Spatiotemporal characteristics and influencing factors of vegetation water use efficiency on the Tibetan Plateau in 2001-2020

HE Chenyang, WANG Yanjiao, YAN Feng, LU Qi

Journal of Geographical Sciences ›› 2025, Vol. 35 ›› Issue (1) : 39-64.

PDF(159087 KB)
PDF(159087 KB)
Journal of Geographical Sciences ›› 2025, Vol. 35 ›› Issue (1) : 39-64. DOI: 10.1007/s11442-025-2312-9
Special Issue: Climate Change and Water Environment

Spatiotemporal characteristics and influencing factors of vegetation water use efficiency on the Tibetan Plateau in 2001-2020

Author information +
History +

Abstract

Water use efficiency (WUE), as a pivotal indicator of the coupling degree within the carbon-water cycle of ecosystems, holds considerable importance in assessment of the carbon-water balance within terrestrial ecosystems. However, in the context of global warming, WUE evolution and its primary drivers on the Tibetan Plateau remain unclear. This study employed the ensemble empirical mode decomposition method and the random forest algorithm to decipher the nonlinear trends and drivers of WUE on the Tibetan Plateau in 2001- 2020. Results indicated an annual mean WUE of 0.8088 gC/mm∙m2 across the plateau, with a spatial gradient reflecting decrease from the southeast toward the northwest. Areas manifesting monotonous trends of increase or decrease in WUE accounted for 23.64% and 9.69% of the total, respectively. Remarkably, 66.67% of the region exhibited trend reversals, i.e., 39.94% of the area of the Tibetan Plateau showed transition from a trend of increase to a trend of decrease, and 26.73% of the area demonstrated a shift from a trend of decrease to a trend of increase. Environmental factors accounted for 70.79% of the variability in WUE. The leaf area index and temperature served as the major driving forces of WUE variation.

Key words

water use efficiency / spatiotemporal characteristic / influencing factor / Tibetan Plateau

Cite this article

Download Citations
HE Chenyang, WANG Yanjiao, YAN Feng, LU Qi. Spatiotemporal characteristics and influencing factors of vegetation water use efficiency on the Tibetan Plateau in 2001-2020[J]. Journal of Geographical Sciences, 2025, 35(1): 39-64 https://doi.org/10.1007/s11442-025-2312-9

1 Introduction

Water and carbon cycling are paramount coupled ecological processes within terrestrial ecosystems, playing pivotal roles in the surface fluxes of matter and energy (Mo et al., 2005; Lu et al., 2019; Bai et al., 2020). The equilibrium of water and carbon within such a system is intricately tied to exchanges between the soil, vegetation, and atmosphere (Ji et al., 2021). Vegetation is an integral component of terrestrial ecosystems and it acts as an important indicator in various geographical settings, orchestrating and modulating material and energy dynamics across soil, water, and air interfaces (Sun et al., 2015; Gao et al., 2020; Lian et al., 2022). Water use efficiency (WUE) is a pivotal metric delineating the intertwined relationship between the carbon and water cycles of an ecosystem. It encapsulates the productivity response of plants under varied water conditions, and it is characterized as the quantum of carbon or biomass production attainable per unit of water expended (Yang et al., 2016). WUE provides critical insight into the interplay between plant hydration strategies and ecosystem-wide water utilization efficacy. When confronted with water scarcity, vegetation must adeptly manage resource allocation to harmonize the concurrent demands of water and carbon, potentially altering ecosystem productivity dynamics. Within the broader narrative of global climatic shifts, discerning the spatiotemporal nuances of WUE and the driving factors becomes imperative. Such understanding can fortify ecosystem adaptability measures, temper climatic and ecological vulnerabilities, and amplify comprehension of terrestrial ecosystem reactions to evolving climatic patterns (Zhang et al., 2016).
Climate-induced variations exert direct influence on the evolutionary growth trajectories of vegetation (Nemani et al., 2003). Concurrently, vegetation can dynamically modulate climatic fluctuations through intricate adjustments of the carbon-water cycle and energy fluxes (Bonan, 2008; Ma et al., 2022). The methodological framework for discerning WUE at the level of both the leaf and the individual plant predominantly encompasses techniques such as gas exchange analyses, stable isotope diagnostics, and direct field assessments (Dawson et al., 2002; Barbour, 2007; Donovan et al., 2007). As the depth of understanding concerning ecosystem WUE increases, traditional estimation paradigms, anchored in the metrics of plant productivity and evapotranspiration (ET), appear less adept at addressing the nuanced demands of expansive spatial and extended temporal investigations of terrestrial ecosystem WUE. The rapid development of remote sensing technologies has ushered in an era of comprehensive, long-term terrestrial ecosystem monitoring (Tang et al., 2015; He et al., 2022a), thereby bolstering endeavors aimed at deciphering the intricacies of ecosystem WUE across vast scales and over prolonged temporal horizons (Ji et al., 2021).
Ecosystem WUE is modulated by various environmental drivers that encompass climatic variables, biotic elements, and soil attributes (Liu et al., 2015). The importance of such factors on WUE varies in magnitude across temporal and spatial dimensions (Du et al., 2019). In high-latitude regions, temperature exerts a pronounced positive impact on WUE. As surface temperatures rise, the duration of the growing season increases (Körner and Basler, 2010; Gunderson et al., 2012), resulting in elevated gross primary production (GPP) and ET. Concurrently, although the rate of photosynthesis is directly augmented by surface warming, the transpiration rate tends to remain stable owing to stomatal regulation (Gunderson et al., 2000; Flanagan and Syed, 2011). Conversely, in humid regions, temperature imparts a detrimental effect on WUE. With ample precipitation, elevated temperatures in such zones amplify ET, predominantly by mitigating the latent heat of vaporization (Zhu et al., 2011). In both high-latitude and moisture-abundant regions, precipitation is negatively correlated with WUE, whereas a positive relation is found in water-scarce zones (Nemani et al., 2003; NIU et al., 2011; Sun et al., 2016). Puma et al. (2013) and Li et al. (2018) reported that an enhanced leaf area index (LAI) bolsters absorption of photosynthetically active radiation, thereby invigorating GPP and elevating WUE, especially in biotic communities with sparse LAI. Zhu et al. (2016) highlighted that a continuous uptick in global LAI reduces the proportion of solar radiation reaching the soil surface, thereby curtailing bare soil evaporation and attenuating ET (Huang et al., 2015).
The Tibetan Plateau is recognized as a critical freshwater reservoir, playing an indispensable role in water resource provisioning for Asia (Ren et al., 2024). This plateau ranks highly among those regions displaying heightened sensitivity to global warming (Chen et al., 2020). The biologically constrained hydrothermal conditions of the Tibetan Plateau render its ecosystem exceptionally vulnerable. Recent years have witnessed profound impacts of global warming on terrestrial ecosystems (Zhou et al., 2014). As per the latest assessment from the Intergovernmental Panel on Climate Change, the global mean temperature has escalated by approximately 0.85℃ since 1850. Furthermore, anthropogenic activities have driven amplification in the concentration of atmospheric carbon dioxide of approximately 40% since 1750, increasing from approximately 280 ppm to around 400 ppm and marking the peak over the past 800,000 years. Delving deep into the spatiotemporal characteristics of WUE by vegetation on the Tibetan Plateau against the backdrop of global warming, together with consideration of its climatic effects, is of paramount importance for judicious resource management, enhancing the efficacy of water utilization, and supporting the stability and sustainability of the plateau’s ecosystems.
Earlier research on vegetation WUE predominantly employed linear regression methodologies, ascertaining long-term unidirectional trends for individual grids at a constant rate, thereby reflecting the spatiotemporal patterns of vegetation WUE within a specified region. This approach involves fitting a linear trend line to time-series data for each grid cell, allowing researchers to detect overall increases or decreases in WUE over time. These methods have been particularly useful for identifying broad spatial patterns and understanding the general behavior of WUE across different regions (Sun et al., 2018; Li et al., 2021). However, reality suggests that vegetation WUE trajectories are nonlinear and nonstationary. A simple linear regression approach omits intrinsic breakpoints and trend shifts, thereby obscuring authentic variations in vegetation WUE. The ensemble empirical mode decomposition (EEMD) approach delineates nonstationary time series into finite components with decreasing frequencies and a long-term trend (Wu and Huang, 2009). The extrapolated long-term trend is either monotonic or encapsulates a singular extremum, with its rate subject to temporal modulation. Notably, EEMD trends are unencumbered by prescribed a priori assumptions and are resilient to time series expansions (Ji et al., 2014). This methodology, which is proficient in unveiling fundamental information on nonlinear and nonstationary time series (Pan et al., 2018), exhibits efficacy in monitoring prolonged time series of vegetation growth dynamics and trend transitions (Yin et al., 2017; Xu et al., 2020). Current research into the impact of environmental changes on vegetation WUE often resorts to using statistical models for characterizing environmental determinants. Traditional approaches, such as correlation and partial correlation analyses, have been widely employed to explore the relationships between WUE and individual environmental variables (Gao et al., 2019; Zheng et al., 2019; Xue et al., 2022). Given the intricate, nonlinear interrelation between vegetation WUE and environmental shifts, a singular regression equation is inadequate for detailing their symbiotic nexus (Zheng et al., 2018).
Therefore, this study focused on the WUE of vegetation on the Tibetan Plateau to address the following: (1) to determine the spatiotemporal distribution patterns of WUE and its variations across different vegetation types and elevational gradients; (2) to assess the contemporary trends in vegetation WUE on the plateau; and (3) to identify the determinants influencing the shifts in WUE of the vegetation on the plateau and to elucidate their respective contributions.

