Using biochar for environmental recovery and boosting the yield of valuable non-food crops: The case of hemp in a soil contaminated by potentially toxic elements (PTEs)

2.2. Physicochemical characteristics of soil

At the end of the incubation period, the main physicochemical characteristics of soil mesocosms were determined. Soil pH and electric conductivity (EC) were quantified on 1:2.5 (w/v) soil:water suspensions [25] (Table 1). Dissolved organic carbon (DOC) was determined following Manzano et al. [24], total organic C and N were quantified with a CHN analyzer (Leco CHN628) and Soil LCRM Leco part n° 502–697 as calibration sample. Extractable P was quantified with the Olsen method [25], while CEC and exchangeable bases (i.e., Na, Ca, K, Mg) were determined with the BaCl₂-triethanolamine method [25].

The pseudo-total concentration of PTEs (i.e., Cd, Cr, Cu, Ni, Pb, Sb, and Zn) was determined by an Inductively Coupled Plasma Optical Emission Spectrometer (PerkinElmer Optima 7300 DV ICP-OES) after microwave digestion (ultraWave Microwave Milestone) of soil samples (0.2 g) with 2 mL of suprapure H₂O₂, and 4 mL of a HNO₃ + HCl (3:1 v/v ratio) mixture. The NIST-SRM 2711 certified reference soil was included for quality assurance, with PTE recoveries between 8.5 and 10% of the certified values. The detection limits of the analyzed elements were between 2.63 μg L⁻¹ (As) and 0.14 μg L⁻¹ (Cd).

2.3. PTEs mobility in soil mesocosms

Only the mobility of those PTEs whose concentration exceeded the contamination thresholds established by the Italian regulations [26], i.e., Sb, Cd, Pb, and Zn (Table 1), was assessed.

To quantify the Sb mobility in soil mesocosms (after the incubation period), the sequential extraction of Wenzel et al. [27] was used. In brief, the following Sb fractions were quantified: the water-soluble and non-specifically sorbed Sb (Fraction 1; extracted with a 0.05 M (NH₄)₂SO₄ solution); the specifically sorbed Sb (Fraction 2; extracted with a 0.05 M NH₄H₂PO₄ solution); the Sb bound to amorphous and poorly crystalline Fe and Al hydroxides (Fraction 3; extracted with a 0.2 M NH₄⁺-oxalate buffer solution); the Sb bound to well-crystallized Fe and Al hydroxides (Fraction 4; extracted with a 0.2 M NH₄⁺-oxalate buffer+ 0.1 M ascorbic acid solution). Additional methodological details can be found elsewhere [28].

The sequential extraction of cationic PTEs was carried out following the protocol described by Basta and Gradwohl [29]. In brief, the following PTEs fractions were quantified: the water-soluble and readily exchangeable PTEs (Fraction 1; extracted with a 0.5 M Ca(NO₃)₂ solution); PTEs forming weak surface complexes (Fraction 2; extracted with a 1 M NaOAc solution at pH 5.0); surface complexed and precipitated PTEs (Fraction 3; extracted with a 0.1 M Na₂EDTA solution). Additional methodological details can be found elsewhere [30].

The residual PTEs fractions were determined, for both the sequential extraction procedures, after the mineralization of the soil samples resulting from the different extraction with 2 mL of suprapure H₂O₂ and 4 mL of a mixture of HNO₃ + HCl (3:1 v/v ratio), and microwave digestion (ultraWave Microwave Milestone). The PTEs present in each liquid fraction were quantified by ICP-OES (PerkinElmer Optima 7300 DV ICP-OES). The NIST-SRM 2711 certified reference soil was included for quality assurance, with PTE recoveries between 8.5 and 10% of the certified values. The detection limits of the analyzed elements were between 2.63 μg L⁻¹ (As) and 0.14 μg L⁻¹ (Cd).

