Impact of biochar on carbon sequestration in permafrost region of Northeast China

Abstract

Background

Biochar effects on soil organic matter stability in permafrost regions remains poorly understood. To address this knowledge gap, two-cycle incubation experiments using representative forest and peatland soils were conducted from Daxing’anling permafrost region. Soils with corn straw-derived biochar (pyrolyzed at 450 °C, 2 h) were amended at 8% w/w of dry soil weight and systematically measured soil organic carbon (SOC), total nitrogen (TN), dissolved organic carbon (DOC), carbon fractions, microbial community, carbon emission, and CO₂ isotope content.

Results

The research indicated that biochar amendment improved physicochemical properties in both soil types. Electrical conductivity increased by 166.62% (forest) and 223.79% (peatland), while SOC increased by 60.57% (forest) and 5.64% (peatland). Mineral-associated organic carbon increased significantly (particularly in forest soils), which also exhibited increases in TN (32.15% at 180 days) and DOC (197.50% at 90 days). Biochar addition reduced the diversity and richness of bacterial communities in forest soils, but had no significant effect on peatlands.

Conclusion

Biochar promoted soil aggregate formation, improved soil carbon sequestration capacity, and reduced CO₂ emissions by 19.37% (forest) and 9.70% (peatland). These findings confirmed the dual functionality of biochar in increasing soil carbon storage and reducing carbon emissions. The study provides valuable insights for enhancing carbon management strategies in vulnerable permafrost ecosystems, emphasizing the potential of biochar in soil management.

Introduction

The Daxing’anling permafrost region is located at the southern edge of Eurasia high-latitude permafrost belt, which is second-largest permafrost zone of China [1] and characterized by discontinuous permafrost transitioning to continuous formations in northern sectors [2]. Climate warming has elevated temperatures in these permafrost areas, which directly influences active layer thickness dynamics and altered soil moisture-heat transport mechanisms, thereby modifying carbon cycling processes [3]. Soil carbon cycles critically influence both soil quality and greenhouse gas (GHG) emissions [4]. Under combined climate change and human disturbances, the soil organic carbon (SOC) depletion in northeastern the black soil region of China has exceeded 50% [5], which accelerates SOC loss and elevating GHG emissions [6]. Consequently, SOC conservation represents a crucial strategy for climate change mitigation [7] and soil quality enhancement [8].

Biochar application enhances SOC sequestration [9] by improving physicochemical and microbial properties [10], stabilizing carbon, reducing GHG emissions [11], and increasing nutrient use efficiency [12]. Long-term SOC increases derive from the inherent stability of biochar and suppression of organic matter mineralization [13, 14]. Interactions between biochar and soil organic matter (SOM) promote organo-mineral complex formation [15] and inhibit SOC mineralization through negative priming effects (PEs) [16]. However, the influence of biochar on carbon emissions remains context-dependent (soil type, biochar feedstock, pyrolysis conditions), and environmental factors [17] may cause increases [18], reductions [19], or neutral effects [20]. Although labile biochar carbon can stimulate microbial co-metabolism and increase emissions transiently [21], long-term mitigation relies on inhibitory mechanisms involving soil property modification, gas adsorption, and microbial activity suppression [22, 23].

Forests and peatlands dominant the Daxing’anling region, which form high-latitude carbon-dense ecosystems [24]. Peatlands exhibit higher SOC density than forests due to temperature- and topography-driven waterlogging that inhibits decomposition [25]. Biochar-induced PEs are regulated by inherent soil properties with SOC content, texture (silt/clay), and pH as key drivers [26–28]. Weaker PEs occur in high-SOC versus low-SOC soils [29], which indicate pre-existing SOC pool size governs responsiveness. Sandy soils exhibit stronger positive PEs than clayey soils due to enhanced microbial accessibility to SOC [27, 30], and biochar-induced pH elevation stimulates microbial activity and intensifies PEs [31]. Identifying such soil-specific mechanisms is essential for developing tailored carbon management strategies.

Converting straw residues into biochar provides a sustainable agricultural waste management strategy [32], which mitigates air pollution and resource waste from open burning. Pyrolysis transforms most organic carbon into recalcitrant aromatic structures, while syngas (e.g., CO, H₂, CH₄) can be utilized for energy to offset carbon emissions [33]. Current permafrost carbon emissions studies primarily focus on three critical seasonal stages (spring thaw period, summer growth season, and autumn freeze period). This research is mainly conducted through short-term experiments examining climate warming [34] and natural freeze-thaw cycles [35] on the GHG fluxes in permafrost regions. However, few studies have systematically monitored carbon emissions across all three periods in permafrost regions. Continuous multi-year (≥ 2 years) data are essential to elucidate long-term carbon emission patterns. To address this gap and evaluate the impact of biochar on carbon dynamics in permafrost, a controlled indoor simulation experiment was conducted encompassing three key seasonal phases.

Clarifying biochar-mediated carbon sequestration requires tracking distinct carbon sources. Soil humus exhibits characteristic carbon isotopic (δ¹³C) differences between C₃ and C₄ plant origins, while straw-derived biochar possesses a distinct isotopic signature. Carbon isotopes can be used to track carbon sources, the whereabouts of SOC and plant residual carbon under the influence of biochar, and the contribution of biochar to CO₂ emissions [36]. Stable carbon isotope analysis with mixing models were combined to identify different carbon pools in the soil and quantify the impact of biochar on carbon mineralization [37].

Existing biochar research focuses predominantly on agricultural soils, and leaving critical knowledge gaps in permafrost regions are (1) the ability of biochar to enhance SOC stability and reduce carbon emissions, (2) comparative studies on soils with different SOC levels (low-carbon forests versus high-carbon peatlands), and (3) direct evidence (e.g., δ¹³C-CO₂) for biochar-induced suppression of native SOC mineralization. The objectives address these gaps by (1) characterizing carbon emission dynamics in permafrost soils of forests (low SOC) and peatlands (high SOC) modified with biochar, (2) elucidating the physicochemical and microbial mechanisms that inhibit soil carbon mineralization, and (3) employing δ¹³C tracing to differentiate CO₂ sources into natural SOC and biochar. This approach establishes a scientific basis for optimizing biochar-mediated carbon management in permafrost ecosystems.

