrelation coefficient. Probability maps of mercury distribution in the field and pollution indicators were elaborated in ArcGIS 10.4.1. using Inverse Distance Weighted interpolation. Analysis of the spatial correlation of Moran’s I showed a lack of spatial dependence between tested points, which may evidence that the motorway does not affect mercury contents in the soil. The elevated mercury content at a single test point may indicate a random event unrelated to the motorway’s operation.

1. Introduction

Mercury is considered to be a major hazardous substance due to its toxicity, mobility and long atmospheric residence time [1,2,3]. The presence of mercury in the soil environment may be caused by natural or anthropogenic factors [4]. The source of mercury in urbanised areas is pollution derived from the burning of fuels and industrial and transport waste, while in agricultural areas, it is constituent in mineral fertilisers and crop protection products. The activity of industrial facilities such as coal-fired power plants, coking plants and cement works constitute, in combination with car traffic, the anthropogenic sources of mercury in road dust [5]. Civilisational development and improvements in living conditions entail the need for a road network in agricultural areas, which causes sources of pollution to accumulate. Vehicle exhaust fumes are a stress factor that can negatively impact soil productivity. Higher vehicle velocities increase exhaust emissions, which contain heavy metals, among other compounds [6]. Interior vehicle lights and high-intensity discharge-type headlamps are also responsible for mercury pollution in roadside soil [7]. Studies about Hg content in soils from urbanized areas are scarce, but different works have found high levels of Hg in soils adjacent to roads. Moreover, the presence of mercury in crude oil and its derivatives that has been indicated by researchers [8,9] justifies research into its contents and spatial distribution in soils adjacent to motorways. The average share of particulate mercury depends on the research areas. Its concentration is significantly higher in urbanized areas than in agricultural areas [10,11,12]. Deposition of particulate mercury is an important part of its migration in the environment. In 2008, the following classification was adopted in a UNEP report (2008) [13], which distinguished three emission sources: current emissions from natural sources, current emissions from anthropogenic sources, and re-emissions from historical deposits from natural and anthropogenic sources. Secondary enrichment mechanisms-sorption, conversion at the solid-air interface are thought to have a significant impact on the concentration and speciation of particulate mercury in dust. Directive 2010/75/UE [14] has been developed, where regulations 595/2009 and 459/2012 specify emission requirements for motor vehicles (Euro 6). The mercury content in crude oil can even exceed 2 mg·kg⁻¹ [15,16]. The main task for improving our knowledge of mercury sources, fates, impacts, and emission control options was defined at the 14th ICMGP Conference in 2019 as bridging knowledge on global mercury with environmental responsibility, human welfare and policy response. Mercury, listed as one of the top ten chemicals of concern (WHO, 2017) is a potentially harmful trace element in the environment.

The research hypothesis assumed that road traffic and the noise barrier would influence the content and spatial distribution of mercury in surrounding soils.

The research objective was to evaluate the content and spatial distribution of mercury in arable soils adjacent to the A1 motorway in Poland.

2. Materials and Method

2.1. Study Area and Sampling Procedur

The research area is located in central Poland (53°21′14.6″ N 18°42′13.1″ E), in the vicinity of the A1 motorway, along international route E75 (Figure 1).

The motorway was opened to traffic on 14 October 2011. According to 2015 data from the Directorate General, traffic volume is in excess of 17,000 vehicles per day. The wind direction in the region varies throughout the year: about 35% westerlies, 25% easterlies, 20% northerlies and 20% southerlies [17]. The investigated agricultural area is separated from the motorway by a noise barrier of steel plates that is about 650 m long and 4 m high. The investigated area is covered by Luvisols (IUSS, 2014) formed from glacial till [18]. The presented area was not previously included in the assessment of the impact of anthropogenic factors on the mercury content in soils. This is of key importance due to the intensive agricultural production conducted in this area.

The research material consisted of 40 soil samples taken from 20 test points on four transects at distances of 5, 10, 25 and 50 m from a noise barrier and in the direction of an arable field, and 10 m from the noise barrier in the direction of the motorway. Soil samples were taken from the surface humic horizon at a depth of 0–20 cm and from the subsurface layer at 20–40 cm. Additionally, in order to determine the geochemical background, six samples were taken from a depth of 150 cm (parent rock) at six points randomly located within the research area. These samples were the parent material, in which the content of Fe and Hg was determined. In the year of sampling, intensive farming was being conducted in the study area (wheat, sugar beet and maize in rotation).

