Saline water irrigation is widely used to alleviate water shortage in arid regions. The Taklimakan Desert Highway Shelterbelt (TDHS) has been drip-irrigated long-term using high saline groundwater. However, the effects of saline irrigation on the spatiotemporal dynamics of soil CO2 flux along the TDHS and its underlying mechanisms remain unclear. In this study, we systematically monitored soil CO2 flux along the TDHS and explored its underlying mechanisms under saline drip irrigation with varying salinity levels over two years. The results demonstrated that high-salinity irrigation significantly reduces soil CO2 flux, exhibiting a distinct spatiotemporal variation pattern characterized by a diurnal single-peak curve peaking around 18:00. Within the same irrigation cycle, soil CO2 flux gradually declined over time. Seasonally, soil CO2 flux was significantly higher in summer than in spring or autumn. Spatially, soil CO2 flux in the inter-canopy zone (Z1) was lower than in the under-canopy zone (Z2). These variations were mainly attributed to the effects of irrigation water salinity and sampling location on soil carbon pools, enzyme activity, and soil physicochemical properties. Thus, we conclude that high saline drip irrigation significantly restrains soil CO2 flux through changing soil physicochemical and biological properties. This study provides a scientific basis for accurately assessing the soil carbon sink function of artificial desert shelterbelts in extreme arid areas, while establishing a theoretical foundation for sustainable management and utilization of desert shelterbelts.

Understanding the competitive-synergistic mechanisms between clay minerals and humic substances in uranium (U) immobilization is critical for mitigating U pollution, yet remains elusive. Here, we employed molecular dynamics simulations to unravel how montmorillonite (MMT) and humic acid (HA) regulate uranyl (UO22+) fate via composition-dependent clustering and interfacial coordination. Systems with varying MMT-HA ratios were simulated to quantify adsorption dynamics, cluster evolution, and energy-driven pathways. Key findings reveal that MMT and HA synergistically immobilize UO22+ through hierarchical "mineral-uranyl-organic" bridging networks, yet their efficiency depends critically on concentration ratios. Excess HA disrupts uranyl self-aggregation via competitive carboxylate coordination, while MMT dominates initial uranyl anchoring via basal/edge-surface sites. Notably, HA promotes MMT agglomeration via sodium-carbonate-uranyl bridges, while carboxyl groups (-COO-) dictate uranyl-HA binding. Energy analyses confirm HA's thermodynamic advantage in trapping uranyl, though MMT surfaces provide preferential adsorption sites. This study provides molecular-scale insights into the dynamic competition and ratio-dependent synergy at mineral-organic interactions. By moving beyond static adsorption models and simulating the co-evolution of components, we establish that optimal immobilization is governed not by individual component capacity but by interfacial stoichiometry and cluster-mediated bridging. This work provides a mechanistic foundation for advancing predictive migration models and guiding ratio-based remediation strategies in complex environmental systems.

Restoring ecosystem functions in sandy saline-alkali soils is intrinsically linked to microbial dynamics. This study investigated how organic and biological amendments (manure, biochar, and a microbial inoculant) influence soil properties, enzyme activities, and microbial communities across a 0–60 cm profile to identify depth-dependent restoration mechanisms. A field experiment was conducted using sheep manure, biochar, and a Bacillus-based inoculant. We integrated physicochemical analyses, extracellular enzyme assays, and high-throughput sequencing to evaluate microbial assembly and co-occurrence networks across stratified soil layers. The 20–60 cm subsoil emerged as a critical microbial restructuring hotspot. Amendments significantly reduced electrical conductivity by up to 51.1

The unsaturated transport of salt solutions with clay particle capillary pores holds significant relevance across soil processes and function. However, the transport dynamics, interfacial interactions, and specific behaviors remain elusive. To unravel these complexities, molecular dynamics simulations combined with infiltration experiments were employed. Notably, during capillary transport, the solution formed a curved meniscus; infiltration depth followed the Washburn equation with ion-specific deviations, linked to interfacial slip behavior (a key nanoscale process unstudied in illite before). Infiltration ability of salt solutions decreased with increasing ionic concentration, following the sequence Cs+ < K+ approximate to Na+ < Ca2+. This phenomenon is governed by two primary mechanisms: 1) driving effect of water molecules on cations and 2) the competition between ion hydration and clay particle attraction. Distinct behaviors emerged during unsaturated capillary transport, including the initial meniscus recession prior to capillary entry and distinct diffusion dynamics versus infiltration dynamics. While MD simulations differed slightly from infiltration experiments, they uniquely illuminated visual infiltration processes, dynamics, and microscopic mechanisms inaccessible through experiments alone. These results not only fill gaps in understanding illite-based unsaturated transport but also provide a quantitative framework for refining soil water models-advancing the field beyond saturated, single-method, or montmorillonite-centric studies.