2 Study area and data

2.1 Study area

The Tibetan Plateau in southwestern China (25°-40°N, 73°-105°E; Figure 1) covers an area of approximately 2.5×106 km2. Characterized by its complex topography, the plateau maintains an average elevation of >4000 m that generally exhibits a topographical trend of decrease from the west toward the east. Most regions of the plateau register an annual average temperature of <5℃. The higher western plateau, with elevations commonly surpassing 5000 m, has an average annual temperature of approximately −10℃, whereas the annual average temperature of the northern and eastern sectors of the plateau is in the range 0-10℃ (Sun et al., 2015). Annual precipitation across the plateau is generally <400 mm, with the major proportion (60%-90%) occuring during the summer months between June and September (Hu et al., 2012; Tong et al., 2014; You et al., 2015). The plateau boasts a unique ecosystem and myriad vegetation types. Varying moisture and thermal conditions dictate the sequential appearance of forests, shrubs, alpine meadows, alpine grasslands, and desert vegetation from the southeast toward the northwest. Notably, alpine meadows and grasslands predominate, accounting for over half of the area of the plateau. Serving as one of Asia’s largest freshwater reservoirs, the Tibetan Plateau is hydrologically diverse, encompassing the world’s third-largest inland lake (i.e., Qinghai Lake) and providing the source of iconic Asian rivers such as the Yangtze, Yellow, and Lancang rivers.
Figure 1 Location and main vegetation types on the Tibetan Plateau

Full size|PPT slide

2.2 Data

We retrieved ET and GPP data from the National Tibetan Plateau Data Center (https://data.tpdc.ac.cn/home). With temporal resolution of 1 day and spatial resolution of 500 m, these datasets comprise GPP and ET on the daily scale estimated using the coupled carbon-water Penman-Monteith-Leuning model version 2. This model has been efficiently calibrated using observational data from 26 eddy covariance flux towers spanning 9 plant functional types in China. The datasets chiefly encompass GPP, soil evaporation, canopy interception evaporation, plant transpiration, and evaporation from water and snow/ice surfaces (He et al., 2022b). For this study, soil evaporation, canopy interception evaporation, and plant transpiration were integrated to represent ET data, and daily data from 2001 to 2020 were aggregated into annual data.
MOD15A2H-LAI data from 2001 to 2020 with 500-m spatial resolution were sourced from NASA’s Land Processes Distributed Active Archive Center. MOD15A2H provides LAI data aggregated over an 8-day period, selecting the highest quality observation for each pixel within each period. This 8-day composite LAI product from MODIS is suitable for monitoring global vegetation conditions and assessing vegetation growth and photosynthetic intensity (https://ladsweb.modaps.eosdis.nasa.gov/). The maximum value composite method was used to calculate the annual maximum LAI to scrutinize the influence of LAI on WUE.
Vegetation type data were acquired from the Resource and Environmental Science and Data Center of the Chinese Academy of Sciences (https://www.resdc.cn), and digitized from the 1:1 million vegetation atlas. For this research, vegetation types were consolidated based on the actual vegetation distribution on the Tibetan Plateau into forest, meadow, grassland, shrubs, cultivation, alpine, desert, and swamp.
Surface net solar radiation (SSR), 2-m temperature (TMP), and soil temperature level 2 (STMP) data were derived from the ERA5-Land monthly averages (https://cds.climate.copernicus.eu). Published by the European Centre for a more nuanced understanding of terrestrial climatic and hydrological variations, this dataset amalgamates surface observations, satellite remote sensing, and meteorological model forecasts(Copernicus Climate Change Service, 2019). It provides monthly averages for surface variables such as temperature, precipitation, snow depth, and ET with 0.1° spatial resolution. For this study, monthly averages of January-December spanning 2001-2020 were amalgamated into yearly data. Given the substantial diurnal variations in temperature on the Tibetan Plateau, with superficial soil temperatures being notably susceptible to ambient conditions (Xu et al., 2011), soil temperatures at depth of 7-20 cm were adopted to represent ground temperatures.
Annual cumulative precipitation data (PRE) with 1-km spatial resolution were procured from the National Tibetan Plateau Data Center (https://data.tpdc.ac.cn/home). This dataset was produced by applying the Delta spatial downscaling scheme to global climate datasets released by the Climate Research Unit and WorldClim Global Climate Data. The reliability of this dataset, validated against 496 independent meteorological observation points, makes it widely applicable to meteorological, hydrological, and environmental science research (Peng et al., 2017; 2019). This study chiefly used monthly averages of January-December 2001-2020 to compose the annual cumulative precipitation data.

3 Methodologies

The workflow of this study is illustrated in Figure 2.
Figure 2 Flowchart of methodology

Full size|PPT slide

3.1 Vegetation water use efficiency

Vegetation WUE is computed using the ratio of GPP to ET:
WUE=GPPET
(1)
where WUE represents water use efficiency (gC/mm∙m2), and GPP and ET are the gross primary productivity (gC m−2 a−1) and evapotranspiration (mm), respectively.

3.2 Coefficient of variation

In quantitative evaluation of datasets, the standard deviation and the coefficient of variation (CV) are essential indicators used to measure the degree of dispersion. The standard deviation is a measure of how a set of values deviates from the mean, and the CV is a statistical metric that assesses the extent of the variability among the observed values and represents the ratio of the standard deviation to the mean (Pearson, 1896). This coefficient can mitigate the influence of different units or averages when comparing the dispersion of two or more datasets, thus reflecting the dispersion per unit mean. A smaller CV indicates a lower degree of data dispersion, suggesting greater relative stability of the data (Yan et al., 2015). The standard deviation and CV can be expressed as follows:
σ=1ni=1n(xix¯)2 (i=1,2,3,...)
(2)
CV=σx¯
(3)
where σ is the standard deviation, xi is the WUE for the i-th year, x¯±s is the mean WUE during 2001-2020, and i and n represent the years from 2001 to 2020.

3.3 Slope trend analysis

Trend analysis based on the linear regression method can effectively reflect the spatiotemporal patterns of change in the time series in the study area (Peng et al., 2012; Liu et al., 2021b). This study used linear regression to analyze the trend of WUE on the Tibetan Plateau during 2001-2020:
Slope=ni=1n(iWUEi)(i=1ni)(i=1nWUEi)ni=1ni2(i=1ni)2
(4)
where n is the total number of years in the study period, WUEi is the WUE of the i-th year, and Slope represents the change rate of the regression equation. A value of Slope > 0 (Slope < 0) indicates that WUE has a trend of increase (decrease).

3.4 Mann-Kendall trend test

The Mann-Kendall (MK) trend test is a nonparametric statistical method (Mann, 1945; Kendall, 1948). Owing to its ability to function without requiring sample data to adhere to a specific distribution and its resilience to outliers, it can effectively test the statistical significance of change in a time series trend. In this study, the MK test was applied to assess the statistical significance of the Slope trend. Within the MK trend test, the null hypothesis posits that time series data consist of a series of independent, identically distributed random variables. The statistical metric Z is used as an indicator of time series change, with its specific calculation formula expressed as follows:
Z={S1var(S),S>00,S=0S+1var(S),S<0
(5)
S=k=1n1j=k+1nsgn(xjxk)
(6)
sgn(xkxj)={1,xkxj>00,xkxj=01,xkxj<0
(7)
var(S)=n(n1)(2n+5)/18
(8)
where xk and xj are samples from the time series dataset, n represents the length of the time series, and sgn is the sign function. Given a statistical significance level of α, a value of ∣Z∣> Z1−α/2 indicates that the null hypothesis of no trend is rejected, implying a statistically significant change in the trend of the time series. In this study, a value of ∣Z∣>1.96 was adopted to signify that the time series passed the test of statistical significance.

3.5 Ensemble empirical mode decomposition method

The EEMD method is a signal processing technique that decomposes complex nonlinear and nonstationary signals into several intrinsic mode functions (IMFs) and a residual long-term trend (Rn) based on empirical mode decomposition (Wu and Huang, 2009; Ji et al., 2014). The extracted long-term trend is either monotonic or contains only one extremum, facilitating applications such as signal denoising, analysis, and forecasting. The IMFs derived from the EEMD represent oscillatory patterns across different temporal scales of the original signal, enhancing the understanding and processing of time series data. This study decomposed the nonlinear trend of WUE and its changes over time using the EEMD method. The specific steps of the EEMD method are outlined in the following.
Step 1: Add Gaussian white noise w1(t) to the original time series data x(t). The amplitude of the white noise is 0.2 times the standard deviation of the original data:
x1(t)=x(t)+w1(t)
(9)
Step 2: Identify all the maxima and minima in the data and connect them using a cubic spline curve to form the upper and lower envelopes of the new time series data x1(t). Then subtract the mean m1(t) of the upper and lower envelopes from the new time series data x1(t):
f1(t)=x1(t)m1(t)
(10)
Step 3: Determine whether m1(t) satisfies the stopping criterion (close to 0 at any point). If it does, halt the decomposition to get the first IMF. If not, treat f1(t) as the new time series and repeat Step 2 until the stopping criterion is met. Eventually, obtain the first IMF: imf1(t):
f2(t)=f1(t)m2(t)
(11)
imf1(t)=fk(t)=fk1(t)mk(t)
(12)
Step 4: Subtract the first IMF, imf1(t), from x1(t) to get the residue R1(t). If R1(t) still contains oscillatory components, take R1(t) as the new time series and repeat Steps 2 and 3:
R1(t)=x1(t)imf1(t)
(13)
Rn(t)=Rn1(t)imfn(t)
(14)
At this point, x1(t) has been decomposed into a series of decreasing frequency IMFs and a monotonic trend or a trend with at most one extremum:
x1(t)=j=1nimfj(t)+Rn(t)
(15)
Step 5: Repeat Steps 1 to 4 a total of i times (here i is set to 5000). Reconstruct the original data multiple times by adding different Gaussian white noise sequences to the original data and perform EEMD decomposition to obtain IMFs and intrinsic trends. The final result is obtained by averaging the outcomes from these multiple decompositions.
On the basis of the monotonicity and extremum characteristics of the intrinsic trend, the WUE time series was categorized into four trends: Monotonically Increasing (In to In), Monotonically Decreasing (De to De), Increasing to Decreasing (In to De), and Decreasing to Increasing (De to In).

3.6 Random forest model

The random forest (RF), which is an ensemble learning algorithm composed of multiple classification and regression decision trees, is suitable for effective handling of high-dimensional data and multisample classification problems (Breiman, 2001). The central idea behind the RF approach is to classify or regress data using multiple decision trees, thereby combining the results of multiple models to enhance accuracy. The RF algorithm not only reduces the computational demand but also improves the predictive accuracy. Moreover, it exhibits robustness to missing and imbalanced data and can handle datasets with thousands of explanatory variables, rendering it insensitive to multicollinearity (Franke et al., 2019; Assal et al., 2021). The RF algorithm uses IncNodePurity as an estimate of predictor variable importance. IncNodePurity is measured by calculating the difference in the residual sum of squares (RSS) for each variable at each node. When constructing classification trees, observations are allocated to different nodes based on the values of each variable. For each node, the RSS of all observations on that node is computed, and then the sum of the RSS of the left and right child nodes is subtracted. This difference represents the IncNodePurity of that variable on that node. A higher value of IncNodePurity indicates that the variable has greater contribution to the splits in the classification tree, i.e., the variable is of higher importance. Two crucial parameters of the RF method are the number of variables preselected for tree nodes (mtry) and the number of trees in the RF (ntree). In this study, mtry was set to 3 and ntree was set to 500.

