1 Introduction

Mercury (Hg) and cadmium (Cd) are trace elements that are potentially harmful to organisms and ecosystems because of their high toxicity, carcinogenicity, and non-biodegradability (Shekhawat et al. 2022). Active heavy metals (HMs) in soil can readily affect human health through the food chain. Excessive consumption of Hg and Cd has been extensively recognized to cause severe harm to the human kidneys, immune system and nervous system (Moutcine et al. 2020). Therefore, it is vital to identify suitable methods to remediate Hg- and Cd-contaminated farmlands and prevent the uptake of HMs by plants.

In situ immobilization/stabilization method is considered the most effective measure for remediating HMs contaminated soils due to its benefits of rapid implementation, easy operation and cost-effectiveness (Liang et al. 2019). Various novel soil amendments have been reported, including phosphorous-containing materials (Wu et al. 2022), biochar (Wang et al. 2020b), clay minerals (Wang et al. 2021b), metal oxides (Huang et al. 2022b), and organic fertilizers (Wang et al. 2020a), etc. Numerous researchers have extensively emphasized the potential of biochar (BC) for the remediation of polluted soil and agricultural environmental benefits (Golia et al. 2022). Biochar amendments can alter soil physicochemical properties (including soil pH, organic carbon (C) composition, aggregate structure, and nutrient availability), and reduce greenhouse gas emissions (Situ et al. 2022; Yang et al. 2022c). It has a large specific surface area and contains abundant functional groups, through which it can decrease the bioavailability of HMs in soil–plant systems. According to previous studies, the proposed mechanisms of biochar action on HMs in soil include both direct and indirect interactions. The major mechanisms of direct interactions include ion exchange, electrostatic interactions, precipitation and surface complexation (Li et al. 2017; Gong et al. 2022). Furthermore, the bioavailability and transportation of HMs may be indirectly affected by biochar-driven biogeochemical redox processes, such as the redox elements sulfur (S) and iron (Fe) cycling (Yang et al. 2021a).

Generally, pristine biochar has a poor adsorption capacity for HMs owing to the constraints related to the surface structure. Therefore, a variety of surface modification approaches, such as acid/alkali modification (Wang et al. 2022), and loading with metal oxides (Feng et al. 2020), minerals (Wang et al. 2015), and nano-particles (Yan et al. 2015), have been used to enhance the adsorption capabilities of biochar. In the past decade, the loading of FeOx/MnOx particles onto biochar surfaces has attracted considerable attention as it effectively increases its adsorption capacity for HM ions by promoting the creation of numerous functional groups and improving pore structures (Wang et al. 2018; Xiao et al. 2020). Lin et al. (2019) discovered that Fe–Mn oxide–biochar composite supplementation lowered the concentration of available arsenic (As) and increased soil enzyme activity through pot trails. Nevertheless, most recent studies focused on the remediation of single HM contaminated soils, and only a few examined the immobilization impact of modified BC in soil co-contaminated with two or more HMs.

In the previous study, we have revealed that the novel Fe–Mn oxide modified biochar (FMBC) exhibited good adsorption performance for Hg and Cd in aqueous systems (Sun et al. 2023). In this context, FMBC was used as an amendment to remediate Hg and Cd co-contaminated paddy soils, and we aimed to further understand the stabilizing effect of FMBC in Hg and Cd co-contaminated soils and elucidate the associated mechanisms. In this present study, we addressed the following: (i) the immobilization potential of FMBC in Hg and Cd co-contaminated paddy soils, (ii) Hg and Cd accumulation and transfer in rice plants, (iii) the role of FMBC in the formation of root IMP, (iv) the change of soil organic C fractions, and (v) the impact of FMBC on bacterial community composition and structure.

2 Materials and methods

2.1 Preparation of biochar (BC) and Fe–Mn modified biochar (FMBC)

Corn straw was rinsed with distilled water to remove specks of dust and pyrolyzed at 600 ℃ for 2 h in a muffle furnace under an N2 condition after being air-dried. The BC sample was cooled and sieved through a 100-mesh sieve (0.15 mm). FMBC was produced by partitioning 100 g of BC into 1 L Fe(NO3)3·9H2O and KMnO4 mixed solution with a mass ratio of Fe:Mn:BC of 0.5:3:10 (the optimal proportion was selected according to the previous study, w/w) (Sun et al. 2023). The specific procedures for the preparation and physicochemical properties of the BC and FMBC are shown in Fig. S1 and Table S1.

2.2 Experimental design

A soil pot experiment was conducted in the greenhouse of Agro-Environmental Protection Institute, MARA (39° 09′ 79″ N, 117° 15′ 26″ E) from May 2021 to Oct. 2021. Surface soil from a depth of 0–20 cm was collected from a Hg and Cd co-contaminated paddy field in Tongren City, Guizhou Province, China (109°14′ E, 27° 30′ N). The physicochemical parameters of the paddy soils are listed in Table S2. We used the following seven treatments, each with three replicates: one control, and three BC and three FMBC treatments with respective application dosages of 0.5%, 1%, and 2% (w/w), henceforth referred to as 0.5%BC, 1%BC, 2%BC, 0.5%FMBC, 1%FMBC, and 2%FMBC, respectively. Additionally, chemical fertilizers containing nitrogen (N), phosphorus (P), and potassium (K) at 0.2, 0.3, 0.4 g kg–1, respectively, were applied in the form of CO(NH2)2 and KH2PO4. The pesticide application amount, flooding conditions and pest control were executed according to normal rice planting protocols. To stimulate rice growth, anaerobic conditions were maintained with water levels of up to 3 cm until the late grain-filling stage, and intermittent flooding conditions were adopted during the maturity stage.

