Microbially Driven Iron Cycling Facilitates Organic Carbon Accrual in Decadal Biochar-amended Soil

: Soil organic carbon (SOC) is pivotal for both agricultural activities and climate change mitigation


1
ABSTRACT: Soil organic carbon (SOC) is pivotal for both agricultural activities and climate change mitigation, 2 and biochar stands as a promising tool for bolstering SOC and curtailing soil carbon dioxide (CO 2 ) emissions.
3 However, the involvements of biochar in SOC dynamics and the underlying interactions between biochar, soil 4 microbes, minerals (notably Fe oxides), and fresh organic matter (FOM, such as plant debris) remain largely unknown, 5 especially in agricultural soils after long-term biochar amendment.We therefore introduced FOM to soils with and 6 without a decade-long history of biochar amendment, performed soil microcosm incubations, and evaluated carbon

INTRODUCTION 24
Terrestrial soils harbor the largest reservoir of active carbon in the Earth system, holding an estimated 1400-1800 Pg 25 of organic carbon within the first meter. 1Minor alterations in soil organic carbon (SOC) can profoundly influence 26 atmospheric carbon dioxide (CO 2 ) levels and consequently climate change. 2,3 ver the last 200 years, significant 27 SOC loss from agricultural land has incurred a substantial carbon debt in soils, correlating with the dramatic 28 atmospheric CO 2 elevation. 4,5 lants (e.g., litter debris and root exudates), microbes, and minerals are pivotal in 29 determining SOC dynamics.7][8][9] Meanwhile, these reports also emphasized 32 the importance of mineral-associated organic matter (MAOM) in preserving microbial-derived organic carbon in soil 33 matrix over a long period, owing to a series of physico-chemical reactions between minerals and soil organic matter, 34 and microbes. 10,11 owever, the protective effect minerals have on SOC could be counteracted by root exudates, 35 which stimulate the microbial co-metabolism and the release of organic compounds from MAOM. 12 36 Biochar, the naturally buried pyrogenic carbon, or solid carbonaceous residue derived from pyrolyzed or 37 hydrolyzed organic materials under low oxygen conditions, shows promise in building soil carbon pools and 38 mitigating climate change by reducing greenhouse gas emissions. 13,14 6][17][18] Notably, an increase 41 in MAOM around biochar and plant roots has been documented in a previous study, where applying biochar to the 42 ryegrass field over a decade stabilized both newly formed and native SOC. 191][22] The varied impacts of biochar on SOC dynamics are attributed to the characteristics of biochar and soils, as well as to the 46 co-occurrence of other organic matter and the duration of amendment, which can shape the microbial community and their survival strategies. 23,24 dditionally, increasing studies discussed the critical role of biochar as electron 48 transfer catalysts (i.e.6][27] While numerous studies have 49 documented alterations in the microbial diversity, specific communities, and the fungi-to-bacteria ratio within biochar 53 necessitate further investigation on the direction and magnitude of the priming effect, as well as the transformation 54 and activity of soil microbes in naturally biochar-amended soils. 23,29 Iron (Fe) is the most prevalent redox-active metal in subsurface environments, and its biogeochemical cycle, 56 particularly the redox reactions between Fe(II) and Fe(III) species, is closely interconnected with SOC turnover. 30,31 Recent studies showed that Fe does not inherently protect organic matter in soils, 32 especially in environments with 58 frequent redox fluctuation and heterogenous aerobic-anaerobic microsites. 33,34 he reductive dissolution of Fe oxides 59 associated with organic matter, a typical component of the MAOM, can release previously protected SOC. 32,35 he 60 oxidation of aqueous Fe(II) species can lead to the precipitation of Fe oxides exhibiting high affinities for organic 61 carbon, as well as the coprecipitation of Fe oxides and organic matter.107 At first, air-dried and sieved soils (< 2 mm, every 300 g) were placed into 1000-mL bottles and supplemented 108 with ultrapure water (18.2MΩ cm -1 ) to achieve a water content of 13%, for a14-day preincubation aiming for soil 109 microbial community resuscitation.Subsequently, 30 g of the pre-incubated soil (equivalent to 22.7 g dry weight) 110 was placed into 125 mL glass bottles, followed by either 0% or 2% (w/w) FOM addition.Ultrapure water was added 111 to achieve a soil water content of 32% (equivalent to 60% water holding capacity).The mixture was thoroughly 113 days.All bottles were sealed with microporous sealing films to maintain the oxygen supply for aeration condition 114 and covered with aluminum foils to protect from light exposure.All soils were incubated at room temperature (~24℃, 115 Figure S2a) in the dark.
