Effect of Arbuscular Mycorrhiza Fungus Diversispora eburnea Inoculation on Lolium perenne and Amorpha fruticosa Growth, Cadmium Uptake, and Soil Cadmium Speciation in Cadmium-Contaminated Soil

Cadmium (Cd) pollution has become aggravated during the past decades of industrialization, severely endangering human health through its entry into the food chain. While it is well understood that arbuscular mycorrhizal fungi (AMF) have a strong ability to regulate plant growth and Cd uptake, studies investigating how they affect soil Cd speciation and influence Cd uptake are limited. We designed a pot experiment comprising two AMF-inoculant groups (inoculation with Diversispora eburnea or no inoculation), three Cd concentration levels (0, 5, and 15 mg/kg), and two plant species (Lolium perenne and Amorpha fruticosa) to study the effect of AMF Diversispora eburnea on plant growth, Cd uptake, and Cd speciation in the soil. The results revealed that L. perenne exhibited higher productivity and greater Cd uptake than A. fruticosa, regardless of AMF D. eburnea inoculation. However, AMF D. eburnea significantly altered soil Cd speciation by increasing the proportion of exchangeable Cd and decreasing residual Cd, resulting in Cd enrichment in the plant root organs and the elimination of Cd from the polluted soils. Our experiments demonstrate that inoculating plants with AMF D. eburnea is an effective alternative strategy for remediating Cd-contaminated soil.


Introduction
With the rapid increase in the rate of industrialization, increasing amounts of waste gas, waste water, and waste residue that frequently contain heavy metals are being discharged from industries [1][2][3]. The heavy metal cadmium (Cd) has attracted considerable attention because of its strong mobility and high toxicity. Reports from multiple countries (including Japan and China) have shown that Cd enters the human body through the food chain, where it can cause renal dysfunction and other diseases [4][5][6]. Physical and chemical methods are commonly used in Cd remediation [7], but the costs involved in the processes and the adverse effects they have on soil biological properties have caused controversy, and their residues can cause secondary pollution [8,9]. Therefore, biotechnological solutions, such as phytoremediation, are favored because they are environmentally friendly and do not require the use of high-cost technology [10].
Arbuscular mycorrhizal fungi (AMF) are an important functional group of soil microorganisms that can form a symbiotic relationship with 80% of vascular plant species in terrestrial ecosystems [11]. Previous studies have shown that AMF play an active role in removing Cd from soil [12,13], as AMF mycelium significantly increase the absorption of mineral ions (such as phosphorus) and water in plants, which increases the plant biomass, Table 1. Initial parameters of the soil and sand used for the pot experiment (means ± SE, n = 3). MBC, microbial biomass carbon; MBN, microbial biomass nitrogen; A-P, available phosphorus.

Host Plants and AMF Inocula
Two plant species, L. perenne and A. fruticosa, were selected for use in this study. Their seeds were provided by the Muyang Gramineae Seed Technology Limited Company, China. The plant seeds were surface-sterilized in 75% alcohol (v:v) for 5 min before being rinsed five times with distilled water. They were then placed on wet filter paper in a Petri dish for germination. The AMF species D. eburnea used in the present study was supplied by the Bank of Glomeromycota in China, Beijing Academy of Agriculture and Forestry, China ([BGC] HK02C). The inoculum was obtained after six months of co-culture with corn (Zea mays L.) in autoclaved sand/vermiculite (1:1, v:v). It comprised a mixture of spores (40 spores g −1 inoculum), mycelia, root fragments, and soil.