Rice seeds (Oryza sativa L., Huanghuazhan) were provided by Hunan Golden Nongfeng Seed Industry Co., Ltd., China. Rice seeds were surface-sterilized with 30% hydrogen peroxide (H2O2) for 20 min and cleaned with deionized water, after which they were soaked in saturated calcium chloride (CaCl2) solution at 25 °C in the dark for one night. The seeds were spread evenly on a sterile plastic mesh and transferred to a smart greenhouse after germination. Twenty days later, rice seedlings of the same growth size were selected and transferred to test pots. Each experimental treatment was replicated in triplicate, with three plants per pot. The test pots were rotated intermittently to ensure consistent growth conditions. After harvesting, the roots, stems, leaves, and grains were collected and rinsed thoroughly using distilled water.

2.3 Measurements and analyses

2.3.1 Soil physicochemical properties

Soil pH and EC were determined using a pH meter (NY/T 1377-2007) and a conductivity meter (FE38–Meter); Soil Eh was determined using a redox potentiometer (S8–Field Kit, Mettler Toledo).

2.3.2 Soil organic carbon (SOC) fractions

The total organic carbon (TOC) in soil was measured by the potassium dichromate oxidation spectrophotometric method (HJ-615-2011). The readily oxidized carbon (ROC) was determined by 333 mmol L–1 KMnO4 oxidation procedure at 565 nm with a TU-1810 visible spectrophotometer (Beijing Purkinje General Instrument Co., Ltd., China) (Blair et al. 1995). The dissolved organic carbon (DOC) was obtained by extracting soil samples with deionized water (1:5, w/v) (Li et al. 2019). The fumigation-extraction method was used to determine the microbial biomass carbon (MBC) (Vance et al. 1987; Feyzi et al. 2020). The extraction procedure was performed on the unfumigated soil samples, and the filtered extract was determined on a TOC analyzer (Vario TOC select, Elementar, Germany). The content of soil MBC was obtained by the fumigated SOC content minus the unfumigated SOC content, and then divided by the conversion factor (kEc = 0.45).

The C-pool index (CPI), C-pool activity (CPA), C-pool activity index (CPAI), and C-pool management index (CPMI) were effective indexes to evaluate the quality of soil carbon pool (Jiang et al. 2021). The calculation formulas can be expressed as:

$${\text{CPI}} = {\text{TOC}}\;{\text{in}}\;{\text{the}}\;{\text{treatment}}/{\text{TOC}}\;{\text{in}}\;{\text{the}}\;{\text{reference}}$$
$${\text{CPA}} = {\text{ROC}}/{\text{TOC}} - {\text{ROC}}$$
$${\text{CPAI}} = {\text{CPA}}\;{\text{in}}\;{\text{the}}\;{\text{treatment}}/{\text{CPA}}\;{\text{in}}\;{\text{the}}\;{\text{reference}}$$
$${\text{CPMI}} = {\text{CPI}} \times {\text{CPAI}} \times 100$$

2.3.3 Hg and Cd concentrations in soil and Oryza sativa L.

The available Hg and Cd were extracted by Na2S2O3 and DTPA solution, and obtained Na2S2O3–Hg and DTPA–Cd, respectively. DTPA–Cd was determined in a 1:5 (w:v) ratio of soil to DTPA solution (0.005 M DTPA, 0.01 M CaCl2, and 0.1 M triethanolamine adjusted to pH 7.3 with HCl) (Lindsay and Norvell, 1978). Na2S2O3–Hg was determined in a 1:10 (w:v) ratio of soil to 0.01 M Na2S2O3 solution (Chang et al. 2020). The soil total Hg content was determined by 10 mL concentrated acid boiling (HNO3:HCl:H2O = 1:3:4 (v/v/v)) water digestion for 2 h (GB/T 22105.1-2008). And the total Cd content was determined by digestion of 8 mL HNO3 and 4 mL HF. The Hg and Cd concentrations in rice tissues were measured by digestion after the addition of 10 mL concentrated acid (HNO3:HCl:H2O = 1:3: 4 (v/v/v)) and HNO3, respectively.

The Hg and Cd fractions in soil of different treatments were measured using the sequential extraction procedure of Tessier et al. (1979), and the five fractions were exchangeable fraction (EXC), carbonate-bound fraction (CAR), Fe/Mn oxide-bound fraction (OX), organic matter-bound fraction (OM), and residual fraction (RES), respectively (Supp. Text S1).

The calculation formulae of bioconcentration factor (BF) and translocation factor (TF) are shown below:

$${\text{BF}}_{{{\text{root}}}} = {\text{C}}_{{{\text{root}}}} /{\text{C}}_{{{\text{soil}}}}$$
$${\text{BF}}_{{{\text{stem}}}} = {\text{C}}_{{{\text{stem}}}} /{\text{C}}_{{{\text{soil}}}}$$
$${\text{BF}}_{{{\text{grain}}}} = {\text{C}}_{{{\text{grain}}}} /{\text{C}}_{{{\text{soil}}}}$$
$${\text{TF}}_{{\text{root - stem}}} = {\text{C}}_{{{\text{stem}}}} /{\text{C}}_{{{\text{root}}}}$$
$${\text{TF}}_{{\text{root - grain}}} = {\text{C}}_{{{\text{grain}}}} /{\text{C}}_{{{\text{root}}}}$$
$${\text{TF}}_{{\text{stem - grain}}} = {\text{C}}_{{{\text{grain}}}} /{\text{C}}_{{{\text{stem}}}}$$

where, Csoil is the Hg and Cd contents (mg kg–1) in soil of different treatments after rice harvest; Croot, Cstem, and Cgrain are the Hg and Cd contents (mg kg–1) in the root, stem, and grain of rice, respectively.