116 Throughout the incubation, the three of day-56 bottles were designated for non-destructive gas sampling on days  only to rise between day 21 and day 28, eventually settling between 40-100 mg kg -1 (Figure 1c).The sharp incline 166 of DOC for CK soils on day 28 might be correlated to the raised room temperature (~26 ℃) (Figure S2a), and we

172
The MBC contents generally showed a reverse trend to DOC contents, which declined as DOC increased (Figure 1c,d).This might signify an initial rapid uptake of DOC for microbial metabolism, followed by an increased 174 decay and lysis of microbes in subsequent phases, indicating interactions among the active carbon pools.Except for 175 day 56, B-W and B-S consistently exhibited the highest MBC contents, which plateaued after an initial surge within 176 the first 7 days.This elevation could be evoked by the readily available carbon from FOM and the beneficial 177 conditions from the longstanding biochar amendment.Regarding SOC contents, BC soils contained significantly 178 elevated SOC contents (B, 31.9 ± 0.53 mg C kg -1 soil) than CK soils (C, 12.9 ± 0.2 mg C kg -1 soil) (Figure S3).After 179 a 56-day incubation with wheat straw, the B-W group exhibited a more pronounced increase of SOC (8.5 ± 0.87 mg  S6a) but a higher fungal α-diversity (Figure S6b) compared to CK soils, linking a potentially 195 diverse microbial functional characteristics function in the two types of soils.
196 Specifically, the relative abundance of Actinobacteria and Firmicutes phyla (Figure 2a) was higher in BC soils 197 than in CK soils.These phyla are known for their proficient capability in aerobically degradation of various labile 198 and complex organic molecules, particularly Streptomyces and bacillus genera, 63 which also exhibited increased 199 abundance in BC soils (Figure 2b).This suggests that biochar may facilitate the thriving of potent decomposers in 200 BC soils with limited bioavailable carbon, owing to biochar's adsorption and occlusion functions as well as its 201 inherent recalcitrance.Additionally, Proteobacteria and Actinobacteria (copiotroph, r-strategists) thrive better in 202 nutrient-rich habitats than Acidobacteria (copiotrophs, K-strategists). 64,65 uch shift in taxonomic composition could 203 reflect the change in soil nutrients conditions in BC soils.Moreover, the elevated Proteobacteria while decreased 204 relative abundance of Acidobacteria in BC soils, both predominant as illustrated in Figure 2a, might be a potential 205 reason for the lower qCO 2 observed in microbial carbon mineralization (Figure S4).These microbial and 206 physiological results, together with genome-scale models predicting a higher potential CUE of Proteobacteria 207 compared to Acidobacteria, 66 could further underscore the enhanced microbial CUE in current biochar-amended soils.

208
In addition to bacteria, soil fungi are adept at degrading aromatic carbon in biochar-amended soils, 67 especially 209 Ascomycota and Basidiomycota phyla that also dominate in the present study (Figure S7a), which may stem from 210 their well-developed hyphae and extracellular polymeric substances that could penetrate biochar and promote 211 aggregation. 61,63 esides, fungi usually comprise a large portion of microbial biomass and exhibit lower qCO 2 than 212 bacteria, 61 thus mightily contributing to the increased MBC while decreased qCO 2 in BC soils.To sum up, long-term 213 biochar application may enhance biotic degradation and assimilation of newly inputted plant debris through the 214 transformation of bacterial and fungal community structure, as well as the microbial metabolic traits.

