Cunninghamia lanceolata and Quercus acutissima are common plantation species in the hilly regions of southern China. To understand the responses of radial growth to drought under the background of climate warming, we analyzed the radial growth and intrinsic water use efficiency (iWUE) of both species from 1968 to 2024, ecolo-gical resilience indices during the 1994-1995 drought event, and physiological strategies under the 2024 drought in Mopanshan Forest Farm, Jurong City, Jiangsu Province. Several variables related with dendrochronology and leaf physiological processes were measured. The results showed that: 1) The radial growth level of C. lanceolata was higher than that of Q. acutissima in the young forest stage (1973-1981), with the standard chronology index reaching a maximum of 1.7, after which it stabilized between 0.9 and 1.1, and was surpassed by Q. acutissima multiple times in the mature forest stage (after 2010). 2) In all the years except 1978, the iWUE of C. lanceolata was consistently higher than that of Q. acutissima. The sensitivity of radial growth and iWUE to climatic factors was higher in C. lanceolata than in Q. acutissima. The iWUE of C. lanceolata showed significant positive correlations with the mean temperature from the previous September-October and the current February-June, the mean maximum temperature from the previous September and the current February-April, as well as the annual mean minimum temperature, indicating that its iWUE was driven by rising temperature. 3) In historical drought events, the resistance of both species was less than 1 and recovery was greater than 1. The resilience of C. lanceolata was greater than 1 while that of Q. acutissima was less than 1, indicating that the ecological resilience of C. lanceolata was superior to that of Q. acutissima. 4) Under drought conditions, C. lanceolata sustained low stomatal conductance and transpiration rates to maintain stable photosynthesis, whereas Q. acutissima adopted a more aggressive physiological strategy, with photosynthetic rate declining rapidly under high temperatures and strong light at noon. In summary, C. lanceolata maintained stronger hydraulic safety and ecological resilience through a conservative water use strategy, while the aggressive strategy of Q. acutissima, increased drought risks. Water use strategy was the primary factor leading to the differences in radial growth and drought response between the two species.

The expansion of Moso bamboo (Phyllostachys edulis) into native forests is a globally prevalent form of forest change, with significant but uncertain consequences for soil carbon and nutrient cycling. This study elucidates the underlying mechanisms by examining a typical encroachment front in subtropical China, where Moso bamboo is replacing chestnut broadleaf forests. Using a space-for-time (SFT) approach across a gradient of forest types (broadleaf, mixed, pure bamboo), this study quantified changes in topsoil physical structure, aggregate stability, root biomass, and ecological stoichiometry (C, N, P, K). Partial least squares structural equation modeling (PLS-SEM) was applied to identify the dominant pathways driving soil organic carbon (SOC) dynamics. Bamboo expansion stimulated a more than sixfold increase in root biomass, which remodeled the soil physical environment by reducing bulk density and enhancing aggregate stability (mean weight diameter increased by similar to 30%). Chemically, it increased soil pH, total SOC, nitrogen (N), phosphorus (P), and potassium (K) pools, yet markedly depleted available P and K, resulting in elevated C/P and C/K ratios. PLS-SEM revealed that SOC accumulation was co-driven by (1) a biochemical pathway via increased pH and enhanced N cycling, and (2) a physical protection pathway via root-induced aggregate stabilization. The results demonstrate that Moso bamboo encroachment can enhance topsoil carbon sequestration through coupled physicochemical mechanisms, but concurrently triggers a strong trade-off that indicates an emerging phosphorus and potassium limitation signal. This carbon-nutrient imbalance highlights a critical constraint for the sustainable management of expanding bamboo forests globally.

