Colletotrichum – current status and future directions

Abstract

A review is provided of the current state of understanding of Colletotrichum systematics, focusing on species-level data and the major clades. The taxonomic placement of the genus is discussed, and the evolution of our approach to species concepts and anamorph-teleomorph relationships is described. The application of multilocus technologies to phylogenetic analysis of Colletotrichum is reviewed, and selection of potential genes/loci for barcoding purposes is discussed. Host specificity and its relation to speciation and taxonomy is briefly addressed. A short review is presented of the current status of classification of the species clusters that are currently without comprehensive multilocus analyses, emphasising the orbiculare and destructivum aggregates. The future for Colletotrichum biology will be reliant on consensus classification and robust identification tools. In support of these goals, a Subcommission on Colletotrichum has been formed under the auspices of the International Commission on Taxonomy of Fungi, which will administer a carefully curated barcode database for sequence-based identification of species within the BioloMICS web environment.

INTRODUCTION

The genus Colletotrichum includes a number of plant pathogens of major importance, causing diseases of a wide variety of woody and herbaceous plants. It has a primarily tropical and subtropical distribution, although there are some high-profile species affecting temperate crops. Fruit production is especially affected, both high-value crops in temperate markets such as strawberry, mango, citrus and avocado, and staple crops such as banana. Colletotrichum species cause devastating disease of coffee berries in Africa, and seriously affect cereals including maize, sugar cane and sorghum. The genus was recently voted the eighth most important group of plant pathogenic fungi in the world, based on perceived scientific and economic importance (Dean et al. 2012).

As plant pathogens, Colletotrichum species are primarily described as causing anthracnose diseases, although other maladies are also reported such as red rot of sugar cane, coffee berry disease, crown rot of strawberry and banana, and brown blotch of cowpea (Lenné 2002). Anthracnose disease symptoms include limited, often sunken necrotic lesions on leaves, stems, flowers and fruit, as well as crown and stem rots, seedling blight etc. (Waller et al. 2002, Agrios 2005). A range of disease symptoms is illustrated in Fig. 1. Many species may be seed-borne and can survive well in soil by growing saprobically on dead plant fragments, and may be spread via water-splash dispersal of conidia and air transmission of ascospores from the sexual morph (Nicholson & Moraes 1980). Infection occurs via an appressorium that develops from the germinating spore on the plant surface, followed by turgor-driven penetration of the cuticle (Deising et al. 2000) and in some cases also of epidermal cells by infective hyphae (Bailey et al. 1992). Establishment within plant tissues is aided via production by the fungus of host-induced virulence effectors (Kleeman et al. 2012, O’Connell et al. 2012). Nascent colonies in most cases then enter a biotrophic phase with infected tissues remaining externally symptomless and which may be short (1–3 d; O’Connell et al. 2000) or extended and presumably involving dormancy (Prusky & Plumbley 1992). Then, the fungus enters a necrotrophic phase that results in significant death of plant cells and the emergence of pathogenic lesions. This delayed onset of disease symptoms may lead to significant post-harvest losses, with apparently healthy crops degenerating in storage (Prusky & Plumbley 1992). The biotrophic life strategies adopted by Colletotrichum species may also contribute to their prominence as symptomless endophytes of living plant tissues (Lu et al. 2004, Joshee et al. 2009, Rojas et al. 2010, Yuan et al. 2011). There are no comprehensive modern reviews of the biology, pathology and host/parasite interactions of Colletotrichum species, but useful information can be found in Bailey & Jeger (1992) and Prusky et al. (2000).

Fig. 1A–L.

(see page 182). Disease symptoms caused by Colletotrichum species. The causal organisms have in most cases been identified to species complex level only. A. Anthracnose on strawberry fruit caused by C. nymphaeae (acutatum clade). B. Leaf spot of Brachyglottis repanda caused by C. beeveri (boninense clade). C. Anthracnose symptoms on leaves of Tecomanthe speciosa caused by C. boninense agg. D. Anthracnose of onion bulb caused by C. circinans (dematium clade). E. Anthracnose of banana caused by C. musae (gloeosporioides clade). F. Coffee berry disease caused by C. kahawae subsp. kahawae (gloeosporioides clade). G. Leaf anthracnose of yam caused by C. gloeosporioides agg. H. Anthracnose of aubergine (eggplant) fruit caused by C. gloeosporioides agg. I. Blossom blight of mango caused by an undetermined Colletotrichum sp. J. Anthracnose of mango caused by C. gloeosporioides agg. K. Leaf blight of maize caused by C. graminicola (graminicola clade). L. Anthracnose of bean pod caused by C. lindemuthianum (orbiculare agg.). A, © Ulrike Damm/CBS. B, C, D, H © Landcare Research, New Zealand. E, F, K © Jim Waller/CABI. G © Paul Cannon/CABI. I, J © Barbara Ritchie/CABI. L © Lu Guo-zhong, Dalian, China.

