“…Consequently, there is also substantial debate over which sequence similarity level gives the biologically most meaningful units (see references above). Typically, clustering thresholds in the range 95%–99% are used in microbial studies (Bonito et al., ; Geml et al., ; Lorberau et al., ; Nemergut et al., ; Vik et al., ). Several studies have addressed the performance of clustering algorithms, including choice of clustering thresholds, in alpha diversity (species richness) estimation (e.g., Koeppel & Wu, ; Tikhonov et al., ; Yamamoto & Bibby, ), while the relationship between OTU similarity thresholds and patterns of beta diversity (compositional turnover) has so far been addressed in few studies (Bik, Halanych, Sharma, & Thomas, ; Frøslev et al., ; Koeppel & Wu, ; Lekberg, Gibbons, & Rosendahl, ).…”