Forest ecosystems regulate microclimates through hydrothermal processes, yet the mechanisms driving phase transitions between cooling and warming states remain poorly quantified, particularly with respect to interactive canopy-atmosphere controls on threshold dynamics. Through six years in situ monitoring in subtropical montane forests, we demonstrate that relative humidity (RH) modulates thermal buffering transitions. We found an asymmetric buffering pattern where intensified nocturnal warming outweighs moderate daytime cooling, culminating in net annual warming. Crucially, RH predominates over soil moisture as the primary thermal regulator, masking soil moisture cooling effects. We further quantify nonlinear threshold responses, showing that rising humidity initially enhances cooling below critical points but suppresses evaporative cooling above these values, thereby triggering warming at seasonally specific thresholds. These biophysical switches provide new empirical evidence of RH-mediated phase transitions in forest microclimate and offer a scalable framework for climate-adaptive hydroclimate management in humid regions.

Climate warming is reshaping forest ecosystems by altering both carbon and water cycles, yet it remains poorly understood whether rapid plantation expansion can match the ecological efficiency of natural forests. This knowledge gap is critical, given global afforestation efforts amidst escalating climate stress. Here, we challenge the simplistic 'high carbon gain-high water demand' paradigm by positing that plantations and natural forests operate under fundamentally distinct biogeochemical regimes. Across Northeast China (2005-2020), we integrated multi-source remote sensing and meteorological data to quantify gross primary productivity (GPP), evapotranspiration (ET), and water-use efficiency (WUE). Both forest types increased in GPP, ET, and WUE, yet plantations consistently showed higher GPP and ET but lower WUE. In long-term stable stands, mean WUE was 1.84 g C m-2 mm-1 in natural forests, higher than 1.80 g C m-2 mm-1 in plantations. Mean GPP was 1067.72 g C m-2 in plantations, exceeding 965.42 g C m-2 in natural forests, and its annual increase was markedly greater in plantations (6.04 vs. 5.63 g C m-2 yr-1). ET increased significantly in plantations (1.89 mm yr-1) but not in natural forests. Structural equation and random forest models revealed contrasting association patterns: stand structural and compositional factors made the largest contribution to WUE in both natural and planted forests, accounting for 60.9% and 52.7% of the total normalized importance, respectively, while soil factors also played an important role in planted forests, contributing 30.1%. These results indicate that plantation expansion in temperate humid forests can enhance carbon uptake, but often was associated with greater water consumption and lower WUE.

The invasion of Phyllostachys edulis leading to the gradual replacement of broad-leaved forest species by P. edulis has occurred more frequently in recent years. Thus, it is critical to understand and quantify how invasive P. edulis affects water-soluble nutrients in broad-leaved forests. In this study, we quantified the water sources of P. edulis and the broad-leaves forest species Quercus acutissima using stable isotope technology and analyzed the impact of invasion on their strategies for uptake of water-soluble nutrients. We observed that the soil water storage capacity progressively decreased, while the water storage capacity of litter proportionally increased with invasion intensity. Phyllostachys edulis mostly took up shallow soil water with high nutrient content through its well-developed shallow root system and exhibited a consequently high resource acquisition rate. Q. acutissima that originally absorbed shallow soil water were forced to uptake water from deep and stable water sources (with low water-soluble nutrient content) after being invaded by P. edulis, and exhibited positive feedback adaptation. After P. edulis invasion, although the resource acquisition ratio of Q. acutissima was low, a stable water source can ensure a sustained water supply for Q. acutissima. The ecological niche overlap index also indicated that P. edulis has a high ecological niche competitive advantage in shallow layers owing to its strong tillering ability. Therefore, the invasion of Q. acutissima by P. edulis was achieved through the capture of resources by its root systems and high ecological niche overlap. These findings suggest plant-soil positive feedback temporarily maintaining stable states in the invasion of ecosystems.