2.3.4 Extraction of iron–manganese plaques from roots

Iron–manganese plaques (IMP) formed on the surface of fresh rice roots were extracted using dithionite–citrate–bicarbonate (DCB) method (Taylor and Crowde 1983). The specific method is listed in Supp. Text S2.

2.3.5 Bacterial community analysis

Fresh soil samples were collected and kept in a refrigerator at – 80 ℃ for microbial sequencing. The microbial sequencing service was commissioned by Shanghai Majorbio Bio-Pharm Technology Co., Ltd. The details of high-throughput sequencing analysis are provided in Supp. Text S3.

2.4 Statistical analysis

Standard soil reference materials (SRM 2586, NIST) and standard rice flour reference materials (GBW(E)100561 and GBW(E)100349, National Institute of Metrology, China) were used to assess the quality of experimental data. One-way analysis of variance (ANOVA) was used to examine differences between treatments, and multiple comparisons of means were performed using Duncan’s test (P < 0.5). Data analysis and visualization were performed using DPS 9.01 and Origin 2021 software. The Mantel tests and partial least square path modelling (PLSPM) analysis were conducted using the “linkET” and “plspm” packages with R software (v.4.2.1).

3 Results and discussion

3.1 Soil physicochemical properties under different treatments

3.1.1 Soil solution pH and redox potential (Eh)

Figure 1 depicts the dynamics of soil solution pH and soil Eh values throughout the rice growth stages. Overall, the soil solution pH first increased and then steadily decreased with the extensive of soil incubation time. BC and FMBC applications enhanced the pH of the soil solution, which was associated with the strong alkalinity of BC (pH 10.06) and FMBC (pH 10.24). In contrast, BC and FMBC addition reduced soil Eh, and the soil Eh increased and subsequently decreased with the rice growth stage, reaching a peak at the grain-filling (GF) stage. The extensive pore structure and surface area of biochar were beneficial for improving soil aeration and water-holding capacities, whereas the input of labile organic compounds in biochar enhanced microbial respiration and accelerated soil O2 consumption, thereby reducing soil Eh (Zimmerman 2010; Yin et al. 2017). In addition, compared to the BC treatment, the soil Eh decreased more sharply in the FMBC treatment, indicating that the FMBC-incorporated paddy soil experienced a considerable reduction reaction throughout the rice growth (He et al. 2022). The variations in the soil solution pH and soil Eh confirmed here were consistent with findings reported by Liu et al. (2021), who demonstrated that ferromanganese could release higher concentrations of Mn2+ than pristine biochar, acting as a strong reductant that lowers soil Eh. Higher pH can promote the adsorption of HMs by soil surface colloids, while lower Eh is conducive to the formation of HgS and CdS deposits, thus reducing Hg and Cd mobility in soil (Sun et al. 2024).

Fig. 1
figure 1

Dynamics of soil solution pH (a) and soil Eh (b) at different rice growth stages. SS seeding stage, TS tillering stage, BS booting stage, GF grain filling, RS ripening stage

3.1.2 Soil pH and electrical conductivity (EC)

Soil pH and EC are crucial factors influencing the bioavailability of HMs (Fig. S2). Previous studies have found that increasing soil pH reduces the bioavailability of soil HMs (Sun et al. 2021). In the current study, the pH of the control soil was 6.19 (weakly acidic), whereas BC and FMBC application increased it by 0.1–0.36 and 0.24–0.75 units, respectively (P < 0.05). This may be because BC and FMBC materials contain alkaline substances and salt–based ions, such as K+, Ca2+ and Na+ which are released into the soil where they increase the soil pH (Jia et al. 2022). Additionally, the net negative charge on the biochar surface can bind to H+ to reduce soil acidity (Qiao et al. 2018). In accordance with the results of recent studies (Noronha et al. 2022), the application of both BC and FMBC also effectively increased soil EC, which may be due to the large amounts of soluble ions and ionizable groups in BC and FMBC (Lin et al. 2022a).

3.2 Effects of soil TOC, LOC fractions and CMPI under different treatments

Soil TOC is an essential biomarker for measuring the soil fertility level and maintaining agricultural ecosystem productivity (Zhu et al. 2015). Previous studies have found that the application of biochar has a considerable influence on soil TOC content (Dong et al. 2016; Munda et al. 2018). Shi et al. (2020) found that the addition of BC in a decade field experiment could increase SOC storage, and had a positive priming effect on native SOC which may help ameliorate climate change. As shown in Fig. 2a, the soil TOC content had a dose–effect relationship with BC and FMBC application, and reached the maximum value at a dosage of 2%. Compared with the control, the content of soil TOC significantly increased by 22.60–58.58% and 24.61–44.21% under the BC and FMBC treatments, respectively (P < 0.05). This may be because BC and FMBC are C-rich materials containing large amounts of condensed aromatic compounds, which are highly resistant to degradation and facilitate the accumulation of SOC (Jiang et al. 2022). Furthermore, BC and FMBC materials can promote soil aggregate formation by enhancing organic–mineral interactions, thereby preventing SOC decomposition by microbes (Weng et al. 2017). However, FMBC increased the organic C content to a lesser extent than the BC treatment, which may be related to the low C concentrations and high ash content of FMBC (Table 1).