Materials and methods

Tested soils

The biochar was derived from corn straw, which pyrolyzed at 450 °C for 2 h under a nitrogen (N₂) atmosphere. The properties of biochar can be found in the BC0 experimental group of published literature [38]. Key properties are detailed in Table 1. Topsoil (0–10 cm depth) from the larch (Larix gmelinii) forest (122.864 °E, 51.850 °N) and peatland (124.136 °E, 50.424 °N) were collected in the mixed forest-wetland ecosystems of Daxing’anling (Fig. 1a). Sloping terrain supported forests with loamy soil, while low-lying permafrost sustained peat swamps, with a summer water table depth of about 10 cm [39]. Three independent sampling points per site were homogenized from five diagonal subsamples. Samples were transported and stored at 4 °C, then processed through 2-mm sieving to remove stones and roots, followed by moisture content determination. The soils were classified as dark brown Alisols (FAO), with texture of 12.45% sand, 58.89% silt, 28.66% clay for the forest, and 72.95% sand, 27.05% silt for the peatland.

Table 1.

Physicochemical properties of corn straw-derived Biochar

	pH		Bulk density (g cm−3)			C (wt%)		N (wt%)	Mad (wt%)	Aad (wt%)	Vad (wt%)
	8.10 ± 0.16		0.257 ± 0.006			77.30 ± 1.39		0.87 ± 0.13	5.07 ± 0.12	10.11 ± 0.89	7.88 ± 0.55
Biochar	Metallic elements (wt%)						Functional groups	Particle size (mm)	SSA (m2 g−1)	Pore volume (cm3 g−1)	Permeability coefficient (cm·s−1)
	Mg	K	Ca		Cr	Fe
	0.84 ± 0.09	1.09 ± 0.12	1.20 ± 0.17		0.37 ± 0.06	0.21 ± 0.05	12	< 2	305.52 ± 9.23	0.098 ± 0.011	0.026 ± 0.008

(1) wt% is the percentage of mass; M_ad is sample moisture; V_ad stands for volatile matter; A_ad is ash content. ad is air drying base; SSA is the specific surface area (SSA) of biochar

(2) The functional groups of biochar mainly included -OH, C-H, C = O, -COOH, etc

Fig. 1.

Location of the research area and experimental flow chart

Experimental description

The optimal amount of biochar is usually less than 10% of the soil dry weight [40, 41]. Therefore, the study established four experimental treatments, each replicated three times: FB0 (forest control), FB8 (forest with 8% biochar, w/w), PB0 (peatland control), and PB8 (peatland with 8% biochar, w/w). Fresh forest soil (957.25 g wet weight; 700 g dry weight) and peatland soil (1391.85 g wet weight; 270 g dry weight) were packed into polymethyl methacrylate (PMMA) columns of uniform volume (Fig. 1b). Soil moisture was maintained at field capacity with ultrapure water additions every three days. Pre-incubation was conducted at 10 ℃ for 15 days under dark conditions to stabilize microbial activity [42].

Incubation occurred in a temperature-controlled, light-proof chamber (Thermo Fisher Scientific, Waltham, MA, USA). Temperature regimes were programmed based on field data recorded at 5 cm soil depth, simulating seasonal phases were spring thaw (0–8 °C), summer growth (10–13-10 °C), and autumn freeze (10–0 °C). Temperatures followed a stepwise progression (0, 1, 3, 5, 6, 8, 10, 12, 13, 11, 11, 10, 10, 7, 7, 3, 1, and 0 °C), with each phase maintained for 10 days. The experiment comprised two consecutive 180-day cycles (total 360 days). Gas was sampled at 5-day intervals during cycle 1 and 10-day intervals during cycle 2, following system stabilization. Prior to sampling, column lids were opened in a well-ventilated area to equilibrate the headspace. Emissions were quantified by collecting 20 mL headspace samples before and after 5-h containment periods. The carbon isotopic composition of respired CO₂ (δ¹³C_CO2) was analyzed at 60-day intervals (Day 60, 120, 180, 240, 300, and 360).

Analysis methods

Determination of soil physicochemical properties

SSA were determined with an SSA and aperture analyzer (ASAP2460, Micromeritics Instrument Corp, USA). Elemental composition was analyzed using an elemental analyzer (Flash2000, Thermo Fisher Scientific, Italy). Soil physicochemical properties were measured at 90-day intervals (0, 90, 180, 270, 360 days), quantifying total porosity by the ring knife method and electrical conductivity (EC) with a DDBJ-351 L conductivity meter. Soil texture followed USDA classification: sand (> 50 μm), silt (2–50 μm), and clay (< 2 μm). Organic matter removal involved sequential pretreatment with 5% H₂O₂ and 5% HCl, followed by sodium hexametaphosphate dispersion and particle size analysis (Mastersizer 3000, Malvern Panalytical, UK).

Soil permeability was assessed via variable head method with a T55 permeameter (2 m initial head height, ultrapure water interpolation), calculated using Eqs. (1) and (2).

1

2

where, k_T is the temperature-dependent permeability coefficient, cm s⁻¹; A is the specimen cross-sectional area, cm; a is the variable head pipe cross-sectional area, cm²; L is the height of the sample, cm; H₁, H₂ is the initial/final water head heights, cm; k₂₀ is the permeability coefficient of the sample at standard temperature (20 °C), cm s⁻¹; η_T, η₂₀ is the water dynamic viscosity coefficient at T °C and 20 °C, 10⁻⁶ kPa·s.

