A review of the Ustilago-Sporisorium-Macalpinomyces complex.

The fungal genera Ustilago, Sporisorium and Macalpinomyces represent an unresolved complex. Taxa within the complex often possess characters that occur in more than one genus, creating uncertainty for species placement. Previous studies have indicated that the genera cannot be separated based on morphology alone. Here we chronologically review the history of the Ustilago-Sporisorium-Macalpinomyces complex, argue for its resolution and suggest methods to accomplish a stable taxonomy. A combined molecular and morphological approach is required to identify synapomorphic characters that underpin a new classification. Ustilago, Sporisorium and Macalpinomyces require explicit re-description and new genera, based on monophyletic groups, are needed to accommodate taxa that no longer fit the emended descriptions. A resolved classification will end the taxonomic confusion that surrounds generic placement of these smut fungi.

INTRODUCTION

Three genera of smut fungi (Ustilaginomycotina), Ustilago, Sporisorium and Macalpinomyces, contain about 540 described species (Vánky 2011b). These three genera belong to the family Ustilaginaceae, which mostly infect grasses (Begerow et al. 2006) and have teliospores that germinate to produce phragmobasidia (Bauer et al. 2001, Begerow et al. 2006). Ustilago and Sporisorium were shown to form a monophyletic group within the Ustilaginaceae after molecular phylogenetic analyses (Begerow et al. 1997, 2006, Stoll et al. 2003, 2005). The systematic position of Macalpinomyces is ambiguous within the Ustilaginales (Begerow et al. 2006).

Many taxa within Ustilago, Sporisorium and Macalpinomyces share two or more morphological characters indicative of the different genera. This makes taxonomic placement of species within genera problematic. The original characters used to identify genera were not sufficiently robust to encompass the full morphological diversity of novel species that have since been discovered. Taxa within Ustilago, Sporisorium and Macalpinomyces are part of a systematically unresolved complex (Vánky 2002a, Stoll et al. 2003, 2005, Piepenbring 2004, Vánky et al. 2006, Vánky & Shivas 2008). Three further genera, Anomalomyces, Melanopsichium and Tubisorus, are considered to be distinct, well-defined members of this complex.

Attempts to reconcile the taxonomy of this complex using either morphology (Vánky 1991, Piepenbring et al. 1998) or molecular phylogenetics (Stoll et al. 2003, 2005) have been unsuccessful. This paper reviews chronologically changing generic concepts in the Ustilago-Sporisorium-Macalpinomyces complex and presents an approach for resolving systematic anomalies.

TAXONOMIC HISTORY

Ustilago

Ustilago, derived from the Latin ustilare (to burn), was named by Persoon (1801) for the blackened appearance of the inflorescence in infected plants, as seen in the type species U. hordei. According to Clinton (1906), Persoon adopted the name Ustilago from Johann Bauhin’s 1651 edition of ‘Historia plantarum universalis’. Persoon (1801) created Ustilago as a subgenus of Uredo in his ‘Synopsis Methodica Fungorum’. He described Uredo, now classified within the rust subphylum Pucciniomycotina (Aime et al. 2006), as lacking a peridium and having spores that were powdery, loose, uniform and mostly globose. Ustilago, now classified in the smut subphylum Ustilaginomycotina, was separated from Uredo by possessing black to brown powdery spores that parasitize mostly plant inflorescences. Ustilago was promoted to the level of genus by Roussel (1806). Ustilago became a catch-all genus for a diversity of smut fungi. Many taxa currently regarded as belonging to Ahmadiago, Antherospora, Anthracoidea, Aurantiosporium, Bambusiomyces, Bauerago, Cintractia, Eriocaulago, Exoteliospora, Farysia, Farysporium, Liroa, Macalpinomyces, Melanopsichium, Microbotryum, Parvulago, Pericladium, Schizonella, Sporisorium, Thecaphora, Tilletia, Tranzscheliella, Ustanciosporium, Vankya, Websdanea and Yelsemia were originally described as members of Ustilago (e.g. Piepenbring et al. 1996, Vánky 1998a, 1999a, b, 2002a, b, 2003b, 2004b, 2011a, b, Bauer et al. 1999, 2007, 2008, Piepenbring 2000, Vánky et al. 2008).

