Method or madness : does OTU delineation bias our perceptions of fungal ecology?

Commentary: Arbuscular mycorrhizal (AM) fungi are among the most widespread and studied terrestrial mutualisms (Smith & Read, 2008). As molecular tools have become widely available, the application of these tools to characterize communities of AM fungi has met with significant obstacles and complications. The low number of ‘known’ AM fungal species is largely derived from morphology based on spore traits from culturable isolates, but molecular techniques are starting to reveal cryptic and unculturable diversity necessitating the creation of ‘virtual taxa’, currently without morphological analogues (Opik et al., 2006). In the absence of an operational species concept, delineating ‘species’ appropriately is essential for mycorrhizal ecologists seeking to compare the diversity of AM fungi both in nature and experiments, particularly for studies using high-throughput DNA sequencing. However, not all approaches for delineating environmental DNA sequences into operational taxonomic units (OTUs) are equivalent and each can result in substantially different descriptions of the fungal communities under study. In this issue of New Phytologist, Lekberg et al. (2014, pp. 1101–1104) compare two methods with high through-put AM fungal DNA sequences and ask whether the methodology used impacts our inferences about the processes driving AM fungal community structure.