Total nitrogen (TN) was quantified by modified Kjeldahl method, while total phosphorus (TP) was quantified by alkali fusion-Mo-Sb anti-spectrophotometry. Soil ammonia nitrogen (NH₄⁺-N) and nitrate nitrogen (NO₃⁻-N) were extracted with 2 M KCl and measured using spectrophotometrically. All analyses were conducted immediately after extraction with a SAN + + continuous-flow analyzer (Skalar Analytical, Netherlands). Soil samples were ground to < 125 μm and analyzed SOC content including the δ¹³C. Carbonate removal was achieved through 2 M HCl pretreatment, followed by SOC quantification using Multi N/C 2100 analyzer. The δ¹³C values was determined using a MAT-253 stable isotope ratio mass spectrometer with zero-blank autosampler. Standard deviations of internal standards were better than ± 0.1‰ for δ¹³C values and ± 2% of the value for abundance measurements. Soil dissolved organic carbon (DOC) was analyzed directly by Multi N/C 3100 analyzer, while particulate organic carbon (POC) and mineral-associated organic carbon (MAOC) were separated by 18-h sodium hexametaphosphate dispersion and wet sieving. Then the POC fraction (> 53 μm) and MAOC fraction (< 53 μm) were separated by sieving. Soil of different fractions were dried and ground for organic carbon analysis. Soil pH was measured with a calibrated meter.

Microbiological analysis

Amplicon sequencing of bacterial 16 S rRNA genes (V3-V4 region) and fungal ITS regions were conducted by Majorbio Bio-pharm Technology Co., Ltd (Shanghai, China). Total genomic DNA was extracted from soil samples using the E.Z.N.A.^® Soil DNA Kit (Omega Bio-tek, USA). DNA quality was assessed by 1% agarose gel electrophoresis with concentration and purity measured using a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). Bacterial V3-V4 regions were amplified with barcoded primers 338 F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′), while fungal ITS1 regions were amplified with primers ITS1F (5′-CTTGGTCATTTAGAGGAAGTAA-3′) and ITS1R (5′-GCTGCGTTCTTCATCGATGC-3′). Triplicate PCR reactions per sample minimized technical variation. PCR products were separated on 2% agarose gel electrophoresis, purified using a PCR Clean-Up Kit (YuHua, China), and quantified with a Qubit 4.0 Fluorometer (Thermo Fisher Scientific, USA).

Respiration measurements

Concentrations of CO₂ and CH₄ were quantified using gas chromatography (Agilent 7890B GC System, USA). Calibration occurred every 12th sample using certified GHG standards (National Reference Materials Center, China). Soil GHG emission rates (F, mg·kg⁻¹ h⁻¹) were calculated using Eq. (3) [43]. Cumulative soil respiration was calculated using Eq. (4) [44].

3

4

where, ρ is the calibration gas concentration, g·L⁻¹; ∆C is the GHG concentration difference relative to the blank, mL·m⁻³·h⁻¹; V is the chamber volume, mL; T is the incubation temperature, ℃; W is the soil dry weight, kg. Positive and negative F values indicate net GHG emissions and absorption, respectively.

The δ¹³C_CO2 was analyzed using a CO₂ isotope analyzer (CCIA, ABB, Canada), calibrated with certified reference gas (401.7 ppm CO₂, δ¹³C = −20.82‰). Analytical precision (± 0.13‰) was confirmed through repeated measurements. The δ¹³C_CO2 was determined using Eq. (5).

5

where R_sample and R_sample represent the ¹³C/¹²C ratios of samples and Vienna Pee Dee Belemnite (VPDB), respectively.

Biochar-derived carbon contributions were quantified using Eq. 6:

6

where, f_biochar is the biochar-carbon fraction in CO₂; δ¹³C_B0, δ¹³C _B8 is the control and biochar-amended soil isotope signature; δ¹³C_biochar, δ¹³C_S is the biochar and native soil isotope signatures.

Statistical analysis

Statistical analysis of microbiome community data was conducted on the Meggie Bioinformatics Cloud Platform. The alpha diversity of the microbiome was calculated using Mothur (version 1.30.2) software, measuring observed species count, Shannon diversity index, and Chao1 richness estimate. Statistical analyses used IBM SPSS Statistics 26. Prior to analysis, data normality was verified using the Shapiro-Wilk test (p > 0.05), and homogeneity of variance was confirmed with Levene’s test (p > 0.05). When data violated these assumptions, appropriate transformations (e.g., logarithmic or square root) were applied to meet ANOVA requirements. Two-way analysis of variance (ANOVA) was performed, followed by Fisher’s Least Significant Difference (LSD) post-hoc test to evaluate the effects of biochar application on soil properties. A significance threshold was applied with results deemed statistically significant at P ≤ 0.05. Effect sizes (η²) were calculated to quantify the magnitude of significant effects, with η² ≥ 0.01, ≥ 0.06, and ≥ 0.14 interpreted as small, medium, and large effects, respectively.

Result

Physicochemical properties

The elemental composition and pore structure characteristics of soil samples are summarized in Table 2. Forest control groups (F0, FB0-M and FB0-E) exhibited a biphasic trend, characterized by an initial increase followed by a decline over the experimental period. Corresponding SSA and pore volume showed parallel temporal patterns. Biochar significantly increased C and N content in forest soils, while other elements remained stable. In peatland controls treatments displayed slight reductions in C, H, and O content, though SSA and pore volume were unaffected. Conversely, biochar-amended peatland soils maintained stable elemental composition throughout incubation period. However, their SSA and pore volume exhibited an initial increase during early phases, followed by a gradual reduction in later stages.

Table 2.