216
Iron Redox Cycling and Hydroxyl Radical Production.The contents of highly reactive iron (Fe HCl ), extracted 217 with 0.5 M HCl, were notably lower in BC soils (~1200-2100 mg kg -1 ) than in CK soils (~2000-2800 mg kg -1 ) 218 (Figure 3a).This suggests that biochar plays a role in the sequestration of Fe in soils, either through adsorption of 219 Fe(II/III) species on biochar or due to aggregation of Fe oxides, as previously reported interactions between biochar 220 and minerals. 68Intriguingly, B-W and B-S exhibited higher ratios of Fe(II) HCl to Fe HCl at median values of 0.  2c).These distinctions illustrate that the microbially 289 mediated Fenton(-like) reactions occurred dramatically in the upland soils when FOM was introduced.Among the 290 FeRB, Bacillus, which belongs to the Firmicutes phylum, was dominant in both CK and BC soils.Recognized as a 291 facultatively anaerobic bacterium, Bacillus commonly exists in aerobic agricultural environments. 57Other FeRB 292 identified in our soils typically prefer anaerobic conditions and use Fe(III) species for anaerobic respiration. 84Their 293 relative abundance was greater in CK soils than in BC soils, suggesting that biochar may improve the aeration of BC 294 soils due to the high porosity structures.Fungi (e.g., Ascomycota, Basidiomycota) are also known for their important 295 roles in iron cycling, ROS production, and plant debris degradation. 81,85 ith abundant Ascomycota and 296 Basidiomycota fungi in BC and CK soils (Figures S6b and S7a), fungi could also play a role in Fenton(-like) reactions 297 in upland soils.

298
Overall, the relatively higher abundance of FeRB, stronger correlation between Fe(II) HCl and FeRB, and between 299 • OH and MBC in BC soils compared to CK soils collectively indicate the improved electron transfer ability of the 300 microbial community under long-term biochar amendment.This was further supported by PICRUSt2 function 301 prediction, which revealed that BC soils exhibited a heightened abundance of cytochrome-c oxidase (Figure S13), a 302 typical enzyme facilitating electron transfer through microbial cells. 86As illustrated by the FTIR spectrum (Figure 303 S1), biochar particles picked from BC soils exhibited redox functional groups (ketones, aromatic carbon, and 304 carboxyl groups at around 1700 and 1600 cm -1 ) and highly condensed aromatic sheets (overall high ratio of aromatic 305 C-H to aromatic C=C) favoring electron transfer, similar with previous studies. 45Moreover, the electron conductivity 306 of current soils showed a higher electron transfer capacity of BC soils (Table S5), determined by the conductivity 307 meter (Leici DZS-706).As previous studies reported that biochar can enhance microbial activity and soil 308 bioelectricity by electrochemical assessments, 45, 87, 88 our observations in microbial analysis and soil properties • OH in FOM-containing soils, coupled with enhanced soil CO 2 emissions (Figures 1a, 3c, and 4b), implies that • OH can break down recalcitrant organic matter, liberating bioavailable carbon with essential nutrients for soil 320 microbial growth. 42,89 herefore, in BC soils, the surplus organic carbon released as a result of iron cycling and • OH 321 attacking, was likely effectively assimilated by microbes and subsequently incorporated into MAOM, instead of 322 being mineralized thoroughly.Due to the short lifespan of • OH (half-life <10 -9 s) that limits diffusion to microbial cells and the protection effect of MAOM on microbes, the detrimental effect of • OH on microbes 42, 90 might be 324 negligible or obscured in BC soils with higher MBC contents (Figure 1d).

7
and iron cycling as well as microbial properties.Biochar amendment resulted in 2-fold SOC accrual over a decade 8 and attenuated FOM-induced CO 2 emissions by approximately 11% during a 56-day incubation through diverse 9 pathways.Notably, biochar facilitated microbially driven iron reduction and subsequent Fenton-like reactions, 10 potentially having enhanced microbial extracellular electron transfer and the carbon utilization efficiency in the long 11 run.Throughout iron cycling processes, physico-chemical protection by minerals could contribute to both microbial 12 carbon accumulation and plant debris preservation, alongside direct adsorption and occlusion of SOC by biochar 13 particles.Furthermore, soil slurry experiments, with sterilization and ferrous iron stimulation controls, confirmed the 14 role of microbes in hydroxyl radical generation and biotic carbon sequestration in biochar-amended soils.Overall, 15 our study sheds light on the intricate biotic and abiotic mechanisms governing carbon dynamics in long-term biocharsoil organic carbon, biochar, iron cycling, fresh organic matter, biotic-abiotic processes, upland 19 soil 20 21 SYNOPSIS: This study elucidated the intricate biotic-abiotic interactions of iron and carbon that underlie SOC 22 accrual in upland soils subjected to decadal biochar amendment.
