关键词: 农户, 农户, 社会网络关系, 评价体系, 变异系数法

Abstract: To define the differences and influencing factors of phenol release amounts in mainstream smoke of flue-cured tobacco leaves among different parts and different producing areas, tobacco leaves of three stem positions (upper, middle and lower) from five main tobacco producing areas of China were sampled, the release amounts of phenol in mainstream cigarette smoke and some chemical components of tobacco leaves were determined, variance analysis, correlation analysis, multiple linear regression analysis were conducted. On the basis of correlation analysis, adding experiments with the chemical components which were highly correlated with phenol release amounts in mainstream cigarette smoke were done. Results showed that the flue-cured tobacco leaves from Huanghuai region had the highest phenol release amount in the mainstream cigarette smoke, the tobacco leaves from the southeast region had the second, and those from the other three tobacco regions had much lower phenol release amounts. The release amounts of phenol in the mainstream cigarette smoke from leaves of different stem positions had an order as: upper leaves> middle leaves> lower leaves. The release amount of phenol in the mainstream cigarette smoke had highly and significantly negative correlation with total sugar, reducing sugar, nonvolatile organic acids, hexane extract, K/Cl and pH value. The release amount of phenol in the mainstream cigarette smoke had highly and significantly positive correlation with nicotine, total alkaloids, volatile bases, total nitrogen, ammonia, free amino acid, chlorogenic acid, higher aliphatic acid and total solanesol. The release amount of phenol in the mainstream cigarette smoke had significantly positive correlation with cellulose, scopoletin, rutin, and had no obvious correlation with tobacco pectin and ether extract. The regression model of the phenol release amounts in the mainstream cigarette smoke was established by multiple regression analysis, which could explain 80.4% changes of the dependent variable, and could accurately predict the phenol release amount in mainstream smoke. Experimental results of sucrose, glucose, malic acid, citric acid, chlorogenic acid, nicotine and solanesol as additives were similar to the correlation analysis data. However, the experimental result of proline as the additive disagreed with the correlation analysis data.