2.4. Enzyme activities, soil basal respiration and microbial carbon

After incubation, selected biochemical and microbiological characteristics were determined in soil mesocosms. In particular, the dehydrogenase (DHG), urease (URE), β-glucosidase (GLU), and acid and alkaline phosphomonoesterase (PHA and PHB) activities were quantified colorimetrically following the procedures described by Alef and Nannipieri [31].

To monitor soil basal respiration (CO₂ efflux), two polyvinyl chloride (PVC) collars (20 cm inner diameter) were placed in each mesocosm (n = 12 total PVC collars), and CO₂ efflux was quantified for each collar with a portable Li-8100 (LI-COR Inc., Lincoln, NE, USA). Measurements took 105 s per collar and were taken weekly throughout the incubation time between 5:00 and 7:00 p.m. (local time, GMT + 1) to avoid direct radiation from reaching the collars and to minimize soil temperature variability during sampling. The temperature at a 5 cm depth adjacent to the collars was recorded simultaneously with CO₂ efflux measurements using a probe connected to the LI-COR 8100. Each collar’s headspace was measured and considered to quantify CO₂ efflux rates as μmol CO₂ m⁻² s⁻¹.

After soil incubation with biochar, soil microbial biomass C (SMB-C) was quantified in triplicate soil aliquots (40 g) from each mesocosm using the chloroform fumigation extraction method as previously reported [32]. Briefly, 20 g of each soil sample were fumigated in the dark at 25 °C with ethanol-free chloroform for 24 h, while the remaining 20 g were directly extracted with 80 mL of 0.5 M K₂SO₄ for 1 h (orbital shaker, 40 rpm). The same extraction procedure was applied to chloroform-fumigated samples after chloroform evaporation in a fume hood. All soil extracts were centrifuged and UV absorbance at 280 nm (A₂₈₀) measured.

2.5. Community-level physiological profile, culturable microorganisms, and molecular analysis of soil bacterial community

After the incubation time, the community-level physiological profile (CLPP) of soil microbial communities was obtained for each mesocosm by means of Biolog EcoPlates (Biolog Inc., Hayward, CA). The protocol described by Garau et al. [6] was followed to extract microbial communities from the different soil sample and subsequent inoculation into the Biolog EcoPlate wells (i.e., a 96-wells microtiter plate containing a sole C source in each well, and a control well with no C, and three replicated wells per C source). The tetrazolium violet present in each well allows the monitoring C source consumption which was recorded every 24 h by measuring the optical density of each well at 590 nm (OD₅₉₀). Principal component analysis (PCA) was carried out (using the correlation matrix) on the normalized OD₅₉₀ values [6] to reduce multidimensional data and allow for easier interpretation of C source consumption data. Percent utilization of the different C source guilds within each Biolog EcoPlate (sugar and sugar derivates, sugar phosphates, carboxylic acids, amino acids, polymers) was also determined. Data related to C source consumption and PCA analysis refer to the 96-h time, since this time point allowed for the best treatment’s distinction.

After the incubation period, total culturable heterotrophic bacteria, actinomycetes, fungi, and heat-resistant bacterial spores (e.g., Bacillus spp.) were quantified in soil mesocosms using the ten-fold serial dilution and spread plate method previously described [33]. To quantify the different microbial populations the following growth media were used: AIA (Actinomycete Isolation Agar, Difco, Milan, Italy) for actinomycetes; 1:10 TSB (Tryptic Soy Broth, Microbiol, Cagliari, Italy) + agar (1.5%) for soil heterotrophic bacteria and aerobic bacterial spores; GYEP pH 4.7 (Glucose Yeast Extract Peptone medium) for soil fungi. Microbial counts were conducted after incubation of the plates at 25 °C for 72 h and reported as Log₁₀ CFU (Colony Forming Unit) g⁻¹ soil (dry matter basis).