7, 14, 21, 28, 35, 42, 49, and 56.Detailed gas sampling and the calculation for the cumulative CO 2 emissions 118 were presented in Text S1.Sacrificed soil samples were divided into three subsets for different analyses.One subset 119 was subjected to immediate analysis for their physiochemistry properties including carbon and Fe-related indices.120 Microbial biomass carbon (MBC) was determined using the chloroform fumigation-extraction method. 50Dissolved 121 organic carbon (DOC) was obtained using the same extraction procedure excluding the fumigation step.Since 122 previous studies demonstrated that only highly reactive iron species (extracted with 0.5 M HCl) positively correlated 123 to bio-reduced Fe(II) and • OH yield, 37, 51-53 therefore, Fe HCl extracted with 0.5 M HCl was used as highly reactive 124 iron species to understand its critical role in reactive soil iron cycling.Detailed extraction processes and the 125 determination of iron species through ferrozine method 54 were presented in Text S2.The potential for • OH formation 126 was evaluated by employing 10 mM sodium benzoate as a probe for a 10-hour reaction. 55, 56Details of these analyses 127 are presented in Texts S2 and S3.Soil pH measured at a soil-to-water ratio of 1:5 was recorded in Figure S2b.Another 128 subset was preserved in a -80 ℃ freezer for DNA extraction, which was subjected to 16S rRNA and ITS rRNA 129 amplification and Illumina sequencing for microbial community analysis as described in Text S4, with a specific 130 focus on enumerating Fe-reducing bacteria (FeRB) referring to previous reports.

167 would only focus
on the overall trends of different treatments at the moment.BC soils displayed an overall lower 168 DOC content with relatively fewer variations than CK soils.Interestingly, a legacy effect in BC soils was observed, 169 with the incline of DOC content delayed until days 30-40.These subtle differences in fluctuations hinted that biochar 170 might hinder the release of DOC derived from FOM and microbes, potentially altering the degradation pattern of 171 FOM.
(Figure3b), despite having relatively lower absolute amounts of Fe(II) HCl compared to the C-W and C-S (Figure222S8b).Considering that ferrous iron can be readily re-oxidized and transformed to amorphous Fe oxides in neutral and 223 oxic conditions, 69, 70 these results point to a higher extent of iron reduction in B-W and B-S groups, which further 224 imply an intensified in situ iron cycling near plant debris and biochar.This is similar to previous study showing 225 increased Fe(II) content and Fenton reactions at the straw-soil interface, possibly influenced by microbial activities, 71 226 and with studies reporting that the conductivity of biochar facilitated to transfer electrons for iron species in solution 227 systems. 72, 73Additionally, through SEM-EDS mapping, a stronger signal of Fe in soils with the addition of FOM 228 and/or biochar were detected (Figure S9).These findings could also indicate the formation of MAOM around the 229 FOM and biochar, with a prevalence of particles smaller than 50 μm in BC soils (Figure S10).Besides, a previous 230 study has shown the accelerated formation of organo-mineral coating on biochar surfaces and pores in the decadal 231 biochar-amended grasslands. 74Hence these increases in MAOM could be ascribed not only to physical adsorption 263 formation is primarily driven by reduced substances (e.g., Fe(II) species and humic substance), which is dominant in 264 the environments experiencing redox fluctuations. 55, 78As regards the biotic ROS formation, Han et al. have 265 illustrated both exogenetic iron-dependent and iron-independent • OH formation during the microbially mediated 266 redox transformation of Fe oxides. 79During the exogenetic iron-dependent • OH formation, the microbial excretions, 267 such as extracellular ROS and enzymes, would also play a role in Fe(II) formation and subsequent abiotic Fenton(-268 like) reactions. 80, 81Given the static aerobic conditions during our soil incubation, the abiotic process might not fully 269 explain the substantial • OH formation observed.Thus, microbial activities and microbially initiated abiotic Fenton(-270 like) reactions could predominantly account for • OH formation in our soils.271 When comparing the data for CK and BC soils, distinct linear relationships emerged with noticeably different 272 Pearson' r values all higher than 0.5 (Figure 4), indicating strong positive correlations across variables. 