Juliohirschhornia was proposed for its pattern of spore germination, which was considered to be intermediate to the Ustilaginaceae and Tilletiaceae (Hirschhorn 1986). Vánky (2002a) noted that Juliohirschhornia was an invalid genus and further considered its spore germination represented only a variant of the Ustilago-type. Several other genera were regarded by Vánky (2002a) as synonymous with Ustilago, including Crozalsiella, Necrosis, Pericoelium and Ustilagidium.

Two attempts have been made to subdivide Ustilago, although the proposed classifications have not been widely accepted. Firstly, Brefeld (1912) proposed the genus Mycosarcoma for Ustilago maydis. Brefeld (1912) based Mycosarcoma on the structure of the peridium, incubation time in the host, localized infection and development of aerial conidia. Generic placement of U. maydis within the complex is contentious (Piepenbring et al. 2002, Stoll et al. 2005) and until the complex is resolved, this taxon is best left within Ustilago because of its importance as a model plant pathogen.

Another attempt to subdivide Ustilago was made in 1949 by the mycologist Tchen Ngo Liou, who considered that the basidia of U. esculenta differed from the type species of Ustilago (cited in Piepenbring et al. 2002). Liou erected the genus Yenia, with Y. esculenta as the type, and transferred seven additional Ustilago species into the new genus (Liou 1949). Vánky (2002a) considered that the eight taxa Liou selected were very different in biology, soral structure, spore morphology and germination patterns, and did not constitute a natural group. Piepenbring et al. (2002) in their single-locus phylogenetic analysis found that U. esculenta was sister to 21 species of Ustilago and Sporisorium, accepting that U. esculenta belonged in a separate genus to Ustilago. Stoll et al. (2005) did not support the separation of U. esculenta from Ustilago on the basis of a molecular phylogenetic analysis, which included this and 97 other Ustilago, Sporisorium and Macalpinomyces species.

Beck (1894) introduced the genus Melanopsichium for a taxon first described as Ustilago austro-americanum on Polygonum. The genus was characterised by compact, hard, irregularly lobed galls in the inflorescence, stems and leaves (Halisky & Barbe 1962, Vánky 2002a). Weiss et al. (2004), Begerow et al. (2004) and Stoll et al. (2005) concluded that Melanopsichium represented an example of a host jump from Poaceae to Polygonaceae, as M. pennsylvanicum belonged to the Ustilago clade. Begerow et al. (2006) consequently rejected the family Melanopsichiaceae proposed by Vánky (2001a).

Langdon & Fullerton (1975) studied the soral ontogeny of six Ustilago species. Their revised concept of Ustilago included taxa that colonised host plants with hyphae that destroyed parenchymatous tissue to then become spores, without forming fungal peridia, columellae, sterile cells or spore balls.

The gross morphology of Ustilago is variable (Fig. 1). Piepenbring (2004) recorded 14 different soral morphologies for Ustilago in her treatise of the sori found in the Ustilaginomycotina. Some taxa, such as U. sparsa and U. trichophora, occurred as localised galls on the host plant, inducing hypertrophied ovaries rather than destroying the entire inflorescence. Ustilago altilis and U. esculenta infected the culms of the host, and some species occurred in the leaves, for example U. calamagrostidis and U. striiformis. Vánky (2002a) considered Ustilago as occurring solely on hosts in the Poaceae, accepting 174 species (Vánky 2011b).

Fig. 1

Diversity of soral morphology in Ustilago. a. Ustilago spinificis on Spinifex longifolius; b. Ustilago xerochloae on Xerochloa barbata; c. Ustilago drakensbergiana on Digitaria tricholaenoides; d. Ustilago tritici on Triticum aestivum; e. Ustilago bouriquetii on Stenotaphrum dimidatum; f. Ustilago altilis on Triodia sp.; g. Ustilago phragmitis on Phragmites karka; h. Ustilago cynodontis on Cynodon dactylon.