In the context of global climate change, seasonal droughts significantly impact tree water use and soil water source contribution (WSC) in forest ecosystems. However, studies on the threshold effects of the soil water content (SWC) and previous precipitation on tree water uptake are scarce. In the study, the focus was placed on the two tree species Platycladus orientalis and Quercus variabilis, which grow in the mountainous areas of northern China. Using eco-hydrological monitoring data from 3 years, we investigated the nonlinear threshold relationship between the SWC and the WSC in different precipitation treatments (zero, half, natural, and double precipitation). A combination of structural equation modeling, random forest modeling, and S-shaped curve threshold analysis was applied to evaluate the indirect effects of previous precipitation on the SWC and WSC. Sshaped threshold analysis identified SWC thresholds spanning approximately 3-26 %; Quercus variabilis showed a consistent deep-layer transition at low SWC (phi 1 = 6.3-8.4 % across zero, half and natural treatments), whereas Platycladus orientalis exhibited clear cross-layer thresholds mainly under half precipitation (phi 1 = 7.0-9.8 %). Thresholds generally shifted upward under wetter conditions. Based on the results, SWC, as the dominant factor influencing the WSC, exhibited complex and significant threshold effects at different soil depths. The two species displayed contrasting water use strategies at similar threshold levels, effectively reducing competition for water. The indirect influence of previous precipitation on the SWC and WSC also varied significantly with soil depth and precipitation amount. The results of this study highlight the complex threshold effects of the SWC on WSC in different precipitation scenarios, providing a scientific basis for understanding forest water dynamics under climate change and developing adaptive management strategies.

Frequent droughts, an increasingly common challenge under global climate change, severely threaten forest health and productivity. While most previous studies have focused on single-trait responses to drought, the interrelationships between stomatal regulation and hydraulic architecture remain unclear. In particular, substantial uncertainties persist regarding whether trees can physiologically adapt to drought through coordinated aboveground (i.e., stomatal regulation-hydraulic architecture) and belowground (i.e., root morphology) traits. With coniferous Cunninghamia lanceolata and broadleaved Phoebe zhennan seedlings from southern China plantations, we conducted a pot experiment to examine drought effects on photosynthetic physiology, hydraulic architecture, root plasticity, growth, and their coordination. Under drought, C. lanceolata reduced its stomatal conductance (gs) by 56.3–79.2% via early stomatal closure, exhibiting a conservative strategy. It maintained high Huber value, ensuring hydraulic safety, and adopted an acquisitive root strategy with high specific root length and fine roots (diameter < 0.05 mm, accounting for 55.6% of total root length), forming a conservative-aboveground–acquisitive-belowground strategy. In contrast, P. zhennan displayed an acquisitive stomatal strategy (gs decreased by 43.1%–77.8%) to sustain photosynthesis, supported by high hydraulic conductivity and high sapwood-specific conductivity; it developed a conservative root strategy with high root tissue density and high specific root area, realizing an acquisitive-aboveground–conservative-belowground strategy. These results reveal distinct multi-trait coordinated drought responses and inform the prediction of seedling growth stability under drought conditions.

Global climate change has led to frequent extreme drought events, introducing substantial uncertainty regarding how tree water source utilization strategies regulate forest growth. This study integrated tree-ring chronologies, stable isotope (delta 13C, delta 2H, delta 18O) analyses, and photosynthetic physiological data to examine the effects of water source plasticity on radial growth (Rg) in Cunninghamia lanceolata and Quercus acutissima from subtropical plantations. C. lanceolata exhibited densely distributed shallow fine roots but poorly developed deep roots, and relies predominantly on shallow soil water in both dry and wet seasons, with limited access to intermediate-depth water. This species displayed a shallow-dominated water use pattern with low plasticity, which failed to mitigate drought-induced physiological constraints, leading to a suppressed photosynthetic rate (Pn) and slower radial growth. In contrast, Q. acutissima possesses a well-developed deep coarse root system and moderately distributed shallow fine roots, enabling efficient uptake of shallow water in wet seasons and flexible utilization of stable deep water sources in dry seasons, thus showing a deep-shallow synergistic water use strategy with high plasticity. Under drought stress, Q. acutissima maintained a higher photosynthetic rate (Pn) and leaf water potential, thereby reducing carbon loss caused by frequent stomatal closure, and had significantly greater growth resistance and resilience than C. lanceolata. Our findings highlight that species-specific root architecture is closely associated with interspecific differences in water source plasticity, (as reflected by the shallow water availability elasticity index, SWEI), which is closely linked to drought resistance and growth stability. This integrated framework (combining tree-ring, stable isotope, and physiological data) offers an empirical foundation for forecasting the responses of subtropical forest tree species to ongoing climate change.

