The black soil region of Northeast China is characterized by flat terrain, thick humus layer, sustains critical agricultural productivity due to its high organic matter content and exceptional fertility, functioning as an indispensable granary for national food security^1,2. However, decades of intensive farming have led to significant degradation: conventional tillage practices (e.g., ploughing and rotary tillage) have disrupted topsoil structure, while excessive fertilizer application has accelerated soil acidification. These factors have sharply increased the mineralization rate of organic matter, degraded pore continuity, reduced biodiversity, and contributed to the formation of restrictive layers^3,4. Over the past 30 years, the thickness of the black soil layer has declined at an average annual rate of 0.32 cm⁵. In some sloping farmlands, irreversible silica neoformation accumulation layers have already developed^6–9. These layers severely hinder root development and the transport of water and nutrients, thereby threatening the region’s role in sustaining national food security¹⁰.

Recent soil investigations have revealed widespread occurrence of a silica neoformation (silica powder)-enriched horizon (designated as the silica neoformation accumulation layer) in the black soil region of Northeast China, particularly pronounced in soils developed from ancient sediments. This horizon exhibits variable occurrence depths and silicate powder contents across different regions and soil types. The silica neoformations typically manifest as angular blocky structures with grayish-white powdery appearance, predominantly coating ped surfaces. In the "Field Soil Description and Sampling Manual"¹¹, this horizon is denoted by lowercase "q" according to Chinese Soil Taxonomy¹² referring to "x" in Soil Taxonomy¹³ and often co-occurs with clay illuviation, forming silica-clay composite layers^8,9,14. Field identification relies on observing grayish-white surface coatings on structural units after rapid non-carbonate tests. When silica accumulation reaches significant levels, this horizon becomes compacted, impeding root penetration and water movement, which frequently leads to perched water tables and exacerbates slope erosion. Current understanding of the spatial distribution and genesis of this horizon remains insufficient.

The formation mechanisms of the silica neoformation accumulation layer remain complex and controversial, with ongoing debates between chemical weathering and biotic enrichment hypotheses. Early studies proposed the chemical weathering hypothesis, suggesting that silicate dissolution and translocation driven by groundwater¹⁵, along with clay mineral desilication¹⁶, are the primary causes of silica enrichment. In contrast, the biological enrichment hypothesis emphasizes the role of plant Si uptake and recycling through mineralization^14,17. Recent advancements in sequential extraction techniques have clarified the characteristics of Si fractions: although crystalline Si accounts for more than 96% of the total Si pool in Northeast China’s black soils, Si_noncry fractions (e.g., mobile Si (Si_l), adsorbed Si (Si_ad), Si bound to organic matter (Si_org), Si associated with pedogenic oxides and hydroxides (Si_occ), biogenic and minerogenic amorphous silica (Si_ba and Si_ma)) display marked spatial heterogeneity. The interplay between Si fractions transformation and pedogenic horizon development underscores the complexity of biogeochemical cycling in these soils.

Si, the second most abundant element in the Earth’s crust¹⁸, has a profound impact on the functioning of soil systems through its geochemical behavior. The activation, migration, and transformation of Si in soils are governed by multiple synergistic processes: Si_l forms Si_ad and Si_occ via adsorption-desorption and precipitation-dissolution reactions; a portion complexes with organic matter to form Si_org, and another portion is transformed into Si_ba through biological uptake and mineralization^19,20. The dynamic equilibrium of these Si fractions both influences soil structural stability and affects the global carbon cycle through mechanisms such as phytolith carbon sequestration^21,22. However, the formation of silica neoformation accumulation layers disrupts this balance: high-density siliceous particles reduce soil pore connectivity through physical barrier effects, while silicate cementation exacerbates compaction, together forming a “silica lock-in” effect⁸.

