Evolutionary Characteristics Of Gross Primary Productivity And Its Response To Water Constraints In China’s Grassland Ecosystems

At the ecosystem level,Gross Primary Productivity（GPP）is defined as the organic compounds formed by plants that absorb atmospheric CO₂through photosynthesis and sequester the carbon in plant bodies.Fossil fuel combustion emits large amounts of CO₂into the atmosphere,which is the main reason for the imbalance of the Earth’s carbon balance,and terrestrial ecosystems are changing from a weak carbon sink that was close to equilibrium to a core carbon sink that is increasing.How to accurately assess the space-time evolution of ecosystem carbon indicators has become a critical issue to be addressed.Chinese grasslands are located in a typical arid and semi-arid climate zone,are sensitive to various aspects of global change,as well as responding strongly to land-atmospheric water constraint events.Global environmental changes and drought events will inevitably have serious impacts on the function and structure of Chinese grasslands.Based on this,our paper took Chinese grassland ecosystems as the research area,used multi-source GPP data from terrestrial ecosystem model simulations and remote sensing satellite observations,combined trend analysis and breakpoint tests to quantitatively analyze the space-time evolution patterns of Chinese grassland GPP over the past 40 years;and used comprehensive attribution analysis to compare the relative contributions of different aspects of global environmental change and individual seasons to long-term changes in Chinese grassland GPP;the Copulas function and Bayesian equation were introduced to explore the response of Chinese grassland GPP to land-atmospheric moisture constraints under atmospheric drought,soil drought,and compound drought.The main results are as follows.（1）Modeled and remotely estimated GPP results showed that more than 80%of Chinese grasslands show a significant increasing trend over the past 40 years,with growth rates ranging from 0.68 to 3.13 g C m^-2year^-1.GPP_maxalso shows that more than 80%of Chinese grasslands are growing rapidly,with overall growth rates exceeding 0.1 g C m^-2year^-1across regions.The overall long-term trends and interannual variability of multi-source GPP and GPP_maxare generally consistent,but vegetation dynamics remain uncertain in some areas.The breakpoint test showed that‘monotonic increase’was the most significant breakpoint type in Chinese grassland GPP（33.09%）,and the change direction of Chinese grassland GPP before and after the breakpoint was also increasing.The peak photosynthetic growth and the length of the phenological period jointly control the interannual variability of GPP in Chinese grasslands.（2）Based on a multi-model analysis of the impacts of various aspects of global change（rising CO₂concentration,nitrogen deposition,climate change,and land use change）on Chinese grasslands,we found that interannual variability of climate change plays a dominant role in the GPP long-term change,with a relative contribution of 64%.In particular,the interannual variability of climate change in summer could explain at least 40%of GPP change.The trend and variance terms of carbo-nitrogen fertilization effects（increased CO₂concentration and increased N deposition）contributed 29%and 26%,respectively,to the relative contribution of increased GPP in Chinese grasslands,with the summer trend and variance terms contributing 12.83%and5.19%,respectively.The overall negative contribution of land use change to the long-term trend of GPP,was-31%.（3）Analyzing the space-time dynamics of saturated vapor pressure（VPD）before and after100 years,we found that the growth rate of VPD in Chinese grasslands was 3.992 k Pa 10^-3year^-1and 17.969 k Pa 10^-3year^-1in the historical period and medium-high emission future scenarios,respectively,and the atmosphere tended to be dry since 1980 and will continue until the end of the21st century.VPD is an important water constraints that regulates interannual variability of GPP in grasslands in China,and a continuously increasing atmospheric drought will break the carbon balance of the ecosystem,leading to a high probability of deficit in GPP（20.31-61.36%）.The correlation between VPD and GPP was mostly negative（r<-0.4）in the warm season,and the correlation was even more negative in the future period,representing that rising VPD limits GPP increase,and water constraint will be stronger in the future.Mild atmospheric drought is more likely to initiate the threshold of mild vegetation loss.A more intense atmospheric drought would be required to activate the threshold for severe vegetation loss.（4）Soil moisture（SM）is another water constraints that regulates the interannual variability of GPP in Chinese grasslands,and soil water shortage can directly cause the loss of ecosystem GPP.SM was strongly positively correlated with GPP（r=0.48）,and the positive correlation between shallow SM（0-50 cm）and GPP was higher（r=0.62）.It is clear that GPP decreases when SM decreases in Chinese grasslands,and especially the shallow soil water decrease has a greater effect on GPP.The probability of GPP decline due to soil drought was overall higher than that of atmospheric drought in the historical period（1.78-8.19%）,but the probability of atmospheric drought-induced GPP deficit was significantly higher in the future and became a key water constraint to inhibit GPP accumulation in some regions（e.g.,Loess Plateau）.The probability of occurrence of compound drought events（i.e.,land-atmospheric water constraints）is 3-4 times higher than that of stochastic drought events,and the frequency,intensity,and affected-area of compound drought occurrence are significantly higher than that of single drought.Compound droughts caused a decline of up to 20.27%in GPP of grassland ecosystems in China,while the decline of single atmospheric drought or soil drought was only 12.34%and 14.32%,respectively.This is due to the fact that VPD and SM are a set of strongly coupled bivariate variables,and the continuous strengthening of the land-atmosphere feedback causes a higher probability of occurrence of compound drought events and a greater impact on ecosystem GPP.（5）Uncertainty analysis of the multi-source GPP dataset used in our paper was conducted using cross-validation,standard error statistics,and ensemble empirical modal decomposition.We found that GPPs from different sources are consistent in portraying the spatio-temporal pattern of GPP in China’s grasslands,but there are still extensive uncertainties in some regions due to many differences in model structure,parameterization,and driving data.Uncertainties are higher in the future scenario than in the historical period,and GPP uncertainties are much higher in the high-emissions scenario than in the medium-low emission scenario.