Element content and pore structure in different experimental groups

Experimental group	Element content				Pore structure
	C (wt%)	O (wt%)	N (wt%)	H (wt%)	BET (m2ּּ g−1)	Micropore volume (cm3ּ g−1)
F0	4.43 ± 0.07	12.93 ± 0.07	0.19 ± 0.03	1.63 ± 0.06	153.07 ± 10.26	0.065 ± 0.009
FB0-M	4.72 ± 0.02	13.02 ± 0.08	0.19 ± 0.03	1.71 ± 0.13	160.42 ± 11.17	0.082 ± 0.015
FB0-E	4.53 ± 0.06	12.36 ± 0.53	0.18 ± 0.03	1.57 ± 0.07	66.56 ± 0.08	0.026 ± 0.004
FB8-M	7.67 ± 0.22	12.76 ± 0.45	0.26 ± 0.01	1.62 ± 0.05	274.97 ± 25.61	0.133 ± 0.023
FB8-E	7.84 ± 0.08	12.68 ± 0.41	0.25 ± 0.01	1.60 ± 0.02	257.31 ± 23.51	0.121 ± 0.019
P0	40.74 ± 0.24	35.60 ± 0.25	2.06 ± 0.04	5.18 ± 0.16	165.22 ± 28.16	0.071 ± 0.011
PB0-M	40.56 ± 0.18	35.00 ± 0.50	1.93 ± 0.05	5.13 ± 0.21	152.29 ± 12.39	0.064 ± 0.017
PB0-E	38.93 ± 0.44	34.36 ± 0.41	2.13 ± 0.07	5.07 ± 0.29	162.19 ± 13.07	0.066 ± 0.023
PB8-M	40.17 ± 0.28	33.58 ± 1.18	2.11 ± 0.07	4.87 ± 0.19	570.02 ± 35.21	0.346 ± 0.052
PB8-E	41.96 ± 0.26	32.31 ± 0.50	2.00 ± 0.04	4.94 ± 0.05	272.68 ± 19.35	0.123 ± 0.029

(1) F0 and P0 represent initial forest and peatland soil states, respectively; (2) FB0-M, FB8-M, PB0-M, and PB8-M represent soil mid-cycle elemental content; (3) FB0-E, FB8-E, PB0-E and PB8-E represent soil end-cycle elemental content

Post-incubation, FB8 exhibited EC increases of 166.62% relative to FB0, while PB8 showed increases of 223.79% compared to PB0 (Table 3). Compared to FB0, the permeability of FB8 reduced by 73.10% in cycle 1 and 73.76% in cycle 2. In contrast, peatland soil exhibited only transient permeability reductions, with a 57.00% decrease during the initial 180 days, but no sustained long-term effects (Table 3). USDA texture analysis classified forests as silty loam and silty clay, whereas peatland soils were dominated by loam and sandy loam (Fig. 2b). Biochar induces changes in soil particle size distribution: in forest soils, sand content increased by 92.83%, while clay content decreased by 68.92%. In peatland soil, biochar addition increased sand content by 5.63% (Fig. 2a).

Table 3.

Permeability coefficient and EC in different experimental groups

Experimental group	EC (µs cm−1)	Permeability coefficient (cm s−1)	Experimental group	EC (µs cm−1)	Permeability coefficient (cm s−1)
F0	24.78 ± 5.47	0.000403 ± 0.000021	P0	126.70 ± 10.59	0.000257 ± 0.000017
FB0-1	39.9 ± 8.52a	0.000394 ± 0.000040a	PB0-1	46.27 ± 5.67a	0.000493 ± 0.000052a
FB8-1	122.77 ± 13.74b	0.000106 ± 0.000024b	PB8-1	112.07 ± 5.45b	0.000212 ± 0.000068b
FB0-2	36.37 ± 7.64a	0.000282 ± 0.000005a	PB0-2	59.47 ± 5.33a	0.000273 ± 0.000106a
FB8-2	96.97 ± 8.33b	0.000074 ± 0.000001b	PB8-2	192.63 ± 9.25b	0.000212 ± 0.000053a

F0 and P0 represent initial forest and peatland soil; FB0-1, FB8-1, PB0-1, and PB8-1 represent soil EC and permeability coefficient of cycle 1; FB0-2, FB8-2, PB0-2 and PB8-2 represent soil EC and permeability coefficient of cycle 2. Lowercase letters denote significant differences (P < 0.05) between biochar-amended and control soils

Fig. 2.

Changes in soil particle size distribution: a. Relative proportion changes across treatment groups post-incubation (cycles 1 and 2), b. USDA textural classification via ternary diagram of treatment groups post-incubation (cycles 1 and 2)

Relative to FB0, pH decreases of 4.60% (cycle 1) and 5.59% (cycle 2) were observed in FB8. In contrast, peatland soils-maintained pH stability regardless of biochar application. Compared to FB0, FB8 exhibited TP increases of 26.80 and 34.84% in cycle 1, while PB8 showed greater increases of 38.85 and 53.36% relative to FB0 (Fig. 3e). However, heterogeneous variance in cycle 2 rendered TP changes statistically insignificant. At 90, 180, 270, and 360 days, the TN content in FB8 increased by 16.4%, 32.15%, 27.25%, and 7.23%, respectively, compared to FB0 (Fig. 3f). The C/N ratio of FB8 increased by 101.52%, 110.45%, 38.58%, and 50.28% at 90, 180, 270, and 360 days, respectively, compared to FB0. In contrast, PB8 showed increases of 7.09%, 10.29%, 12.34%, and 1.45% at the corresponding time points relative to PB0. Furthermore, the addition of biochar significantly affects the soil C/N ratio (Table 4).

Fig. 3.

Soil chemical properties of SOC (a), DOC (b), MAOC (c), POC (d), TP (e), TN (f), NH₄⁺-N (g) and NO₃⁻-N (h) accumulated changes from forests and peatland under biochar addition for 360 days incubation

Table 4.

