将赤潮藻生长过程中产生的荧光溶解有机物(fluorescent dissolved organic matter,FDOM)的三维荧光光谱与主成分分析相结合,尝试建立了我国沿海10种常见赤潮藻的识别测定技术.用主成分分析提取三维荧光光谱第一主成分载荷谱作为识别特征谱,建立了浮游植物荧光特征谱库,在此基础上利用Bayesian 判别方法进行识别测定.甲藻在种水平上的正确识别率≥92%,硅藻在属水平上的正确识别率≥75%,其中褐指藻属和角毛藻属的正确识别率≥90%.从结果可以看出,利用赤潮藻生长过程中的溶解有机物三维荧光光谱结合主成分分析的方法实现对赤潮藻在属水平上的识别测定是可行的.%The identification techniques for 10 species of red tide algae often found in the coastal areas of China were developed by combining the three-dimensional fluorescence spectra of fluorescence dissolved organic matter (FDOM) from the cultured red tide algae with principal component analysis. Based on the results of principal component analysis, the first principal component loading spectrum of three-dimensional fluorescence spectrum was chosen as the identification characteristic spectrum for red tide algae, and the phytoplankton fluorescence characteristic spectrum band was established. Then the 10 algae species were tested using Bayesian discriminant analysis with a correct identification rate of more than 92％ for Pyrrophyta on the level of species, and that of more than 75％ for Baeillariophyta on the level of genus in which the correct identification rates were more than 90％ for the phaeodaetylum and ehaetoeeros. The results showed that the identification techniques for 10 species of red tide algae based on the three-dimensional fluorescence spectra of FDOM from the cultured red tide algae and principal component analysis could work well.
展开▼
