New Evidence of Semi-Mangrove Plant Barringtonia racemosa in Soil Clean-Up: Tolerance and Absorption of Lead and Cadmium

Mangrove plants play an important role in the remediation of heavy-metal-contaminated estuarine and coastal areas; Barringtonia racemosa is a typical semi-mangrove plant. However, the effect of heavy metal stress on this plant has not been explored. In this study, tolerance characteristics and the accumulation profile of cadmium (Cd) and lead (Pb) in B. racemosa were evaluated. The results indicated that B. racemosa exhibited a high tolerance in single Cd/Pb and Cd + Pb stress, with a significant increase in biomass yield in all treatment groups, a significant increase in plant height, leaf area, chlorophyll and carotenoid content in most treatment groups and without significant reduction of SOD, POD, MDA, proline content, Chl a, Chl b, Chl a + b, Car, ratio of Chl a:b and ratio of Car:Chl (a + b). Cd and Pb mainly accumulated in the root (≥93.43%) and the content of Cd and Pb in B. racemosa was root > stem > leaf. Pb showed antagonistic effects on the Cd accumulation in the roots and Cd showed antagonistic or synergistic effects on the Pb accumulation in the roots, which depended on the concentration of Cd and Pb. There was a significant synergistic effect of Cd and Pb enrichment under a low Cd and Pb concentration treatment. Thus, phytoremediation could potentially use B. racemosa for Cd and Pb.


Introduction
Mangrove forests-woody plant communities that grow in the intertidal zone of tropical and subtropical coastlines [1]-provide a natural barrier to the coast [2]. Mangrove forests have an important ecological service value in the estuarine and coastal zones, where they contribute to plant community composition, play a role as a windbreak and in wave elimination, and contribute to phytoremediation and the removal of heavy metals [1,[3][4][5].
Heavy metal pollution in mangrove forests has been widely reported. In the coastal zone of Olkhon Island, cadmium (Cd) and lead (Pb) could rise to dangerous levels [6]. Araujo et al., found the metal concentrations in soil fractions and mangrove organisms in the Botafogo estuary in Brazil exceeded by up to 2.6-fold the geochemical background, and 64% of the soils were contaminated with Pb [7]. Chen et al., found that Cd in sediments of mangrove forests could cause serious ecological risks [8]. El Ashmawy et al., indicated the contamination of sediment by Pb, Cu and Mn on the Egyptian Red Sea coast [9]. The presence of heavy metals in mangrove wetlands can affect soil microbial communities, soil organic matter decomposition and greenhouse gas emissions [10] and B. racemosa seeds were collected from the natural B. racemosa forest in Danzhou city, Hainan Province, China, in October 2019. The seeds were sown in a sand bed in a greenhouse (22 • 64 N, 110 • 14 E) in the southeast Guangxi Flower Breeding Base of Yulin Normal University. Upon reaching a height of about 10 cm, the seedlings were transplanted into experimental pots (30 cm diameter and 28 cm height) containing a 1.5 kg mixture of culture substrate; the culture substrate was a soil, coconut bran and nutrient soil mixture at a ratio of 2:1:1 (v:v:v), with 1 plant per pot. The soil was taken from the 20 cm surface soil on the shore of Lake Yulan of Yulin Normal University, air dried, handpicked to remove plant detritus and stones, then used for plant culture. The physicochemical properties of the soil were as follows: organic matter, 30.52 g kg −1 ; total N, 2.04 g kg −1 ; total P, 0.72 g kg −1 ; total K, 7.47 g kg −1 ; pH, 7.62; Cd, 0.149 mg kg −1 ; Pb, 3.64 mg kg −1 . The seedlings were used for the experiment after half a year of normal growth.