Sporisorium

Ehrenberg described Sporisorium in a letter to Link, based on a collection he had made of S. sorghi on the cultivated grass Sorghum in the Poaceae (Link 1825). Sporisorium was described as unique because it possessed columellae of equal length as the glumes, formed agglutinated spores and mutilated floral parts. Sporisorium also had sterile partitioning cells in groups or chains and a peridium (Link 1825, Langdon & Fullerton 1978).

Four years after the description of Sporisorium, Rudolphi (1829) described the confusingly named Sorosporium from Saponaria officinalis in the Caryophyllaceae. Many authors subsequently chose Sorosporium for smut taxa with peridia and spore balls including those that infected grasses (Poaceae). Sporisorium was overlooked for about 150 years until Langdon & Fullerton (1978) re-established the name. Many of the species described in Sorosporium and Ustilago have since been reclassified in Sporisorium. More precisely, Sporisorium contains at least 60 taxa originally classified as Ustilago, and about 170 taxa described as Sorosporium (Robert et al. 2005).

Sorokin described Endothlaspis in 1890 for two smuts, on Sorghum and Melica, of which the respective types have been lost (cited in Langdon & Fullerton 1978, Vánky 2002a). Langdon & Fullerton (1978) believed the description and illustrations of Endothlaspis were vague and poorly executed. Vánky (2002a) considered that Endothlaspis was a synonym of Sporisorium and that the type species was based on a host misidentification.

Lavrov (1936) and Ciferri (1938) divided Sorosporium into two subgenera depending on whether they infected hosts in Poaceae or Caryophyllaceae (cited in Vánky 2002a). Langdon & Fullerton (1975) noted that Sorosporium species on Poaceae differed in soral ontogeny and structure to species on Caryophyllaceae, essentially in that Sorosporium on Caryophyllaceae lacked a well-defined sorus. Langdon & Fullerton (1975) suggested that smuts on Poaceae should be grouped in a separate genus, but did not make any taxonomic revisions at that stage. Vánky (1998b) considered Sorosporium to be a synonym of Thecaphora after an examination of the types of both genera revealed no essential morphological differences. This decision was subsequently supported by molecular phylogenetic analyses (Vánky et al. 2008).

Sphacelotheca was established by de Bary (1884) for Sph. hydropiperis on Polygonum. Sphacelotheca was defined as having a membrane or peridium enclosing the spores and a columella (cited in Langdon & Fullerton 1978). Clinton (1902) transferred 10 taxa from Ustilago to Sphacelotheca, including Sporisorium sorghi, which he referred to as Ustilago sorghi. Clinton did not mention Sporisorium, but he attributed the authorship of U. sorghi to Link, indicating that he was aware of Sporisorium as an earlier described genus. Aside from a brief mention of the characters of Sphacelotheca, Clinton gave no reason why the 10 taxa would be better suited to Sphacelotheca. Clinton’s transferral of taxa in Sporisorium to Sphacelotheca sensu Clinton was precedent for over 110 subsequent descriptions of species of Sphacelotheca on grasses (Robert et al. 2005).

Langdon & Fullerton (1978) ascertained that the columellae in Sphacelotheca species on Polygonaceae and Poaceae were not homologous. Sphacelotheca formed a columella from fungal cells adhering to one another on hosts in the Polygonaceae, whereas columellae were derived from host material in the Poaceae. Langdon & Fullerton (1978) also noted differences in the peridium and the development of the spore mass between Sphacelotheca in the Polygonaceae and Poaceae. Sphacelotheca occurred only on hosts in the Polygonaceae and has been shown by Bauer et al. (1997) to belong to the Microbotryales in the Pucciniomycotina. This systematic placement was confirmed by molecular analyses (Weiss et al. 2004, Kemler et al. 2006).

