The Sanjiangyuan region, located in the hinterland of the Qinghai-Tibet Plateau(QTP), is highly sensitive to global climate change. Reconstructing its Holocene January and July temperatures is crucial for studying climate change and guiding ecological conservation in alpine regions. Current research on paleoclimate changes in Sanjiangyuan region primarily focuses on small subregions, limiting holistic understanding of regional climate.This study utilizes fossil pollen data, for the first time, integrates the Dynamic Multi-proxy Fusion and Scaling(DMFS) model to reconstruct the Holocene January and July temperature change sequences, thereby exploring temperature variations in the Sanjiangyuan region during 12.5 ka BP. The results indicated:12.5-6.0 ka BP: Both January and July temperatures showed a gradual increase,marking climatic improvement. 6.0-4.0 ka BP, both January and July temperatures remained at high levels, despite their fluctuations. During this period, temperatures reached their peak, reflecting a warm, humid, and most hospitable climate.4.0-2.5 ka BP: Both January and July temperatures showed declining trends to varying degrees, the climate became cold and dry. Post 2.5 ka BP: Both January and July temperatures rebounded. Comparisons with other high-resolution environmental records from the QTP confirmed consistent trends and synchronic dry-wet events. This study contributed essential fossil pollen data and paleotemperature records to the Sanjiangyuan region. This will fill a critical gap in paleoclimate research for the Sanjiangyuan region and provide valuable insights for long-term paleoclimate studies.

Pollen ratios have been widely employed to infer vegetation and climate in arid and semi-arid regions globally. However, the reliability of pollen ratios in differentiating vegetation types and indicating climate characteristics has not been validated using vegetation data, as the relationship between modern pollen and vegetation is not directly equivalent. In this study, we examined six commonly used pollen ratios-Artemisia/Chenopodiaceae (A/ C), Artemisia/Cyperaceae (A/Cy), (Artemisia + Chenopodiaceae)/Poaceae ((A + C)/Po), Cyperaceae/Poaceae (Cy/Po), Poaceae/Asteraceae (Po/As), and the sum of relative abundance of xerophilous taxa (Chenopodiaceae, Ephedra, Nitraria, Tamaricaceae, and Zygophyllaceae, Sumxero)-using 1385 modern pollen samples and 2855 vegetation plot records from northern and western China. Vegetation plot data, in addition to modern pollen data, were used to validate the pollen ratios as well. The relationships between pollen ratios, vegetation, and climate were further analyzed numerically to assess the reliability of pollen ratios in distinguishing six vegetation types and in reflecting climatic characteristics. The results demonstrated that pollen-based and vegetation-based pollen ratios exhibited varying degrees of effectiveness in distinguishing vegetation types and reflecting precipitation condition. Pollen ratios derived from pollen assemblage and vegetation composition exhibited inconsistency in distinguishing most vegetation types and in representing precipitation. Certain pollen ratios demonstrated a stronger ability to distinguish vegetation types than to reflect precipitation. The pollen A/C ratio and Sumxero can be effectively used to differentiate temperate steppe from desert, as well as alpine desert from other alpine vegetation types, respectively. Most pollen ratios are not suitable for precipitation indication; only the A/C, A/Cy, (A + C)/Po and Sumxero exhibit limited sensitivity to precipitation within specific climatic ranges for the entire study region. However, when separating vegetation types, the pollen ratios and precipitation relationships are much weaker and even ineffective to some extent, than those for the whole region. The application of pollen ratios in distinguishing vegetation types and reflecting climatic characteristics should therefore be approached with caution and must take vegetation composition and the climatic context into account.

Evergreen and deciduous broadleaved mixed forests (EDMF), which exhibit high sensitivity to climate fluctuations, are important transitional vegetation types in subtropical regions and possess the potential for palaeoenvironmental reconstruction. Nevertheless, at present, there is a dearth of relevant research, resulting in uncertainties regarding the internal succession characteristics and external driving patterns of mountain broadleaved forests. In this study, we reconstructed pollen-based biome changes since 5.8 ka with a resolution of 18 years from a mountain wetland in southwestern (SW) China. Based on the comparison between modern pollen and vegetation quadrats, it is inferred that since the mid-Holocene, the study area has been predominantly characterized by EDMF, and there have been significant changes in the internal taxa composition. From 5.8 to 3.4 ka, 2.3 to 1.6 ka, and 0.6 to 0 ka, the vegetation was classified as the deciduous (D)-type EDMF, mainly composed of deciduous taxa such as Quercus (D), Fagus, and Corylus. Conversely, from 3.4 to 2.3 ka and 1.6 to 0.6 ka, the proportion of evergreen taxa like Castanopsis, Cyclobalanopsis, and Ilex was relatively high, leading to a vegetational shift to the evergreen (E)-type EDMF. Through comparison with the TraCE simulations, it is demonstrated that the summer temperature and precipitation may jointly affect the internal transformation between D-type and E-type EDMF. On a larger spatial scale, it is inferred that this transformation may exhibit a teleconnection with ENSO frequency. Specifically, variations in ENSO frequency may exert selective effects on the internal taxa changes of subtropical EDMF.