Soil stoichiometric characteristics serve as key indicators for assessing soil nutrient status and quality. Previous studies have predominantly focused on surface soil (0-40 cm), with limited understanding of deep soil (>40 cm) stoichiometric traits and their underlying drivers. In this study, we provide a case study of three typical restoration stands from China's Loess Plateau, which compared differences between surface and deep soil layers in stoichiometric traits and influencing factors. Soil samples were systematically collected at 20 cm intervals down to bedrock to analyze the reserves and stoichiometric differences in C, N, and P between surface and deep soil layers, and to identify relevant environmental influencing factors. The results showed that: (1) Within this Loess Plateau case study, deep soil accounted for 33%-47% of the total profile storage of C, N, and P, representing a critical nutrient reservoir. (2) The differences from China's average in C:N and C:P were markedly greater in deep soil (15.55 and 68.62, respectively) than in surface soil (11.63 and 8.31, respectively), indicating more pronounced nitrogen and phosphorus limitations in deep soil. (3) The factors influencing surface soil stoichiometry were mainly climate-related and biological interaction (altitude, soil water content and pH), while those for deep soil layers were factors related to nutrient storage and transport (soil thickness, soil bulk density and altitude). These results highlight that neglecting deep soil can lead to substantial underestimation of ecosystem nutrient reserves and misinterpretation of soil stoichiometry and its drivers. Therefore, we advocate incorporating deep soil into sampling designs in stoichiometric studies and attach more research attention to deep soil's stoichiometry and its role in biogeochemical cycling.

Background: Seasonal precipitation variability significantly affects water use in forests; however, whether water uptake is adapted to changes in precipitation, particularly whether it could affect the coexistence of tree species, has rarely been quantified in forest systems. Method: In this study, dual stable isotopes and the Li-6400 portable photosynthesis system were used to determine the water sources of a mixed conifer (Pinus massoniana) and broadleaf (Quercus acutissima) forest and changes in hydraulic characteristics during the dry and wet seasons in a southern hilly region of China. Results: Although the hydraulic characteristics of P. massoniana were lower than those of Q. acutissima, it maintained a stable water source from the deep soil layer and a higher stomatal conductance (Gs), leading to a higher transpiration rate (Tr) during the growing seasons. Q. acutissima mainly absorbed water from deeper soil layers in the dry season and took up from shallow soil layers in the wet season. Its Gs values exhibited sensitivity to precipitation, while it maintained a lower Tr value during the growing seasons. The excessive water-use strategy observed in P. massoniana may confer weak drought-tolerance during higher frequency and more intense extreme precipitation events, whereas Q. acutissima may exhibit better ecological adaption to precipitation changes. Conclusions: The overlap of water niches in mixed forests did not appear to affect the coexistence of tree species. The present study provides insights into reforestation and water management in the southern hilly regions of China.