DNA was extracted from each soil sample (∼500 mg) with the PowerSoil DNA isolation kit (Mo Bio laboratories, Carlsbad, CA). DNA extracts were used as templates for PCR amplification of the V4–V5 (primer-set: 515f 5′-GTGYCAGCMGCCGCGGTAA-3′, 926r 5′-CCGYCAATTYMTTTRAGTTT-3′) regions of the 16S rRNA gene [34,35], and 300bp paired-end sequencing was carried out at the Integrated Microbiome Resource (IMR, Halifax, NS, Canada) facilities according to their in-house protocol [36].

The sequencing data can be accessed with the Bioproject number PRJNA930723 at the Sequence Read Archive of the National Center for Biotechnology Information (https://www.ncbi.nlm.nih.gov/bioproject/).

2.6. Bioinformatics data analysis

Sequencing data were quality assessed and controlled, and microbial sample to amplicon sequence variant (ASV) matrices were generated with the dada2 v1.22.0 package [37] of the R v4.1.3 software, based on previously published protocols [38] as follows. Data were quality assessed and sequences with unknown nucleic acid bases were removed. Sequencing error rates were estimated and read-pairs were assembled to their amplicons of origin. Read-pairs with mismatches at the ‘3-end (error prone) read areas, preventing the reconstruction of the original sequences, were removed from downstream analysis. Chimeric amplicons (according to de novo chimera search) and off-target sequences (based on taxonomic annotation with the Silva v138 16S rRNA gene database; [39]) were also removed.

2.7. Pot experiment and plant analysis

At the end of soil-biochar incubation, six pots (40 cm diameter, 30 cm height), each containing about 25 kg of soil, were set up for each treatment (two for each mesocosm); 12 pots in total were therefore arranged. Twenty hemp plantlets (C. sativa L.) cv. Tiborszallasi were initially planted in each pot and thinned to 6 after 2 weeks. Plants were cultivated for 5 months (from April to September 2022) in a greenhouse under controlled conditions (25–35 °C air temperature and 60–70% relative humidity). The experiment was stopped when plant seeds were fully mature. At harvest, seeds, shoots and roots were separated, carefully and vigorously washed with deionized water, to effectively remove adhering soil particles from the roots, and dried at 55 °C for 72 h; then plant tissues were mineralized by microwave digestion (ultraWave Microwave Milestone) with a solution containing 2 mL of suprapure H₂O₂ and 4 mL of a mixture of HNO₃ and ultrapure H₂O (ratio 1:1). Afterwards, total PTEs concentration (i.e., Sb, Cd, Pb and Zn) in mineralized roots, shoots and seeds samples was determined by ICP-OES (PerkinElmer Optima 7300 DV). Peach leaves (NIST-SRM 1547) were used as standard reference material for quality assurance, with PTE recoveries between 8.5 and 10% of the certified values. The detection limits of the analyzed elements were between 2.63 μg L⁻¹ (As) and 0.14 μg L⁻¹ (Cd).

The root (MM_R), shoot (MM_A), and seed (MM_S) PTE mineralomasses were calculated as roots, shoots, or seeds biomass x PTE concentration in roots, shoots, or seeds [40].