Pot Experiment
A total of 13 treatments were set up as described in Table 1. The concentrations of Cd and Pb added were based on the pre-experimental results, while the soil environmental quality risk control was according to the standard for soil contamination of agricultural land (Beijing, China, GB 15618-2018) [32,33]. Cd and Pb solutions were prepared by dissolving Cd (NO 3 ) 2 ·4H 2 O and Pb (NO 3 ) 2 in deionized water, respectively, then applying 400 mL of different treatment solutions to each pot with seedlings. To prevent the leakage of Cd and Pb solutions from the pots, all pots were placed in plastic trays, with three pots per tray. There was a total of 15 B. racemosa seedlings per treatment (1 seedling per pot), with 3 replicates. The soil was air dried, ground and passed through a 0.149 mm (100 mesh) nylon sieve. Next, 50 g of the soil was analyzed for heavy metal content as previously described [34,35]. After 80 days of exposure to Cd and Pb stress treatments, plant height and ground diameter were determined by the method described by Liang et al., [29]. B. racemosa leaves and root morphology were scanned on the Microtek ScanMaker i800 plus system (WSeen, Hangzhou, China) and leaf area and root surface area were calculated using a LA-S Leaf Area and Root Morphology Analysis software (WSeen, Hangzhou, China), respectively.
One seedling from each pot was divided into its root, stem and leaf parts. Fresh tissue was weighed, dried at 105 • C for 30 min and samples were dried at 70 • C to a constant weight (approximately 48 h). Next, dry weights (DW) of the samples were recorded and the total biomass as well as root:shoot ratios (R/S) were calculated using the following formulae [18]: where DW root , DW stem and DW leaf represent the dry weight of the roots, stems and leaves, respectively.

Determination of Superoxide Dismutase and Peroxidase Activity and Quantification of Malondialdehyde and Proline Levels
Three mature leaves of each plant were collected after 80 days of exposure to Cd and Pb stress treatments. The leaf samples (0.5 g) were ground under liquid nitrogen and homogenized in 10 volumes of ice-cold 50 mM sodium phosphate buffer (pH 7.8). The contents were centrifuged at 15,000× g at 4 • C for 20 min and the resulting supernatants were collected and used for enzyme activities. Superoxide dismutase (SOD) activity was determined using the modified nitroblue tetrazolium (NBT) method, peroxidase (POD) activity was determined by the guaiac wood phenol method, while enzyme activity was measured as described by Saleem [36]. Quantification of malondialdehyde (MDA) content was determined using the thiobarbituric acid (TBA) reaction method, as described by Atabaki et al. [37], while proline content was measured by the acid ninhydrin method [38]. The optical density (OD) values of SOD, POD, MDA and proline were measured with a Varioskan Lux Multimode Microplate Reader (Thermo Fisher Scientific, Singapore) [29].

Photosynthetic Pigments Determination
Chlorophyll a (Chl a), chlorophyll b (Chl b) and carotenoid (Car) were extracted from 0.5 g of fresh leaves using 80% (v/v) acetone and quantified as previously described [39]. Extraction procedures were done at room temperature, under dark conditions for 24 h. Then chlorophyll a, b and Car content in the supernatant were measured using the Varioskan Lux Multimode Microplate Reader (Thermo Fisher Scientific, Singapore) at 649, 665 and 470 nm wavelengths, respectively. Concentrations of chlorophyll a, b, chlorophyll (a + b) and carotenoids in the leaf were calculated using the following equations [40,41]:  [42] , followed by the determination of the Cd and Pb concentration using a direct-reading inductively coupled plasma optical emission spectrometer ((ICP-OES) Optima8000, PerkinElmer Inc., Waltham, MA, USA) [14,43].
The translocation factor (TF) and bioenrichment factor (BCF) were calculated as follows: Translocation factor (TF) = Cs/Cr Bioenrichment factor (BCF) = Cp/Cso (8) where Cs is the metal concentration in the aerial part (shoot) of the plant, Cr is the concentration in the roots, Cp is the metal concentration in the plant (shoot) and Cso is the total metal concentration in the soil [44][45][46].

Statistical Analysis
Differences between different Cd, Pb and Cd + Pb treatments were assessed by a one-way analysis of variance (ANOVA), which was implemented in SPSS 19.0 software (SPSS, Inc., Chicago, IL, USA) with multiple comparisons performed using the Duncan's test at a 5% significance level (p < 0.05). Regression analysis was performed to develop models for the absorption of Cd and Pb using OriginPro 2021 9.8.5.201 (OriginLab Inc., Northampton, MA, USA).

