Descriptive statistical analyses.
Trace metals.
The descriptive statistics of the content of trace metals in the vegetable field soils from the Eastern Nile Delta are given in Table 1. The descriptive statistics results show that the mean concentrations of Cd, Co, Cr, Cu, Fe, Mn, Ni, Pb and Zn were 2.36, 16.12, 151.56, 9.85, 20972.74, 479.63, 27.25, 28.31 and 100.01, with a median concentration of 2.78, 19.98, 178.90, 10.48, 25634.40, 576.30, 29.68, 31.30 and 116.91 mg kg−1, respectively. The SD values were high because the concentrations of trace metal shad quite high heterogeneous distribution in the studied area.
Table 1
Descriptive statistics of trace metal concentrations (mg kg−1) and soil properties (clay %, organic matter % (OM), pH, and electrical conductivity (EC) dS m−1) in the soils of the study area. K–S: Kolmogorov–Smirnov test, CV: coefficient of variation, BG: Background as reported by Kabata-Pendias (2011), Avg: average, sd: standard deviation, cv: coefficient of variation, max: maximum value, min: minimum value, skew: skewness, Q1: lower quartile, middle quartile (median), Q3: upper quartile, MAC: Ranges of Maximum Allowable Concentrations for trace metals in agricultural soils (mg kg−1) (Kabata-Pendias, 2011).
Variables | Minimum | Q1 | Median | Mean ± S.D. | Q3 | Maximum | CV % | Skewness | Kurtosis | K-S | MAC | Background values |
Cd | 0.17 | 2.06 | 2.78 | 2.36±1 | 3.02 | 4.86 | 47.26 | -0.53 | -0.19 | 0.19 | 1-5 | 0.41 |
Co | 1.65 | 4.32 | 19.98 | 16.12±9 | 22.55 | 28.43 | 55.00 | -0.66 | -1.24 | 0.24 | 20-50 | 11.2 |
Cr | 37.27 | 58.19 | 178.90 | 151.56±70.16 | 205.24 | 255.05 | 46.29 | -0.60 | -1.22 | 0.21 | 50-200 | |
Cu | 2.09 | 5.86 | 10.48 | 9.85±5 | 13.33 | 18.04 | 45.95 | -0.08 | -1.01 | 0.11 | 60-150 | 38.9 |
Fe | 2162.20 | 6201.24 | 25634.40 | 20973±11553 | 29016.30 | 38350.88 | 55.09 | -0.70 | -1.13 | 0 .24 | - | - |
Mn | 68.28 | 135.24 | 576.30 | 479.63±267 | 677.72 | 935.98 | 55.67 | -0.53 | -1.23 | 0 .19 | - | 480 |
Ni | 12.18 | 18.74 | 29.68 | 27.25±8 | 33.06 | 40.04 | 29.02 | -0.33 | -1.23 | 0.15 | 20-60 | 29 |
Pb | 0.00 | 17.11 | 31.30 | 28.31±16 | 40.13 | 60.90 | 56.28 | -0.06 | 0.84 | 0.10 | 20-300 | 27 |
Zn | 0.00 | 51.18 | 116.91 | 100.01±41 | 126.76 | 170.30 | 41.15 | -0.62 | -0.50 | 0.18 | 100-300 | 70 |
Clay | 0.70 | 8.15 | 38.40 | 31.17±19 | 44.81 | 61.74 | 60.52 | -0.57 | -1.12 | 0.18 | - | - |
OM | 0.42 | 0.83 | 1.14 | 1.21±0.5 | 1.65 | 2.21 | 40.99 | 0.23 | -1.04 | 0.10 | - | - |
pH | 7.10 | 7.80 | 8.10 | 8.02±0.4 | 8.30 | 8.70 | 4.69 | -0.70 | -0.07 | 0.15 | - | - |
EC | 0.98 | 1.42 | 1.85 | 2.07±0.9 | 2.42 | 5.02 | 42.94 | 1.49 | 2.22 | 0.13 | - | - |
Omran17 found that the ranges of Co (70.5-113.8), Cu (7.7-280.3), Ni (61.3-88.7) and Zn (39.7-215.6) concentration (mg kg−1 soil) in the soils of Bahr El Baqar in the Eastern Nile Delta tended to be upper than the ranges found in this study, while the ranges of Cr (84.9-134.1) and Pb (16.4-52.4) tended to be low. The close difference between the lower quartile and the upper quartile points to the distribution and uniform source of trace metal in soil. Among nine trace metals, Cd had a relatively close difference between these quartiles. This may be due to more adding of P-fertilizers that contain a significant amount of Cd, about 12.45 mg kg−1 18. Moreover, cadmium may also be added to soils adjacent to roads28.
