傅里叶变换红外光谱结合化学计量学建立快速鉴别牛肝菌种类及测定牛肝菌中总Hg含量的方法。 采集15种共48份云南常见牛肝菌的红外光谱信息并用冷原子吸收光谱-直接测汞仪测定牛肝菌的总Hg含量, 根据FAO/WHO规定的每周Hg允许摄入量（provisional tolerable weekly intake, PTWI）评价牛肝菌的食用安全性; 采用Norris平滑、 多元散射校正、 二阶导数、 正交信号校正-微波压缩等方法对牛肝菌的红外光谱进行优化处理, 优化处理后的数据进行主成分分析、 偏最小二乘判别分析建立快速鉴别牛肝菌种类及牛肝菌总Hg含量的预测模型。 结果显示: （1）主成分分析的前三个主成分累积贡献率为77.1%, 不同种类牛肝菌在主成分得分图中能够明显区分开, 表明不同种类牛肝菌的化学组分或含量具有差异; （2）不同产地、 种类牛肝菌总Hg含量差异明显, 其总Hg含量在0.17～15.2 mg·kg-1 dw之间; 若成年人（60 kg）每周食用300 g新鲜牛肝菌则少数牛肝菌摄入的Hg超过PTWI的限量标准, 食用有一定风险; （3）牛肝菌红外光谱数据与总Hg含量拟合, 进行偏最小二乘判别分析, 能快速区分总Hg含量低（≤1.95 mg·kg-1 dw）、 中（2.05～3.9 mg·kg-1 dw）、 高（≥4.1 mg·kg-1 dw）的牛肝菌样品, 并且Hg含量差异越大, 越易于区分; 进一步建立牛肝菌总Hg含量预测模型, 训练集的R2为0.911 4, RMSEE为1.09, 验证集的R2和RMSEP分别为0.949 7和0.669 5, 牛肝菌总Hg含量预测值与测定值比较接近, 模型预测效果良好。 红外光谱结合化学计量学方法能快速鉴别牛肝菌种类, 区分不同总Hg含量的牛肝菌样品并对Hg含量进行准确预测, 为野生牛肝菌的质量控制和食用安全评估提供快速、 简便的方法。

Fourier transform infrared spectroscopy combined with chemometrics was used to establish a method for rapid identification of different species of bolete mushrooms and determination of total mercury (Hg). In this study, 15 species of bolete mushrooms were used and the information of infrared spectra of 48 samples was collected. Meanwhile, the total Hg was determined with cold-vapour atomic absorption spectroscopy and direct mercury analyzer. The food safety of bolete mushrooms was evaluated according to provisional tolerable weekly intake (PTWI) for Hg recommended by the United Nations food and agriculture organization and the World Health Organization (FAO/WHO). The original infrared spectra were optimized with Norris smooth, multiplicative signal correction (MSC), second derivative, orthogonal signal correction and wavelet compression (OSCW). The spectra data were analyzed with principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA) after the optimal pretreatment. Then the discrimination model for different species of bolete mushrooms and prediction model of Hg content were established, respectively. The results showed that: (1) The cumulative contribution of first three principal components of PCA was 77.1%. Different species of boletes can be obviously distinguished in principal component score plot. It indicated that the chemical composition or contents were different in these species of boletes. (2) There were significant differences in total Hg contents in different samples and the total Hg content in the boletes were 0.17～15.2 mg·kg-1 dry weight (dw). If adults (60 kg) ate 300 g fresh bolete mushrooms a week, Hg intakes in a few samples were higher than the PTWI standard with potential risks. (3) The infrared spectra data in combination with the total Hg content was performed by partial least squares discriminant analysis. The mushroom samples with low (≤1.95 mg·kg-1 dw), medium (2.05～3.9 mg·kg-1 dw) and high (≥4.1 mg·kg-1 dw) total Hg content could be discriminated. Moreover, the more different the Hg content was, the more easily to distinguish. In addition, the prediction model of total Hg content of boletes was established. The R2 and RMSEE of the training set were 0.911 4 and 1.09, respectively while R2 and RMSEP of validation set were 0.949 7 and 0.669 5, respectively. The predictive values of total Hg content in boletes were approximate to the measured values which showed that the model has good predictive effect. Infrared spectroscopy combined with chemometrics can be used for rapid identification of bolete species and discrimination of bolete samples with different contents of total Hg. Furthermore, the total Hg content could also be predicted, accurately. This study may provide a rapid and simple method for quality control and edible safety assessment of wild-grown bolete mushrooms.