Studies in the Lasiosphaeriaceae. Monographs of two key genera and a family-level phylogeny

Title Page | Introduction | Morphological Characters | Bibliography | Key to Genera |

Project Goals and Background Information



Table of Contents:
I. Introduction to the Lasiosphaeriaceae
II. Taxanomic Background of the Lasiosphaeriaceae
III. Known distribution and abundance
IV. Biogeographic questions being addressed

The Study Group and Conceptual Issues

Ascomycetes (Subdivision Ascomycotina) constitute the largest known group of fungi (Hawksworth et al., 1983, Hawksworth, 1991) and occur in all ecosystems and geographical areas. Many are small and inconspicuous, parasitic on plants and evident only from the effects they produce upon their hosts. Others are saprobic, living on the soil, decaying logs, leaf mold or dung. The ascomycetes have two distinct reproductive phases, the teleomorph (sexual or meiotic state) and the anamorph (asexual or mitotic state), which will most often be found growing and propagating independent of one another in time and space (Fig. 4). This two-phase life-cycle poses special problems in understanding the biology, ecology, classification and taxonomy of these organisms. Ascomycetes have been included in recent phylogenetic studies (Fig. 1) and have been estimated to have diverged from the basidiomycetes about 400 million years ago, with all the major lineages established by the end of the Mesozoic (Berbee and Taylor, 1994) (Fig. 2).
The order Sordariales is one of several taxonomically diverse and biologically and ecologically important groups within the Subdivision Ascomycotina, Class Hymenoascomycetes (superordinal classification follows Barr, 1990) (Table 1). Resolving phylogenetic relationships within this order will help elucidate relationships within the class as a whole. Important taxa in this order include Neurospora, a widespread group of molds and soil inhabitants including N. crassa, the model organism widely used in molecular and genetic studies (Perkins, 1992; Taylor et al, 1993); Sordaria fimicola the coprophilous organism, used to demonstrate gene segregation in spore color; and Chaetomium Kunze, an important cellulolytic organism, responsible for the destruction of paper and fabrics. Of considerable economic importance are a number of Ceratocystis Ellis and Halst. species that are aggressive plant pathogens in both agricultural and natural ecosystems (Kile, 1993).

According to one recent compilation (Eriksson and Hawksworth, 1993) the Sordariales contains eight families (Table 2). Barr and Cannon (1994) suggested that two of these families belong in other orders. The presumed synapomorphies supporting the order Sordariales are the ascomal ontogenetic characters involved in centrum development (centrum = structures within the fruitbody such as asci and surrounding sterile elements). Recent cladistic analyses of molecular data have shown the Sordariales s. lat. to be non-monophyletic (Fig. 3). The four genera, representing four families, out of the 99 genera of Sordariales sensu Eriksson and Hawksworth (1993), that were included in Spatafora and Blackwell's (1994) cladistic analyses of the ascomycetes segregated into two different clades. Three of these genera, Neurospora (Sordariaceae), Chaetomium (Chaetomiaceae), and Cercophora Fuckel (Lasiosphaeriaceae) formed a strongly supported clade, with a sister group that included the Diaporthales and Ophiostomatales (Fig. 3). These genera represent a core group of typical Sordariales. The other genus of Sordariales included in their study, Melanospora Corda (Ceratostomataceae), has been placed variously in different orders depending on the investigator's emphasis of different morphological characters. The molecular data are concordant with the notion that the genus is best removed from the Sordariales.
The Sordariaceae is the taxonomically best studied family in the order. It primarily contains coprophilous taxa, including Neurospora and Sordaria. Over the years, as taxa have been removed, the circumscription of the family has narrowed. Many taxa have been placed in the Lasiosphaeriaceae. The Lasiosphaeriaceae, as it stands now, is widely circumscribed and needs detailed study to assess the true affinities of its members. The various genera within the Lasiosphaeriaceae exhibit a broad range of developmental and morphological characters in both the teleomorph and anamorph that are currently used as important indicators of phylogenetic relationships within the ascomycetes.
The family level--Historically, the taxa that make up the present Sordariaceae and Lasiosphaeriaceae have been combined under one family (Munk, 1957; Dennis, 1968, Carroll and Munk, 1964; Barr and Cannon, 1994), separated into two families with the addition of subfamily groupings (Lundqvist, 1972), or split off into a third family, Tripterosporaceae (Barr, 1990). Hawksworth and Eriksson (1993) place the relevant taxa in two families; that hypothesis is used here with the understanding that the two families are probably closely related if not monophyletic. Assessment of the monophyly of these taxa, and the relationship of the Lasiosphaeriaceae with other families in the Sordariales is an important aspect of our monographic work.

