水华暴发期蓝藻胞外有机物释放过程的指纹荧光光谱解析及其在水质安全预警中的应用

Based on the understanding that cyanobacterial cell extracellular organic matter (EOM) and algal toxins, odor substances belong to the same release process and their release ability is affcted by similar factors, a novel idea was proposed that algal toxins concentration in water can be retrieved or predicted based on EOM components analysis. The release process of EOM components from microcystis aeruginosa in water bloom period will be analyzed by fluorescence fingerprint spectra technique combined with chemometrics methods; those EOM components that significant correlated with algal toxins concentration currently and on time series will be screened out, and concentration retrieving and predicting models for algal toxins will be builded based on linear mathematical model (multiple linear regression model, vector autoregression model) and nonlinear mathematical model(BP NN model). With actual measurement data in the water field where cyanobacterial water bloom actually occured, the concentration retrieving and predicting models for algal toxins will be further improved combing water quality and weather parameters. In this project, fluorescence fingerprint spectra that can be determined on-line and in time was applied for analyzing EOM components and concentration, therefore on-line monitoring and early-warning for water quality damage caused by cyanobacterial bloom can be realized with the retrieving and predictiving model.

基于蓝藻细胞胞外有机物（EOM）与藻毒素、异味物质同属一个动态释放过程的认识，提出了一种基于EOM组分分析的水体藻毒素浓度反演及预测的新思路。利用指纹荧光光谱技术，结合化学计量学方法解析水华暴发期铜绿微囊藻EOM组分的释放过程；通过探索EOM各组分浓度与藻毒素浓度间的相关性，筛选获得与藻毒素浓度当期和时间序列上显著相关的EOM组分；尝试采用线性数学模型（多元线性回归模型和向量自回归模型）以及非线性数学模型（BP神经网络模型）构建藻毒素浓度的反演及预测模型；根据野外蓝藻水华实际发生水域的实际测量数据，结合水质、气象参数，对藻毒素浓度反演及预测模型进行修正。本课题利用可在线实时测量的指纹荧光光谱解析EOM组分及浓度，结合相关模型应用，为蓝藻水华水质致害过程的在线监测与预警提供了可能。

本课题利用三维荧光光谱技术，结合化学计量学方法，对蓝藻胞外有机物（EOM）的指纹荧光光谱特征、光谱组分信息及其释放过程进行了分析，解析了蓝藻生长过程中EOM的动态释放规律；通过对微囊藻毒素MC-LR浓度与EOM光谱组分间动态释放关联性的分析，筛选出了与MC-LR浓度显著相关的EOM组分；构建了一种新颖的水体MC-LR浓度反演模型。研究结果表明，蓝藻释放的EOM主要包括类蛋白和类腐殖质两类组分，二者具有相似的动态释放规律，在蓝藻衰亡期出现集中大量释放，而结合态胞外有机物（bEOM）中类腐殖质含量的快速上升可以作为蓝藻细胞结构破碎的前兆标志；溶解态胞外有机物（dEOM）中类蛋白荧光峰T和类腐殖质荧光峰C是与MC-LR浓度最显著相关的两种EOM组分；以dEOM中类蛋白T或者类腐殖质C的荧光强度，结合叶绿素a浓度，可以快速准确地反演蓝藻释放的MC-LR浓度。相关的研究结果，可以为利用荧光技术在线监测蓝藻水华致害过程，保障饮用水的安全提供理论和方法基础。