Application of stochastic Bayesian modeling to assess quality and safety profile of tea in China market

Abstract

Limited information was available on the quality and safety profile of tea in China, even though food quality and safety of tea is an issue of great concern to consumers and the public. Stochastic Bayesian modeling was employed in this study to investigate food quality and safety profile of tea in China market. Results indicated that the overall nonconforming rate (95 % CI, confidence interval) of tea was 2.6 ‰ (1.3 ‰ to 4.5 ‰), indicating a high level for the quality and safety status of tea in market. Pesticide residues exceeding maximal limit (MRLs) was the major cause leading to nonconformity of tea, with estimated incidences (95 % CI) of 11.7 ‰ (0.5 ‰ to 53.2 ‰), 1.1 ‰ (0.4 ‰ to 2.0 ‰), 2.3 ‰ (0.6 ‰ to 5.5 ‰), and 0.55 ‰ (0.06 ‰ to 1.51 ‰), respectively, for Yellow tea, Green tea, and Oolong tea; whereas illegal application of additives was more likely to occur in black tea, with an incidence of 1.1 ‰ (0.4 ‰ to 2.7 ‰), despite that an estimated incidence of 2.9 ‰ (0.4 ‰ to 16.1 ‰) was generated for Yellow tea due to fewer inspects and small sample sizes involved for analysis. The difference in quality and safety status of tea between provinces was statistically insignificant, as indicated by overlapped 95 % CIs of nonconforming rates. Results of this study provided reliable information on quality and safety profile of tea in China market, and suggested that application of pesticides during tea plantation and illegal use of additives during black tea production worth more attention in respect of quality and safety of tea.