Field Verication of Low-Level Biochar Applications as Effective Ameliorants to Mitigate Cadmium Accumulation into Brassica Campestris L from Polluted Farmland Soil

Farmland soils in China have been reported to be diffusely contaminated, Cd has been recognized as a signicant contributor to this issue and biochars have been reported to be effective in mitigating soil Cd pollution. However, previous studies have shown contradictory outcomes. Furthermore, in general, laboratory experiments and unrealistically large amounts of biochar (>10 t/ha) have been used. In this study, three biochars: rice straw biochar (RS), pig manure biochar (PM) and rice husk biochar (RH) were produced from readily available farm residues and characterized. These were used in a eld experiment, at low applications rates of 1.8 and 3.6 t/ha, with rape (Brassica campestris L.). Batch adsorption experiments indicated Cd adsorption in the order RS biochar > PM biochar > RH biochar. Field experiment indicated biochar amendments to slightly changes in soil pH and CEC; yet led to considerable and signicant decreases in extractable Cd concentrations (reductions of: 43%-51% (PM), 29%-35% (RS) and 17%-19% (RH)). Reduced extractable Cd correlated with lower Cd concentrations in rape plants. PM and RS biochars were the most effective in decreasing Cd phytoaccumulation into edible parts of rape (>68% reduction). It is highlighted that biochars were produced using a pyrolysis unit with an output of 20 ton/yr. Thus, assuming a working application rate of 2 ton/ha, the pyrolysis unit could service 10 ha/yr. While at a modest scale, this research demonstrates the genuine reality of biochar-based remediation solutions to contribute to the mitigation of diffuse Cd contamination in some of China’s impaired farmland. Cd in soil Doubling eld application rates of biochar increased Cd immobilization by 10%


Introduction
Due to industrialization and urbanization, soil cadmium (Cd) pollution is a global issue (Liao et al. 2015).
Often this pollution is localized in areas of industrial activity. However, long range transport from industrial sources and the use of tainted water resources for irrigation have resulted in farmland accumulating high levels of contaminant (Lu et al. 2015a). In China the latest o cial soil surveys conclude that 19.4% of national farmland soils (based on the sampling points) has been contaminated (MEP 2014). Cd, along with Hg, As, Cu, Pb, Cr, Zn and Ni, are recognized to be priority risk drivers. Cd can migrate from contaminated soils to food crops, which can signi cantly increase human health risk associated with Cd poisoning (Ochoa et al. 2020). There is therefore an urgent need to develop e cient and scalable techniques for the remediation of farmland diffusely polluted with potentially toxic elements (PTEs).
Biochar is the product of the pyrolysis of organic solid wastes under anoxic or limited oxygen conditions. Biochars can be produced from a wide range of organic feedstocks including crop straw, manure, wood chip, and sewage sludge . Generally, biochar is alkaline, and contains abundant surface functional groups (Kloss et al. 2012). Several types of biochars have been reported to be e cient in immobilizing heavy metals through mechanisms such as adsorption, surface precipitation and electrostatic interaction (Chen et al. 2015; Alam et al. 2018). Biochar in uence on Cd availability has been extensively studied over the last decade. For example, Lu  concentrations from soils amended with orchard residue biochar to decrease by > 70%, while for r tree biochar, Cd extractability decreased by only 12%.
As brie y outlined above, most of these studies were conducted in laboratory or greenhouse settings. However, there is to date little information and guidance on eld-scale in-situ applications of biochar. Moreover, where eld trial results have been reported they have shown inconsistent outcomes, (Cui et Rondon et al. 2014). Given the availability of biochar resource and the associated cost of biochar application, such large rates of biochar application at a eld scale would likely be unfeasible. Importantly, it has been reported that decreases in metal availability, following biochar application, are not proportional to the amount of biochar applied (Bian et al. 2014); and, in Jeffery's meta-analysis (1483 studies), a poor relationship between biochar application rate and crop yield was reported (r 2 = 0.1; Jeffery et al. 2011). It might therefore be concluded that high application rates of biochar may not be necessary to achieve desired outcomes. Thus, there a need to evaluate biochar e cacy to mitigate Cd phytoaccumulation using low application rates of biochar under eld relevant conditions.
In this study, three biochars including rice straw biochar (RS), rice husk biochar (RH) and pig manure biochar (PM) were prepared and evaluated for their Cd adsorption capacities in the laboratory. The biochars were then used, in a eld experiment at two low application rates (1.8 and 3.6 t/ha), under a crop of rape (Brassica campestris L.), to: i) evaluate their in uence upon the accumulation of Cd from soils into the rape roots and edible parts; ii) to explore the mechanisms in uencing Cd phytoaccumulation, and; iii) to provide evidence to assist the selection of biochar feedstock and biochar application rate for utilisation to abate risks associated with Cd pollution in soil.

