Abstract: In order to compare the microbial communities on tobacco leaves between clone library method and highthroughput sequencing method, we analyzed the diversity of microbial communities based on bacterial 16S rDNA and fungal ITS on tobacco leaves collected from Fujian province in China. The results showed that more numbers of bacterial and fungal sequences were detected by highthroughput sequencing than clone library. The number of operational taxonomic units (OTUs) clustered from highthroughput sequencing method was almost the same as that from clone library method in bacterial communities, while that was higher from highthroughput sequencing in fungal communities. The rarefaction curves drawn from highthroughput sequencing method tended to approach the saturation plateau. The expected community diversity indices (Chao1 and Ace) were higher in clone library method, whereas the observed community diversity indices (Simpson and Shannon) suggested that higher bacterial diversities were detected through clone library method and higher fungal diversities were detected through highthroughput sequencing method. There were more kinds of bacteria and fungi genera identified through highthroughput sequencing method. The dominant genera of bacteria were similar detected from both of the methods, including Pseudomonas, Acinetobacter and Sphingonomonas. However, the fungal genera detected through these two methods were significantly different. Overall, novel highthroughput sequencing methods outperform clone library approaches in terms of resolution and magnitude. They enable identification and relative quantification of community members of tobacco leaves and offer new insights into environmental microbiology. 

Key words: clone library, highthroughput sequencing, bacteria, fungi, tobacco leaves