Fig. 2
figure 2

The concentrations of soil organic carbon (SOC) (a), microbial biomass carbon (MBC) (b), dissolved organic carbon (DOC) (c) and easily oxidized organic carbon (ROC) (d) in different treatments. Notes: Error bars represent SD of three replicates. Different letters indicate significant difference (P < 0.05) from each other according to the Duncan's test

Table 1 Effects of BC and FMBC on soil C-pool activity (CPA), the C-pool index (CPI), the C-pool activity index (CPAI), and the C-pool management index (CPMI) in paddy field

The SOC pool is composed mainly of stable and labile organic carbon (Han et al. 2020). LOC fractions mainly include microbial biomass C (MBC), dissolved organic C (DOC), and readily oxidized carbon (ROC), which can directly influence soil C cycling by reflecting variations in soil C pools (Li et al. 2021c; Zhao et al. 2022). Soil supplementation with FMBC increased the soil LOC fraction (DOC, ROC and MBC) concentrations compared to the control (Fig. 2), and similar findings have been described previously (Jiang et al. 2022). As an important form of the LOC, although soil DOC content accounts for a small proportion of soil organic carbon pools, it plays a crucial role in soil ecosystems (Mustapha et al. 2021; Lin et al. 2022a). There are different opinions about the effects of biochar on soil DOC content, and previous research has revealed that biochar addition has no obvious influence on soil DOC, or even shows a decreasing trend (Yang et al. 2018; Jiang et al. 2022). Nevertheless, some studies have confirmed that biochar can enhance soil DOC concentration. For example, Li et al (2022c) observed that the DOC content significantly increased with increasing doses of biochar, and DOC could be immobilized on the biochar surface by complexing with the mobile forms of soil HMs. In this study, compared with the control, the incorporation of BC and FMBC significantly enhanced the content of oil DOC, which reached the maximum value (248.50 mg kg–1) under the 2%FMBC treatment (P < 0.05). The porous structure of BC and FMBC provided an ideal habitat for soil microbes, and the microorganism metabolism and the efficiency of C utilization were improved, so part of the LOC was decomposed by microorganisms, which increased the DOC concentration (Meng et al. 2018). The portion of soil C that is readily oxidized by KMnO4 is referred to as the ROC (Li et al. 2021b). Compared with control, the ROC concentration was significantly increased by 25.45–44.99% and 15.61–23.97% after adding BC and FMBC, respectively (P < 0.05), which may be related to the decomposition of easily oxidized components on the surface of biochar materials, as well as the increase in root exudates and microbial activity (Yang et al. 2018). Soil MBC is crucial for regulating nutrient cycling and microbially driven biogeochemical processes (Joergensen and Wichern 2018). After adding BC and FMBC, the MBC content significantly increased by 13.04–45.50% and 8.72–27.11%, respectively (P < 0.05). This could be because the amendments enhanced the C and N input into the soil, improved soil properties and fertility, and provided favorable habitat conditions for microbial community development and reproduction (Lin et al. 2022a).

According to the contents of SOC and its fractions, the indices of CPA, CPAI, CPI and CPMI were evaluated as shown in Table 1. The CPMI can effectively represent the variability of the soil organic C pool after agricultural management practices, which is influenced by the comprehensive effects of the soil C pool and C pool activity (Blair et al. 1995; Lin et al. 2022b). Generally, a higher CPMI reflects a higher quality and quantity of soil C pools, and can promote soil C sequestration (Sainepo et al. 2018; Ma et al. 2021). In the present study, the effects of the BC and FMBC treatments on soil CPA and CPAI were not significantly different from those in the control treatment (P > 0.5). After adding BC and FMBC, the soil CPI and CPMI were significantly greater than those of the control treatment (P < 0.05), and CPMI increased by 25.64–43.16% and 19.98–21.03% in BC and FMBC treatments, respectively. These results are in line with those of Ren et al. (2021). Furthermore, the soil CPMI values of BC and FMBC treatments were all greater than 100, indicating that BC and FMBC can enhance soil C storage and soil nutrient cycling function, and improve the beneficial development of soil (Zhao et al. 2022). Studies have shown that biochar can effectively enhance the levels of persistent organic C and protect newly produced C through the organic–mineral interactions in the soil (Wang et al. 2024). Moreover, BC and FMBC application may improve the formation of organic–mineral interactions, thereby enhancing the stabilization of SOC, which is of great significance for soil C sequestration (Nan et al. 2022; Giannetta et al. 2024).

3.3 Hg and Cd availability in soils and their accumulation in Oryza sativa L.

3.3.1 Soil Hg and Cd bioavailability and fractionation

The concentrations of available Hg and Cd leached by Na2S2O3 and DTPA solutions are shown in Fig. 3a and b. The incorporation of both FMBC and BC significantly reduced soil Na2S2O3–Hg and DTPA–Cd contents, which gradually decreased with an increase in the applied dose. Furthermore, FMBC incorporation resulted in a lower content of Na2S2O3–Hg and DTPA–Cd than BC incorporation at the same dosage, indicating that FMBC was more effective in soil Hg and Cd stabilization than BC. Compared with the control, the contents of Na2S2O3–Hg and DTPA–Cd significantly decreased by 41.49–81.85% and 19.47–33.02% in the FMBC treatments, respectively (P < 0.5).

