基于稳定同位素特征与土壤性质的有机茶身份辨识研究

为探究云南普洱地区有机茶身份辨识特征,完善高原有机茶身份辨识技术,选取云南省普洱市的5个代表性有机种植认证茶园（ZX、LS、CYT、YS、HL）和相邻的5个常规茶园（C-1、C-2、C-3、C-4、C-5）,采集300余份茶叶和土壤样本,结合稳定同位素比例质谱仪、元素分析仪等测定茶叶和土壤的δ¹⁵N丰度、δ¹³C丰度、全氮、全碳、有机碳、碱解氮等指标,通过因子分析和判别分析的方法构建了有机茶叶同位素-土壤性质判别模型,并使用Holdout法进行验证。结果表明：不同茶园有机茶叶与相邻常规茶叶相比未表现出一致的δ¹⁵N,δ¹³C值差异规律,因此,使用单一稳定同位素因子难以鉴别有机茶。通过进一步对茶叶和土壤性质进行因子分析,筛选出6个适合判别的特征向量,构建了基于同位素特征和土壤性质的多元有机判别模型,其判别函数方程为：y=-0.081（茶叶δ¹⁵N）＋0.819（茶叶δ¹³C）＋0.093（土壤全氮）＋2.117（土壤全碳）－0.155（土壤有机碳）＋2.870（土壤pH）＋4.232,y<0.5为常规茶,判别准确度为92.3%。综上,基于同位素特征值和土壤性质参数指标构建的多元有机判别模型可实现云南省普洱市高原有机茶身份的准确判别。

Identification of organic tea based on stable isotope characteristics and soil properties

In order to explore the identity identification characteristics of organic tea in Pu''er area of Yunnan Province and improve the identity identification technology of organic tea on the plateau， 5 representative organic planting certified tea gardens （ZX、LS、CYT、YS、HL） and adjacent conventional tea gardens （C-1、C-2、C-3、C-4、C-5） in Pu''er City of Yunnan Province were selected， and more than 300 tea and soil samples were collected. δ¹⁵N abundance， δ¹³C abundance， total nitrogen， total carbon， organic carbon， alkali hydrolyzable nitrogen and other indexes of tea and soil were determined by stable isotope mass spectrometer and element analyzer. The isotope-soil property discrimination model of organic tea was constructed by factor analysis and discriminant analysis， and verified by Holdout method. The results show that there is no consistent difference in δ¹⁵N and δ¹³C values between organic tea and adjacent conventional tea in different tea gardens， so it is difficult to identify organic tea by using a single stable isotope factor. Through further factor analysis of tea and soil properties， six feature vectors suitable for discrimination were selected， and a multivariate organic discrimination model based on isotope characteristics and soil properties was constructed， and the function is as follows： y=-0.081 tea δ¹⁵N+0.819 tea δ¹³C+0.093 soil total nitrogen +2.117 soil total carbon -0.155 soil organic carbon +2.870 soil pH+4.232， y＜0.5 is convention tea.The model achieves an impressive accuracy rate of 92.3%. In summary， the multivariate organic discrimination model constructed based on isotopic characteristic values and soil property parameter indicators can accurately discriminate the identity of organic tea from the Pu''er plateau in Yunnan.