4 Results

4.1 Spatial variation in water use efficiency

On the Tibetan Plateau, the spatial variation in WUE during 2001-2020 exhibited a distinct pattern of general reduction from the southeast toward the northwest (Figure 3). The spatial distribution and growth status of vegetation types play a pivotal role in driving the spatial disparity in WUE across the plateau. In southeastern and eastern regions of the Tibetan Plateau, forests, marshland, and cultivated plants manifest relatively higher WUE. This is primarily attributed to the humid climatic conditions prevalent in these zones, ensuring relatively abundant moisture resources. Plants, in adapting to these moist environments, require enhanced mechanisms for water uptake and transport regulation, thereby elevating WUE. Furthermore, the topographical and soil conditions favor soil moisture retention and storage, bolstering vegetation WUE. In contrast, the central region of the Tibetan Plateau, dominated by grasslands, meadows, and shrubs, presented a comparatively low WUE. The arid climate and scarce water resources in this region force plants to consume substantial quantities of water to sustain their growth and metabolic processes, leading to a relative dip in WUE. Dominated by deserts and mountains, the northwestern part of the plateau also reflects lower WUE. The underlying reason is the combination of the arid climate and higher wind speeds that enhances the evaporation rate. Concurrently, plants in these areas, in adapting to the arid conditions, mitigate water loss by minimizing stomatal opening and leaf surface area, thereby resulting in a subsequent reduction in WUE.
Figure 3 Spatial distribution of water use efficiency on the Tibetan Plateau in 2001-2020

Full size|PPT slide

Of the different vegetation types, forests exhibited the highest WUE with an average value of 1.85 gC/mm∙m2 (Figure 4). Predominantly found in areas of lower elevation with relatively moist conditions, forests naturally possess heightened WUE. Moreover, the larger LAI and relatively slow ET rate of forests are key factors contributing to their superior WUE. Cultivated plants displayed a considerably high WUE, averaging 1.41 gC/mm∙m2, which likely can be attributed to the typical irrigation practices adopted for these plants that bolster their overall WUE. The WUE of swamp and shrubs was determined as 1.35 and 1.12 gC/mm∙m2, respectively. Swamp, characterized by abundant water resources, naturally sustains high WUE. Conversely, despite their smaller individual size, shrubs possess a high LAI that allows them to use water resources efficiently. Meadows and grasslands presented relatively low WUE, averaging 0.77 and 0.43 gC/mm∙m2, respectively. Such vegetation types are predominantly found in high-elevation regions characterized by arid conditions and scant rainfall. The combination of the lower LAI and faster ET rate curtails their ability to effectively harness water resources. Vegetation in alpine and desert areas exhibited the lowest WUE values, averaging 0.32 and 0.39 gC/mm∙m2, respectively. This can be linked to the challenging and adverse climatic conditions in such areas, which restrict vegetation growth and productivity, making it challenging for plants to maintain high WUE.
Figure 4 Distribution of mean water use efficiency for different vegetation types in 2001-2020

Full size|PPT slide

Analysis revealed gradual decline in WUE with increasing elevation (Figure 5a). Examination of the distribution of WUE across different elevations on the Tibetan Plateau (Figure 5b) indicated that areas with elevation of <3500 m, located primarily in southwestern, eastern, and northern regions of the plateau, predominantly harbor forests, shrubs, and meadow vegetation. The relatively low elevation of these regions means that they enjoy a milder climate with abundant rainfall, leading to higher vegetation WUE. The average WUE of such regions was determined as 1.68 gC/mm∙m2. Areas with elevation of 3500-4000 m mainly comprise alpine meadows and grasslands. The climate of such areas is generally drier and vegetation growth is somewhat stunted, resulting in reduced WUE, i.e., averaging 1.30 gC/ mm∙m2. The land use of areas with elevation of 4000-4500-m is primarily dominated by alpine grasslands. The WUE for most vegetation in such regions was determined as <1 gC/mm∙m2 (average: 0.85 gC/mm∙m2). Areas with elevation of 4500-5000 m were found to have an average WUE of 0.52 gC/mm∙m2, and those with elevation of 5000-5500 m registered an average WUE of 0.34 gC/mm∙m2. In areas with elevation of >5500 m, the mean WUE diminished to 0.23 gC/mm∙m2. The climatic conditions at the highest elevations are exceptionally harsh, rendering them nearly inhospitable for any form of vegetation growth, thereby resulting in the lowest WUE of the region.
Figure 5 Distribution of water use efficiency on the elevation gradient in 2001-2020

Full size|PPT slide

4.2 Temporal variation in water use efficiency

Analysis of the interannual fluctuations in WUE revealed a mean WUE for the plateau of 0.8088 gC/mm∙m2 (Figure 6). Years with a positive WUE anomaly (subsequently denoted as ΔWUE) primarily included 2015, 2016, 2017, 2018, and 2020; the most pronounced positive anomalies of 0.0840, 0.0789, and 0.0542 gC/mm∙m2 were in 2018, 2020, and 2015, respectively. Years with a negative ΔWUE predominantly included 2002, 2003, 2004, 2005, and 2008. Of these, 2008, 2004, and 2005 had the greatest negative ΔWUE values of −0.0637, −0.0623, and −0.0603 gC/mm∙m2, respectively. During 2001-2020, WUE increased at the rate of 0.0047 gC/mm∙m2∙a-1 (p<0.001). To better encapsulate the nonlinear and nonstationary trends in the WUE time series, results derived from the EEMD analysis indicated that 2005 was a turning point in the annual average WUE. A trend of decline in WUE was evident during 2001-2005, which subsequently shifted to a trend of increase after 2005.
Figure 6 Annual variation of water use efficiency in 2001-2020

Full size|PPT slide

Fluctuation analysis can enhance the comparability of the maximum fluctuation degree of time series variables over different spatial scales. To reflect the volatility of WUE on the Tibetan Plateau during 2001-2020, we used the CV as an indicator to quantitatively assess the degree of WUE variation. Upon processing the 2001-2020 WUE time series for the Tibetan Plateau, we obtained the spatial distribution of the CV (Figure 7a). Analysis revealed that the pixel-scale CV across the plateau during 2001-2020 predominantly ranged between 0.06 and 0.18, with a peak value at 0.1073. On the basis of the computed results, the CV values were categorized into five classes. The areas with variation coefficients within the ranges of 0≤CV<0.06, 0.06≤CV<0.12, 0.12≤CV<0.18, 0.18≤CV<0.24, and CV≥0.24 accounted for 1.31%, 23.05%, 29.01%, 19.64%, and 26.99% of the total area of the plateau, respectively. Notably, the range of 0.06≤CV<0.18 represents the largest proportion of fluctuation on the Tibetan Plateau, encompassing 52.06% of the area, mainly distributed in central and eastern regions of the plateau.
Figure 7 Fluctuation characteristics of water use efficiency in 2001-2020

Full size|PPT slide

Analysis of the fluctuation characteristics of WUE for various vegetation types on the Tibetan Plateau during 2001-2020 (Figure 7b) revealed distinct patterns of volatility across vegetation types. Alpine vegetation thrives in cold and snowy environments. The unique habitat conditions result in a short growing season and slow growth rate for alpine vegetation, causing pronounced fluctuations in its WUE. Swamp and meadow vegetation, with a relatively abundant water supply, showed less notable variations in their WUE. Grassland, forest, and shrubs presented similar CV values, suggesting that the magnitude of WUE fluctuation across these vegetation types is alike. This similarity might arise because these vegetation types grow under analogous climatic and topographical conditions; therefore, they are influenced by comparable environmental factors that cause their WUE to vary in a similar manner.

4.3 Trends in water use efficiency

Trend analysis offers comprehensive understanding of the spatiotemporal variations in WUE across the study area. Using a univariate linear regression method, we assessed the changes in WUE trend on the Tibetan Plateau over the past 20 years (Figure 8a). Findings indicated that the area of the Tibetan Plateau with increased WUE (Slope > 0) during 2001-2020 substantially surpassed that with diminished WUE (Slope<0). The region with increased WUE covered 1,417,696.64 km2, representing 62.97% of the total area, whereas the region with diminished WUE covered 836,077.50 km2, accounting for 37.03% of the entire plateau area. Statistical analysis of the linear trends in WUE for different vegetation types on the Tibetan Plateau during 2001-2020 (Figure 8b) revealed distinct trends among the different vegetation types. Notably, forest vegetation presented the most pronounced trend of increase, with an average Slope value of 0.0162, indicative of a higher rate of improvement in WUE for forests. However, there was also considerable fluctuation in the Slope value, likely attributable to various factors that influence the WUE of forests, e.g., climate change and anthropogenic activities. Cultivated plants and desert vegetation also both displayed notable upward trends in WUE, with mean Slope values of 0.0105 and 0.0072, respectively. Human interventions, such as irrigation and crop improvement, might have contributed to these trends. In contrast, shrubs and grasslands exhibited less distinct upward trends in WUE, recording average Slope values of 0.0040 and 0.0039, respectively. Alpine, swamp, and meadow vegetation types presented only marginal increases in WUE, with Slope values of 0.0014, 0.0012, and 0.0010, respectively.
Figure 8 Trend characteristics of water use efficiency slope in 2001-2020

Full size|PPT slide

Application of the MK test to verify the statistical significance of the Slope trend (Figures 8c and 8d) revealed that 23.11% of the region manifested a statistically significant WUE trend, whereas 76.89% of the region showed a non-significant trend. Areas with a notable trend of increase in WUE, constituting 18.43% of the total area, were predominantly located in southeastern, northeastern, and northwestern parts of the Tibetan Plateau. In contrast, areas with pronounced reduction in the trend of WUE, comprising 4.68% of the total area, were mainly in central and southwestern parts of the plateau.
Trends in time series are often nonlinear and nonstationary, and the use of a univariate linear regression approach can often mask underlying trend transitions. The EEMD method was employed in this study to conduct a nonlinear trend test on the WUE of the Tibetan Plateau for the period 2001-2020. In terms of long-term trends, the spatial distribution of the EEMD intrinsic trend in WUE is similar to that of the linear regression analysis, i.e., areas with increased and decreased WUE accounted for 50.73% and 49.63% of the total, respectively. Regions with WUE increase were predominantly located in southeastern and northwestern parts of the Tibetan Plateau, whereas areas with reduction in WUE were mainly in southwestern parts of the region (Figure 9a). Unlike linear regression analysis, the EEMD method not only detects the monotonic change in WUE but also identifies trend transitions from increase to decrease, and vice versa. Areas with a monotonic increase in WUE accounted for 23.64% of the total area, which is a reduction of 39.33% compared with that detected using the linear regression method. The areas with a monotonic decrease in WUE accounted for 9.69%. Trend shifts were identified in 66.67% of areas, with inflection points predominantly occurring between 2007 and 2012 (49.32%) (Figure 9b). Of these, 39.94% of areas (mainly in southwestern and northeastern regions of the plateau) experienced a shift from a trend of increase to a trend of decrease; conversely, 26.73% of areas (primarily in the southern part of the Tibetan Plateau) experienced a shift from a trend of decrease to a trend of increase.
Figure 9 Spatial distribution of water use efficiency trends (a) and the timing of change points (b)