Two-way ANOVA results of soil properties after biochar addition after two incubation cycles

Soil types	Experimental conditions	SOC		DOC		POC
		P	η2	P	η2	P	η2
Forest	Biochar	0.000*	0.881	0.827	/	0.000*	0.569
	Incubation time	0.377	/	0.000*	0.991	0.803	/
	Biochar × incubation time	0.001*	0.091	0.000*	0.008	0.001*	0.407
Peatland	Biochar	0.020*	0.104	0.000*	0.074	0.000*	0.577
	Incubation time	0.000*	0.847	0.000*	0.953	0.024*	0.024
	Biochar × incubation time	0.813	/	0.000*	0.064	0.000*	0.575
Soil types	Experimental conditions	MAOC		TN		C/N
		P	η2	P	η2	P	η2
Forest	Biochar	0.000*	0.945	0.000*	0.453	0.000*	0.868
	Incubation time	0.968	/	0.002*	0.294	0.221	/
	Biochar × incubation time	0.31	/	0.013*	0.14	0.004*	0.081
Peatland	Biochar	0.000*	0.607	0.887	/	0.077*	0.031
	Incubation time	0.001*	0.057	0.000*	0.975	0.000*	0.867
	Biochar × incubation time	0.000*	0.514	0.72	/	0.137	/

The “*” indicated significant effect (P < 0.05)

The SOC in FB8 increased by 135.94%, 177.97%, 82.18%, and 60.57% at 90, 180, 270, and 360 days, respectively, compared to FB0 (p < 0.05). In contrast, PB8 exhibited smaller SOC increases of 0.64%, 10.22%, 4.42%, and 5.64% at the same intervals relative to PB0 (Fig. 3a). The DOC content was significantly affected by incubation time in both ecosystems (Table 4, p < 0.05). Specifically, DOC in FB8 increased by 197.50%, 30.62%, 23.73%, and 0.62% at 90, 180, 270, and 360 days, respectively, relative to FB0 (Fig. 3b). The impact of biochar on DOC in forest attenuated temporally, exhibiting a progressive reduction toward the end of incubation.

Biochar application exhibited a significant positive correlation with MAOC in all tested soils (Table 4). FB8 showed increases of 103.37% in cycle 1 and 175.89% in cycle 2 relative to FB0. Peatland MAOC (PB8) displayed smaller but consistent increases of 41.76% (cycle 1) and 3.05% (cycle 2) compared to PB0. Conversely, the impact of biochar on POC diverged markedly between ecosystems: forest soils demonstrated significant positive correlations with POC content, whereas peatland soils exhibited negative trends (Table 4; Fig. 3d). Specifically, POC content in FB8 (cycle 1) and FB8 (cycle 2) increased by 50.33% and 4.59% respectively, relative to FB0. In contrast, PB8 (cycle 1) and PB8 (cycle 2) decreased by 40.03% and 0.05%, respectively, compared to PB0.

Microbial abundance and diversity

Microbial communities were consolidated by merging OTUs with relative abundances below 0.05% into “Others”, which yielded eight bacterial and five fungal phyla (Fig. 4a). In forest soils, Pseudomonadota, Acidobacteriota, Actinomycetota, Chloroflexota, and Gemmatimonadota dominated the bacterial community, collectively comprising > 75% of total sequences, which reflected typical aerobic and litter-input ecosystems. Peatland soils exhibited significant enrichment of Verrucomicrobiota and Bacteroidota, which functionally replaced Gemmatimonadota, indicating anaerobic, organic-rich conditions promoted methanotrophs (Verrucomicrobiota) and fermenters (Bacteroidota). Fungal communities in both habitats were dominated by Ascomycota and Mortierellomycota. Biochar amendment significantly increased the abundances of Pseudomonadota, Actinomycetota, Chloroflexota, Gemmatimonadota, and Ascomycota abundances in forest soils while reducing Acidobacteriota, Verrucomicrobiota, and Mortierellomycota. In contrast, peatland soils showed no significant alterations except decreased Ascomycota and increased Mortierellomycota abundances.

Fig. 4.

The relative abundance and community diversity of soil bacteria and fungi in the different treatment groups of forests and peatlands at the end of the cultivation period ((a) Changes in the relative abundance of bacteria and fungi at different treatment levels. (b) and (c) represent the changes in the diversity indices of soil bacterial and fungal communities (Shannon and Chao 1) for each treatment group, respectively. * Indicates significant differences at the 0.05 level, ** indicates significant differences at the 0.01 level, and *** indicates significant differences at the 0.001 level.)

Bacterial and fungal community diversity responded differentially to biochar amendment in forest versus peatland soils (Fig. 4b and c). Peatlands exhibited significantly higher bacterial Shannon diversity and Chao1 richness than forests (P < 0.05). Although fungal Shannon index was higher in peatlands, their fungal Chao1 index was significantly lower than in forests. In biochar-amended forest soils (FB8 vs. FB0), bacterial Shannon and Chao1 indices declined by 6.57% and 11.73% (P < 0.05), respectively, with no significant changes in fungal diversity.

Soil carbon emission

CO₂ emission rates and cumulative emissions in biochar-amended soils exhibited pronounced seasonal variability (Fig. 5). In FB8, cumulative CO₂ emissions decreased by 33.61% (Spring 1), 13.80% (Summer 1), and 29.64% (Autumn 1) during cycle 1, and by 65.53% (Spring 2), 14.64% (Summer 2), and 19.37% (Autumn 2) during cycle 2, relative to FB0. The mitigation efficacy of biochar was most pronounced during the spring thaw period, followed by the autumn freeze and summer growth periods. In PB8, cumulative CO₂ emissions showed smaller, nonsignificant reductions: 13.82% (Spring 1), 8.05% (Summer 1), and 4.95% (Autumn 1) in cycle 1, and 8.86% (Spring 2), 12.33% (Summer 2), and 9.70% (Autumn 2) in cycle 2 compared to PB0.

Fig. 5.

CO₂ and CH₄ emission dynamics in forest and peatland soils under biochar addition across two incubation cycles (Error bars indicate standard error (SE), and color-coded shading distinguishes between cycles.)