Abstract The climatic transition from Marine Isotope Stage 3 (MIS3) to the Last Glacial Maximum (LGM) had caused widespread vegetation change. Despite the dynamic equilibrium between vegetation and climate, the specific role of functional composition in vegetation’s response to climate change was inadequately understood. Here, we analyzed the long-term trajectories of palynological diversity, vegetation coverage and community-weighted-mean (CWM) functional traits based on EH22 pollen record (35–18 cal ka BP) from Erhai Lake, southwestern China. The results disclosed a vegetation transition from temperate deciduous broadleaf forest dominance in late MIS3 to cold coniferous and mixed broadleaved/coniferous forests in LGM. This vegetation dynamic involved functional composition shifts from competitive-driven functional convergence to partial recovery via niche differentiation during the late MIS3, and finally to a low-diversity but functional differentiation state through trait complementarity and diversification strategies during the cold LGM. Our results likely support a function-mediated climate filtering process whereby climate change regulated long-term vegetation dynamics during the MIS3–LGM transition primarily through shifts in CWM functional composition. These findings underscore the potential of pollen-based trait approaches to reconstruct ecosystem properties and advance our understanding of ecosystem change over decadal to millennial time-scales.

The Hengduan Mountains (HDM) in southwestern (SW) China, characterized by complex terrain and diverse vegetation types, offer an ideal environment for the reconstruction of paleovegetation and palaeoclimate. Nevertheless, due to the absence of detailed local vegetation investigations corresponding to pollen samples, the comprehension of certain pollen-based paleovegetation results within the region still remains controversy. In this study, we conducted a thorough investigation of 66 vegetation quadrats (700-4700 m a.s.l.) in the HDM, and performed corresponding pollen-based biomes. Results show that 1. pollen-based biomes match local vegetation well in alpine shrublands and meadows (ALSM) and evergreen sclerophyll Quercus forests (ESQF). However, more than half of the cool conifers mixed forest (COMX) was reconstructed as deciduous broadleaved forest (DBLF), because the high proportion of exotic Alnus pollen. Owing to the high proportion of Poaceae resulting from human activities, a small number of evergreen broadleaved forest (EBLF) have been reconstructed into ALSM; 2. The six pollen-based biomes are clearly fewer than the ten investigative vegetation types. This disparity is probably due to the high pollen productivity of high-altitude conifers and the incapability of collecting entomophilous pollen from lowland tropical taxa. Redundancy analysis reaffirms that annual/seasonal climate has a clearly impact on most vegetation types. Nevertheless, DBLF might be more reliant on climatic stability than on the climatic threshold. This study can provide novel data to further clarify the relationship among local vegetation, pollen-based biome, and climate in HDM. Based on this data, the reliability of paleovegetation reconstructions within the region can be improved.

Abstract The accelerating global climate change has been triggering large-scale vegetation reorganizations, yet our understanding of how mountain ecosystems respond to rapid climatic oscillations is critically constrained. Here, we present a high-resolution palynological record from Erhai Lake, southwestern China, revealing eight episodes of rapid vegetation reorganizations in the Hengduan Mountains (HMs) over the past 35,200 years. These reorganizations closely tracked the rhythms of rapid climate oscillations, particularly during the Last Glacial Maximum and the Last Deglacial Period. We find that while the timing of rapid vegetation reorganizations were synchronous with Atlantic Meridional Overturning Circulation (AMOC) anomalies that modulated global climate variability, the magnitude of these reorganizations did not exhibit a linear correlation with AMOC strength; instead, they were governed by local heat and moisture availability mediated through teleconnections. This demonstrates a strong natural regulatory capacity of mountain vegetation in HMs, enabling resilience to intense climatic fluctuations. However, when using the Erhai record as a benchmark, we project that rapid reorganizations under the high-emission pathway (SSP585) will likely surpass the intensities observed during historical events. These findings reveal the high climatic sensitivity and strong natural regulatory capacity of mountain ecosystems, highlighting the critical necessity of climate mitigation actions and nature-based solutions to safeguard subalpine and alpine biodiversity against unprecedented future climate change.

