分类单元
生物
线性判别分析
主成分分析
分类学(生物学)
数值分类学
古生物学
判别函数分析
植物
进化生物学
人工智能
数学
统计
计算机科学
作者
Jingjing Wang,Cunlin Xin,Lu-Han Wang,Yamei Zhang,Zhi-Peng Jiao,Guo-Yun Di,Song-xin Liu
标识
DOI:10.1080/08912963.2021.1999939
摘要
The categories of Ginkgophyta fossil plants are varied and distributed all over the world. Due to wide variations in leaf division along with the relatively similar epidermal structure between genera and species, it is difficult to identify Ginkgophyta fossils accurately. At this point, the current study sheds light on the quantitative numerical taxonomy of these plant fossils, depending on characterisation of fossil morphology and epidermal microstructures. Through principal component analysis and correlation analysis of specific characters, 15 traits are determined as the key characters in the identification of Ginkgophyta fossils. Cluster analysis was also used to divide the operational taxonomic units into 3 categories and 16 representative groups, which mostly correspond to the traditional classification, with some inconsistencies. Consequently, 8 genera of Ginkgophyta fossils are established as two new taxa: Ginkgophyllum, Rhipidopsis, Sinophyllum,Psygmophyllum and Pseudotorellia, Vittifoliolum, Eretmophyllum, Glossophyllum. In addition, Dicranophyllum should be assigned to Coniferophytina, and Saportaea nervosa can be temporarily attributed to Ginkgo.Finally, the Bayes discriminant model was established for the 38 species of Ginkgophyta fossil plants in 7 genera with relatively certain classification positions, and the discriminant model was tested using literature data. It is believed that the discriminant model has certain accuracy and can be applied to the classification and identification of Ginkgophyta fossil plants in the future.
科研通智能强力驱动
Strongly Powered by AbleSci AI