A similar number of OTUs (30–32) was identified for each diet. Good’s coverage of the combined library was 91.1%, while the coverage for the alfalfa, orchardgrass and concentrate libraries was 83.8%, 88.1% and 85.2%, respectively (Table 3). Although the Chao1 estimation was lower for the orchardgrass, the predicted OTUs and the overall level of diversity estimation by the Shannon index were higher for the alfalfa and orchardgrass hay libraries (Table 3), which correlated with the DGGE observation IWR-1 in vivo (Fig. 1). Among the 77 (24.6%, 2 OTUs) clone sequences that showed 97% or more sequence similarity with cultured Treponema, 76 were related to T. bryantii. Only a single sequence related to T. zioleckii

and no sequences having 97% or more similarity

with T. saccharophilum were found. The majority of clones (236 clones, 75.4%) were related to uncultured Treponema, irrespective of diet (Table 3). Among the uncultured Treponema, 70 clones had 97% or more similarity with sequences of uncultured Treponema clones, while 166 clones showed 86–96% similarity (Table 3) with any sequence in the NCBI database. Pairwise comparison of each 16S rRNA gene library using web-libshuff confirmed that the libraries were significantly (P=0.001) different from one another selleckchem (data not shown). The results of a phylogenetic analysis of the 67 OTUs identified among the combined 16S rRNA gene sequences from the three libraries are shown in Fig. 3. The phylogenetic tree (Fig. 3) was divided into two major clades

(clades I and II). Additionally, clade II was further categorized in to subclades (a–e), although this was not supported by higher bootstrap values. The distribution of clones in the different clades was shown by pie charts with the size of the pie charts corresponding to the size of the clones in each clade. In clade I, 59 clones (58.4%) were from the concentrate PAK5 library, while in clade II 185 clones (87.3%) were from the hay libraries. 16S rRNA gene-based clone libraries constructed using universal PCR primers have been used to monitor the entire rumen bacterial community (Whitford et al., 1998; Tajima et al., 1999; Koike et al., 2003; Sundset et al., 2007). However, such universal libraries do not sufficiently represent the diversity of specific groups of bacteria in a complex gut environment (Li et al., 2008). Our recent analysis of the rumen Prevotella community based on group-specific clone libraries showed the abundance of novel rumen Prevotella previously undetected (Bekele et al., 2010), indicating the advantage of this approach. In the present study, we focused on Treponema, a frequently detected rumen bacterial group that has been implicated in the degradation of fiber (Koike et al., 2003; Shinkai et al., 2010). A Treponema group-specific primer was successfully developed and used to illustrate the diversity and molecular ecology of rumen Treponema.