**3. Results**

The ECs of the Industrial estate was the highest (3.9 dSm−1; Table 1) and was significantly higher than those of the Khan Village (1.4 dSm−1) and Vehari road (1.5 dSm−1; *p* = 0.001, *p* = 0.019, respectively) sites, in which the two sites had significantly lower ECs than the Sameej abad site (3.5 dSm−1; *p* = 0.016). All other ECs comparisons did not significantly differ. The pHs of the sites were not significantly different, except for the Industrial estate pHs (8.4; Table 1), which was significantly higher than that of the Sameeja abad site (8.2; *p* = 0.005).

The ECiw of Industrial estate was also the highest (3.6 dSm−1; Table 1) and was significantly higher than those of the Khan Village (0.3 dSm−1) and Vehari road (0.3 dSm−1; *p* = 0.001, *p* = 0.017, respectively) sites, in which the two sites had significantly lower ECiw than the Shujabad site (2.8 dSm−1; *p* = 0.017). All other comparisons did not significantly differ. The pHiw of the sites were not significantly different with the exceptions of the Khan Village pHiw (7.2; Table 1), which was significantly higher than those of the Industrial

estate (6.8; *p* = 0.001) and Shujabad (6.9; *p* = 0.017) where the Shujabad site had significantly lower pHiw than that of the Soraj miani site (7.1; *p* = 0.017).

Overall, the study sites (soils) were significantly different in chemistry and heavy metal concentration (Table 1; ECs: *p* = 0.001; pHs: *p* < 0.001; Zn: 0.036; Cu: *p* = 0.002; Fe: *p* < 0.001; Mn: *p* < 0.001; Cd: *p* = 0.006; Cr: *p* = 0.003; Ni: *p* = 0.005; and Pb: *p* < 0.001). All respective irrigation waters were also significantly different, except for Cr (ECs: *p* = 0.019; pHs: *p* < 0.001; Zn: 0.031; Cu: *p* = 0.005; Fe: *p* = 0.012; Mn: *p* = 0.003; Cd: *p* = 0.004; Cr: *p* = 0.058; Ni: *p* = 0.002; and Pb: *p* = 0.013). It is to be noted that Mn is not exactly a heavy metal but toxic when absorbed or present in excessive amounts.

Soil heavy metal concentrations were significantly different (*p* < 0.05; Table 1) between sites with the highest concentrations of all metals at the Industrial estate (Zn = 3.9, Cu = 2.6, Fe = 17.3, Mn = 4.0, Cd = 4.8. Cr = 4.4, Ni = 1.7, and Pb = 3.4 mg kg−1), and the lowest at the Khan Village site (Zn = 1.8, Cu = 0.9, Fe = 6.8, Mn = 2.9, Cd = 0.7, Cr = 1.1, Ni = 0.4, and Pb = 1.4 mg kg−1). All other sites had a mix of higher or lower concentrations of the heavy metals compared to each other.

There were significant (*p* < 0.05) main (site/water and vegetable) and interactive (soil/water \* vegetation) effects on phytoextraction of all metal concentrations. At the site level, the Cu, Mn, Cd, and Ni phytoextraction values were significantly different, except for the Cu and Cd values at Khan Village that did not differ from the corresponding values at Shujabad road and Vehari road (both; *p* = 1.000, *p* = 0.073, respectively; Figures 1 and 2), Cd concentrations at Sameeja abad did not differ from those at the Khan village and Vehari road values (both *p* = 1.00, respectively Figures 3 and 4). The phytoextraction concentrations of Mn, Cd, and Ni also significantly differed between all four vegetables; however, the lettuce Zn concentration did not differ from that in brinjal (*p* = 1.00), the cauliflower Cu value did not differ from that in spinach (0.189), and the lettuce and cauliflower values did not differ from those in brinjal and spinach (both; *p* = 1.00; Figures 1 and 2)

