Physico-chemical properties of the Bahr El Baqar soils
The pH, CEC, clay and organic matter contents (OM) are the principal soil characteristics that determine the capacity to retain heavy metal pollutants. The average of pH in Bahr El Baqar soils is about 8.00, because of the presence of calcareous parent material (Table 2). The pH of the soil solution maintained at alkaline condition showed low mobility of all heavy metals. It could be attributed to the presence of carbonates at a high concentration. The high pH value measured in soils potentially limits the risk of metal mobilization. The relatively high content of OM (1.6 %) is mainly related to the high organic matter flux to the soil due to direct discharge of domestic and industrial wastewaters. Soil organic matter is a key for sorbing phase of metals. The dissolution of humic acid at high pH is responsible for the dissolution of Cu and Pb from the soil. Organic matter is important for the retention of metals by soil solids, thus decreasing mobility and bioavailability.
The summary statistics results indicate that the mean value of soil Pb, Cu, Cr and Zn concentrations is 36.64, 65.70, 106.96, and 90.56 mg kg−1, respectively and the range between the minimum and maximum values is 27.98, 272.58, 49.14, and 175.91 mg kg−1, respectively, which is large. There are great variations and high skewness for Cu and Zn (2.27 and 0.76). The CV % values reflect the mean variation of each sampling site in the population. The order of the CV %s for each element, from high to low, was Cu > Zn > Fe > Mn > Co > Cr > Ni > Pb > Cd. This result showed that the variation of Cu and Zn in the soil was larger than other metals.
The mean concentration of metals at selected locations and their world surface rock average is given in Table 3. The mean value of soil Cu, Ni, Pb, Cd, Co and Cr concentrations is 65.70, 73.22, 36.64, 14.69, 89.72, and 106.96 mg kg−1, respectively, which is much higher than the US EPA threshold level (16, 16, 20, 0.6, 23, and 26 mg kg−1, respectively) presented in Table 3. This result shows that Pb, Cd, Cu, Co, Cr and Ni concentrations in Bahr El Baqar soil are higher than those in the reference soil. These levels are far above the average concentrations in the earth’s crust (Hasan 2007; Yobouet et al. 2010) and the threshold concentration of European Union Standards. Table 3 indicates these soils are heavily polluted with heavy metals that are part of the most dangerous industrial and municipal waste (Hasan 2007). The ranking order of mean values of the heavy metals in the Bahr El Baqar soils followed the sequence: Cr > Zn > Co > Ni > Cu > Mn > Fe > Pb > Cd.
Multivariate statistical analysis
Multivariate analysis (i.e., Principal component analysis; and correlation) has been proved to be an effective tool for providing suggestive information regarding heavy metal sources and pathways (Hu et al. 2013).
Table 4 shows a factor analysis, which was, performed on raw data in an attempt to further clarify the major controlling factors that determine the heavy metal’s distribution in the Bahr El Baqar soils. Four of the factors, which account for most of the variability in the 17 variables, were obtained. Four factors have been extracted which had eigenvalues greater than or equal to 1. Together they account for 78.11 % of the variability in the original data (Table 4). The types of factoring have been selected as principal components. Metal loadings of the factors have been given in Table 4. Factor1 accounts for 47.24 % of the variability in the original data. The first group of variables can be described as an anthropogenic assemblage composed of mainly Fe, Co, Cd, Cu and OM. This factor reflects the binding of heavy metals to organic matter. Factor 2 accounts for 13.65 % of the variability in the original data. The second group of anthropogenic variables composed of mainly Ni, Cr and Pb. Factor 3 accounts for 10.72 % of the variability in the original data and explains metal sorbtion pools: carbonates, alumino silicates (clay minerals), pH and CEC. The first two of them resulted from terrestrial sources. Both of them are conservative components and they lose some trace metal contents via resuspension by the winds. Factor 4 accounts for 6.49 % of the variability in the original data and is composed of mainly Mn and Zn.