In FB8, CH₄ uptake decreased by 64.62% (Spring1), 48.28% (Summer1), and 32.30% (Autumn1) during cycle 1, and by 44.07% (Spring2), 26.84% (Summer2), and 46.94% (Autumn2) during cycle 2 compared to FB0 (Fig. 5). The most significant reduction occurred during the spring thaw period of cycle 1. In PB8, CH₄ emissions decreased by 29.59% (Spring1), 22.19% (Summer1), and 52.70% (Autumn1) in cycle 1, and by 34.57% (Spring2), 33.62% (Summer2), and 28.07% (Autumn2) in cycle 2 relative to PB0. The inhibitory effect of biochar on CH₄ uptake peaked during spring thaw in forest soils, whereas the strongest suppression in peatland soils occurred during autumn freeze (cycle 1). Stabilized inhibition effects were maintained across both ecosystems in cycle 2, which suggested sustained biochar efficacy.

Carbon isotopes

GHG emissions analysis revealed variable decomposition rates in active SOC pools over the 12-month incubation. Temporal δ¹³C_CO2 shifts reflected evolving carbon source contributions (Table 5). Initial δ¹³C_CO2 enrichment was observed, peaking at −3.072 (FB0) and − 0.236 (FB8) by day 120 in forest soils, and − 5.796 (PB0) and − 4.902 (PB8) by day 180 in peatlands. Subsequent depletion occurred, which reached minimum values of −10.659 (FB0), −10.056 (FB8), −10.165 (PB0), and − 9.469 (PB8) by day 360. The contribution of biochar-derived carbon followed a consistent trend across the entire ecosystem (an initial increase peaking at day 120). Among these, the total CO₂ emission rate for forest FB8 was 0.6628 mg kg⁻¹h⁻¹ with 40.6% derived from biochar, amounting to 0.2691 mg kg⁻¹h⁻¹. The CO₂ emission rate from soil SOC was 0.3937 mg kg⁻¹h⁻¹. For peat PB8, the total CO₂ emission rate was 1.7991 mg kg⁻¹h⁻¹ with 17.8% coming from biochar, which was 0.3202 mg kg⁻¹h⁻¹. While the CO₂ emission rate from soil SOC stood at 1.4789 mg kg⁻¹h⁻¹. Following this, a decline stabilized around 10%.

Table 5.

The δ¹³C_CO2 values of soil CO₂ emissions under different treatment groups with the addition of biochar, as well as the CO₂ emission, soil-derived CO₂ emissions, and biochar-derived CO₂ emissions during the incubation experiment

Soil types	Time (day)	δ13CCO2-B0 (‰)	δ13CCO2-B8 (‰)	CO₂ emission rate (mg CO₂-C kg−1h−1)	Soil - CO₂ emission rate (mg CO₂-C kg−1h−1)	Biochar - CO₂ emission rate (mg CO₂-C kg−1h−1)	fbiochar
Forest	60	−8.424	−7.513	0.1957	0.1702	0.0254	0.130
	120	−3.072	−0.236	0.6628	0.3937	0.2691	0.406
	180	−4.225	−3.641	0.3536	0.3239	0.0297	0.084
	240	−7.098	−6.632	0.1574	0.1469	0.0105	0.067
	300	−10.191	−9.376	0.5641	0.4981	0.0660	0.117
	360	−10.659	−10.056	0.0580	0.0530	0.0050	0.086
Wetland	60	−8.055	−7.868	1.5038	1.4752	0.0286	0.019
	120	−6.850	-5.116	1.7991	1.4789	0.3202	0.178
	180	−5.796	−4.902	0.9611	0.8726	0.0884	0.092
	240	−6.749	−5.878	0.6811	0.6205	0.0606	0.089
	300	−9.732	−8.672	0.4934	0.4396	0.0538	0.109
	360	−10.165	−9.469	0.2817	0.2617	0.0200	0.071

δ¹³C_CO2-_B0 and δ¹³C_CO2-_B8 represent the soil CO₂ isotope measurement values for the control group and the biochar treatment group, respectively

For FB8, the cumulative CO₂ emission reached 1218.540 mg kg⁻¹ at the end of incubation (Table 5), of which 260.536 mg kg⁻¹ originated from biochar-derived carbon. Given a soil dry weight of 0.7 kg, this translates to an absolute biochar-CO₂ emission of 182.375 mg. Based on the biochar carbon content (77.3 wt%, Table 1), the mineralized biochar-C during the 360-day incubation was calculated as 0.064 g. With 56 g of biochar-C initially applied in FB8, the annual mineralization rate was extrapolated to 0.065 g year⁻¹. Consequently, the theoretical persistence of biochar in forest soil was estimated at approximately 875 years. Similarly, for PB8, biochar-C consumption was 0.041 g over 360 days, against an initial application of 21.6 g biochar-C. This yielded an extrapolated annual mineralization rate of 0.042 g year⁻¹ and a projected biochar persistence of 527 years.

Discussion

Responses of soil properties to Biochar addition

Biochar significantly increased EC in both soils, possibly due to their inherent high EC and water-soluble cation content. The increased concentration of soluble salts reduces soil colloid dispersion and promotes flocculation [45], a process that is hypothesized to enhance the aggregation mechanism in this unstable state, thereby improving the stability of soil aggregates [46, 47]. Soil permeability, a critical indicator of plant root respiration and soil quality, was significantly reduced by biochar application (Table 3). The applied biochar exhibited a highly porous structure, with a mean pore volume of 0.098 ± 0.011 cm³ g⁻¹, porosity of 60.94 ± 6.62%, and SSA of 305.52 ± 9.23 m² g⁻¹ (Table 1). Biochar did not function act as a soil skeleton with particle sizes (< 2 mm). However, its high SSA facilitated clay particle absorption [48], thereby reducing soil pore space. Biochar addition altered soil particle size distribution, with pronounced effects in forest soils but minimal impacts in peatlands.