Effects of Cd and Pb on Plant Growth
Cd stress increased plant height in B. racemosa ( Figure 1A), while only the Cd4 treatment exhibited a significant increase (increased by 86.9%). Only the Cd1 treatment exhibited a significantly higher ground diameter (increased by 43.31%) relative to the controls ( Figure 1B). Treatments with Cd3 and Cd4 resulted in a significant increase in leaf area, increasing by 49.27% and 56.20%, respectively, compared with the control ( Figure 1C). Moreover, Cd1 and Cd2 resulted in a significantly higher root surface area, relative to the control, with the greatest effect observed in Cd1 (increased by 151.58%) ( Figure 1D). With regard to biomass, B. racemosa plants under four Cd treatments increased significantly ( Figure 1E). In addition, a significantly lower root:shoot ratio in plants under Cd3 and Cd4 treatments was found, relative to the control ( Figure 1F).
For Pb stress, Pb3 and Pb4 treatments resulted in a significantly higher plant height (increased by 117.60% and 148.17%) relative to the control. Ground diameter significantly increased by 27.56% under the Pb1 treatment and decreased by 29.13% under the Pb4 treatment, compared with the control ( Figure 1B). Moreover, plants exposed to three Pb treatments (Pb2, Pb3 and Pb4) had a significantly higher leaf area compared with the control ( Figure 1C). Pb2 and Pb3 treatments resulted in a significantly lower root surface area ( Figure 1D) and root:shoot ratio ( Figure 1F), while plants treated with Pb4 exhibited a significant increase by 96.82% in root surface area ( Figure 1D). Plants treated with four Pb concentrations exhibited a significant increase in biomass ( Figure 1E) For Pb stress, Pb3 and Pb4 treatments resulted in a significantly higher plant heig (increased by 117.60% and 148.17%) relative to the control. Ground diameter significan increased by 27.56% under the Pb1 treatment and decreased by 29.13% under the P treatment, compared with the control ( Figure 1B). Moreover, plants exposed to three treatments (Pb2, Pb3 and Pb4) had a significantly higher leaf area compared with the co trol ( Figure 1C). Pb2 and Pb3 treatments resulted in a significantly lower root surface ar ( Figure 1D) and root:shoot ratio ( Figure 1F), while plants treated with Pb4 exhibited For Cd + Pb treatment, the height of B. racemosa plants increased, with an increase in Cd + Pb concentration, compared with the control. The plant height under M2, M3 and M4 significantly increased; the highest value was observed in M4, evidenced by a 173.17% increment ( Figure 1A). Exposure to M1 and M2 treatments resulted in a significant increase in ground diameter, 29.92% and 43.31% increments, respectively, relative to the control plants ( Figure 1B). Plants treated with four Cd + Pb concentrations exhibited a significantly higher leaf area than the controls ( Figure 1C). Moreover, plants treated with M1, M2 and M3 exhibited a significantly higher root surface area than their control ( Figure 1D). B. racemosa plants exposed to four Cd + Pb treatments recorded a significantly higher biomass than the control; the highest value was observed in the M2 treatment ( Figure 1E). Furthermore, exposure to four Cd + Pb treatments resulted in significantly lower root:shoot ratios compared with the control ( Figure 1F).