Fig. 3
figure 3

The bioavailability concentration of Na2S2O3–Hg (a), DTPA-Cd (b) and soil Hg (c) and Cd (d) fractions under different treatments

Figure 3c and d shows the changes in soil Hg and Cd fractions under the BC and FMBC treatments. The soil Hg fraction was mainly the residual fraction, and the proportion of RES–Hg exceeded 85% in all treatments. In addition, FMBC addition increased the OM–Hg percentage, but had no significant effect on the other Hg fractions. For the Cd fractions, the percentages of EXC–Cd, CAR–Cd, OX–Cd, OM–Cd, and RES–Cd in the untreated soil were 30.47%, 14.86%, 16.88%, 3.21% and 34.58%, respectively. Compared to the control, the incorporation of FMBC reduced the EXC–Cd and CAR–Cd contents by 19.08–35.81% and 28.74–42.13%, respectively, whereas the proportion of OX–Cd, OM–Cd and RES–Cd contents increased by 14.02–23.07%, 6.97–23.40% and 7.61–37.73%, respectively. These results suggest that FMBC can stabilize soil Hg and Cd, and reduce their bioavailability, which is similar with the previous observation of Yang et al. (2022b), who discovered that Fe–Mn oxide modification increased the adsorption and immobilization ability of biochar for soil Cd and significantly decreased the content of available Cd. In our previous study, we found that FMBC has a higher pH and a larger specific surface area than pristine biochar, which can undergo a series of physicochemical reactions with Hg and Cd, such as chemical complexation, cation exchange electrostatic adsorption, precipitation and metal–π interaction (Sun et al. 2023). After incorporated on the surface of biochar, Fe/Mn oxides can provide more electrons to promote electron transfer and form monodentate or multidentate inner-sphere complexes with Hg and Cd ions (Yin et al. 2020). Moreover, the addition of FMBC increased soil pH and decreased soil Eh, further reducing the availability and mobility of HMs, thereby inhibiting the migration of Hg and Cd from the soil to rice plants (Li et al. 2022a). Furthermore, reductions in available Hg and Cd may be associated with the soil microbial community (Chen et al. 2022a).

3.3.2 Rice growth and HM content

Crop productivity is necessarily related to soil quality, hence well-managed soils can lead to sustainable production and improved rice grain yield (Zhu et al. 2015). The grain biomass under both BC and FMBC treatments was higher than that under the control treatment, with an increase of 19.02–57.38% and 9.98–45.23%, respectively (Table 2). However, the effects of the BC and FMBC treatments on the dry weights of the roots, stems and leaves were not significant. BC and FMBC addition provided high amounts of nutrients to the soil (e.g., C, N, P, and K) and facilitate the utilization and uptake of nutrients by crops, thereby promoting crop growth and enhancing yields (Yu et al. 2017). However, the increase in rice yield did not correspond with the rate of FMBC addition, indicating that excessive FMBC addition may not result in a higher rice yield; therefore the FMBC addition rate should be managed at the optimal level for practical application.

Table 2 Effects of BC and FMBC application on the dry weights of rice plant

Under long-term flooding conditions, Fe and Mn ions in the soil are oxidized in the rice rhizosphere, forming a reddish-brown IMP on the rice root surface (Zhou et al. 2022). Many factors affect IMP production on the root surface, including soil pH, Eh, radial oxygen loss (ROL), rice rhizosphere microbial activities, and soluble Fe and Mn ion contents (Wu et al. 2016). Previous research has shown that the IMP on the root surface can serve as an important barrier to prevent HMs from entering rice plants, and it plays a crucial role in reducing the accumulation of HMs in rice plants (Huang et al. 2022a; Zhou et al. 2022).

As shown in Fig. S3, high levels of IMP formed on the root surface following BC and FMBC supplementation. Among them, Fe was the main component of the IMP (13.21–27.31 g kg–1), which was much more abundant than Mn (46.06–1191.53 mg kg–1). The DCB–Fe concentration was greatly increased by 18.27–41.53% and 46.54–106.73% after BC and FMBC application in comparison with the control, respectively. Nevertheless, there was no significant difference in the DCB–Mn concentration between the BC and CK treatments (P > 0.5), whereas the DCB–Mn concentration significantly increased by 5.72-, 8.12- and 18.35-fold in the 0.5%FMBC, 1%FMBC and 2%FMBC treatments, respectively, in a dose-dependent manner (P < 0.5). This may have been caused by the release of Fe/Mn ions from FMBC, and these soluble Fe/Mn ions could have reacted with oxygen or oxidants released from rice roots to generate oxidized precipitates, thus promoting the production of IMP on the rice roots surface (Zhou et al. 2018b). After the addition of FMBC, a higher pH and the decrease in Eh favored the reductive dissolution of Fe/Mn oxides and enhanced the contents of Fe2+/Mn2+ (Wang et al. 2021a). Moreover, soil supplementation with FMBC can affect the activity of microorganisms by altering the soil microecological environment, and increasing the abundance of Fe-reducing bacteria, thereby promoting Fe3+ reduction and accelerating the development of iron plaques (Yu et al. 2016; Liang et al. 2022).

