Assessment of diversity (including both species richness and intraspecific variation) within a habitat may often be difficult when morphological criteria are used for identification and classification of isolates. Here we describe how combined fatty acid and sterol profiles (FAST-profiles) can be used for classification of fungal isolates into FAST-groups (i.e. operational chemotaxonomic units) according to a defined upper variation limit. Data analysis includes a two stage statistical approach. First, the isolates are grouped into taxonomic species using a discriminant model based on well identified ‘model’ isolates. These species groups together with one group that includes isolates which could not be identified with the discriminant model are further divided into operational chemotaxonomic units using a FAST-profile mismatch threshold. This method is demonstrated with Norway spruce needle endophytes.

The model isolates were obtained from fallen spruce needles and identified by morphological criteria. Five species fruited on the needles: Lophodermium piceae, Rhizosphaera kalkhoffii, Lirula macrospora, Tiarosporella parca, and Thysanophora penicillioides. Additionally, mycelia of Sclerophoma pythiophila were frequently noted. These model isolates were separated by their FAST-profiles by applying a multivariate discriminant model into six distinct non-overlapping clusters corresponding to their species. This model was used for identification of 1740 isolates obtained from surface sterilized Norway spruce needles. As an example, isolates identified as T. parca were further divided into FAST-groups by applying a FAST-profile mismatch threshold value. This threshold value was the average within-species mismatch-value obtained for isolates identified by morphological criteria. This method allows a division of any batch of fungi cultivable in vitro into operational chemotaxonomic units according to a defined upper variation limit.