Among all trace metals, Fe has the largest mean value (20973 mg kg−1), while Cd has the lowest mean value of 2.36 mg kg−1. These results were consistent with the mean values of Fe and Cd found in agricultural soils in Egypt22. The mean concentrations of Cd, Co, Cr, Cu, Ni, Pb and Zn were below the maximum value of the range of Maximum Allowable Concentrations (MAC) mentioned by Kabata-Pendias29. The mean concentrations of Cu, Mn and Ni were lower than their corresponding background values. However, the mean concentrations of Cd, Co, Pb and Zn were exceeded their background values, which can demonstrate that the soils of the studied area are polluted with these metals by the anthropogenic activities23. In the same line, higher contents of Cd, Pb, Cu, Zn were observed in vegetable fields in China30. The main contribution of trace metals to agricultural soils is the use of chemical fertilizers, pesticides and organic fertilizers5,17,30. Thus, Cd, Co, Pb and Zn require more attention in the future, and continuous monitoring of the application of agrochemicals in the production of vegetables is recommended to decrease the probable ecological risks caused by these trace metals in soils of the Eastern Nile Delta.
The coefficient of variation (CV) measures the variability degree of the trace metal concentrations in soils. A high value of CV indicates that the contamination of soils by trace metal is produced mainly from human activities. A low CV value indicates that trace metal pollution is due to natural sources1,2,31. The high concentrations of metals coupled with high variation suggest that anthropogenic inputs may be their primary source in the study area. The highest CV values were recorded for Cd (47.26), Co (55.00), Cr (46.29), Cu (45.95), Fe (55.09), Mn (55.53), Pb (56.28) and Zn (41.15). The high CV value confirms that these metals have a wide concentration range and is an indication of the potential impact of human activities on them. Lower values of CV for Ni (29.02) confirm that human activities are less effective on it31.
The skewness values of Cu and Pb were near zero and came near a normal distribution, while the other metals had slightly negative skewness, indicating that the center of distribution is shifted to the right. The kurtosis of Pb was greater than zero, indicating the distribution has a towering shape. The kurtosis values of other elements were negative, indicating that the distribution has a flat shape. The results of the Kolmogorov–Smirnov test (p < 0.05) confirm that the concentrations of Cu, Ni and Pb are nearly normal distribution, while the concentrations of other metals did not follow a normal distribution.
Soil physicochemical parameters.
The descriptive statistics analyses of soil physicochemical parameters, i.e., clay, OM, pH and EC are given in Table 1 show that the study area had a wide range of contents of clay (0.70–61.74; median: 38.40) and EC (0.98–5.02; median: 1.85 dS m−1). The soils were moderately alkaline with mean pH of 8.02, with low OM (mean 1.21), as typically found in the Mediterranean environments. The CV values were highest for clay (60.52%), OM (40.99%) and EC (42.94%), indicating a non-homogenous distribution. The soil pH had a homogeneous distribution of variables, as indicated by the low value of CV of 4.69%. These results are in line with those found by Aitta et al.20.
The soil clay and pH observations skewed to the right, while the OM and EC observations skewed to the left. The Kurtosis results showed that clay and OM parameters revealed a flat shape, while the distribution of EC values exhibited a towering shape. The pH observations were subject to normal distributions. The K-S test suggested that clay values were not normally distributed, while OM, pH and EC values approached a normal distribution.
Multivariate statistical analysis.
Correlation analysis.
The correlation between different variables is presented in Table 2. The relationship that occurs between trace metals in the soils is usually due to parent material, the influence of pedogenic process, and the effect of human activities32. Very significant positive correlations were found between all trace metals except Cu, which indicated that Cd, Co, Cr, Fe, Mn, Ni, Pb and Zn could be largely derived from the same sources and spreading32. The Cu concentration was positively and significantly correlated with the concentration of Fe, Mn and Ni. Non-significant correlations of Cu with Cd, Co, Pb and Zn may be due to the various processes like external inputs and biological effects. A similar trend was reported in the Southwestern Nile Delta, where the associations between Cr, Cu, Ni, Pb and Zn and scavenger metals (Fe and Mn) had significant high associations23.
