Abstract Freeze–thaw periods represent transient yet climatically significant episodes of greenhouse gas (GHG) exchange in seasonally frozen ecosystems. Projected changes in winter snowfall are expected to alter the balance among carbon dioxide (CO 2 ), nitrous oxide (N 2 O) and methane (CH 4 ) fluxes, yet these short‐lived non‐growing season pulses remain poorly constrained because winter GHG measurements are technically and logistically challenging, hindering accurate assessment of their aggregate impact on ecosystem climate forcing across environmental gradients. We combined a 1500‐km transect spanning arid to meadow grasslands with a controlled snow manipulation incubation experiment using intact soil cores to disentangle snow depth effects on freeze–thaw‐induced GHG balance and assess their consequences for ecosystem global warming potential (GWP) across ecoregions. Deepened snow increased freeze–thaw GWP primarily through increases in water‐filled pore space (WFPS), which strongly stimulated N 2 O emissions (980%–1589%) while exerting weaker and site‐dependent effects on CO 2 (32%–54%) and CH 4 (25%–100%). As WFPS increased, the relative contribution of N 2 O to total GWP surged from 1.4% to 82.8%, while that of CO 2 decreased from 98.3% to 17.1%. Meanwhile, CH 4 shifted from a weak sink (−0.6%) to a net source (1.7%). This pattern reflected a moisture‐driven trade‐off between CO 2 and N 2 O‐derived contributions to freeze–thaw GWP. Crucially, plant species richness exerted dual control over GHG emissions: it promoted CO 2 emissions through fungal‐mediated increases in enzyme activity, particularly associated with Basidiomycota, while mitigating N 2 O emissions by reducing nitrate (NO 3 − ‐N) availability and arbuscular mycorrhizal fungi‐associated suppression of denitrification. Synthesis : Our findings reveal that snow‐driven changes in soil moisture reactivated a cross‐seasonal plant–fungal legacy that shaped microbial GHG dynamics. Thus, integrating vegetation–microbe linkages alongside winter climate processes is critical to improving the prediction of non‐growing season carbon–nitrogen feedback under climate change.

Ecosystem carbon use efficiency (CUE) is a key indicator of an ecosystem's capacity to function as a carbon sink. While previous studies have predominantly focused on how climate and resource availability affect CUE through physiological processes during the growing season, the role of canopy structure in regulating carbon and energy exchange, especially its interactions with winter climate processes and nitrogen use efficiency (NUE) in shaping ecosystem CUE in semi-arid grasslands, remains insufficiently understood. Here, we conducted a 5-year snow manipulation experiment in a temperate grassland to investigate the effects of deepened snow on ecosystem CUE. We measured ecosystem carbon fluxes, soil nitrogen concentration, species biomass, plants' nitrogen concentration, canopy height and cover and species composition. We found that deepened snow increased soil nitrogen availability, while the concurrent rise in soil moisture facilitated nutrient acquisition and utilization. Together, these changes supported greater biomass accumulation per unit of nitrogen uptake, thereby enhancing NUE. In addition, deepened snow favoured the dominance of C3 grasses, which generally exhibit higher NUE and greater height than C3 forbs, providing a second pathway that further elevated community-level NUE. The enhanced NUE, through both physiological efficiency and compositional shifts, promoted biomass production and facilitated the development of larger canopy volumes. Larger canopy volumes under deepened snow increased gross primary production through improved light interception, while the associated increase in autotrophic maintenance respiration was moderated by higher NUE. Besides, denser canopies reduced understorey temperatures throughout the day, particularly at night, thereby suppressing heterotrophic respiration. Ultimately, deepened snow increased ecosystem CUE by enhancing carbon uptake while limiting respiratory carbon losses. Synthesis. These findings demonstrated the crucial role of biophysical processes associated with canopy structure and NUE in regulating ecosystem CUE, which has been largely overlooked in previous studies. We also highlight the importance of winter processes in shaping carbon sequestration dynamics and their potential to modulate future grassland responses to climate change.