Effects of Cd and Pb Stress on Physiological Characteristics
No significant differences were found in the superoxide dismutase (SOD) and proline concentrations across all Cd treatments and controls (Figure 2A,D). The plants treated with Cd3 and Cd4 had significantly higher peroxidase (POD) and malondialdehyde (MDA) than those in the control ( Figure 2B,C), and POD under Cd4 stress was significantly higher than under the Cd3 treatment. There were no significant differences in MDA contents between plants treated with Cd3 and Cd4.
Cd + Pb concentration, compared with the control. The plant height under M2, M3 a M4 significantly increased; the highest value was observed in M4, evidenced by a 173.1 increment ( Figure 1A). Exposure to M1 and M2 treatments resulted in a significant crease in ground diameter, 29.92% and 43.31% increments, respectively, relative to control plants ( Figure 1B). Plants treated with four Cd + Pb concentrations exhibite significantly higher leaf area than the controls ( Figure 1C). Moreover, plants treated w M1, M2 and M3 exhibited a significantly higher root surface area than their control ( Fig  1D). B. racemosa plants exposed to four Cd + Pb treatments recorded a significantly hig biomass than the control; the highest value was observed in the M2 treatment (Figure 1 Furthermore, exposure to four Cd + Pb treatments resulted in significantly low root:shoot ratios compared with the control ( Figure 1F).

Effects of Cd and Pb Stress on Physiological Characteristics
No significant differences were found in the superoxide dismutase (SOD) and prol concentrations across all Cd treatments and controls (Figure 2A,D). The plants trea with Cd3 and Cd4 had significantly higher peroxidase (POD) and malondialdehy (MDA) than those in the control ( Figure 2B,C), and POD under Cd4 stress was sign cantly higher than under the Cd3 treatment. There were no significant differences in M contents between plants treated with Cd3 and Cd4.  Similarly, there were no significant differences in SOD and proline concentrations across all Pb treatments and controls (Figure 2A,D). The plants treated with Pb2, Pb3 and Pb4 exhibited significantly higher POD levels than the controls, with the highest POD under the Pb3 treatment ( Figure 2B). Exposure to Pb1 resulted in a significant increase in concentrations compared with the control ( Figure 2C).
Significant lower SOD and POD concentrations were observed only after treatment with M2 compared with the control (Figure 2A,B). The plants treated with M1 and M4 had significantly higher MDA concentrations than the control ( Figure 2C). No significant differences were observed in the proline concentration across the treatments ( Figure 2D).

Effects of Cd and Pb Stress on Photosynthetic Pigments
B. racemosa treated with Cd1 exhibited significantly lower chlorophyll (Chl) a, Chl b and Chl (a + b) concentrations than the control ( Figure 3A-C). Cd1, Cd2 and Cd4 treatments significantly increased chlorophyll a:b ratio and Car:Chl (a + b) ratio ( Figure 3E,F). B. racemosa treated with Pb1 exhibited significantly higher chlorophyll (Chl) a, Chl b, Chl (a + b) and Car concentrations than the control (Figure 3A-D). Pb3 and Pb4 treatments significantly increased Chl a (Figure 3A), Pb3 treatment increased Chl a + b ( Figure 3C) and Pb2 and Pb3 treatments increased Car ( Figure 3D). Pb3 and Pb2 treatments increased Chl a:b ratio and Car:Chl (a + b) ratio, respectively ( Figure 3E,F). Under Cd + Pb treatments, compared with the control, M2, M3 and M4 significantly increased Chl a and chlorophyll a:b ratio ( Figure 3A,E), M4 significantly increased Chl b (Figure 3B), M2 and M4 significantly increased Chl a + b ( Figure 3C) and M1, M2, M3 and M4 significantly increased Carotenoid (Car) ( Figure 3D). M1 significantly increased Car:Chl (a + b) ratio ( Figure 3F).

Concentration and Distribution of Cd and Pb in Plants
The Cd concentration and distribution in roots, stems and leaves of B. racemosa pl after 80 days of treatment are shown in Table 2. More than 93.43% of Cd in the pl accumulated in the roots under the Cd and the Cd + Pb treatments, indicating that Cd not effectively translocate to stems and leaves. Cd accumulated less in the roots under