Langdon & Fullerton (1978) resurrected Sporisorium after showing that Sphacelotheca and Sorosporium were not suitable genera for smut fungi on grasses. They designated a new type specimen of Sporisorium sorghi from an Australian collection on Sorghum leiocladum, which Vánky (1990) believed to represent S. cruentum. Vánky (1990) proposed a new neotype from an Egyptian collection of S. sorghi. The neotype originally proposed by Langdon & Fullerton (1978) appeared to belong to a distinct species, S. australasiaticum (Vánky & Shivas 2001).

Langdon & Fullerton (1978) outlined the characteristics of Sporisorium based on their neotype of Sporisorium sorghi. Characters of importance included a ‘hyphal peridium, columella composed of host tissues and hyphae, and spores intermixed with partitioning (sterile) cells’. These characters are variable among other Sporisorium species (Fig. 2).

Fig. 2

Diversity of soral morphology in Sporisorium. a. S. cenchri-elymoidis on Cenchrus elymoidis; b. S. cryptum on Yakirra sp.; c. S. heteropogonicola on Heteropogon contortus; d. S. bothriochloae on Dichanthium sericeum; e. S. tumefaciens on Chrysopogon sp.; f. S. iseilematis-ciliati on Iseilema sp.; g. S. themedae on Themeda triandra; h. S. aristidicola on Aristida sp.; i. S. likhitekerajae on Ischaemum sp.; j. S. doidgeae on Capillipedium parviflorum; k. S. sacchari on Saccharum sp.; l. Ustilago scitaminea on Saccharum officinarum; m. Sporisorium caledonicum on Heteropogon contortus; n. S. ischaemum on Ischaemum indicum; o. S. holwayi on Andropogon bicornis.

The morphological variation of peridia, columellae, sterile cells and dimorphic spores in Sporisorium has led to different interpretations by mycologists. For example, Langdon & Fullerton (1975) described the presence of a columella in Sporisorium consanguineum, but it was later reported absent by Vánky & Shivas (2008). A columella was not described by Langdon (1962) in Ustilago porosa, but this species was regarded to have one by Vánky & Shivas (2001). The presence or absence of columellae, peridia, sterile cells and dimorphic spores has formed the taxonomic boundary between Sporisorium and Ustilago, and interpretations of these structures must be consistent before the complex can be resolved.

Another character used to define Sporisorium was that spores were often compacted in permanent (or semi-permanent) spore balls (Vánky 2002a, Vánky & Shivas 2008). Vánky (1998c) considered spore balls to be homoplasious in the Ustilaginomycotina and they do not occur across all taxa in Sporisorium. Vánky (2011b) recognised 326 species of Sporisorium.

Macalpinomyces

Langdon & Fullerton (1977) established Macalpinomyces to accommodate M. eriachnes, which they considered as distinct from Sporisorium and Ustilago. Macalpinomyces lacked columellae, produced sterile cells and the spores were uniformly ornamented and polyangular or subpolyangular (Langdon & Fullerton 1977, Vánky 1996).

The nomenclatural history of M. eriachnes epitomises the confusion caused by many taxa in the Ustilago-Sporisorium-Macalpinomyces complex. The original collection of M. eriachnes in Australia by the botanist Ferdinand von Mueller, was divided and sent to two mycologists, Mordecai Cooke in England and Felix von Thümen in Germany. Two new fungal taxa were described based on this single collection, Sorosporium eriachnes by Thümen in 1878 and Ustilago australis by Cooke in 1879 (Langdon & Fullerton 1977). Langdon & Fullerton (1977) later transferred this smut to a new genus, Macalpinomyces, nearly a century after the specimen was first described.

Vánky (1996) broadened the concept of Macalpinomyces to include taxa that lacked a columella but possessed sterile cells, which are morphological features shared by both Sporisorium and Ustilago. This led to numerous taxonomic combinations, for example, M. bursus, M. neglectus and M. spinulosus. The broadened concept of Macalpinomyces allowed for a variety of gross morphologies to be included, ranging from localised or systemic galls in the ovaries, to longitudinally hypertrophied sori up to 16 cm long in M. chrysopogonicola (Fig. 3).

