Analysis On The Spatiotemporal Changes Of Livestock In Qinghai Grassland In Recent 30 Years

The Qinghai-Tibet Plateau,as an important barrier for China’s and even the world’s ecological security,is vital to the global ecological security.Qinghai Province,as an important part of the Qinghai-Tibet Plateau,is vital to the protection of the environment,to ensure the ecological security of our country,there is an unshirkable responsibility and obligation.Grassland is the largest ecosystem type in Qinghai Province and an important resource for human survival in the plateau.The change of livestock number is an important issue for sustainable grassland management.It is helpful to understand the impact of human activities on grassland ecosystem to master the characteristics of livestock spatial distribution.In this paper,the characteristics of the time series change of the stock quantity in Qinghai Province in the past 30 years are analyzed,and the livestock quantity in Qinghai Province is spatialized by using the random forest model and zoning density mapping method,and the grass-livestock balance in Qinghai Province is evaluated,the main conclusions reached are as follows:(1)In the past 30 years,the stock quantity of livestock in Qinghai Province showed a decreasing trend,the fastest falling speed was horse,and the center of weight of stock quantity moved from southwest to northeast,the average change rate of stock yield was 5%.Among them,large livestock and cattle generally decline slowly,showing a first decline after the slow growth of the trend,sheep showed the first wave growth,the middle cliff-like decline,and then continued to decline slowly characteristics,the decline trend is the most obvious horse.At the end of the year,the number of sheep in semi-agricultural and semi-pastoral areas such as Haibei Prefecture,Haidong Prefecture and Hainan Tibetan Autonomous Prefecture Prefecture in Qinghai Province is increasing,while the number of large livestock is decreasing,which is also related to the increase in the number of agricultural machinery.In pure animal husbandry areas such as Haixi Mongol and Tibetan Autonomous Prefecture,Golog Tibetan Autonomous Prefecture and Yushu,the year-end stock of sheep is on a downward trend.(2)By comparing with the data sets of GLW2 and GLW3 in Qinghai Province,we found that the overall precision of the data spatialization in this paper is higher.The decision coefficients of fitting the GLW2 and GLW3 data sets with county statistical data are 0.50 and 0.52 respectively,and the R2 in this paper is 0.99.The explanation degree of the random model variable(% var explained)for the prediction of livestock stocking quantity at the end of four periods was 97%,96.84%,96.99% and 96.77% respectively.Mean of squared residuals is 0.0111331,0.009525385,0.00943594 and 0.009216907,respectively.In Qinghai Province in 2017,more than 40% of the regions had a livestock density of less than 20 sheep units / KM2,and the remaining 30% had a livestock density of 20-100 sheep units / km2.By ranking the importance of independent variables,it is found that the density of residential area,road density,population density and precipitation are important indicators of livestock distribution in Qinghai Province,and play an important role in the model.At the same time,the spatialization of high spatial resolution livestock data can be realized by using random forest model combined with multiple impact factors,which can provide an important data source for the balance management of Alpine Grassland and livestock,it can also support relevant policy decision-making.(3)The grassland biomass of Qinghai Province is from low to high in northwest,southwest,Central,south and east.The overloading of grassland livestock is common in Qinghai Province,which is more than 50% of the total number of counties in Qinghai Province.The areas where overloading is more serious are Haidong and Hainan Tibetan Autonomous Prefecture,and the grass and livestock in Haibei and Haixi can basically be in balance,other areas are in overload state,but in the past 30 years most areas of Qinghai Province overload rate has decreased.