Full size|PPT slide

Among the different vegetation types, the EEMD trends of vegetation WUE were found to vary (Table 1). Desert vegetation exhibited the highest proportion of a monotonic trend of increase at 44.32%. Forest vegetation had the highest proportion of a trend transitioning from decrease to increase, accounting for 48.07%. Grassland, cultivated plants, meadows, shrubs, swamp, and alpine vegetation predominantly displayed a trend of transition from increase to decrease with proportions of 40.14%, 33.28%, 46.20%, 36.69%, 48.33%, and 40.91%, respectively.
Table 1 Proportions of EEMD trends in water use efficiency for different vegetation types
Types In to In De to De De to In In to De
Desert 44.32 3.36 26.43 25.89
Grassland 32.74 7.44 19.68 40.14
Cultivation 29.29 8.93 28.49 33.28
Forest 21.10 8.59 48.07 22.23
Meadow 20.25 11.27 22.28 46.20
Shrubs 18.41 10.75 34.14 36.69
Swamp 17.65 10.12 23.90 48.33
Alpine 20.03 9.95 29.10 40.91

4.4 Factors influencing water use efficiency

LAI is a critical indicator of vegetation health and productivity, which can directly influence transpiration and photosynthesis. As a primary source of water input, precipitation is a key factor influencing soil moisture availability and, consequently, plant water uptake and WUE. Surface net solar radiation affects the energy available for photosynthesis and evapotranspiration, directly impacting vegetation growth and WUE. It is crucial for understanding how thermal conditions affect vegetation WUE. Temperature influences metabolic rates, photosynthesis, and evapotranspiration. Soil temperature impacts root activity, soil moisture dynamics, and microbial processes. Therefore, we chose these five factors to analyze the effect on WUE.
The LAI displayed a highly significant positive correlation with WUE (R2=0.74), followed by that of PRE, TMP, and STMP (Figure 10), which also manifested strong positive correlations with WUE (R2 values of 0.35, 0.31, and 0.29, respectively). Conversely, SSR demonstrated a significant negative relationship with WUE (R2=0.05).
Figure 10 Scatter plots between water use efficiency and environmental factors in 2001-2020

Full size|PPT slide

The Tibetan Plateau has complex topography with substantial variation in hydrothermal conditions that lead to pronounced differences in the response of vegetation to environmental factors (Figure 11). Analysis of the correlation between PRE and WUE during 2001-2020 revealed that areas with positive correlation occupied 32.60% of the study area, and that they were primarily situated in the northern parts of the Tibetan Plateau such as Hainan Tibetan Autonomous Prefecture and Huangnan Tibetan Autonomous Prefecture. In contrast, areas with negative correlation comprised 67.40% of the study region, and these were predominantly in areas such as Nagqu in the northern part of the plateau. Regions that exhibited positive correlation between SSR and WUE covered 71.94% of the study area, and were mainly in sparsely vegetated zones in the southwestern part of the plateau. Analysis of the relationship between LAI and WUE indicated that regions with positive correlation occupied 67.50% of the study area, and that these areas were mostly in northern parts such as Hainan Tibetan Autonomous Prefecture and Huangnan Tibetan Autonomous Prefecture. Conversely, regions with negative correlation accounted for 32.50% of the study area, and they were mainly in western regions such as Nagqu and Haixi Prefecture. Areas with positive correlation between STMP and WUE and between TMP and WUE accounted for 68.40% and 74.05% of the study area, respectively. In central parts of the Tibetan Plateau, where WUE values are relatively high, strong correlation was found between air temperature and WUE. As elevation increases, temperatures gradually decrease, making ground temperature one of the primary influencing factors of vegetation growth and ET.
Figure 11 Spatial distributions of the correlations between water use efficiency and environmental factors in 2001-2020

Full size|PPT slide

The RF model was used to analyze the spatial response of vegetation WUE to environmental factors (Figure 12). Discretizing the WUE and environmental factor data for 2001- 2020 into 20 gridded periods, with WUE serving as the dependent variable and environmental factors serving as the independent variables, RF regression was used to model WUE. This provided qualitative analysis of the spatial impact of environmental factors on WUE. For model building, 70% of the data were randomly selected, and the remaining 30% of the data were used for model accuracy validation. Within the RF model, the coefficient of determination between the simulated and actual data was in the range 0.5671-0.7646, with an average regression coefficient explanatory rate of 70.79%. The climatic factors and LAI effectively explained the spatial distribution differences in vegetation WUE, fulfilling the requirements for quantitative analysis of the driving factors behind the changes in vegetation WUE.
Figure 12 Scatter plots of the validation of the predicted values of random forest water use efficiency in 2001-2020

Full size|PPT slide

Determination of the importance of the variables derived using the RF model serves to highlight the contribution of predictor factors in governing WUE, with contributions subsequently normalized. Using the contribution rates of various climatic factors based on the RF model, we analyzed the impact of the climatic factors and LAI on the WUE of the Tibetan Plateau during 2001-2020 (Figure 13). Overall, LAI had the greatest influence on the spatial distribution of WUE in the study area, with an average contribution rate of 50.98%. It suggests that the leaf area of plants has notable impact on the change in WUE, which is potentially related to the area and number of leaves determining the photosynthetic capability and transpiration processes of a plant. The second-highest contribution was from TMP (average contribution rate: 20.61%), highlighting the pivotal role of temperature in regulating the physiological processes of plants and moisture evaporation. The average contribution rate of PRE was only 14.21%, possibly owing to the substantial spatiotemporal variations in precipitation on the Tibetan Plateau, making its influence on WUE less pronounced than that of other factors. Additionally, STMP and SSR had average contribution rates of 9.41% and 7.66%, respectively. It suggests that both soil temperature and solar radiation exert some influence on plant growth and water use, even though their contribution rates are relatively low.
Figure 13 The explanation degree of the stochastic forest regression model in the current year (a) and the importance of environmental factors (b)

Full size|PPT slide

5 Discussion

The findings of this study have significant implications for the Tibetan Plateau and the global community. On a regional level, understanding the drivers of WUE aids in formulating effective land and water management practices tailored to the unique environmental conditions of the Tibetan Plateau, crucial for sustaining its fragile ecosystems. Nationally, given the role of the Tibetan Plateau as a critical water source for major rivers in China, improving WUE through better management of vegetation can enhance water conservation efforts, benefiting downstream regions. Globally, enhancing WUE is a key strategy for mitigating the impacts of climate change. Efficient water use in vegetation can reduce water stress and improve carbon sequestration, contributing to global efforts to combat climate change.

5.1 Characteristics of spatiotemporal variation in water use efficiency

Between 2001 and 2020, the annual mean WUE of vegetation on the Tibetan Plateau was 0.8088 gC/mm∙m2, with a gradual trend of increase at the rate of 0.0047 gC/mm∙m2∙a-1. Influenced by the pattern of distribution of both water and heat (Zheng et al., 2020), the WUE of vegetation on the Tibetan Plateau exhibits a spatial pattern of reduction from the southeast toward the northwest. Wang (2020) indicated that WUE on the Tibetan Plateau increased slowly at the rate of 0.004 gC/mm∙m2∙a-1 during 1982-2015, suggesting an overarching upward trend in WUE over nearly half a century. However, this trend is not constant. The EEMD trend analysis in this study revealed initial decline in WUE followed by a gradual increase, with a turning point observed in 2005.
The WUE of different vegetation types on the Tibetan Plateau exhibited pronounced disparity, ranking from highest to lowest for forest, cultivated, marshland, shrub, meadow, grassland, alpine, and desert. Several other studies also found that forests and croplands typically exhibit higher WUE compared with that of grasslands(Chen and Yu, 2019; Bai et al., 2020; Veldkamp et al., 2023). The dense canopy structure of forests can lower surface temperatures, reduce wind speeds, and minimize evaporation (Gao et al., 2017; Liu et al., 2022), thereby creating conditions favorable for photosynthesis. Moreover, forest trees, with their deeper root systems, can access stored moisture from deeper soil strata (Schenk and Jackson, 2002; Pierret et al., 2016), ensuring continued growth even during periods of low precipitation. These attributes render forests more efficient and adaptable in terms of water use. Cultivated vegetation on the Tibetan Plateau predominantly consists of dryland crops such as barley, wheat, and maize, often irrigated using glacier meltwater or groundwater. These plants adopt efficient water-use strategies such as smaller leaves, deep root systems, and high stomatal densities to maximize their water use (Ai et al., 2020). Similarly, plateau marshland vegetation also adopts effective water-use strategies, including a high LAI and a deeper root system to optimize the absorption and use of moisture. Meadows and grasslands are mainly located in central and western regions of the Tibetan Plateau (Zhou et al., 2023), which are characterized by higher elevations and cold, dry climates. With relatively low vegetation cover and a low LAI (Wang et al., 2022), the photosynthetic capability of such vegetation is constrained. Although vegetation transpiration is minimal, elevated soil evaporation results in comparatively low WUE.
Vegetation distribution is intricately tied to elevational gradients (Piao et al., 2011). Elevation indirectly governs vegetation growth by influencing climatic variables such as temperature and precipitation (Liu et al., 2021a). Many preceding studies attributed the relationship between elevation and WUE to shifts in vegetation types (Zhu et al., 2015; Wang et al., 2023). With rising elevation, ecological alterations necessitate transitions from forests to shrubs and then to grasslands to adapt to cooler climates (He et al., 2018; Li et al., 2021).