82Specifically, 273 the correlation between • OH formation potential and Fe(II) HCl was stronger in CK soils (0.827, P<0.001) than in BC 274 soils (0.547, P<0.001), with a lower slope in CK soils (Figure 4d).Meanwhile, BC soils demonstrated a stronger 275 positive relationship between • OH formation potential and MBC content (0.835, P<0.001) than CK soils (0.571, 276 P<0.001) (Figure4c).These findings indicate underlying differences in • OH formation between CK and BC soils.277 While biochar could act as an electron shuttle or conductor to activate pre-existing ROS (i.e., microbial-derived ROS, 278 and abiotic ROS from Fenton(-like) reactions) to yield more • OH, 75, 83 it might also initially foster the generation of 279 microbial-derived Fe(II) and • OH in BC soils.Consequently, the elevated • OH formation potentials along with higher 280 Fe(II) HCl and MBC contents would be attributed a lot to microbially mediated Fenton(-like) reactions, particularly in 281 the biochar-amended soil.282 To investigate the involvement of microbes in the generation of Fe(II) species, iron-reducing bacteria (FeRB), 283 which have a remarkable ability for extracellular electron transfer, were enumerated separately from microbial 284 communities. 52While the correlation between Fe(II) HCl content and MBC content was not outstanding (CK, 0.471, 285 P<0.001; BC, 0.380, P<0.01), a significant relationship between the relative abundance of FeRB and Fe(II) HCl content 286 (CK, 0.951, P<0.0001; BC, 0.843, P<0.01) (Figure S12) underscores the potential role of FeRB in the generation of 287 Fe(II) HCl and • OH detected in our system.Specifically, soils treated with FOM (C-W, B-W) exhibited a higher relative 288 abundance of FeRB than those without FOM input (C, B) (Figure 309 reaffirm the acclimated extracellular electron transfer capacity in the field-collected BC soils.The facilitated biotic-310 abiotic electron transfer, consequently, may have contributed to the intensified hotspots of microbially mediated 311 Fenton(-like) reactions in the biochar-amended soil.312 313 Microbial assimilation and Physical Association for SOC Accrual in Biochar Soil.The substantial presence of 314

624 Figure 1 .Figure 2 . 645 Figure 3 .Figure 4 .
Soil carbon dynamics during 56-day microcosms incubation.(a) Soil carbon loss 625 represented by cumulative CO 2 emissions and (b) the corresponding priming effects; (c) DOC 626 contents, and (d) MBC contents.C, empty green circle, control soil without FOM input; B, filled 627 green circle, biochar-amended soil without FOM addition; C-W and B-W, empty and filled triangles, 628 soils incubated with 2% wheat straw input; C-S and B-S, empty and filled squares, soils incubated 629 with 2% soybean straw input.The error bars represent the standard deviation from triplicate 630 experiments.Notes: The cumulative CO 2 emissions in (a) were calculated by integrating the CO 2 631 emission rate at each examination time interval.The cumulative priming effects in (b) were 632 calculated by the ratio of the difference between the treatment groups and the control groups to the 633 control groups, detailed in Text S1. 634 636 Microbial responses of bacteria at (a) phyla and (b) genus level, and (c) identified Fe-637 reducing bacteria at genus level.Only the top 10 identified bacteria or fungi are presented in (a), 638 and those bacteria genera with relative abundance less than 0.025 were cleared in (b).C and B, 639 control soils and biochar-amended soils without FOM input; C-W and B-W, soils incubated with 2% 640 wheat straw addition; -0d, soil samples that after pre-incubation but before incubation; -28d, soil 641 samples on day 28.Different lowercase letters above the columns in (b) indicate significant 642 differences (P<0.05)among soil samples.644 Highly reactive iron species and hydroxyl radical indexes.