Temporal And Spatial Distribution Of Stellera Chamaejasme And Its Relationship With Climate Factors In Qinghai Province

Stellera chamaejasme L.,also known as steamed bread flower,belongs to family Thymelaeaceae and genus Stellera,it is a perennial herb with deep roots.In recent years,under the interference of climate change and human activities,natural grassland in the Qinghai-Tibet plateau has degraded to different degrees.One of the main characteristics of grassland degradation was the proliferation of poisonous grass.Qinghai province,located in the northeast of the Qinghai-Tibet plateau,is the "disaster area" of S.chamaejasme distribution.According to the field survey of grassland station in Qinghai province in 1998,the infectious area of S.chamaejasme accounted for 1/4 of the total area of poisonous grass in Qinghai.The occurrence of S.chamaejasme on a large scale has seriously affected the livelihood of pastoralists,the sustainable development of local economy and the stability of the Qinghai-Tibet plateau ecosystem.At present,the distribution area of S.chamaejasme in Qinghai Province mainly comes from the grassland survey of the provincial grassland station from 1997 to 1998.Field investigations are time-consuming and labor-intensive,which brings great inconvenience to research.The development of remote sensing technology has made it possible to identify a large range of S.chamaejasme.However,there are no remote sensing studies on the distribution of S.chamaejasme in Qinghai Province.In recent years,in high altitude areas,the study of grassland vegetation under climate change has become a hot spot.Some scholars have done research on the distribution of S.chamaejasme and climatic factors in Qinghai Province,but the factors involved are mainly temperature and precipitation.There is no comprehensive study about what climatic factors have a greater impact on the distribution of S.chamaejasme.In our study,the distribution of S.chamaejasme in Qinghai during 2013-2016 was extracted from MOD09A1 image by specific spectral characteristic parameters of S.chamaejasme.On the other hand,the advantages and disadvantages of ANUSPLIN and Universal Kriging methods in interpolating different climatic elements were compared,and the interpolation method with high precision was selected to interpolate seven meteorological elements in Qinghai.Then,the relationship between the distribution of S.chamaejasme and climate factors was analyzed from two aspects: one was the quantitative study of the relationship between the distribution of S.chamaejasme and climate factors by GAM model,and the other was the study about the relationship between the potential distribution of S.chamaejasme and climate factors by Max Ent model.The main conclusions were as follows:(1)The spectral characteristics of S.chamaejasme in Qinghai province before and after the poisonous flower period were analyzed,and the characteristic parameters of distribution extraction of S.chamaejasme in Qinghai province were constructed according to the different regions of flowering period.The distribution information of S.chamaejasme in Qinghai province from 2013 to 2016 was extracted.The overall accuracy of extraction was 69.7%,and kappa coefficient was 0.376.Compared with the field survey data,our results showed that most of the extraction results were close to the statistical data except some counties.(2)The spatial correlation analysis of S.chamaejasme distribution from 2013 to 2016 showed that the distribution of S.chamaejasme had noticeable spatial correlation.From the spatial distribution visualization of S.chamaejasme,we observed that the distribution of S.chamaejasme in Qilian,Tianjun,Gangcha,Menyuan and other places in the north of Qinghai province is the most,and showing a strong state of aggregation.(3)The accuracy of interpolation results of climatic factors in Qinghai province showed that ANUSPLIN method was obviously superior to Universal Kriging method,which verifies the superiority of ANUSPLIN method in interpolation of complex terrain area.(4)The relationship between the distribution of S.chamaejasme and climatic factors was analyzed quantitatively by generalized additive model.The results showed that the main factors affecting the distribution of S.chamaejasme in Qinghai were average air temperature,average relative humidity and sunshine duration.(5)Based on the Maxent model,the niche of S.chamaejasme distribution in Qinghai was predicted.The results showed that the AUC values of training data and validation data were both greater than 0.98,which indicated that the Max Ent model based on climate had better prediction effect on S.chamaejasme distribution in Qinghai.The results of the model show that the contribution of water vapor pressure,average temperature and precipitation is the highest among the climatic factors affecting the potential distribution of S.chamaejasme.(6)By comparing the results of Max Ent model prediction with the results of remote sensing distribution extraction,we found that Xinghai,Gangcha,Qilian,Haiyan,Gonghe,Menyuan,Henan and other counties with the highest probability of potential distribution also have higher distribution in remote sensing extraction results,which indicated that the maximum probability of potential distribution has indicative effect on the distribution of S.chamaejasme in Qinghai Province.(7)Combined with the two models,it is found that the main climatic factors limiting the distribution of S.chamaejasme in Qinghai Province are air temperature and water vapor pressure.