2.2. Laboratory Analysis:

2.2.1. Selected Physical and Chemical Properties

The following laboratory analyses were carried out in soil samples after initial preparation in accordance with the standard (ISO 11464:2006) [19]: particle-size composition using soil texture by sieving and sedimentation (ISO 11277:2009) [20]; total organic carbon content (TOC) by sulfochromic wet oxidation in a potassium dichromate solution (K₂Cr₂O₇) (ISO 14235:1998) [21]; potentiometric reaction in a solution of 1 mol·dm⁻³ KCl and H₂O according to ISO 10390:2005 [22] using a CPC-551 pH meter and calcium carbonate by field method and by the Scheibler volumetric method according to ISO 10693:1995 [23]. All determinations were made with three replicates.

2.2.2. Mercury Content

Total mercury content was determined by atomic absorption spectrometry using an AMA 254 analyser (Altec, Czech Republic). The analyser used for the measurements is a single-beam atomic absorption spectrometer with a serial arrangement of two cuvettes. The apparatus is designed for a limit of detection of 0.01 ng of mercury in a sample. The reproducibility of test results is assessed at <1.5%. The mercury content results are the arithmetic means of replicates not exceeding 5% of the coefficient of variation. Mercury compounds are highly volatile when heated, which makes analyses hard to conduct correctly. Therefore, to verify the reliability of the mercury analysis procedure, TILL-3 certified reference material (Certificate of Analysis, 1995) was also used for measurements.

2.3. Statistical and Geostatistical Analyses

Descriptive statistics parameters (minimum, mean and maximum values, standard [SD], kurtosis and coefficient of variation [CV]) were presented as the sum of mercury content (mg·kg⁻¹). To determine the correlation between total mercury content and chemical properties, Pearson’s correlation coefficient was used (p < 0.05). The results were statistically analysed in Statistica 13.1 software (TIBCO Software Inc.) for Windows PL.

The spatial relationship between adjacent observations of variables was assessed using Moran’s I overall autocorrelation coefficient [24]. Moran’s I global analysis involves identifying spatial patterns in soil chemistry that may be random, scattered or clustered, depending on location and sample values. The research hypothesis was that the mercury content in the entire research area was randomly distributed. Positive values of Moran’s I indicate that neighbouring values are similar, suggesting a positive spatial correlation, while negative values indicate that neighbouring values are different, suggesting an inverse correlation. A Moran’s I of zero indicates no spatial pattern. Probability maps of mercury distribution in the field and pollution indicators were elaborated in ArcGIS 10.4.1 (Esri). using Inverse Distance Weighted interpolation [25].

2.4. Pollution Indices

In order to assess the sources of mercury in the studied soils, the indices used were pollution and the geo-accumulation index. Pollution indices can assess the quality of soils exposed to heavy-metal contamination [8,26]. These indices are used in studies of mercury in dust [27] and in soil [28]. To determine the level of mercury contamination in the tested soil samples, the geo-accumulation index (Igeo) and the enrichment factor (EF) were used. These indexes are used to assess the presence and intensity of anthropogenic contaminant deposition on soil.

The geo-accumulation index [29] is used to define the contamination level of a single element. It is used to compare contents of metallic elements in tested soil material against natural metal concentrations. The formula used was:

Igeo = log [Cn/(1.5 × Bn)]

(1)

where:

Cn = Measured concentration of heavy metal

Bn = Geochemical background value

Cn refers to calculated values and Bn to background values. The factor 1,5 is applied to order control the variations of Bn values caused by the environment. The factor 1.5 allows for involving natural fluctuations in the concentration of a given substance in the environment and very small anthropogenic influences. The applied geochemical background values were derived from the analysis of the determined content in the parent material.

