Identification accuracy and diversity reproducibility associated with internal transcribed spacer-based fungal taxonomic library preparation

This study investigated analytical parameters that are inherently relevant to identifying and quantifying fungal communities based on polymerase chain reaction amplicons. Specifically, we evaluated the accuracy of the BLASTn-based identification for internal transcribed spacer (ITS) sequences generated from pure cultures, and quantified the reproducibility of relative abundances as well as and diversity measurements using duplicated environmental samples. The BLASTn-based method produced accurate fungal identification for the pure culture sequences at the genus rank. Percentages of the sequences assigned to correct genera were 99.8%, 99.8% and 99.9% for Alternaria alternata, Cladosporium cladosporioides and Epicoccum nigrum respectively. These fractions were smaller for Aspergillus fumigatus and Penicillium chrysogenum, which have dual nomenclatures or sibling species that are indistinguishable by ITS sequences. Our duplicate environmental analyses demonstrated that diversity and relative abundance levels were reproducible (r2>0.9), that variability decreases with increased sequence quantity, and that the differences in distinct environmental samples were larger than differences in replicate samples ( diversity). These results serve to better characterize the identification and quantification limits of ITS-based fungal taxonomic studies, and demonstrate that while diversity quantification is reproducible, limitations in ITS-based taxonomic identification and dual nomenclature conventions are current barriers to identification accuracy.