论文部分内容阅读
为研究夜蛾右前翅的数学形态特征作为分类依据的可行性,以夜蛾科20种夜蛾(每种20个)的右前翅为对象,应用BugShape软件提取其偏心率、球状性、似圆度、矩形度、延长度、叶状性和不变矩Hu1、Hu2等13项数学形态特征,对13项数学形态特征平均值进行方差分析,结果表明:13项数学形态特征参数组间差异极显著(P<0.01),都可以对夜蛾科进行分类鉴定;400个样本13个特征参数统计分析后,其中矩形度、偏心率、似圆度、延长度、叶状性、Hu1、Hu6、球状性、Hu4和Hu7等10个可以作为分类依据的指标,其贡献大小为矩形度>偏心率>似圆度>Hu1>Hu4>球状性>延长度>叶状性>Hu6>Hu7。用筛选出的10个指标建立判别函数和分层聚类,其判别正确率为84.3%,聚类分析的亲缘关系图谱和传统分类学分类结果基本相符。夜蛾右前翅的数学形态特征可以作为夜蛾科进行分类鉴定的依据。
In order to study the feasibility of the morphological characteristics of the right front wing of the noctuid moth as a basis for classification, the right front wing of 20 species of noctuid moth (20 species each) was selected as the object. BugShape software was used to extract the eccentricity, Degree, squareness, extension, leafiness and moment invariant Hu1, Hu2 and other 13 mathematical morphological characteristics of the average of 13 mathematical morphological characteristics of variance analysis showed that: 13 mathematical morphological characteristics of the difference between the parameters Hu1, Hu6, Hu1, Hu1, Hu1, Hu1 and Hu1 were all significantly different (P <0.01). After the statistical analysis of 13 characteristic parameters of 400 samples, Hu4 and Hu7, which can be used as index of classification, the contribution is rectangle> eccentricity> circularity> Hu1> Hu4> globality> elongation> leafiness> Hu6> Hu7. The discriminant function and hierarchical clustering were established with the selected 10 indexes, the discrimination accuracy was 84.3%. The phylogenetic relationship of the cluster analysis was basically consistent with the traditional taxonomic classification results. The mathematical morphology of the right wing of the noctuid may be used as the basis for classification and identification of the noctua.