Within the context of global change, the increasing frequency and co-occurrence of different types of droughts raise concerns regarding compound drought impacts. However, the relative influences of different drought types (meteorological, agricultural, and hydrological) on compound drought remains poorly understood. Focusing on the Yellow River Basin (YRB), this study integrates hydrological modeling, copula functions, and machine learning interpretability techniques to investigate drought propagation pathways, compound drought characteristics, and the varying influences of different drought types across subregions. The results indicate that drought propagation pathways differ spatially: meteorological drought propagates faster to hydrological drought in parts of the upper and middle reaches (III and IV) but propagates faster to agricultural drought in other regions (I, II, V, and VI). Spatially, compound drought was most severe in parts of the upper reaches (III), and mildest in the lower reaches (VI). Across the YRB and its upper and middle reaches (I-V), meteorological drought generally had the strongest influence on compound drought during the identified compound drought months. However, hydrological drought was the most influential factor in the lower reaches. Agricultural drought typically ranked second (I-IV) or third (V and VI) in terms of influence. Furthermore, the analysis revealed the nonlinear influence of precipitation minus potential evapotranspiration (PPET), soil water content (SW), and water yield (WYLD) on compound drought and their regionally variable thresholds (PPET: -10 to -59 mm; SW: 4 to 50 mm; WYLD: 0.2 to 4.9 mm). For the YRB, compound drought tended to be alleviated when PPET > -35.18 mm, SW > 38.62 mm, and WYLD > 2.76 mm. These findings highlight the spatial heterogeneity of compound drought drivers and provide quantitative, threshold-based insights that are crucial for developing targeted drought management strategies. The integrated framework employed is scalable and provides valuable guidance for a compound drought analysis in other regions.

Long-term pure forest (PF) management and successive planting has result resulted in "low-efficiency artificial forests" in large areas. However, controversy persists over the promoting effect of introduction of broadleaf tree species on production efficiency of PF. This study hypothesised that introduced broadleaf tree species can significantly promote both water-nutrient use efficiency and gross primary productivity (GPP)of PF. Tree ring chronologies, water source, water use efficiency and GPP were analysed in coniferous Cunninghamia lanceolata and broadleaved Phoebe zhennan growing over the past three decades. The introduction of P. zhennan into C. lanceolata plantations resulted in inter-specific competition for water, probably because of the similarity of the main water source of these two tree species. However, C. lanceolata absorbed more water with a higher nutrient level from the 40-60-cm soil layer in mixed forests (MF). Although the co-existing tree species limited the basal area increment and growth rates of C. lanceolata in MF plots, the acquisition of dissolved nutrients from the fertile topsoil layer were enhanced; this increased the water use efficiency and GPP of MF plots. To achieve better ecological benefits and GPP, MFs should be constructed in southern China.

The spatiotemporal patterns of water travelling through the soil-plant-atmosphere continuum (SPAC) have received much attention due to the frequency of extreme weather and water scarcity. However, the interconnections and transboundary transport of specific compartments within the hydrologic cycle are poorly understood. We portrayed the propagation paths of water isotope signals by analyzing isotopes of precipitation, soil water, and xylem water. The isotope sine curves analysis method was improved to quantify water transfer efficiency, mean transmission time from precipitation to soil (MTT1), mean transmission time from soil to trees (MTT2), mean residence time within soil (MRTSW), and mean residence time within tree xylems (MRTXY). The temporal trajectory of water traveling through SPAC was depicted, and its influencing factors and intrinsic mechanisms were explored. Our results showed that (1) water isotopes were blocked when crossing water pools (precipitation, soil water, and xylem water) and the transfer efficiency was progressively weaker (85.5 %-> 13.5 %). (2) The MTT1 was significantly and positively correlated with soil porosity (Spearman correlation coefficient, 0.551). Its spatiotemporal pattern (0.8-15.9d) indicated that the transport and mixing of soil water involved two modes: displacement flow and bypass flow. The MTT2 (-13.2d-4.3d) of Platycadus orientalis was in general smaller than that of Quercus variabilis (-4.7d-11.5d). (3) The MRT (35.2d-343.8d) was influenced by a combination of soil physicochemical properties, root distribution, and tree morphological traits. Our study shows that connectivity between pools is better reflected with transfer efficiency. Comparing MTT and MRT of P. orientalis and Q. variabilis, tree hydrodynamic processes were quantified, and P. orientalis was more sensitive to seasonal moisture variability. Additionally, our study improved the understanding of the "black box" within the hydrological interface of plant-soil interactions.