Fig. 3

Diversity of soral morphology in Macalpinomyces. a. M. ewartii on Sorghum timorense; b. M. arundinellae-setosae on Arundinella setosa; c. M. mackinlayi on Eulalia mackinlayi; d. spores of M. mackinlayi; e. M. siamensis on Coelorachis striata; f. M. eriachnes on Eriachne helmsii; g. spores of M. eriachnes. — Scale bars: d, g = 10 μm.

Molecular phylogenetic analysis has shown that Macalpinomyces is polyphyletic. The type species, M. eriachnes, is sister to all other taxa in the complex, and forms a monotypic genus within the Ustilaginaceae (Stoll et al. 2005). Begerow et al. (2006), in their phylogenetic study of the Ustilaginomycotina, proposed that M. eriachnes might not belong to the Ustilaginaceae as it did not occur in the clade containing Sporisorium, Ustilago and Moesziomyces.

Species of Macalpinomyces have sterile cells, a peridium derived from host material, and lack true spore balls (Vánky 2011b). Vánky (2011b) accepted 46 species of Macalpinomyces.

Relationships within the Ustilago-Sporisorium-Macalpinomyces complex

Taxa within the Ustilago-Sporisorium-Macalpinomyces complex often possess morphological characters that occur in more than one genus. Overlapping characters create uncertainty for species placement, as illustrated by Macalpinomyces eriachnes, which was independently placed in both Ustilago and Sorosporium. In a comprehensive taxonomic study over the course of eight years, Vánky (1996, 1997, 1998d, 2001c, 2002b, 2003a, b, 2004a, b) and Vánky & Shivas (2001, 2003) combined over 30 smut species that possessed a combination of Sporisorium and Ustilago characters into Macalpinomyces. Taxonomic shuffling occurred later with many species described before 1978 as Ustilago and that were subsequently moved to either Macalpinomyces or Sporisorium. The result was that many taxa have been moved back and forth among genera without systematic evidence that they constituted natural, monophyletic groups.

New genera have been raised for some smuts that differed subtly from the type descriptions of Ustilago, Sporisorium and Macalpinomyces. Endosporisorium (Vánky 1995a), Lundquistia (Vánky 2001b), Anthracocystis (Brefeld 1912), Yenia (Liou 1949) and Tubisorus (Vánky & Lutz 2011) are examples of genera that were proposed to subdivide Ustilago and Sporisorium. The description of new genera or placement of taxa in poorly defined genera, has contributed to systematic confusion within the complex.

Vánky (1995a) described Endosporisorium to accommodate Sorosporium capillipedii (type) and Sorosporium loudetiae and later added two other smut taxa (Vánky 1995b). This genus differs from Ustilago in having sterile cells and ephemeral spore balls, and from Sporisorium in lacking columellae and a fungal derived peridium. The sori of Endosporisorium were described from the stems rather than the inflorescences. After Vánky (1996) emended Macalpinomyces to encompass more taxa, he subsequently synonymised Endosporisorium with Macalpinomyces, preferring a large, well-delimited genus, rather than many monotypic and closely related genera (Vánky 1997).

Vánky (2001b) originally established Lundquistia for L. fascicularis (syn. L. panici-leucophaei), and later added three other taxa (Vánky 2004c), which were transferred from either Sporisorium or Ustilago. The emended Lundquistia (Vánky 2004c) differed from Ustilago in having spore balls and sterile cells; from Sporisorium in lacking peridia and columellae; and from Macalpinomyces in having permanent or ephemeral spore balls. Molecular phylogenetic analyses showed that Lundquistia was a synonym of Sporisorium as it occurred in the Sporisorium clade (Cunnington et al. 2005, Stoll et al. 2005). Cunnington et al. (2005) included four Lundquistia species in their phylogenetic analysis using the ITS region and demonstrated that it was a polyphyletic group. Vánky (2001b) described Lundquistia as lacking true columellae, whereas, Piepenbring (1999) considered the fascicular vascular bundles mixed with fungal material as columellae in Sporisorium panici-leucophaei (syn. L. panici-leucophaei).

