Spatial Analysis Of Pest Distribution And Prediction Of Potential Geographical Distribution About Prosodes (Lioprosodes) Dilaticollis Motschulsky In Xinjiang Grassland

There are 5.72588 × 105 km 2 natural grassland area and 4.80068 × 105 km 2available grassland area in Xinjiang,which is not only one of China's animal husbandry bases,but also is one of the serious damage regions suffered grassland pests and diseases.In 2013,there was 2.3073×104 km2 pest damage area in Xinjiang grassland,in which was 1.0100×104 km2 serious damage area.The harmful species are mainly include locusts,Prosodes(Lioprosodes)dilaticollis Motschulsky,Loxostege sticticalis and so on.P.dilaticollis is a very harmful underground pest in the grassland of Xinjiang.In order to carry out early-warning,prevention and control for P.dilaticollis effectively in Xinjiang,We went to typical damaged area Manasi County— Hutubi County grassland to conduct field sampling surveys from April 13 th to 15 th,April28th to 29 th,May 19 th to 20 th and June 2nd to 3rd in 2016.And we obtained the insect density data,soil factors data(soil temperature,soil humidity and soil conductivity)and GPS data and so on about P.dilaticollis' different occurrence periods(early period,peak period and the end of the period)from the four times wild sampling.The criteria about determining the occurrence density and damage degree about P.dilaticollis were established based on expert consultation method.Combined with "3S" technology(Geographic Information System,GIS;Global Navigation Satellite System,GNSS),Landsat8 satellite images of three periods were used to interpret the pest distribution using maximum likelihood supervision classification(MLC).The remote sensing interpretation accuracy can be used for area estimation basically.The results about the damage changes of P.dilaticollis in Manasi County—Hutubi County grassland generally showed a trend of moderate damage(10 heads / m2<x?20 head / m2)to mild damage(0 head / m2<x ? 10 heads/ m2)in 2016.The highest damage degree of P.dilaticollis generally showed a trend of serious damage(x>20 head / m2)downgrade to moderate damage(10 heads / m2<x?20head/m2)and each damage level area and its proportion had been decreasing from2013 to 2016.Meanwhile four different spatial interpolation methods(Inverse Distance Weighted,IDW;Radial Basis Functions,RBF;Local Polynomial Interpolation,LPI;Kernel Smoothing,KS)were used to interpolate based on measured population density data from the four times wild sampling.Kernel Smooth(KS)interpolation accuracy is generally higher than others by cross validation method.We concluded that the MLC-KS method could reflect the pest distribution and the pest level.The results showed that the pest distribution and the pest level were anastomotic by two methods.From the whole Xinjiang scale,potential geographic distribution of P.dilaticollis was predicted and analyzed in current climate condition and future climate scenarios by using GPS data of actual distribution,biological climatic data(1950-2000),future bioclimatic data(Beijing Climate Center_Climate System Model version1.1,BCC_CSM1.1 model)in Xinjiang based on Max Ent model.The soil factors data and meteorological data and so on in typical damaged area were used to verify the impact factors that Max Ent model obtained,and the results were also validated by ROC(receiver operating characteristics)curve analytical method.The results showed that AUC(area under the ROC curve)of the training data and testing data in current climate condition and future climate scenarios were all greater than or equal 0.990,which showed that the predicted results had achieved the level of excellence.The potential geographical distribution of P.dilaticollis in Xinjiang was located in Tianshan Mountains region,mainly in Ili River valley and northern slope of the Tianshan Mountains.The high-level suitable distribution region and moderate suitable distribution region of P.dilaticollis were in Ili River valley such as the grassland of Xinyuan County,Nileke County,Zhaosu County,Gongliu County,Tekesi County and Chabuchaer County as well as northern slope of the Tianshan Mountains such as grassland of Manasi County,Hutubi County and Changji City in Changji Prefecture,Shanwan County in Tachen Prefecture,Jinhe County and the middle of Bole City in Bortala Mongolian Autonomous Prefecture as well as Dabancheng District in Urumqi City.And its potential geographical distribution was also predicted using BCC-CSM1.1 in three periods,2021-2040(2030s),2041-2060(2050s)and2061-2080(2070s),under three climate change scenarios(RCP2.6,RCP4.5 and RCP8.5),which were adopted by IPCC AR5(Intergovernmental Panel on Climate Change Fifth Assessment Report).The results showed the distribution pattern remained unchanged in the future.In the Tacheng Prefecture,the range of suitable habitat in Tuoli County,Emin County and Hebukesaier County had been reduced,and the range of suitable habitat in Bole City and Jinghe County had been reduced.In addition,the RCP8.5 scenario in 2070s(2061-2080),there are some suitable habitat to be reduced in Xinjiang,which showed the relationship between the suitable habitat and the change of climate warming.The formation about geographical distribution of P.dilaticollis is related to many environmental factors.Precipitation was the main factor for the potential distribution of suitable habitat based on the result of jackknife in Max Ent model.Especially the rainfall of May,June,November and December had the greatest impact on the geographic potential distribution of P.dilaticollis.Historical precipitation of the typical damaged area verified the accuracy of response curves about environmental variables,which can be used as a reference for monitoring and forecasting the pest preliminarily in Xinjiang grassland.When soil moisture was in the range of 15% ~30%,30 cm deep soil moisture of field sampling and population densities showed the most significantly week positive correlation,and soil conductivity at 10 cm of field sampling and population densities showed significantly week positive correlation,other deep(10cm,20 cm,30cm)soil environmental factors(soil moisture,soil temperature,soil conductivity)and insect population density didn't show significant relationship,which all proved the accuracy of the model predictions.The conclusion showed that amount of precipitation and intensity of precipitation played an important role in the survival and occurrence of adult and larvae because they were the main and direct natural factors affecting the soil temperature,soil humidity and soil oxygen content.