Biochar preparation
Rice straw biochar (RS), rice husk biochar (RH) and pig manure biochar (PM) were produced using pilotscale biochar pyrolysis equipment (L 4.5 m ⋅ W 0.8 m ⋅ H 2 m). The pyrolysis equipment comprised a pyrolysis chamber, a heating chamber, a feedstock supply system and a biochar output system. The equipment had an overall production capacity of 20 ton per year. Feedstocks were pyrolyzed in the pyrolysis chamber at 500°C for 1 hour under limited oxygen conditions. Thereafter, the biochars were ground through a 2 mm sieve. The physico-chemical properties of the three biochars were assessed using methods described in previous work (Zhang and

Field study
The eld experiment was conducted on an upland farm located in Zhouzai village (24°23′26.08″N, 117°43′26.25″E), Zhangzhou city, Fujian province in southern China. Local soils had been contaminated with potential toxic elements (PTEs); Cd being the primary pollutant. The concentration of Cd in the soil was 0.38 mg/kg; this concentration exceeded the regulatory limit for agricultural soils (0.30 mg/kg; MEP 2018).
In November 2016, soil (0-20 cm) was ploughed. Biochars were then spread on the soil surface, and thoroughly mixed with the soils using a tillage machine (1GQN-120; Weifang shengxuan machines corporation). The applications of biochar were 0, 1.8 and 3.6 t/ha, respectively. Each treatment was produced in triplicate. The area of each plot was 4 m 2 . Two weeks after biochar application, the land was rolled, and rape (Brassica campestris L.) seeds were sown. Irrigation and fertilizer management were performed according to the conventional practices of the local farmers and were identical on all plots.

Sampling and analysis
In January 2017, the rape plants were harvested. Three composite rape samples, each consisting of 10 plants randomly selected from each plot, were collected (resulting in nine composite rape samples for each treatment regime). Following their transfer to the laboratory, the rape samples were washed with deionized water. Thereafter, the rape samples were cut into two parts (the edible part and the root), and dried in the oven at 90°C. For the analysis of Cd concentrations in the plants, subsamples of the edible part, or root, were crushed, ground, and passed through a 0.2 mm sieve. Samples of the edible part or root were then digested with HNO 3 (65%, Merck, EMSURE™) using a protocol modi ed from Zhu et al. (2008).
For quality control, plant reference material (GBW 10015; purchased from the National Research Center for Standards in China) was digested and measured alongside the experimental samples, as were reagent blanks. The concentration of Cd in the digestates were analysed using an inductively coupled plasma mass spectrometry (ICP-MS 7500cx; Agilent Inc.). Concentrations of Cd were reported on dry weight basis. The recovery of Cd from the reference material was 108-113%.
Nine soil samples (0-20 cm) were collected from each treatment. Soil samples were air-dried in the laboratory at room temperature, and then ground and passed through a 2 mm sieve. Subsamples of soils were ground and passed through a 0.15 mm sieve. The pH of soil samples was measured using a pH meter at the ratio of 1 g soil : 2.5 mL deionized water. The cation exchangeable capacity (CEC) of soil was measured using the NH 4 OAc-NaOAc method (U.S. Environmental Protection Agency 1986). The concentrations of available Cd in soils were determined using a 0.01 mol/L CaCl 2 extraction method (Houba et al. 1996). Brie y, soil samples were extracted with 0.01 mol/L CaCl 2 solution at the ratio of 1 g soil : CaCl 2 solution (10 mL). After 2 hours shaking, the suspensions were centrifuged, and the concentrations of Cd in the supernatant were measured using inductively coupled plasma mass spectrometry (ICP-MS 7500cx; Agilent Inc.).

Data management and Statistic analysis
All data were analysed using Microsoft Excel. The tting of batch adsorption data was performed with OriginLab 2018 software (OriginLab Corp., USA). Statistical analysis of the data was performed with IBM SPSS Statistics 22.0 software (IBM Corp., USA). Signi cant differences among treatments were analyzed using one-way analysis of variance (ANOVA) (at p < 0.05 level).

