3.1 Effects of amendments on fresh weight
Table 4 showed the fresh weight of pak choi under different treatments. The application of sepiolite increased the fresh weight of pak choi by 38.1%, 20.8% under FWHC, NWHC, respectively. Because sepiolite alleviated Cd toxicity on pak choi. Liang et al.(2017) reported that sepiolite increase the total antioxidant capacity and nonprotein thiols contents, and reduce the MDA content of rice root to alleviate the stress of Cd in rice. In unamended soil, the application of goat manure increased fresh weight of pak choi by 34.9%-88.4% under FWHC. In SP-amended soil, the application of goat manure slightly decreased fresh weight of pak choi at different application rate under NWHC treatment, whilst significantly increased fresh weight of pak choi when applied at a concentration of 2.0%. The results indicated that the coupling effect of water management and organic fertilizer reduce the toxicity of Cd on plant growth. Our results are in the agreement with previous studied reported by Lahori et al. (2017) found that the combined application of tobacco biochar and mineral additives increased soil pH and dry biomass production of Chinese cabbage.
Table 4
Water Management
|
|
Treatments
|
|
CK
|
G1
|
G2
|
G3
|
FWHC
|
Unamended
|
23.51±1.84 e
|
31.71±2.19 c
|
44.30±3.35 a
|
28.19±0.83 d
|
SP-Amended
|
32.46±2.31 c
|
28.72±1.25 d
|
28.75±1.20 d
|
40.58±4.27 b
|
NWHC
|
Unamended
|
19.35±0.21 fg
|
21.51±1.81 ef
|
17.59±1.67 gh
|
16.44±0.88 h
|
SP-Amended
|
23.38±1.31 e
|
20.12±0.06 fg
|
16.59±1.33 h
|
21.45±0.28 ef
|
3.2 Effects of amendments on soil pH and DOM
Figure 1 showed the effect of goat manure on soil pH and DOM of unamended and SP-amended soil under NWHC and FWHC. The application of goat manure and sepiolite alone increased soil pH by 0.30-0.63, 1.0-1.1 respectively. Fard et al. (2011) found a negative relationship between pH and Cd sorption capacity of biosolids and a similar pattern is seen here. Soil pH was relatively low under NWHC treatment. As a consequence, available Cd extracted by DTPA remained at a low level under FWHC compared with NWHC treatment. The application of sepiolite decreased DOM by 31.5%-32.6%. Because, clay minerals, especially secondary mineral, tend to adsorbed organic matter to form stratified structure through ligand exchange, complexation, cationic bridge, hydrogen bond, and Van der Waals forces. Thus, the clay mineral increased thermal and chemical stability of organic matters and resulted in the rearrangement of organic microstructures(Kleber et al., 2007, Chasse et al., 2015). The application of goat manure increased DOM by 1.14-4.75, 1.16-2.9 under NWHC, FWHC treatment. Goat manure increased the content of humic acid and fulvic acid, because they are main components of DOM in soil. In SP-amended soil, soil DOM increased with the increase of goat manure, however, the changes of water contents had no effect on soil DOM.
3.3 Effects of amendments on Cd contents of pak choi
Figure 2 showed the effect of goat manure in unamended and SP-amended soil under NWHC and FWHC treatment. Compared to NWHC treatment, Cd content in shoot and root decreased by 43.9%, 28.0% under FWHC, respectively. When applied sepiolite alone, Cd content of shoot and root decreased by 37.3%-75.3%, 36.5%-62.9%, respectively. When applied at the concentration of 0.5%, coupling effect of goat manure and NWHC reduced the accumulation of Cd in shoot of pak choi compared with the goat manure and FWHC treatment. When applied at the concentration of 1.0% and 2.0%, coupling effect of goat manure and FWHC reduced the accumulation of Cd in shoot of pak choi compared with the goat manure and NWHC treatment. Whether in the unamended or SP-amended soil, coupling effect of organic manure and water management have the same tendency.
3.4 Effects of amendments on available Cd
Figure 3 showed the effect of goat manure and water management on available Cd extracted by DTPA. Under NWHC, the application of sepiolite decreased DTPA-Cd by 74.5%. Sun et al. (2016) found that the addition of sepiolite was effective in immobilizing Cd in pot and field applied research, because sepiolite decreased the content of Cd extracted by TCLP. Under FWHC, the application of sepiolite increased DTPA-Cd by 79.2%, and the application of goat manure increased DTPA-Cd by 42.0%-68.9%.