The Hengduan Mountains (HDM) constitute one of the world's richest biodiversity regions and are designated as a top-tier priority for ecological conservation. Vegetation investigation can help with the design and implementation of biodiversity conservation in this region. Here we present the HDM-Plot, a plot-based vegetation dataset compiled from 314 plots surveyed during four campaigns between 2022 and 2024 across the Hengduan Mountains and adjacent regions, across major vegetation types from lowland dry-hot valleys to alpine areas spanning altitudes of 754-4932 m. Each plot records detailed species-level information, including scientific name, growth form, life form, number of individuals or clumps, plant height, diameter at breast height or at base, crown width, and coverage, along with geographic coordinates and hierarchical vegetation classification. In total, the dataset comprises 14 113 individual records belonging to 1127 species from 379 genera and 117 families. The dominant families are Rosaceae (133 species), Ericaceae (93), Fabaceae (66), Asteraceae (63), and Fagaceae (37), and the dominant genera are Rhododendron (75), Berberis (34), Cotoneaster (30), Salix (24), and Quercus (22), with composition varying among vegetation types. Growth forms are mainly composed of shrubs (46.0 %), trees (27.3 %), and herbs (23.6 %). Herbs are dominated by perennial (92.1 %), shrubs are mainly deciduous broadleaf (59.7 %), and trees are primarily deciduous broadleaf (46.8 %) and evergreen broadleaf (41.6 %). Species richness, growth forms, and life forms show clear elevational changes within the HDM-Plot dataset. Floristically, genus-level areal-types in the HDM-Plot dataset are dominated by temperate elements (54.1 %), followed by tropical elements (35.4 %). 314 plots can be assigned to three vegetation formation groups, 18 vegetation formations, 142 alliance groups, 209 alliances, 238 association groups, and 299 associations. The HDM-Plot dataset provides an updated and standardized baseline for quantitative analyses of mountain vegetation, biodiversity assessment, and vegetation classification and mapping in southwestern China. Such information can be future used in the revisions of China's vegetation classification scheme and Vegegraphy of China. The dataset is available from Figshare (Jin et al., 2026a; 10.6084/m9.figshare.32706207) and through the National Tibetan Plateau/Third Pole Environment Data Center (Jin et al., 2026b; 10.11888/Terre.tpdc.303394).

Plant functional trait databases are indispensable tools in modern ecology and have been widely employed in research on plant life history strategies, community species composition and structural dynamics, and ecosystem responses to environmental change. Karst regions are characterized by complex terrain, high habitat heterogeneity, and unique vegetation types, and each of these environmental factors independently and interactively produce distinctive communities and traits along the landscape. Currently, most investigations into plant functional traits within karst regions have focused exclusively on leaf traits of a small number of dominant species within communities. This narrow focus has resulted in a significant gap in the available data regarding plant functional traits in karst environments. This study presents data on ten morphological traits from 3,661 individuals of 152 plant species across 90 genera and 65 families. All individuals had a diameter at breast height of ≥1 cm and were collected from twelve plots in the Maolan National Nature Reserve, southwestern China. Utilizing this database, we analyzed trait correlations and both interspecific and intraspecific variations. Most traits exhibited significant correlations with one another (P < 0.01). Trait variability differed markedly, with twig traits showing lower variation than leaf and bark traits. Intraspecific, interindividual variation described a significant proportion of variation, from 39.42% to 52.49%. This database consolidates extensive plant functional trait data from karst regions, supporting future research on plant adaptive strategies in these unique habitats and improving our understanding of local community assembly and maintenance.

