Fungal Diversity RReevviieewwss,, CCrriittiiqquueess aanndd NNeeww TTeecchhnnoollooggiieess Biology and recent developments in the systematics of Phoma, a complex genus of major quarantine significance Aveskamp, M.M.1*, De Gruyter, J.1, 2 and Crous, P.W.1 1CBS Fungal Biodiversity Centre, P.O. Box 85167, 3508 AD Utrecht, The Netherlands 2Plant Protection Service (PD), P.O. Box 9102, 6700 HC Wageningen, The Netherlands Aveskamp, M.M., De Gruyter, J. and Crous, P.W. (2008). Biology and recent developments in the systematics of Phoma, a complex genus of major quarantine significance. Fungal Diversity 31: 1-18. Species of the coelomycetous genus Phoma are ubiquitously present in the environment, and occupy numerous ecological niches. More than 220 species are currently recognised, but the actual number of taxa within this genus is probably much higher, as only a fraction of the thousands of species described in literature have been verified in vitro. For as long as the genus exists, identification has posed problems to taxonomists due to the asexual nature of most species, the high morphological variability in vivo, and the vague generic circumscription according to the Saccardoan system. In recent years the genus was revised in a series of papers by Gerhard Boerema and co-workers, using culturing techniques and morphological data. This resulted in an extensive handbook, the “Phoma Identification Manual” which was published in 2004. The present review discusses the taxonomic revision of Phoma and its teleomorphs, with a special focus on its molecular biology and papers published in the post-Boerema era. Key words: coelomycetes, Phoma, systematics, taxonomy. Article Information Received 4 February 2008 Accepted 12 February 2008 Published online 31 July 2008 *Corresponding author: Maikel M. Aveskamp; e-mail: [email protected] Introduction species. Several molecular methods to detect quarantine species have been developed in the The genus Phoma is geographically past decade, e.g. for P. macdonaldii (Miric et widespread and consists of a large group of al., 1999), P. cucurbitacearum (Somai et al., fungi that are found in numerous ecological 2002b; Koch and Utkhede, 2004), P. foveata niches. Besides several harmless saprobic (Macdonald et al., 2000; Cullen et al., 2007) species, Phoma has also been shown to be an and P. tracheiphila (Balmas et al., 2005; important fungal plant pathogenic genus occur- Licciardella et al., 2006), but the validation of ring on economically important crops. Several these methods is heavily questioned as the Phoma species are also of quarantine signifi- species concepts of these taxa are not yet fully cance, posing serious problems to organisations understood. In this literature review we will that are involved in plant health quarantine provide an outline of the biology of the genus, regulation. Identification of isolates found on and circumscribe its taxonomic boundaries. We possible infected material is frequently carried will also focus on the methodologies that can be out under extreme time constraints. Because used to obtain a better understanding of morphological studies are time-consuming, speciation within Phoma. expensive, and require highly skilled person- nel, the chance of successfully identifying Phoma biology strains to species level in such laboratories is The generic name Phoma was in the first often low. Therefore, there is a need for the instance solely reserved for plant stem development of fast, reliable molecular methods pathogens (Saccardo, 1880), but nowadays the for the detection of quarantine actionable Phoma genus comprises pathogens, opportunists as 1 well as saprobes from a much wider range of Fitt et al., 2006). Although recent estimations substrates. The more ubiquitous species such as of financial losses are rare, the impact of this P. herbarum, P. glomerata, P. pomorum var. genus on agriculture is highly significant. pomorum and P. eupyrena have been found on Phoma lingam is regarded to be the most inorganic materials; isolates are known from important pathogen of oilseed rape in the asbestos, cement, oil-paint and plaster (P. Northern Hemisphere, with yield losses of up herbarum), chemicals, paint (P. glomerata) and to 25% being recorded (Fitt et al., 2006). crockery (P. pomorum var. pomorum), along Besides the direct costs due to yield loss, with many other inorganic substrates. These indirect losses occur due to import and export ubiquitous fungi probably also play an important restrictions to prevent introductions of possible role in the degradation of organic materials, pathogenic or quarantine relevant Phoma