ABSTRACT
The taxonomic usefulness of carbon substrate utilization patterns was assessed for 81 strains (including 13 duplicate strains) of Penicillium species representing 18 morphologically defined species. Standardized conidial suspensions were inoculated into BiologrmYT= plates, and growth for each substrate was assessed after 4 days incubation using a microplate reader. The resulting dendrograms were inconsistent with accepted infrageneric classifications, demonstrating that physiological similarities do not necessarily reflect phylogenetic relationships. Responses for replicates of any strain are consistent at a 90-100% level of similarity. Some species are well-defined at an approximately 80% similarity level, but strains of other species exhibit more variation and do not cluster together. In the Penicillium aurantiogriseum group, the amount of variation between strains of any one species seems to overshadow differences among species, although some species, such as P. melanoconidium, appear well-defined. Cultures tentatively assigned to the Penicillium glabrum/spinulosum appear to be divisible into at least four distinct clusters, with the majority of the variation occurring in P. spinulosum. The sensitivity of this technique makes it a useful adjunct to morphological analysis, by rapidly providing a completely independent data set that can be used as an aid for interpreting colony or micromorphological variation. The probability that the technique could be adapted as an efficient identification system for Penicillium is discussed.
