Response Of Temperate Grasslands To Climate Change In The Northeast China Transect

The temperate grasslands in China are one of the most ecologically sensitive regions worldwide to climate variability.Hence,it is of great importance to investigate the response of temperate grasslands to climate change for protecting ecological security and ensuring sustainable use of grasslands.The grasslands located in the northeast China transect?NECT?are selected as the study area with its advantages in studying the response of temperate grasslands to climate change:there is an evident gradient of precipitation in the NECT,which can represent different water conditions of most temperate grasslands in China.Based on both statistic methods and a land surface model,we aim to figure out the response patterns and mechanisms of grasslands in the context of climate change.Firstly,this study analyzes the response of the growing season normalized difference vegetation index?NDVI?to climatic variables?i.e.,growing season precipitation and temperature?during 1982-2014 in the grasslands of NECT,using partial correlation and linear regression methods.The results show that precipitation is the dominant factor of grass growth at both spatial and temporal scales.And the spatial sensitivity of growing season NDVI to precipitation is 12-15×10^-4?mm^-1,much greater than the temporal sensitivity(1.8-4.8×10^-4?mm^-1).The land surface model,named as Community Land Model 4.5?CLM?,is employed to explore the response mechanism of grasslands to climate change.We evaluate the temperate grassland phenology simulated by CLM for the NECT with the remote sensing leaf area index?LAI?data,and figure out two deficiencies:a decay of end of growing season and an unrealistic senescence period existed in summer.By parameter sensitivity tests,we substantially improve the phenology algorithm in CLM for temperate grasslands at both intra-and inter-annual scales.Based on the adjusted CLM,we claim that the difference between spatial and temporal sensitivities of grass growth to precipitation result from 1)the“memory effect”,which means the effects of the past precipitation on grass growth,reduces the temporal sensitivity;2)the spatial sensitivity includes both the shift in the dominant grass species and precipitation change along precipitation gradients;3)the spatial sensitivity also represents the effects of spatial heterogeneity of other climatic and environmental factors?e.g.,temperature,topography and soil texture?on grass growth.Furthermore,we predict responses of grass growth as well as regional carbon and water fluxes using CLM under ideal climate change scenarios.When precipitation increases by 10%-20%and temperature increases by 1-2 ^oC in the future,the activity of grasslands in the NECT will be enhanced comparing with current status:mean annual LAI will increase by 10.0%-21.9%,mean annual evapotranspiration will increase by 23.45-51.98 mm and mean annual carbon sink will increase by 0.25-2.07 g C?m^-2.For a typical region in Horqin Sandy Land where grass growth is significantly influenced by groundwater availability,the land surface model with lateral groundwater flow?CLM-GW?is employed to simulate grass dynamics.It is proved that CLM-GW can capture the soil moisture,LAI as well as sensible and latent heat fluxes for grass regions much better than CLM.The results indicate an important role of groundwater for the grassland ecosystems in the typical region.The groundwater stimulates grass growth during growing season by providing sufficient water supply:LAI increases by 28.9%,evapotranspiration increases by 43.5%,gross primary productivity increases by 80.6%,and water use efficiency at the ecosystem scale increases by 25.2%.