Concentration and Distribution of Cd and Pb in Plants
The Cd concentration and distribution in roots, stems and leaves of B. racemosa plants after 80 days of treatment are shown in Table 2. More than 93.43% of Cd in the plants accumulated in the roots under the Cd and the Cd + Pb treatments, indicating that Cd did not effectively translocate to stems and leaves. Cd accumulated less in the roots under the Cd + Pb treatment significantly than under the Cd treatment, exhibiting antagonistic effects. However, in the stems, Cd accumulated more under the M2 treatment significantly than under the Cd2 treatment, exhibiting synergistic effects. In the leaves, Cd accumulated more under the M4 treatment significantly than under the Cd4 treatment, exhibiting synergistic effects. Similarly, Pb primarily accumulated in the roots (more than 95.20% of Pb in the plants) compared with the stems and leaves of B. racemosa seedlings under Pb and Cd + Pb treatments ( Table 3). The results also indicated that Pb primarily accumulated in the roots and found it difficult to translocate to stems and leaves. Pb accumulation in the roots under M1 and M4 was higher than that under Pb1 and Pb4, respectively, which showed significant synergistic effects, while that under M2 and M3 in the roots was significantly lower relative to the plants exposed to Pb2 and Pb3, respectively, indicative of antagonistic effects. On the contrary, in the stems, Pb accumulation under M1 and M4 treatments were significantly lower than under Pb1 and Pb4 treatments, respectively, exhibiting antagonistic effects. Pb accumulation under M2 and M3 treatments were significantly higher than under Pb2 and Pb3 treatments, respectively, exhibiting synergistic effects. In the leaves, Pb accumulation under the M2 treatment was significantly lower than under the Pb2 treatment, exhibiting antagonistic effects. The Pb accumulation under the M1, M3 and M4 treatments were significantly higher than under the Pb1, Pb2 and Pb4 treatments, respectively, exhibiting synergistic effects.
Significantly high Cd and Pb enrichment coefficients were found under the M1 treatment (Tables 2 and 3); the enrichment coefficient of Cd and Pb were 28.58 and 11.63, respectively. The translocation factor of Cd (0.35) and Pb (0.63) in the control was dramatically higher than under single Cd/Pb and Cd + Pb treatments (range 0.01-0.07 for Cd and 0.01-0.05 for Pb).

Cd and Pb Absorption Models
Results from the regression analyses revealed that the concentration of Cd and Pb in the soil under single Cd or Pb or Cd + Pb treatments were strongly correlated with (R 2 ≥ 0.93) Cd or Pb in the roots ( Figure 4A-D). The Cd concentration in the roots and the Cd concentration added in the soil fitted the quadratic equation under the Cd and Cd + Pb treatments. The Pb concentration in the roots and the Pb concentration added in the soil fitted the quartic equation in the Pb and Cd + Pb treatments.

Cd and Pb Absorption Models
Results from the regression analyses revealed that the concentration of Cd and Pb the soil under single Cd or Pb or Cd + Pb treatments were strongly correlated with (R 0.93) Cd or Pb in the roots (Figure 4A-D). The Cd concentration in the roots and the C concentration added in the soil fitted the quadratic equation under the Cd and Cd + P treatments. The Pb concentration in the roots and the Pb concentration added in the so fitted the quartic equation in the Pb and Cd + Pb treatments.

Discussion
Tolerance, absorption and migration of mangrove plants to heavy metals is im portant for using mangrove plants for phytoremediation [15]. In the long evolutiona process, plants have formed their own internal anti-stress mechanism to reduce or elim nate the damage caused by various stresses. Mangrove plants Kandelia obovate and B. gy norrhiza are tolerant to multiple heavy metals, such as Cd 2+ , Pb 2+ and Hg 2+ [19,47]. Ma grove plant A. marina is used in the phytoremediation process, owing to its ability in t deposition of heavy metals and the transport to upper plant parts, hence preventing po lution of the nearby areas with heavy metals [9].
Biomass, plant height and leaf area are important indicators of plant tolerance [4 In the present study, typical semi-mangrove plant B. racemosa seedlings exposed to sing Cd/Pb and Cd + Pb combined stress for 80 days did not exhibit heavy metal poisoning either shoots or roots. Biomass was significantly increased under all single Cd/Pb and C + Pb stresses. Plant height and leaf area were significantly increased under most sing Cd/Pb and Cd + Pb stresses. Ground diameter and root surface area did not significant reduce under most single Cd/Pb and Cd + Pb stresses (Figure 1). These results, especial