Geo-accumulation index [29]

Igeo ≤ 0—Unpolluted

0 < Igeo < 1—Unpolluted to moderately polluted

1 < Igeo < 2—Moderately polluted

2 < Igeo < 3—Moderately to highly polluted

3 < Igeo < 4—Highly polluted

4 < Igeo < 5—Highly to extremely polluted

Igeo ≥ 5—Extremely polluted

The normalised enrichment factor is used to determine differences in contents between metals from anthropogenic and natural environmental sources. It is also used to assess the distribution of metallic elements in different genetic layers of soils. It involves comparing the concentration of one element to another that is particularly stable and poorly mobile in soil. The selected component should also be bound to finer particles (depending on grain size), and its concentrations should be immune to human impact [30]. The conservative element most commonly used in soil contamination research is iron [31]. Other elements used in many studies are Al, Mn and Rb, and total organic carbon [30,32]. The calculations employ the content of individual metals defined as the background value. The formula used was:

$$\mathrm{EF}=\frac{\left(\mathrm{M}/\mathrm{Fe}\right) \mathrm{in} \mathrm{a} \mathrm{genetic} \mathrm{layer}}{\left(\mathrm{M}/\mathrm{Fe}\right) \mathrm{background} \mathrm{value}}$$

(2)

where:

M—total metal content,

Fe—total iron content.

According to Chen et al., EF values were divided into seven classes in which:

EF < 1 does not indicate enrichment;

EF < 3 is slight enrichment;

EF = 3–5 is moderate enrichment;

EF = 5–10 is moderately severe enrichment;

EF = 10–25 is severe enrichment;

EF = 25–50 is very severe enrichment;

EF > 50 is extremely severe enrichment.

An enrichment factor above 1.5 indicates anthropogenic accumulation of metallic elements [33].

3. Results

Grain-size composition tests were used to determine groups and subgroups of soil material in the surface and subsurface layers of the analysed soils. The soils were sandy loam and sandy clay loam (USDA, 2012). The sand fraction is the largest grain-size fraction in both layers, followed by silt and clay. The analysed samples contained 57.0–72.0% sand fraction, 16.0–26.0% silt and 7.0–21.0% clay. By agrotechnical soil weight category (PTG, 2008), they are classed as medium soils (

Table 1. Selected physical and chemical properties of soils.

No	Depth [cm]	pH H2O	pH KCl	TOC [g.kg−1]	CaCO3 [%]
1	0–20 20–40	7.6 7.6	7.2 7.2	3.8 2.6	<1 <1
2	0–20 20–40	7.0 7.3	6.5 6.8	8.6 5.4	<1 <1
3	0–20 20–40	7.1 7.0	6.5 6.3	4.2 2.7	<1 <1
4	0–20 20–40	7.1 7.2	6.5 6.8	2.7 9.3	<1 <1
5	0–20 20–40	7.4 7.5	7.1 7.2	9.3 5.6	1.3 1.4
6	0–20 20–40	7.4 7.1	7.1 6.4	3.9 2.0	<1 <1
7	0–20 20–40	7.4 7.1	7.0 6.3	2.9 1.9	<1 <1
8	0-20 20–40	7.3 7.2	6.9 6.8	2.1 3.2	<1 <1
9	0–20 20–40	7.6 7.7	7.2 7.3	1.8 1.5	4.6 5.3
10	0–20 20–40	7.6 8.0	7.3 7.4	4.6 0.9	1.4 5.6
11	0–20 20–40	7.2 7.1	6.9 6.6	5.5 4.0	<1 <1
12	0–20 20–40	7.3 7.3	7.1 7.0	7.7 6.5	<1 <1
13	0–20 20–40	7.5 7.5	7.1 7.2	6.3 7.6	1.3 1.5
14	0–20 20–40	7.3 7.2	6.8 6.8	4.6 4.9	<1 <1
15	0–20 20–40	7.8 7.7	7.3 7.3	1.9 1.5	2.3 2.8
16	0–20 20–40	7.1 7.2	6.8 6.8	4.3 5.7	<1 <1
17	0–20 20–40	6.0 6.4	5.3 5.6	6.9 5.6	<1 <1
18	0–20 20–40	7.4 7.7	7.2 7.3	4.9 6.2	1.3 3.9
19	0–20 20–40	6.9 7.4	6.8 6.6	10.7 3.9	<1 <1
20	0–20 20–40	7.6 7.9	7.2 7.3	3.6 2.9	2.0 6.4

). In the samples, pHH2O ranged from 6.0 to 8.0, while pHKCl ranged from 5.3 to 7.4, indicating acid to alkaline reaction. CaCO3 content ranged from <1 to 6.4% [31] (Table 1). Total organic carbon content (TOC) ranged from 1.8 to 10.7 g·k⁻¹ in surface layer samples and from 0.9 to 9.3 g·k⁻¹ in subsurface samples [34].