Studying the factors influencing ecosystem regulation services in southwestern Zhejiang is of great significance for formulating reasonable pricing strategies for forest ecosystem regulation services and optimizing ecological security. This study constructed a theoretical framework for analyzing forest ecosystem regulation services and assessed the spatiotemporal evolution and influencing factors of forest ecosystem regulation services using InVEST model calculations and spatial autocorrelation analysis. The results showed that all ecosystem services of forests in the study improved from 2000 to 2019, with the exception of soil conservation. The water conservation function increased significantly from 2000 to 2019, with an overall increase of 3.53%. The biodiversity conservation function in 2019 also increased significantly, with an average increase of 2.16% compared with 2000. The synergies mainly occurred between water source regulation and soil conservation, soil conservation and biodiversity, and forest recreation and carbon storage. Forest Reserve was precipitation, canopy closure, elevation, and soil texture, and their driving forces differed at different time scales. The trade-offs mainly occurred between soil conservation and forest recreation, forest recreation and biodiversity, and carbon storage and biodiversity. The research results provide a reference for achieving ecological protection and high-quality development in the southwestern region of Zhejiang.

Although optimizing nitrogen management (Nitrogen fertilizer reduction, NFR; organic fertilizer substitution, OFS) is the feasible solution for reducing pollutant emissions (NH3 volatilization, greenhouse gas emissions, and N leaching and runoff) in rice cultivation, the different revenue and cost results in an indefinite net ecosystem economic benefit (NEEB) is the major hindrance when promoting these nitrogen strategies. Investigating and assessing the NEEB could exhibit the suitability of different nitrogen management. In a two-year field experiment, the conventional nitrogen management (N180), NFR (25% nitrogen reduction based on N180), and OFS (25% N substituted with organic fertilizer based on N180) were compared under continuous flooding irrigation (CF) and alternating wet and dry irrigation (AWD). Both NFR and OFS showed significant reductions in NH3 volatilization (25.0% and 18.9%), GHG emissions (35.4% and 28.6%), and total N leaching and runoff (39.0% and 28.1%), compared to N180. However, NFR resulted in a 19.0% reduction in rice yield and a lower NEEB (15.5 x 103 CNY ha-1) than N180 (27.4% lower). Similarly, OFS produced an identical NEEB (15.6 x 103 CNY ha-1) as NFR due to the high cost of organic fertilizer in rice production. Furthermore, AWD irrigation showed better environmental and economic benefits than CF, reducing CH4 emissions and conserving irrigation water and electricity. While NFR and OFS reduced the NEEB in rice cultivation, the recommendation is coupling NFR with other field practices to improve yield revenue, reducing organic fertilizer production costs and maintaining long-term application, both of which will potentially enhance the NEEB when applying these optimizing nitrogen managements.

The process of plant water use is complex and changeable, which is affected by various factors. Exploring the sources and influencing factors of plant water use can provide reference for clarifying the mechanisms of forest water adaptation under climate change. We chosen the typical forest communities in the hilly region of Sou-thern China, Pinus massoniana and Quercus acutissima mixed forest as the research object. By analyzing water sources of plants in different seasons, the factors affecting the changes of water sources were explored in combination with soil water, precipitation, and plant roots. The results showed that water use characteristics of P. massoniana and Q. acutissima were similar and both mainly utilized 0-40 cm soil water during the dry season, with proportions of 60.0% and 66.6%. During the rainy season, as soil water content of deep layers increased, the main water sources of both gradually shifted towards deep soil. The similarity proportion indices of P. massoniana and Q. acutissima were above 60%, indicating that there was an obvious water competition between them. Root system of Q. acutissima had plasticity in water absorption, and played a dominant role in absorbing shallow water during the dry season. Water was the main driving factor for water source transformation of Q. acutissima and P. massoniana during the rainy season. Compared with P. massoniana, Q. acutissima was more sensitive to the changes of water sources. Under the background of future warming and drying, the competition between the two species for shallow water sources might be intensified. Those two species should be sparsely planted or thinned to optimize forest structure to cope with water stress.