Discussion
Tolerance, absorption and migration of mangrove plants to heavy metals is important for using mangrove plants for phytoremediation [15]. In the long evolutionary process, plants have formed their own internal anti-stress mechanism to reduce or eliminate the damage caused by various stresses. Mangrove plants Kandelia obovate and B. gymnorrhiza are tolerant to multiple heavy metals, such as Cd 2+ , Pb 2+ and Hg 2+ [19,47]. Mangrove plant A. marina is used in the phytoremediation process, owing to its ability in the deposition of heavy metals and the transport to upper plant parts, hence preventing pollution of the nearby areas with heavy metals [9].
Biomass, plant height and leaf area are important indicators of plant tolerance [48]. In the present study, typical semi-mangrove plant B. racemosa seedlings exposed to single Cd/Pb and Cd + Pb combined stress for 80 days did not exhibit heavy metal poisoning in either shoots or roots. Biomass was significantly increased under all single Cd/Pb and Cd + Pb stresses. Plant height and leaf area were significantly increased under most single Cd/Pb and Cd + Pb stresses. Ground diameter and root surface area did not significantly reduce under most single Cd/Pb and Cd + Pb stresses (Figure 1). These results, especially for biomass, indicated that B. racemosa exhibited strong stress tolerance against Cd and Pb stresses; probably, the increase of leaf area was for the purpose of absorbing more light energy and obtaining more organic matter to cope with the stress.
The antioxidant system in plants is a well-known defensive mechanism against reactive oxygen species and POD activity in roots and leaves maybe serve as a biomarker of heavy metal stress in K. candel [19,47] [49,50]. MDA is an important product of cell membrane lipid peroxidation and it is an effective indicator of the degree of cell membrane damage [51]. In this study, compared with the control, the activity of POD increased significantly under high Cd and Pb concentration treatments, MDA significantly increased under Cd3, Cd4, Pb1, M1 and M4 ( Figure 2) and B. racemosa showed strong resistance to high Cd and Pb stress. Significant lower POD concentrations were observed only under treatment with M2. SOD only under M2 treatment was significantly lower than the control group, which indicated that there were other ways or other cell osmotic substances to alleviate the toxicity of Cd and Pb.
Heavy metal stress affects the water balance in plants and induces a significant increase in the level of proline, which is involved in the osmotic regulation in cells [52]. However, there was no significant difference in the proline content in all treatment groups compared with the controls in the present study, probably because B. racemosa alleviates the damage of cell membrane permeability mainly through the enzyme system rather than by increasing the proline content. This also could be the strategy for the underlying response to heavy metals Cd and Pb in the B. racemosa species.
Photosynthetic pigments are essential for capturing light energy in plants, thus playing a key role in plant photosynthesis [53]. The chlorophyll content in photosynthetic pigments can directly reflect the growth status and the photosynthetic capacity of plants. Carotenoids are photosynthetic pigments that play important antioxidant roles; carotenoids can absorb residual energy, quench reactive oxygen species and prevent membrane lipid peroxidation [40]. Car/Chl (a + b) play key roles in protecting PSI and PSII photosystems [54]. Increasing the chlorophyll content possibly could reduce oxidative damage of the cell membrane by regulating the activities of antioxidant enzymes in wheat seedlings under Cd stress [53]. Previous findings indicated that heavy metal Pb and Cd stress reduces the chlorophyll content in the leaves of the mulberry (Morus alba L.) [55]. In the current study, the contents of Chl a, Chl b and Chl (a + b) significantly reduced only under Cd1 stress; Car, Chl a/b and Car/Chl (a + b) did not reduce under any Cd, Pb or Cd + Pb stress (Figure 3). These results could contribute to the tolerance to Cd and Pb contamination; Cd, Pb or Cd + Pb contamination could not reduce the antioxidant level or photosynthetic capacity of B. racemosa.