3.2. Influence of biochar on Sb mobility

The results of the sequential extractions showed that in the untreated soil, the percentage of water-soluble and exchangeable Sb (i.e., the labile Sb fraction, Fraction 1), although low compared to the pseudo-total Sb concentration in soil (i.e., 0.65%, Fig. 1A), was very high in absolute terms (>10 mg kg⁻¹). Labile Sb decreased by 1.14-fold after biochar addition (Fig. 1A). This is relevant from an environmental viewpoint since this fraction is recognized to greatly impact soil functionality [27,30]. Moreover, from a remediation viewpoint, the 3% biochar addition led to a reduction of approx. 6 kg of labile Sb per ha (assuming 0.3 m depth; 1300 kg m³ soil density) which is remarkable. This reduction can be attributed to the formation of more stable interactions between the Sb(OH)₆^ˉ oxyanion (the dominant Sb species in aerobic soils; [28]) with different functional groups of biochar such as hydroxyl groups [8], its occlusion within biochar pores, and/or its precipitation as insoluble compound [45]. The content of Sb chemically bound to biochar surfaces through inner-sphere complexes (Fraction 2, Fig. 1A) was 2.48% and 2.22% of the pseudo-total Sb in C and C + B soil respectively, with a 1.1-fold reduction in the amended soil. A similar trend was observed for the Sb associated to poorly crystalline and amorphous Fe and Al (hydr)oxides (Fraction 3, Fig. 1A). In particular, biochar addition decreased by 1.1 times the Sb extracted in Fraction 3. The Sb associated to well-crystallized Fe and Al oxides (Fraction 4, Fig. 1A) was not affected by biochar addition while the residual Sb (i.e. that strongly retained by soil colloids and/or precipitated) significantly increased in amended soil (Fig. 1A). This fraction was the most relevant in both soils, with ∼47 and 49% of the pseudo-total Sb in C and C + B respectively (Fig. 1A). These results agree with Hua et al. [46], who reported a reduction of Sb labile pools and an increase of residual Sb in a contaminated soil treated with straw and apple tree-derived biochar. Wang et al. [11] also reported a reduction of Sb mobility in a contaminated soil treated with low rates (i.e. between 0.5 and 2%) of tea branches-derived biochar and Fe–Mn-modified biochar. On the whole, adding biochar favoured a redistribution of soil Sb, i.e. a reduction of soluble, exchangeable and chemically bound species and a parallel increase of more stable (and hardly bioavailable) ones. Such Sb immobilizing effect was attributable to the high concentration of metal cations in biochar (e.g. exchangeable Ca; Table S1), which could have favoured the formation of insoluble Ca–Sb precipitates, as well as to the presence of Fe and Mn oxides that can fix this PTE increasing its residual fraction [8,46,47].

Although the influence of biochar in reducing labile antimony was significant, substantial amounts of potentially bioavailable Sb were still present in C + B treatment. The high mobility of the Sb(OH)₆^ˉ anion may still cause a serious threat to the abundance, diversity, and functionality of the soil microbial community as well as to the soil’s fertility. These aspects should be carefully investigated in order to fully understand the biochar potentials to recover Sb-contaminated soils.

3.3. Influence of biochar on the mobility of cationic PTEs

The water-soluble and readily exchangeable Cd and Pb were under the detection limit (<0.2 μg kg⁻¹; Fraction 1, Fig. 1B and C), while labile Zn (Fraction 1, Fig. 1D) did not change after biochar addition. The concentration of Zn in Fraction 1 of treated and untreated soils was small compared to its pseudo-total concentration (0.26% and 0.29% in C and C + B soils, respectively; Fig. 1D). Probably, the neutral soil pH, and the use of charcoal during the smelting process [22] promoted the retention of soluble PTEs by the carbonaceous material through surface adsorption and internal diffusion processes. It is noteworthy to highlight that the low or no content of Cd, Pb and Zn in Fraction 1 implies a low environmental risk due to the high mobility and potential bioavailability (for plants and microorganisms) of this fraction [6,30].

The fractions of PTEs extracted with NaOAc (Fraction 2) were 3.71%, 0.38%, and 10.52% of the pseudo-total Cd, Pb, and Zn respectively. These fractions decreased by 1.09-, 1.18- and 1.27-fold for Cd, Pb and Zn respectively in the biochar-amended soil (Fig. 1B, C and D). The PTEs extracted in Fraction 3 (i.e., those poorly mobile and bioavailable) were 2.66, 14.10 and 7.85% of the pseudo-total Cd, Pb and Zn respectively (Fig. 1B, C and D). The EDTA-extractable (i.e., Fraction 3) Pb and Zn decreased by 1.15- and 1.19-fold in biochar-amended soil respectively compared to the control, while it remained unchanged for Cd. A significant increase of the residual Cd, Pb and Zn pool (+1.05-, +1.02- and +1.04-fold compared to C, respectively), which accounted for more than 80% of the pseudo-total PTEs concentration was recorded in biochar treated soil (Fig. 1B, C and D). The pH increase in the C + B soil, the high content of available P in biochar (which can favour the formation of stable PTEs precipitates, e.g. PTEs-phosphates), as well as the presence of amorphous Fe and Mn oxides and organic functional groups in biochar (e.g. carboxylic and phenolic) contributed to reduce the mobility of PTEs in the amended soil and to increase their residual fraction, in agreement with previous studies [24].

