| Ambrosia artemisiifolia L.and Ambrosia trifida L.are the first and second batches of severely invasive species announced in China,respectively.Which not only seriously threaten agricultural and animal husbandry production,causing biodiversity to decline,and its pollen have highly allergenic.At present,these two species have invaded in the Central Asian hub of the “Belt and Road” – the Yili Valley in Xinjiang,and widely spreading in the farmland,forest area,grassland,roadside,wasteland.However their potential development trend is hard to estimate.It will be of great significance for the early warning and specific prevention and control of invasive species to clarify which regions are suitable for the invasive species to survive,how their suitability degree are and how their response to climate change.In this paper,firstly,the Receiver Operating Characteristic curve(ROC curve)was used to evaluate the simulation effects of the three niche models(Bioclim,ENFA,and Maxent).Then,a relatively better model among them was used to predicted the potential distribution of A.artemisiifolia and A.trifida in the current climate based on two scales data(Global scale and Xinjiang scale),and then divided the continuous suitability probability map into a binary(presence / absence)distribution map,finally we divided the suitability and unsuitability areas into four levels respectively by using frequency-based statistics method.In order to explore the effects of climate change on the potential habitat of A.artemisiifolia and A.trifida in Xinjiang,17 sets of GCMs data were used to predict in two periods,2050s(2041-2060)and 2070s(2061-2080),under two climate change scenarios,RCP4.5(Representative Concentration Pathway 4.5)and RCP8.5(Representative Concentration Pathway 8.5),then a majority voting methods were used to determine their future suitable habitats.Finally,the distribution changes of A.artemisiifolia and A.trifida in different periods were calculated and analyzed from the aspects of area change and range change.The result showed that:(1)The Maxent model has a relatively better prediction effect.The AUC values of the ENFA,Bioclim,and Maxent models were 0.898,0.921,and 0.965,respectively.The model performance showed that Maxent was the model with the best simulation effect among them.In addition,Maxent model output the continuous existence probability map,value range from 0~1,which can better distinguish the suitable degree of species in different regions.(2)Major environmental impact factors.Precipitation-related variable was the primary limiting factor for the distribution of A.artemisiifolia in the world and Xinjiang,with cumulative contribution rates of 64.7% and 39.9%,respectively.The Precipitation of Driest Month was the factor that contributes most to A.artemisiifolia's distribution in both scales(63.5% and 27.3%,respectively).Precipitation-related(cumulative contribution rate was 36.0%)variable was the primary limiting factor for A.trifida's distribution in Xinjiang,and temperature-related(cumulative contribution rate was 56.0%)variable was the primary limiting factor for A.trifida's distribution in the world.But in global scale,the contribution rate of the Precipitation of Driest Month was the highest(36.2%),and the Temperature Seasonality was the second(29.1%),indicating that the Precipitation of Driest Month still was the factor that contributes most to A.trifida's distribution on global scale,but as a whole,the restriction of temperaturerelated variable about distribution was higher than that of precipitation-related variable.Farmland and construction land are the areas with the highest potential invasion risk of A.artemisiifolia(accounted for 58.8%,58.0%,respectively)and A.trifida(accounted for 40.4%,42.1%,respectively).Irrespective of topographic variables will underestimates its suitable habitats in places where with low precipitation(such as Hami and Aksu).The contribution of topographical catchment to these areas with low precipitation,such as arid areas,cannot be ignored,especially for these harmful invasive species.(3)The potential habitats was predicted with two scales data under current climate.Under the current climate,the global data predicted the total suitable habitats of A.artemisiifolia and A.trifida in Xinjiang(accounting for about 17.78% and 13.71% of the total area of Xinjiang,respectively),which were much larger than that of the study area data(accounting for about 0.77% and 0.13%,respectively).The suitable habitats of A.artemisiifolia predicted by the study area data were mainly distributed in Yili,Bole,Tacheng,Altay,Shihezi,Changji,Urumqi,while A.trifida were mainly distributed in Yili,Tacheng,Changji.The suitable habitats of both A.artemisiifolia and A.trifida predicted by the global data were distributed all over in northern Xinjiang,and in southern Xinjiang were mainly distributed in Aksu,Atushi.We suggest that the prediction results of global data which covering the richer niche information of A.artemisiifolia and A.trifida should be used as the basis for current long-term prevention and control and early warning.In addition,when predicting the potential distribution of these species which with short invasion times and imbalanced between distribution and environment,data including their native range and invasion range should be used.(4)Future changes of potential suitable habitats.Under the RCP4.5 scenario,the potential suitable habitats of A.artemisiifolia and A.trifida increased by 16% and 56% in the 2050 s period,respectively,and increased by 17% and 63% in the 2070 s period,respectively.Under the RCP8.5 scenario,the potential suitable habitats increased by 17% and 64% in the 2050 s period,respectively,and increased by 18% and 79% in the 2070 s period,respectively.Under the RCP4.5 and the RCP8.5 scenarios,as time goes on,the potential suitable habitats of A.artemisiifolia and A.trifida both showed an increasing trend,and the changes under RCP8.5 scenario were more drastic than those under RCP4.5 scenario.The increasing range of A.trifida's potential suitable habitats is larger than that of A.artemisiifolia,so future climate change will be more conducive to A.trifida's invasion and expansion in Xinjiang.The expansion and contraction positions of A.artemisiifolia and A.trifida were roughly the same,and both showed a trend of northward expansion and transfer,and their shrinkage positions were mainly concentrated in the Junggar Basin.(5)Comparison of the potential suitable habitats of A.artemisiifolia and A.trifida.In current climate,the area where A.artemisiifolia is more likely to grow than A.trifida were mainly located in Altay(1.81% more),Tacheng(0.73% more),Yili(0.34% more),Changji(0.32% more),Kashi(0.23% more),Bole and Aksu(both 0.18% more),Atush(0.15% more),Urumqi and Hami(both 0.11% more),Bayin Guolen(0.01% more),Shihezi was equal,and there's no suitable habitats for A.artemisiifolia and A.trifida in Hotan.In Turpan and Karamay,the suitable habitats for A.trifida were slightly more than A.artemisiifolia,0.01% and 0.10%,respectively.In the RCP4.5 and RCP8.5 scenarios,as time goes on,A.trifida will narrowed the gap with A.artemisiifolia in Altay,and surpass A.artemisiifolia in Tacheng,Changji,Kashi,Aksu,and Bole. |
PDF Full Text Request