Climate change threatens global food security, yet the long-term effects of climate variability on agriculture remain poorly understood in high-altitude regions such as the Tibetan Plateau, where millions of people depend on pastoralism and small-scale agriculture. Here we show how mid- to late Holocene temperature variability shaped agricultural transformation across the Tibetan Plateau. By integrating newly generated and published proxy records and employing Monte Carlo simulations to account for uncertainties, we reconstruct regional temperature variability. Our results show that mid- to late Holocene cooling drives a shift from thermophilic millet-based cultivation to a cold-resilient wheat- and barley-dominated system. We further find that the rate of temperature change exerts a primary control on agricultural stability and human responses. Episodes of heightened climate variability correspond with declines in crop productivity, whereas more stable conditions are associated with agricultural expansion and sociopolitical florescence. These findings imply that the rate of environmental change is an important determinant of resilience in highland agroecosystems. Mid–late Holocene cooling on the Tibetan Plateau shifted farming from millet to wheat–barley, and temperature changes affected agricultural productivity and society, according to an analysis that integrates temperature proxies and uses Monte Carlo simulations.

Knowledge of the long-term land-use history and human-environment interaction processes in broad valleys is essential for the harmonious development of the Tibetan Plateau. In this study, we analyzed high-resolution multi-proxy records derived from Ruba Lake in the middle reaches of the Yarlung Zangbo River, including pollen, grain-size, End-member modelling, total organic carbon, total nitrogen, black carbon (BC), and delta 13C of black carbon (delta 13CBC), to reconstruct the valley agriculture history and its environmental background over the last 2600 years. There is possible evidence that agriculture was practiced in the Ruba Basin since ca. 2000 cal yr BP, while it flourished after ca. 500 cal yr BP, as represented by increases in the pollen abundance and frequency of cereal Poaceae. Increased Artemisia and Nitraria indicate a regional aridification trend on the southern Tibetan Plateau, while elevated BC concentrations and enriched delta 13CBC values reflect intensified human activities. The increase in coarse-grained sediments further suggests enhanced soil erosion under these conditions. Our results confirm that regional drought at ca. 500 cal yr BP (middle Ming Dynasty) did not apparently affect the valley agriculture directly because of advanced cultivation technology and the relatively plentiful water resources.

Pollen serves as an important proxy in the reconstruction of past climate and vegetation. However, the relationships between pollen, vegetation, and climate on the eastern Tibetan Plateau (TP) remain insufficiently validated. This study aims to investigate these relationships based on pollen assemblages from 181 surface samples. Discriminant analysis and clustering reveal that pollen assemblages represent the altitudinal vegetation zones well. Redundancy analysis (RDA) indicates that mean temperature of the coldest month (Mtco) is the primary climate factor explaining the variation of pollen assemblages. Weighted average ecological optima and tolerances of common pollen taxa along environmental variables are consistent with their modern distributions, suggesting they are strongly indicative of vegetation and climate. Among the methods used for Mtco prediction, random forests (RF) showed the best performance, making it suitable for past climate reconstruction. Some pollen taxa have limited indicator capacity, requiring cautious interpretation to minimize potential bias. Overall, this study fills a gap in regional research on the eastern TP, demonstrates a robust and consistent correspondence between pollen assemblages, vegetation, and climatic factors, and provides a reliable reference for past environmental reconstruction in this region.