5.2 Characteristics of the changing trend of variation in water use efficiency

The CV serves as a statistical measure to gauge the relative variability of a distribution, aptly facilitating comparison of datasets with differing means (Yan et al., 2015). This study found evident variability in WUE across different vegetation types on the Tibetan Plateau. The discrepancies in the degree of fluctuation in WUE among the various vegetation types might be attributed to their adaptability to local hydrothermal conditions. Additionally, diverse vegetation types might harbor distinct physiological mechanisms for modulating WUE, thereby influencing its volatility. Trends in WUE are jointly influenced by natural elements (e.g., the hydrothermal status) and anthropogenic factors, with pronounced differences in water and thermal conditions across years. Between 2001 and 2020, 62.97% of the area of the Tibetan Plateau demonstrated a gradual trend of increase in WUE. Nonetheless, certain regions (primarily in southwestern and central parts of the plateau) exhibited decline.
Traditional linear trend analysis, which can reveal the monotonic trend of a time series, overlooks the transitions in WUE on the Tibetan Plateau over the study period. The EEMD method, however, can detect changes in the trend of vegetation WUE, irrespective of whether from increase to decrease, or vice versa. In regions where the EEMD analysis identified monotonic trends, there was considerable overlap with areas for which the traditional methodologies pinpointed notable trends. This congruence underscores the robustness of our research findings. Results revealed that areas with a monotonic increase in WUE accounted for 23.64% of the total study area, a reduction of 39.33% compared with that detected using linear regression analysis, whereas areas with a monotonic decrease in WUE accounted for 9.69% of the total area. Remarkably, 66.67% of the area experienced a trend transition during the study period.
The decline in vegetation WUE primarily occurred in the Lhasa-Namtso Lake Basin in the southwestern Tibetan Plateau and the Three-River Headwaters Region in the central region, likely related to the shrubs expansion and desertification of the Tibetan Plateau (Brandt et al., 2013). The southwestern region of the Tibetan Plateau, characterized as a frost desertification degradation zone, predominantly encompasses the Northern Tibet Plateau and the Gangdise Mountain range. Owing to the considerable distance from the ocean and the blocking effect of the vast mountain ranges, the influence of warm and moist air currents from the southwestern Indian Ocean is substantially reduced, leading to a dry climate with intense radiation and relatively high environmental temperatures (Zhang et al., 2023). This harsh climatic setting results in a vulnerable ecosystem with low vegetation cover that is susceptible to external environmental changes. Over the past two decades, the region has predominantly experienced a warming and drying trend. Such a climatic shift can lead to thickening of the seasonal frost layer, surface soil desiccation, intensified frost weathering, and greater freeze-thaw action. This subsequently induces vegetation degradation, weakening the protective role of vegetation over soil. Such a scenario triggers a positive feedback loop that ultimately results in frost desertification. The decline in vegetation WUE in the Three-River Headwaters Region may be attributed to vegetation degradation in the central area (He et al., 2022). This degradation is primarily driven by anthropogenic factors such as population growth and overgrazing (Li et al., 2013; Mipam et al., 2019; Gu et al., 2023). Furthermore, the activities of certain small mammal species may also exacerbate local vegetation degradation (Miehe et al., 2019).

5.3 Factors influencing water use efficiency

We used a combination of correlation analysis and RF regression in this study to elucidate the spatiotemporal determinants of WUE. These complementary approaches allowed us to capture the linear relationships between WUE and its environmental determinants, as well as to identify and quantify the importance of these determinants in a non-linear, multi-dimensional context. Earlier studies identified that the LAI is an environmental factor of paramount importance for driving WUE variations (Hu et al., 2018; Cao et al., 2020). An elevated LAI not only augments photosynthesis but also influences WUE by affecting both interception of solar radiation by plants and soil evaporation (Beer et al., 2009; Huang et al., 2015). Temporally, a statistically significant correlation was observed between the LAI and meadow vegetation in northeastern and central parts of the Tibetan Plateau. The RF model further accentuated the importance of the LAI as a pivotal environmental element shaping the spatial pattern of WUE. Nonetheless, in specific scenarios where leaves with lower carbon assimilation predominate, an increase in the LAI could diminish WUE (Wang et al., 2018). Zhao (2022) suggested that the relationship between the LAI and WUE is not plainly linear, potentially explaining the subdued correlation between the LAI and forest vegetation in the southern Tibetan Plateau.
Our temporal model depicted a positive precipitation response across most of the Tibetan Plateau, indicating that increased rainfall generally promotes vegetation growth, particularly in the more arid and semi-arid regions where water availability is a primary constraint (Sun et al., 2019). However, in the southwestern zones, where vegetation degradation is prevalent, the response to precipitation was negative. This anomaly can be attributed to several factors, including soil erosion, overgrazing, and the cumulative effects of long-term droughts, which have degraded the soil structure and reduced its capacity to retain moisture (Lin et al., 2023). The recent warming of the climate has significantly altered the ecological dynamics of the Tibetan Plateau, leading to varied and complex responses in vegetation across the region (Zhong et al., 2019). In regions of the central plateau with extensive vegetation cover, strong positive correlation between temperature and vegetation was noted, potentially attributable to the heightened sensitivity of alpine shrubs to temperature (Martin et al., 2017). The shrubs, adapted to cold climates, benefit from moderate warming, which can extend the growing season and enhance photosynthetic activity, thereby improving WUE (Körner, 2021). As elevation increases, the importance of soil temperature becomes more pronounced (Yan et al., 2023). Central and eastern parts of the Tibetan Plateau primarily feature sparse grasslands and alpine meadows, where plant root systems predominantly inhabit the upper 10 cm of soil, thereby rendering them directly impacted by soil temperature. Elevated soil temperatures expedite the thawing of the permafrost that is widespread throughout the region (Gu et al., 2005), indirectly fostering vegetation growth. Yang identified soil temperature as the primary determinant affecting vegetation growth in the source regions of the Yellow and Yangtze rivers (Yang et al., 2006). Xu (2011) asserted that soil temperature might confer more consequential effects on vegetation variation and directly influence root growth, nutrient uptake, and microbial activity, all of which are critical for sustaining vegetation in high-altitude ecosystems, aligning more closely with the conclusions of this study. In arid and semiarid regions, soil moisture stands as the key factor restricting vegetation growth and transpiration. Higher solar radiation can promote vegetation transpiration, thereby enhancing WUE. In humid settings, the influence of solar radiation on vegetation WUE is comparatively trivial, with water not serving as the principal growth constraint.
The RF model exhibited superior fitting performance in areas with lower WUE values, but it was less effective in regions with higher WUE. This might arise from the greater sample density in areas with lower WUE and comparative scarcity of samples in high-elevation regions, causing the model to accentuate the features of the former while insufficiently comprehending the latter, resulting in underestimations. Integration of the climatic parameters and LAI into the model might not adequately encapsulate WUE. In areas of higher values, additional factors such as soil moisture distribution and plant root structure, which remained unconsidered, could have contributed to the predictive underestimation.

6 Conclusions

During the research period, the WUE of vegetation on the Tibetan Plateau exhibited an average value of 0.8088 gC/mm∙m2. Spatially, a clear gradient was observed, declining from the southeast toward the northwest, closely mirroring the variation in elevation. Annually, a statistically significant inflection in WUE growth was detected around 2005, showing an overall gradual upward trend. Diverse vegetation types exhibited notable differences in their WUE; forests recorded the highest values, while alpine vegetation registered the lowest. The observed trend in WUE was not monotonous, i.e., a shift was noticed in 66.67% of the regions under study. The LAI and temperature both demonstrated strong correlation with WUE. Furthermore, the RF model indicated that environmental factors could explain approximately 70.79% of the variations in WUE, with the LAI and temperature emerging as the primary drivers of spatial distribution of WUE. The observed fluctuations in WUE provide invaluable insights into the responses of terrestrial ecosystems under changing climatic scenarios and anthropogenic pressures.

References

[1]
Ai Z P, Wang Q X, Yang Y H et al., 2020. Variation of gross primary production, evapotranspiration and water use efficiency for global croplands. Agricultural and Forest Meteorology, 287: 107935.
[2]
Assal T J, Steen V A, Caltrider T et al., 2021. Monitoring long-term riparian vegetation trends to inform local habitat management in a mountainous environment. Ecological Indicators, 127: 107807.
[3]
Bai Y J, Zha T S, Bourque C P et al., 2020. Variation in ecosystem water use efficiency along a southwest-to-northeast aridity gradient in China. Ecological Indicators, 110: 105932.
[4]
Barbour M M, 2007. Stable oxygen isotope composition of plant tissue: A review. Functional Plant Biology, 34(2): 83.
With the development of rapid measurement techniques, stable oxygen isotope analysis of plant tissue is poised to become an important tool in plant physiological, ecological, paleoclimatic and forensic studies. Recent advances in mechanistic understanding have led to the improvement of process-based models that accurately predict variability in the oxygen isotope composition of plant organic material (δO). δO has been shown to reflect the isotope composition of soil water, evaporative enrichment in transpiring leaves, and isotopic exchange between oxygen atoms in organic molecules and local water in the cells in which organic molecules are formed. This review presents current theoretical models describing the influences on δO, using recently published experimental work to outline strengths and weaknesses in the models. The potential and realised applications of the technique are described.
[5]
Beer C, Ciais P, Reichstein M et al., 2009. Temporal and among-site variability of inherent water use efficiency at the ecosystem level. Global Biogeochemical Cycles, 23(2).
[6]
Bonan G B, 2008. Forests and climate change: Forcings, feedbacks, and the climate benefits of forests. Science, 320(5882): 1444-1449.
The world's forests influence climate through physical, chemical, and biological processes that affect planetary energetics, the hydrologic cycle, and atmospheric composition. These complex and nonlinear forest-atmosphere interactions can dampen or amplify anthropogenic climate change. Tropical, temperate, and boreal reforestation and afforestation attenuate global warming through carbon sequestration. Biogeophysical feedbacks can enhance or diminish this negative climate forcing. Tropical forests mitigate warming through evaporative cooling, but the low albedo of boreal forests is a positive climate forcing. The evaporative effect of temperate forests is unclear. The net climate forcing from these and other processes is not known. Forests are under tremendous pressure from global change. Interdisciplinary science that integrates knowledge of the many interacting climate services of forests with the impacts of global change is necessary to identify and understand as yet unexplored feedbacks in the Earth system and the potential of forests to mitigate climate change.
[7]
Brandt J S, Haynes M A, Kuemmerle T et al., 2013. Regime shift on the roof of the world: Alpine meadows converting to shrublands in the southern Himalayas. Biological Conservation, 158: 116-127.
[8]
Breiman L, 2001. Random forests. Machine Learning, 45(1): 5-32.
[9]
Cao R C, Hu Z M, Jiang Z Y et al., 2020. Shifts in ecosystem water use efficiency on China’s Loess Plateau caused by the interaction of climatic and biotic factors over 1985-2015. Agricultural and Forest Meteorology, 291: 108100.
[10]
Chen J H, Yan F, Lu Q, 2020. Spatiotemporal variation of vegetation on the Qinghai-Tibet Plateau and the influence of climatic factors and human activities on vegetation trend (2000-2019). Remote Sensing, 12(19): 3150.
[11]
Chen Z, Yu G R, 2019. Spatial variations and controls of carbon use efficiency in China’s terrestrial ecosystems. Scientific Reports, 9(1): 19516.
Carbon use efficiency (CUE), one of the most important eco-physiological parameters, represents the capacity of plants to transform carbon into new biomass. Understanding the variations and controls of CUE is crucial for regional carbon assessment. Here, we used 15-years of continuous remote sensing data to examine the variations of CUE across broad geographic and climatic gradients in China. The results showed that the vegetation CUE was averaged to 0.54 ± 0.11 with minor interannual variation. However, the CUE greatly varied with geographic gradients and ecosystem types. Forests have a lower CUE than grasslands and croplands. Evergreen needleleaf forests have a higher CUE than other forest types. Climate factors (mean annual temperature (MAT), precipitation (MAP) and the index of water availability (IWA)) dominantly regulated the spatial variations of CUE. The CUE exhibited a linear decrease with enhanced MAT and MAP and a parabolic response to the IWA. Furthermore, the responses of CUE to environmental change varied with individual ecosystem type. In contrast, precipitation exerted strong control on CUE in grassland, while in forest and cropland, the CUE was mainly controlled by the available water. This study identifies the variations and response of CUE to environmental drivers in China, which will be valuable for the regional assessment of carbon cycling dynamics under future climate change.
[12]
Copernicus Climate Change Service, 2019. ERA5-Land monthly averaged data from 2001 to present. ECMWF(2019) [2023-08-12].
[13]
Dawson T E, Mambelli S, Plamboeck A H et al., 2002. Stable isotopes in plant ecology. Annual Review of Ecology and Systematics, 33(1): 507-559.
[14]
Donovan L A, Dudley S A, Rosenthal D M et al., 2007. Phenotypic selection on leaf water use efficiency and related ecophysiological traits for natural populations of desert sunflowers. Oecologia, 152(1): 13-25.
Plant water-use efficiency (WUE) is expected to affect plant fitness and thus be under natural selection in arid habitats. Although many natural population studies have assessed plant WUE, only a few related WUE to fitness. The further determination of whether selection on WUE is direct or indirect through functionally related traits has yielded no consistent results. For natural populations of two desert annual sunflowers, Helianthus anomalus and H. deserticola, we used phenotypic selection analysis with vegetative biomass as the proxy for fitness to test (1) whether there was direct and indirect selection on WUE (carbon isotope ratio) and related traits (leaf N, area, succulence) and (2) whether direct selection was consistent with hypothesized drought/dehydration escape and avoidance strategies. There was direct selection for lower WUE in mesic and dry H. anomalus populations, consistent with dehydration escape, even though it is the longer lived of the two species. For mesic H. anomalus, direct selection favored lower WUE and higher N, suggesting that plants may be "wasting water" to increase N delivery via the transpiration stream. For the shorter lived H. deserticola in the direr habitat, there was indirect selection for lower WUE, inconsistent with drought escape. There was also direct selection for higher leaf N, succulence and leaf size. There was no direct selection for higher WUE consistent with dehydration avoidance in either species. Thus, in these natural populations of two desert dune species higher fitness was associated with some combination direct and indirect selection for lower WUE, higher leaf N and larger leaf size. Our understanding of the adaptive value of plant ecophysiological traits will benefit from further consideration of related traits such as leaf nitrogen and more tests in natural populations.
[15]
Du X Z, Zhao X, Zhou T et al., 2019. Effects of climate factors and human activities on the ecosystem water use efficiency throughout northern China. Remote Sensing, 11(23): 2766.
[16]
Flanagan L B, Syed K H, 2011. Stimulation of both photosynthesis and respiration in response to warmer and drier conditions in a boreal peatland ecosystem: Peatland carbon dioxide exchange. Global Change Biology, 17(7): 2271-2287.
[17]
Franke A K, Feilhauer H, Bräuning A et al., 2019. Remotely sensed estimation of vegetation shifts in the polar and alpine tree-line ecotone in Finnish Lapland during the last three decades. Forest Ecology and Management, 454: 117668.
[18]
Gao J B, Jiao K W, Wu S H, 2019. Investigating the spatially heterogeneous relationships between climate factors and NDVI in China during 1982 to 2013. Journal of Geographical Sciences, 29(10): 1597-1609.

