Vegetation Change And Its Response To Climate And Grazing In The Three-river Headwaters Region,China

Located at the third pole of the Earth,the Three-River Headwaters Region(TRHR)is regarded as the “Source of Rivers” and “China Water Tower”.It is of great importance to the water resources conservation and ecological security to China.Due to significant climate change and intensified human activities,the grassland ecology and hydrology condition have gone through significant changes.This research aims at investigating the response of vegetation to climate(i.e.precipitation and temperature)and grazing.The study is of great importance to reveal the influence of climate change and human activities on the fragile TRHR,and will be useful in land use management in this area.Firstly,based on data from gauge stations and remote sensing,we analyze the spatiotemporal distribution and change in temperature,precipitation,potential evapotranspiration,runoff,NDVI and grazing intensity,systematically.From 1960 to2018,mean annual temperature,annual precipitation,annual evapotranspiration increased significantly with change rate at 0.036?/yr,0.95mm/yr,0.48m/yr,respectively.The trends showed spatiotemporal heterogeneity.From 1956 to 2018,the runoff in the Yangtze River Source increased with a change rate of 68 million m~3/yr,while it decreased in the Yellow River Source.The NDVI in the TRHR showed an upward trend,accompanied by intensified human activities.Secondly,we investigate the response of vegetation towards climate condition and grazing activity,and investigate the regulation effects on vegetation among temperature,precipitation and solar radiation as well as those between climatic factors and anthropogenic factors.The results reveal that the vegetation activities in the TRHR were mainly driven by temperature and precipitation for the period from 1982 to 2015.The increase in temperature and precipitation has benefited the grassland.The regulation effects of temperature on precipitation contributed to vegetation growth,while human activities offseted the positive contribution of warming.We set up the Non-Homogenous Hidden Markov Model to model NDVI series in the TRHR,based on variables selection and model selection.The relative error for the model is beyond 10%,and Nash-Sutcliffe Coefficient of Efficiency is above 0.975.To attribute grazing effects on vegetation,we conduct scenario comparison method.We find that grazing has seasonal impacts on NDVI,i.e.,negative effects for the middle growing season(July,August,September),while positive effects for the early growing season(May,June)and late growing season(October).This is due to modifications of grazing on the vegetation greenness,soil properties,vegetation phenology.Because of this seasonal effect,the grazing impacts might be underestimated if we apply trend analyses with the growing season NDVI or annual NDVI.Furthermore,we find that the grazing effects continued to climb for the period from 1982 to 2009,and whether grazing exerts pressure on vegetation depended on climate condition.The research will be a useful tool for grazing management practice in the TRHR.