Lake sediments on the Tibetan Plateau (TP) are key archives for understanding paleoenvironmental changes and have been extensively used to investigate climate change since the Last Deglaciation. Long-term landscape changes such as vegetation cover are influenced by climate, and can, in turn, affect fluvial and sedimentation processes, and thus the interpretation of records. Such changes are insufficiently studied. In this study, we present multi-proxy records covering the past 46 cal ka BP, including elemental composition, carbon and nitrogen content, and grain size from Lake Koucha on the northeastern TP. By integrating these data with vegetation cover inferred from another core (KC-19) from the same lake, we examine the potential influence of vegetation cover on hydrological processes. The results indicate that during the interval from 46 to 14 cal ka BP, exogenous clastic material was mainly transported into the lake via runoff triggered by monsoonal precipitation. However, the lake experienced a substantial reduction in detrital input thereafter, despite the climate shifting to warmer and wetter conditions. This decline may be attributed to the vegetation cover surpassing a threshold at approximately 14 cal ka BP, which effectively buffered erosion and sediment transport processes within the basin. Comparable patterns in published records from the central TP support our results. Our study provides a regional case from the northeastern TP, illustrating how changes in vegetation cover influence the interpretation of lake sedimentary proxies over glacial-interglacial timescales, and suggesting that some previous paleoenvironmental studies on the TP may need to be re-examined.

Vegetation dynamics on the central Tibetan Plateau (TP) over the past two millennia provide critical insights into ecological responses to future warming. In this study, a 42-cm-long sediment core from Jiaruo Co on the central TP, spanning the last two millennia, was analyzed for pollen, charcoal, and geochemical proxies to reconstruct past vegetation and its underlying drivers. From 60 to 550 CE, alpine steppe prevailed around the lake, as evidenced by relatively high pollen percentages of Artemisia, Cyperaceae, and Poaceae, alongside low Amaranthaceae abundance. After 550 CE, a pronounced turnover in pollen assemblages was recorded, characterized by a marked rise in arid-tolerant taxa (e.g., Amaranthaceae, Ephedra, and Rosaceae) and a concurrent decline in Artemisia, Ranunculaceae, and Lamiaceae, indicating a shift toward drought-adapted composition. In addition, both decreased total pollen concentration and increased Pinus abundance indicate reduced vegetation coverage. This functional (increased drought-adapted composition) and structural (reduced coverage) vegetation turnover was primarily driven by aridification, as corroborated by our multiproxy records (e.g., decreasing TOC/TN ratios and increasing Rb/Sr ratios) and independent evidence from previous regional records. We thus infer that this aridification was a regional phenomenon, resulting from the weakening of the Indian Summer Monsoon (ISM). Our study highlights the substantial ecological risks posed by ongoing warming.

Clarifying the relationship between soil-surface phytolith assemblages and land-use types in valley agricultural areas is essential for reconstructing the origins and spread of agriculture on the Tibetan Plateau. However, research on modern phytoliths related to wheat (Triticum sativa L.) and barley (Hordeum vulgare L.) cultivation at high elevations remain limited. In this study, we analysed 40 farmland and 17 natural vegetation soil-surface samples from the Hehuang Valley, a typical valley agricultural area. The results show that phytolith assemblages generally reflect the cold, arid climate of the region but also vary across land-use types. Wheat farmland can be distinguished from natural vegetation, with ELONGATE DENDRITIC >= 1.0% as a diagnostic indicator for wheat and CRENATE <3.8% for natural vegetation. These findings highlight the potential of phytolith analysis for land-use reconstruction and establish an operational framework and threshold standards for identifying wheat farmland in the Hehuang Valley.

Abstract. Modern marine pollen records can be interpreted reliably only when the main source-to-sink processes between terrestrial vegetation, rivers, and marine depocentres are documented by comparable modern data. Here, we present and quality-assess a paired marine-fluvial pollen dataset for the Bohai Sea, China, consisting of 430 marine surface-sediment samples and 104 fluvial surface-sediment samples from 16 major inflowing rivers. The dataset includes sample coordinates, water depth for marine stations, river and regional identifiers, pollen and spore counts, pollen concentrations, and grain-size parameters. In total, 104 pollen and spore morphotypes were identified. Marine assemblages are dominated by arboreal pollen, especially Pinus (mean 59 %), whereas nearshore and river-influenced settings contain higher proportions of herbaceous taxa, particularly Chenopodiaceae/Amaranthaceae-type (mean 14 %), Artemisia, and Poaceae. Redundancy analysis after variance-inflation-factor screening indicates that water depth is the strongest retained predictor of marine pollen composition (λ1/λ2 = 0.303), followed by geographic position and sediment texture. In river samples, geographic position, water discharge, and pollen concentration together explain 20.53 % of the compositional variance. Along the 0–35 m depth interval, arboreal pollen and Pinus increase with water depth, whereas herbaceous taxa decrease, providing an internally consistent modern analogue for distinguishing nearshore fluvial influence from more distal aeolian influence. The dataset is intended primarily as a reusable calibration and quality-assessment resource for palaeoenvironmental studies in the Bohai-Yellow Sea region. The archived data are available at the National Tibetan Plateau Data Center (TPDC; Li et al., 2026; https://doi.org/10.11888/Paleoenv.tpdc.303516).

