Beak identification of four dominant octopus species in the East China Sea based on traditional measurements and geometric morphometrics

文献类型： 外文期刊

第一作者： Fang, Zhou

作者： Fang, Zhou;Fan, Jiangtao;Chen, Xinjun;Chen, Yangyang;Fang, Zhou;Fan, Jiangtao;Fang, Zhou;Chen, Xinjun;Fang, Zhou;Chen, Xinjun;Fang, Zhou;Chen, Xinjun

作者机构：

关键词： Octopus; Species identification; Beak measurement; Geometric morphometircs; Machine learning

期刊名称：FISHERIES SCIENCE （ 影响因子：1.617； 五年影响因子：1.531 ）

ISSN： 0919-9268

年卷期： 2018 年 84 卷 6 期

页码：

收录情况： SCI

摘要： Octopus is the most abundant genus in the family Octopodidae and accounts for more than half of the total cephalopod landing in neritic fisheries. A taxonomic problem still exists due to synonymous scientific names and limited genetic information. The cephalopod beak is a stable structure that allows an effective solution to the problem of the species and stock identification. Beak shape variation has been more widely considered than beak measurements in recent years. In this study, with the beak as the experimental material, we combined geometric morphometrics (GM) with machine learning methods and compared the discrimination results obtained by traditional and GM methods in four Chinese neritic octopus species (Amphioctopus fangsiao, Amphioctopus ovulum, Octopus minor and Octopus sinensis). According to our analyses, Octopus sinensis has the larger beak size [both upper beak (UB) and lower beak (LB)] than other species. The results of ANOVA showed that all beak measurements differed significantly among the four species. Significant differences in both UB and LB shapes among four species were identified in MANOVA analysis based on the GM method. The results of GM-based discriminant analysis were better than those of traditional measurements, and machine learning methods also showed the higher correct classification rates than linear discriminant analysis. GM is a useful method to reconstruct the shape cephalopod beak and can also effectively distinguish different species. We should improve classification accuracy with machine learning methods for determining species structure in the future.

分类号：

  • 相关文献
作者其他论文 更多>>

微信小程序