Grain size is widely used for environmental reconstruction in various depositional settings, yet its precise interpretation remains challenging. In shallow lakes within humid region, interpretation of grain size is particularly hindered by complex variations in climate and environments. Here, we conducted grain size and End Member Modeling Analysis (EMMA) on a 4.56 m sediment core from Yilong Lake, southwestern China, to clarify the response of grain size composition to different climatic and environmental factors. Four end members (EMs) were fitted from grain-size distributions. EM1 comprises mainly clay and fine silt, and EM2 shows a bimodal structure concentrated in fine-silt and sand fractions. Both EM1 and EM2 likely formed under shallow-lake conditions, indicating low-energy deposition and episodic flood deposits, respectively. EM3 is composed of well-sorted silt, and EM4 comprises relatively coarse particles. EM3 and EM4 reflect high-hydrodynamic regimes in humid climates, corresponding to deep-lake deposition under abundant precipitation and to high-energy inputs generated by strong hydrodynamic disturbances. Different EMs exhibit diverse variabilities, showing complex responses to multiple factors of precipitation, water level, and vegetation coverage since the Last Glacial Maximum (LGM). Notably, grain-size distributions reverted to LGM-like features, with increased proportions of EM1 and EM3, under intensified human activities during the late Holocene. Our findings show that climate, vegetation, lake state, and human activities jointly impacted depositional processes in Yilong Lake over the past 27,000 years. Thus, in humid-region shallow lakes, precise interpretation of grain size distributions requires robust priori knowledge and careful consideration of climate and environmental trajectories.

Abstract. Alpine treelines are expected to respond to climate warming, but the magnitude and direction of treeline shifts often vary across mountain landscapes. In the Hengduan Mountains, river corridors may shape how forest expansion is expressed as lateral and elevational treeline shifts, with important implications for future alpine habitat loss. Here, we modified and applied the spatially explicit individual based model LAVESI to simulate treeline dynamics along four major river corridors: the Dadu, Lancang, Nu, and Yalong rivers. Simulations covered the historical period from 1940 to 2020 and the future period from 2020 to 2100 under SSP1-2.6, SSP2-4.5, and SSP5-8.5. During the historical period, simulated treeline shifts were gradual but river specific. Lateral changes were strongest in the Lancang and Nu rivers, intermediate in the Yalong River, and weak in the Dadu River. Elevational changes were more limited, with the Lancang River showing the clearest upward shift and the Nu River remaining close to stable. After 2020, lateral treeline shifts became stronger and more divergent among river corridors and scenarios. In most rivers, future lateral advance increased, with SSP2-4.5 generally producing relatively high cumulative expansion. Elevational shifts followed a related but not identical pattern: the Lancang River maintained the strongest upward trend, the Dadu and Yalong rivers showed moderate increases, and the Nu River remained weakly responsive. The relationship between lateral and elevational changes therefore varied among rivers, indicating that horizontal boundary reorganization did not always translate into comparable upslope shift. Treeline invasion potential was also uneven. The Dadu River frequently reached the predefined upper limit, although this result should be interpreted in relation to the shorter simulation extent. In contrast, most simulations for the Lancang, Nu, and Yalong rivers remained below the upper limit by 2100, suggesting incomplete occupation of the available treeline tundra ecotone. Overall, our results indicate that future treeline shifts in the Hengduan Mountains are likely to remain spatially heterogeneous across river corridors, with different implications for alpine habitat vulnerability. This vulnerability should therefore be evaluated at the river corridor scale, where lateral expansion, elevational advance, and local topographic and ecological settings jointly shape the potential for future forest expansion.

Climate warming significantly exacerbates hydrological droughts in the tropics, threatening shallow lake ecosystems through increased vulnerability to drying and degradation. However, the ecological impacts of rising temperature on the network stability of these ecosystems remain poorly understood. Here, we analyze a high-resolution (∼26 yr/cm) diatom record from Manxing Lake, a tropical shallow lake in southwestern China, to investigate temporal changes in co-occurrence network stability over the past two millennia and explore the underlying mechanisms linked to community assembly and network topology. Our findings reveal that during the Medieval Warm Period (1097–1520 CE), hydrological droughts were prevalent in the study area, resulting in network configurations with high vulnerability and moderate robustness. A comparison of the goodness of fit (R2) of neutral community models with network vulnerability suggests that stochastic processes (high R2) generally produce network structures with high modularity that is sensitive to the removal of keystone species. The highest robustness and the lowest vulnerability were observed during the Little Ice Age (1540–1750 CE), when the network exhibited a high average degree and low degree skewness. This supports the hypothesis that increased network complexity enhances stability through structural redundancy that preserves overall system functionality. During the Current Warm Period (1773–2024 CE), the network was reconfigured into a structure with relatively low average degree and high degree skewness, ultimately leading to a noticeable decline in network robustness and a slight increase in vulnerability. These results underscore that future warming-induced droughts, coupled with intensified human impacts, may drive substantial ecological reorganization and destabilization in tropical shallow lakes.