Abstract Forest expansion and retreat are key drivers of alpine ecosystem dynamics, yet it remains unclear how forest expansion shapes alpine plant communities and whether alpine assemblages that re-emerge after forest decline resemble those of the Late Glacial. Here, we integrate sedimentary ancient DNA (sedaDNA) and pollen records from Zhagaer Co on the eastern Tibetan Plateau to reconstruct vegetation changes over the past ∼14 ka. Using Picea abundance as a proxy for forest cover, we apply constrained ordination of sedaDNA-inferred plant communities to identify plant taxa as “winners” and “losers” of forest expansion and compare the composition of loser taxa between the late Glacial and late Holocene cold-open phases. Our results show that forest expansion during the early to mid-Holocene favoured woody and forest-margin taxa (e.g. Rhododendron , Salicaceae), while suppressing alpine forbs and graminoids (e.g. Carex , Thalictrum ), consistent with patterns expected under ecological filtering. However, late Holocene reopening did not fully restore a late Glacial-like alpine community; instead, it was characterised by a stronger contribution of alpine meadow and shrub taxa. This difference may reflect contrasting environmental backgrounds, including higher atmospheric CO2 levels during the late Holocene, ecological legacies of prior forest expansion, and increasing human influence. These findings suggest that forest expansion may lead to long-term restructuring of alpine plant assemblages, and that late Holocene cooling did not simply restore late Glacial alpine communities but instead produced a distinct alpine ecosystem state. Together, these results highlight the long-term legacy of treeline dynamics in shaping alpine ecosystem trajectories.

Understanding spatial differences in the millennial-scale environmental evolution on the central-western Tibetan Plateau is essential for revealing mechanisms of climate changes. In this study, a 787-cm sediment core covering the last 18 thousand years (ka) from Ngamring Co was analyzed for pollen to infer long-term vegetation changes. In addition, a modern pollen dataset (n = 964) and fossil pollen records (n = 6) from the western-central Tibetan Plateau were used to investigate past precipitation changes, using the quantitative method Weighted Averaging Partial Least Squares (WA-PLS) and the Cyperaceae/Chenopodiaceae ratio (Cy/C). Pollen spectra from Ngamring Co indicate that arid alpine desert was distributed around the lake from 18 to 14.6 cal ka BP as revealed by low pollen concentration and a high proportion of exogenous arboreal pollen, followed by a shift to alpine steppe until 10.3 cal ka BP. Alpine shrub-steppe (dominated by Cyperaceae, Artemisia, Poaceae, and Hippophae) emerged afterwards and was progressively supplanted by drier alpine steppe with sparse vegetation cover after 3.5 cal ka BP. Quantitative reconstructions and the pollen ratio reveal a distinct transition (characterized by progressive precipitation reduction and amplified aridity) since the Mid-Late Holocene across the central-western Tibetan Plateau, gradually weakening from west to east. This aridification is likely linked to the weakening summer monsoon. In addition, pollen and observational data suggest a recent shift in desertification trends on the centralwestern Tibetan Plateau.