Nitrogen and phosphorus contents are critical to soil nutrient cycling and SOC stabilization. Biochar increased the TP content in both soils during early incubation stages but had no significant effect in the later stages (Fig. 3e). The observed increase in soil TP following biochar addition has been widely documented [49, 50]. Biochar also significantly increased the TN content of forest soils (Fig. 3f). The soil C: N ratio increased with biochar addition, attributable to its inherently high carbon content.

Responses of soil carbon fractions to Biochar addition

Biochar amendment increased SOC content. Biochar pyrolyzed at 400 and 600 ℃ was shown to convert over 82% of labile carbon into aromatic structures [51]. It enhanced soil cation exchange capacity, water retention, and organic matter content, which promoted soil aggregate formation [52]. These aggregates function as primary SOC reservoirs by physically isolating and encapsulating organic carbon, thereby inhibiting microbial decomposition [53]. Furthermore, microaggregate organo-mineral complexes enhance long-term carbon stabilization [54]. Soil DOC content is influenced by SOC and moisture. Labile carbon derived from biochar contributes to higher DOC levels [55], which are readily utilized by microorganisms [56]. However, prolonged incubation reduces this effect as biochar-derived DOC decomposes and adsorption capacity equilibrates [57]. DOC concentration is positively correlated with soil moisture [58], which accounts for elevated DOC levels in peatlands due to their higher moisture content. However, excessive moisture in peatlands limits the ability of biochar to regulate DOC dynamics [59].

Biochar modified soil carbon fractions, MAOC and POC increased in forest soils, whereas peatlands showed MAOC enhancement with POC reduction (Fig. 3c and d). This divergence reflects ecosystem-specific mechanisms. In low-SOC forests, biochar was classified as POC or a substance with similar particle size [60] and promoted aggregate formation [61], which physically protected native POC while reducing decomposition [62]. Meanwhile, the long-term regulatory effect of biochar on soil carbon components may be closely related to its chemical aging process. Subsequent soil oxidation increased biochar’s carboxyl (-COOH) and carbonyl (C = O) groups [63], which enhances mineral complexation, facilitated the formation of organic-mineral complexes [64, 65], and transformed POC into MAOC. This explained greater MAOC increase in Cycle 2 compared to Cycle 1 in forest soils. Conversely, in anaerobic peatlands, the porous structure of biochar enhanced local aeration [66], which accelerated aerobic decomposition of protected POC (e.g., partially decomposed fibers/lignin) [67]. Concurrently, aging-induced pore collapse [68, 69] reduced SSA in PB8 from 570.02 to 272.68 m² g⁻¹ (Table 2), which weakened physical unstable carbon protection (Fig. 3d). Future studies should quantify aging rates by tracking functional group evolution during freeze-thaw cycles, which fragment particles and expose fresh oxidative surfaces.

While biochar universally enhances carbon stabilization, its net effect depends on ecosystem properties. For low-SOC forests, the addition of biochar shows significant potential to simultaneously increase both labile (POC) and stable (MAOC) carbon pools, which made it an efficient carbon sequestration and emission reduction strategy. However, it risks accelerating labile carbon decomposition in high-SOC peatlands, which causes transient carbon loss. When biochar was applied in such ecosystems, the potential risks of carbon loss against the long-term net carbon sequestration benefits were thoroughly evaluated.

Carbon emissions as influenced by Biochar application

Biochar induced sustained reduction in forest soil CO₂ emissions over 360 days, significantly reducing cumulative emissions (Fig. 5). This aligns with mitigation effects in silt loam forests [70], with quantified potential reaching up to 0.92 Gt CO₂ yr⁻¹ [71]. The macropore structure and adsorption capacity of biochar stabilize SOC through pore occlusion and organo-mineral binding. This reduces the CO₂ mineralization rate by inhibiting and altering microbial activity [72]. Increased soil C: N ratios following biochar amendment (Table 4) were correlated with reduced decomposition rates [73], further suppressing mineralization. A negative correlation between biochar C: N ratios and cumulative CO₂-C emissions was confirmed [74]. Reduced soil permeability (Table 3) and restricted gas diffusion due to biochar-induced water retention alterations were identified as contributing mechanisms. Enhanced microbial carbon use efficiency following biochar addition was shown to decrease primary SOC mineralization [75].

A study based on Florida temperate soils found that biochar produced from herbaceous materials at low pyrolysis temperatures (such as 250 °C and 400 °C) typically promotes CO₂ emissions particularly during the early stages of incubation, which attributed to the unstable components of biochar and the co-metabolism of microorganisms [21]. However, the study significantly inhibited CO₂ emissions at all seasonal stages, even in the early stages of incubation (Fig. 5). The difference is mainly attributed to key variations in experimental conditions and environmental contexts, the soil in this study has a high SOC content (Fig. 3), where organic-mineral complexes formed in high SOC environments may simultaneously stabilize both the existing SOC and the newly added biochar carbon [15], thereby masking potential positive stimulation effects in the early stages of incubation. Furthermore, the low incubation temperatures (0–10 °C) during early period suppressed microbial activity, which inhibited the likelihood of increased CO₂ emissions in the early stage.

δ¹³C analysis quantified the contribution of biochar to soil GHG emissions. Lignin and lipids in soil were ¹³C depleted, while sugars, amino acids and hemicellulose were ¹³C enriched [76, 77]. Glucose and sucrose, which are ¹³C rich substrates, were preferentially degraded [78, 79]. Their increasing contribution to CO₂ emissions during the early incubation period resulted in a gradual rise in δ¹³C_CO2 values. δ¹³C_CO2 reached its peak at 120 days in forest soil, but this peak was delayed until 180 days in peatland due to the high porosity and water retention capacity of soil that extended oxygen diffusion pathways. These peaks represented the maximum contribution of ¹³C enriched substrates to CO₂ production. As these readily degradable substrates became depleted, microbial communities increasingly decomposed ¹³C depleted organic matter, which led to a subsequent decrease in δ¹³C_CO2 values. In biochar-amended soils, the instability of organic carbon components in biochar led to its mineralization during the early stages of incubation [80], which resulted in a peak of δ¹³C_CO2 in forest and peatland soils at 120 days. Among this, f_biochar in forest soil accounted for 0.406, supporting the assertion. This stabilized near 10% after depletion of unstable fractions, which confirmed minimal emissions from stable carbon.