(a) Fe HCl contents, (b) Fe(II) HCl 646 to Fe HCl ratios, and (c) • OH formation potentials.C, soil without any treatment as control; B, soil 647 with 10-year continuous biochar amendment; -W, treatment with 2% wheat straw addition; -S, 648 treatment with 2% soybean straw addition.Solid lines inside the boxes represent the median values.649 Squares represent the mean values, while circles within each treatment depict data derived from 650 triplicate experiments at 6 time-points.The bottom and top edges of the boxes represent the 25th 651 and 75th percentiles, respectively.Whiskers demarcate minimum and maximum data points within 652 1.5× of the interquartile range.Different lowercase letters above the boxes indicate significant 653 differences (P<0.05)among separate treatments.655 Pearson correlation analyses between (a) CO 2 emission rate and MBC content, (b) CO 2 656 emission rate and • OH formation potential, (c) • OH formation potential and MBC content, and (d) 657 • OH potential and Fe(II) HCl content measured from control soils (CK, blue symbols) and biochar-658 amended soils (BC, pink symbols) soils.Blue and pink lines indicate linear fitting to the data from 659 CK and BC soils, respectively.r, Pearson correlation coefficient; P, significant level; R 2 , the 660 coefficients of determination.661 To access the differences in Fe HCl contents, and • OH formation potentials across various 135 microcosms, as well as the variations in the relative abundance of FeRB and microbial alpha diversity indices, we As no significant differences were revealed between soils with soybean and wheat straw 143 inputs, the following discussions would mostly focus on the overall impacts of FOM inputs.In the absence of FOM, 144 a comparison of the cumulative soil CO 2 emissions (CCE) from biochar-amended soil (B) to the control soil (C) 52, 57The remaining soils were stored 131 in a -20 ℃ freezer before being freeze-dried and grinned for solid morphology and chemical composition analysis 132 by scanning electron microscope with an energy dispersive spectrometer (SEM-EDS, ZEISS Sigma 300, Germany).133 134 Statistical Analysis.154 Such findings coincide with a previous study on a 9.5-year biochar-amended field, which indicated that long-term 155 biochar amendment diminished soil rhizosphere priming and enhanced underground SOC sequestration of new plant-156 derived carbon by 20% relative to the control field. 19In contrast, another study with soils after a 3.5-year biochar 157 amendment, reported a negligible suppression in the priming effect induced by cane sucrose and beet sugar. 20158 Different from the low molecular weight organic carbon compounds that are readily utilized by microbes, the FOM 159 used in the current study is relatively complex original plant debris.Nonetheless, the presence of biochar aids in the 160 preservation of newly added plant debris in this study, possibly through physical adsorption and occlusion of 161 decomposed FOM particles and FOM-derived DOC molecules, 60 which could be supported by the dynamics of DOC, 162 MBC, and SOC.163 Dynamic trends in DOC concentrations during the incubation period indicate the role of external factors and 164 biochar amendments.Throughout the incubation, DOC contents in all groups initially declined over the first 10 days, 165 Upon closer SEM-EDS examination, the 325 facilitated biosynthesis, microbial growth, and thus MAOM, were supported by fine particles embedded within the Figure5a), indicating microbial respiration as the primary contributor to the loss of SOC.Moreover, 334 after 14-day incubation, γCW and γBW showed negligible • OH formation potentials (<1.94 μM kg -1 soil) (Figure 5b), 335 confirming the critical role of microbial activities in ROS production.Owing to the necessary roles of both Fe(II) 336 species and H 2 O 2 for • OH formation via the Fenton(-like) reactions, in under dark and alkaline conditions, γ-sterilized 337 soils with low levels of Fe(II) HCl would hardly generate • OH without biogenic H 2 O 2 (FigureS14).even with comparable or lower • OH levels.Besides, γBW with aqueous Fe(II) addition showed 344 lower • OH formation but higher CO 2 emission rates compared to BW.Therefore, these disproportional findings reinforce the crucial role of microbes in BC soils, where a portion of organic carbon released by • OH attack was qCO 2 , reactive Fe and • OH, SEM-EDS characterization images, thank Lu Lu and Ruiyu Yang at China West Normal University for their assistance in microbial community analysis 395 and lab group member Gaosheng Xi for laboratory assistance.