Taxanomic Background of the Lasiosphaeriaceae


In a recent compilation, the Lasiosphaeriaceae is considered to include 51 genera with 43 synonyms (Eriksson and Hawksworth, 1993). Relationships of genera within the family are based on a number of morphological characters of the teleomorph, for example, ascomal wall structure, wall surface covering, ascus shape and apical ring structure; and the hyphomycetous development of the anamorph. However, this information is not known for all of the members and the number of taxa that really belong in the group is unknown.
Of the 51 genera in the Lasiosphaeriaceae, 19 are predominantly coprophilous and these are the only recently studied members of the family (Cain, 1962; Lundqvist, 1972). The important plant pathogenic genus Ceratocystis is currently included in the family, although this has recently been questioned (Rogers, 1994; Spatafora and Blackwell, 1994). The family also contains members that have been untouched monographically. Chaetosphaeria and Lasiosphaeria occur mostly on decaying wood, herbaceous stems and leaves, not on dung, and are among the most species-rich genera in the family, yet there are no comprehensive treatments or keys to species.
Characteristics of some genera currently included in the Lasiosphaeriaceae suggest relationships to other families in the order or even to other orders. Ascospore germination in Lasiosphaeria and Lasiosphaeriella Sivan. suggests a relationship with the Nitschkiaceae. Chaetosphaeria has long been allied with the Trichosphaeriales, while the iodine-induced amyloid, apical, ascus ring in Lasiobertia Sivan. suggests a connection to the Xylariales. Taxa with long, narrow ascospores, Acrospermoides Miller and Thomps. and Ophioceras Sacc., show affinities with the Diaporthales and Mycomedusiospora Carroll and Munk with light colored ascomata may show a relationship with the Hypocreales.
Phylogenetic analysis of the genera currently maintained within the family is therefore long overdue and clearly warranted. This rigorous assessment of relationships needs to incorporate both morphological and molecular data. As currently circumscribed, the Lasiosphaeriaceae is almost certainly not monophyletic.
The genus level--The two predominantly lignicolous genera, Chaetosphaeria and Lasiosphaeria, are chosen for monographic revision because they have never been treated systematically even though they are key to understanding phylogenetic relationships in the family and are apparently ecologically and economically important. Species in the two genera have been described primarily from north temperate regions. However, preliminary evidence suggests that these genera are a dominant part of the neotropical mycota involved with litter decomposition since they were among the most commonly collected taxa in several recent collecting trips to French Guiana (Table 1 and Table 2).
According to the Index to Saccardo's Sylloge Fungorum (Reed and Farr, 1993), there are 47 names described in Chaetosphaeria and 171 in its synonym Zignoella. From 1920 to the present, there have been 52 Chaetosphaeria species and 11 Zignoella species described (Index of Fungi), making a total of 281 described taxa under Chaetosphaeria. The genus was originally described for a hyaline-spored species with smooth ascomata (Tulasne and Tulasne, 1863), but when the genus was taken up by Saccardo (1883) it was conceived as having smooth or tomentose ascomata and hyaline or brown ascospores. Zignoella was placed into synonymy with Chaetosphaeria by Booth (1957, 1958). Over the years taxa with many different character combinations have been included. Ascospore coloration is still maintained as an important generic character in some groups, and this character will probably prove important for circumscribing monophyletic units out of the species currently included in Chaetosphaeria. Species originally described under Zignoella have hyaline spores, but these spores are multi-septate while the original circumscription of Chaetosphaeria only included species with one septum. A number of Chaetosphaeria names have been transferred to or linked with names in Lasiosphaeria. This is because both share the general characteristics of superficial, ovoid-shaped ascomata and hyaline septate ascospores. The two genera can be separated by critical examination of the teleomorph and anamorph characters.
Chaetosphaeria occurs mainly on decaying twigs or wood, however, a few species are described from leaves and other substrates. Chaetosphaeria was one of the first genera in which a teleomorph and anamorph were recognized as belonging to the same fungus (Tulasne and Tulasne, 1863). Species of Chaetosphaeria have been linked to several phialidic anamorphs belonging to a various genera including Chloridium Link and Gonytrichum Nees (Gams and Holubova-Jechova, 1976). Gams and Holubova-Jechova noted that the sections recognized in Chloridium based on morphology corresponded to sections in Chaetosphaeria based on ascospore morphology. Some species of the Chloridium anamorph have cellulolytic activity occurring in advance stages of decay on wood, and a species of the Gonytrichum anamorph has the ability to decompose lignin (Gams and Holubova-Jechova, 1976).
Lasiosphaeria is a species-rich genus which will be crucial to resolve relationships in the family. A history of the generic name and its relationship to coprophilous taxa was given by Lundqvist (1972), without further species treatments. The genus was originally based on 7 species (Cesati and de Notaris, 1863). As it was revised over the years, with changes in its concepts (Fuckel, 1870; Saccardo, 1883), it gradually became packed with a variety of probably unrelated elements. The Index to Saccardo's Sylloge Fungorum (Reed and Farr, 1993), lists 131 names in Lasiosphaeria and there are 43 names for species described after 1920 (Index of Fungi) making a total of 174.
Lasiosphaeria is closely linked to Cercophora and Bombardia (Fr.) Karst. by similarities in anamorph development, ascomal, ascus and ascospore morphology. Although Cercophora is considered to be mainly coprophilous, lignicolous species have been transferred among these three genera without resolving a clear generic concept in any of them (Hilber and Hilber, 1979; Hilber et al, 1987). The Nordic coprophilous species of Cercophora have been outlined with excellent descriptions and illustrations (Lundqvist, 1972), however, no monograph of either Cercophora or Bombardia exists. While this project is not intended as a monograph of these two genera, the type species and other representative species of each must be reassessed and their relationship to Lasiosphaeria clearly defined. Only after this is done can we understand the extent of revision of Cercophora and Bombardia that is necessary within the context of a monograph of Lasiosphaeria.
There has never been a monograph for the genus. For identification we have had to rely on scattered works by few authors treating limited number of species (Seaver, 1912; Carroll and Munk, 1964; Hilber and Hilber, 1983, Hilber et al, 1987; Hilber and Hilber, 1979, make reference to an unpublished dissertation of Lasiosphaeria which the PI has been unable to locate). In recent years, a few species have been segregated in other genera e.g. Lasiosphaeriella based mostly on ascospore characters (Sivanesan, 1975). However, other characters, such as its anamorph development suggest that these species should be included in a reassessment of the group.
Lasiosphaeria has been conceived in a broad sense regarding ascospore types and ascomal wall structure (Carroll and Munk, 1964; Barr, 1990). Variation includes ascomata with assorted types of vestiture and variation in ascus shape, apical apparatus and dehiscence. The few reported anamorphs for species of Lasiosphaeria are phialides that form directly on germinating ascospores (Hughes, 1951; Gams, 1973; Hilber et al, 1987), a pattern of development that is not found in many ascomycete genera (Hanlin, 1994; Huhndorf and Glawe, 1990; Samuels, pers. comm.). However, some species of Cercophora and Lasiosphaeriella, as well as one species of Bertia de Not. in the Nitschkiaceae, also have this type of anamorph development (Sivanesan, 1978). A presumed apomorphy which supports the genus Lasiosphaeria is the precocious development of the ascospores, where the species have become physiologically mature in the hyaline stage (Barr and Cannon, 1994). This could also be the case for Chaetosphaeria.
It is not known at this time how well the number of species described reflects the actual number of species that will remain in Lasiosphaeria and Chaetosphaeria after detailed examination. Based on the PI's monographic studies in other groups (Huhndorf, 1992c), it is estimated that two-thirds of the described species will belong to other genera, some closely related but others more distantly so. The type collections of all the species must be examined to determine the number of actual species known at this time, however the type collections for the most part will not provide any information about anamorphs. For this, new collections and pure culture studies will be necessary. Previously unknown species will undoubtedly be encountered and will be described.