In the present study, more than 93.43% of Cd and Pb accumulated in the root system of the plants under Cd, Pb or Cd + Pb stress, and the content of Cd and Pb in B. racemosa was root > stem > leaf. Therefore, Cd and Pb primarily accumulated in the roots. The translocation factors (TF) of Cd and Pb were less than 1 under Cd, Pb and Cd + Pb stress (Tables 2 and 3), the transport capacity of Cd and Pb was weak and Cd and Pb were difficult to transfer from the root to the aerial parts. However, the Cd and Pb enrichment ability of the root system of B. racemosa is stronger, which could be one of the reasons that B. racemosa was tolerant to Cd and Pb.
The coexistence of heavy metals is an important factor that affects the adsorption and transport of heavy metals in plants [9,46,56]. Liu et al. [57] reported that Cd inhibited the accumulation of Pb in Impatiens balsamina, whereas Pb promoted the accumulation of Cd. Wang et al., found that the added Pb may enhance the efficiency of Cd phytoextraction in wheat and maize [58]. Lou et al., found that the presence of a low Cd level impacted positively towards tall fescue growth under Pb stress, while a high level of Cd impacted negatively [59]. In this study, Pb inhibited Cd accumulation in the root (Table 2). However, Pb accumulation in the root showed significantly synergistic effects under M1 and M4, and antagonistic effects under M2 and M3 (Table 3). Therefore, the effect of Cd on the Pb accumulation depended on the concentration of Cd and Pb, probably because the effect of Cd on the Pb accumulation performed differently under different levels of Cd + Pb stress, and different Cd concentrations could induce different mechanisms of Pb accumulation, which is worth further investigation. Under each single Cd/Pb or Cd + Pb treatments, there were differences of the effect of Pb on the Cd accumulation, and the effect of Cd on the Pb accumulation were found between the roots, stems and leaves, which indicated that the interactions of Cd and Pb could be opposite in different organs. In addition, the highest enrichment coefficient of Cd and Pb were 28.58 and 11.63 under the M1 treatment, respectively (Tables 2 and 3), which was about 10-20 times higher than that under other treatments. This result indicated that the synergistic effect of Cd and Pb enrichment under a low Cd and Pb concentration treatment (M1) was significantly stronger than under other treatments. Therefore, the phytoremediation effect of B. racemosa will be better in a low concentration of Cd and Pb polluted areas.
In this study, the absorption model of Cd and Pb in B. racemosa was simulated to analyze their absorption rules. The Cd concentration in the roots and the Cd concentration added in the soil of B. racemosa fitted the quadratic equation in Cd and Cd + Pb treatments ( Figure 4A,B); roots will accumulate more Cd under higher Cd and Cd + Pb stress. The Pb concentration in the roots and the Cd concentration added in the soil fitted the quartic equation in Pb and Cd + Pb treatments ( Figure 4C,D), though the concentration of Pb fluctuated with Pb concentration increasing in the soil and the Pb concentration would possibly increase in the roots. This indicated that B. racemosa potentially has the ability to hyperaccumulate Cd and Pb in the roots.

Conclusions
B. racemosa showed a high tolerance of single Cd/Pb and Cd + Pb stress, with a significant increase in biomass yield in all treatment groups, with significant increases in plant height, leaf area, chlorophyll and carotenoid content in most treatment groups, and without significant reduction of SOD, POD, MDA, proline content, Chl a, Chl b, Chl a + b, Car, ratio of Chl a:b or ratio of Car:Chl (a + b).
The B. racemosa plant accumulates high amounts of Cd and Pb in the roots and the root system plays an important role in ameliorating Cd and Pb toxicity. There were antagonistic effects of Pb on Cd accumulation and the effect of Cd on Pb accumulation depended on the concentration of Cd and Pb. The synergistic effect of Cd and Pb enrichment under a low Cd and Pb concentration treatment was significant. B. racemosa potentially has a significant ability to hyperaccumulate Cd in the roots and it has a high potential for the phytoremediation of Cd-and Pb-polluted soil.