By applying the Pearson (parametric) rank order correlations (Table 5), the results revealed that the Fe is well correlated with Mn and moderately correlated with Cu and Zn (r = 0.819,0.488, 0.471 respectively). Furthermore, there is a good correlation between Ni and Co (r = 0.966). There are significant correlations between Cu and each of Zn, Pb, (r = 0.678, 0.544, respectively) as shown in Table 5. This result revealed that these metals have the same source of contamination. The results in Table 5 also showed that Ni, Fe, Mn, Cu, Zn and Co are positively correlated with OM, pH and clay, CaCO3, and CEC and of course negatively correlated with sand. Clay highly correlated with Fe, Mn, Cu, and Zn (0.773, 0.658, 0.786, and 0.709 respectively).
Quantification of heavy metals accumulation
Table 6 shows the results of contamination factor (CF). As shown in Table 6, average CF values for heavy metals have an order Cd > Cr > Co > Ni > Cu > Pb > Zn > Mn > Fe, suggesting that soil samples was extremely high enrichment with Cd, while Pb exhibit significant enrichment. In contrast, the rest of the metals show moderate or minimal enrichment in the study area. With respect to specific sites, high CF values for Cd (e.g., 31.79 were found in samples 10, 18, and 23. High CF values for Cr (5.16) were found in sample 5. High CF values for Co (4.95) were found in samples 6 and 19. However, for Cu, most of the locations have moderate to considerable contamination except samples of 26 and 27 which very highly contaminated. These locations were located at the downstream and continuously receive a vast amount of wastewater and other wastes of the city. Results in Table 6 show the CF values of most of the Mn, Zn, and Fe metals in the study area, which are low degree of contamination. Nevertheless, CF values for metals like Ni, Cd and Cr shows considerable degree. This is due to the influence of external discrete sources like industrial activities, agricultural and other anthropogenic inputs. Only, Cd shows a high degree of contamination.
Table 7 presents the geo-accumulation index for the quantification of heavy metal accumulation in the study area. The I-geo grade for the study area varies from metal to metal and location to location (across metals and locations). Fe remains in grade 0 (unpolluted) in all locations suggesting that the study area soils are in background value with respect to this metal. The I-geo for Mn and Zn attain grade 0 in few locations (unpolluted), while, attain in grade 1 in other soils which indicates that these soils were slightly polluted by Mn and Zn. I-geo index showed that most of (Ni, Co, Cr) heavy metals are in grade 3 (Table 7). Pb and Cu are in grade 2, however, Cd is in grade 6. This suggests that the soils of Bahr El Baqar are having background concentrations and these elements are practically changed by anthropogenic influences, while the concentration of Cd, Ni Co, and Cr exceeded the average standard value. These dangerous metals may be derived from industrial waste and gasoline additives used, in the factories and cars (Mwamburi 2003). These elements may also be derived through corrosion of the numerous abandoned launches along the drain and agricultural activities.
Table 8 shows different combined indices for heavy metal accumulation in the study area. DC values characterize a very high pollution for all of the Bahr El Baqar soils, reflecting the changes in soil occupation and the intensity of economic activities. The soils no. 26 and 18 is the most contaminated soils which shows the highest DC value of all studied areas (55.41 and 52.61) which classified as very high. The soils no. 31 and 17 show lowest DC values (37.54 and 37.65). Pollution severity and its variation along the sites was determined with the use of pollution load index. This index is a quick tool in order to compare the pollution status of different places. The values of the Pollution Load Indexes (Table 7) were found to be generally polluted (>1) in all the studied soils. The difference in indices results due to the difference in sensitivity of these indices towards the soil pollutants (Praveena et al. 2007). These confirmed that Bahr El Baqar drain is facing probable environmental pollution, especially with dangerous heavy metals (Pb, Cd, Co, Cr and Ni) which result from increased rate of non-treated industrial waste which are discharged to Bahr El Baqar drain. If this combined index (MPI) is above 1, the concentrations of trace metals would be considered elevated and ecosystem could be regarded as “polluted”.
The potential ecological risk index
Table 9 summarizes the results for PERI calculation at the studied areas. CF increment generates an equal increase of DC, since DC = Σ CF, but PERI increment will depend on which metal has this higher CF because it is specific for each one which were classified in decreasing order of toxicity (Cd = 30 > Cu = Pb = 5 > Cr > Zn = 1).