Humic acids (HAs) extracted from soils under different vegetation restoration stages exhibit distinct molecular fingerprints, yet how these features interact with pH to regulate HA adsorption onto clay mineral such as montmorillonite, and the implications for soil carbon stability, remains poorly understood. Here, we combine molecular dynamics simulations, spectroscopy (XPS, FTIR), and batch adsorption experiments to unravel the mechanism by which humic acids (HA) extracted from soils under herbaceous, shrub, and arboreal vegetation exhibit distinct molecular fingerprints, thereby controlling pH-dependent adsorption behaviors. We show that arborealderived HA possesses the highest carboxyl content, whereas shrub-derived HA is enriched in aromatic and hydroxyl groups. Under acidic conditions (pH 5.3), carboxyl abundance governs adsorption via ligand exchange and hydrogen bonding. In contrast, at alkaline pH (8.7), electrostatic repulsion reduces overall adsorption but reverses the affinity order, with shrub-derived HA exhibiting the greatest stability due to compensatory hydroxyl H-bonding, Ca2+ bridging, and it-it interactions. Molecular dynamics simulations visually corroborate the dispersion of carboxylrich HA versus the sustained adsorption of aromatic/hydroxyl-rich HA under high pH. Our results demonstrate that the vegetation-driven carboxyl/hydroxyl ratio dynamically shifts the dominant adsorption mechanism from carboxyl coordination to hydroxyl-aromatic stabilization along a pH gradient. This work provides a molecular-scale framework for understanding and predicting SOM persistence in mineral-organic associations under environmental change.

Aims Legume-grass intercropping is an important strategy for restoring degraded grasslands and improving productivity in Southwest China.Arbuscular mycorrhizal fungi(AMF)play a key regulatory role in nutrient uptake within these systems.However,the mechanism by which AMF regulates nitrogen uptake and allocation in intercropping communities composed of plants with varying rooting depths remains unclear. Methods Based on differences in nitrogen-fixing traits and root depths,this study selected a deep-rooted legume Medicago sativa and a shallow-rooted legume Trifolium repens,intercropped with two grass species,Dactylis glomerata and Lolium perenne.Monocultures and mixtures were established with or without AMF.Using 15N labeling at two soil depths 3 cm shallow and 25 cm deep,we explored the effects of AMF on the nitrogen acquisition strategies and community functions of plants with different root depths. Important findings Intercropping significantly increased the total community of the plant community.The combination of T.repens+D.glomerata+L.perenne performed best,increasing yield by 54.78％and 76.58％compared to T.repens or L.perenne monocultures,respectively.This highlights a significant belowground niche complementarity between shallow-rooted legumes and grasses,which enhanced resource use efficiency.Conversely,the deep-rooted M.sativa showed spatial root overlap with grasses,intensifying interspecific competition and diminishing in insignificant intercropping advantages.AMF inoculation significantly promoted biomass accumulation in the strongly mycorrhizal-dependent legumes while moderating the competitive dominance of the less-dependent grasses,thereby increasing the legume proportion within the community.Furthermore,AMF induced plasticity in root morphology,significantly increasing the root length of T.repens and L.perenne.15N tracing revealed that intercropping promoted legume nitrogen fixation rates;AMF inoculation further strengthened this process and reduced the uptake of soil 15N by legumes.Meanwhile,AMF narrowed the difference in 15N uptake by grasses across soil layers,promoting more efficient utilization of deep-soil nitrogen.AMF achieved optimal resource allocation at the community level by enhancing legumes nitrogen fixation efficiency and promoting deep soil nitrogen acquisition by grasses.This study verifies that the shallow-rooted legume+grass mixture combined with AMF inoculation is the optimal strategy for achieving belowground spatial partitioning and high aboveground yield,providing a theoretical reference for the establishment of artificial grasslands in Southwest China.