Table 2
The Pearson correlation coefficients between trace metals and soil properties. *p < 0.05, **p < 0.01
Variables | Cd | Co | Cr | Cu | Fe | Mn | Ni | Pb | Zn | Clay | OM | pH |
Cd | 1 | | | | | | | | | | | |
Co | 0.780** | 1 | | | | | | | | | | |
Cr | 0.763** | 0.957** | 1 | | | | | | | | | |
Cu | 0.097 | 0.178 | 0.202* | 1 | | | | | | | | |
Fe | 0.804** | 0.986** | 0.951** | 0.216* | 1 | | | | | | | |
Mn | 0.735** | 0.979** | 0.953** | 0.201* | 0.982** | 1 | | | | | | |
Ni | 0.665** | 0.942** | 0.905** | 0.202* | 0.948** | 0.952** | 1 | | | | | |
Pb | 0.576** | 0.664** | 0.668** | -0.017 | 0.579** | 0.601** | 0.496** | 1 | | | | |
Zn | 0.589** | 0.676** | 0.665** | 0.161 | 0.651** | 0.614** | 0.583** | 0.655** | 1 | | | |
Clay | 0.732** | 0.895** | 0.880** | 0.212* | 0.905** | 0.895** | 0.816** | 0.634** | 0.709** | 1 | | |
OM | 0.556** | 0.774** | 0.763** | 0.321** | 0.777** | 0.775** | 0.796** | 0.504** | 0.549** | 0.688** | 1 | |
pH | 0.288** | 0.410** | 0.483** | -0.059 | 0.396** | 0.402** | 0.356** | 0.497** | 0.510** | 0.531** | 0.345** | 1 |
EC | 0.371** | 0.400** | 0.455** | 0.017 | 0.387** | 0.416** | 0.359** | 0.456** | 0.240* | 0.320** | 0.251* | 0.365** |
The correlation among trace metal concentrations and soil properties is presented in Table 2. Clay % had very significant positive correlations with Cd (0.732), Co (0.895), Cr (0.880), Fe (0.905), Mn (0.895), Ni (0.816), Pb (0.634) and Zn (0.709), and had significant positive correlations with Cu (0.212). The high degree of correlation endorse that, the clay is acting as metal carrier and play a vital role in the distribution pattern of trace metals23. The positive relationships between all studied trace metals and organic matter percentage were very highly significant. The soil properties especially soil organic matter and clay particles content effectively adsorb trace metals23,40. Soil organic matter is one of the main soil properties with dual effects on trace metals mobility5. This may be because organic matter decomposition produces organic chemicals into soil solution that might act as chelates and raise the bioavailability of trace metals. However, the organic matter of the clay fraction might also decrease the trace metal bioavailability by forming stable complexes with humic substances or through adsorption33. Both pH and EC correlated positively with Cd, Co, Fe, Mn, Ni, Pb and Zn. Significant correlations were not found between Cu and soil pH and EC. A similar finding has been found by Aitta et al.20.
Factor analysis.
Factor analysis was carried out to identify the sources of pollution. The results of factor analysis using the varimax rotation method are presented in Table 3. Four factors with eigenvalues > 1.0 were selected for the retention data of trace metals in the studied area, which contributed to approximately 87.41% of the total variance. The first factor explains 50.43% of the total variance with negative loadings on all elements. The first factor was characterized by high loadings (≥ 0.72) of Co, Cr, Fe, Mn and Ni, and moderate loading (0.520) of Cd. Several studies found that most of these metals originated from parent material in agricultural soils32. Thus, this factor may represent a natural source. Factor 2 accounting only for about 11.50% of the total variability. It was heavily loaded (≥ 0.99) with Cu with positive loading. The increase Cu level in soil most probably due to the agricultural activities such as fertilizer and fungicide application34,35. Factor 3 explained 13.28% of the total variance. It mainly condensed the information of Pb with negative loadings (- 0.867). The higher content of Pb might have come from anthropogenic activities (2). The fertilizers and traffic emissions that can be the major source of the amount and distribution of Pb in agricultural soils31. Factor 4 exhibits 12.20% of the total variance with positive loading (0.852) on Zn. Organic manure and phosphate fertilizers cause a significant expansion in levels of Zn36.
Table 3
Factor loadings rotated matrix for trace metals.
