Abstract: As a highly hazardous invasive alien species, it is important to identify the dominant environmental factors and potential habitat areas that affect the Solanum rostratum in the agro-pastoral ecotone of northern China in order to rationally control and prevent it. In this study, based on the Maximum Entropy Model (MaxEnt) optimized by the R language ENMeval data package and ArcGIS, combined with the distribution data of S. rostratum and 9 environmental factors in agro-pastoral ecotone of northern China, we predicted the potential suitable growth area of S. rostratum in the agro-pastoral ecotone of northern China. According to the results of the model, the potential distribution area of the current simulation was basically consistent with the actual distribution. The area under the receiver operating characteristic curve (AUC value) was 0.868, and the prediction result was better. The results showed that the mean temperature of wettest quarter, aridity Index, precipitation variation coefficient and altitude were the main environmental variables affecting the potential distribution of S. rostratum. The highly suitable area of S. rostratum in the agro-pastoral ecotone of northern China accounted for about 6.82×10⁴ km², which was mainly distributed in the eastern part of the agro-pastoral ecotone of northern China.

Key words: Solanum rostratum, Agro-pastoral ecotone of northern, Suitable distribution area, Maximum entropy model (MaxEnt)

摘要： 黄花刺茄(Solanum rostratum)作为危害性极强的外来入侵物种，明确影响其在中国北方农牧交错带分布的主要环境因子以及潜在适生区对合理防治和管控具有重要意义。本研究以中国北方农牧交错带为研究区域，基于R语言ENMeval数据包优化的最大熵模型(MaxEnt)和地理信息系统(ArcGIS)，结合黄花刺茄在中国北方农牧交错带的分布数据及9个环境因子，分析黄花刺茄在中国北方农牧交错带的潜在适生区。模型预测结果显示，当前模拟的潜在分布区与实际分布基本吻合，受试者工作特征曲线下面积(AUC值)为0.868，预测结果较好。结果表明，最湿季平均温、干旱度指数、降雨量变异系数、海拔是影响黄花刺茄分布的主要环境变量；黄花刺茄在中国北方农牧交错带的高适生区面积约占6.82×10⁴ km²，主要分布在中国北方农牧交错带的东部地区。

关键词: 黄花刺茄, 北方农牧交错带, 适生区, 最大熵模型