Experimental Design
The experiment used a 2 × 3 × 2 factorial design with two plant species (L. perenne and A. fruticosa), three levels of Cd concentration (0, 5, and 15 mg/kg), and two AMF treatments (D. eburnea and a non-mycorrhizal control) in a completely randomized design. Each treatment had four replicates, yielding a total of 48 units (pots). Each plastic pot (upper diameter, 14 cm; lower diameter, 10 cm; height, 10 cm) was filled with 1 kg of soil. Before sowing the seeds, 30 g mycorrhizal inoculum (approximately 1200 spores) and soil were mixed and placed in a pot. For the non-mycorrhizal control treatment, 30 g of sterilized mycorrhizal inoculum was added.
A 10 mL AMF-free filtrate was applied to minimize differences in microbial communities between mycorrhizal and non-mycorrhizal pots. Thirty germinated seeds of L. perenne or ten germinated seeds of A. fruticosa were sown in each pot. After two weeks of germination, thinning was performed, leaving 15 seedlings for L. perenne and 5 seedlings for A. fruticosa. Seedlings were arranged under natural light and temperature conditions in a greenhouse and irrigated with tap water to maintain the soil at a water holding capacity of 70% during the growing period. The greenhouse temperature ranged from 18 • C to 34 • C during the day. For the first two weeks, pots received 10 mL of a full-strength Hoagland nutrient solution per week and were rotated weekly (Table S1) [35]. The greenhouse pot experiments began on 23 August 2020, and ended on 22 October 2020.

Harvest and Measurements
At harvest, the plant shoots and roots were harvested separately. Roots were thoroughly cleaned with tap water to remove attached soil particles and then carefully cleaned with deionized water. Shoots and one root subsample were dried at 70 • C for 72 h and weighed. The AMF infection rate was determined using the other root subsample, and root biomass was calculated using the ratio of fresh and dry root weights. To determine the Cd concentration in the plants, the dried shoots and roots were ground and digested with HNO 3 -HClO 4 (4:1, v:v). The digestion solution was fixed to 25 mL with 2% HNO 3 and filtered after acid removal, and the Cd content was determined using graphite furnace atomic absorption spectrometry (GF-AAS, Perkin Elmer, Waltham, MA, USA). Soil pH was measured using a pH meter in a 1:2.5 (w:v) suspension of soil and deionized water. The NH 4 + -N and NO 3 − -N concentrations were measured using dual wavelength spectrophotometry and the indophenol blue method, and quantified using SmartChem 200 discrete auto analyzer (AMS Group Westco, Rome, Italy). Available phosphorus (A-P) was extracted using 0.5 M NaHCO 3 at a pH of 8.5 and quantified using the SmartChem 200 discrete auto analyzer. The activities of soil alkaline phosphatase and neutral phosphatase were determined using respective assay kits (Solarbio, Beijing, China, BC0285, and BC0465  4 and HCl, high-purity, 0.5 h) fractions using Tessier's five step extraction method [36]. Then, the contents of different Cd fractions were determined using GF-AAS.
The AMF structure of the roots was stained according to the improved Kormanik, method [37]. The cleaned fresh root segments (1 cm slices, 50 fragments randomly selected for each sample) were decolorized in 10% KOH solution (92 • C, 30 min). Roots were acidified with 2% HCl solution at 20 • C for 5 min and stained with 0.05% acid fuchsin solution (92 • C, 20 min). After the root fragments were stained, slides were prepared for observation. The structures of AM fungi in roots, such as arbuscule, vesicle, hyphal coil and non-septate hypha, were observed in each sample. The AMF colonization rate per unit root length was calculated using the grid-crossing method [

Statistical Analysis
The two-way ANOVA was used to examine the effects of plant species, Cd, and their interaction on the D. eburnea colonization rate. The one-way ANOVA was used to examine the effects of AMF and Cd addition on soil properties, plant biomass, Cd uptake, and heavy metal tolerance for each plant species. Duncan's test was used to compare the means among different treatments when the homogeneity of variance was satisfied; otherwise, the data were log or square root transformed. When the variances remained unequal, Tamhane's T2 test was used as a post hoc test for multiple comparisons. A structural equation model (SEM) was also established using AMOS 24.0 to investigate the hypothetical approach associated with the effect of D. eburnea and Cd addition on the D. eburnea colonization rate and the bioconcentration factor for each plant species. Statistical analyses were performed using SPSS version 23.0, and the significance level was set at p < 0.05.