Climate change is a major driver of vegetation activity, and thus their complex processes become a frontier and difficulty in global change research. To understand this relationship between climate change and vegetation activity, the spatial distribution and dynamic characteristics of the response of NDVI to climate change were investigated by the geographically weighted regression (GWR) model during 1982 to 2013 in China. This model was run based on the combined datasets of satellite vegetation index (NDVI) and climate observation (temperature and moisture) from meteorological stations nationwide. The results showed that the spatial non-stationary relationship between NDVI and surface temperature has appeared in China: the significant negative temperature-vegetation relationship was located in Northeast, Northwest and Southeast China, while the positive correlation was more concentrated from southwest to northeast. By comparing the normalized regression coefficients from GWR model for different climate factors, it presented the regions with moisture dominants for NDVI were in North China and the Tibetan Plateau, and the areas of temperature dominants were distributed in East, Central and Southwest China, where the annual mean maximum temperature accounted for the largest areas. In addition, regression coefficients from GWR model between NDVI dynamics and climate variability indicated that the higher warming rate could result in the weakened vegetation activity through some mechanisms such as enhanced drought, while the moisture variability could mediate the hydrothermal conditions for the variation of vegetation activity. When the increasing rate of photosynthesis exceeded that of respiration, the positive correlation between vegetation dynamics and climate variability was reflected. However, the continuous and dynamic process of vegetation activity response to climate change will be determined by spatially heterogeneous conditions in climate change and vegetation cover. Furthermore, the dynamic description of climate-induced vegetation activity from its rise to decline in different regions is expected to provide a scientific basis for initiating ecosystem-based adaptation strategies in response to global climate change.

[19]
Gao M D, Wang X H, Meng F D et al., 2020. Three-dimensional change in temperature sensitivity of northern vegetation phenology. Global Change Biology, 26(9): 5189-5201.
[20]
Gao Y, Markkanen T, Aurela M et al., 2017. Response of water use efficiency to summer drought in a boreal scots pine forest in Finland. Biogeosciences, 14(18): 4409-4422.
[21]
Gu C J, Liu L S, Zhang Y L et al., 2023. Understanding the spatial heterogeneity of grazing pressure in the Three-River-Source Region on the Tibetan Plateau. Journal of Geographical Sciences, 33(8): 1660-1680.

Elucidating the distribution of the grazing pressure requires an understanding of the grazing activities. In this study, we analyzed the grazing behavior of yaks in Three-River- Source Region (TRSR) and identified the main factors influencing the distribution of grazing intensity (GI) using trajectory data and remote sensing datasets. Our results revealed that a semi-resident transhumance strategy is employed in this region. The average grazing time (GT) of four GPS collars over the year was 11.84 h/day (N6), 11.01 h/day (N11), 9.25 h/day (N18), and 11.61 h/day (N24). GT was generally higher in warm seasons (summer and autumn) than in cold seasons (spring and winter). The average daily moving speed was found to be closely related to the pasture size of different herders and the seasons. Geodetector analysis identified the distance to camp (DOC) as the most important single factor influencing the distribution of GI, explaining up to 52% of the GI variations. However, relying solely on this factor may not accurately depict the actual GI distribution. When pairwise factors interacted, the explanatory power of the model increased, ranging from 34.55% to 63.26%. Our study highlights the importance of considering multiple factors when predicting grazing intensity, as grazing activities tend to cluster near settlements, but other factors may also be influential.