3.4. Influence of biochar on soil enzyme activities, microbial carbon and soil basal respiration

Soil enzyme activities (i.e., DHG, URE, GLU, PHA and PHB) are sensitive to environmental changes, such as pH variation or nutrient availability in soil [48]. However, they are also sensitive to PTEs, which could influence their activity (e.g., by denaturation or inactivation) and/or their abundance in soil (e.g., by exerting a toxic effect on soil microbial populations; [30,47]).

The DHG activity, due to intracellular enzymes whose activity is used to estimate the oxidative capability of soil microbial communities [49], decreased by 2.1-fold in C + B soil compared to C (Fig. 2). This is in line with Demisie et al. [50] and Khan et al. [51], who observed a reduction of DHG activity in a soil treated with low rates (between 1 and 2.5%) of bamboo- and wheat straw-derived biochar. Likewise, Ali et al. [52] reported a reduction of DHG activity in a Zn and Cd polluted soil treated with 5 and 10% of apricot shell- and apple tree-derived biochar. Also, Kaurin et al. [53] observed reduced DHG activity in a Cd, Pb and Zn polluted soil treated with microbially enriched biochar. It is possible that the biochar used in this study, characterised by a high specific surface and adsorbing capacity [8,24], retained part of the triphenyl-tetrazolium chloride (TTC) or triphenylformazan (TPF) used to quantify DHG, leading to an underestimation of the enzyme activity in biochar-treated soil [54]. Alternatively, the reduced DHG activity could be due to a reduction of the microbial abundance in the amended soil which led to reduced DHG activity.

PHA and PHB activities in soil are due to extracellular enzymes released by soil microorganisms (and particularly sensitive to soil pH) that catalyse the hydrolysis of phosphate anhydrides and esters [31]. Biochar addition did not affect PHA, while PHB increased by 1.18-fold in C + B soil compared to the control (Fig. 2). This could be explained by the different activity of PHA and PHB at different pH values, being the former more active at acid and sub-acid pH, and the latter at alkaline ones, like those recorded in C + B soil (Table 1; [55]).

URE activity, due to extracellular enzymes capable of catalysing the hydrolysis of urea to CO₂ and NH₄⁺, increased by 1.22-fold in C + B compared to C (Fig. 2), in line with previous findings [52,55]. These authors reported an increase of URE activity in PTEs-contaminated soils treated with biochar from different sources (apricot shell, apple tree and corn stalk). In our case, this could be due to an increased microbial activity in the amended soil, resulting in increased synthesis of URE to meet the available nitrogen demand of the microbial community [55]. If it was the case, the reduced DHG activity recorded in C + B was therefore attributable to biochar adsorption of TTC and/or TPF.

GLU activity, which catalyses the hydrolysis of non-reducing terminal glucosyl residues from polysaccharides, was not affected by biochar addition (Fig. 2). This agrees with Ali et al. [52] who observed that biochar derived from apricots and apple trees, when applied at low doses (i.e., 1% or 2.5%), did not affect or slightly reduced GLU activity in Cd- and Zn-contaminated soils.