Despite the recognized role of vegetation in shaping soil dissolved organic matter (DOM), a systematic understanding of how major vegetation classes (arbors, shrubs, grasses, crops) differentially drive DOM molecular diversity and thereby control heavy metal complexation remains lacking. This study fills this critical gap by employing an integrated multi-spectroscopic and molecular dynamics (MD) simulation approach to comprehensively characterize DOM molecular signatures across four vegetation types and to elucidate their distinct cadmium (Cd) complexation mechanisms. Results reveal that vegetation type fundamentally structures DOM chemistry: arbor-derived DOM is relatively enriched in hydrophilic, oxygen-containing functional groups (e.g., carboxyl) that form dispersed molecular clusters via hydrogen bonding, maximizing carboxyl site accessibility and enhancing Cd complexation capacity (critical coagulation concentration = 12.56 mM). In contrast, grassland DOM contains a higher proportion of hydrophobic aromatic structures that aggregate into compact clusters, increasing steric hindrance and reducing carboxyl accessibility, thereby weakening Cd immobilization (CCC = 10.6 mM). MD simulations further confirm that Cd forms stable inner-sphere coordination primarily with carboxyl-O, with coordination numbers following grassland (1.57) > cropland (1.33) > arbors (1.17) > shrubs (1.0). This work establishes the first mechanistic framework linking vegetation type to DOM molecular diversity and subsequent Cd binding behavior, providing a molecular-level basis for optimizing vegetation design in contaminated soil remediation and improving predictive assessment of heavy metal fate in terrestrial ecosystems.

Arid and semi-arid ecosystems store substantial amounts of soil organic carbon (SOC), critical to the global carbon cycle and climate mitigation. In the Mu Us Desert (MUD), one of China’s four main sandy regions, the spatial distribution and controlling factors of SOC, especially in deep soils, are not adequately understood. This study integrated high-density SOC sampling from 0–200 cm with comprehensive environmental data, including soil properties, geographic features, meteorological conditions, and vegetation characteristics. The Random Forest Model (RFM) simulated the distribution of SOC in different layers, and its %IncMSE, along with correlation analysis and a Structural Equation Model (SEM), clarified the mechanisms regulating SOC patterns. SOC was enriched in shallow soils and showed greater accumulation in deep soils of the western regions. Total storage of SOC was 101.50 Tg, exhibiting a depth-wise distribution with a peak at 0–10 cm, a minimum at 20–40 cm, and a secondary peak in the deeper layers. Correlation analysis, IncMSE%, and SEM consistently indicated that SOC distribution was mainly controlled by clay and silt (Csil) contents. In the 0–40 cm layer, soil inorganic carbon (SIC), electrical conductivity (EC), and pH directly and significantly affected SOC, while meteorological factors influenced SOC indirectly via vegetation and SIC. In the 40–200 cm layer, Csilt remained the dominant controlling factor, while the influences of pH and SIC weakened. These findings provide a scientific basis for carbon management and ecological restoration in the MUD and support sustainable development of arid and semi-arid ecosystems.

The influence of biochar on N cycle is well-known, but specific biological and chemical mechanisms under biochar with different C:N ratios and soil types, remain inadequately unclear. This study aimed to explore the changes in N cycle after adding N-enriched biochar (NB) with varying C:N ratios to loamy and sandy soils. Experiment included nine treatments: three NB treatments (NB1, NB2, and NB3) and a pristine biochar (PB), applied at 20 t ha(-1) (L1) and 40 t ha(-1) (L2), along with a control (CK). NB significantly increased N fractions, nitrification, ammonification, mineralization, and soil enzymatic activities in loamy soil than in sandy soil. NH4+-N and NO3--N were maximum in NB treatments during first 15 days, while NO3--N levels were higher in CK at later stages. NB effectively increased soil TN, SOM, AK, and AP compared to PB and CK in both soil types. In loamy soil, NB increased cumulative N2O emissions by 157.3 % to 229.5 %, while PB reduced emissions by 14.7 % at L2. In sandy soil, PB and NB significantly reduced cumulative N2O emissions, with the greatest decrease (39.2 % to 86.1 %) at L2. Structural analysis showed that N fractions significantly influence N transformation in loamy soil, whereas soil properties and N fractions affect N pathways in sandy soil. We thus demonstrate biochar's C:N ratio and soil type are crucial in influencing N transformations and N2O emissions. These findings are crucial for developing targeted biochar application strategies to enhance soil fertility and reduce greenhouse gas impacts with potential applications across global agroecosystems.