together with other, more specialized species. species. Rigorous measures, such as refusal or In contrast to these harmless saprobes, more even destruction of a shipment that is suspected than 50% of the species described thus far are of infection with such quarantine organisms, known to be able to occur in living tissue, are the cause of high additional costs to both either as opportunists or as primary pathogens. the exporting and importing traders. These Phoma infections commonly occur in quarantine organisms include amongst others humans and animals. Zaitz et al. (1997) and De P. andigena and P. tracheiphila in the Hoog et al. (2000) referred to nine species that European and Mediterranean regions (Smith et were isolated from humans. Recently, an al., 1992), P. foveata (also known as P. exigua additional species, P. exigua, was added to the var. foveata) in Southern America (Mendes et list of human pathogens (Balis et al., 2006). al., 2007), and P. macdonaldii in Australia Several severe vertebrate diseases are also (Miric et al., 1999). associated with Phoma, such as bovine mycotic Only in some cases the pathogenic nature mastitis (Costa et al., 1993) and fish-mycosis of Phoma is regarded as helpful, by means of in salmon and trout (Ross et al., 1975; Hatai et biocontrol agent of weeds and plant pathogens. al., 1986; Voronin, 1989; Faisal et al., 2007). The ubiquitous species P. herbarum, P. exigua Soil associated organisms such as arthropods and P. macrostoma may play a role as and nematodes can also be subject to Phoma bioherbicide, effective against broadleaf weeds, infections. Chen et al. (1996) listed 11 Phoma such as dandelion (Taraxacum spp.), and species found in association with cyst chickweed (Caryophyllaceae) (Stewart-Wade nematodes. Furthermore, Phoma species have and Boland, 2004, 2005; Zhou et al., 2004), been found parasitizing other fungi and clematis (Clematis vitalba) (Paynter et al., oomycetes (Hutchinson et al., 1994; Sullivan 2006) and salal (Gaultheria spp.) (Zhao et al., and White Jr., 2000). Although lichens have 2005, 2007). The antagonistic effect of P. been poorly studied, a key to 14 lichenicolous glomerata and P. etheridgei, respectively against Phoma species was provided by Hawksworth Microsphaera penicillata and Phellinus and Cole (2004). Still, the vast majority of the tremulae, has been studied in detail. However, species appear to only colonise plant material. their application as biocontrol agent is deba- Some species are only known from decaying table, due to application problems and a low leaves or wood, whereas others play a role as impact on the disease spread (Hutchinson et secondary invaders of weakened plant tissue. al., 1994; Sullivan and White Jr., 2000). However, more than 110 species are known to be primary plant pathogens, mainly speciali- The life cycle sing on a single plant genus or family. Although the life cycle is influenced by The economically most important patho- the occupied niche, it is relatively similar for gens include the widespread species P. all plant pathogenic Phoma species. Primary medicaginis (Broscious and Kirby, 1988) and infection of hosts may occur through wounds the two species that are involved in Black Leg that are caused by cultivation practices, weather in Brassicaceae: P. lingam and to a lesser conditions, or interaction with other organisms. extent the unnamed Phoma anamorph of Lepto- In several species, entering a host plant through sphaeria biglobosa (Gugel and Petrie, 1992; stomata or directly through the epidermis also 2 Fungal Diversity may occur (Williams, 1992; Agrios, 1997; sex in vitro. Roustaee et al., 2000; West et al., 2001; Van The abundant production of conidia, de Graaf et al., 2002). Initially, the fungal together with a relatively fast growth rate and hyphae grow intercellulary through plant the capability to invade a large number of hosts tissues (Hammond and Lewis, 1987). Follow- and substrates, are the main reasons why the ing this symptomless stage, the fungus genus is so cosmopolitan in distribution. The becomes necrotrophic. Host cells are subject to most frequent occurring species, P. eupyrena, phytotoxification or to a hypersensitive response, P. exigua, P. glomerata, P. herbarum, and P. after which the fungus has acces to the macrostoma are found worldwide, irrespective resources of the dead plant tissue. This can be of the climatic conditions. Phoma herbarum is observed macroscopically as the formation of probably also indigenous to Antarctica (McRae lesions. After a short period, often dark- and Seppelt, 1999; Tosi et al., 2002). Several coloured, mostly