As a predict model, the PERI is dependent on calibration with bio-indicators in order to evaluate its effectiveness for potential risk of soil heavy metals. The PERI application in Bahr El Baqar ecosystems was successful, demonstrating that the environmental variables used in the algorithm proposed by (Hakanson 1980) are the main integrator parameters of biogeochemical processes in this ecosystem. In addition, the relationship among these variables shows a logical synthesis of biogeochemical processes that influence metal behavior in the soils. In this study, spatial distribution of Cd in all study areas was investigated. Cd was chosen because it represents up to 90 % of PERI (Table 9) values for almost all study areas, being representative of metal bioavailability and their risks in these areas.
Spatial distribution of soil heavy metals
Table 10 and Fig. 3 list the cross validation and fitted parameters results in examining the validity of the different models and parameters of semivariograms (e.g., Cd parameters, single and combined index). Cross-validation was used for comparing the interpolation methods. Three indices were calculated from the measured and interpolated values at each validation sample site. The mean error (ME), the mean absolute error (MAE) and the root mean square error (RMSE) are determined from the measured values. The ME is a measure of the bias of the interpolation, which should be close to zero for unbiased methods, and the MAE as well as RMSE are accuracy measures of the interpolation, which should be as small as possible for accurate interpolation.
For the DC index the best fit is the K-Bessel model (SME 0.0051) and Exponential model for MPI and PLI index with a 0.0034 and −0.0032 SME respectively which is closest to zero. The RMSS values for MPI and PLI are 0.9625 and 0.9603, respectively, which are closest to 1. The best fit for the mean cadmium parameter is the Pentaspherical model (SME −0.0057) and Rational Quadratic model for cadmium (by CF) with a −0.0076 SME which is closest to zero. The best fit for the PERI index is the Pentaspherical model (SME −0.0037). When the average estimated prediction standard errors are close to the root-mean-square prediction errors from cross-validation, you can be confident that the prediction standard errors are appropriate (Johnston et al. 2001).
Table 10 lists cross validation results in examine the validity of the fitting models and parameters of semivariograms for heavy metals. Table 10 shows the most suitable models and their prediction error values for each parameter. Table 10 also shows that for the different parameters, different models may give better results. For heavy metals, RMSS range from 0.9018 to 0.9862. Figures 4 and 5 shows the spatial distribution maps of different indices (e.g., DC, PLI, MPI, PERI….) in the study area based on these interpolations.
From the point of integrated assessments view, the ecological risk of heavy metals in the surface soils for the study area indicating a high contamination risk which was dominated by Cd (Fig. 5).
Recall, the principal objective of this study was to assess soil contamination in Bahr El Baqar agricultural area. The discharge of industrial, agricultural and municipal wastewaters in Bahr El Baqar drain has contaminated the environment surrounding Bahr El Baqar areas. The fieldwork was conducted during the dry season in order to obtain maximal heavy metal concentration from the soil. (Yahaya et al. 2009) confirmed that the concentration of heavy metal in soil is higher in the dry season than in a rainy season because heavier metals are lost in the soil due to run-off and infiltration in a rainy season. The accumulation of heavy metals in these soils is a serious concern due to their persistence and toxicity. Thirty-four samples were collected at 0-20 cm and evaluated for heavy metals using geoaccumulation index (I-geo), enrichment factor (EF) and contamination factor (CF), pollution load index (PLI), and metal pollution index (MPI), etc.