Limited topographic connectivity to refugia can delay plant community responses to climate warming, but its influence on long-term community assembly remains unclear. Here we reconstruct plant community dynamics on the eastern Tibetan Plateau over the past 12,000 years using sedimentary ancient DNA records from Dangzi Co and four published datasets, with zeta diversity quantifying shared taxa across sites and time-slices. We examine the roles of climate and topographic connectivity in shaping communities. Warming before 8.5 ka (ka = thousand years before present) triggered upslope migration of lowland woody taxa, but variable topographic connectivity among lakes caused asynchronous assembly. Under a stable mid-Holocene climate (8.5–5.5 ka), niche differentiation increased, promoting community similarity. Subsequent cooling (5.5–3.5 ka) facilitated alpine taxa expansion, maintaining high taxa sharing among lakes. Reduced convergence since 3.5 ka likely reflects cooling and land use. Our findings highlight how topographic connectivity shapes millennial-scale community assembly and biodiversity. Topographic connectivity to refugia shaped millennial plant community assembly, with delayed and spatially asynchronous responses to climate revealed by zeta diversity analysis of sedimentary ancient DNA from five lakes over the past 12,000 years.

The saline lake ecosystem in the Qaidam Basin is notoriously fragile and sensitive to global climate change. Over the past two decades, climatic humidification and brine extraction have together driven significant shifts in local vegetation and hydrological patterns around these lakes. Thus, studying long-term vegetation succession here is crucial for projecting future ecosystem trajectories. This study presents a high-resolution pollen analysis of a sediment core (XT) from the Xitaijnar Saline Lake (central Qaidam Basin). The biome method was employed to quantitatively reconstruct regional vegetation succession and lake evolution since the late MIS 3. The results indicate that during the late MIS 3 (similar to 36.4-30.0 cal kyr BP), temperate steppe and meadow (dominated by Artemisia, Poaceae, Chenopodiaceae, and Cyperaceae pollen) expanded significantly under relatively warm and humid conditions. During MIS2 (similar to 30.0-10.0 cal kyr BP), a marked increase in pollen from Ephedra, Tamaricaceae, and Nitraria points to a substantial expansion of desert shrubland under extreme aridity. The Holocene climate was characterised by aridity in the early-mid stages (similar to 10.0-4.0 cal kyr BP), during which desert shrubland persisted. Subsequently, in the late Holocene (similar to 4.0-1.0 cal kyr BP), temperate steppe and meadow increased as the climate shifted toward humidity. The hydroclimatic changes recorded in the central Qaidam Basin since the late MIS 3 correspond to patterns seen across the Westerlies-dominated Northern Hemisphere. We infer that vegetation succession in the basin has primarily been governed by variations in Northern Hemisphere mid-to-high latitude solar radiation and North Atlantic moisture transport. These findings clarify the response of the Qaidam Basin's ecosystem to orbital-scale climate forcing and establish a foundation for modelling its future trajectory.

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.

Identifying agricultural signals in sediments using phytoliths is effective in tracing early subsistence strategies. However, on the Tibetan Plateau this work remains limited by the absence of a systematic modern phytolith reference framework for the Plateau's characteristic valley agricultural regions. In this study, we analyzed 77 farmland and 37 non-farmland topsoil samples collected in the middle reaches of the Yarlung Tsangpo River Valley to evaluate phytolith assemblage responses to different land-use types. The results show clear and statistically significant differences in phytolith assemblages between farmland and non-farmland sites. Farmland samples are characterized by elevated abundances of inflorescence-derived morphotypes, including Elongate dendritic, Papillate radiate, and Papillate circular, together with a greater degree of homogeneity. Assemblages from non-farmland sites are dominated by short-cell morphotypes and exhibit greater compositional variability. Machine-learning classification and indicator species analyses (ISA) demonstrate that land-use type can be distinguished based on phytolith assemblages, with Elongate dendritic showing a clear threshold (2.8%) response that effectively separates farmland from non-farmland samples. Crop types had a weak influence on variation in phytolith assemblages, explaining only 8% of the total variance. The result might indicate that phytoliths were mixed across multiple cropping cycles due to crop rotation and shallow tillage, which further suggests that topsoil phytolith assemblages mainly reflect long-term, integrated land-use history rather than a single season's or the current crops. This study provides a modern phytolith reference for valley agriculture on the southern Tibetan Plateau, offering a potential basis for identifying sedimentary agricultural signals and supporting methodological approaches in high-altitude archaeological and paleoenvironmental research.