globose or flask-shaped plant pathogenic species have been transmitted pycnidial conidiomata can be observed within worldwide with the cultivated host on seeds the lesions, which are embedded in the plant’s and other plant material. epidermis. These pycnidia contain numerous conidia, which ooze in a pale white to pinkish The Phoma generic concept coloured matrix. A detailed description of the The first descriptions of fungi that belong spore production is provided by Boerema to the genus Phoma date back to 1821 (Sutton, (1965), Brewer and Boerema (1965) and 1980). Nevertheless, it took until 1880 before Boerema and Bollen (1975). In some cases the generic name was officially introduced by extradermal mycelium may be formed. Conidia, Saccardo, and was later emended by Boerema and in some species mycelial fragments, disperse and Bollen in 1975. In the Saccardoan system, easily by water-splash, misting or wind, and the genus name Phoma applied to filamentous can thus infect new host plants. Also birds and fungi which were capable of forming pycnidial insects may act as vectors (Perrotta and Graniti, conidiomata with aseptate, hyaline conidia that 1988). In absence of a suitable host, due to crop could inhabit plant stems. Fungi with the same harvesting for example, most species persist as morphological appearance, but found in saprobes on decaying organic material in the association with leaf spots, were placed in the soil. In most cases, this material is the residue genus Phyllosticta (Van der Aa and Vanev, of plants that were previously infected. In this 2002; Boerema et al., 2004). mode, the mould may survive periods of stress, Identification of fungi occurred mainly in such as drought or extreme cold. The structures vivo until the fourth decade of the 20th century. which are most suitable for long-term survival Wollenweber and Hochapfel (1936) were the are the conidiomata, and the uni- or multicellu- first to recognise the advantages of using lar chlamydospores, which are formed only by growth characteristics on artificial substrates in a small number of Phoma species. From this Phoma taxonomy, whereas Dennis (1946) material infection may occur in newly planted launched the use of basic morphology, physio- crops, and from the exuding conidiomata new logy and biochemical tests in vitro. In a study conidia can emerge to facilitate onward dispersal. on hyphomycetes, Hughes (1953) introduced conidiogenesis as an important taxonomic Sometimes also a meiotic cycle can be criterion, a feature that was later applied to all observed, besides the asexual one described conidial fungi including species of Phoma above, in which ascospores are formed in (Sutton, 1964). In addition, new criteria for pseudothecial ascomata. If such sexual struc- morphological differentiation were introduced tures are found in nature, they are mostly to distinguish Phoma species from other observed in the saprobic part of the life cycle coelomycetes such as Ascochyta (Boerema and (Williams, 1992; West et al., 2001). However, Bollen, 1975), Phyllosticta (Van der Aa et al., such a sexual stage is quite uncommon within 1990), Coniothyrium and Paraconiothyrium this genus, and is presently not known for more (Verkley et al., 2005). than 40 taxa, of which many produce the sexual In 1964, eight years after the designation reproductive structures only in vivo. Unfortu- of Phoma as a conserved genus name in the nately, little is known about the induction of 3 International Code of Botanical Nomenclature spora (Chaetothyriales), a fungus associated (Lanjouw et al., 1956), the designated type with Petri disease of grapevines (Mostert et al., species, Phoma herbarum, was restudied and 2006). This complexity in taxonomy is a futher lectotype material was selected (Boerema, complicating factor for the identification and 1964). Morphology, synonymy and ecology of differentiation of members of the genus this type species were later extensively described Phoma. by Boerema (1970) and Morgan-Jones (1988a). Amongst the many scientists that attemp- Currently Phoma species can be defined ted to order the numerous species belonging to as filamentous fungi that produce pycnidial Phoma are Saccardo (1884), Grimes et al. conidiomata with monophialidic, doliiform to (1932), Wollenweber and Hochapfel (1936), flask-shaped conidiogenous cells. A collarette Dennis (1946), Sutton (1964) Monte et al. is present at the apex of those cells after the (1990, 1991) and Rai and Rajak (1993). production of the first conidium. In vitro the Morgan-Jones and co-workers described some hyaline conidia are mainly single-celled, important species