The concentration of cadmium ranges from 10.27 to 19.07 mg/kg with a mean concentration of 14.69 mg/kg. The calculated geo-accumulation index (I-geo) for cadmium indicates that the soils are extremely contaminated and very high contamination degree with the CF index. Cd is regarded as one of the most toxic trace elements in the environment. Cd is higher in the study area because of the uses of phosphate fertilizers, irrigation by untreated wastewater of Bahr El Baqar darin. Therefore, the water from this drain is not suitable for agricultural purposes. Manganese ranged from 13.73 to 98.73 mg/kg. The mean was 58.98 mg/kg. The calculated I-geo value gave a value that indicates uncontaminated to moderate and low contamination degree with the CF index. Manganese can be adsorbed onto soil depending on organic content, pH, grain-size and cation exchange capacity of the soil and this can be exemplified by the strong positive correlation with organic matter (<0.01 level). The concentration of copper varied from 7.73 to 280.30 mg/kg with an average value of 65.70 mg/kg. A moderately positive, high correlation with lead and Zinc was established (<0.01 level). Although zinc remains adsorbed to soil, leaching has been reported at waste disposal sites. The lower concentrations of the Zn might be due to the continuous removal of heavy metals by the crops grown and due to leaching of heavy metals into the deeper layer of the soil and in to the ground water. Chromium concentration ranges from 84.92 to 134.06 mg/kg with a mean value of 106.96 mg/kg. Chromium may be lower in some sites due to the continuous removal of heavy metals by the crops and due to leaching of heavy metals into the deeper layer of the soil and to the ground water. No correlation was found with other metals and its concentration falls within the contaminated. Nickel measured concentrations (61.32–88.73 with mean 73.22) are above the average reference abundance in an uncontaminated soil. A moderate positive correlation with Zn was noted at <0.05 level. The results show that lead concentration ranged from 24.42 to 52.40 mg/kg with a mean concentration of 36.64 mg/kg. Though there was an observed strong correlation with Cu (<0.01 level), its concentration is within the level of uncontaminated soil. Zn concentration varied between 39.67 mg/kg and 215.58 mg/kg with an average concentration of 90.56 mg/kg. These values are found to be low the average abundance for an uncontaminated soil. The I-geo concentration lies below the range of uncontaminated soil. Zinc had very strong positive correlation with Cu and Pb (<0.01 level) and moderate positive correlation with Ni (<0.05 level). The irrigation by Bahr El Baqar wastewater and uses of agricultural fertilizers led to the increasing the Ni concentrations. Cobalt concentration ranged 70.51–113.82 mg/kg with a mean value of 89.72 mg/kg. The measured concentrations of Co are in range of contaminated soil. The irrigation of agricultural lands with untreated water led to the accumulation of Co in the soils. Among significant variables that control the distribution and enrichment of heavy metals in the soils are soil pH, grain size of the soil, amount of organic matter in the soil and the cation exchange capacity (Huang and Lin 2003; Lin et al. 2002). The soil pH is generally high (7.58–8.82) while CaCO3 (0.61–29.2) characterize the top soil and these condition enhances the precipitation and bio-accumulation of heavy metals in soil. Heavy metals have a strong affinity for organic content, clay and silt fraction because of their high cation exchange capacity (Bodur and Ergin 1994; Zonta et al. 1994). The topsoil comprises organic content (0.22–3.55), clay and silt fraction.
Four principal components (Eigenvalues >1) emerged accounting for 78.11 % of the cumulative variance from the principal component analysis. The first principal component (PC-1) loading with 47.24 % variance showed higher loading for Cd, Cu, Co, Fe and organic matter. The second principal component (PC-2) has loading 13.65 % of the total variance, had high loading. The pH of the soil could have contributed to Pb and Zn retention in the soil, resulting in low mobility of the metals (Amadi et al. 2012; Yoshida et al. 2002). The third principal component (PC-3) explains 10.72 % of the total variance and consists of silt, clay, CaCO3, CEC, OM and pH. The fourth principal component (PC-4) has a moderate loading for Mn, and Zn, which accounts for 6.49 % of the total variance. Industrial activities domiciled in the area may be responsible for the presence of Mn and Zn. The physico-chemical properties of clay could have encouraged their availability in the soil. By applying the Pearson rank order correlations, Fe is well correlated with Mn and moderately correlated with Cu and Zn. There is a good correlation between Ni and Co. There are significant correlations between Cu and each of Zn and Pb. Ni, Fe, Mn, Cu, Zn and Co are positively correlated with OM, pH, clay, CaCO3, and CEC and of course negatively correlated with sand. These results revealed that these heavy metals in the soils have the same source of contamination, which is Bahr El Baqar drain.