D. eburnea Colonization Rate
The colonization rates of D. eburnea ranged from 11.00% to 29.60% for L. perenne, and 19.13% to 64.28% for A. fruticosa ( Figure 1). The two-way ANOVA showed that the D. eburnea colonization rate was significantly influenced by the plant species and Cd, but not by their interaction (Figure 1). The addition of Cd had no effect on the root colonization rate of L. perenne ( Figure 1). However, when compared with the Cd-free (0 mg/kg) treatment, the root colonization rate of A. fruticosa in low-Cd (5 mg/kg) and high-Cd (15 mg/kg) soil was significantly decreased by 45.15% and 26.10%, respectively ( Figure 1).

Plant Biomass
Regardless of D. eburnea inoculation, the shoot biomass of L. perenne and A. fruticosa in the high-Cd treatment was lower than that in the Cd-free treatment (Figure 2a). Furthermore, in the absence of D. eburnean, A. fruticosa shoot biomass was significantly decreased in the low-Cd treatment. However, D. eburnea inoculation had no significant effect on shoot biomass of L. perenne or A. fruticosa in any of the Cd treatments ( Figure 2a).
When compared with Cd-free plants, the root biomass of L. perenne was significantly increased by 62.81% in the high-Cd treatment when D. eburnea was present, while the total biomass of L. perenne was significantly decreased by 43.13% in the high-Cd treatment in the absence of D. eburnea (Figure 2b Without D. eburnea inoculation, the root-to-shoot rate of L. perenne was significantly higher under the low-Cd treatment than under the control and high-Cd treatments. By contrast, with D. eburnea inoculation, the root-to-shoot rate of L. perenne was significantly higher under the high-Cd treatment than under the Cd-free or low-Cd treatments ( Figure  2d). However, Cd addition had no significant effect on the root to shoot rate of A. fruticosa, regardless of D. eburnea inoculation. D. eburnea inoculation significantly decreased the root to shoot rate of L. perenne by 27.31% under the low-Cd treatment, but significantly increased the root shoot ratio of L. perenne by 145.03% under the high-Cd treatment. Moreover, the root to shoot rate of A. fruticosa was significantly increased by D. eburnea inoculation by 58.21% and 77.00% under the control and low-Cd treatments, respectively (Figure 2d).

Plant Biomass
Regardless of D. eburnea inoculation, the shoot biomass of L. perenne and A. fruticosa in the high-Cd treatment was lower than that in the Cd-free treatment ( Figure 2a). Furthermore, in the absence of D. eburnean, A. fruticosa shoot biomass was significantly decreased in the low-Cd treatment. However, D. eburnea inoculation had no significant effect on shoot biomass of L. perenne or A. fruticosa in any of the Cd treatments ( Figure 2a).
When compared with Cd-free plants, the root biomass of L. perenne was significantly increased by 62.81% in the high-Cd treatment when D. eburnea was present, while the total biomass of L. perenne was significantly decreased by 43.13% in the high-Cd treatment in the absence of D. eburnea (Figure 2b,c). D. eburnea inoculation significantly increased the root and total biomass of L. perenne by 206.47% and 103.72%, respectively, in the high-Cd treatment (Figure 2b,c). Regardless of D. eburnea inoculation, Cd addition had no significant effect on the root and total biomass of A. fruticosa (Figure 2b,c).
Without D. eburnea inoculation, the root-to-shoot rate of L. perenne was significantly higher under the low-Cd treatment than under the control and high-Cd treatments. By contrast, with D. eburnea inoculation, the root-to-shoot rate of L. perenne was significantly higher under the high-Cd treatment than under the Cd-free or low-Cd treatments ( Figure 2d). However, Cd addition had no significant effect on the root to shoot rate of A. fruticosa, regardless of D. eburnea inoculation. D. eburnea inoculation significantly decreased the root to shoot rate of L. perenne by 27.31% under the low-Cd treatment, but significantly increased the root shoot ratio of L. perenne by 145.03% under the high-Cd treatment. Moreover, the root to shoot rate of A. fruticosa was significantly increased by D. eburnea inoculation by 58.21% and 77.00% under the control and low-Cd treatments, respectively ( Figure 2d).