Biochar characterisation
Properties varied across the three biochar types (Table 1). pH was always alkaline and ranged between 9.87 and 10.7; RS biochar had the highest pH (10.7). Carbon content in PM biochar was 20%, while the C content in RS biochar or RH biochar was much higher (> 40%). N and O contents in PM biochar were the highest amongst the three biochars (1.5% and 20%, respectively), while the N and O contents in RH biochar were lowest (0.6% and 11%, respectively). The ash content of PM biochar was particularly high (68%; this accounting for the lower C content observed). The ash contents in the RS and RH biochars were much lower, 36% and 35%, respectively. The surface topography and pore structure of the biochars also varied (Table 1 and Figs. S1-S2). The surface area of biochars followed the order of RS biochar > PM biochar > > RH biochar, and the average pore volume of biochars followed the order of PM biochar = RS biochar >>> RH biochar. The average pore size of PM biochar (12.7 nm) was the highest of the three biochars; average pore size in the RS and RH biochars were similar (~ 9.5 nm). Based on these chemical and physical properties, it was hypothesised that the RS biochar (with high pH, high carbon content and  Results of the batch equilibrium experiment indicated that Cd adsorption capacity of the three biochars followed the order: RS biochar > PM biochar > RH biochar ( Fig. 1 and Table S1). Cd sorption onto both PM biochar and RH biochar were better tted by the Freundlich isotherm model (rather than Langmuir isotherm model), while the RS biochar was tted well with both the Langmuir and Freundlich isotherm models (Table S1). Given the model agreements, it was concluded that Cd adsorption onto the three biochars was likely attributable to chemical adsorption on heterogeneous surfaces (Zhang and Luo 2014). The value of 1/n in the Freundlich isotherm model followed the order of RS biochar < PM biochar < RH biochar. Since 1/n represented the adsorption a nity of metal ions onto adsorbent, this result con rmed that the adsorption of Cd 2+ ion onto RS biochar was the greatest, while RH biochar showed lowest a nity for Cd ions.
Available Cd concentration in the control soil was 21.5 ± 1.9 µg/kg. Following biochar amendment, the available Cd concentrations were signi cantly lowered in all treatments (p < 0.05). However, greatest reductions in bioavailable Cd in soil were observed with PM amendment (43% and 51% reduction at 1.8 and 3.6 t/ha) and RS amendment (29% and 35% reduction at 1.8 and 3.6 t/ha) (Fig. 3). In the treatments with RH biochar amendment, bioavailable Cd concentration was reduced by only 17% and 19% in the 1.8 t/ha and 3.6 t/ha treatments, respectively.

Effects of biochar amendment on the growths of rape plants in led
Biochar amendment in uenced the growth of rape plants. Increases in dry weight of edible parts (Fig.  S3a) and dry weight of roots (Fig. S3b) were observed in all instance following biochar amendment.
However, these increases in biomass were marginal, and none were signi cantly different (p > 0.05) when compared to the control (except for the dry weight of root, in the treatment with 1.8 t/ha amendment of PM biochar).

Effects of biochar amendment on Cd accumulation into plants
In all instances, biochar amendment resulted in a decrease in the concentration of Cd in rape plant (Fig. 4). For the edible part of rape (Fig. 4a), the concentrations of Cd in the treatments with PM biochar were signi cantly (p < 0.05) decreased to 0.52 ± 0.08 mg/kg (1.8 t/ha) and 0.50 ± 0.03 mg/kg (3.6 t/ha); these levels were 70% and 68% of the control value. In the RS treatments, the Cd concentrations were signi cantly (p < 0.05) decreased to 0.51 ± 0.09 mg/kg (1.8 t/ha) and 0.51 ± 0.02 mg/kg (3.6 t/ha), respectively; these values were 70% of the control value. Following RH amendment, the concentrations of Cd in the edible part of rape were 0.66 ± 0.04 mg/kg (1.8 t/ha) and 0.67 ± 0.03 mg/kg (3.6 t/ha); these values were approximately 89% of the control value.
Regarding Cd concentrations in roots, treatments with PM biochar indicated decreased concentrations (0.60 ± 0.12 mg/kg (1.8 t/ha) and 0.55 ± 0.04 mg/kg (3.6 t/ha), compared to 0.74 ± 0.30 mg/kg in the control (Fig. 4b). Similarly, the Cd concentrations in the treatments with RS biochar and RH biochar were also observed to decrease (to 0.50 mg/kg − 0.52 mg/kg and 0.51 mg/kg − 0.65 mg/kg, in 1.8 t/ha and 3.6 t/ha, respectively). Where application rate was equivalent, no signi cant differences (p > 0.05) were observed across the three biochar types.