In the present study, the incorporation of FMBC increased the content of DCB–Cd in IMP by 21.55–60.79%, possibly inhibiting the uptake of Cd in rice. Furthermore, the content of DCB–Cd was significantly correlated with DCB–Fe and DCB–Mn in the IMP (P < 0.5) (Fig. S4). This result was similar to that reported by Liu et al. (2021), who discovered that the DCB–Cd content increased significantly with the enhancement of the IMP after the incorporation of ferromanganese biochar (FM–CSB) (P < 0.5), and Cd transfer from the soil to root tissue was obviously inhibited. This may be because FMBC boosted the formation of IMP, and more available Cd in the soil could be sequestered on the root surface by generating inner–sphere complexes (Zhou et al. 2018a; Zhang et al. 2020).

3.3.3 HM accumulation and translation in Oryza sativa L.

To further explore the immobilization efficiency of Hg and Cd by FMBC, the uptake of Hg and Cd by rice under different treatments was explored. The Hg and Cd concentrations in different tissues of rice showed obvious differences, in the following order: root > stem > leaf > panicle rachis ≈ rice huck, for all treatments (Fig. 4a and b). Rice roots intercepted most of the Hg and Cd, which is in accordance with the results of previous studies (Marrugo–Negrete et al. 2016; Chen et al. 2022b). Overall, soil application of BC and FMBC decreased both Hg and Cd contents in rice tissues, and the FMBC treatment had a more pronounced effect, suggesting a linear increase in food security, compared to the control plants. The Hg content in roots, stems, leaves, panicle rachis and rice husks decreased by 27.30–47.39%, 19.33–77.36%, 9.45–32.73%, 7.51–56.67% and 14.52–67.35%, respectively, and the Cd content decreased by 60.86–64.59%, 59.55–72.23%, 47.77–61.73%, 42.02–60.71% and 40.92–71.08%, respectively.

Fig. 4
figure 4

Concentrations of Hg and Cd in different parts of rice plant (Oryza sativa L.)

Figure 4c and d shows the Hg and Cd contents in brown rice under the BC and FMBC treatments. All treatments reduced the Hg and Cd contents in brown rice, with the most significant reduction in the 2%FMBC treatment. When compared with the control treatment, it was significantly (P < 0.05) reduced by 18.32%, 63.71% and 71.16%, respectively, for Hg content in brown rice, and 59.52%, 69.62% and 72.11%, respectively, for Cd content in brown rice under 0.5%FMBC, 1%FMBC and 2%FMBC. The reduction of Hg and Cd contents in grain under the FMBC treatments was obviously greater than that under the BC treatments, implying that FMBC was more effective in inhibiting Hg and Cd uptake in brown rice, thereby lowering phytotoxicity. These results also confirmed previous findings, as Irshad et al. (2022) found that goethite–modified biochar (GB) application markedly suppressed the uptake of As and Cd in rice grains compared to CK and BC treatments. Moreover, the Cd content in brown rice was lower than the permissible limit value after FMBC treatments (0.2 mg kg–1 Cd, GB 2762-2017), but the Hg concentrations still exceeded the national limit value (0.02 mg kg–1 Hg, GB 2762-2017) due to the severe exceedance of total Hg concentrations in the study area. Therefore, the application of FMBC combined with other remediation methods could be considered in future studies to reduce the total Hg concentration in brown rice to below the limit.

In addition, the bioconcentration factor (BF) and translocation factor (TF) were used to assess the migration characteristics of Hg and Cd from the soil to rice (Table 3). The incorporation of both BC and FMBC into the soil decreased the BFroot, BFstem and BFgrain, with the latter producing more prominent effects, and the magnitude of the reduction became more obvious with increasing passivator application amounts. The BFgrain of Hg and Cd was significantly reduced by 11.64–70.14% and 57.63–72.35% under FMBC treatments, respectively, compared to the control (P < 0.05). High-dose FMBC (2%FMBC) significantly inhibited the TFroot-stem of Hg and Cd, but had no significant impact on TFstem-grain relative to the control treatment. This study revealed that the decrease in brown rice content was attributable primarily to the fact that FMBC could reduce the content of available soil Hg and Cd which could be directly absorbed by the rice, and boost the adsorption performance of soil, thereby limiting the transfer of Hg and Cd from soil to rice plants. Furthermore, the addition of FMBC stimulated the generation of more IMP on the rice root surface, which helped to block HMs and reduce Hg and Cd uptake by rice plants (Sui et al. 2021). In conclusion, the incorporation of FMBC into Hg and Cd co-contaminated soils proved to be an effective strategy for inhibiting plant-available Hg and Cd uptake and transport to brown rice.