Understanding plants responses to drought stress is crucial for selecting appropriate species for shelter-forest construction in arid and semi-arid regions. Calligonum caput-medusae, one of the most planted shrubs along the Taklimakan Desert Highway Shelterbelt (TDHS), contributes significantly to maintaining the highway’s ecological stability. This study aimed to investigate the physiological responses of biennial C. caput-medusae seedlings to drought stress by monitoring changes in soil moisture and chlorophyll fluorescence parameters [actual photo chemical efficiency of PSII (Y(II)), unregulated energy dissipation quantum yield (Y(NO)), non-photochemical quenching coefficient (NPQ), and regulatory energy dissipation quantum yield (Y(NPQ))] under controlled conditions. The results showed that soil moisture declined progressively with prolonged drought stress. Although the photosystem II (PSII) reaction centers of the seedlings experienced some stress after 30 days of drought, no irreversible photodamage occurred. However, the risk of photoinhibition and damage to the photosynthetic apparatus increased with prolonged drought, as evidenced by an increase in NPQ. These findings suggest that C. caput-medusae seedlings adapt to drought stress by modulating their chlorophyll fluorescence characteristics, enhancing our understanding of its drought adaptation mechanisms and highlighting the need for future research on its long-term physiological responses under field conditions and varying drought intensities.

Role of varying C:N ratios of biochar to enhance soil health and impacts across soil types remain inadequately understood. Therefore, this study aimed to elucidate the mechanisms of change in soil organic carbon (SOC) fractions and greenhouse gas (GHG) emissions under nitrogen‐enriched biochar (NB) application in sandy soil and loamy soil. A control (CK; without biochar) and four biochar types, including one pristine biochar (PB) and three types of NB (NB1, NB2, and NB3), were applied at two rates (20 t ha −1 [L1] and 40 t ha −1 [L2]). Biochar types and application levels significantly influenced CH₄ and CO₂ emissions. NB effectively reduced CH₄ while increased CO₂ emissions in both soil types. NB enhanced the SOC pools, which is primarily attributed to increased mineral‐associated and particulate SOC in both sandy and loamy soils. β‐glucosidase, cellobiohydrolase, and β‐xylosidase were significantly enhanced by NB, particularly at a higher application rate of biochar (L2) compared to CK, with more pronounced increases in loamy soil. Structural equation modeling showed that biochar types and application levels significantly influenced CH₄ and CO₂ emissions and SOC. The results provide valuable insights for guiding biochar applications aimed at reducing CO₂ and CH₄ emissions while improving soil fertility, with potential benefits for diverse agroecosystems and farming communities worldwide.

Intensive rice cultivation to meet the rising global demand depend heavily on fertilizers and water application which leads to environmental pollution, affects human and aquatic lives, and inhibits crop nitrogen (N) use efficiency. To overcome these challenges, it is essential to implement strategies that improve nutrient use efficiency (NutUE) while optimizing crop growth and physiological performance. Therefore, a pot trial was executed to explore the potential of various organic and inorganic amendments to improve NutUE and physiological attributes of rice. Treatments included biochar (BC), compost (Comp), two types of bio-gypsum (Biogyp1 and Biogyp2), gypsum (Gyp), and a Ca waste by-product (Cawp), which were applied at a rate of 1

Wind erosion leads to significant loss of soil and soil organic carbon in arid and semi-arid desert ecosystems, posing a serious threat to sustainable agriculture and ecological security. Soil salt crusts (SSCs),1 widely distributed on soil surfaces in these regions, have a notable impact on surface processes, including the initiation, release, and transport of soil particles. However, there is limited understanding of the role of SSCs in reducing wind erosion and PM10 emission, as well as the surface changes of crust-covered soils during wind erosion. In this study, we used wind tunnels to evaluate the influence of artificial SSCs on wind erosion rate (WER),2 surface characteristics, and PM10 emission during wind erosion processes. The results showed that as the salt concentration of irrigation water increased, the hardness and shear resistance of SSCs also increased, the structure became denser, but the thickness decreased. The effectiveness of SSCs in reducing WER and inhibiting PM10 emission also improved with the increasing salt concentrations. Additionally, SSCs delayed and reduced the geomorphological changes in the soil surface during wind erosion. We conclude that irrigation water salinity affects SSC formation, altering the structure and strength of the soil surface. Once SSCs are destroyed, WER and PM10 emissions may increase significantly. Thus, in addition to supporting plant growth, saline water irrigation in arid and semi-arid regions can form soil salt crusts (SSCs) that serve as an effective measure against wind erosion.

