ANAMORPHS AND TAXONOMY OF THE AMPHISPHAERIAGEAE (SENS. LAT.)

The Amphisphaeriaceae (Sphaeriales), in its broadest sense (Müller & Arx 1962) is usually negatively defined to include all the nonxylariaceous and nondiatrypaceous fungi that have an amyloid apical ring in their asci. The familial concept has been refined by Barr (1975, 1976) and Krug (1977) on perithecial characters.

Unfortunately very few species of the family, in its broadest sense, have been linked to anamorphs (see reviews in Barr 1975, Brockmann 1975, Nag Raj & Kendrick 1985). Amphisphaeria umbrina (Fr.) de Not., the type species of the family, has not been linked to an anamorph and no species of Amphisphaeria has been unequivocally linked to an anamorph through pure culture techniques. We have grown two New Zealand collections of Amphisphaeria multipunctata (Fuckel) Petrak (PDD 36845, 36846) ["Amphisphaeria millepunctata" (Fuckel) Petrak and the so-called basionym, "Diaporthe millepunctata" Fuckel, that Petrak published (Ann. Mycol. 21: 329. 1923) were in error and the erroneous spelling has been perpetuated. The original spelling reads: Diaporthe multipunctata Fuckel, Jahrb. Nassauischen Ver. Naturk. 27/28: 37. 1873.] from ascospores; both isolates produced perithecia on sterilized apple twigs but neither formed an anamorph. This is not surprising as other members of the Amphisphaeriaceae (e.g. Cainia Arx & Muller, Leiosphaerella Höhnel) have failed to form anamorphs in culture. Nag Raj (1977) found pycnidia of Bleptosporium pleurochaetum (Speg.) Sutton intimately associated with perithecia of A. argentinensis Nag Raj, a species that he considered to be closely related to A. umbrinum.

One consequence of this general lack of information about anamorphs in the Amphisphaeriaceae is that clues manifested by anamorphs and used to clarify the taxonomy of other orders of ascomycetes (e.g. see Hughes 1976, Samuels & Rossman 1979, Samuels & Seifert, In press) are not available for an assessment of amphisphaeriaceous species. In the present series of papers (Samuels et al. 1987, Samuels & Rossman 1987 and herewith) we have documented anamorph-teleomorph connections for several amphispheriaceous ascomycetes. From these observations and from previously published work (Glawe & Rogers 1982a,b; 1986 and summaries in Rogers 1979, Barr 1975, Brockmann 1975, Nag Raj & Kendrick 1985) the anamorphs indicate at least two lines of development among the teleomorphs. Conidia produced holoblastically on sympodially proliferating cells are common to both lines. The first line is characterized by Pestalotiopsis and similar anamorph genera, all anamorphs of a closely-knit group of teleomorph genera that includes, among others, Pestalosphaeria, Lepteutypa, Discostroma and Broomella. If, as surmised by Nag Raj (1977), Amphisphaeria argentinensis is closely related to A. umbrinum and indeed has Bleptosporium pleurochaetum as its anamorph then it fits into this group and this group could be more precisely defined as the Amphisphaeriaceae (sensu str.). These fungi are plant parasites, mostly found within spots on leaves of dicotyledonous plants, less frequently on woody substrates. Conidiomata are formed and are acervular or pycnidial. Conidiogenous cells are phialides that do not proliferate, or that proliferate percurrently giving a conspicuously annellate aspect. Conidia of most of the anamorph genera in this group are pigmented and often versicolorous, transversely septate, and variously appendaged. "Spermatia" produced holoblastically on sympodially proliferating conidiogenous cells are found in conidiomata of the Seiridium anamorph of Lepteutypa cupressi (Swart 1973). Shoemaker & Muller (1963) found morphologically similar conidia in cultures of Broomella species (anamorph = Pestalotia) but the method of production was not described.

Perithecia in this first line are immersed, nonstromatic and clypeate or nonclypeate. The perithecial wall is narrow and comprises cells that are elliptical in section. Paraphyses are unbranched and are apparently apically free throughout their development. The ascal apex has a conspicuous amyloid ring (except for species of Broomella, where the apical ring is nonamyloid) and the ascospores are usually septate and are often pigmented.

If the Amphisphaeriaceae is thus narrowly defined, there remain several genera that will have to be redisposed in new or existing families. These genera include, among others, Cainia, Oxydothis Penz. & Sacc., Leiosphaeriella Petrak, Monographella Petrak, Induratia, Iodosphaeria, Collodiscula and Pseudomassaria Jacz. These fungi are loosely united by a tendency to form an amyloid apical ring in the ascus, and to have conidia produced holoblastically on sympodially proliferating conidiogenous cells.