Table 3 Effects of different treatments on the BF and TF of Hg and Cd

3.4 Soil bacterial community under different treatments

3.4.1 Bacterial community composition and abundance

Soil microbial communities are recognized as one of the key indicators for assessing soil quality, and the diversity of soil microorganisms reflects the degree and function of biological activity in soil (Asma et al. 2024; Yang et al. 2021b). Previous research showed that biochar can affect the structure and activities of soil bacterial communities, with diverse effects depending on microbial communities according to the type of soil and the properties of biochar (Li et al. 2022b). As shown in Fig. 5a, all treatments shared 2439 core operational taxonomic units (OTUs), accounting for 49.06%, 48.00%, 50.51% and 53.82% of the total amount of OTUs in the control, 0.5%FMBC, 1%FMBC and 2%FMBC treatments, respectively. To evaluate variations in soil bacterial community structure between treatments, we examined beta diversity using principal coordinate analysis at the OTU level (Fig. 5b). The total variance explained by the principal coordinate axes PC1 and PC2 was 39.67% and 20.75%, respectively, for the bacterial communities in the seven treatments. The distance between the BC and FMBC treatment groups was relatively small when compared with the control treatment, suggesting that the bacterial community composition in the BC- and FMBC-amended treatments was more similar. However, the bacterial community composition and structure of the low-dose biochar and non-amended soils were similar, and the largest shifts were observed in the 2%FMBC treatment compared to the control. This was in accordance with previous findings (Li et al. 2022b), which showed that high-dose biochar was more likely to alter the bacterial community composition and structure than low-dose biochar.

Fig. 5
figure 5

Venn diagram showing differences in microbial community composition among different treatments (a); PCoA exhibiting the differences among different treatment groups (b); Relative abundance of bacterial composition at phyla level under different treatments (c); Heatmap of dominant bacterial genera under different treatments (d)

The present investigation found that the incorporation of BC and FMBC resulted in variations in the abundance of bacterial communities at the phylum and genus levels in the soil (Fig. 5c and d). The dominant phyla observed across all treatments included Firmicutes, Chloroflexi, Actinobacteria, Proteobacteria, Acidobacteria, and Bacteroidetes, respectively (relative abundance > 5%), which was consistent with the results of previous studies on rice paddy soils (Ji et al. 2022; Xu et al. 2022). These phyla represented more than 80% of the total bacterial total sequences, with relative abundances of 21.56–25.25%, 14.68–17.25%, 10.45–16.03%, 11.82–14.80%, 8.09–10.86%, and 3.62–7.47%, respectively, across all treatments. Overall, soil supplementation with FMBC enhanced the abundance of Firmicutes, Proteobacteria and Actinobacteria, and reduced the abundance of Acidobacteriota, Chloroflexi and Bacteroidota. The relative abundance of Firmicutes, Proteobacteria and Actinobacteria was the highest under 2%FMBC treatment, which increased by 52.21%, 16.43% and 30.54%, respectively, compared with the control. Firmicutes is a eutrophic bacterial group that exists in anoxic environments and can effectively decompose anaerobic and facultative anaerobic microorganisms into organic matter. The highest abundance of Firmicutes in all samples may be attributed to the constant anoxic conditions of the rice paddies under flood irrigation. The addition of BC and FMBC provided additional C and N energy sources for soil microorganisms, creating more favorable conditions for Firmicutes growth. Additionally, an increase in the abundance of Firmicutes was conducive to soil C and N cycling processes (Tan et al. 2019; Ji et al. 2022). Proteobacteria were regarded as a group of highly metal-tolerant bacteria that prefer to survive in nutrient-rich soil environments (Song et al. 2024). The addition of FMBC input labile C and fertilizer into the soil to provide nutrients for the growth and survival of microbes, thereby boosting the abundance of Proteobacteria (Yao et al. 2017; Dong et al. 2018). The phylum Actinobacteria was also one of the main bacterial taxa under each treatment and is typically associated with biogeochemical cycling processes of soil C. Compared with the control, the higher abundance of Actinobacteria in BC- and FMBC-amended soils may be related to their capacity to break down recalcitrant carbon compounds (Yan et al. 2021). Acidobacteria is a group of acidophilic bacteria that play a vital role in the re-immobilization of HMs (Wang et al. 2020c; Zhang et al. 2022). The application of BC and FMBC led to a decrease in the relative abundance of Acidobacteria, which may help to inhibit Hg and Cd activation, because Acidobacteria can activate heavy metals by secreting acidic substances (Xu et al. 2022). In addition, a clustered heat map of the bacterial community at the genus level showed that the application dose drastically affected the bacterial community composition and structure, and the high-dose biochar and modified biochar treatments (2%BC and FMBC) had a greater effect at the genus level in comparison with the control (Fig. 5d). After the addition of BC and FMBC, the dominant genera of Bacillus, Clostridium, Bacteroidetes, Vicinamibacterales and Anaerolineaceae changed markedly. Clostridium is one of the most prevalent iron–reducing genera in soil, predominantly using starch or glucose as a C source (Lyu et al. 2022). The reduction in the relative abundance of Clostridium may be mainly due to changes in the soil pH, Eh, and organic matter concentration (Li et al. 2021a). Moreover, Bacillus can produce extracellular polymers to adsorb heavy metals in soil. Our results showed that high–dose modified biochar (2%FMBC) increased the relative abundance of Bacillus, which was beneficial to alleviating the toxicity of heavy metals (Sneh et al. 2021; Ma et al. 2023).

3.4.2 Bacterial community diversity

Microbial diversity affects the structure, function and stress resistance of subsurface ecosystems, and its variation is influenced by various stress variables such as soil nutrients and environmental conditions (Yang et al. 2022a). Figure 6 shows the effects of applying BC and FMBC on the bacterial alpha diversity index. The coverage of all treatments ranged from 95.03 to 96.15%, indicating that the current sequencing depth of gene sequences in the soil samples is representative of the real situation of the microbial community (Liu et al. 2020). For example, Azeem et al. (2022) found that the incorporation of cow bone-derived biochar (CB) reduced the alpha diversity of soil bacterial communities but increased the biomass and genetic abundance of specific functional bacteria at the phylum level, thereby promoting their growth and development. However, other studies suggested that biochar can increase the diversity and abundance of microbial communities to some extent (Yao et al. 2017; Sheng and Zhu 2018). In our study, the addition of BC and FMBC had no significant effect on the ɑ-diversity index of bacterial communities compared with the control treatment (P > 0.05), where the Simpson, ACE and Chao1 indices slightly decreased, while the Shannon index was slightly increased, indicating that BC and FMBC addition may be beneficial to the diversity and evenness of soil bacterial communities, but not its richness.