Brefeld (1912) described Anthracocystis for a smut on Panicum miliaceum, which is currently named Sporisorium destruens. He considered it different from Ustilago due to the peculiar formation of its soral peridium, which developed from the floral envelopes. Soral structures such as columellae and spore balls were not included in the protologue (Brefeld 1912). Vánky (2002a) erroneously considered Anthracocystis a nomen nudum and thereby an illegitimate name according to the then ‘International Code of Botanical Nomenclature’. However, Anthracocystis is a validly published name, as it contained a diagnosis and was described in 1912, before Latin was required in taxonomic descriptions.

Vánky et al. (2006) described Anomalomyces as a monotypic genus with shared characters of Ustilago, Sporisorium and Macalpinomyces, but with a unique partitioning of the sorus and two types of sterile cells. They established a new genus based on the peculiar morphology and a phylogenetic analysis that placed Anomalomyces in a polytomy with the Sporisorium groups and the Ustilago group occurring on pooid grasses. Anomalomyces differed from Ustilago by possessing a peridium, spore balls and sterile cells, but did not fit into Sporisorium as it lacked columellae. It differed from Macalpinomyces by possessing genuine spore balls.

Some species fit unambiguously into Sporisorium and Ustilago. Molecular phylogenetic analysis has shown many morphologically similar smut species to be sister to the types of Sporisorium and Ustilago (Stoll et al. 2005). Macalpinomyces was resolved as a monotypic genus (Stoll et al. 2005). The difficulty with the Ustilago-Sporisorium-Macalpinomyces complex has been that many species do not sit strictly within the boundaries of the genera as defined by the types. To resolve this problem, the genera Ustilago and Sporisorium must be re-described and new genera, based on monophyletic groups, must be established to accommodate taxa not included in the emended genera.

DETERMINING A NATURAL CLASSIFICATION OF THE USTILAGO-SPORISORIUM-MACALPINOMYCES COMPLEX

Studies based on spore and ultrastructural morphologies were unable to resolve the Ustilago-Sporisorium-Macalpinomyces complex (Vánky 1991, Piepenbring et al. 1998). Langdon & Fullerton (1975) used soral ontogeny as a means to separate Sporisorium (as Sorosporium) and Ustilago. Molecular phylogenetic analyses showed that there were several monophyletic groups within the Ustilago-Sporisorium-Macalpinomyces complex, but there was no correlation between these groups and their morphological traits (Stoll et al. 2003, 2005). Stoll et al. (2005) noted strong evidence that smuts had co-evolved with their grass hosts, and sister taxa usually occurred on closely related grasses.

Stoll et al. (2005) considered the morphology of columellae, peridia, sterile cells, spore balls and the classification of the hosts (tribe or subtribe) in their molecular phylogenetic analysis of the Ustilago-Sporisorium-Macalpinomyces complex. They mapped these characters onto the hypothesised phylogeny, but none appeared consistently within the monophyletic groups. Stoll et al. (2005) concluded that soral morphology was unsuitable for delimiting genera and resolving the classification of the Ustilago-Sporisorium-Macalpinomyces complex.

Taxa in the Ustilago-Sporisorium-Macalpinomyces complex should not be unified under Ustilago: a case study with smuts on Themeda

Themeda belongs to the grass tribe Andropogoneae in the subfamily Paniceae. Themeda is parasitized by 17 species in the Ustilago-Sporisorium-Macalpinomyces complex, which includes four types of soral morphology (Fig. 4). Several taxa, for example, Sporisorium themedae (Fig. 2g), S. exsertum and S. benguetense (Fig. 4a), infect all the spikelets in an inflorescence, but leave the inflorescence architecture otherwise intact. These species also possess stout or woody columellae. Sporisorium anthistiriae (Fig. 4b) and S. holstii infect individual spikelets in an inflorescence. Species such as Sporisorium enteromorphum (Fig. 4c) and S. langdonii, destroy entire racemes with sori that have several filiform columellae. Macalpinomyces bursus (Fig. 4d) occurs localised in hypertrophied ovaries.