Rural non-point source pollution is primarily driven by rainfall runoff. The heterogeneous landscape of rural catchments, along with variable contributing source areas, result in complex pollutant export processes during rainfall events. Consequently, making accurate simulation of pollutant export during rainfall is both essential and challenging for effective water quality management. Existing Rainfall Runoff Model (RRM) assumes uniform contributing areas with exponential pollutant wash-off during rainfall, ignoring rural landscape complexity and dynamic expansion of contributing areas, leading to systematic errors during initial and peak rainfall periods. In this study, we propose a Modified Exponential Model (MEM) that integrates Variable Source Area (VSA) theory, using two parameters, α and β, which represent the maximum extent of VSA-contributing areas and their dynamic activation rates, respectively. The model was validated and applied to dissolved nitrogen (DN) and dissolved phosphorus (DP) modelling in rural residential catchments in the Yangtze River basin. We show that the MEM significantly outperformed RRM, particularly during medium and long-duration rainfall events (Regimes II and III), with Nash-Sutcliffe efficiency (NSE) coefficients increasing by 0.15–0.24 for DN and 0.31–0.42 for DP. We also observed distinct parameter characteristics between pollutants: DN exhibited higher contributing area values (α: 0.076–0.924) than DP (α: 0.022–0.299). SHAP analysis demonstrated that runoff characteristics primarily modulate DN transport, while precipitation features more strongly influence DP dynamics. These findings not only deepen our understanding of rural pollutant export processes but also deliver a practical, interpretable modeling framework for water quality management in rural areas.

Intensified aridity beyond a critical threshold could disrupt vegetation, microbial, and soil processes, reshaping the mechanisms controlling soil carbon (C) storage in drylands. However, the aridity threshold at which the transition occurs and how the controls over different soil C fractions shift remain unclear. Here, we conducted a 2400 km transect survey across 45 sites spanning a broad aridity gradient in temperate grasslands of China. We identified a pronounced shift in the dominant drivers of soil C storage at an aridity threshold of 0.749. Below this threshold, complex vegetation structures enhanced soil C by promoting microbial activity and mineral abundance, which stimulated the accumulation of both POM and MAOM, with a stronger effect on POM. Above the threshold, fine roots dominated soil biochemical processes, sustaining microbial activity and mineral formation that indirectly stabilized SOC, particularly via MAOM. Across the entire aridity gradient, vegetation structure mediated surface soil susceptibility to wind erosion with complex structures providing effective protection, while simpler structures offered limited buffering. These findings highlight the dual role of vegetation-mediated C input and wind erosion protection in sustaining soil stocks in drylands, underscoring the need to account for canopy and root structure when species are selected for dryland restoration.

Grasslands store approximately one-third of terrestrial carbon (C) with most of it located belowground as soil organic carbon (SOC). Preserving and restoring SOC is essential for sustaining grassland ecosystem health and mitigating climate change. However, the effectiveness of soil management is constrained by limited understanding of where, how much, and how additional C can be stabilized in mineral-associated forms. Here, we combined a 2000-km field survey across temperate grasslands in China with machine learning to map the spatial distribution of mineralogical C deficit, defined as the unfilled capacity for long-term C stabilization through the formation of mineral-associated organic carbon (MAOC). We further conducted a 13C-labeled laboratory incubation experiment to identify key drivers of SOC formation efficiency. Random forest analysis revealed that mineralogical C deficit was primarily controlled by fine particle content, followed by mean annual temperature (MAT). Structural equation modeling showed that human disturbance indirectly reduced the deficit by reducing fine particles, while MAT increased it by altering soil chemistry and reducing plant cover. The largest deficits occurred in degraded and arid regions, totaling 0.78 ± 0.08 Pg C within the top 15 cm of soil. Isotope tracing further demonstrated that MAOC formation efficiency declined with increasing C deficit, as high-deficit soils were also biogeochemically degraded, characterized by lower SOC, nitrogen content, and microbial biomass, which limit C stabilization. Collectively, our study provides a spatially explicit assessment of soil C sequestration potential, highlights how degradation constrains SOC formation, and identifies priority areas for targeted restoration and climate mitigation.

Precipitation fluctuations strongly influence biomass production and its stability of terrestrial ecosystems. However, our understanding of the extent to which plant communities adjust their water-use strategies in response to non-growing season precipitation variations remains limited. Our 5-year snow manipulation experiment in a semi-arid grassland, complemented with paired stable isotope measurements of δ18O and δ13C for all species within the community, demonstrated that the impact of snowmelt on plant physiological activities extended into the peak growing season. Deepened snow enhanced ecosystem water use efficiency (WUE), biomass production, and its temporal stability. We further examined whether the observed increase in biomass stability was associated with the functional diversity of plant water-use strategies. Plant cellulose Δ18Ocell analysis revealed that both community-weighted mean and functional dispersion of stomatal conductance were positively associated with biomass production and its stability. The δ13C results further indicated that even with increased stomatal conductance, grasses were able to maintain their high intrinsic WUE by increasing photosynthesis more than transpiration. This resulted in higher biomass and greater dominance of high-WUE functional groups under deepened snow. In addition, we also found that deepened snow increased root biomass, particularly in the 0- to 5-cm and 20- to 40-cm soil layers. This increase in root biomass enhanced the uptake of snowmelt from both surface and deep soil layers, further contributing to community stability. Overall, our study demonstrates that plant communities can optimize water acquisition and utilization, thereby enhancing the stability of biomass production through coordinated changes in plant physiology, species reordering, and root distribution under altered snow regimes.