Variables | Factor 1 | Factor 2 | Factor 3 | Factor 4 |
Cd | - 0.520 | 0.017 | - 0.227 | 0.217 |
Co | - 0.876 | 0.065 | - 0.287 | 0.250 |
Cr | - 0.847 | 0.093 | - 0.305 | 0.232 |
Cu | - 0.120 | 0.991 | 0.028 | 0.051 |
Fe | - 0.885 | 0.130 | - 0.186 | 0.233 |
Mn | - 0.915 | 0.087 | - 0.241 | 0.186 |
Ni | - 0.942 | 0.081 | - 0.128 | 0.200 |
Pb | - 0.346 | - 0.051 | - 0.867 | 0.278 |
Zn | - 0.377 | 0.078 | - 0.302 | 0.852 |
Eighen value | 4.54 | 1.035 | 1.20 | 1.10 |
Variance% | 50.43 | 11.50 | 13.28 | 12.20 |
Cumulative (%) | 50.43 | 61.93 | 75.21 | 87.41 |
The spatial and vertical distribution of trace metals levels are influenced by soil properties, for example, content of clay and OM. Industrial activity, agricultural practices and intensive urbanization are the primary anthropogenic sources of trace metal pollution. Cr, Cu, Ni, Pb and Zn are derived from uncontrolled utilization of phosphate fertilizers, and by atmospheric deposition from industrial activity and urban areas2,37, while Co is enhanced by the organic manure application to the agricultural soil51,55. Cu and Cr are produced from reactions involving Cu SO4, which mainly originated from disease prevention34,35. The results of factor analysis also indicate that Fe and Mn came from different potential sources. Agricultural soils can be polluted with Co, Cr, Ni and Zn that might be mainly originated from the application of fertilizers and organic manure17,29,51,55. Moreover, it can be polluted with Cr, Cu, Ni, Pb and Zn that result from industrial activity and atmospheric deposition24,55. Atmospheric deposition from industrial and urban areas represented 80.36% of Zn, 76.55% of Pb, 67.48% of Cu, 62.23% of Cr and 37.79% of Ni entering agrarian soil from anthropogenic activities37. It appears to be that anthropogenic and lithologic practices are the chief sources of Cd accumulation in soil. Cadmium is a metal marker of agricultural production activity, and it can be come from atmospheric deposition30,34.
Cluster analysis.
The results of the cluster analysis show that nine trace metals in the soils of the Eastern Nile Delta were divided into five categories (Fig. 1). The first cluster contained Cu. The second cluster contained Zn. Cluster third contained Pb. The fourth cluster contained Cd. The fifth cluster contained Mn, Fe, Co and B. The clustering results were in agreement with the factor analysis results.
Spatial distribution of trace metals.
The spatial distribution maps of trace metal contents in the vegetable field soils of the Eastern Nile Delta are presented in Fig. 2. Generally, the concentrations of all studied trace metals were the highest in the northwest part of the region. The possible reason for this is that some vegetable fields are close to industrial regions and intensive traffic activity, and irrigated with wastewater, which leads to the higher soil trace metal contents. Additionally, very high values of Co, Cr, Cu, Fe, Mn and Ni were also recorded in the central part. The sites with the lowest concentration of most trace metals are located in the southeast part of the region, which suggests that vegetable soils in this part might not have been affected by industrial activities. The distribution pattern of the Cd and Cu are mainly similarly distributed over the study region. This suggests the primary role of agricultural activities such as fertilizer and pesticide application in the vegetable fields as the main pollutant sources.
Pollution assessment of trace metals.
The EF values for the studied metals are presented in Fig. 3. The EF values indicated that Co, Cu, Fe, Mn and Ni < 2, showing no enrichment. The metals of Cr, Pb and Zn indicated moderate enrichment with EF mean of 4.856 4.773 and 3.874 respectively. Khalifa and Gad23 and Abou El-Anwar21 found similar results. The maximum EF values was 22.687 for Cd and consequently signifying very high enrichment. The EF values < 2 point to the trace metal is completely come from a geological origin, but the EF values > 2 indicate that the trace metal possibly derives from anthropogenic activities38.
The values of geo-accumulation index (Igeo) for the studied metals are illustrated in Fig. 3. Generally, the positive Igeo values indicate that the metal contamination is related to anthropogenic activities39. The mean values of Igeo increased in the order of Fe (0.0002) < Mn (0.007) < Zn (0.0450) < Cu and Ni (0.046) < Co (0.123) < Pb (0.145) < Cd (2.047) < Cr (3.114). The range of Igeo values for individual metal is as follows: Cr (0.039 - 0.059), Cd (5.705 - 5.068), Co (0 - 0.169), Cu (0.016 - 0.062), Fe (0.00016 - 0.00022), Mn (0.005- 0.008), Ni (0.035- 0.052), Pb (-0.012 - 0.198) and Zn (0 - 0.052). Based on the mean values of Igeo, Cr had trace contamination, and Cd had moderately to heavily contaminated. However, the other metals showed uncontaminated to moderately pollution load. This result is consistent with one previous review on agricultural soils, which shown that the Igeo values of Cr and Cd were above 16,23,34.