During the first 2 weeks of incubation soil basal respiration increased in both treatments (C and C + B) then steadily decreased (Fig. 3A). However, between week 5 and 8 respiration of C + B was always higher (approx. 1.22-fold) compared to C (Fig. 3A). Similar results were reported in a meta-analysis of 18 studies considering several types of biochar and soils, and were explained as a short-term “priming effect” of biochar [56]. The new source of C brought by the biochar addition may have increased CO₂ emissions in the amended soil, as evidenced by the grand mean of soil respiration (Fig. 3B). At the end of the incubation period, the SMB-C reflected a similar trend, i.e. it was 1.85-fold higher in C + B compared to C (Fig. 3C). It is likely that the reduction of labile PTEs (and consequent reduced bioavailability) resulted in a beneficial effect on the soil microbial community whose size and activity increased as evidenced by the SMB-C content and CO₂ emission data. This contrasts with the reduced DHG recorded in C + B soil, which, in light of these findings, can be explained by the TTC and/or TPF adsorption by the biochar rather than a reduced microbial content in the amended soil. These results showed that the studied biochar can increase soil microbial abundance and general activity while stimulating selected soil enzyme activities. It remained to clarify if this was also accompanied by a change of the soil microbial community structure.

3.5. Influence of biochar on culturable soil microbial communities and bacterial diversity

Apart for total culturable bacteria, whose decrease in C + B was not statistically significant, the size of the other targeted culturable microbial populations significantly increased (i.e., between 1.57- a 1.98-fold) in biochar amended soil (Fig. 4A). The reduced PTEs mobility (Sb especially) and the higher P availability recorded in the amended soil favoured the growth of culturable actinomycetes, fungi and spore-forming bacteria.

The Biolog CLPP highlighted a different use of C sources by the microbial communities of different soils. This was shown by the PCA analysis of carbon source utilization data (Fig. 4B) which accounted approx. for 67% of total variance and separated well C and C + B microbial communities. Tween 80 (r = 0.85), d-cellobiose (=0.86), d-glucosaminic acid (r = 0.98), d-malic acid (r = 0.83), l-phenylalanine (r = −0.82) and l-threonine (r = 0.86) were identified as the C sources highly correlated with PC1, while β-methyl-D-glucoside (r = 0.65), D,L-α-glycerol phosphate (r = 0.62), γ-hydroxybutyric acid (r = 0.66), itaconic acid (r = 0.88) and l-arginine (r = 0.81) were highly correlated with PC2. Overall, C + B microbial communities showed a higher utilization of carboxylic acids (i.e., +4%) and a reduced consumption of sugar and sugar derivates (i.e., −5%) compared to those of the unamended C soil (Table S3). These results pointed out that biochar not only led to an increase in size of the microbial community but also influenced its diversity.

The influence of the biochar on microbial diversity was investigated by the molecular DNA analysis of soil bacterial population. Quality assessment and quality control results are presented in Table S4. The analysis of α-diversity indices showed statistically significant differences between the two treatments in all dominance levels and corresponding indices, and indicated a clear role of biochar in the promotion of soil bacterial diversity (Fig. 5A).

NMDS results also showed a clear sample grouping according to the treatment, with the PERMANOVA results showing that 61.14% of the observed variance coincided with the treatment groupings (Fig. S1). In the PERMANOVA model, although the R² was rather high (showing a strong effect along the treatment gradient), the model P-value was not significant (P = 0.1). However, it should be noted that the reason for this is purely technical since PERMANOVA requires at least 4 replicates for 2 treatment levels to enable P-values smaller than 0.05 [57].

Moreover, differential abundance analysis aiming to identify prevalent genera across treatments (present in all replicates of at least one treatment group), showed 20 genera being differentially abundant among treatments (Fig. 5B). Such analysis, as also suggested by NMDS, indicated a significant impact of biochar on the bacterial community composition. In particular, the abundance of selected Proteobacteria, Actinobacteriota and Acidobacteriota was stimulated by biochar while Nitrospirota and Myxococcota were significantly reduced (Fig. 5B). Overall, these results (Fig. 5) show that biochar derived from softwood was able to increase soil bacterial diversity and to significantly change the structure of the bacterial community inhabiting the PTEs-contaminated soil. This could be relevant in terms of plant growth as different authors highlighted a positive relationship between soil fertility and microbial diversity [58], while the influence of soil microbial community on plant growth is widely recognized [59]. However, these aspects need further evaluation and experimental evidence from specific plant growth experiments.