Symbioses with mycorrhizal fungi and nitrogen-fixing bacteria (NFB) enhance nitrogen (N) acquisition in host plants and may promote N transfer to neighbouring plants through mycorrhizal networks (MN). Nevertheless, the extent and mechanisms of this transfer remain unclear. On the basis of a synthesis of 15N labeling studies, we show that MN and NFB synergistically enhanced interplant N sharing. In the presence of MN, N transfer from N-fixing donors to non-N-fixing receivers increased by an average of 9.7-fold, accounting for 5.61% of the total N in receiver plants. Moreover, greater amounts of N were transferred from N-fixing plants towards their phylogenetically distant plants. Source-sink gradients driven by differences in N content between neighbouring plants further promoted N transfer. Together, our findings highlight the ecological significance of an expanded MN framework in explaining interplant N sharing and provide new insights into how symbiotic guild interactions promote species coexistence and biodiversity maintenance.

Introduction Traditionally, Silicon (Si) fertilizers are applied externally to the soil or sprayed on the leaves to reduce rice’s arsenic (As) accumulation. However, in practical agricultural production, the economic benefits of lowering As in rice grains often fail to outweigh the production costs. Objectives This study investigated an innovative technology—Si-rich seedlings (endogenous Si)—that is simple to implement, cost-effective, and highly effective in reducing As accumulation in rice grains. Methods The effects of exogenous and endogenous Si supplementation on As dynamics in rice were investigated in hydroponics in field experiments. Results Both methods significantly reduced As accumulation in the rice grains (22.23 % and 17.70 %, respectively). There was no significant difference in the treatment effects. However, endogenous Si supplementation demonstrated a significantly lower cost per unit of As reduction (88.61 % decrease), and it was easier to implement when compared with exogenous Si supplementation. The main processes by which various Si treatment techniques decrease As accumulation in rice differed, even though Si was crucial to As absorption in rice through a combination of mechanisms, such as gene regulation, iron plaque inhibition, and node I sequestration. The fundamental mechanism was Si/As antagonism, which regulated Si absorption through gene regulation in both methods. Interestingly, the exogenous Si treatment improved the barrier function of the root surface iron plaque against As. At the same time, laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) demonstrated that endogenous Si had a more pronounced impact on rice node I. Furthermore, experiments conducted at multiple scales (including variety, pot, and field experiments) validated the stability and reliability of the Si-rich seedling technology under complex environmental conditions. Conclusion Si-rich seedlings provide a cost-effective, stable, and practical solution for mitigating As contamination in paddy fields. This approach holds significant potential for enhancing soil health and improving food safety, contributing to the sustainable development of rice cultivation.

Freeze-thaw-induced N2O pulses could account for nearly half of annual N2O fluxes in cold climates, but their episodic nature, sensitivity to snow cover dynamics, and the challenges of cold-season monitoring complicate their accurate estimation and representation in global models. To address these challenges, we combined in situ automated high-frequency flux measurements with cross-ecoregion soil core incubations to investigate the mechanisms driving freeze-thaw-induced N2O emissions. We found that deepened snow significantly amplified freeze-thaw N2O pulses, with these ~50-day episodes contributing over 50% of annual fluxes. Additionally, freeze-thaw-induced N2O pulses exhibited significant spatial heterogeneity, ranging from 3.4 to 1184.1 μg N m-2 h-1 depending on site conditions. Despite significant spatiotemporal variation, our results indicated that 68%-86% of this variation can be explained by shifts in controlling factors: from water-filled pore space (WFPS), which drove anaerobic conditions, to microbial constraints as snow depth increases. Below 43% WFPS, soil moisture was the overwhelmingly dominant driver of emissions; between 43% and 66% WFPS, moisture and microbial attributes (including denitrifying gene abundance, nitrogen enzyme kinetics, and microbial biomass) jointly triggered N2O emissions pulses; above 66% WFPS, microbial attributes, particularly nitrogen enzyme kinetics, prevailed. These findings suggested that maintaining higher soil moisture served as a trigger for activating microbial activity, particularly enhancing nitrogen cycling. Furthermore, we showed that hotspots of freeze-thaw-induced N2O emissions were linked to high root production and microbial activity in cold and humid grasslands. Overall, our study highlighted the hierarchical control of WFPS and microbial processes in driving freeze-thaw-induced N2O emission pulses. The easily measurable WFPS and microbial attributes predictable from plant and soil properties could forecast the magnitude and spatial distribution of N2O emission "hot moments" under changing climate. Integrating these hot moments, particularly the dynamics of WFPS, into process-based models could refine N2O emission modeling and enhance the accuracy of global N2O budget prediction.