Figure 2.
Effects of AMF and Cd addition on shoot biomass (a), root biomass (b), total biomass (c), and root to shoot rate (d) of L. perenne and A. fruticosa (Means ± SE, n = 4). Different letters indicate significant differences among treatments for each plant species according to Duncan's or Tamhane's T2 test at p < 0.05. −AMF, not inoculated with D. eburnea; +AMF, inoculated with D. eburnea; 0, no Cd added to soil; 5, 5 mg/kg Cd added to soil; 15, 15 mg/kg Cd added to soil.

Soil Parameters
Compared with the Cd-free treatment, the pH of L. perenne soil was significantly increased in the low-Cd treatment without D. eburnea inoculation, whereas that of A. fruticosa soil was not significantly affected by D. eburnea inoculation or Cd addition (Table 2).
When D. eburnea was not inoculated, the available phosphorus (A-P) concentration in L. perenne soil in the low-Cd treatment was significantly lower than that under Cd-free or high-Cd treatments. However, compared with the Cd-free treatment, the A-P concentration in L. perenne soil was significantly increased by 96.70% under the low-Cd treatment when D. eburnea was inoculated. The A-P concentration in A. fruticosa soil was significantly increased in the high-Cd treatment, with or without D. eburnea ( Table 2).
The high-Cd treatment significantly increased the NH4 + -N and NO3 − -N concentrations in L. perenne soil, compared to the low-Cd treatment after D. eburnea inoculation. However, Cd addition had no significant effect on the NH4 + -N or NO3 − -N concentrations in A. fruticosa soil, regardless of D. eburnea inoculation. D. eburnea inoculation had no significant effect on NH4 + -N concentration in L. perenne or A. fruticosa soils in any of the Cd treatments (Table 2). However, D. eburnea inoculation significantly increased the NO3 − -N concentration in L. perenne and A. fruticosa soils by 66.51% and 65.80%, respectively, under the high-Cd treatment ( Table 2).
Neutral phosphatase activity in L. perenne and A. fruticosa soils and alkaline phosphatase activity in L. perenne soil were not significantly affected by Cd addition or D. eburnea inoculation ( Table 2). Compared with Cd-free soil, the alkaline phosphatase activity in A. fruticosa soil was significantly increased by high-Cd treatment by 21.79% in the absence of D. eburnea, while it was significantly decreased in the high-Cd treatment by 24.35% in the presence of D. eburnea (Table 2). Alkaline phosphatase activity in A. fruticosa soil was significantly increased by D. eburnea inoculation under the Cd-free treatment, but was decreased by D. eburnea inoculation under the high-Cd treatment ( Table 2).

Soil Parameters
Compared with the Cd-free treatment, the pH of L. perenne soil was significantly increased in the low-Cd treatment without D. eburnea inoculation, whereas that of A. fruticosa soil was not significantly affected by D. eburnea inoculation or Cd addition (Table 2). When D. eburnea was not inoculated, the available phosphorus (A-P) concentration in L. perenne soil in the low-Cd treatment was significantly lower than that under Cd-free or high-Cd treatments. However, compared with the Cd-free treatment, the A-P concentration in L. perenne soil was significantly increased by 96.70% under the low-Cd treatment when D. eburnea was inoculated. The A-P concentration in A. fruticosa soil was significantly increased in the high-Cd treatment, with or without D. eburnea ( Table 2).
The high-Cd treatment significantly increased the NH 4 + -N and NO 3 − -N concentrations in L. perenne soil, compared to the low-Cd treatment after D. eburnea inoculation. However, Cd addition had no significant effect on the NH 4 + -N or NO 3 − -N concentrations in A. fruticosa soil, regardless of D. eburnea inoculation. D. eburnea inoculation had no significant effect on NH 4 + -N concentration in L. perenne or A. fruticosa soils in any of the Cd treatments (Table 2). However, D. eburnea inoculation significantly increased the NO 3 − -N concentration in L. perenne and A. fruticosa soils by 66.51% and 65.80%, respectively, under the high-Cd treatment ( Table 2).
Neutral phosphatase activity in L. perenne and A. fruticosa soils and alkaline phosphatase activity in L. perenne soil were not significantly affected by Cd addition or D. eburnea inoculation ( Table 2). Compared with Cd-free soil, the alkaline phosphatase activity in A. fruticosa soil was significantly increased by high-Cd treatment by 21.79% in the absence of D. eburnea, while it was significantly decreased in the high-Cd treatment by 24.35% in the presence of D. eburnea (Table 2). Alkaline phosphatase activity in A. fruticosa soil was significantly increased by D. eburnea inoculation under the Cd-free treatment, but was decreased by D. eburnea inoculation under the high-Cd treatment ( Table 2).