Glycyrrhiza glabra L., a commercially important licorices species, is rich in flavonoids with significant medicinal properties. Phytohormone jasmonates play a pivotal role in modulating flavonoid biosynthesis, though their specific impact on distinct flavonoid subclasses in G. glabra remains unclear. This study investigates the effects of methyl jasmonate (MeJA) and salicylhydroxamic acid (SHAM) on flavonoid biosynthesis in G.glabra hairy roots using transcriptomic and metabolomic analyses. MeJA treatment significantly upregulated key enzymes and transcription factors involved in flavonoid biosynthesis, leading to increased levels of specific flavonoids such as 8-prenylnaringenin. Conversely, SHAM treatment downregulated these genes and reduced flavonoid content. Notably, GurMYB04 emerged as a key regulator of flavonoid biosynthesis, showing contrasting expression patterns under MeJA and SHAM treatments. These findings highlight the divergent role of jasmonate signaling in flavonoid biosynthesis and provide insights for targeted metabolic engineering in G. glabra.

Aeolian sandy soils are an important carbon sink for dissolved inorganic carbon (DIC), however, the underlying mechanisms are still mystery. To solve this issue, the adsorption of HCO3- and CO32- on aeolian sandy soil particles was studied using molecular dynamics simulation and adsorption experiments. Both HCO3- and CO32- can be adsorbed, albeit with varying adsorption capacity, distributions, and local environment, which was influenced by the pH value and concentration. The adsorption mechanisms of DIC were though hydrogen bonds or cation bridges, as well as in the form of single or multiple molecular clusters, which were regulated by the pH level. At pH=7.5, HCO3- was mainly captured by hydrogen bond and in a form of single molecule; at pH=9.5 and 11, both HCO3- and CO32- were mainly adsorbed as DIC clusters through the linkage of Na+ bridge. The HCO3- cluster mainly found at pH=9.5, comprising of HCO3--Na+-CO32- complex, whereas the CO32- cluster mainly observed at pH=11, comprising of CO32--Na+-CO32-. The CO32- cluster exhibited greater stability compared to the HCO3- cluster. These results are valuable in enhancing our understanding of the carbon cycle process and carbon balance dynamics in arid and semi-arid regions.

Microbial necromass carbon (MNC) is the most important component of stable soil organic carbon (SOC). Vegetation restoration affects SOC stocks, yet the response of MNC and associated microbial mechanisms remain understudied in coal mining areas of semi-arid deserts. Here, we investigated the influence of vegetation type and soil depth on MNC accumulation and its possible drivers in a desert mining area after 11 years of revegetation. The results showed that the contents of SOC and MNC, particularly fungal necromass carbon (FNC), increased in shrubland (SL) plots compared to grassland (GL) and mixed vegetation (GS) plots. MNC contributed an average of 38.5% to SOC in the 0-100 cm soil profile (primarily topsoil) of SL plots, greater than that of GL (28.1%) and GS (28.6%) plots. Vegetation type strongly shaped soil fungal diversity, but not bacterial diversity. SL plots were enriched with Bacteroidota (r-strategists), whereas Actinomycetota, Chloroflexota, Gemmatimonadota, and Basidomycota (K-strategists) were depleted. Fungal pathotrophs, symbiotrophs, and saprotroph-symbiotrophs also increased in SL plots, alongside heightened complexity and robustness of networks dominated by positive links. Fungal diversity, community composition, and network complexity directly drove MNC accumulation, with FNC as a principal contributor to SOC. Soil depth indirectly influenced the necromass accumulation coefficient by altering MNC content through its negative effect on microbial diversity and community composition. These findings indicate that fungal rather than bacterial communities govern MNC accumulation during long-term vegetation restoration. This study provides mechanistic insight into microbial-mediated MNC dynamics under in-situ revegetation, supporting land degradation control in semi-arid deserts.