Rural areas lacking essential sewage treatment facilities and collection systems often experience eutrophication due to elevated nutrient loads. Understanding nitrogen (N) sources and transport mechanisms in rural catchments is crucial for improving water quality and mitigating downstream export loads, particularly during storm events. To further elucidate the sources, pathways, and transport mechanisms of N from a rural catchment with intensive agricultural activities during storm events, we conducted an analysis of 21 events through continuous sampling over two rainy seasons in a small rural catchment from the lower reaches of the Yangtze River. The results revealed that ammonia-N (NH4+-N) and nitrate-N (NO3--N) exhibited distinct behaviors during rainstorm events, with NO3--N accounting for the primary nitrogen loss, its load being approximately forty times greater than that of NH4+-N. Through examinations of the concentration-discharge (c-Q) relationships, the findings revealed that, particularly in prolonged rainstorms, NH4+-N exhibited source limited pattern (b = -0.13, P < 0.01), while NO3--N displayed transport limited pattern (b = -0.21, P < 0.01). The figure-eight hysteresis pattern was prevalent for both NH4+-N and NO3--N (38.1% and 52.0%, respectively), arising from intricate interactions among diverse sources and pathways. For NO3--N, the hysteresis pattern shifted from clockwise under short-duration rainstorms to counter-clockwise under long-duration rainstorms, whereas hysteresis remained consistently clockwise for NH4+-N. The hysteresis analysis further suggests that the duration of rainstorms modifies hydrological connectivity, thereby influencing the transport processes of N. These insights provide valuable information for the development of targeted management strategies to reduce storm nutrient export in rural catchments.

In this study, we have reported spatial and temporal variation in particulate matter (PM), sulfur dioxide (SO2), nitrogen dioxide (NO2), carbon monoxide (CO) and ozone (O3) over five provincial capital cities in northwestern China during 2013–2020. Regarding the seasonal variation, all pollutants (except ozone) exhibited the lowest concentration during summer and the highest concentration during winter, which could be attributed to increased anthropogenic activities (like coal burning) and conducive meteorological features. The highest monthly mean concentrations were primarily observed during December-February, whereas ozone exhibited the highest concentration during April-August, with different cities experiencing the highest concentration during different seasons. Regarding the diurnal variation exhibited by the pollutants, the lowest concentration of pollutants (except O3) was observed during the late afternoon (17:00–18:00) period. Ozone posed the urban site diurnal variation characteristic (peak during afternoon hour) over all sites. Urumqi had the highest PM2.5/PM10 ratio during November-March and the lowest ratio during April-October. Compared to the WHO revised guideline, the annual mean PM2.5 concentration was about 8–12 times higher, whereas the annual PM10 concentration was exceeded by a factor of up to 7. Most pollutants exhibited reduced concentration during the spring festival period. Analysis using HYSPLIT back trajectories indicated that the air masses affecting the five sites primarily originated from the northwestern area of China, although the impact of long-range pollution transport from remote regions should not be overlooked.

Understanding and predicting the aging process of exogenous selenium (Se) in soil is crucial for Se biofortification. However, the long-term aging of selenite in various soils has rarely been reported, and the key factors influencing this aging process remain unclear. Our study involved nineteen typical Chinese soils with varying physiochemical properties, all spiked with potassium selenite (1.0 mg kg(-1) Se) and incubated for 180 days. Soil available Se extracted using a 0.1 M K2HPO4-KH2PO4 solution was measured through the whole aging process. The average available Se% (the percentage of available Se in aged soils to total added Se) of all soils decreased from 55.4 % on the day 1 to 32.6 % on day 60, remaining stable thereafter. Pseudo-second-order equation provided the optimal fit (R-2 > 0.989, P < 0.01) for characterizing the dynamic process of selenite aging in soil, indicating that chemisorption, rather than internal diffusion, controlled the main rate-limiting step in the selenite aging process. Both machine learning and traditional correlation analysis indicated aging time was the most critical feature and the key soil property that contributed to available Se was pH. Empirical models incorporating soil properties and aging time were developed to predict changes of available Se in soil during aging under aerobic conditions. The reliability of the prediction model was further validated using data collected from previous studies. The developed aging model could potentially be used to scale biofortification data of Se generated from different soils under different aging times.

