摩洛哥南部沿岸两种沙丁鱼耳石形态识别的初步研究

Using otolith morphology to identify Sardinella aurita and Sardina pilchardus in south coastal waters of Morocco

摘要:

根据2011年6-8月我国拖网船在摩洛哥南部沿岸采集的46尾金色小沙丁鱼（Sardinella aurita）和53尾沙丁鱼（Sardina pilchardus）的耳石样本，利用传统形态测量法和傅立叶分析法分别进行分析，对不同的耳石形态进行判别。传统形态测量法测量了6个耳石形态参数，并利用t检验比较两种沙丁鱼耳石形态差异；耳石轮廓经图像化处理后利用软件转化为椭圆傅立叶标码（EFDs），最后分别采用6个耳石形态参数和标准化后的77个傅立叶系数进行主成分分析和判别分析。结果表明，两种沙丁鱼耳石形态在全长（TL）、宽度（TW）、背宽（DW）、腹噱长（VL）中存在显著差异（P<0.01），背长（DL）和翼叶长（WL）不存在差异（P>0.05）；主成分分析表明，传统测量法的前两个主成分累积贡献率达77.0%，第一、二主成分的最高变量分别为全长（TL/FL）和背宽（DW/FL），在散点图中显示出了很好的区分性；而傅立叶分析法中，前16个主成分累积贡献率为82.8%，前两个主成分因子仍有较大的重叠，区分性相对较差。判别分析结果认为，传统测量法中的3个参数值进入了逐步判别分析中，总判别正确率为83.76%；傅立叶分析法的77个傅立叶系数中，有9个系数进入了逐步判别分析中，总判别正确率为92.02%。总体来看，相比传统形态测量法，傅立叶方法在鱼类耳石的判别分析中更为有效。

Abstract:

Based on the 46 samples of Sardinella aurita and 53 samples of Sardina pilchardus taken from the Chinese trawling vessel working in the Morocco coastal waters from June to August in 2011, their otolith morphology was analyzed based on traditional measurement and Fourier analysis. Six of variables were measured on the otolith and differences were compared by t-test; digitalized photographs of otolith were transformed to Ellipse Fourier descriptors (EFDs). Finally, six variables and 77 standardized EFDs were analyzed by principal component analysis (PCA) and stepwise discriminant analysis (SDA). The results showed that the total length (TL), total width (TW), dorsal width (DW), and ventral length (VL) had significant differences between two species (P<0.01), whereas dorsal length (DL) and wing length (WL) had no difference (P>0.05). Result of PCA showed that the first and second components made total contribution of mutation for 77.0% by traditional measurement, and the highest variables were total length/fork length (TL/FL) and dorsal width/fork length (DW/FL) respectively, the scatter-plot reveal a good dispersion; there were 16 principal components with 82.8% cumulative contribution by Fourier analysis, but the first two components had too much overlap with a bad separation. SDA showed that three variables of traditional measurement entered in the SDA, total correct classification was 83.76%; 9 harmonics of 77 EFDs entered in SDA with Fourier analysis, and the total correct classification was 92.02%. Overall, Fourier analysis is more effective than traditional measurement in the fish otolith morphology classification.