The tested soils have low TOC content: in most samples, TOC is below 2.2% value, the assumed average value for Poland [35,36].

Total mercury content ranged from 0.013 to 0.041 mg·kg⁻¹ in surface layers and from 0.013 to 0.059 mg·kg⁻¹ in subsurface layers (

Table 2. The content of Hg and Fe. Calculated geo-accumulation indexes (Igeo) and enrichment factor (EF).

No.	Depth (cm)	Hg (mg·kg−1)	Fe	Igeo Index	EF
1	0–20	0.039	962.30	−0.23	3.90
	20–40	0.027	957.80	−0.61	2.65
2	0–20	0.025	907.20	−0.65	2.41
	20–40	0.025	908.40	−0.68	2.33
3	0–20	0.021	894.70	−0.82	2.00
	20–40	0.022	877.20	−0.80	2.01
4	0–20	0.027	861.30	−0.59	2.41
	20–40	0.033	866.50	−0.39	2.97
5	0–20	0.039	975.20	−0.23	3.93
	20–40	0.016	1000.60	−1.14	1.63
6	0–20	0.025	1114.10	−0.66	2.93
	20–40	0.028	1116.00	−0.56	3.24
7	0–20	0.024	1115.50	−0.70	2.81
	20–40	0.025	1070.50	−0.66	2.82
8	0–20	0.024	981.90	−0.70	2.47
	20–40	0.025	959.40	−0.65	2.54
9	0–20	0.021	1082.00	−0.86	2.32
	20–40	0.021	1040.20	−0.85	2.25
10	0–20	0.024	946.90	−0.70	2.39
	20–40	0.022	1053.60	−0.78	2.46
11	0–20	0.032	935.10	−0.41	3.16
	20–40	0.036	923.90	−0.30	3.48
12	0–20	0.028	981.50	−0.56	2.85
	20–40	0.029	992.90	−0.52	3.01
13	0–20	0.025	1086.60	−0.66	2.86
	20–40	0.034	981.90	−0.37	3.45
14	0–20	0.022	971.70	−0.80	2.21
	20–40	0.018	960.60	−0.97	1.84
15	0–20	0.025	991.60	−0.67	2.59
	20–40	0.013	972.00	−1.30	1.34
16	0–20	0.029	886.70	−0.51	2.69
	20–40	0.037	921.90	−0.26	3.60
17	0–20	0.032	831.40	−0.42	2.78
	20–40	0.033	829.80	−0.39	2.85
18	0–20	0.041	1001.20	−0.18	4.25
	20–40	0.059	1024.30	0.19	6.31
19	0–20	0.025	1067.20	−0.66	2.80
	20–40	0.030	1107.30	−0.49	3.43
20	0–20	0.013	942.70	−1.29	1.31
	20–40	0.018	952.50	−0.97	1.83

). Soil is a vital resource for humans because its chemical and physical conditions affect agricultural production and the quality of its products. Mercury entrance in the food chain represents a geochemical risk because of toxicity and bioaccumulation phenomena.

The elevated mercury content in dust indicated by researchers [5,27] in an area influenced by vehicular traffic may be the cause of the even distribution of mercury contamination in the soil in the immediate vicinity of the motorway. The data from the present study have a fairly uniform distribution (heavily right-skewed), which differs from a normal distribution. A significant data amplitude and high skewness and kurtosis coefficient values were noted (Figure 2).

Mercury contents were lower and less variable in the surface layer than in the subsurface layer. The correlation coefficients (

Table 3. Statistical parameters.

Element Hg mg/kg	Statistical Parameters
	Average	Min	Max.	SD	Skewness	Kurtosis	CV (%)
0–20 cm	0.028	0.013	0.046	0.007	0.82	1.37	26.61
20–40 cm	0.028	0.013	0.059	0.01	1.58	4.45	36.03

) confirmed a significant positive relationship between total carbon content and total mercury content (r = 0.3878; p < 0.05), which is confirmed by previous studies [28,37]. A positive correlation between the level of Hg and the amount of organic matter has been confirmed by studies conducted on natural (non-agricultural) soils, while in cultivated soils no such correlation has been found, which may be due to the impact of agrotechnical treatments, which generally reduce the organic matter content of soil and increase exposure to air.