in detail in several papers although in several species a small percentage published in Mycotaxon (Morgan-Jones and of transversely septate conidia may also be White, 1983; White and Morgan Jones, 1983, observed. This definition applies to more than 1986, 1987; Morgan-Jones and Burch 1987a,b, 220 inter- and intraspecific taxa (Boerema et 1988a,b; Morgan-Jones, 1988a,b). The greatest al., 2004), but it should be borne in mind that effort made on Phoma systematics was by this definition does not necessarily apply to Boerema and co-workers, who published samples in nature, as bicelled or even pigmen- contributions towards a genus concept includ- ted conidia occur more often in vivo. ing numerous species descriptions and syno- The genus Phoma is generally considered nyms in a series of papers in Persoonia (De by most modern mycologists to be taxonomi- Gruyter and Noordeloos, 1992; Boerema 1993, cally problematic due to ambiguous morpholo- 1997, 2003; Boerema et al., 1994, 1996, 1997, gical criteria, but also due to uncertain phylo- 1999; Boerema and De Gruyter 1998, 1999; De genetic affinities. The genus is currently linked Gruyter et al., 1998, 2002; Van der Aa et al., to three different teleomorph genera. If teleo- 2000; De Gruyter, 2002; De Gruyter and morphs are known, they reside in Didymella, Boerema, 2002). As a final product of their 40- Leptosphaeria, or occasionally Pleospora year study on Phoma taxonomy, a taxonomic (Table 1, Fig. 1). It should be noted that not all handbook was written based on the papers species in these genera form Phoma ana- published in the previous decades (Boerema et morphs; this is a feature that may have been al., 2004). The Phoma generic concept, and lost over time, or developed multiple times in therefore also the identification key is based on multiple ancestors. In literature, Phoma species the establishment of nine sections within the have also been associated with other genus (Boerema, 1997), which are listed in teleomorph genera within the Dothideomy- Table 1. cetes, including Mycosphaerella (Punithalin- gam, 1990; Corlett, 1991; De Gruyter et al., Table 1. The nine sections of the genus Phoma, with their type species and associated teleo- 2002), Belizeana (Kohlmeyer and Volkmann- morph genera. Kohlmeyer, 1987), and Fenestella, Cucurbi- taria, Preussia, and Westerdykella (Von Arx, Section Type species Associated 1981). As many of these connections are based teleomorph on association, they remain to be confirmed in genus culture. Furthermore, a range of synanamorphs Heterospora P. heteromorphospora - of Phoma spp. have been recognised in Macrospora P. zeae-maydis Didymella Paraphoma P. radicina - Stagonosporopsis, Epicoccum, Phialophora Peyronellaea P. glomerata - and Sclerotium (Boerema and Bollen, 1975; Phoma P. herbarum Didymella Sutton, 1977; Boerema, 1993; Boerema et al., Phyllostictoides P. exigua var. exigua Didymella 1994, 1997; Arenal, 2000, 2004). Crous and Pilosa P. betae Pleospora Gams (2000) also described a phoma-like Plenodomus P. lingam Leptosphaeria Sclerophomella P. complanata Didymella synanamorph of Phaeomoniella chlamydo- 4 Fungal Diversity Fig. 1 A-D. Teleomorphs of Phoma. Asci and ascospores of A-B. Didymella zeae-maydis (anam. P. zeae-maydis) C. Leptosphaeria maculans (anam. P. lingam) D. Pleospora herbarum, lectotype of the genus Pleospora. Scale bars = 10 µm. Although the key is helpful for identi- Towards a generic concept for Phoma fication of species, it is still uncertain if this Macromolecular approaches to systema- division into sections can be considered natural tics were introduced in mycology in the late from an evolutionary perspective. In this 80’s (for a review see Bruns et al., 1991). classification system, most sections are to be These techniques, although accurate, were based on morphological characters that imply a providing only a limited number of relevant certain evolutionary relationship, or comprise characters. The introduction of nucleotide species that share a teleomorph in the same sequences as phylogenetic characters greatly genus. Unfortunately, several characters that advanced fungal systematics (Bridge, 2002; are linked to a certain section, sometimes also Malgoire et al., 2004). Within the genus seem to occur in species that are placed in other Phoma this gave rise to several novelties, and taxonomic groups. As an example, Punitha- to new insights on the delimitation of Phoma lingam (2004) mentioned the slightly pilose and its teleomorphs, as described below. species, P. anserina and P. leonuri that are not The present classification system has accommodated in either section