Alpine Grassland Ecological Security Assessment For The Headwaters Of The Yellow River,China

With the outbreak of the global ecological environment crisis,ensuring ecosystem security is a new theme for the sustainable development of the whole society.The construction of the ecological security evaluation system has become a hotspot and difficult issue in international ecosystem research.Grassland,which is ecologically fragile and provides important ecological security barriers,is one of the most important ecosystems on land.Therefore,taking the lead in carrying out ecological security assessment in grassland areas is not only conducive to regional ecological security management,but also beneficial to provide useful support for the development of ecological security assessment schemes with independent intellectual property rights.The Headwater of the Yellow River(HYR),known as the"Chinese Water Tower",is an important ecological barrier in China.As the main body of the HYR,alpine grassland plays an important role in the regional ecological security.Due to the overgrazing and climate change,the degradation of alpine grassland in the HYR has been more and more serious.Significant changes have taken place in grassland net primary productivity(NPP),soil organic carbon content(SOC),soil erosion degree and water conservation function,which pose a serious threat to the ecological security of the HYR.Therefore,it is urgent to quantitatively and accurately evaluate the ecological security pattern and dynamic changes of alpine grassland ecosystem in the HYR,which can benefit not only the decision-making of ecological protection in this region,but also the ecological security management and regional development in the middle and lower reaches of the Yellow River.Based on the grassland interface theory,this paper puts forward an evaluation system of alpine grassland's ecological security considering the demand of ecosystem services,grassland ecological health and grassland ecological risk.Under this framework,grassland net primary productivity(NPP)and grassland soil organic carbon content(SOC)were taken as the core indicators affecting grassland health,grassland water conservation function was used to evaluate the demand for ecosystem services,and grassland soil erosion to evaluate the risk of stress on alpine grassland ecosystem.The authenticity of the evaluation indices is very critical to the accuracy of the results.Therefore,through systematic sampling in the HYR,293 measured data of biomass,628 measured data of SOC and 521 measured data of ¹³⁷Cs were collected.Based on the measured data,the machine learning algorithm was used to construct and verify the constructed models of NPP,SOC and soil erosion.The simulation results of the optimal model were selected as the index value.The water yield(WY)of the HYR from 2001-2020 was quantitatively evaluated by the In VEST(Integrated Valuation of Ecosystem Services and Trade-offs)model.On this basis,combined with the entropy method to determine the index weight,the ecological security pattern and dynamic changes of the HYR were comprehensively evaluated.The main results are as follows:(1)Grassland NPP in the HYR has increased,and there is a large spatial heterogeneity.R² of the simulated NPP of the optimal model and the measured value was 0.73,and the root mean square difference(RMSE)was 102.97 g C m^-2,indicating the model accuracy was better than that of the CASA model(R²=0.43,RMSE=155.56g C m^-2 yr^-1)and MOD17A3 products(R²=0.18,RMSE=196.35 g C m^-2 yr^-1);the spatial pattern distribution of NPP decreased from southeast to northwest;during the period from 2001-2020,the total NPP in the HYR increased by 12.07%and to a large extent after 2015;NPP in 92.35%of the HYR showed an increasing trend,while in lowland meadows and shrubs decreased slightly.(2)The spatial distribution of SOC in the HYR showed an obvious spatial heterogeneity,and the spatial pattern of SOC decreased from southeast to northwest.R²of the simulated SOC of the optimal model and the measured value was 0.71,and the RMSE was 2.05 g/kg;the average value of SOC was 73.2 g/kg;from 2001-2020,SOC in 65.52%of the HYR showed an increasing trend,while 32.27%a downward trend.Where SOC decreased was mainly alpine grassland.(3)Soil erosion in the HYR showed a decreasing trend,but the spatial distribution pattern was obviously different.The accuracy of soil erosion modulus simulated by the Ridge regression-Random Forest algorithm was the highest with R²=0.41 and RMSE=6.09 t ha^-1 y^-1,which was better than the traditional RUSLE model(R²=0.12,RMSE=15.66 t ha^-1 y^-1);spatially,soil erosion increased gradually from southeast to northwest,and the average soil erosion rate and total soil loss were 19.57 t ha^-1 y^-1 and229.16×10⁶ t y^-1,respectively;from 2001-2020,soil erosion has been reversed in 61.08%of the HYR,and aggravated in 37.32%.(4)Water yield(WY)in the HYR decreased spatially from southeast to northwest,showing a step-by-step distribution.The annual WY is about 5.09X10¹⁰m³ on average.From 2001 to 2020,WY showed an increasing trend wholly in the HYR,of which 9.44%increased significantly while in the other 90.56%a slight increase.(5)Construct a local ecological security evaluation model with independent intellectual property rights under the guidance of grassland ecosystem theory.The model expression is:ESI=NPP_normalX0.396+SOC_normalX0.219+WY_normalX0.246+SE_normalX0.139The results of the model show that from 2001 to 2020,the ecological security index decreases spatially from southeast to northwest,and the average level of ecological security index in the HYR is 0.58.The area at the safe and relatively safe level is as high as 57.38%;the unsafe and unsafe areas account for 22.21%of the total area of the HYR.Areas with improved ecological security account for 39.30%of the total area,while areas with deterioration for 32.73%.The results of ecological security assessment provide a data basis for ecological protection decision-making in the HYR and the middle and lower reaches of the Yellow River.