Changes in pH under different cultivation strategies
Compared with the initial soil pH (6.1 ± 0.2), the pH in the S. alfredii monoculture (5.8 ± 0.3) and C. arietinum L. (6.3 ± 0.2) monoculture did not vary significantly, while the intercropping treatment (5.5 ± 0.2) had a significantly lower soil pH than the initial value. The change in the pH of 0.6 units in the intercropping system could mobilize soil metals, especially Cd (Burges et al., 2017; Strobel et al., 2005). The influence of soil pH on Cd desorption was evaluated by Strobel et al. (2005), who suggested that the distribution coefficients of Cd had log-linear relationships with pH, which can predict the solubility of Cd during soil acidification. Based on Eq. 4 in that study, Kd (Cd) was reduced from 3,419.79 kg− 1 at the beginning of the experiment to 2,523.48 and 1,862.09 in the S. alfredii monoculture and intercropping systems, respectively, indicating that the change in soil pH was a crucial factor for Cd and Cu mobilization, particularly in the intercropping system.
DOM characterization
The soil DOM level in the soil before transplantation was 136.1 ± 10.2 kg− 1. The S. alfredii monoculture, C. arietinum L. monoculture, and intercropping system obviously enhanced the content of DOM in the soil by 27.3%, 16.9%, and 62.9%, respectively (Fig. 1). Under different cultivation strategies, the S. alfredii monoculture and intercropping system resulted in significantly higher levels of DOM than the C. arietinum L. monoculture (Fig. 1).
The characteristics of DOM fractionations under different planting strategies are provided in Fig. 2. The highest levels of the hydrophilic fractions, including HIA, HIB, and HIN, were found in the intercropping system, which were significantly higher than those in the two monocultures. In addition, the concentrations of these three hydrophilic fractions in the S. alfredii monoculture were greater than those in the C. arietinum L. monoculture. Conversely, the levels of the hydrophobic fractions, such as HOA, HOB, and HON, in the C. arietinum L. monoculture were significantly higher than those in the S. alfredii monoculture and intercropping system, although the total DOM content in the former was significantly lower.
Moreover, the levels of the acid fractions, such as HIA and HOA, in the intercropping system were significantly higher than those in the two monocultures. However, the concentrations of the acid fractions in the S. alfredii monoculture were not significantly different from those in the C. arietinum L. monoculture (Fig. 2). The DOM was dominated by acid fractions, which constituted 79.3%, 83.2%, and 76.0% of the total DOM in the S. alfredii monoculture, C. arietinum L. monoculture, and intercropping system, respectively.
Metal extraction capacity of DOM
The metal (Cd, Cu, and Pb) extraction capacity of DOM was significantly stronger than that of deionized water, regardless of the planting strategy. The DOM from the intercropping system had the strongest Cd extraction capacity, followed by that from the S. alfredii monoculture and C. arietinum L. monoculture (Fig. 3). For Cu and Pb, the DOM from the intercropping system and S. alfredii monoculture had a similar extraction capacity, which was significantly higher than that from the C. arietinum L. monoculture.
Metal concentrations and fractions
As shown in Table 1, all the treatments did not vary the concentrations of Cd, Cu, and Pb in the rhizosphere significantly, compared with their initial concentrations. This result agrees with Martínez-Alcalá et al. (2016) who reported that N. caerulescens failed to change the content of Cd in soil during a 132-day phytoextraction.