Biochar enriches Actinobacteria with complex organic matter degradation capabilities in forest soils [81, 82], and this was possibly through specialized niches provided by the biochar itself or the adsorbed complex carbon sources. The competitive advantage of Actinobacteria suppresses other bacterial groups through resource competition (carbon/nitrogen/space) and antibiotic secretion [81], which leads to a decrease in bacterial diversity (Fig. 2b and c). Although the reduction in diversity indicates functional simplification, Actinobacteria effectively utilize carbon derived from biochar and recalcitrant organic matter, which reduces the mineralization demands of native SOM [83]. This altered metabolic strategy interacts synergistically with the improved soil properties conferred by biochar, which enhanced carbon retention and establishing Actinobacteria as core biotic factors in carbon mitigation in forest soils. Cross-domain interactions further regulate microbial balance, decomposition intermediates derived from Actinobacteria may provide substrates for fungi such as Mortierellomycota [84], while the reduction in bacterial diversity alleviates competition pressure between bacteria and fungi.

Biochar reduced the cumulative oxidation of CH₄ in forest soils throughout the incubation period (Fig. 5). Soil TN content was increased by biochar amendment (Table 4; Fig. 3f). The presence of metal ions and organic matter in biochar (Table 1) was found to inhibit CH₄ oxidation through interactions with methanotrophic bacteria. The CH₄ oxidation rate was reduced or maintained at baseline levels due to inhibitory compounds in biochar, which included organic matter and metals [85]. Soil moisture alterations were shown to affect the exchange of CH₄ and O₂ between soil and the atmosphere [86], modify oxidation-reduction potentials [87], and regulated methanogen and methanotrophic bacterial activity that governs CH₄ oxidation. Biochar application was associated with reduced soil permeability (Table 3) and enhanced the water retention capacity.

Biochar reduced the cumulative emissions of CH₄ from peatland soils (Fig. 5). Soil NO₃⁻-N content was found to increase (Fig. 3h), which correlated with reduced CH₄ production. High soil moisture restricted CH₄ and O₂ diffusion, thereby limiting CH₄ depletion rates. Biochar-induced pore space expansion was counteracted by water saturation, which resulted in minimal aeration improvements. Elevated moisture levels suppressed microbial activity through restricted O₂ diffusion, delaying organic matter decomposition derived from biochar and extending CH₄ oxidation duration. In waterlogged conditions, CH₄ emissions were further reduced through methanogen inhibition, methanotroph stimulation, and enhanced CH₄ adsorption [88].

Limitations and prospects

This study systematically evaluated biochar-induced changes in permafrost soil properties and carbon mitigation mechanisms under simulated seasonal conditions, yet experimental limitations persist. Despite efforts to approximate field conditions, the indoor incubation system cannot fully replicate complex natural ecosystems due to the absence of live root systems with associated rhizosphere processes (carbon inputs, microbial interactions). The inability to reproduce physical freeze-thaw disturbances (e.g., soil structure disruption and protected organic matter release) despite temperature simulation, and the exclusion of natural variables including precipitation dynamics, wind-driven gas exchange, faunal activity, and soil-permafrost spatial heterogeneity. Future research should therefore integrate controlled experiments with long-term field monitoring to (1) quantify multi-year GHG fluxes, (2) assess biochar-vegetation interactions, and (3) evaluate freeze-thaw impacts on biochar stability.

Additionally, optimizing and modifying biochar research holds promise for further enhancing its efficacy in reducing soil carbon emissions, and future studies should strengthen experimental exploration targeting modified biochar. Future field studies should pay special attention to the influence of particle size on the mobility and spatial distribution of biochar. For field application, high-pressure injection of granular material into permafrost soils has been proposed as a technique for large-scale soil amendment [89]. However, complex terrain (slopes, obstacles) impedes equipment transport and risks root damage. Biochar application is better suited to specific scenarios such as ecological restoration (e.g., post-fire, post-logging, eroded areas), contaminated soil remediation, or soil improvement in production-oriented plantations.

Conclusions and recommendations

A 360-day incubation study in permafrost forest and peatland soils showed that biochar application increased SOC content, reduced permeability, and enhanced silt fraction stabilization. Biochar effectively reduced soil CO₂ emissions, forest soil CH₄ uptake, and peatland soil CH₄ emissions. Actinobacteria enrichment and the resulting decline in bacterial diversity constitute the central biological driver of carbon mitigation in forest soils. For soil gas isotope analysis, the early contribution of biochar to soil GHG emissions was related to the unstable C in biochar, which was preferentially mineralized to a certain extent compared with primary SOC. But this tends to stabilize after 120 days of incubation. This confirms the potential of biochar to reduce soil carbon emissions and enhance carbon storage in permafrost areas. The study highlights the potential of biochar to mitigate permafrost carbon-climate feedback by strategically modulating soil physicochemical properties (e.g., C: N ratio, aggregate formation). These findings advocate for optimized biochar deployment as a key strategy for regulating carbon emissions in vulnerable permafrost ecosystems.

Author contributions

Haiyin Wu: Data curation, Formal analysis, Investigation, Methodology, Resources, Software, Supervision, Visualization, Writing – original draft. Shuying Zang: Funding acquisition, Methodology, Resources, Validation, Writing – review & editing. Hanxi Wang: Conceptualization, Methodology, Funding acquisition, Software, Project administration, Supervision, Writing – review & editing. Dianfan Guo: Formal analysis, Investigation, Software, Validation, Writing – original draft. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

Funding

The present study was gratefully supported by the Science & Technology Fundamental Resources Investigation Program (No. 2022FY100701) and High-level Talent Foundation Project of Harbin Normal University (No. 1305124219).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.