Fig. 4

Four smuts that occur on Themeda. a. Sporisorium benguetense; b. S. anthistiriae; c. S. enteromorphum; d. Macalpinomyces bursus.

Vánky (2001a, 2002a) and Piepenbring (2004) believed one of two approaches were needed to resolve the Ustilago-Sporisorium-Macalpinomyces complex. The first was to synonymise all of the genera under the earliest name, Ustilago, and the second was to split the three genera into smaller genera and subgenera. Unification of the smuts on Themeda into one genus would provide a natural classification, albeit not a very useful one, and to group them based on what appear to be convergent characters would exacerbate taxonomic problems within the complex.

There has been a view that host anatomy dictates the soral morphology of smut taxa (Piepenbring 2004, Stoll et al. 2005). Holton et al. (1968) argued that gross morphology was determined by genotypic or inherently permanent factors. The gross morphology of an infection will be influenced to some extent by environmental factors (Fullerton 1975), but as in the case of the smuts on Themeda, the morphology of the sorus will be distinctive for different species rather than dependant on the structure of the grass.

A diverse range of soral morphologies occur in the Ustilago-Sporisorium-Macalpinomyces complex on other andropogonoid grasses, for example, in Bothriochloa, Sorghum and Heteropogon, which are host to 15, nine and eight smuts, respectively. It will be possible to distinguish genera if soral morphology is synapomorphic. We consider that this diversity necessitates the recognition of new genera or subgenera, rather than the unification of current genera in the complex into Ustilago.

CONCLUSION

Is there a solution to the Ustilago-Sporisorium-Macalpinomyces complex?

It has been approximately 200 years since the genera Ustilago and Sporisorium were first described. These genera contain a diversity of taxa that do not strictly conform to the original generic descriptions. In particular, the genus Macalpinomyces contains many species that have specific characters from both Sporisorium and Ustilago. A stable and workable taxonomy needs to be developed for these important plant pathogens.

Vánky (2002a), Stoll et al. (2005) and Vánky et al. (2006) suggested that analysing additional molecular loci could resolve the Ustilago-Sporisorium-Macalpinomyces complex. It is important to relate synapomorphic characters to monophyletic groups in order to create a meaningful taxonomy (Mooi & Gill 2010). Resolution of the complex will depend on a combined analysis of morphological and molecular characters.

Inclusion of morphological data will help to determine synapomorphies that can be used to define groups within the Ustilago-Sporisorium-Macalpinomyces complex. To accomplish this, a more detailed examination of the soral structures and their development is warranted. Langdon & Fullerton (1975) identified different soral development patterns in several species of Sporisorium, but lacked the advantage of molecular phylogenetic analysis on which to base a new classification. Stoll et al. (2005) considered the presence or absence of columellae and peridia in their study, but did not identify synapomorphies. It is premature to dismiss characters that were thought to be homoplasious, for example spore balls, as a means to delimit genera in the Ustilaginaceae. It is possible that spore balls have evolved independently within monophyletic groups in the Ustilago-Sporisorium-Macalpinomyces complex. Because there are limited morphological characters that can be examined it is necessary to include all the available characters to determine their systematic potential.

Generic concepts of Ustilago, Sporisorium and Macalpinomyces have been refined over the last 30 years, although they still remain polyphyletic genera. The diversity of taxa within the complex requires further delimitation rather than unification of all smuts under Ustilago. Ustilago, Sporisorium and Macalpinomyces need to be revised and a new classification established based on the synapomorphic characters found in monophyletic groups.