Salinization is a leading threat to soil degradation and sustainable crop production. The application of organic amendments could improve crop growth in saline soil. Thus, we assessed the impact of sugarcane bagasse (SB) and its biochar (SBB) on soil enzymatic activity and growth response of maize crop at three various percentages (0.5%, 1%, and 2% of soil) under three salinity levels (1.66, 4, and 8 dS m-1). Each treatment was replicated three times in a completely randomized block design with factorial settings. The results showed that SB and SBB can restore the impact of salinization, but the SBB at the 2% addition rate revealed promising results compared to SB. The 2% SBB significantly enhanced shoot length (23.4%, 26.1%, and 41.8%), root length (16.8%, 20.8%, and 39.0%), grain yield (17.6%, 25.1%, and 392.2%), relative water contents (11.2%, 13.1%, and 19.2%), protein (17.2%, 19.6%, and 34.9%), and carotenoid (16.3, 30.3, and 49.9%) under different salinity levels (1.66, 4, and 8 dS m-1, respectively). The 2% SBB substantially drop the Na+ in maize root (28.3%, 29.9%, and 22.4%) and shoot (36.1%, 37.2%, and 38.5%) at 1.66, 4, and 8 dS m-1. Moreover, 2% SBB is the best treatment to boost the urease by 110.1%, 71.7%, and 91.2%, alkaline phosphatase by 28.8%, 38.8%, and 57.6%, and acid phosphatase by 48.4%, 80.1%, and 68.2% than control treatment under 1.66, 4 and 8 dS m-1, respectively. Pearson analysis showed that all the growth and yield parameters were positively associated with the soil enzymatic activities and negatively correlated with electrolyte leakage and sodium. The structural equational model (SEM) showed that the different application percentage of amendments significantly influences the growth and physiological parameters at all salinity levels. SEM explained the 81%, 92%, and 95% changes in maize yield under 1.66, 4, and 8 dS m-1, respectively. So, it is concluded that the 2% SBB could be an efficient approach to enhance the maize yield by ameliorating the noxious effect of degraded saline soil.

Drought is a global environmental problem, while the effect of drought-induced unsaturation on the fate of heavy metal ions is still poorly understood, particularly the lack of mechanistic information at the molecular level. This study used molecular dynamics simulations to investigate nanoscale interactions at the montmorillonite surface under different moisture conditions. Compared to the saturated condition, drought increased the amounts and strength of Cd2+ ions adsorbed on the montmorillonite (MMT) surface while decreased the diffusivity, which was especially obvious in extreme drought conditions (θv=21%−7%). This is closely related to the compressed electric double layer, overcompensation of surface charge, and increased ion pair interactions, resulting from the confinement of water films under drought stress. Further analysis showed that the decrease of hydration effect was responsible for the exacerbated cadmium pollution. Therefore, this study may break the stereotypes about the interactions between heavy metal ions and soil minerals. The results suggest that water management (e.g., irrigation) may be prioritized before beginning heavy metal remediation.

Understanding plant water consumption is crucial for artificial afforestation under drought en-vironments and water stress in desert regions. However, the water consumption characteristics of desert species responding to the irrigation regimes are often neglected. By conducting a field test in the Taklimakan Desert Highway shelterbelt, this study examines the sap flow traits of two typical woody halophyte species (Calligonum mongolicum and Haloxylon ammodendron) and how they react to weather conditions and watering practices. Under the same irrigation treatment, the stem flux of C. mongolicum on sunny days was 1.5-5.3 times that on dusty days, while the stem flux of H. ammodendron on sunny days was 3.5-5.5 times that on dusty days. Both species demonstrated some sap flow during the night, representing 14.3%-24.9% and 7.3%-10.4% of the total sap flow for C. mongolicum and H. ammodendron, respectively. H. ammodendron maintained a higher stem flow during daytime and was more drought resistant than C. mongolicum. The daily sap flow patterns of these two species varied, showing both 'single' and 'double peak' curves depending on the watering conditions. A delay was also observed between the sap flow of these two species and the environmental factors. The factors influencing plant sap flow were found to be in the order of solar radiation, temperature, relative humidity, and saturated water vapor pressure difference. A BP-neural network proved highly effective for accurately simulating the sap flow of these two species. This research provides insights into how two common desert tree species adapt their water use in response to drought conditions, which is vital for artificial forest creation in desert areas.

The fate of heavy metal ions in contaminated soil is controlled by its interactions with clay minerals and water molecules. However, previous understanding was mainly based on saturated conditions, while indeed soil is always unsaturated. Using molecular dynamics simulation, substantially different scenery was observed under nonsaturation. Compared with saturated condition, drought increased the amounts of Cd2+ ions adsorbed on the montmorillonite (MMT) surface. Not only that, the diffusivity of Cd2+ was decreased and the adsorption strength was increased. This was especially obvious in extreme drought conditions, i.e., water content θv = 21%-7%. The promotion effect on Cd2+ adsorption is related to the destroy in the connection of clay pore under nonsaturation, which restricts the movement in water films covered on the MMT surfaces. This resulted in a specific unsaturated clay-water interface where the electric double layer was compressed, and ion pair interactions increased. The improved Cd2+ adsorption was fundamentally driven by the decrease of hydration effect, instead of MMT-Cd and Cd-Cl interactions. The results indicate that drought may be considered a predictor for aggravation of heavy metal pollution and water management (e.g., irrigation) may be prioritized before beginning heavy metal remediation.