Colletotrichum species are also extensively studied as model organisms for research into genetics. This work has a long history; the first investigation into mating types in Glomerella was published a century ago (Edgerton 1912, 1914), and genetic mechanisms in G. cingulata were extensively studied in the 1940’s and 50’s (e.g. Andes 1941, Lucas et al. 1944, Wheeler 1950, 1954, Olive 1951).

Research into host/parasite systems has had almost as long a history, originating with work on the C. lindemuthianum/Phaseolus vulgaris interaction by Barrus (1918). Mechanisms of infection and disease development in the same model system were extensively studied in the 1980’s (e.g. Bell et al. 1984, O’Connell et al. 1985, 1986).

Maize anthracnose caused by Colletotrichum graminicola is an economically important disease on a global level, stimulating a further body of research into Colletotrichum genetics, pathology and host-parasite interactions. It has been reviewed by Nicholson (1992), Bergstrom & Nicholson (1999), Vaillancourt et al. (2000) and Crouch & Beirn (2009).

The relationship between Colletotrichum higginsianum and its Brassica hosts has also been the subject of much recent research (Perfect et al. 1999, O’Connell et al. 2004). Huser et al. (2009) discovered pathogenicity genes in C. higginsianum by random insertional mutagenesis. Jaulneau et al. (2010) compared the defence reactions of resistant or susceptible lines of Medicago truncatula to the alfalfa pathogen C. trifolii with reactions of the nonadapted pathogens C. lindemuthianum and C. higginsianum. O’Connell et al. (2012) studied the genomes and transcriptomes of two species, C. higginsianum and C. graminicola with different infection strategies.

Work on the genetics of pathogenicity in the C. orbiculare species aggregate (e.g. Pain et al. 1994, Rodriguez & Redman 2000) led to transformation of pathogenic strains to endophytic forms. These were shown to exhibit mutualistic activity by protection against virulent strains of the same species, and also to Fusarium pathogens. Gene manipulation techniques such as Agrobacterium tumefaciens-mediated transformation or protoplast transformation are established (Tsuji et al. 2003) and for host parasite interaction studies with C. orbiculare, a model plant Nicotiana benthamiana is being used. Several genes involved in signal transduction pathways essential for the formation of infection structures were identified (Takano et al. 1997, Tanaka et al. 2009) and two peroxisome biogenesis genes, PEX6 and PEX13 that are essential for pathogenesis were functionally analysed (Kimura et al. 2001, Fujihara et al. 2010). Asakura et al. (2009) discovered the importance of the pexophagy factor ATG26 for appressorium function.

Whole-genome sequences of C. graminicola and C. higginsianum have been completed (O’Connell et al. 2012) – the latter genome from a pathogen of the model plant organism Arabidopsis thaliana – and projects to sequence several other species are in progress or preparation (Damm et al. 2010). The research to date is already demonstrating step changes in our understanding of host-parasite interactions in Colletotrichum.

Colletotrichum is traditionally recognised as an asexual genus of fungi, with a number of species linked to sexual morphs assigned to the genus Glomerella (Glomerellaceae, Glomerellales; Zhang et al. 2006, Réblová et al. 2011). In the light of recent moves towards a unified nomenclatural system for the Fungi, we will for the most part refer to species using asexual names, which not only have date priority in all cases we have identified, but are much better known in the applied sciences.

HOST RELATIONS AND SPECIFICITY

For many years, Colletotrichum species were assumed to be specific to the plants they infected, leading to large numbers of taxa described with little in the way of distinctive features apart from the identity of their plant partners.