3.5 Microbial community abundance and structural diversity
3.5.1 Soil bacteria alpha diversity
A total of 1,486,094 high-quality sequences and 1598 OTUs for the bacterial 16S gene from all 24 samples were obtained. The Good’s average coverage of all the samples exceeded 99.9%, indicating that the bacterial OTUs for each soil sample were well captured by this sequencing method. Generally, the Shannon and Simpson indexes can be used to evaluate soil bacterial community diversity, and the ACE and Chao1 indexes can be used to evaluate soil bacterial community richness (Shi et al., 2019). The larger the Chao1 index and ACE index, the higher the abundance of the microbial community. The greater the Shannon index, the higher the diversity of microbial community while the larger the Simpson index, the lower the diversity of microbial community.
The soil microbial alpha diversity is shown in Table 5. Compared with the control, the abundance-based coverage estimators (ACE) index, Chao1 index and Shannon index were significantly increased under the G, SP, and SG treatments, whereas the Simpson index had no significant difference between treatments. The results showed that the G, SP, and SG treatments increased the diversity and richness of soil bacteria, especially the simultaneous application of goat manure and sepiolite (SG). Previous studies showed that lime mixed with organic manure can further enhance specific bacterial communities (Ai et al., 2015; Martínez García et al., 2018). In addition, the bacterial diversity is high in neutral soils and low in acidic soils (Fierer and Jackson, 2006; Lauber et al., 2009). Chodak et al. (2013) reported that heavy metals had a negative impact on soil bacterial diversity. Thus, the increase in pH and the inhibition of the heavy metal activity may contribute to the improvements in the richness and diversity of bacterial communities in this study. The soils collected under the G, SP, and SG treatment had a much higher pH (Fig. 1) and a lower Cd bioavailability than CK, which may have caused the significant differences in microbial diversity.
Table 5
Treatments
|
ACE
|
Chao1
|
Shannon
|
Simpson
|
FCK
|
1477
|
1491
|
7.96
|
0.9872
|
FG
|
1533
|
1535
|
8.32
|
0.9905
|
FSP
|
1555
|
1558
|
8.34
|
0.9887
|
FSG
|
1498
|
1503
|
8.21
|
0.9889
|
NCK
|
1479
|
1486
|
8.38
|
0.9917
|
NG
|
1585
|
1588
|
8.66
|
0.9925
|
NSP
|
1576
|
1576
|
8.78
|
0.9922
|
NSG
|
1539
|
1544
|
8.66
|
0.9919
|
3.5.2 Composition of the microbiota
Bacteria are the most important decomposers in soil. The microbiota compositions at phylum, for the 8 soil samples were analyzed. Fig. 4A showed the bacterial composition of the soil under the different treatments and its frequency of bacterial phyla. The soil bacteria in the experimental field were mainly from 10 phyla. Ten types of bacteria accounted for more than 90% of the total bacterial count. Proteobacteria, Actinobacteria, Bacteroidetes, Gemmatimonadetes, Acidobacteria in soil were the dominant community of soil microbes in the current experiment, accounting for 38.46%, 25.75%, 9.47%, 11.05%, 7.10%, respectively. These phyla have also been reported as the dominant groups in heavy metal contaminated soils (Liu et al., 2019; Tipayno et al., 2018).
In general, toxic pollutants reduce the relative abundance of microbial communities in normal soils but increase the relative abundance of heavy metal-resistant microbial communities (Xu et al., 2018). According to reports, Proteobacteria has the vast majority of heavy metal resistance genes, followed by Actinobacteria and Bacteroidetes as shown by metagenomic sequencing (Li et al., 2020).
In this study, the abundance of proteobacteria, which accounted for the largest proportion of soil bacteria, significantly increased under the FG, FSG, FSP treatments compared with control. Besides, abundance of proteobacteria reached a peak in the FSG treatment, which probably attribute to the more nutrient-rich environment present and higher soil pH under the goat manure treatment (Wu et al., 2017; Huang et al., 2018).