Discussion
The C content of the RS and RH biochar (~ 42%) was more than 2 times that of the PM biochar; while the content of N in PM biochar was 1.9 times the content in RS biochar, and 3.3 times the content in RH biochar (Table 1). In agreement with previous reports, the biochars prepared from manure contained lower C and higher N than the biochars prepared from plant residues (Xu et al. 2013;Zornoza et al. 2016). Such differences have previously been attributed to differences in the element composition of feedstocks ). In addition, the three biochars had different surface topography and pore structure (Fig. S1). The surface area, average pore volume and average pore size of PM biochar and RS biochar were all higher than those for RH biochar (Table 1) . As a consequence of differing physical and chemical properties, the sorption capacities of the three biochars varied (Fig. 1). Complexation of metals, through ion exchange interactions, with ionized surfaces and oxygen-containing functional groups (i.e. carboxyl (-COOH), hydroxyl (-OH), phenol (R-OH) groups) has been suggested as an important mechanism for metal sorption by biochar . In this study, RS biochar had highest surface area and highest content of surface functional groups (Table 1), available to prompt interactions with metal ions. Similarly, since the BET surface area and the content of surface functional groups of RH biochar were lowest among three biochars (Table 1), the sorption of Cd 2+ ions onto RH biochar was correspondingly lower. In addition, the pH of three biochars also in uenced the equilibrium pH and thus the adsorption of Cd ions (Fig. S4). The sorption of more Cd 2+ ion with increasing pH is consistent with Zhang and Luo (2014) who also reported this relationship.
In the batch experiments, the addition of RS biochar with highest pH lead to highest equilibrium pH among three biochars (Table S3); this, likely, underpinned the greatest sorption of Cd 2+ ions onto RS biochar. Overall the RH equilibrium pH was lower than RS or PM, and Cd 2+ ion sorption was also lower.
When amended to soil, biochars increased soil pH in all instances, but the increases were not signi cantly different (p > 0.05) among the three biochars (Fig. 2a). In addition, biochar had limited in uenced on soil CEC (Fig. 2b). These outcomes are likely due to the low amendment levels (1.8 and 3.6 t/ha).
Nonetheless, all three biochars signi cantly (p < 0.05) reduced extractable Cd concentrations (decreases followed the order: PM biochar > RS biochar > RH biochar) (Fig. 3). It is suggested, therefore, that changes in Cd availability were most likely linked to Cd ion interaction with biochar (rather than changes to the soil chemical environment). Reduced concentrations of available Cd were translated into observed reductions in Cd content in rape plants ( Fig. 4 and Table S2). Importantly, the three types of biochar led to different outcomes for Cd-plant interactions. Treatment with RH biochar was relatively ineffective, while amendment with PM or RS biochar resulted in much more effective abatement of soil to plant transfer of Cd. Very little difference was observed where 1.8 and 3.6 t/ha application levels of the same biochar were compared. This observation suggesting, even at the lowest application rate (1.8 t/ha), that PM and RS biochars were effective ameliorants. The literature, in many cases, has reported metal-biochar-soil However, in the present study, RS and PM biochars (applied at low-levels: 1.8 and 3.6 t/ha) were established to be effective for the control of Cd phytoaccumulation, while RH biochar was observed to have only a limited effect.
While Cd sorption capacity of RS biochar was higher than that of PM biochar in the batch adsorption experiment, there were no signi cant difference between the decreased magnitude of Cd concentrations in rape plants where RS and PM treatments were compared (Fig. 4). These results highlight that the performance of these biochars in the batch adsorption experiments and in real soil systems were not consistent. This outcome is consistent with Uchimiya et al. (2010) who reported biochars produced from broiler litter manure at 350 °C (350BL) removed more Ni 2+ and Cd 2+ ions than biochars produced at 700°C (700BL); but when applied (at 5% − 10 % (w/w)), to soil the 350BL treatment contained higher soluble metal concentrations than the 700BL treated soils. Although the 350BL had a higher adsorption capacity than the 700BL, its lower ability to increase soil pH underpinned the less effective immobilization of metals in soil by 350BL. Similarly, the pH increase, rather than primary Cd-biochar interaction, has also been proposed by other researchers (Houben et al. 2013;Rees et al. 2014). Thus, it is recommended that, before eld scale deployment, biochar sorption capacity should be established in the presence of the soil it is intended to remediate.
In the present study, the pH of biochar amended soils were increased by 0.2-0.4 unit, while slight (although non-signi cant) changes in soil CEC were observed following biochar amendment. The minimal affect is most likely due to the low level of biochar applied (1.8 t/ha − 3.6 t/ha). Given that soil chemical properties (pH and CEC) were largely unchanged it is suggested that the primary interactions between Cd and the biochars were likely responsible for the outcomes observed.
The application of RS biochar or PM biochar at low level (1.8 t/ha) was effective to mitigate the transfer of Cd from soil into crop. With a doubling in application rates, the decrease of soil available Cd concentrations were increased (on average by approximately extra 10%), but the concentrations of Cd in rape showed limited change. It is highlighted that in the present research, the soil was not heavily contaminated. The soil Cd concentration (0.38 mg/kg) only just exceeding the regulatory limit of 0.30 mg/kg (MEP 2018). It is therefore emphasized that the low application rates of bichar were directed at a small excess of Cd in the soil system (this likely underpinned the successful outcomes observed) and that the lower application rate was su cient to accommodate the excess of Cd. It follows that should soil Cd concentrations are much higher, such an outcome might not transpire and larger applications of biochar could be needed to accommodate a greater excess of Cd in the soil system. This said, Nie et al. In contrast, the present study considered biochars produced using a larger pyrolysis system. The pilot scale pyrolysis system used to produce the biochars for this present research had an output capacity of 20 ton per year. Thus, assuming an application rate of 2 t/ha, such a unit could service 10 ha p.a. It is emphasised that this scale of production, and low application rates, represents a realistic approach to support the production of biochar in quantities that would allow for meaningful eld scale application. Given the extent of diffuse pollution associated with a considerable proportion of China's farmland (Lu et

