How natural should anamorph genera be?

Abstract

The taxonomy of anamorphic fungi has always been artificial, as a concession to the practical needs of identifying and naming important organisms, because of insufficient characters indicative of a more natural classification. Integration of anamorph taxa into a teleomorph classification is best served by retaining anamorph names for fungi that have no teleomorph fructification. Proposals to split more or less heterogeneous anamorph genera such as Aspergillus, Fusarium, Acremonium, Verticillium, and Gliocladium into more natural units have prompted this assay. While subdivision of these genera can be defended only partly, some nematophagous genera with affinities to Arthrobotrys or Monacrosporium are considered for lumping, to achieve a more natural classification. This raises the question of to what extent anamorph genera can be delimited naturally and how far they must reflect teleomorph associations. A consistent application of natural genus concepts for anamorphic fungi, even if that were possible, would completely upset nomenclature. While the segregation of morphologically divergent, obviously unrelated taxa from an anamorph genus is defendable, some admittedly deviating taxa (different teleomorphs) should be tolerated in somewhat artificial genera for the sake of identification. Conversely, lumping all anamorphs associated with one teleomorph genus into one genus is not supported when the criteria used for identification clearly favour splitting; otherwise it would render identification according to morphological criteria impossible. Therefore classification of anamorph genera cannot aim at genera that adequately reflect natural relationships and the most convenient generic delimitation must be considered for individual cases. Key words: anamorph, teleomorph, classification, nomenclature, Deuteromycetes, pleomorphism.