Actinobacteria was the other highly represented phylum in soil. The frequency of actinobacteria significantly decreased with the application of amendments under FWHC, while slightly increased under NWHC. The reason might be the GM and SP amendment had an immobilization effect on the toxicity of heavy metals, thereby increasing the abundance of microbial communities that are not resistant to heavy metals. Gemmatimonadetes and Acidobacteria were basically stable between different treatments.
Figure 4B showed unweighted Pair-group Method with Arithmetic Mean (UPGMA) of the control and amended soil. The closer the samples and the shorter the branch length indicates that the species composition of the two samples are more similar. According to the genetic relationship between samples, all the treatment can be divided into three group. FCK and NCK represented the soil microbial environment without artificial disturbance. Under NWHC treatments, NG, NSP, and NSG indicated soil water capacity was another factor that affect the microbial diversity. The single immobilization treatments, for instance NG and NSP, were more closer than simultaneous application of SP and GM (refers to the NSG). There was similar findings among FG, FSP, and FSG treatments.
The similarity of species composition data among samples in control soil and amended soil was investigated by principal component analysis (PCoA). Fig. 4C showed comprehensive impact of water management and organic manure on the diversity of microbial communities based on the weighted Euclidean distance. This explained about 67.5% of the changes in the microbial community composition (first and second dimensions were 49.43% and 18.07%, respectively), indicating that relevant parameters such as soil moisture and organic manure had a significant impact on the microbial community. As shown in the PCoA analysis plots, the control and amended treatment could be easily separated, with CK results gathered together on the left side of the abscissa PC1, well separated from the NG, FG, NSP, FSP, NSG, and FSG treatments. Besides, the microbial communities in the normal water condition and water-saturated condition were independent of each other.
Furthermore, we could intuitively determine the relative dominance microbe through the comparison of absolute values of abundance with a heat map. Generally, one square represented one kind of bacteria (horizontal range) and its corresponding habitat (vertical range). The redder the square, the more dominant of abundance, and the bluer the square, the less prevalent of the abundance (Uddin et al., 2019). The heat map of the microbial genera after different soil treatments is shown in Fig. 4D. The bacterias were clustered according to the similarity of distribution in eight soil samples. Among the control treatments, the top five dominant bacteria genera (high to low dominance) were Armatimonadetes, Cyanobacteria, Gemmatimonadetes, Chloroflexi, Verrucomicrobia. Among the SP-amended treatments, the top five dominant bacteria genera (high to low dominance) were Firmicutes, Nitrospirae, Rokubacteria, GAL15, Elusimicrobia.
3.5.3 Relationships between bacteria community composition and environmental variables
Understanding the role of microbes in the solubility of Cadmium (Cd) is of fundamental importance for remediation of Cd toxicity. The present study aimed to identify the microbes that involved in regulating Cd solubility and to reveal possible mechanisms.
Previous studies have documented that soil microbes are sensitive to environmental changes (Fierer and Jackson, 2006; Paul, 2014; Zheng et al., 2016). but at the same time they are resilient and have the ability to recover its structure to initial state after the perturbation disappeared (Allison and Martiny, 2008; Sydow et al., 2016). The effect of changed soil physicochemical properties and the available Cd on microbial community composition was performed by RDA (Fig. 5). The results of bacterial community analysis showed that the first two axes of RDA explain 37.24% and 18.53% of the total variation in the data, respectively. The bacterial communities in the soils following the FG, FSG, NG, and NSG treatments were distinct from those in the control, being separated primarily by the first axis, which was positively correlated with pH and DTPA-Cd, negative correlated with DOM.
Figure 5 also showed the relationship between abundant genera and soil physicochemical properties. The results showed that soil microbes such as bacteroidetes, patescibacteria, and proteobacteria were correlated positively with pH, DOM, and DTPA-Cd. Acidobacteria was correlated positively with DTPA-Cd, and negatively with pH and DOM.
The relationship of soil physicochemical properties and the available Cd content to the bacterial community variation and significant correlation indicated that GM or SP mixed additives may indirectly affect the soil microbial community by changing soil physicochemical properties and soil available Cd content (Fig. 5). Soil microbial community of SP treatments were closer than G and SG treatments compared to control.