Global change poses significant threats to the survival and ecological functions of numerous species, particularly endangered species with small populations. Beshanzu fir Abies beshanzuensis , an ancient and Critically Endangered species with only 3 mature individuals remaining in Baishanzu National Park, eastern China, urgently requires improved conservation strategies under rapid climatic change. In this study, an ensemble species distribution model was applied to predict the current and future potential distributions of the Beshanzu fir to evaluate climatic impacts and identify potential refugia. The results indicate that precipitation and temperature, especially their annual means, primarily determine the species’ distribution. Under future scenarios, highly suitable habitats are projected to shift upward in elevation, as well as westward and southward, suggesting notable redistribution trends. A gap analysis identified several climate refugia, encompassing >8000 km 2 , 72% of which lie outside but near the existing conservation network. Based on these findings, 5 priority areas are proposed for adjusting conservation planning: (1) between Zhejiang and Fujian provinces, (2) between Jiangxi and Fujian provinces, (3) between Suichang and Longquan Counties in Zhejiang Province, (4) along the border of Jiangxi and Hunan provinces, and (5) northern Taiwan Island. These insights provide a refined understanding of habitat shifts in the Beshanzu fir and offer valuable guidance for its ex situ conservation, as well as for the management of other critically endangered plant species with small populations worldwide.

The nitrogen (N) vs. phosphorus (P) scaling exponent, indicating the allocation strategies of the two nutrients, plays an important role in predicting plant and ecosystem functioning. Besides the constant scaling relationship of N vs. P, plasticity of scaling exponents has been demonstrated, with proposed potential causes including plant functional groups (PFGs) or environmental factors. However, relatively little is known about whether the N vs. P scaling exponent of different PFGs response similarly to different environmental conditions. We measured N and P concentrations in leaves and twigs of 964 shrub individuals in a secondary subtropical forest in Eastern China and explored how the N vs. P scaling exponent vary with different light or soil nutrient conditions for deciduous and evergreen shrubs. The leaf N vs. P scaling exponent of evergreen but not deciduous shrubs was affected by light condition, with higher scaling exponent of understory shrubs than that of typical shrubs. The twig scaling exponent of understory shrubs was lower than that of typical shrubs for both PFGs. With increasing soil N availability, responses of scaling exponents in both organs for both PFGs consistent with compliant type first and resistant type later. The scaling exponents of deciduous and evergreen shrubs in both organs showed opposite trends at a relatively high soil P availability level. Our results reveal that environmental factors may alter the N vs. P scaling exponent differently across PFGs. These findings have important implications for improving predictions of stoichiometric models and broadening our understanding of plant responses to varying environments. This study investigates the N vs. P scaling exponents of leaves and twigs in deciduous and evergreen shrubs under different light or soil nutrient conditions in a secondary subtropical forest, revealing that environmental factors may alter the scaling exponents differently across plant functional groups. The study provides new insights for improving stoichiometric model predictions and deepening our understanding of plant responses to environmental variability. (sic)(sic):(sic)-(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).(sic)(sic)(sic)(sic), (sic)-(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)-(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic).(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)964(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic),(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)-(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).(sic)(sic)(sic)(sic):1)(sic)(sic)(sic)(sic)(sic)(sic)(sic)-(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).2)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).3)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)-(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic), (sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic)(sic).

As one of three recycling pathways in terrestrial ecosystem, wildfires have long been attributed to climate changes and play an important role in regulating nutrient cycling and terrestrial vegetation composition. Southwestern China is a hotspot of plant diversity and experiences severe wildfires, however, the long-term fire trajectory, variability, and forcings behind wildfire activity are less understood. Here, we use sedimentary charcoals to reconstruct fire regimes in the Erhai Lake region of southwestern China over the past 35,200 years. Our findings illustrate that periods with increased wildfire activity were generally linked to dry and/or cool climate, with clearly increased frequency and severity of fire occurrence during late MIS3, Heinrich-1, and the Younger Dryas. In contrast, levels of biomass burning decreased obviously during the B & oslash;lling-Aller & oslash;d warm period and throughout the Holocene. Long-term trends of wildfire activity were generally consistent with paleofire records in southwestern China. Wildfire frequency and severity in this region should be regulated by relative humidity in the dry season, which is modulated by spring solar insolation. Synchronized variations were observed between Erhai charcoal record and Asian stalagmites delta 18O, reinforcing a suggested impact of the Asian Summer Monsoon on wildfire regimes at multi-centennial scales. In addition, our results support a top-down pattern in which climate-driven wildfires facilitated at different times the development of pine forest and/or grassland around Erhai Lake. These findings highlight the critical role of long-term hydro-climate variability in regulating fire regimes as well as the vegetation dynamics in southwestern China.