Although previous studies have examined the ecological roles of Si or focused on single-species interactions, the coupled mechanisms linking Si fraction transformation with pedogenic processes remain poorly understood. In particular, no quantitative model has yet elucidated climate-biological-geochemical linkages responsible for the genesis of silica neoformation accumulation layers in Northeast China’s black soils. This study addressed that gap by establishing a SBC quantification framework using sequential extraction scheme to quantify Si fractions transformation. Our findings clarified the climate-driven mechanisms that have shaped Si fractions dynamics since the Late Pleistocene. This work provides a theoretical foundation for mitigating compaction-related layers in black soils, advances the biogeochemical understanding of Si cycling, and offers valuable insights for predicting soil evolution under global change scenarios.

Quantitative analysis using a pedogenic reconstruction model revealed significant vertical differentiation in Si fractions indicating migrations and transformations across pedogenic horizons (Fig. 3). All horizons experienced net losses of Si_l relative to the parent material, with the E horizon showing the greatest depletion (0.006 g·(100 cm³)⁻¹), corresponding to a 53.1% reduction from the initial value. In contrast, Si_org exhibited widespread enrichment, peaking in the Ahb horizon with an increase of 0.26 g·(100 cm³)⁻¹, representing a 1.3-fold increase over the parent material.

The dynamics of the Si_tot pool reflected a bidirectional regulatory pattern: the E horizon experienced a net depletion of 2.53 g·(100 cm³)⁻¹, while the Btrq1 horizon showed substantial accumulation of 10.04 g·(100 cm³)⁻¹, indicating a pattern of surface loss coupled with subsurface enrichment. Si_res displayed signs of vertical redistribution. Depletions of 0.68, 1.24, and 2.68 g·(100 cm³)⁻¹ were observed in the Ap1, Ap2, and E horizons, respectively, whereas the Btrq1 horizon exhibited significant enrichment (9.02 g·(100 cm³)⁻¹), highlighting active downward migration and reallocation of silicon within the soil profile. Si_noncry transformation showed a more complex transformation pattern. Apart from a 0.18 g·(100 cm³)⁻¹ loss in the Btrq2 horizon, all other horizons exhibited varying degrees of enrichment, with the Btrq1 horizon increasing by 1.02 g·(100 cm³)⁻¹. Within this horizon, Si_l decreased by 0.003 g·(100 cm³)⁻¹, while Si_ad (0.014 g·(100 cm³)⁻¹), Si_occ (0.52 g·(100 cm³)⁻¹), Si_ba (0.004 g·(100 cm³)⁻¹), and Si_ma (0.42 g·(100 cm³)⁻¹) showed coordinated increases. Comparative analysis revealed that Si_occ experienced its greatest depletion in Btrq2 (0.3 g·(100 cm³)⁻¹), whereas Btrq4 demonstrated the strongest combined enrichment of biogenic and mineral amorphous silica, reaching 0.005 g·(100 cm³)⁻¹ and 0.62 g·(100 cm³)⁻¹, respectively. Notably, specific Si_org enrichment in the Ahb horizon (0.23 g·(100 cm³)⁻¹) further confirmed the prominent role of biochemical Si fixation processes in surface soils.

"SOC" is the abbreviation of Soil Organic Carbon; "BD" is the abbreviation of Bulk Density; "X_lf" represents the Magnetic susceptibility. The gradient-colored column on the right represents the correlation coefficient. The larger the absolute value, the higher the degree of correlation.

Pearson correlation analysis systematically elucidated the relationships between soil physicochemical properties and various Si fractions (Fig. 4). The results demonstrated that soil pH was highly negatively correlated with Si_ma (R = − 0.61, P < 0.001), and positively correlated with Si_res (R = 0.57, P < 0.01), suggesting acidic conditions enhanced Si_ma dissolution and potentially promoted its transformation into residual phases. Organic carbon content showed a highly significant positive correlation with Si_org (R = 0.65, P < 0.001), while displaying significant negative correlations with Si_ad (R = − 0.60, P < 0.001), Si_res (R = − 0.89, P < 0.001) and Si_ba (R = − 0.55, P < 0.01). These patterns indicate that a dual regulatory role of organic matter: it promotes biological fixation of Si through organo-silicon complexation while simultaneously destabilizing other Si fractions. Among physical properties, bulk density was positively correlated with Si_ad (R = 0.46, P < 0.01) but negatively correlated with Si_org (R = − 0.60, P < 0.001), indicating that soil compaction affects the adsorption-desorption equilibrium of Si. Additionally, sand content showed a positive correlation with Si_l (R = 0.41, P < 0.05), suggesting coarse particles facilitate Si migration within the profile.