The soil texture has significant effects on the Hg content in soil. The concentrations of Hg and other heavy metals in soils generally tend to increase with decreasing grain size, which is due to the propensity of metals to bind with finer particles [38].

However, no relationship between mercury content and colloidal clay content was confirmed (Figure 2). According to Polish regulations (Regulation, 2016) [39], the investigated area qualifies as uncontaminated by mercury. Moreover, the values were determined to be typical of uncontaminated soils in Europe [40]. The average for European topsoil Hg concentrations was 0.04 mg kg⁻¹, with a range of 0–159 mg kg⁻¹ [41]. The average mercury content in the analysed samples was lower than that recorded by Kobierski et al. (2016) [28] in agricultural soils in Pomorze and Kujawy. A higher average mercury content was also recorded by Jezierski et al. (2011) [42] in southern Poland. The mercury concentration in the study area was similar to that recorded by Kobierski and Malczyk (2016) [28] in soils of hunting grounds, and in urban soils tested by Różański et al. (2015) [43].

However, it was lower than the mean values (0.316 mg·kg⁻¹) recorded in an industrialised zone of Beijing investigated by Guo-Li et al. (2014), [44] and Luo et al. (2008) [25]. In terms of the vertical distribution of mercury, in the case under study there is no such relationship as that indicated by Wang et al. (2012) [1], who showed that the concentration of mercury was greatest in the arable layer (0–20), which mainly evidenced the external source of this element in the soil. The natural mercury content in Poland, i.e., the background value for soils, is 0.05 mg·kg⁻¹ [45]. Salminen (2005) [40] reports lower background mercury levels (0.017 mg·kg⁻¹) in some regions of Poland. The present analyses indicate that in much of the study area, mercury concentrations do not exceed background levels. At only one test point in the subsurface layer did values slightly exceed background levels (Table 2). The spatial distribution of mercury and the selected indices (Igeo and EF) indicate that mercury contamination is lowest nearest to the motorway, which excludes the road as a source of mercury in this area. The geo-accumulation coefficient shows no harmful values.

Enrichment factors are used to indicate an exogenic origin (atmosphere) to soils with background contents of the metal considered (Hg in this case), that is not enriched due to the lithology, but you can never distinguish from a natural or an anthropogenic origin of that metal [46].

The enrichment factor indicates a natural or moderate source of this element in the field, except at one point, where the contamination factor was very high. To illustrate the spatial differentiation in total mercury content and the selected indices in the studied soils, a map was drawn up, for which the input point data were interpolated using Inverse distance weighting (IDW) (Figure 3).

The spatial distribution is in line with the increased mercury content in the field being a random event and it cannot be associated with the location of the motorway. The lack of spatial correlation is indicated by the Moran analysis performed with the Cluster and Outlier Analysis tool (Anselin Local Morans I) in ArcMap 10.4.1(Esri) (manufacturer or laboratory, city, country).

Hg pollution is a global problem. Mercury is an element related to the intensity of human pressure on the environment. The conducted research allows for the assessment of the impact of road traffic and noise barrier on the level of Hg concentrations in agricultural soils, which indicates the possibility of Hg transfer to the human food chain.

4. Conclusions

• The total mercury content ranged from 0.013 to 0.041 mg·kg⁻¹ in the surface layers and from 0.013 to 0.059 mg·kg⁻¹ in the subsurface layers, which qualifies the study area as uncontaminated by mercury.
• The correlation coefficients confirmed a significant positive relationship between total carbon content and total mercury content.
• The geo-accumulation coefficient shows no harmful values.
• The enrichment factor indicates a natural or moderate source of this element in the field, except at one point, where the contamination factor was very high.
• Analysis of Moran’s I spatial correlation showed a lack of spatial dependence between tested points, which may evidence that the motorway does not affect mercury contents in the soil.
• The elevated mercury content at a single test point may indicate a random event unrelated to the motorway’s operation.
• The geochemical monitoring of soil is important in the aim of evaluating the natural content of mercury in soils, related to parental materials and possible enrichment due to human activities.