Droughts have intensified under climate change, threatening ecosystem stability. While rising atmospheric CO2 concentrations may enhance vegetation drought resistance, the net effect remains uncertain amid concurrent warming. Here we combine ecological modeling with multi-source observations to investigate how CO2 and warming jointly regulate vegetation drought responses on the Qinghai-Tibetan Plateau, a sensitive alpine region exposed to escalating drought threats under changing precipitation regimes. Using factorial scenarios to isolate individual forcings, we show that 40-year CO2 rise mitigated drought-induced productivity losses by 5.7 +/- 0.9% under constant temperature. However, in the presence of warming, rising CO2 intensifies drought stress by 5.2 +/- 0.5%, reflecting increased plant water demand and disrupted regional water supply-demand balance. Permafrost areas experienced the strongest CO2-driven drought alleviation under constant temperature, but also the greatest warming-induced reversal. These findings reveal interacting CO2-warming impacts on alpine vegetation drought responses, highlighting ecological risks for the plateau and other permafrost-dominant regions under future warming.

Evergreen and deciduous broadleaved mixed forests (EDMF), which exhibit high sensitivity to climate fluctuations, are important transitional vegetation types in subtropical regions and possess the potential for palaeoenvironmental reconstruction. Nevertheless, at present, there is a dearth of relevant research, resulting in uncertainties regarding the internal succession characteristics and external driving patterns of mountain broadleaved forests. In this study, we reconstructed the local biome changes over the past 5.8 ka based on a pollen record with an approximately 18–year resolution from a mountain wetland in southwestern (SW) China. The results suggest that since the mid–Holocene, the study area has been predominantly characterized by EDMF, and there have been distinct changes in the internal plant species composition. From 5.8 to 3.4 ka, 2.3 to 1.6 ka, and 0.6 to 0 ka, the vegetation was classified as the deciduous (D)–type EDMF, mainly composed of deciduous taxa such as Quercus (D), Fagus, and Corylus. Conversely, from 3.4 to 2.3 ka and 1.6 to 0.6 ka, the proportion of evergreen taxa like Castanopsis, Cyclobalanopsis, and Ilex was relatively high, leading to a vegetational shift to the evergreen (E)–type EDMF. Therefore, it is postulated that the expansion range of low–altitude evergreen broadleaved forests (EBLF) during this period might have been overestimated previously. Moreover, by comparing the results of pollen analysis with the TraCE simulations, it is shown that the thresholds of annual/summer temperature and precipitation may jointly influence the internal transformation between D–type and E–type EDMF. On a larger spatial scale, it is inferred that this transformation may exhibit a teleconnection with El Niño-Southern Oscillation (ENSO) variability. Specifically, different ENSO states may exert selective effects on the internal dynamic changes of subtropical transitional EDMF.

The influence of human activities on vegetation has been extensively documented in sedimentary pollen records, highlighting the importance of understanding the relationship between modern human activities and land use when reconstructing historical vegetation changes using pollen data. How can modern surface pollen from human disturbed vegetation indicate potential vegetation or land use type, however, still need to pay more attention. This research analyzes pollen assemblages derived from 117 surface soil samples collected in Southwest China. It elucidates the characteristics of these pollen assemblages and their source ranges across various land use types, as well as the indicative significance of predominant pollen types in relation to vegetation composition. The findings reveal that regional pollen assemblages are predominantly comprised of Pinus, Tsuga, deciduous Quercus, evergreen Quercus, and Betula, which collectively dominate the assemblages. Notably, there are marked differences in the characteristics of pollen assemblages associated with distinct land use types. Specifically, cultivated lands and plantation forests are mainly dominated by artificially selected plants, while secondary forests are mainly dominated by naturally distributed vegetation. Within a radius of 0-200 m from the sampling locations, significant variability in vegetation composition is observed among different sites, reflecting local vegetation characteristics. Conversely, at distances ranging from 0.2 to 20 km, as the spatial separation between sampling points increases, the disparities in vegetation composition and the proportional representation of various land use types diminish, leading to a more homogeneous landscape. At this spatial scale, the correlation between pollen assemblages and vegetation is significantly strengthened, with vegetation accounting for over 90 % of the pollen composition. Thus, the pollen dominantly reflects the regional vegetation characteristics. These results underscore the importance of thoroughly considering the relationship between the contribution rate of pollen assemblages and the composition of regional vegetation when reconstructing regional vegetation patterns through pollen analysis.