Soil Cd Fractions
Compared with the low-Cd treatment, exchangeable Cd was significantly increased and carbonate-bound Cd was significantly decreased by the high-Cd addition in soils of both plant species, whether D. eburnea was inoculated or not (Figure 3a,b). Residual Cd was significantly higher under the low-Cd treatment without D. eburnea than under the other treatments (Figure 3a,b). Under the low-Cd treatment, D. eburnea inoculation significantly increased exchangeable Cd by 3.73% in L. perenne soil and significantly decreased residual Cd by 1.56% and 3.90% in L. perenne and A. fruticosa soils, respectively (Figure 3a,b).

Plant Cd Concentration and Accumulation
The shoot Cd concentrations of both plants significantly increased with an increase in the soil Cd concentration, regardless of D. eburnea inoculation (Figure 4a). D. eburnea inoculation significantly decreased the shoot Cd concentration of L. perenne by 28.54% under the high-Cd treatment, but had no significant effect on the shoot Cd concentration of A. fruticosa in any of the Cd treatments (Figure 4a). Regardless of D. eburnea inoculation, the high-Cd treatment significantly increased the root Cd concentration of L. perenne compared with the low-Cd treatment (Figure 4b). Among all the Cd treatments, D. eburnea inoculation had no significant effect on the root Cd concentration of L. perenne. Compared with the low-Cd treatment, the high-Cd treatment significantly increased the root Cd concentration of A. fruticosa by 189.74% under the D. eburnea inoculation treatment (Figure 4b). D. eburnea inoculation significantly increased the root Cd concentration of A. fruticosa by 252.44% under the high-Cd treatment (Figure 4b).
Regardless of D. eburnea inoculation, shoot Cd accumulation in L. perenne was significantly increased by the high-Cd treatment compared to the low-Cd treatment (Figure 4c). Neither D. eburnea inoculation nor Cd addition had a significant effect on shoot Cd accumulation in A. fruticosa (Figure 4c). Compared with the low-Cd treatment, the high-Cd treatment significantly increased root Cd accumulation in L. perenne under D. eburnea inoculation (Figure 4d). D. eburnea inoculation significantly increased root Cd accumulation in L. perenne by 260.98% under the high-Cd treatment. In addition, D. eburnea inoculation significantly increased root Cd accumulation in A. fruticosa by 161.37% and 398.31% under all low-and high-Cd treatments, respectively (Figure 4d). . Different letters indicate significant differences among treatments for each plant species according to Duncan's test or Tamhane's T2 post hoc test at p < 0.05. 5 mg/kg, 5 mg/kg Cd added to soil; 5 mg/kg + AMF, 5 mg/kg cadmium, and D. eburnea added to soil; 15 mg/kg, 15 mg/kg Cd added to soil; 15 mg/kg + AMF, 15 mg/kg cadmium, and D. eburnea added to soil.