The computed results of the contamination factor (CF) for studied trace metals are presented in Fig. 3. The mean CF values of trace metals decreased in the following order: Cd (7.851) > Cr (1.684) > Pb (1.416) > Zn (1.053) > Co (0.848) > Mn (0.564) > Fe (0.444) > Ni (0.401) > Cu (0.219). The mean CF value for Cd indicated a very high contamination level (CF > 6), while the mean CF values for Cr, Pb and Zn showed a moderate contamination level (1 < CF < 3), and the mean CF values for Co, Mn, Fe, Ni and Cu pointed to a low contamination level (CF < 1). CF show minor similarity with Omran’s17 study in the soils of Bahr El Baqar in the Eastern Nile Delta, Shokr et al.26 in the soils of the middle Nile Delta, and Abou El-Anwar21 in the soils of the Upper of Egypt.
Health risk assessment.
The values of HQ, HI, and CR related to nine metals for both adults and children are summarized in Table 4. The HQ values of different metals through three pathways reduced in the order of ingestion > dermal absorption > inhalation. This result implies that ingestion of soil particles is the major route for trace metals that were adverse to human health. Previous studies have obtained similar results10,35. The HI values of soil trace metals for both adults and children were far lower than the safe level (HI ≤ 1) and reduced in the following order: Zn > Cr > Mn > Fe > Cd > Pb > Ni > Co > Cu. These results indicated that the non-carcinogenic threat for children and adults is relatively light across the vegetable field soils of the study area.
Table 4
Health risks of trace metals in vegetable soils from the Eastern Nile Delta.
Metals | HQ ing | HQ derm | HQ inh | HI | CR |
Adults | Children | Adults | Children | Adults | Children | Adults | Children | Adults | Children |
Cd | 1.61E-03 | 3.61E-03 | 1.29E-03 | 8.43E-03 | 3.44E-05 | 8.30E-05 | 2.93E-03 | 1.21E-02 | 2.17E-09 | 5.23E-09 |
Co | 5.52E-04 | 1.24E-03 | 5.51E-06 | 3.61E-05 | 4.12E-04 | 9.95E-04 | 9.70E-04 | 2.27E-03 | 2.31E-08 | 5.57E-08 |
Cr | 3.46E-02 | 7.75E-02 | 1.38E-02 | 9.04E-02 | 7.74E-04 | 1.87E-03 | 4.92E-02 | 1.70E-01 | 9.30E-07 | 2.24E-06 |
Cu | 1.69E-04 | 3.78E-04 | 4.49E-06 | 2.94E-05 | 3.58E-08 | 8.64E-08 | 1.73E-04 | 4.07E-04 | - | - |
Fe | 2.05E-02 | 4.60E-02 | 1.64E-04 | 1.07E-03 | 3.83E-06 | 9.24E-06 | 2.07E-02 | 4.70E-02 | - | - |
Mn | 7.14E-03 | 1.60E-02 | 1.42E-03 | 9.28E-03 | 4.90E-03 | 1.18E-02 | 1.35E-02 | 3.71E-02 | - | - |
Ni | 9.33E-04 | 2.09E-03 | 2.96E-05 | 1.94E-04 | 1.93E-07 | 4.66E-07 | 9.63E-04 | 2.28E-03 | 3.34E-09 | 8.07E-09 |
Pb | 5.54E-03 | 1.24E-02 | 2.95E-04 | 1.93E-03 | 1.17E-06 | 2.84E-06 | 5.84E-03 | 1.43E-02 | 1.65E-07 | 3.70E-07 |
Zn | 2.28E-04 | 5.11E-04 | 9.11E-06 | 5.97E-05 | 4.87E-08 | 1.18E-07 | 9.42E-02 | 2.85E-01 | - | - |
The HI and CR values for adults were lower than values for children. Previous studies reported that children experienced higher hazards by trace metal contamination than adults35. The CR of Cd, Cr, Ni and Pb were low than the safe value (1×10-6) and had no risk40. The CR values for Cr were between 1×10−6 and 1×10−4 for children, which indicate that a lower but elevated carcinogenic risk. Therefore, children have much more chances of carcinogenic risk from Cr exposure in the study area than adults. Similar results were found by Song et al.41, Mo et al.7 and Liu et al.32 who reported that Cr posed a significant carcinogenic risk. Moreover, previous studies have revealed that human exposure to low Cr concentrations for long-term can cause poisonous and cancer-causing impacts in people42. Also, Zhao et al.35 and Liu et al.32 reported the Cr in soil caused a significant carcinogenic risk to adults and children. Thus, particular attention should be paid to Cr pollution. Risks associated with non-carcinogenic and cancer results from the investigation region were overall lower than those found in the rural areas irrigated with wastewater in the Nile Delta40.