Table 1
metal concentrations and fractions in the soils under different planting strategies
| | water soluble | Exchangeable | Carbonate | Fe–Mn oxide | organic matter | Residual | Total |
Cd | before | 0.26 ± 0.03a | 0.37 ± 0.05a | 0.82 ± 0.12a | 0.67 ± 0.09a | 0.49 ± 0.05c | 1.62 ± 0.18a | 4.23 ± 0.52a |
S | 0.17 ± 0.03b | 0.26 ± 0.03b | 0.75 ± 0.13a | 0.61 ± 0.12a | 0.72 ± 0.06b | 1.53 ± 0.21a | 4.04 ± 0.58a |
C | 0.29 ± 0.04a | 0.34 ± 0.05a | 0.79 ± 0.09a | 0.71 ± 0.13a | 0.42 ± 0.07c | 1.66 ± 0.26a | 4.21 ± 0.64a |
S + C | 0.13 ± 0.03b | 0.19 ± 0.02c | 0.77 ± 0.12a | 0.56 ± 0.09a | 0.87 ± 0.09a | 1.59 ± 0.22a | 4.11 ± 0.57a |
Cu | before | 0.65 ± 0.09c | 1.17 ± 0.13c | 3.97 ± 0.21a | 14.49 ± 1.81a | 24.55 ± 2.29b | 50.87 ± 5.69ab | 95.70 ± 10.22a |
S | 2.39 ± 0.25a | 2.87 ± 0.36a | 3.21 ± 0.33b | 10.02 ± 1.87b | 35.81 ± 3.32a | 40.36 ± 5.28b | 94.66 ± 11.41a |
C | 0.60 ± 0.05c | 1.29 ± 0.19c | 4.16 ± 0.37a | 12.96 ± 1.62ab | 26.07 ± 2.72b | 52.19 ± 6.07a | 97.27 ± 11.02a |
S + C | 1.62 ± 0.17b | 1.87 ± 0.27b | 3.62 ± 0.33ab | 10.12 ± 1.26b | 37.67 ± 3.91a | 40.38 ± 4.62b | 95.28 ± 10.56a |
Pb | before | 1.85 ± 0.11c | 3.29 ± 0.51c | 12.71 ± 1.63a | 21.65 ± 2.71a | 8.28 ± 0.59c | 35.78 ± 2.87a | 83.56 ± 8.81a |
S | 2.41 ± 0.16b | 7.32 ± 1.03a | 10.81 ± 1.39a | 17.53 ± 1.79ab | 12.56 ± 1.69a | 34.36 ± 2.61a | 84.99 ± 8.73a |
C | 2.03 ± 0.19c | 4.96 ± 0.52b | 11.92 ± 1.09a | 19.76 ± 1.66ab | 9.72 ± 0.61b | 35.31 ± 2.92a | 83.70 ± 6.99a |
S + C | 3.12 ± 0.29a | 8.31 ± 0.92a | 11.26 ± 1.56a | 15.65 ± 1.79b | 15.22 ± 1.63a | 31.56 ± 2.28a | 85.12 ± 8.47a |
Different letters are significant differences (p < 0.05) in concentrations and chemical forms of Cd, Cu, and Pb in the soils under different planting strategies (mg kg− 1) estimated using Fisher’s LSD post-hoc tests. |
Compared with the initial concentrations, the levels of water-soluble and exchangeable Cd decreased significantly in the S. alfredii monoculture and intercropping system but did not vary significantly in the C. arietinum L. monoculture at the end of the experiment (Table 1). Furthermore, the intercropping system had significantly lower levels of exchangeable Cd than the S. alfredii monoculture.
The concentrations of water-soluble and exchangeable forms of Cu and Pb increased significantly in the intercropping system and S. alfredii monoculture compared with the initial values, while the levels of these two chemical forms in the C. arietinum L. monoculture did not vary significantly (Table 1). The S. alfredii monoculture and intercropping system enhanced the proportion of organic matter Cd, Cu, and Pb from 11.0%, 25.7%, and 9.9% before transplantation to 17.8% and 21.2%, 37.8% and 39.5%, and 14.8% and 17.9%, respectively, at the end of the treatments (Table 1). The C. arietinum L. monoculture did not vary the proportion of organic matter Cd, Cu, and Pb under any planting strategy.
These results suggest that the hyperaccumulator can mobilize the non-hyperaccumulated metals in the soil by decreasing the soil pH and forming the metal-DOM complexes, as manifested in the significantly higher soluble forms of these metals in its rhizosphere soils, but does not absorb them. This argument is supported by the result of the stronger metal extraction capacity of DOM from the intercropping system.
The concentrations and proportions of the other metal chemical fractions did not vary significantly during treatment.
Phytoextraction potential and ecological risk
Compared with the C. arietinum L. monoculture, the intercropping system significantly reduced the shoot Cd concentration in C. arietinum L. by 72.7%, which could be attributed to the efficient consumption capacity of S. alfredii. However, S. alfredii with C. arietinum L. significantly increased the concentrations of Cu and Pb in the aerial parts of C. arietinum L. by 108.3% and 287.5%, respectively (Table 2). This result, which can be explained by the variations in the chemical fractions of metals under different planting strategies, suggests that intercropped S. alfredii decreases the edible safety of C. arietinum L.