Plant Cd Concentration and Accumulation
The shoot Cd concentrations of both plants significantly increased with an increase in the soil Cd concentration, regardless of D. eburnea inoculation (Figure 4a). D. eburnea inoculation significantly decreased the shoot Cd concentration of L. perenne by 28.54% under the high-Cd treatment, but had no significant effect on the shoot Cd concentration of A. fruticosa in any of the Cd treatments (Figure 4a). Regardless of D. eburnea inoculation, the high-Cd treatment significantly increased the root Cd concentration of L. perenne compared with the low-Cd treatment (Figure 4b). Among all the Cd treatments, D. eburnea inoculation had no significant effect on the root Cd concentration of L. perenne. Compared with the low-Cd treatment, the high-Cd treatment significantly increased the root Cd concentration of A. fruticosa by 189.74% under the D. eburnea inoculation treatment ( Figure  4b). D. eburnea inoculation significantly increased the root Cd concentration of A. fruticosa by 252.44% under the high-Cd treatment (Figure 4b).

Metal Tolerance Index, Transfer Factor, and Bioconcentration Factor (BCF)
In the absence of D. eburnea, the metal tolerance index of L. perenne was significantly reduced in the high-Cd treatment by 50.75% compared with the low-Cd treatment, but the metal tolerance index of L. perenne was significantly increased by 19.94% in the presence of D. eburnea. However, D. eburnea inoculation significantly decreased the metal tolerance index of L. perenne by 16.36% under the low-Cd treatment, but it significantly increased the metal tolerance index of L. perenne by 103.72% under the high-Cd treatment (Figure 5a). Compared with the low-Cd treatment, the metal tolerance index of A. fruticosa inoculated with D. eburnea in the high-Cd treatment was significantly reduced by 34.61%, and D. eburnea inoculation significantly increased the metal tolerance index of A. fruticosa by 33.30% under the low-Cd treatment (Figure 5a).
Neither D. eburnea inoculation nor Cd addition had a significant effect on shoot Cd accumulation in A. fruticosa (Figure 4c). Compared with the low-Cd treatment, the high-Cd treatment significantly increased root Cd accumulation in L. perenne under D. eburnea inoculation (Figure 4d). D. eburnea inoculation significantly increased root Cd accumulation in L. perenne by 260.98% under the high-Cd treatment. In addition, D. eburnea inoculation significantly increased root Cd accumulation in A. fruticosa by 161.37% and 398.31% under all low-and high-Cd treatments, respectively (Figure 4d).

Metal Tolerance Index, Transfer Factor, and Bioconcentration Factor (BCF)
In the absence of D. eburnea, the metal tolerance index of L. perenne was significantly reduced in the high-Cd treatment by 50.75% compared with the low-Cd treatment, but the metal tolerance index of L. perenne was significantly increased by 19.94% in the presence of D. eburnea. However, D. eburnea inoculation significantly decreased the metal tolerance index of L. perenne by 16.36% under the low-Cd treatment, but it significantly increased the metal tolerance index of L. perenne by 103.72% under the high-Cd treatment (Figure 5a). Compared with the low-Cd treatment, the metal tolerance index of A. fruticosa inoculated with D. eburnea in the high-Cd treatment was significantly reduced by 34.61%, and D. eburnea inoculation significantly increased the metal tolerance index of A. fruticosa by 33.30% under the low-Cd treatment (Figure 5a).
Compared with the low-Cd treatment, the high-Cd treatment had no significant effect on the transfer factor of L. perenne, but it significantly increased the transfer factor of A. fruticosa by 132.60% under non-AMF D. eburnea inoculation (Figure 5b), whereas D.