Fig. 6
figure 6

Alpha diversity indices of soil bacteria in different treatments

3.4.3 The functional prediction of soil bacterial communities

The PICRUSt bioinformatics software has been widely used to evaluate the potential metabolic potential of bacterial communities in soil (Pang et al. 2022). Overall, the types of predicted functional genes were similar among different treatments (Fig. 7). According to the KEGG database, there were six main functions, i.e., metabolism, genetic information processing, cellular processes, human diseases, organismal systems and environmental information process on KEGG pathway level 1, of which metabolism function accounted for 77.54–77.91% in all treatments. The predictive function of soil bacteria on KEGG pathway level 2 revealed 46 predominant sub-functions, including global and overview maps, signal transduction, degradation and metabolism of exogenous substances, substance dependence, transportation and catabolism, and environmental adaptation. The functions of bacteria related to soil metabolism (such as energy metabolism, amino acid metabolism, glycan biosynthesis and metabolism) were significantly increased in the FMBC-supplemented soil. FMBC application can promote the utilization of soil carbon, thereby improving soil bacterial metabolic activity (Hong et al. 2022). Furthermore, amino acids can immobilize soil-available HMs by forming stable complexes (Li et al. 2020; Zhang et al. 2022). These findings suggest that the improvements in bacterial metabolic function are conducive to the stability and health of the soil ecosystems.

Fig. 7
figure 7

The functional prediction of soil bacterial communities under different treatments (a) KEGG pathway level 1, (b) KEGG pathway level 2

3.5 Correlations between environmental factors and bacterial communities

Mantel tests were performed to investigate the relationship between the bacterial community composition and environmental factors (Fig. 8a). In this study, the soil variables EC, TOC, DOC, Na2S2O3–Hg and DTPA–Cd were the main factors affecting the bacterial community structure and function (P < 0.05). In particular, bacterial communities were significantly correlated with TOC (P < 0.01). Furthermore, the contents of Hg and Cd in grains had a significant negative correlation with soil pH, DOC and DCB–Fe and DCB–Mn content (P < 0.01) and a significant positive correlation with Na2S2O3–Hg and DTPA–Cd contents (P < 0.01). Consistent with previous results, biochar and modified biochar materials can reduce the soil-available HM concentrations and transform them into stable compounds that are not readily adsorbed by crops (Chen et al. 2022a).

Fig. 8
figure 8

Mantel tests depict the association of bacterial community composition and environmental factors (a); The partial least square path models (PLSPM) describing the effects of the key factors on the Hg and concentrations of rice grain (b). Red and green arrows represent negative and positive correlations, respectively; The solid-line path indicates that the effect is significant, and a dashed-line path indicates that the effect has no significance; C fraction refers to soil TOC and labile organic C fractions; Bacterial community data are derived from the relative abundance of top 10 bacterial phyla; GoF the goodness of fit of the model

The partial least square path modeling (PLSPM) is an efficient statistical technique that was used to analyze the direct and indirect effects of FMBC on Hg and Cd concentrations in rice grains (Cui et al. 2016; Xu et al. 2022) (Fig. 8b). PLSPM analysis demonstrated that the incorporation of FMBC changed Hg and Cd contents in rice grains mainly via affecting the contents of Na2S2O3–Hg, DTPA–Cd and IMP in the roots surface, with changes in Na2S2O3–Hg and DTPA–Cd attributed to the variation in pH, EC, TOC, labile organic C fractions, and changes in Na2S2O3–Hg and DTPA–Cd influenced bacterial community composition and function. The above findings show that the FMBC may have greater potential for reducing Hg/Cd uptake in rice tissues through the passivation effect of soil Hg/Cd and the barrier effect of IMP on root surfaces.

4 Conclusion

This study showed that the addition of BC and FMBC significantly decreased the concentrations of Na2S2O3–Hg and DTPA–Cd by 41.49–81.85% and 19.47–33.02% under the FMBC treatments compared with the control, respectively (P < 0.5), whereas the IMP content on the rice root surface increased significantly. Furthermore, FMBC incorporation caused lower Hg and Cd contents in rice grains than BC at the same dosage, indicating that FMBC was more effective in remediation of Hg and Cd polluted soils. Increases in SOC, LOC fractions (DOC, MBC, and ROC) and CPMI could further improve soil C sequestration and stability. Although FMBC had no significant influence on the alpha diversity index of the bacterial communities, it improved the relative abundance of specific phyla and their metabolism potential. Mantel tests results revealed that EC, TOC, DOC, DTPA–Cd and Na2S2O3–Hg were the main factors affecting bacterial communities. The results of our study revealed that FMBC can be employed as an efficient material to remediate Hg and Cd co-contaminated paddy soils and restrict their uptake by rice grains, thereby alleviating food security issues. Further research is needed to explore the long-term effects of FMBC in reducing Hg and Cd contents in paddy soils and rice plants under actual field conditions.