Table 2
Dry weight and metal concentrations of S. alfredii and C. arietinum L. under different planting strategies
| | S. alfredii | C. arietinum L. |
| | roots | shoots | roots | shoots |
Biomass | S | 0.67 ± 0.09b | 8.56 ± 0.91b | - | - |
C | - | - | 3.52 ± 0.59a | 13.29 ± 1.08a |
S + C | 1.16 ± 0.12a | 12.17 ± 1.31a | 2.89 ± 0.43a | 9.87 ± 0.76b |
Cd | S | 24.3 ± 2.2a | 75.6 ± 4.9b | - | - |
C | - | - | 1.59 ± 0.12a | 0.11 ± 0.02a |
S + C | 27.1 ± 3.1a | 109.2 ± 5.9a | 0.76 ± 0.09b | 0.06 ± 0.01b |
Cu | S | 137.1 ± 9.2b | 29.1 ± 1.7b | - | - |
C | - | - | 7.32 ± 1.03b | 5.19 ± 0.68b |
S + C | 182.5 ± 11.3a | 33.5 ± 2.9a | 19.65 ± 2.26a | 10.81 ± 1.23a |
Pb | S | 276.3 ± 10.2a | 68.9 ± 6.1b | - | - |
C | - | - | 4.31 ± 0.62b | 0.16 ± 0.02b |
S + C | 249.6 ± 15.1b | 59.6 ± 5.2b | 6.03 ± 0.49a | 0.42 ± 0.05a |
Different letters are significant differences (p < 0.05) in dry weight (g) and metal concentrations (mg kg− 1) of S. alfredii and C. arietinum L. under different planting strategies (mg kg− 1) estimated using Fisher’s LSD post-hoc tests. |
The metal accumulation in plant tissues was used to evaluate the metal removal efficiency of the plant species under different planting strategies. The concentrations of metals multiplied by the biomass yield of the plant tissues can be regarded as the metal accumulation ability of the plant. The intercropping system significantly increased the dry weights of the roots and shoots of S. alfredii (Table 2) because intercropped C. arietinum L. can fix approximately 30 kg of nitrogen per ha from the atmosphere. However, the shoot dry weight of C. arietinum L. declined significantly in the intercropping system because intercropped S. alfredii drove the species to accumulate significantly higher amounts of Cu and Pb in its shoots. The variations in the oxygen free radical generation, chlorophyll levels, and antioxidant enzyme activities in crops under different planting strategies should be analyzed in future work.
Based on the dry weight and metal concentrations, 2.65, 0.01, and 2.72 mg of Cd, 1.36, 0.19, and 1.40 mg of Cu, and 3.10, 0.03, and 2.05 mg of Pb, were eliminated in the S. alfredii monoculture, C. arietinum L. monoculture, and intercropping system, respectively. The amounts of metals in the substrate were calculated as the soil mass (6 kg) multiplied by the differences between the analyzed Cd (4.2 mg kg− 1), Cu (95.7 mg kg− 1), and Pb (83.6 mg kg− 1) concentrations and their corresponding safe limits (0.3, 50, and 50 mg kg− 1 for Cd, Cu, and Pb, respectively). The required planting cycles for remediating the substrate were the quotients of the excessive amounts of metals in the soil and the corresponding metal accumulation in the plant tissues. Therefore, 9, 2,340, and 9 planting cycles are required to eliminate the superfluous Cd in the substrate in the S. alfredii monoculture, C. arietinum L. monoculture, and intercropping system, respectively. The phytoremediation effect of C. arietinum L. is negligible because it is a commercial crop rather than a hyperaccumulator. Although the Cd accumulation capacity of S. alfredii in the intercropping system was twice that found in the monoculture (1.36 mg vs. 0.66 mg), the remediation efficiency of these two plant strategies was nearly the same, because only half the number of S. alfredii plants was planted in the intercropping system. A suitable planting density would enhance the remediation effect.
The results indicate that 201, 1,443, and 196 planting cycles and 65, 6,720, and 98 planting cycles are required to remove excessive Cu and Pb from the soil in the S. alfredii monoculture, C. arietinum L. monoculture, and intercropping, respectively. The required time is unacceptably long, and the intercropping method did not significantly improve the remediation effect.