Pathways of AMF D. eburnea and Cd Treatment Effects on D. eburnea Colonization Rate and Bioconcentration Factor
A structural equation model (SEM) analysis was conducted to separately evaluate the influence of the addition of AMF D. eburnea and Cd on the AMF colonization rate and bioconcentration factor for each plant species. According to the model, AMF D. eburnea inoculation had a direct and significant positive impact on the root biomass and plant colonization rate (Figure 6a) and a direct and positive impact on the soil neutral phosphatase content, but the impact was not significant (Figure 6a). The addition of Cd indirectly reduced the colonization rate of D. eburnea by increasing the proportion of exchangeable Cd, but it had no direct or significant impact on the colonization rate of D. eburnea ( Figure  6a). Furthermore, D. eburnea inoculation had a direct and significant positive impact on the root biomass of L. perenne (Figure 6b), where the addition of Cd indirectly improved Compared with the low-Cd treatment, the high-Cd treatment had no significant effect on the transfer factor of L. perenne, but it significantly increased the transfer factor of A. fruticosa by 132.60% under non-AMF D. eburnea inoculation (Figure 5b), whereas D. eburnea inoculation significantly decreased the transfer factor of A. fruticosa by 59.34% under the high-Cd treatment (Figure 5b).
Compared with the low-Cd treatment, the bioconcentration factor of A. fruticosa was significantly increased by 8.77% in the high-Cd treatment by 8.77% in the presence of D. eburnea. D. eburnea inoculation had no significant effect on the bioconcentration factor of L. perenne but significantly increased the bioconcentration factor of A. fruticosa by 47.11% and 51.88% under the low and high Cd treatments, respectively (c).

Pathways of AMF D. eburnea and Cd Treatment Effects on D. eburnea Colonization Rate and Bioconcentration Factor
A structural equation model (SEM) analysis was conducted to separately evaluate the influence of the addition of AMF D. eburnea and Cd on the AMF colonization rate and bioconcentration factor for each plant species. According to the model, AMF D. eburnea inoculation had a direct and significant positive impact on the root biomass and plant colonization rate (Figure 6a) and a direct and positive impact on the soil neutral phosphatase content, but the impact was not significant (Figure 6a). The addition of Cd indirectly reduced the colonization rate of D. eburnea by increasing the proportion of exchangeable Cd, but it had no direct or significant impact on the colonization rate of D. eburnea (Figure 6a). Furthermore, D. eburnea inoculation had a direct and significant positive impact on the root biomass of L. perenne (Figure 6b), where the addition of Cd indirectly improved the bioconcentration factor of L. perenne by increasing the proportion of exchangeable Cd in the soil (Figure 6b). AMF D. eburnea inoculation had a direct and significant positive effect on the bioconcentration factor of A. fruticose, where the addition of Cd had a direct and significant negative impact on the root biomass of A. fruticosa (Figure 6c). Moreover, the addition of Cd indirectly reduced the root biomass of A. fruticosa by increasing the soil pH value (Figure 6c).

Cd in Soil Affected the Colonization Rate of AMF D. eburnea
The colonization rate of AMF D. eburnea was decreased with an increase in Cd concentration in soil, and this occurrence has already been verified in other studies [39,40]. The structural equation model proved that Cd increased the proportion of exchangeable Cd in the soil, which increased the biological toxicity of Cd in the soil and inhibited the development of D. eburnea [41]. Moreover, the AMF colonization rate of Leguminosae A. fruticosa was generally higher than that of Gramineae L. perenne. It is considered that AMF provides A. fruticosa with plenty of phosphorus in the process of nitrogen fixation [33,34], while receiving nitrogen nutrition to improve symbiosis with the host plants [42].

AMF Inoculation and Cd Addition Affected Cd Fractions in Soil
The form of Cd in the soil affects its toxicity and uptake by plants [43]. In this study, the addition of Cd significantly increased the proportion of exchangeable Cd and reduced the proportion of carbonate bound Cd. Previous studies have shown that an increase in pH reduces the number of Cd ions in soil [44]. However, when pH increases to a high level, the solubility of Cd is increased [45][46][47], and some of the Cd ions in high-pH soil

Cd in Soil Affected the Colonization Rate of AMF D. eburnea
The colonization rate of AMF D. eburnea was decreased with an increase in Cd concentration in soil, and this occurrence has already been verified in other studies [39,40]. The structural equation model proved that Cd increased the proportion of exchangeable Cd in the soil, which increased the biological toxicity of Cd in the soil and inhibited the development of D. eburnea [41]. Moreover, the AMF colonization rate of Leguminosae A. fruticosa was generally higher than that of Gramineae L. perenne. It is considered that AMF provides A. fruticosa with plenty of phosphorus in the process of nitrogen fixation [33,34], while receiving nitrogen nutrition to improve symbiosis with the host plants [42].