Conclusions
Biochars derived from different feedstocks, under identical pyrolysis conditions, showed different sorption capacities for Cd 2+ ions. Following amendment to soil, even at low application rates (1.8 t/ha), all biochars were observed to reduce the available concentrations of Cd in soil; slightly higher Cd immobilisation was observed at 3.6 t/ha (an average increase of 10%). These changes in Cd availability resulted in decreased Cd concentrations in rape plants. Biochars derived from pig manure (PM) or rice straw (RS) led to much lower Cd concentrations in rape plants when compared to outcomes for rice husk (RH) biochar. Results underline that favorable Cd-biochar sorption capacities, established in the absence of soil under laboratory batch sorption conditions, did not necessarily translate into comparable success in metal pollution amelioration under eld conditions. These results highlight the need to trail biochars, in the presence of the soil to be targeted for remediation, before full scale deployment is undertaken. Importantly, this research has validated an approach, that is relevant in terms of both biochar production rate and application rate, for meaningful engagement with the amelioration of Cd-tainted farmland in China at a realistic scale.

Declarations
Author contribution C Cai and YC Zhang conceived and designed the study; JJ Fan and SN Lin performed the experiment; YW Hou analyzed the data; YC Zhang drafted the manuscript; BJ Reid and F Coulon revised the manuscript; C Cai approved the nal version of manuscript.

Funding
This work was nancially supported by grants from National Key R&D Project (2018YFC1802703, 2016YFD0800706) and the Science and Technology Project of Xiamen city (3502Z20182001).
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
The data that support the ndings of this study are available from the corresponding author upon reasonable request.  Changes in the properties of soil and soil amended with biochars. "*" indicates a value to be signi cantly different to the control (CK) value (p < 0.05). Like uppercase letters indicate no signi cant difference (p > 0.05) between groups with biochar amendment at like application rates.

Figure 3
Available Cd concentrations in soil and soil amended with biochars. "*" indicates a value to be signi cantly different to the control (CK) value (p < 0.05). Like uppercase letters indicate no signi cant difference (p > 0.05) between groups with biochar amendment at like application rates.