Chinese fir in China are generally inefficient plantations with single species, unreasonable stand density, and low productivity. The introduction of broadleaved species is usually adopted as a strategy to improve Chinese fir plantations. Taking the pure forests and mixed forests of the Guanshan Forest Farm in Jiangxi Province as example, we quantified the intrinsic water-use efficiency (iWUE) of trees based on the stable isotope carbon method, as well as its response to meteorological factors, and investigated the improvement of stand quality after introducing Phoebe zhennan into Chinese fir plantation. The results showed that the basal area increment was 0.23 cm2 in pure forest, being higher than that of 0.19 cm2 in mixed forest. The δ13C and iWUE of pure forest were -27.4‰ and 52.9%, respectively, being lower than those of -26.7‰ and 62.8% in the mixed forest. Tree δ13C in pure forest was more sensitive to changes in mean annual precipitation and mean annual relative humidity, while that in mixed forest was not significantly correlated with meteorological factors. Pure forest iWUE was positively correlated with mean annual temperature, mean annual atmospheric CO2 concentration, and mean annual maximum temperature, and negatively correlated with mean annual precipitation and mean annual relative humidity, while mixed forest iWUE was positively correlated with mean annual atmospheric CO2 concentration only. Our results indicated that pure forests was more sensitive to climate than mixed forests.

The intensity and frequency of droughts are projected to rise in recent years and adversely affect forests. Thus, information on plant water use and acclimation during and after droughts is crucial. This study used the stable isotope and thermal dissipation probes to detect the water-use adaptation of mixed forests to drought using a precipitation gradient control experiment in the field. The results showed that Platycladus orientalis and Quercus variabilis mainly absorbed stable water from deep soil layers during the drought (32.05% and 28.2%, respectively). The synergetic nocturnal sap flow in both species replenished the water loss, but P. orientalis experienced a greater decline in transpiration acclimation to drought. The transpiration of Q. variabilis remained high since it was mainly induced by radiation. After short-term exposure to drought, P. orientalis majorly obtained shallow soil water, confirming its sensitivity to shallow water. Contrarily, Q. variabilis mainly absorbed stable water from deep soil layers regardless of the soil water content. Therefore, these findings suggest that Q. variabilis cannot physiologically adjust to extreme drought events, possibly limiting their future distributions and altering the composition of boreal forests.

The interspecific and temporal dynamics of tree water use are poorly understood. We investigated in hightemporal resolution patterns of water use of two tree species, Platycladus orientalis and Quercus variabilis, in a temperate mountainous monsoon forest in northern China. We leverage a unique sampling design where we systematically traced tree water source of ten individuals continuously throughout a three-year period (2015-2017) using stable isotopes of hydrogen and oxygen, coupled with root distribution assessment and measurements of transpiration rates. We observed species-specific patterns in tree water use. Q. variabilis withdrew water evenly from different soil layers without detectable seasonal changes in pattern in water use throughout the different years. By contrast, P. orientalis increases shallow source water uptake during the rainy season in all three years. Isotopic composition of both species plotted along an evaporation line below the local meteoric water line (LMWL). However, during sporadic periods in the rainy and transition season, P. orientalis plotted on the LMWL and showed similar isotopic to spring water. This suggests an ephemeral ecohydrological connectivity between tree water source and spring water during short periods within the wet and transition season. The observed distinct patterns in tree water use between the two species may be associated with root distribution characteristics. Besides root morphological traits, our data also suggest that the influence of overall stem volume (Vob) on water source contribution depends on the rainfall and species. Our results support the seasonality of ecohydrological separation observed in previous studies. Additionally, our data show that that ecohydrological separation may be species dependent in a temperate mountainous monsoon forest.