Pilosa or been criticized for the unclear discrimination Paraphoma. Based on the appearance of their between Phoma and several related or even conidia, these species have been placed in the convergent genera. For example, Grondona et sections Phoma and Plenodomus, respectively. al. (1997) discussed the poor delineation This ambiguous approach has resulted in between the Phoma section Paraphoma and the multiple overlaps between the separate genera Pyrenochaeta and Pleurophoma. sections. Furthermore, the sections Phoma and Furthermore, the relation between Ascochyta Phyllostictoides are also considered to be and Phoma species that produce septate conidia artificial. In stead of comprising species with a remains unresolved. The confusion between the shared feature, both sections seem to be a two genera is illustrated by the large number of repository for species that lack the presence of shared synonyms. Although both genera have good sectional characters. According to teleomorphs in Didymella, Boerema and Bollen Boerema (2004), section Phoma comprises (1975) differentiated Phoma from Ascochyta species “that have much in common with P. on differences in conidiogenesis and conidial herbarum”. But as the majority of the species septation. The conidia of Ascochyta are basi- are accommodated within this section, it shows cally always 2-celled, due to early euseptation the necessity to conduct further research on during conidiogenesis, whereas in Phoma classification within the genus. septate conidia are rare in culture - although it 5 Fig. 2A-D. Pycnidial types. A. Regular glabrous as in P. herbarum B. Setose as in P. carteri C. Pilose as in P. betae D. Pycnosclerotia-like pycnidia as in P. incompta. Scale bars = 100 µm. is not uncommon in nature - and are formed by the link between Plenodomus lingam and the a late euseptation. Spectrometrical analysis of genus Phoma due to differences in pycnidial crystals formed by specimens of both genera development. Later, the authors justified the seemed to support this differentiation (Noorde- recombination based on the similarity in loos et al., 1993). However, later Faris-Mokaiesh conidiogenesis: both taxa produced conidia et al. (1995) showed that P. pinodella and A. from unicellular, flask-shaped phialides, a pinoides probably are closely related, basing feature that is considered to be specific for the their conclusions on a similarity in the banding genus Phoma (Boerema et al., 1981). This pattern of PCR and RFLP products. This decision arose again as point of discussion in relatedness between the two species was 1997, when molecular data of the ITS-region supported by sequence analyses of various revealed that P. lingam and its relative P. gene regions (Barve et al., 2003; Fatehi et al., wasabiae were only distantly related to other 2003; Peever et al., 2007). phomoid fungi (Reddy et al., 1998). This led to A section that has been the subject of the hypothesis that the former genus Plenodo- continuous discussion is Plenodomus. In the mus had been incorrectly reduced to synonymy, mid-70s, the genus Plenodomus was incorpo- and should be reinstated (Reddy et al., 1998). rated into Phoma as Phoma section Plenodo- This idea agrees with other, more recent studies mus (Boerema et al., 1981). Presently 32 taxa using other genes (Pethybridge et al., 2004; are accommodated in this section (Boerema et Torres et al., 2005b; Schoch et al., 2006), al., 1994, 1996, 2004; Torres, 2005b). Already although the interspecific variation in Plenodo- in 1964, even before the official recombination mus-linked Leptosphaeria species is relatively of this group of fungi into the genus Phoma, high (Morales et al., 1995; Câmara et al., Boerema and van Kesteren (1964) questioned 2002). 6 Fungal Diversity The synanamorphs and teleomorphs of tively). The formation of unicellular chlamydo- Phoma have been poorly investigated thus far, spores (Fig. 3F-H) is not regarded as charac- except for the P. lingam complex. Few teristic for any particular section, but can aid synanamorph relationships have thus far been identification at the species level. In the same confirmed by means of molecular techniques. way the characters of swollen hyphae, com- Arenal et al. (2000, 2004) demonstrated P. monly regarded as ancestral to chlamydo- epicoccina and Epicoccum nigrum to be spores, are regarded as informative (Boerema, synanamorphs by employing ITS sequence 1993). data, and synonymised these species after Conidia are formed in pycnidia, and further microscopical studies. conidial shape and size are regarded as the most useful indicators for determining a strain Towards