
基于优化的MaxEnt模型预测赤水蕈树的潜在适宜区.docx
26页基于优化的MaxEnt模型预测赤水蕈树的潜在适宜区 文国卫 叶兴状 施晨阳 赖文峰 刘邦友 蒋天雨 朱晓如 张国防摘 要:赤水蕈樹(Altingia multinervis)具有较高的经济和药用价值,但由于气候变化和人类活动的强烈干扰,使其野生种质资源已近枯竭,被列为国家二级保护植物为重建末次间冰期以来赤水蕈树地理分布格局变化过程、了解环境因子对潜在适宜区的制约机制,为赤水蕈树及其周围野生动植物的栖息地保护和引种栽培提供可靠的科学依据,该文运用MaxEnt模型结合地理信息系统ArcGIS软件模拟赤水蕈树潜在适生区及其空间变化格局,利用贡献率(PC)、置换重要值(PI)和刀切图评估制约现代地理分布的主导因子结果表明:(1)优化的MaxEnt模型预测精度极高,AUC值大于0.97;赤水蕈树现代潜在适生区主要集中分布在贵州、四川和重庆交界的长江流域周围,其中以贵州赤水河流域为最佳适宜区;昼夜温差月均值(bio2)、年降水量(bio12)、降水季节性变化(bio15)和温度年较差(bio7)四个关键环境因子影响赤水蕈树的地理分布2)现代赤水蕈树潜在中高适生区面积2.692 6×104 km2,末次间冰期、全新世中期、未来时间段(2041—2060年、2061—2080年)低浓度(RCP2.6)和高浓度(RCP8.5)的中高适生区面积分别是2.277 3×104、2.831 0×104、2.159 6×104、2.605 1×104、2.330 4×104、2.460 4×104 km2。
3)在未来4种排放情景下,赤水蕈树新增的面积集中在四川和贵州区域,而在过去的全新世中期则集中分布在重庆区域综上结果表明,赤水蕈树分布范围狭窄,而赤水河流域独特的地形优势可能是其避难的主要场所关键词: 赤水蕈树, MaxEnt模型, 潜在适生区, 环境因子, 避难所:Q945:A:1000-3142(2022)03-0363-10Potential suitable area of Altingia multinervispredicted by optimizated MaxEnt modelWEN Guowei1, YE Xingzhuang1, SHI Chenyang1, LAI Wenfeng1, LIU Bangyou2,JIANG Tianyu1, ZHU Xiaoru1, ZHANG Guofang1*( 1. College of Forestry, Fujian Agriculture and Forestry University, Fuzhou 350002, China; 2. GuizhouChishui Alsophila National Nature Reserve administration, Zunyi 564700, Guizhou, China )Abstract:Altingia multinervis has been listed as a second-class protected plant in China., with high economic and medicinal values. However, its wild germplasm resources have been nearly exhausted due to the strong disturbance of climate change and human activities. Here, we aim to simulate the changes in the geographical distribution pattern of A. multinervis since the last interglacial (LIG) and to explore how climatic factors restrict the potential suitable areas, to provide a reliable scientific basis for habitat protection and cultivation of A. multinervis and its surrounding wildlife. In this study, based on optimized MaxEnt model and ArcGIS software, we simulated the potential suitable areas of A. multinervis and its spatial change pattern. The importance of environmental factors that constrains current geographical distribution was evaluated by percent contribution(PC), permutation importance(PI) and Jackknife test. The results were as follows: (1) The prediction accuracy of the optimized MaxEnt model was extremely high, as indicated by the value of the area under the receiver operator characteristic curve over 0.97. The potential suitable areas of A. multinervis for the present distribution were mainly distributed in the Yangtze River basin at the junction of Guizhou, Sichuan and Chongqing, of which the best suitable area was around Chishui River basin in Guizhou. The key environmental factors affecting the potential distribution of A. multinervis were mean diurnal range (bio2), annual precipitation (bio12), seasonality of precipitation (bio15) and range of annual temperature (bio7). (2) The potential middle-high suitable area of A. multinervis in current was 2.692 6×104 km2, in last interglacial (LIG) was 2.277 3×104 km2, in the mid-Holocene (MH) was 2.831×104 km2, in four future scenarios were 2.159 6×104 km2 (2050s RCP2.6), 2.605 1×104 km2 (2050s RCP8.5), 2.330 4×104 km2 (2070s RCP2.6), 2.460 4×104 km2 (2070s RCP8.5), respectively. (3) Under the four future discharge scenarios, the newly increased area of A. multinervis was concentrated in Sichuan and Guizhou, while it was concentrated in Chongqing in the mid-Holocene. All the above results indicate that the distribution range of A. multinervis is narrow and the potential suitable area is very small, and the unique topographic advantage of Chishui River Basin may be the main refuge place of A. multinervis.Key words: Altingia multinervis, MaxEnt model, potential suitable area, environmental factors, refuge近年来,全球气候变化和人类活动对环境、生物圈和生物多样性的影响日益显著,主要表现在降水的次数、强度和频率极不稳定,以及热浪、干旱、雷暴、洪水和飓风等极端事件的频发(D′Amato et al., 2020)。
其中,极端气候事件强度和频率的增加,将直接对植物分布产生较大的负面影响(Mga et al.,2021)由于气候变化的复杂性和不可预测性,以及地理位置的特殊性,植物对气候变化的适应就显得尤为重要(Ceccarelli & Grando, 2020)为应对气候变化,特别是在气候环境恶劣的情况下,微避难所就极有可能是支持当地物种生存的栖息地(Mclaughlin et al., 2017)目前,虽然对一些物种的灭绝机制已有新的进展,但仍未知哪些气候因子变化引起了物种灭绝,以及有多少物种会面临灭绝风險(Roman-Palacios & Wiens, 2020)在气候和人为土地利用变化的双重作用下,植物生存和繁衍所需的栖息地类型大幅减少,并呈现碎片化(Sales et al., 2020; Karl et al., 2020)而减缓气候变化和人类活动对原始森林生态系统的干扰,有益于物种间进行自然更新、基因交流和空间迁移等过程(Matteo et al., 2020)因此,在气候变化的背景下保护生物多样性,需要制定出保护濒危物种切实可行的保护策略(Suggitt et al., 2018)全球气候变暖是21世纪生物多样性保护的重大挑战之一,而生态位模型(environmental niche models,ENMs)对气候变化下物种适宜潜在分布的变化性预测,已成为多个领域研究的热点(Li et al., 2020)。
生态位模型在生态、进化和生物地理学中被广泛应用,因其能够对物种进行连续性预测和潜在适宜分布范围的变化描述(Galante et al., 2017;Mariano et al., 2019)目前常用生态位模型主要有:最大熵模型(MaxEnt)(Sobek-Swant et al., 2012)、规则集遗传算法模型(GARP)(Zhang et al., 2020)、CLIMEX模型(Szyniszewska et al., 2020)等其中,MaxEnt模型是最常用的小生境建模工具之一(Feng et al., 2019),是由于该模型对极少的样本数量(n≥5)就能较好地模拟出物种的适生范围,且成本低、操作简单、运行时间短(Pearson et al., 2007)因此,MaxEnt模型常被国内外研究者应用于珍惜濒危动植物保护(Farashi & Shariati, 2017; Han et al., 2019)、物种入侵(Kariyawasam et al., 2020; 陈剑等,2021)、作物种植区划(Zhang et al., 2016)、病虫害防治(Li 。