In the context of the metal oxide system, total iron (R = − 0.45, P < 0.01) and total aluminum (R = − 0.67, P < 0.001) were negatively correlated with Si_occ, whereas total manganese was positively correlated with both Si_occ (R = 0.41, P < 0.05) and Si_ma (R = 0.44, P < 0.05). These patterns reflect differential affinities of multivalent metal oxides for various Si fractions. The significant positive correlation between magnetic susceptibility and Si_ma (R = 0.37, P < 0.05) further supports the protective role of Fe/Mn oxides in stabilizing amorphous Si phases.

Correlations among Si fractions themselves revealed a highly significant positive correlation between Si_res and Si_l (R = 0.61, P < 0.001), and a highly significant negative correlation between Si_res and Si_org (R = − 0.61, P < 0.001), indicating an equilibrium between labile and recalcitrant Si pools. Additionally, the positive correlation between Si_ba and Si_l (R = 0.45, P < 0.05) further suggests that biological cycling processes contribute to the replenishment of reactive Si pools in the soil system.

The storage of different Si fractions in the black soil profile exhibited significant vertical differentiation (Fig. 5). Si_tot storage ranged from 606.6 to 3062.3 g·m^− 2, with Si_res forming the dominant pool (870.9 − 2947.0 g·m^− 2), accounting for 96.0% to 97.3% of Si_tot. In contrast, the Si_noncry showed substantially lower storage (31.2 − 115.3 g·m^− 2), representing only 2.7% to 4.0% of Si_tot storage. The vertical distribution pattern revealed that Btrq1 and Btrq4 horizons served as core storage zones, with both Si_res and Si_noncry storage levels significantly exceeding those of other genetic horizons (p < 0.01), reflecting the long-term Si retention capacity of mid- to lower-soil horizons.

Within the Si_noncry fraction, the Btrq1 and Btrq4 horizons displayed a bimodal vertical pattern in the storage of Si_l (0.3 and 0.4 g·m^− 2), Si_ad (0.69 and 0.73 g·m^− 2), Si_occ (41.9 and 24.9 g·m^− 2) and Si_ba (0.5 and 0.6 g·m^− 2), with maximum values consistently observed in the Btrq1 horizon. Notably, the Ahb horizon showed a pronounced enrichment of Si_org (10.4 g·m^− 2), which was 0.3 to 3.3 times higher than in other horizons. This highlights the critical role of organic matter in biological Si fixation and preservation, particularly within buried surface soil layers.

In the black soil profile developed from loess-derived parent material (hilltop position, unaffected by groundwater), the transformation of Si fractions reflects multi-scale geo-biological interactions during soil formation. Variations in Si_noncry exhibit a strong positive correlation with the Chemical Index of Weathering (CIW; r = 0.97), confirming its effectiveness as an indicator of weathering intensity. Meanwhile, changes in Si_ba and Si_org reflect the driving force of surface biological processes on Si cycling. Accordingly, this study proposes a novel parameter—the Silicon Biochemical Coupling Ratio (SBC=(Variation of Si_ba + Variation of Si_org)/(Variation of Si_noncry)), to quantify interaction intensity between biological and chemical processes, defined as the ratio of biological enrichment flux to chemical weathering flux (Fig. 6).