a species concept for Phoma up to species level, but are also used for Numerous Phoma strains have been differentiating some sections (Fig. 4). The identified and characterised using mainly average conidium measures ca. 2.5-10 × 1-3.5 morphological characters in vitro. As in most µm. However, in the section Macrospora other fungal genera, these characters include enlarged conidia occur, which are up to 25 × 9 the size and shape of conidiomata, conidia and µm (Fig. 4K-L). The section Heterospora chlamydospores, but also metabolite produc- comprises species that posses both normal tion and growth rates and patterns on various sized conidia as well as so-called macroconidia agar media. (Fig. 4I-J). Especially in these sections there Characters relating to pycnidial conidio- might be a high variability in conidial size, mata are mainly used for sectional differen- even within a single species. Studying cultures tiation (Fig. 2). Pycnidia are highly variable in in vitro should be done under standardised shape and size, but in most species they are conditions: cultures grown on different media either globose or subglobose, or sometimes and under different conditions may prove to be pyriform due to an elongated neck (section highly variable (Rai, 1998). Besides conidial Plenodomus). In older cultures the pycnidia measurements, conidial shape is of primary may aggregate. The colour varies per species importance. Guttule number and size may from yellowish to brown-olivaceous or olivace- provide additional valuable characters for ous-black and depends on the culturing condi- species identification. The section Phyllostic- tions and age. Occasionally the pycnidia are toides (Fig. 4E-H) can be distinguished from setose as in section Paraphoma (Fig. 2B), or the section Phoma (Fig. 4 A-D) by the pilose as in section Pilosa (Fig. 2C). Single or presence of a low percentage of septate conidia multiple ostioles may be observed, although in in pure culture. The septation ratio is highly some species (sections Sclerophomella and variable, even within species and can be Plenodomus) pycnosclerotia are found (Fig. influenced by the growth media. Therefore, this 2D). In contrast to most sections that produce character is often regarded as uninformative thin-walled pseudoparenchymatous pycnidia, (Onfroy et al., 1999). In vivo, the ratio of the species in Sclerophomella are characterised septate conidia may be up to 95%, which has by the ability to produce thick-walled conidio- led to many misidentifications in Ascochyta in mata, whereas species in section Plenodomus the past (Boerema and Bollen, 1975), and a produce scleroplectenchyma in the pynidial large number of synonyms for Phoma species wall. The presence of multicellular chlamydo- in Ascochyta (Van der Aa et al., 2000). spores (Fig. 3) is often a good indication that a Growth characteristics on media, such as strain belongs to the section Peyronellaea, growth rate, pigment formation and colony although this might not always be the case. The outline and pattern can also aid identification. shape of these chlamydospores is in most cases The media types that are most intensively comparable to the multicellular conidia of applied in Phoma identification include oat- Alternaria spp., so-called alternarioid or meal agar, malt extract agar and, to a lesser alternarioid-botryoid (Fig. 3A-D), but in some extent, cherry decoction agar (for protocols see species pseudoscleroid or epicoccoid chlamy- Boerema et al., 2004). Boerema et al. (2004) dospores can be found (Fig. 3E and 3I respec- provide measurements of the diameter of 7 Fig. 3 A-I. Chlamydospore morphology in Phoma. A. Chain of chlamydospores of P. glomerata. B. Alternaroid chlamydospores of P. glomerata. C. Alternaroid chlamydospores of P. jolyana. D. Botryoid chlamydospore as in P. sorghina. E. Unicellular (u) and pseudosclerotoid multicellular chlamydospores (pm) in P. chrysanthemicola. F. Unicellular chlamydospores of P. pinodella G. Chain of unicellular chlamydospores in P. clematidina. H. Chain of unicellular chlamydospores in P. chrysanthemicola. I. Epicoccoid chlamydospores in P. epicoccina. Scale bars: A = 100 µm, G = 50 µm, B-F and H-I = 10 µm. 8 Fungal Diversity Fig. 4 A-L. Conidial morphology in Phoma. A-D. Conidia hyaline, aseptate small-sized as in A. P. herbarum B. P. multirostrata C. P. astragali and D. P. eupyrena. E-H. Conidia hyaline, occasionally septate, small-sized as in E. P. exigua F. P. cucurbitacearum G. P. macrostoma and H. P. polemonii. I-J. Conidia generally hyaline small-sized, but always also large septate and often pigmented condia occur as in I. P. actaeae