Forest expansion and retreat are key drivers of alpine ecosystem dynamics, yet it remains unclear how forest expansion shapes alpine plant communities and whether alpine assemblages that re-emerge after forest decline resemble those of the Late Glacial. Here, we integrate sedimentary ancient DNA (sedaDNA) and pollen records from Zhagaer Co on the eastern Tibetan Plateau to reconstruct vegetation changes over the past similar to 14 ka. Using Picea abundance as a proxy for forest cover, we apply constrained ordination of sedaDNA-inferred plant communities to identify plant taxa as "winners" and "losers" of forest expansion and compare the composition of loser taxa between the late Glacial and late Holocene cold-open phases. Our results show that forest expansion during the early to mid-Holocene favoured woody and forest-margin taxa (e.g. Rhododendron, Salicaceae), while suppressing alpine forbs and graminoids (e.g. Carex, Thalictrum), consistent with patterns expected under ecological filtering. However, late Holocene reopening did not fully restore a late Glacial-like alpine community; instead, it was characterised by a stronger contribution of alpine meadow and shrub taxa. This difference may reflect contrasting environmental backgrounds, including higher atmospheric CO2 levels during the late Holocene, ecological legacies of prior forest expansion, and increasing human influence. These findings suggest that forest expansion may lead to long-term restructuring of alpine plant assemblages, and that late Holocene cooling did not simply restore late Glacial alpine communities but instead produced a distinct alpine ecosystem state. Together, these results highlight the long-term legacy of treeline dynamics in shaping alpine ecosystem trajectories.

Understanding long-term interactions between climate, permafrost, and vegetation provides an essential context for interpreting current Arctic greening. Using 65 fossil pollen records from northern Siberia and a Random Forest model trained on a dataset of 835 modern pollen-climate assemblages, we quantitatively reconstructed mean temperature of the warmest month (Mtwa: mean July temperature) anomalies over the past 40 thousand years (ka) and assessed associated vegetation changes. During the Last Glacial Period, herbaceous taxa overwhelmingly dominated, and warming of similar to 1 degrees C during similar to 40-35 cal ka BP was insufficient to deepen the active layer beyond the threshold required for tree establishment, leaving woody cover minimal. In the early Holocene, sustained warming of nearly 2 degrees C triggered permafrost degradation and active-layer thickening, enabling forest expansion, although tree taxa lagged shrubs by several millennia. These results reveal a clear threshold effect in vegetation-permafrost interactions and show that only sustained warming can overcome permafrost constraints. By providing quantitative temperature estimates, our reconstruction offers critical benchmarks for predicting how ongoing Arctic warming may transform vegetation patterns and permafrost stability.

The influence of paleoclimate in shaping current biodiversity pattern is widely acknowledged. However, it remains unclear how the upper paleo-range limit of trees, which dictated the habitat of endemic alpine species, affects the variability in endemic alpine species composition across space over the Tibetan Plateau. We integrated satellite-derived upper range limit of trees, dendrochronological data, and fossil pollen records with a paleoclimate dataset in a climate-driven predictive model to reconstruct the spatio-temporal upper range limit of trees at 100-year intervals since the Last Glacial Maximum. Our results show that trees distributed at the lowest elevations during the Last Glacial Maximum (~3426 m), and ascended to the highest elevations during the Holocene Climatic Optimum (~4187 m), a level ~180 m higher than the present-day (~4009 m). The temporal fluctuations in paleo-range limits of trees play a more important role than paleoclimate in shaping the current spatial pattern of beta-diversity of endemic flora, with regions witnessing higher fluctuations having lower beta-diversity. We therefore suggest that anthropogenic-caused climate change on decadal-to-centennial timescales could lead to higher fluctuations in range limits than orbitally-forced climate variability on centennial-to-millennium timescales, which consequently could cause spatial homogenization of endemic alpine species composition, threatening Tibetan endemic species pool.