[22]
Gu S, Tang Y H, Cui X Y et al., 2005. Energy exchange between the atmosphere and a meadow ecosystem on the Qinghai-Tibetan Plateau. Agricultural and Forest Meteorology, 129(3-4): 175-185.
[23]
Gunderson C A, Edwards N T, Walker A V et al., 2012. Forest phenology and a warmer climate-growing season extension in relation to climatic provenance. Global Change Biology, 18(6): 2008-2025.
[24]
Gunderson C A, Norby R J, Wullschleger S D, 2000. Acclimation of photosynthesis and respiration to simulated climatic warming in northern and southern populations of Acer saccharum: Laboratory and field evidence. Tree Physiology, 20(2): 87-96.
Physiological acclimation and genotypic adaptation to prevailing temperatures may influence forest responses to future climatic warming. We examined photosynthetic and respiratory responses of sugar maple (Acer saccharum Marsh.) from two portions of the species' range for evidence of both phenomena in a laboratory study with seedlings. A field study was also conducted to assess the impacts of temperature acclimation on saplings subjected to an imposed temperature manipulation (4 degrees C above ambient temperature). The two seedling populations exhibited more evidence of physiological acclimation to warming than of ecotypic adaptation, although respiration was less sensitive to short-term warming in the southern population than in the northern population. In both seedling populations, thermal compensation increased photosynthesis by 14% and decreased respiration by 10% in the warm-acclimated groups. Saplings growing in open-top field chambers at ambient temperature and 4 degrees C above ambient temperature showed evidence of temperature acclimation, but photosynthesis did not increase in response to the 4 degrees C warming. On the contrary, photosynthetic rates measured at the prevailing chamber temperature throughout three growing seasons were similar, or lower (12% lower on average) in saplings maintained at 4 degrees C above ambient temperature compared with saplings maintained at ambient temperature. However, the long-term photosynthetic temperature optimum for saplings in the field experiment was higher than it was for seedlings in either the 27 or the 31 degrees C growth chamber. Respiratory acclimation was also evident in the saplings in the field chambers. Saplings had similar rates of respiration in both temperature treatments, and respiration showed little dependence on prevailing temperature during the growing season. We conclude that photosynthesis and respiration in sugar maple have the potential for physiological acclimation to temperature, but exhibit a low degree of genetic adaptation. Some of the potential for acclimation to a 4 degrees C increase above a background of naturally fluctuating temperatures may be offset by differences in water relations, and, in the long term, may be obscured by the inherent variability in rates under field conditions. Nevertheless, physiologically based models should incorporate seasonal acclimation to temperature and permit ecotypic differences to influence model outcomes for those species with high genetic differentiation between regions.
[25]
He C Y, Yan F, Wang Y J et al., 2022a. Spatiotemporal variation in vegetation growth status and its response to climate in the Three-River Headwaters Region, China. Remote Sensing, 14(19): 5041.
[26]
He S Y, Zhang Y Q, Ma N et al., 2022b. A daily and 500 m coupled evapotranspiration and gross primary production product across China during 2000-2020. Earth System Science Data, 14(12): 5463-5488.
[27]
He Z B, Du J, Chen L F et al., 2018. Impacts of recent climate extremes on spring phenology in arid-mountain ecosystems in China. Agricultural and Forest Meteorology, 260/261: 31-40.
[28]
Hu Y R, Maskey S, Uhlenbrook S, 2012. Trends in temperature and rainfall extremes in the Yellow River source region, China. Climatic Change, 110(1/2): 403-429.
[29]
Hu Z M, Yu G R, Fu Y L et al., 2018. Effects of vegetation control on ecosystem water use efficiency within and among four grassland ecosystems in China. New Phytologist, 123(8): 1609-1619.
[30]
Huang M T, Piao S L, Sun Y et al., 2015. Change in terrestrial ecosystem water-use efficiency over the last three decades. Global Change Biology, 21(6): 2366-2378.
Defined as the ratio between gross primary productivity (GPP) and evapotranspiration (ET), ecosystem-scale water-use efficiency (EWUE) is an indicator of the adjustment of vegetation photosynthesis to water loss. The processes controlling EWUE are complex and reflect both a slow evolution of plants and plant communities as well as fast adjustments of ecosystem functioning to changes of limiting resources. In this study, we investigated EWUE trends from 1982 to 2008 using data-driven models derived from satellite observations and process-oriented carbon cycle models. Our findings suggest positive EWUE trends of 0.0056, 0.0007 and 0.0001 g C m(-2)  mm(-1)  yr(-1) under the single effect of rising CO2 ('CO2 '), climate change ('CLIM') and nitrogen deposition ('NDEP'), respectively. Global patterns of EWUE trends under different scenarios suggest that (i) EWUE-CO2 shows global increases, (ii) EWUE-CLIM increases in mainly high latitudes and decreases at middle and low latitudes, (iii) EWUE-NDEP displays slight increasing trends except in west Siberia, eastern Europe, parts of North America and central Amazonia. The data-driven MTE model, however, shows a slight decline of EWUE during the same period (-0.0005 g C m(-2)  mm(-1)  yr(-1) ), which differs from process-model (0.0064 g C m(-2)  mm(-1)  yr(-1) ) simulations with all drivers taken into account. We attribute this discrepancy to the fact that the nonmodeled physiological effects of elevated CO2 reducing stomatal conductance and transpiration (TR) in the MTE model. Partial correlation analysis between EWUE and climate drivers shows similar responses to climatic variables with the data-driven model and the process-oriented models across different ecosystems. Change in water-use efficiency defined from transpiration-based WUEt (GPP/TR) and inherent water-use efficiency (IWUEt, GPP×VPD/TR) in response to rising CO2, climate change, and nitrogen deposition are also discussed. Our analyses will facilitate mechanistic understanding of the carbon-water interactions over terrestrial ecosystems under global change. © 2015 John Wiley & Sons Ltd.
[31]
Ji F, Wu Z H, Huang J P et al., 2014. Evolution of land surface air temperature trend. Nature Climate Change, 4(6): 462-466.
[32]
Ji Y D, Li Y, Yao N et al., 2021. The lagged effect and impact of soil moisture drought on terrestrial ecosystem water use efficiency. Ecological Indicators, 133: 108349.
[33]
Kendall M G, 1948. Rank correlation methods. Griffin.
[34]
Körner C, 2021. Alpine Plant Life: Functional Plant Ecology of High Mountain Ecosystems. Cham: Springer International Publishing [2024-08-13].
[35]
Körner C, Basler D, 2010. Phenology under global warming. Science, 327(5972): 1461-1462.
[36]
Li G C, Chen W, Li R R et al., 2021. Assessing the spatiotemporal dynamics of ecosystem water use efficiency across China and the response to natural and human activities. Ecological Indicators, 126: 107680.
[37]
Li X, Wu P F, Guo F T et al., 2021. A geographically weighted regression approach to detect divergent changes in the vegetation activity along the elevation gradients over the last 20 years. Forest Ecology and Management, 490: 119089.
[38]
Li X L, Gao J, Brierley G et al., 2013. Rangeland degradation on the Qinghai-Tibet Plateau: Implications for rehabilitation. Land Degradation & Development, 24(1): 72-80.
[39]
Li Y, Shi H, Zhou L et al., 2018. Disentangling climate and lai effects on seasonal variability in water use efficiency across terrestrial ecosystems in China. New Phytologist, 123(8): 2429-2443.
[40]
Lian X H, Jiao L M, Liu Z J et al., 2022. Multi-spatiotemporal heterogeneous legacy effects of climate on terrestrial vegetation dynamics in China. GIScience & Remote Sensing, 59(1): 164-183.
[41]
Lin H H, Cheng X L, Bruijnzeel L A et al., 2023. Land degradation and climate change lessened soil erodibility across a wide area of the southern Tibetan Plateau over the past 35-40 years. Land Degradation & Development, 34(9): 2636-2651.
[42]
Liu C L, Li W L, Wang W Y et al., 2021a. Quantitative spatial analysis of vegetation dynamics and potential driving factors in a typical alpine region on the northeastern Tibetan Plateau using the Google Earth Engine. CATENA, 206: 105500.
[43]
Liu Y B, Xiao J F, Ju W M et al., 2015. Water use efficiency of China’s terrestrial ecosystems and responses to drought. Scientific Reports, 5(1): 13799.
[44]
Liu Y C, Li Z, Chen Y N et al., 2021b. Evaluation of consistency among three NDVI products applied to High Mountain Asia in 2000-2015. Remote Sensing of Environment, 269: 112821.
[45]
Liu Z Q, Ji X F, Ye L M et al., 2022. Inherent water-use efficiency of different forest ecosystems and its relations to climatic variables: Forests, 13(5): 775.
[46]
Lu X H, Ju W M, Jiang H et al., 2019. Effects of nitrogen deposition on water use efficiency of global terrestrial ecosystems simulated using the IBIS model. Ecological Indicators, 101: 954-962.
[47]
Ma Y R, Guan Q Y, Sun Y F et al., 2022. Three-dimensional dynamic characteristics of vegetation and its response to climatic factors in the Qilian Mountains. CATENA, 208: 105694.
[48]
Mann H B, 1945. Nonparametric tests against trend. Econometrica, 13(3): 245.
[49]
Martin A C, Jeffers E S, Petrokofsky G et al., 2017. Shrub growth and expansion in the arctic tundra: An assessment of controlling factors using an evidence-based approach. Environmental Research Letters, 12(8): 085007.
[50]
Miehe G, Schleuss P M, Seeber E et al., 2019. The Kobresia pygmaea ecosystem of the Tibetan highlands: Origin, functioning and degradation of the world’s largest pastoral alpine ecosystem. Science of The Total Environment, 648: 754-771.
[51]
Mipam T D, Zhong L L, Liu J Q et al., 2019. Productive overcompensation of alpine meadows in response to yak grazing in the eastern Qinghai-Tibet Plateau. Frontiers in Plant Science, 10: 925.
[52]
Mo X, Liu S, Lin Z et al., 2005. Prediction of crop yield, water consumption and water use efficiency with a SVAT-crop growth model using remotely sensed data on the North China Plain. Ecological Modelling, 183(2/3): 301-322.
[53]
Nemani R R, Keeling C D, Hashimoto H et al., 2003. Climate-driven increases in global terrestrial net primary production from 1982 to 1999. Science, 300(5625): 1560-1563.
Recent climatic changes have enhanced plant growth in northern mid-latitudes and high latitudes. However, a comprehensive analysis of the impact of global climatic changes on vegetation productivity has not before been expressed in the context of variable limiting factors to plant growth. We present a global investigation of vegetation responses to climatic changes by analyzing 18 years (1982 to 1999) of both climatic data and satellite observations of vegetation activity. Our results indicate that global changes in climate have eased several critical climatic constraints to plant growth, such that net primary production increased 6% (3.4 petagrams of carbon over 18 years) globally. The largest increase was in tropical ecosystems. Amazon rain forests accounted for 42% of the global increase in net primary production, owing mainly to decreased cloud cover and the resulting increase in solar radiation.
[54]
Niu S L, Xing X R, Zhang Z et al., 2011. Water-use efficiency in response to climate change: From leaf to ecosystem in a temperate steppe. Global Change Biology, 17(2): 1073-1082.
[55]
Pan N Q, Feng X M, Fu B J et al., 2018. Increasing global vegetation browning hidden in overall vegetation greening: Insights from time-varying trends. Remote Sensing of Environment, 214: 59-72.
[56]
Pearson K, 1896. IV. Contributions to the mathematical theory of evolution: III. Regression, heredity, and panmixia. Proceedings of the Royal Society of London, 59(353-358): 69-71.
[57]
Peng J, Liu Z H, Liu Y H et al., 2012. Trend analysis of vegetation dynamics in Qinghai-Tibet Plateau using hurst exponent. Ecological Indicators, 14(1): 28-39.
[58]
Peng S Z, Ding Y X, Liu W Z et al., 2019. 1 km monthly temperature and precipitation dataset for China from 1901 to 2017. Earth System Science Data, 11(4): 1931-1946.
[59]
Peng S Z, Ding Y X, Wen Z M et al., 2017. Spatiotemporal change and trend analysis of potential evapotranspiration over the Loess Plateau of China during 2011-2100. Agricultural and Forest Meteorology, 233: 183-194.
[60]
Piao S L, Cui M D, Chen A P et al., 2011. Altitude and temperature dependence of change in the spring vegetation green-up date from 1982 to 2006 in the Qinghai-Xizang Plateau. Agricultural and Forest Meteorology, 151(12): 1599-1608.
[61]
Pierret A, Maeght J L, Clément C et al., 2016. Understanding deep roots and their functions in ecosystems: An advocacy for more unconventional research. Annals of Botany, 118(4): 621-635.
Background Deep roots are a common trait among a wide range of plant species and biomes, and are pivotal to the very existence of ecosystem services such as pedogenesis, groundwater and streamflow regulation, soil carbon sequestration and moisture content in the lower troposphere. Notwithstanding the growing realization of the functional significance of deep roots across disciplines such as soil science, agronomy, hydrology, ecophysiology or climatology, research efforts allocated to the study of deep roots remain incommensurate with those devoted to shallow roots. This is due in part to the fact that, despite technological advances, observing and measuring deep roots remains challenging. Scope Here, other reasons that explain why there are still so many fundamental unresolved questions related to deep roots are discussed. These include the fact that a number of hypotheses and models that are widely considered as verified and sufficiently robust are only partly supported by data. Evidence has accumulated that deep rooting could be a more widespread and important trait among plants than usually considered based on the share of biomass that it represents. Examples that indicate that plant roots have different structures and play different roles with respect to major biochemical cycles depending on their position within the soil profile are also examined and discussed. Conclusions Current knowledge gaps are identified and new lines of research for improving our understanding of the processes that drive deep root growth and functioning are proposed. This ultimately leads to a reflection on an alternative paradigm that could be used in the future as a unifying framework to describe and analyse deep rooting. Despite the many hurdles that pave the way to a practical understanding of deep rooting functions, it is anticipated that, in the relatively near future, increased knowledge about the deep rooting traits of a variety of plants and crops will have direct and tangible influence on how we manage natural and cultivated ecosystems.© The Author 2016. Published by Oxford University Press on behalf of the Annals of Botany Company. All rights reserved. For Permissions, please email: journals.permissions@oup.com.
[62]
Puma M J, Koster R D, Cook B I, 2013. Phenological versus meteorological controls on land-atmosphere water and carbon fluxes. Journal of Geophysical Research: Biogeosciences, 118(1): 14-29.
[63]
Ren S Y, Jing H C, Qian X X et al., 2024. Spatial and temporal coordinated development research on ecosystem services and human well-being in the typical pastoral area of the Qinghai-Tibet Plateau. Journal of Geographical Sciences, 34(2): 252-288.