Known distribution and abundance
Genera of Lasiosphaeriaceae have been described from many parts of the world but many are still known only from single collections. For coprophilous ascomycetes, Lundqvist (1972) suggests there are two distribution patterns, north-temperate and tropical-subtropical, with the possibility of a third, south-temperate group on which information is sparse. Some species have been found only in certain climatic regions, whereas the genera seem not to be limited in this way. Cercophora is best known from the temperate zones but there are a number of records of tropical collections in the literature (Lundqvist, 1972). Coprophilous genera have been described from Venezuela (Jeng and Krug, 1976, 1977), and from India (Subramanian and Lodha, 1968) and one genus appears to be pantropical (Cain and Farrow, 1956).
Data for the lignicolous taxa are too incomplete to discern distribution patterns, and nothing can be said about the abundance of most. However, eight genera have been reported mostly from Europe and North America, including Chaetosphaeria and Lasiosphaeria. Genera have been described from Brazil (Müller and Samuels, 1982a; Samuels and Müller, 1978), India, Jamaica, Japan, New Zealand and United Kingdom (Minoura and Muroi, 1978; Müller and Samuels, 1982b; Samuels et al, 1981; Sivanesan, 1983), Australia (Hyde, 1992a,b) and Egypt (Schatz, 1985).
While working on an ascomycete contribution for a checklist (Courtecuisse et al., in prep.), the PI examined a large number of French Guiana ascomycete collections to assess species richness and abundance of the various litter inhabiting genera and species in this lowland rainforest area (Table 1 and Table 2) The Sordariales, specifically the Lasiosphaeriaceae, was the first and third most-frequently encountered group of ascomycetes in French Guiana in these two surveys where it comprised between 4% and 16% of the ascomycetes encountered. The two other most-frequently encountered ascomycete families belong to the Hypocreales and Xylariales. The Lasiosphaeriaceae were not specifically targeted for collection in these studies. Additionally, no attempt was made at an exhaustive search for coprophilous, soil-inhabiting or aquatic ascomycetes, so taxa occurring exclusively in these habitats are noticeably absent from the collections. Nevertheless, thirteen lignicolous or herbicolous Lasiosphaeriaceae genera were found of which two are new to the Neotropics. The four genera listed in Table 2 were the most abundant. Most of the genera are new records for French Guiana.
Most recognized species of Chaetosphaeria and Lasiosphaeria have been described from temperate Europe and North America but it is unknown how accurately this reflects the global distribution. Approximately 60% of the species have been described from Europe, but only a very few of the same names ever appear in checklists of commonly collected species (Eriksson, 1992). Reports of Chaetosphaeria and Lasiosphaeria from North America are mostly from scattered historical collecting sites in the eastern U.S. (Ellis and Everhart, 1892). Many of these sites no longer exist as natural areas. Most locations in North America have been uncollected for ascomycetes in general, so it is difficult to discern distribution patterns. However, interesting finds of Lasiosphaeria and Cercophora species in recent years (Hilber et al, 1987) suggest that the Appalachians may be a center of diversity. All collections of Lasiosphaeria and Chaetosphaeria (and synonymous names) will be requested from the major North American and European herbaria to supplement the material collected during this study.
Approximately 15% of the total number of species of Chaetosphaeria and Lasiosphaeria have been described from tropical-subtropical regions worldwide. Historical collections are available from Brazil, Ceylon, Cuba, Java and the Philippines. The actual distribution of these genera in the tropics is unknown. Recent studies of lignicolous members of the family include four species of Lasiosphaeria found in Costa Rica (Carroll and Munk, 1964), one species of Cercophora in Venezuela (Hanlin and Tortolero, 1987) and one species of Chaetosphaeria found as an endophyte in French Guiana (Petrini and Dreyfuss, 1981). However our recent data (Table 2) indicate that they are abundant and diverse in the neotropics. The data indicate that Chaetosphaeria, and Lasiosphaeria s. lat. (including Lasiosphaeriella and Cercophora) are the most commonly collected members of the family in French Guiana and together the four genera make up between 3% and 11% of the ascomycetes encountered.

