摘要
为探讨小盾片几何形态特征对金龟子分类鉴定的可行性和可靠性,以金龟子小盾片的几何形态测量特征为依据,对10种鳃金龟进行分类识别,利用TpsDig2软件在10种金龟子各20个样本的小盾片上,分别提取8个标记点,再利用PAST软件进行普氏叠加和薄板样条分析,消除大小、方向等非形状因素的影响后,得到小盾片8标记点形状的扭曲图像,最后利用SPSS 22.0进行判别分析。结果表明:对10种鳃金龟小盾片8个标记点坐标数据进行普氏叠加和薄板样条分析后,得到标记点位置折线图和扭曲图像,可直观看出10种鳃金龟小盾片形态差异主要发生在标记点2和6处;利用小盾片8个标记点坐标数据形成的几何形态特征进行判别分析,原始判别和交叉判别正确率均高达99.0%。因此,利用小盾片的几何形态特征可以对10种鳃金龟进行准确识别。
In order to provide the theoretical basis for the rapid and accurate digital automatic identification of Melolonthidae with the geometric morphometrics of scutellum.A total of 10 species of Melolonthidae with 20 samples of each species are used to extract 8 markers of scutellum by TpsDig2 software.PAST and SPSS software are used for Procrustes superposition,thin plate spline,principal component and discriminant analysis.The results show:The line chart of scutellum with 8 markers after Procrustes superposition and the figure of thin plate spline display the differences between the 10 species of Melolonthidae.The accuracies of both raw discriminant and cross discriminant of 10 species of Melolonthidae,are 99.0%.Therefore,the 10 species of Melolonthidae could be accurately identified by using the geometric morphometrics features of the scutellum.
作者
苏筱雨
蔡小娜
周健
黄大庄
SU Xiaoyu;CAI Xiaona;ZHOU Jian;HUANG Dazhuang(College of Forestry,Agricultural University of Hebei,Baoding071000,China;Department of Basic Course,Hebei Finance University,Baoding071051,China;College of Landscape Architecture and Tourism,Agricultural University of Hebei,Baoding071000,China)
出处
《中国农业大学学报》
CAS
CSCD
北大核心
2019年第8期108-113,共6页
Journal of China Agricultural University
基金
高等学校博士学科点基金(20131302110008)
关键词
鳃金龟科
小盾片
几何形态测量学
分类
Melolonthidae
scutellum
geometric morphometrics
classification