:Xylariaceae: Biscogniauxia

GENERIC AFFINITIES AND EVOLUTION

Biscogniauxia is without doubt a genus of Xylariaceae. The taxa included in the genus have evolved with angiosperms on seasonally dry sites (see earlier herein; Rogers, 1979). The significant number of species with dark ascospores bearing a hyaline cellular appendage supports the argument that the ancestors of Biscogniauxia had two-celled hyaline ascospores. Biscogniauxia anceps, with a great preponderance of two-celled hyaline ascospores, probably represents a neotenic situation, i.e., a reversion of ascospores to a juvenile state that, in fact, recapitulates the mature ancestral ascospores (Rogers et al., 1996; Rogers, 1979). It is probable that the ancestral taxa would fit in family Amphisphaeriaceae sensu Müller and von Arx (1973) and had (have) many of the features of Vivantia J. D. Rogers, Y.-M. Ju, & Cand. (Rogers et al., 1996). A curious collection of a Biscogniauxia sent to us by Françoise Candoussau [GUADALOUPE: Basse Terre, Cabout, XI.1987, Vivant, J., GUAD-213, wood (JDR)] has brown ascospores with broad-based hyaline appendage cells, with narrow-based hyaline appendage cells, and lacking appendages altogether. This fungus appears to be transitional between those Biscogniauxia species with persistent appendages and those without. Like B. anceps, this fungus seems to have a loose or leaky regulatory control over ascospore cytology and final ascospore morphology. Biscogniauxia fuscella (Rehm) San Martín & J. D. Rogers likewise often shows an appendage on some maturing ascospores, but other maturing ascospores lack an appendage altogether.

The degree of carbonization of the stroma surrounding perithecia seems greatest among those taxa with the primitive ascospore trait, i.e., two-celled ascospores (Ju and Rogers, 1996). Species in section Annulata of Hypoxylon seem to show a relationship to Biscogniauxia in the dehiscence of a portion of stroma surrounding the ostiole, i.e., represents the last vestige of an outer dehiscent stromatal layer. Annulate Hypoxylon species generally retain a degree of carbonization surrounding perithecia; this is almost totally lacking in taxa of section Hypoxylon (Ju and Rogers, 1996). Stromata of Biscogniauxia lack discernible pigments in KOH, in contrast with most Hypoxylon species. Such pigments probably are associated with light absorption and ascospore release and have selective value in fungi that are completely erumpent from bark or otherwise exposed to high light intensities, i.e., Hypoxylon, Daldinia, etc.

Ascospores of Biscogniauxia lack dehiscent perispores, an innovation that is apparent in more advanced groups such as Hypoxylon and Daldinia (Ju and Rogers, 1996). These latter two taxa also lack appendaged ascospores altogether (Ju and Rogers, 1996; Ju et al., 1997).

The genus most closely related to Biscogniauxia is Camillea Fr. The latter genus has bipartite stromata as in Biscogniauxia. Stromata of Camillea species, however, are more highly specialized, often having unusual receptacles for ascospore discharge and/or ostioles serving multiple perithecia (Laessøe et al., 1989). Ascospores are pale-colored, ornamented with complex rods and/or reticulations, and lack a germ slit (Laessøe et al., 1989). Ascospores of one species are highly multinucleate (Rogers, 1968). Most of them germinate in low percentages or not at all. Asci tend to have rounded or diamond-shaped apical rings. Anamorphs are referable to Xylocladium Syd. ex Lindau, but, in culture, sometimes appear Nodulisporium-like by failure to produce the characteristic ampullae (Jong and Rogers, 1972). Camillea species are almost entirely found in the Americas, particularly the American tropics (Laessøe et al., 1989), whereas Biscogniauxia is represented by species worldwide.

The actual relationship of Biscogniauxia with Camillea is unknown. It is believed that Biscogniauxia is the older genus, based upon its wider distribution and simpler stromatal configurations. It is probable that Camillea species evolved under pressure from vectors that distribute ascospores, accounting in part for the bizarre and fascinating stromata (Laessøe et al., 1989). Interestingly, one species of Biscogniauxia--B. reticulospora Y.-M. Ju & J. D. Rogers--has reticulate ascospores. Unfortunately, cultures could not be obtained from it.

Another genus with an apparent relationship to Biscogniauxia is the monotypic Obolarina Pouzar (Pouzar, 1986). The Tulasne brothers described the fungus as a Nummularia, but recognized that it is atypical primarily in not producing a conidial state within the stroma (Tulasne and Tulasne, 1863). Candoussau and Rogers (1990) showed that, at least in culture, conidia of O. dryophila (Tul. & C. Tul.) Pouzar are infrequently produced from very rudimentary conidiophores. The asci also lack an apical ring. Ostioles are rudimentary and inconspicuous. It appears that Obolarina illustrates an evolutionary trend in the direction of cleistocarpy.

The Tulasne brothers were among the first to realize that the Diatrypaceae and Biscogniauxia [as Nummularia] had some features in common (Tulasne and Tulasne, 1863). They showed that Diatrype stigma (Hoffm.: Fr.) Fr. [as Stictosphaeria hoffmanni] develops in bark, but were somewhat equivocal on the presence of a dehiscent outer stromatal layer. They discussed and illustrated the moist-spored anamorph that was said to develop in "spermogonia."

Munk (1957) discussed "points" where the Diatrypaceae and Xylariaceae seemed to impinge. He believed that the Xylariaceae was moving toward small asci and pale allantoid ascospores. He merged the families, reducing Diatrypaceae to tribe Diatrypeae. While we agree that there is a good deal of commonality between these families, we would not merge them. We especially doubt that Camarops is closely related to either the Diatrypaceae or Xylariaceae (see Rogers, 1994). Our suspicions of a relationship of Biscogniauxia with the Diatrypaceae comes primarily from culturing a fungus probably equatable with Nummularia viridis Theiss. (Ju et al., 1993). This fungus has fairly small ascospores and whitish tissue surrounding small perithecia. On culturing the fungus yields long, curved conidia in wet masses that, like those of undoubted Diatrypaceae, do not germinate. This fungus, along with several others such as Hypoxylon hypophlaeum (Berk. & Ravenel) S. Teng, is removed from Biscogniauxia to Jumillera and is not considered further herein (Rogers et al., 1997). Moreover, Nummularia microplaca (Berk. & M. A. Curtis) Cooke-a taxon with small ascospores, bipartite stroma, and dry scolecosporous anamorph that inhabits members of the aromatic angiospermous families Lauraceae and Myrtaceae-is placed in Whalleya (Rogers et al., 1997).

It is uncertain if Jumillera and Whalleya represent a convergence with undoubted taxa of the Diatrypaceae or if, in fact, they have a common ancestry and are evolving in parallel (see Rogers, 1994). Another genus that seems to have interfaces with the Xylariaceae and Diatrypaceae is Graphostroma Piroz. This monotypic genus, currently placed in family Graphostromataceae, has stromata reminiscent of Diatrype or Biscogniauxia, but has an anamorph similar to the latter (see Barr et al., 1993). Graphostroma has ascospores that approach an allantoid shape and asci that are somewhat like those of undoubted Diatrypaceae. No undoubted member of the Xylariaceae has allantoid ascospores or long-stipitate asci with the characteristic shape of those of the Diatrypaceae.