Biogeographic questions being addressed
No work has been done on the biogeography of the taxa in the Sordariales. A major precursor to undertaking biogeographic analysis is a good understanding of phylogenetic relationships within the group. It also requires a more complete record of geographic distribution than is presently available.
Our collecting sites in North America, the Caribbean, and Central and South America (see below) were chosen to allow us to contrast the distribution of ascomycetes in the Americas and begin to think about their biogeographic distribution in lowland tropical rainforests. Because the vascular plant flora in these sites is documented, we can test ideas about plant/fungus ratios used in estimating numbers of fungal species (Hawksworth, 1992) and about host specificity as it relates to tropical plant diversity. Whereas Hawksworth (1991) hypothesized a high degree of host specificity worldwide leading to a high global estimate of fungi, May (1991) suggested that fungal species in the tropics may be less host specific because the hosts are sparsely distributed. If this is true then the plant/fungus ratio in the tropics would be lower and worldwide estimates of numbers of fungal species would be substantially reduced.
Host specificity also plays a role in ecosystem management and biogeography. Host-specific fungi may not co-disperse with their host following forest fragmentation, thus altering the susceptibility of both plant and fungi to extinction. We will be able to test ideas about fungal diversity as it relates to habitat and host diversity and to the abundance of resources. From the taxonomic composition based on our U.S. and tropical collecting sites we will be able to contrast temperate and tropical mycotas and suggest distribution patterns and hypotheses on migration versus dispersal that can be tested with the phylogeny that we will propose.


Top of this Page | Introduction | Morphological Characters | Bibliography | Key to Genera |