During the Last Glacial Maximum (29.38 − 22 ka BP; corresponding to BCrq and Btrq4 horizons), cold climates significantly suppressed pedogenesis in Northeast China’s black soil region (Fig. 6). Low temperatures drastically reduced silicate mineral weathering intensity (Si_noncry=0.79 g·(100 cm³)⁻¹), inhibiting Si_l release and leading to the accumulation of Si_res to 623.28 − 635.04 mg·g^− 1. At the same time, sparse herbaceous vegetation limited biological enrichment³⁶, establishing chemical weathering as the dominant process. This regime resulted in a low SBC (0.25), with chemical weathering rate (0.11 g·(100 cm³)⁻¹·ka^− 1) significantly exceeding biological enrichment rate (0.03 g·(100 cm³)⁻¹·ka^− 1). Extensive aeolian deposition of fine particles (silt and clay) during this interval laid the parent material for subsequent black soil development.

During the interstadial phase (22 − 11 ka BP; Ahb–Btrq3 horizons), a transition to warm, humid conditions triggered the recovery of herbaceous vegetation³⁷. Microbial decomposition of herb-derived cellulose/hemicellulose released low-molecular-weight organic acids and sugars, contributing fulvic-acid-dominated humus³⁸. Carboxyl/phenolic hydroxyl groups in the humus formed stable complexes with H₄SiO₄, promoting Si_org accumulation (0.13 − 0.23 g·(100 cm³)⁻¹) and facilitating its downward migration (Ahb − Btrq3 horizons lost 0.001 − 0.002 g·(100 cm³)⁻¹). Simultaneously, glacial-interglacial freeze-thaw transitions increased porosity in the Btrq4 horizon³⁹, enabling Si_l migration into deep fractures where dehydration-precipitation processes led to Si_ma formation. Coupled with efficient biological Si uptake by herbaceous plants (often termed “biological silicon pumping”), these processes elevated the SBC to 0.8 and increased biological enrichment rates to 0.05 g·(100 cm³)⁻¹·ka^− 1.

Climatic warming during the Holocene, characterized by increased precipitation and rising temperatures, enhanced vegetation growth and promoted organic matter accumulation. During the early to mid-Holocene (11 − 4.2 ka BP), warm-humid conditions accelerated vegetation recovery and succession, with a rising proportion of woody plant⁴⁰. The deep root systems of woody vegetation secreted organic acids that strongly enhanced silicate mineral weathering, driving chemical weathering intensity to its peak (Si_noncry=1.35 g·(100 cm³)⁻¹) and laying the foundation for chernozem development. Intense leaching, indicated by low weathering-leaching coefficients (ba = 0.49 − 0.51), led to Si_ba accumulation in the Btrq1 horizon (0.12 mg·g^− 1). Elevated Si_ba levels (0.10 − 0.12 mg·g^− 1) throughout this period further reflect increased plant biomass⁴¹. Dense vegetation not only stabilized aeolian dust but also accelerated soil development and organic accumulation through root exudation and plant litter decomposition. The concurrent accumulation of dust and organic matter resulted in soil aggradation and profile thickening^42–44, forming an overlying loess layer above the buried soil. This phase was characterized by an SBC of 0.2, with high chemical weathering rates (0.20 g·(100 cm³)⁻¹·ka^− 1) but low biological enrichment rate (0.04 g·(100 cm³)⁻¹·ka^− 1), possibly attributable to lower phytolith production and return rates in northern perennial woody plants compared to annual herbs⁴⁵.

By the late Holocene (4.2 ka BP − present), climatic fluctuations shifted toward cooler and drier trends, altering vegetation composition toward herbaceous dominance and reducing woody plant cover. Around 0.42 ka BP, large-scale agricultural reclamation began^46,47. Intensified anthropogenic activities, such as deforestation, tillage, cropping, markedly increased the mineral surface area exposed to weathering. The short life cycles of crops and herbs released greater quantities of organic acids and chelators⁴⁸, enhancing chemical weathering rate to 0.25 g·(100 cm³)⁻¹·ka^− 1. Concurrently, straw return altered soil Si cycle via two synergistic pathways: fragmented straw decomposed by microbes released Si_ba, and the organic matter produced through straw decomposition, under cool and dry conditions that suppress microbial activity, enhanced humus accumulation. During this process, Si_l bound to humic substances, increasing Si_org content. These synergistic processes strengthened biological contributions, increasing biological enrichment rate to 0.09 g·(100 cm³)⁻¹·ka^− 1 and raising the SBC to 0.33.