Forest soils are intricate ecosystems that harbor a diverse array of microorganisms, yet our current understanding of the characteristics and environmental implications of forest soil microbiomes remains limited. Here we present a continental-scale study on soil microbiomes of ten forests in China, spanning latitudes from 18°54′N to 50°48′N, resulting in a comprehensive catalog of bacterial and fungal operational taxonomic units (OTUs). Both bacterial and fungal communities exhibited discernible spatial variations in taxonomic and phylogenetic diversity. The composition of bacterial communities varied distinctively due to climatic zones—cold temperate, temperate, subtropical, and tropical—whereas fungal communities did not exhibit such pronounced distinctions. The co-occurrence network complexity of bacterial communities displayed a decremental trend along the cold temperate, temperate, subtropical, and tropical zones, whereas that of fungal communities displayed an incremental pattern. These results may be attributed to the facts that low temperatures sustain a high biomass of bacteria and foster increased interactions, while high temperatures and precipitation stimulate fungi-plant interactions. Furthermore, community assembly modeling revealed that forest soil microbial communities were dominated by stochastic processes. The bacterial community structure was mainly driven by homogeneous selection (26–41%) and dispersal limitation (35–44%), whereas dispersal limitation (51–56%) had the greatest impact on fungal community structure. Notably, bacterial functional genes associated with carbon fixation were more abundant in cold temperate soils compared to tropical soils. This study sheds light on spatial variations of forest soil microbiomes in China, enhancing further understanding of their response to global changes and implications for soil organic carbon cycles.

Soil organic carbon (SOC), as the largest terrestrial carbon pool, plays an important role in global carbon cycling, which is significantly impacted by agricultural practices. However, our ability to accurately detect and predict the impacts of fertilization and tillage on SOC dynamics is still limited. Investigating the effects of fertilization and tillage on different SOC fractions [i.e. mineral-associated organic carbon (MAOC), particulate organic carbon (POC), free POC (frPOC), occluded POC (oPOC), coarse POC (cPOC), and fine POC (fiPOC)]can aid in the understanding of overall SOC accumulation and stabilization. To this end, we evaluated the fertilization and tillage influences on SOC fractions through a global meta-analysis. We also quantified the role of environmental and agronomic factors in modulating these effects. Fertilization increased SOC fractions by mean percent change (MPC)13 %-77 %, while tillage decreased by MPC 4 %-63 %. Among them, cPOC was the most sensitive to fertilization, while frPOC had the highest sensitivity to tillage. MAOC was the least sensitive to both practices. The application of organic fertilizer increased MAOC, SOC, and POC the most (MPC 20 %-77 %), while mineral- organic fertilizer increased frPOC, oPOC, fiPOC, and cPOC the most (MPC 81 %-126 %). Fertilization in alkaline soils with warm and humid (MAT=16-24 degrees C, MAP>1000 mm) climate could maximally increase SOC contents from various fractions in surface layer (<20 cm depth), particularly when the altitude was 500-1000 m. However, tillage in acidic soils with low temperatures and rainy climate (MAT=8-16 degrees C, MAP>1000 mm) reduced the contents of SOC fractions the most in deep layer (>40 cm depth), especially at altitudes greater than 2000 m. Whether under fertilization or tillage, POC (occupying 62 %-74 %) consistently contributed more to SOC than MAOC (26 %-38 %). Overall, we suggest that SOC fractions should be prioritized over total SOC when evaluating the effects of site-specific management strategies on carbon sequestration in agricultural lands.

Increased atmospheric nitrogen (N) deposition significantly disturbs ecosystem N cycle. Although foliar interception and uptake of N deposition can provide an important alternative N supply to forest ecosystems, the mechanisms regulating foliar N uptake from wet deposition are not fully understood. Here, we selected 19 woody species with a wide range of plant traits from different functional groups and conducted a 15 N isotope labelling experiment through brushing 15 NH 4 + and 15 NO 3 − solution on canopy leaves. Our findings demonstrate that leaves can directly absorb N from wet deposition within a few hours. The average leaf 15 N recoveries were 10% and 28% under 15 NH 4 + and 15 NO 3 − treatments across species, respectively, while twig N recoveries were only 1%–7% of leaf N recoveries. Differences in foliar N uptake efficiency among species were closely associated with leaf traits but were little influenced by meteorological conditions or soil nutrient status. Specifically, plants with higher leaf N concentration, larger specific leaf area and lower wax concentration exhibited higher leaf N recovery. Our results indicated that tree canopies could directly absorb N from atmospheric deposition. We highlight the critical role of leaf traits in determining canopy foliar N uptake, which may consequently influence plant competition under elevated N deposition.