基于仪器检测和大数据分析的云南大叶种茶叶滋味和香气致味物质组成模型的研究

The main tea product of Yunnan big-leave tea is Pu-erh tea. However, there is no quantitative and qualitative way or model to evaluate the quality of Pu-erh tea at present yet. And the mechanism of the formation and transformation of tea taste, flavor, etc., are largely unknown yet. These constraints are critical for the quantitative evaluation, standardizing processing techniques and hierarchization of Pu-erh tea. The quality of the tea, including taste and aroma, are linked to hundreds of mixed chemical compounds, and they are dramatically and continuously changing during processing steps, fermentation and post-fermentation, and their contribution to the final taste and quality of the tea is very complicated. Previously, we established a high-efficient HS-SPME/GC–MS with high resolution and high sensitivity to analyze the volatile compounds in teas. By using that method and combined with big data analysis methods, we have made dozens of research works, e.g. the aging of Pu-erh tea, the influence of processing techniques on the volatile chemicals, patterns of Pu-erh tea from typical tea mountains, etc. In this project, we will use neural network analysis techniques to find out the material composition pattern for the taste and odor of Yunnan big-leave tea from large scale data, for quantization of quality control and evaluation of tea. Four main questions will be discussed in this project: 1. The taste and flavor of tea are the final expression of hundreds of mixed chemical compounds in tea infusion. The contribution of these compounds for the final presentation is complicated and influenced by each other and largely unclear. We will try to draw some patterns for the contribution of main compounds to the final taste and odor in tea infusion. 2. The formation, transformation mechanism of these tea chemicals during processing, fermentation, and post-fermentation. 3. The influence of environmental factors such as soil and climate on the quality of final tea products, which can provide references to the selection and control of quality types at tea tree variety and cultivation. 4. The quantitative and qualitative model for evaluation the taste and flavor of some typical Yunnan tea products by using analytical instruments.

云南大叶茶的风味化学基础及其与加工、陈化关系等尚未有基础模型，严重制约了普洱茶品质评价定量化、加工标准化、产品等级化。困难在于茶叶有数百种致味物质，存在复杂的相互作用。本项目组建立了茶叶高分辨质谱检测方法并获得茶叶挥发性物质大数据，在致香组成特征与茶叶陈化年份、产地、加工工艺的影响等方面的研究上取得了一系列成绩，并在应用神经网络分析方法用于复杂多维度生物学数据处理方面做了一些尝试和积累。本项目在此基础上开展云南大叶种茶叶滋味和香气的物质基础模型构建研究，重点探讨以下4个方面的问题：①致味和致香是复杂体系的表观呈现，感官感受和各类主要物质组合规律之间的关系；②致味物质、致香物质在加工和陈化各环节中的变化规律；③土壤、气候等环境因素对茶叶品质影响的规律，从而能在茶树源头上对品质类型的选择及控制提供参考；④个别典型茶叶滋味和茶香质量控制关键物质及量化规律。

普洱茶是云南临沧市、普洱市、西双版纳州等地区的重要甚至支柱产业，并形成了一定地域特征的茶叶产品风格。但消费者喜欢的普洱茶品质风格形成与种植、栽培管理因素的研究一直缺乏，也对于茶叶品质特征在土地选择、土壤优化等的优化尚属空白。本项目“基于仪器检测和大数据分析的云南大叶种茶叶滋味和香气致味物质组成模型的研究”，通过对土壤气候、茶叶水溶性致味物质、挥发性香味物质的关联分析，发现了一定的规律：土壤总肥力高，会提升含糖量，也会导致茶多酚物质相对较低，茶叶相对甜度更高但刺激感更低，这是普洱市茶叶普遍的特征；更高的紫外线强度会提升茶叶总多酚含量，同时晒青环节也会提高茶叶具有接近茶叶烘焙的成分，形成临沧高海拔所谓“高原糖香味”，导致茶叶滋味饱满度、回甜感受强烈。普遍认可的古树茶有更好的滋味感受，我们的研究表明，这很可能与古茶园更高的土壤重金属及其构成有关，古茶园，可以理解为较少的人工干预，较少的生物量移除，因此土壤重金属处在相对较高的水平。植物生长在含有重金属的土壤中，由于植物根系分泌的有机酸，以及根系微生物的生命活动和降水淋洗等因素，以矿物质形态存在的重金属化合物会缓慢溶解释放，被植物根系吸收、转运到植物的各个部位，茶树为抵抗重金属的胁迫，形成了一定特色的体内次生代谢产物的积累状态。现代茶园，会频繁采摘、修剪枝条等形势移除生物量，经过数年乃至数十年的过程，土壤中的重金属含量及可溶解释放的量逐步降低到一个更低的水平。比如易武古茶园土壤重金属Fe、Mn、Cu 、As显著高于其他地区。当然，古茶园土壤重金属含量相对较高并不意味着茶叶本身的重金属超标，而且现代茶叶的冲饮方式茶叶中重金属的浸出比例低于30%，使得饮用更加安全。基于以上发现，我们在普洱市的一个茶园进行了尝试，通过改变土壤肥力结构，现在提高了茶叶的香气感受、滋味饱满度。