基于最大熵模型的新疆主要有毒杂草分布区预测  

Prediction of distribution area of main noxious and miscellaneous weeds in Xinjiang based on MaxEnt model

作　　者：姚政宇 韩其飞 林彬[4] YAO Zhengyu;HAN Qifei;LIN Bin(School of Geographic Sciences,East China Normal University,Shanghai 200241,China;School of Geographical Sciences,Nanjing University of Information Science&Technology,Nanjing 210044,China;Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters,Nanjing University of Information Science&Technology,Nanjing 210044,China;The 7th Institute of Geology&Mineral Exploration of Shandong Province,Linyi 276000,China)

机构地区：[1]华东师范大学地理科学学院,上海200241 [2]南京信息工程大学地理科学学院,南京210044 [3]南京信息工程大学气象灾害预报预警与评估协同创新中心,南京210044 [4]山东省第七地质矿产勘查院,临沂276000

出　　处：《生态学报》2023年第12期5096-5109,共14页Acta Ecologica Sinica

基　　金：江苏省基础研究计划自然科学基金面上项目(BK20201393);国家自然科学基金面上项目(42271493)。

摘　　要：近年来,在气候变化和人类放牧活动的共同作用下,新疆地区有毒杂草发展迅速,草地退化严重,影响了新疆草地生态系统及草地畜牧业的发展。了解新疆主要有毒杂草的时空分布及影响其分布的主要生态因子,对新疆畜牧业可持续发展具有重要意义。使用最大熵模型(MaxEnt),预测了不同气候情景下新疆主要有毒杂草的潜在分布区。结果表明:1)使用MaxEnt最大熵模型在三种气候情景下的模拟精度值均大于0.8,处于“好”水平,表明模型模拟结果有较高的可信度;2)白喉乌头受等温性和最干月降水量的影响,主要分布在阿勒泰等地区;无叶假木贼主要受到放牧率和最干月降水量的影响,沿天山山脉以及塔里木盆地西、北部分布;影响纳里橐吾的是平均气温日较差和等温性,分布在巴音郭楞州等地;小花棘豆主要受到平均气温日较差和降水量季节性变化的影响,集中分布在新疆西北部地区;苦豆子受海拔和最冷季降水量影响,主要分布于塔里木盆地等地区;3)有毒杂草入侵概率较低的区域(非适生区)占比51.69%;入侵的高危区面积占比5.62%,主要集中在阿勒泰、塔城、昌吉自治州、五家渠、博尔塔拉州、伊犁自治州、石河子市、乌鲁木齐市、阿克苏、克孜勒苏州和喀什;4)未来气候情景下,新疆主要有毒杂草呈现进一步扩散趋势;SSP245情境下有毒杂草扩张蔓延趋势比在SSP126情境下更为明显,这说明在较高辐射强迫的气候情境下,有毒杂草表现出较高的适应能力;未来平均气温日较差变小的区域更易被有毒杂草入侵,而最干月份降水量明显增加的区域,有毒杂草入侵的可能性较小,其面积发生退缩。In recent years,under the influences of climate change and grazing activities,grassland has been seriously degraded,which has greatly affected the development of grassland ecosystem and animal husbandry in Xinjiang.Therefore,it is great significance to understand the spatio-temporal distribution of main noxious and miscellaneous weeds and the main ecological factors affecting their distribution for the sustainable development of animal husbandry in Xinjiang.In this paper,MaxEnt was adopted to predict potential distribution areas of main noxious and miscellaneous weeds in Xinjiang under different climate scenarios.The results showed that:1)the simulation accuracy values of MaxEnt in the three scenarios were all greater than 0.8,which was at the good level,indicating that the model results have a high degree of confidence.2)In the historical period,Aconitum leucostomum was mainly influenced by isothermality and precipitation of driest month,mainly distributed in areas such as Altai.Anabasis aphylla was mainly influenced by grazing rates and precipitation of driest month,distributed along the Tianshan Mountains and around the western and northern parts of the Tarim Basin.The main influencing factors for the Ligularia narynensis were mean diurnal range and isothermality,mainly distributed in areas such as Bayin Goleng Prefecture.Oxytropis glabra was mainly influenced by mean diurnal range and precipitation seasonality,and distributed mainly in the northwestern part of Xinjiang.Sophora alopecuroides was affected by altitude and precipitation of coldest quarter and mainly found in areas such as the Tarim Basin.3)The area with low probability of invasion of noxious and miscellaneous weeds accounted for 51.69%.The area with high risk of invasion accounted for 5.62%,mainly concentrated in Altay,Tacheng,Changji Autonomous Prefecture,Wujiaqu,Boltala,Yili Autonomous Prefecture,Shihezi,Urumqi,Aksu,Kizilsu and Kashgar.4)In the future climate scenarios,the main noxious and miscellaneous weeds in Xinjiang show a trend of further

关 键 词：有毒杂草 最大熵(MaxEnt)模型 生态因子 时空分布 

分 类 号：S451[农业科学—植物保护]