During periods of dust deposition, the overlying soil section exerted a “spatiotemporal sequestration effect” on the underlying buried section. Progressive dust accumulation buried the original grassland vegetation, while dense dust layers restricted oxygen penetration and surface runoff. This dual protective mechanism shielded humic substance-Si complexes in the buried section from microbial decomposition, leading to an accumulation of Si_org storage of 9.61 g·m^− 2. It also suppressed organic mineralization rates, thus maintaining the long-term stability of humic acid-Si complexes. Additionally, seasonally leached humic substances migrated into the Ahb horizon, where they complexed with Si_l, further elevating Si_org content by 0.23 g·(100 cm³)⁻¹. Importantly, leached Si_l, Si_ad, and Si_ba that migrated into buried section remained protected from erosion and loss due to the overlying dust barrier. Collectively, these processes established a consistent pattern in which Si storage across all fractions in the buried section exceeded those in the overlying section.

This study reveals the vertical differentiation and biogeochemical drivers of Si fractions in a black soil profile, showing a pattern of Si_res >Si_ma >Si_occ >Si_org >Si_ba >Si_ad >Si_l. Within this framework, Si_res, which constitutes the dominant Si pool (96.02%–97.34%), forms the baseline reserve due to its association with primary silicate minerals. In contrast, Si_l functions as the central cycling nexus, regulating fluxes among all Si fractions through dissolution-precipitation equilibria. This pattern arises from synergistic interactions among key soil properties, including pH, organic carbon content, and bulk density.

Revealing the cross-scale regulatory effects of climatic evolution and pedogenesis on Si cycling. Climatic cycles, through their influence on hydrothermal regimes and vegetation patterns, have profoundly shaped Si migration and transformation in both the overlying and buried sections. Fluctuations from the Late Pleistocene (29.38 − 11 ka BP) to the Holocene (11 − 0 ka BP) drove a multi-stage evolution of Si fractions: Glacial aridity (29.38 − 22 ka BP) enhanced Si_res enrichment under weak chemical weathering and low biological activity (SBC = 0.25). Interstadial warming (22 − 11 ka BP) accelerated Si bio-cycling via intensified biological Si uptake (SBC = 0.81). While early- to mid-Holocene conditions (11 − 4.2 ka BP) supported the development of characteristic black soil Si pools through coupled strong chemical weathering and moderate biological enrichment (SBC = 0.20). Crucially, Holocene dust accumulation exerted a “spatiotemporal sequestration effect” that significantly enhanced silicon preservation in the buried section. Dense sedimentary layers simultaneously blocked oxidative erosion and suppressed microbial decomposition, maintaining the stability of humus-Si complexes and ensuring silicon storage across all fractions states consistently exceeded those in overlying section.

As a result of synergistic chemical weathering and biological enrichment, silica-enriched horizons exhibit high biogeochemical reactivity (SBC = 0.06 − 0.87) and distinctive migration and transformation characteristics. These findings provide new insights into the dynamic equilibrium of Si pools in cold-region soils.

While this study provides systematic quantification of Si fractions in black soil and elucidates the coupling among climate, pedogenesis, and Si cycling, several limitations remain. Notably, the lack of differentiation between microbially derived Si_ba and other sources may have led to underestimation of its contribution. Additionally, mineralogical characteristics and formation mechanisms of ganister sand were not fully explored. Future research should incorporate isotopic tracing and microscale characterization techniques to further resolve biogenic contributions to Si cycling and mineral transformation pathways—advancements critical for refining biogeochemical Si models in cold-region soils.