In this study, the interplay between ecosystem services and human well-being in Seni district, which is a pastoral region of Nagqu city on the Qinghai-Tibet Plateau, is investigated. Employing the improved InVEST model, CASA model, coupling coordination model, and hierarchical clustering method, we analyze the spatiotemporal patterns of ecosystem services, the levels of resident well-being levels, and the interrelationships between these factors over the period from 2000 to 2018. Our findings reveal significant changes in six ecosystem services, with water production decreasing by 7.1% and carbon sequestration and soil conservation services increasing by approximately 6.3% and 14.6%, respectively. Both the habitat quality and landscape recreation services remained stable. Spatially, the towns in the eastern and southern areas exhibited higher water production and soil conservation services, while those in the central area exhibited greater carbon sequestration services. The coupling and coordination relationship between ecosystem services and human well-being improved significantly over the study period, evolving from low-level coupling to coordinated coupling. Hierarchical clustering was used to classify the 12 town-level units into five categories. Low subjective well-being townships had lower livestock breeding services, while high subjective well-being townships had higher supply, regulation, and support ecosystem services. Good transportation conditions were associated with higher subjective well-being in townships with low supply services. We recommend addressing the identified transportation disparities and enhancing key regulatory and livestock breeding services to promote regional sustainability and improve the quality of life for Seni district residents, thus catering to the diverse needs of both herdsmen and citizens.

[64]
Schenk H J, Jackson R B, 2002. Rooting depths, lateral root spreads and below-ground/above-ground allometries of plants in water-limited ecosystems. Journal of Ecology, 90(3): 480-494.
[65]
Sun S B, Song Z L, Wu X C et al., 2018. Spatio-temporal variations in water use efficiency and its drivers in China over the last three decades. Ecological Indicators, 94: 292-304.
[66]
Sun W C, Wang Y Y, Fu Y H et al., 2019. Spatial heterogeneity of changes in vegetation growth and their driving forces based on satellite observations of the Yarlung Zangbo River Basin in the Tibetan Plateau. Journal of Hydrology, 574: 324-332.
[67]
Sun W Y, Song X Y, Mu X M et al., 2015. Spatiotemporal vegetation cover variations associated with climate change and ecological restoration in the Loess Plateau. Agricultural and Forest Meteorology, 209/210: 87-99.
[68]
Sun Y, Piao S L, Huang M T et al., 2016. Global patterns and climate drivers of water-use efficiency in terrestrial ecosystems deduced from satellite-based datasets and carbon cycle models. Global Ecology and Biogeography, 25(3): 311-323.
[69]
Tang X G, Li H P, Desai A R et al., 2015. How is water-use efficiency of terrestrial ecosystems distributed and changing on earth? Scientific Reports, 4(1): 7483.
[70]
Tong K, Su F G, Yang D Q et al., 2014. Evaluation of satellite precipitation retrievals and their potential utilities in hydrologic modeling over the Tibetan Plateau. Journal of Hydrology, 519: 423-437.
[71]
Veldkamp E, Schmidt M, Markwitz C et al., 2023. Multifunctionality of temperate alley-cropping agroforestry outperforms open cropland and grassland. Communications Earth & Environment, 4(1): 20.
[72]
Wang L M, Li M Y, Wang J X et al., 2020. An analytical reductionist framework to separate the effects of climate change and human activities on variation in water use efficiency. Science of The Total Environment, 727: 138306.
[73]
Wang L X, Good S P, Caylor K K, 2018. Global synthesis of vegetation control on evapotranspiration partitioning. New Phytologist, 123(8): 6753-6757.
[74]
Wang X, Chen G, Wu M Q et al., 2023. Differences in the patterns and mechanisms of leaf and ecosystem-scale water use efficiencies on the Qinghai-Tibet Plateau. CATENA, 222: 106874.
[75]
Wang Y F, Lv W W, Xue K et al., 2022. Grassland changes and adaptive management on the Qinghai-Tibetan Plateau. Nature Reviews Earth & Environment, 3(10): 668-683.
[76]
Wu Z H, Huang N E, 2009. Ensemble empirical mode decomposition: A noise-assisted data analysis method. Advances in Adaptive Data Analysis, 1(1): 1-41.
[77]
Xu W X, Gu S, Zhao X Q et al., 2011. High positive correlation between soil temperature and NDVI from 1982 to 2006 in alpine meadow of the Three-River Source Region on the Qinghai-Tibetan Plateau. International Journal of Applied Earth Observation and Geoinformation, 13(4): 528-535.
[78]
Xu X J, Liu H Y, Jiao F S et al., 2020. Time-varying trends of vegetation change and their driving forces during 1981-2016 along the silk road economic belt. CATENA, 195: 104796.
[79]
Xue Y Y, Liang H B, Zhang B Q et al., 2022. Vegetation restoration dominated the variation of water use efficiency in China. Journal of Hydrology, 612: 128257.
[80]
Yan F, Wang Y J, Lu Q et al., 2023. Seasonally freeze-thaw changes on the Qinghai-Tibet Plateau and their possible causes. International Journal of Climatology, 43(5): 2110-2126.
[81]
Yan F, Wu B, Wang Y J, 2015. Estimating spatiotemporal patterns of aboveground biomass using Landsat TM and MODIS images in the Mu Us Sandy Land, China. Agricultural and Forest Meteorology, 200: 119-128.
[82]
Yang J P, Ding Y J, Chen R S, 2006. Spatial and temporal of variations of alpine vegetation cover in the source regions of the Yangtze and Yellow rivers of the Tibetan Plateau from 1982 to 2001. Environmental Geology, 50(3): 313-322.
[83]
Yang Y T, Guan H D, Batelaan O et al., 2016. Contrasting responses of water use efficiency to drought across global terrestrial ecosystems. Scientific Reports, 6(1): 23284.
[84]
Yin Y H, Ma D Y, Wu S H et al., 2017. Nonlinear variations of forest leaf area index over China during 1982-2010 based on EEMD method. International Journal of Biometeorology, 61(6): 977-988.
[85]
You Q L, Min J Z, Lin H B et al., 2015. Observed climatology and trend in relative humidity in the central and eastern Tibetan Plateau. Journal of Geophysical Research: Atmospheres, 120(9): 3610-3621.
[86]
Zhang Q, Shen Z X, Pokhrel Y et al., 2023. Oceanic climate changes threaten the sustainability of Asia’s water tower. Nature, 615(7950): 87-93.
[87]
Zhang T, Peng J, Liang W et al., 2016. Spatial-temporal patterns of water use efficiency and climate controls in China’s Loess Plateau during 2000-2010. Science of The Total Environment, 565: 105-122.
[88]
Zhao F B, Wu Y P, Ma S et al., 2022. Increased water use efficiency in China and its drivers during 2000-2016. Ecosystems, 25(7): 1476-1492.
[89]
Zheng H, Lin H R, Zhou W J et al., 2019. Revegetation has increased ecosystem water-use efficiency during 2000-2014 in the Chinese Loess Plateau: Evidence from satellite data. Ecological Indicators, 102: 507-518.
[90]
Zheng Y T, Han J C, Huang Y F et al., 2018. Vegetation response to climate conditions based on NDVI simulations using stepwise cluster analysis for the Three-River Headwaters region of China. Ecological Indicators, 92: 18-29.
[91]
Zheng Z T, Zhu W Q, Zhang Y J, 2020. Seasonally and spatially varied controls of climatic factors on net primary productivity in alpine grasslands on the Tibetan Plateau. Global Ecology and Conservation, 21: e00814.
[92]
Zhong L, Ma Y M, Xue Y K et al., 2019. Climate change trends and impacts on vegetation greening over the Tibetan Plateau. Journal of Geophysical Research: Atmospheres, 124(14): 7540-7552.
[93]
Zhou G S, Ren H R, Liu Tong et al., 2023. A new regional vegetation mapping method based on Terrain-climate-remote sensing and its application on the Qinghai-Xizang Plateau. Science China Earth Sciences, 66(2): 237-246.
[94]
Zhou L M, Tian Y H, Myneni R B et al., 2014. Widespread decline of Congo rainforest greenness in the past decade. Nature, 509(7498): 86-90.
[95]
Zhu Q, Jiang H, Peng C H et al., 2011. Evaluating the effects of future climate change and elevated CO2 on the water use efficiency in terrestrial ecosystems of China. Ecological Modelling, 222(14): 2414-2429.
[96]
Zhu X J, Yu G R, Wang Q F et al., 2015. Spatial variability of water use efficiency in China’s terrestrial ecosystems. Global and Planetary Change, 129: 37-44.
[97]
Zhu Z C, Piao S L, Myneni R B et al., 2016. Greening of the earth and its drivers. Nature Climate Change, 6(8): 791-795.
Global environmental change is rapidly altering the dynamics of terrestrial vegetation, with consequences for the functioning of the Earth system and provision of ecosystem services(1,2). Yet how global vegetation is responding to the changing environment is not well established. Here we use three long-term satellite leaf area index (LAI) records and ten global ecosystem models to investigate four key drivers of LAI trends during 1982-2009. We show a persistent and widespread increase of growing season integrated LAI (greening) over 25% to 50% of the global vegetated area, whereas less than 4% of the globe shows decreasing LAI (browning). Factorial simulations with multiple global ecosystem models suggest that CO2 fertilization effects explain 70% of the observed greening trend, followed by nitrogen deposition (9%), climate change (8%) and land cover change (LCC) (4%). CO2 fertilization effects explain most of the greening trends in the tropics, whereas climate change resulted in greening of the high latitudes and the Tibetan Plateau. LCC contributed most to the regional greening observed in southeast China and the eastern United States. The regional effects of unexplained factors suggest that the next generation of ecosystem models will need to explore the impacts of forest demography, differences in regional management intensities for cropland and pastures, and other emerging productivity constraints such as phosphorus availability.

Funding

National Nonprofit Institute Research Grant of CAF(CAFYBB2018ZA004)
National Nonprofit Institute Research Grant of CAF(CAFYBB2023ZA009)
Fengyun Application Pioneering Project(FY-APP-ZX-2023.02)
PDF(159087 KB)

365

Accesses

0

Citation

Detail

Sections
Recommended

/