摘要
Paragorgia arborea及Primnoa resedaeformis是北大西洋最常见的两类大型冷水柳珊瑚,该两种柳珊瑚显著增加了底栖环境的生境复杂度。其分布信息对资源管理及保护极其重要,但难以获取。本文基于随机森林模型及最大熵模型预测其在北大西洋Traena区域的潜在分布。预测精度评价表明随机森林模型较最大熵模型具有更好的预测性能。平均曲率(90 m尺度)是随机森林模型预测两种柳珊瑚分布的主导因子,表明其与两种柳珊瑚的分布具有较强生态相关性。预测结果显示两种柳珊瑚趋向于分布在珊瑚礁体上。本预测成果可为该区域冷水珊瑚保护提供决策辅助信息。模型对比研究可为大型底栖动物分布建模的模型选择提供依据。
Paragorgia arborea and Primnoa resedaeformis are two large cold-water coral species commonly occurring in North Atlantic,which significantly increase the complexity of the benthic habitat.Mapping their distribution is fundamental for resource management and conservation,but is difficult（given their remoteness）. In this study,their potential distribution at Traena Reef complex of the North Atlantic were predicted based on random forest（RF） and Max Ent models,respectively. The RF prediction was shown to outperform the Max Ent prediction. Mean curvature at an analysis scale of90 m is the most useful terrain variable in RF prediction,which is ecologically correlated to the distribution of the two species. The two species were predicted to occur on the reefs. The result can contribute to cold-water coral protection at the study area. Comparing model performances is useful to provide for distribution modelling of large benthic species.
作者
佟瑞菊
袁玉珠
Tong Ruiju Yuan Yuzhu(School of Transportation, Fujian University of Technology, Fuzhou 350118, China)
出处
《福建工程学院学报》
CAS
2017年第4期393-398,共6页
Journal of Fujian University of Technology
基金
福建省中青年教师教育科研项目(JAT160323)
福建工程学院科研启动基金(GY-Z15123)
关键词
冷水柳珊瑚
潜在分布建模
随机森林模型
最大熵模型
cold-water gorgonians
potential distribution modelling
random forest（RF） model
Max Ent