There were significant (*p* < 0.05) main (site and vegetable) and interactive (soil \* vegetation) effects on vegetable transfer factor (VTF) for all metals, except no site/water \* vegetable interaction for the Cr VTF was found (*p* = 0.585; Tables 2 and 3). VTF comparisons revealed that spinach was the best phytoextractant with the highest phytoextraction values (spinach: Zn = 20.2 ( ±3.6), Cu = 12.3 ( ±8.3), Fe = 17.1 ( ±8.1), Mn = 30.3 ( ±16.5), Cd = 6.1 (±4.5), Cr = 7.6 ( ±5.7), Ni = 9.2 ( ±3.1), and Pb = 6.9 ( ±1.7)—Tables 2 and 3) followed by cauliflower and brinjal, while lettuce had the lowest VTF values for phytoextraction (lettuce: Zn = 20.2 ( ±4.3), Cu = 12.3 ( ±4.0), Fe = 17.1 ( ±2.0), Mn = 30.3 ( ±2.4), Cd = 6.1 (±4.7), Cr = 7.6 ( ±5.1), Ni = 9.2 ( ±2.6), and Pb = 6.9 ( ±1.3)—Tables 2 and 3) under main (site and vegetable) or interactive (site \* vegetable) effects. Values in brackets show the standard deviation.

The soil metal concentration–VTF correlations were significant (*p* < 0.05; Figures 3 and 4) for brinjal and lettuce for Zn; for brinjal and cauliflower for Cu; for brinjal only for Fe; all for Mn and Ni; for spinach for Cd and Cr; and for brinjal, lettuce, and spinach for Pb (Figures 3 and 4).

**Figure 1.** Zn, Cu, Fe, and Mn concentrations (mg kg−<sup>1</sup> dry weight) in brinjal (*Solanum melogena*) and spinach (*Spinacia oleacea*) harvested in April 2016, and lettuce (*Lactuca sativa* L.) and cauliflower (*Brassica oleracea*) harvested in December 2015 from the Khan Village, Vehari road, Shujabad road, Industrial estate, Soraj miani, and Sameej abad study sites.

**Figure 2.** Heavy metal (Cd, Cr, Ni, and Pb) concentrations (mg kg−<sup>1</sup> dry weight) in brinjal (*Solanum melogena*) and spinach (*Spinacia oleacea*) harvested in April 2016, and lettuce (*Lactuca sativa* L.) and cauliflower (*Brassica oleracea*) harvested in December 2015 from the Khan village, Vehari road, Shujabad road, Industrial estate, Suraj miani and Sameej abad study sites.

**Figure 3.** Relationships between the transfer factors (TF) of the vegetables (brinjal, cauliflower, lettuce, and spinach) and heavy metal (Zn, Cu, Fe, Mn, Cd, and Cr) concentrations (mg kg−<sup>1</sup> dry weight) in the respective soils of the Khan Village, Vehari road, Shujabad road, Industrial estate, Soraj miani, and Sameej abad study sites. A relationship is significant at *p* < 0.05. Only significant relationships are retained to sustain the brevity of the manuscript.

**Figure 4.** Relationships between the transfer factors (TF) of the vegetables (brinjal, cauliflower, lettuce, and spinach) and heavy metal (Ni and Pb) concentrations (mg kg−<sup>1</sup> dry weight) in the respective soils of the Khan Village, Vehari road, Shujabad road, Industrial estate, Soraj miani, and Sameej abad study sites. A relationship is significant at *p* < 0.05. Only significant relationships are retained to sustain the brevity of the manuscript.



\* Sites comparison: soil–plant transfer of Fe was significantly higher at the industrial estate site than that at the Khan Village site (F5, 24 = 2.51, *p* = 0.048, R<sup>2</sup> = 0.41). The Ni transfer factor at the Industrial site was significantly higher than all sites except the Shujabad road site (F5, 24 = 5.51, *p* = 0.003, R<sup>2</sup> = 0.61).


**Table 3.** Results of a two-way MANOVA of the soil–vegetable heavy metal transfer factor for six study sites in Multan, Pakistan.

Bold values are significant at the =0.05 level. R<sup>2</sup> shows the relationship between the soil and plant heavy metal concentrations.