AMF Inoculation and Cd Addition Affected Cd Fractions in Soil
The form of Cd in the soil affects its toxicity and uptake by plants [43]. In this study, the addition of Cd significantly increased the proportion of exchangeable Cd and reduced the proportion of carbonate bound Cd. Previous studies have shown that an increase in pH reduces the number of Cd ions in soil [44]. However, when pH increases to a high level, the solubility of Cd is increased [45][46][47], and some of the Cd ions in high-pH soil form Cd hydroxide precipitation [48,49]. In the present study, the AMF treatment increased the proportion of exchangeable Cd and decreased the residual Cd in the soil. This phenomenon has also been observed in previous studies [20]. It is considered that this occurred because D. eburnea inoculation increased the plant root biomass, and the amount of plant root activated soil Cd during the plant growth process, which resulted in an increase in the proportion of exchangeable Cd [50].

AMF Impact on Plant Biomass and Cd Uptake
D. eburnea inoculation increased the total biomass and root shoot ratio of both L. perenne and A. fruticosa under the high-Cd treatment, and this may have occurred because D. eburnea improved the soil environment (e.g., NO 3 − -N) and promotes plant growth [15]. The results of the structural equation model showed that D. eburnea inoculation had a positive effect on the root biomass, and the increased soil neutral phosphatase content significantly increased the plant root biomass (Figure 6a). The A-P concentration in soil was decreased in this study when D. eburnea was inoculated with a high-Cd treatment. This may be that the content of available phosphorus in the soil itself was relatively low (Tables 1 and 2), and because D. eburnea inoculation increased the root biomass and the soil neutral phosphatase content, which enabled greater amounts of A-P to be absorbed by the plants [51]. Under the high-Cd treatment, D. eburnea reduced the transfer factor by 50.08% for L. perenne and 59.34% for A. fruticosa, which showed that AMF inhibited the transport of Cd to the plant aerial parts at high-Cd concentrations, thus reducing the damage caused by Cd to aerial parts and improving the resistance of plants. This phenomenon has also been observed in other studies [52,53], and it could result from AMF changing the intracellular distribution and chemical form of Cd in the roots and it increases the cellulose in root cell walls to bind Cd ions [25,54].
Some studies have shown that AMF can promote the absorption of Cd by plants [20,55], but other studies have reported opposite results [12,56]. The mechanism underlying this difference remains unclear. However, according to the findings of our study and those of others, it is possible that different AMF species, plant species, soil properties, and Cd concentrations have different effects on the efficiency of AMF in promoting Cd uptake. The differences between the plant species in this study were mainly evident with respect to the metal tolerance of L. perenne and A. fruticosa. In this respect, L. perenne showed a high tolerance to heavy metals in low-Cd soils and a low tolerance in high-Cd soils, whereas A. fruticosa showed a low tolerance to heavy metals in all of the Cd-treated soils, and the results are consistent with those of previous studies [57]. In summary, inoculation with D. eburnea increased the biomass of plant roots, increased the Cd content of plant roots, and improved the enrichment capacity of plants.

Conclusions
In our study, the biomass and Cd absorption capacity of L. perenne were higher than those of A. fruticosa regardless of D. eburnea inoculation. Under the high-Cd treatment, D. eburnea increased the root biomass and root shoot ratio of both L. perenne and A. fruticosa, and it also increased the Cd absorption and bioconcentration factor of L. perenne and A. fruticosa by altering the soil Cd speciation by increasing exchangeable Cd and decreasing residual Cd. However, Cd was mainly accumulated in the plant roots. Our study indicates that the combined use of D. eburnea and remediation plants may be a viable alternative for remediating Cd-contaminated soil. However, the current data were obtained under greenhouse conditions over a two-month period, and this may not be representative of the influence of AMF in a natural environment. In addition, soil heavy metal pollution under natural conditions can involve significant heavy metal content. Therefore, it is necessary