3.6. Influence of biochar on hemp growth

Hemp successfully grew in the PTEs-contaminated soil and no apparent phytotoxicity symptoms were detected. Biomass production of hemp grown in the polluted soil (Fig. S2) was similar to that observed by Cosentino et al. [60] and Flajšman and Kocjan Ačko [61], who cultivated the cv Tiborszallasi (the same used in our study) in uncontaminated agricultural soils. On the other hand, comparable shoot and seed biomass was also observed by Ref. [18] when cannabis (cv. USO 31) was grown in a Cd-, Ni- and Pb-contaminated soil; while higher biomass yields were reported by Meers et al. [20] for hemp (cv. Fibranova, Chameleon and Codimono respectively) grown in a multi-PTEs-contaminated soil where, however, Sb was not present. These results suggest that hemp variety can be relevant to optimize biomass production in PTEs-contaminated soils.

Biochar did not significantly affect root growth but did significantly increase cannabis aboveground biomass. Shoot biomass in C + B increased by 1.67-fold, and seed number and weight increased by ∼2-fold compared to C (Fig. S2). The enhanced soil fertility (particularly due to the increase of the available P and CEC; Table 1), the reduced content of labile Sb and other PTEs, the higher soil microbial content (i.e., SMB-C) as well as the community structure and activity within biochar amended soil, were likely responsible for the higher yield of hemp in the amended soil. A positive impact of biochar on hemp biomass was previously observed by Loffredo et al. [62] for the cv. Kompolti (i.e., 78 and 59% for roots and shoots respectively) grown in a soil contaminated by organic pollutants. Apart from that, we are not aware of other studies addressing biochar impact on hemp growth in polluted soils. Overall, these results highlight that hemp can be successfully grown in soils heavily contaminated by Sb and other co-occurring PTEs and that biochar enhanced plant biomass in such harsh environments. This latter biomass could be used for bioenergy production and/or bio-based products (biofuels, bioplastics and fibres), thus opening the possibility of converting unproductive contaminated soils into exploitable and productive ones (e.g., Ref. [3]). However, the role of hemp as PTEs phytostabilizer or phytoextractant in such soils remains to be ascertained.

3.7. Influence of biochar on PTEs uptake by hemp

Apart from Sb, the other PTEs (i.e. Cd, Pb and Zn) were mainly concentrated in root tissues, regardless of the treatment applied (Fig. 6); importantly, biochar affected in different ways PTEs uptake and distribution in hemp. The concentration of Sb (the most abundant PTE in soil) decreased significantly in the roots of plants grown in C + B soil (approx. by 35% compared to those grown on C soil), while Sb concentration in shoots and seeds increased by 24 and 27%, compared to control plants, respectively (Fig. 6A).

The concentration of cationic PTEs increased in roots of cannabis grown in C + B soil (compared to C), i.e. by approx. 14, 32 and 77% for Cd, Pb and Zn respectively (Fig. 6B, C and D). On the contrary, Cd and Zn concentration in shoots decreased for plants grown in C + B soil (i.e., by 24 and 8% respectively) vs control plants (Fig. 6B–and D), while Pb concentration substantially increased (∼87%). Also, PTEs concentration in seeds was significantly affected by biochar, i.e. Cd and Pb concentrations in seeds of plants grown in C + B were under the detection limit (<0.2 μg kg⁻¹), while those of control plants were 0.12 and 1.09 mg kg⁻¹ for Cd and Pb respectively (Fig. 6B and C). On the other hand, biochar addition did not affect Zn concentration in seeds (Fig. 6D).

Overall, these results show for the first time that biochar can significantly increase hemp phytostabilisation potentials with respect to Cd, Pb and Zn. At the same time, biochar revealed very active at increasing hemp phytoextracting capacity with respect to Sb. This could be explained by a biochar-mediated improvement of the soil physico-chemical environment which eventually led to a better plant growth (Fig. S2) and a possible metabolic adaptation of hemp (to the harsh environment) which favoured PTEs uptake and/or their translocation [62].