and J. P. schneiderae. K-L. Conidia hyaline, occasionally septate, but all relatively large as in K. P. rabiei and L. P. xanthina. Scale bars = 10 µm. 9 Phoma colonies after 7 d of incubation at 20- all the species-specific characters. As a result, 22ºC in complete darkness, and complete up to 2003 almost one fifth of the total number colony descriptions after a further incubation of Phoma strains of which sequences were period of 7d at the same temperature, but at a submitted to public sequence databases such as UV : dark interval of 11:13 (Boerema et al., GenBank were found to be incorrectly 2004). Unfortunately, within a single species, identified (Bridge et al., 2003). This trend is wide variability may be observed, especially not unique to Phoma, however, as this the presence of sector formations as mycelial or percentage is slowly increasing over the years, pycnidial zones. Again, standardisation is and by 2006, close to 27% of the total public essential, as slight alterations in the media may fungal sequences were derived from incorrectly be the cause of observed differences in growth identified strains (Nilsson et al., 2006). This characteristics (Rai, 1998). may also be due to the existence of so-called The use of biochemical reactions and species complexes, in which multiple physiological tests to indicate the presence of morphologically indistinguishable taxa are certain metabolites was common practice in occupying the same ecological niche, but are Phoma systematics (Wollenweber and Hochap- only distantly related from an evolutionary fel, 1936; Dennis, 1946; Boerema and point of view. Further, it seems to have become Höweler, 1967; Dorenbosch, 1970; Monte et common practice to sequence fungal strains al., 1990, 1991; Noordeloos et al., 1993). The without even identifying them morphologically application of alkaline reagents (KOH, NaOH) (Hyde and Soytong, 2007). Bridge et al. (2004) on fresh cultures is still used as it may change suggested that the nucleotide sequences the colour of pH dependent metabolites and deposited as P. herbarum in GenBank probably pigments (Boerema and Höweler, 1967; Doren- represent at least two different taxa. Because P. bosch, 1970). Furthermore, the biochemical herbarum is the type species of the genus analysis of dendritic crystals formed in older Phoma (Boerema, 1964; Morgan-Jones, 1988a), cultures can be used for the identification of a it should be unambiguously clarified to which strain up to species level (Noordeloos et al., taxon the name P. herbarum should be applied. 1993). Although the development of such Several other species complexes have techniques in mycology is not yet fully been revealed using molecular techniques. In optimised, the interest in such methods for this context the P. lingam species complex can species identification is decreasing as such be mentioned. A natural variance in virulence specific methods can only be applied on a and pathogenicity within P. lingam was limited scale. Instead, molecular labs are more observed in the United States (Pound, 1947) commonly equipped to perform more routine and Canada (McGee and Petrie, 1978). These and popular DNA-based methods, which are findings were found to occur worldwide, as thought to be more consistent. high genetic diversity amongst isolates of this species was observed by e.g. Johnson and DNA-based approaches in Phoma taxonomy Lewis (1990), Schäfer and Wöstemeyer (1992), Modern DNA-based techniques can Morales et al. (1993), Pongam et al. (1999), greatly contribute to identification and taxono- Williams and Fitt (1999), Purwantara et al. my of fungal species. Nevertheless, the actual (2000), and Voigt et al. (2001). Molecular number of studies using macromolecular typing tools aided in distinguishing two species approaches to define new species in Phoma and within P. lingam (teleomorph: L. maculans), resolve species complexes within this genus is which could hardly be separated using solely relatively low (e.g. Shoemaker and Brun, 2001; morphological characters. The teleomorph of Bridge et al., 2003, 2004; Torres et al., 2005a, the weakly aggressive variant was described as b). L. biglobosa with an unnamed anamorph in the To identify strains up to species level, a genus Phoma (Shoemaker and Brun, 2001). high level of expertise is required. Never- Leptosphaeria biglobosa comprises 3 to 5 theless, misidentifications in morphological separate taxa itself, whereas the remaining taxonomy of Phoma species cannot always be isolates in P. lingam also seem to be a avoided, as not all isolates may fully express heterogeneous assemblage of cryptic taxa 10