Our current understanding of the extent that Colletotrichum species exhibit host specificity is imperfect. This is due to a number of factors, including incomplete sampling, restriction of data largely to populations affecting crop or ornamental plants, and poor knowledge of pathogenic effects. Information on most strains in culture collections indicates an association with a particular plant species, but rarely provides details of the interaction. Many studies on Colletotrichum are restricted to strains affecting single crop species (e.g. Buddie et al. 1999, González et al. 2006, Gazis et al. 2011), significantly reducing the extent of the gene pool being sampled. Mackenzie et al. (2007) demonstrated gene flow between populations of C. acutatum from native plants and those from adjacent strawberry crops, demonstrating the limitations of host-restricted studies.

The ability of many Colletotrichum species to exist as endophytes adds extra complication to our understanding of host specificity (Lu et al. 2004, Liu et al. 2007, Rojas et al. 2010). Isolation from living plant tissue does not necessarily imply that the species is a latent pathogen with a hemibiotrophic phase (Latunde-Dada 2001, Peres et al. 2005, Münch et al. 2008), and distinguishing between the two life strategies is problematic. Freeman & Rodriguez (1993) and Redman et al. (1999) demonstrated that a single disruption event of a pathogenicity gene transformed a pathogenic strain of Glomerella magna from Citrullus lanatus into an endophyte that conferred protection for the host plant against wild type strains and other pathogens. Similar single gene effects on pathogencity are documented from the interaction between C. graminicola and maize (Thon et al. 2000, 2002). Research into the molecular basis of host-parasite interactions in Colletotrichum is currently highly active (see O’Connell et al. 2012), and such approaches will dominate research in the future into the extent of host specificity exhibited by Colletotrichum species.

We are not aware of any major group of angiosperms that does not harbour endophytic Colletotrichum colonies. There are also well-documented cases of Colletotrichum living as endophytes and disease agents of conifers (Dingley & Gilmour 1972, Wang et al. 2008, Joshee et al. 2009, Damm et al. 2012a) and ferns (Leahy et al. 1995, MacKenzie et al. 2009). Species are associated widely with both herbaceous and woody plants, though the latter appear mainly to contain colonies in fruits, leaves and other non-lignified tissues.

There are isolated accounts of Colletotrichum species causing infections of insects, including C. fioriniae on hemlock scale insects in New England and a claimed member of the C. gloeosporioides aggregate on citrus scale insects in Brazil (Marcelino et al. 2008). Infection mechanisms are not fully understood; under experimental conditions the insects became infected after being sprayed with a conidial suspension (Marcelino et al. 2009). In the field it seems possible that endophytic colonies of the fungus are ingested via the insect mouth-parts, the reverse of a process that has been shown in members of the Clavicipitaceae to infect plants via the stylets of sap-sucking insects (Torres et al. 2007, Tadych et al. 2009).

In rare instances, Colletotrichum species have been implicated in human disease, causing keratitis and subcutaneous infections (e.g. Ritterband et al. 1997, Guarro et al. 1998, Shiraishi et al. 2011, Shivaprakash et al. 2011). A single occurrence of disseminated mycotic infection of a sea turtle has also been recorded (Manire et al. 2002). Cano et al. (2004) reviewed the identification procedures for Colletotrichum species of clinical interest.

Some Colletotrichum clades appear to contain species that show at least a degree of host specificity, though these data may be linked to incomplete sampling and/or species concepts that assume specificity. The orbiculare clade is a case in point; here species seem to be restricted to individual host genera (Liu et al. 2007). That clade is a basal group (see Fig. 2), which might suggest that the extraordinary flexibility in host preference demonstrated by most other clades evolved subsequent to appearance of the genus itself. The graminicola group contains several species that are limited to host genera within the Poaceae (Crouch et al. 2009a). Colletotrichum cereale, a grass-inhabiting taxon which occupies a separate clade from the graminicola aggregate, does not appear to show genus-level specificity, though all strains to date derive from the same family (Crouch et al. 2009c). Here, population-level specificity is found in some cases, though the basal lineage is plurivorous, suggesting that host specialisation is in the process of development.

Fig. 2.

Phylogenetic tree derived from a Bayesian analysis of an alignment of ITS (599 bp) sequences, run for 1×10⁷ generations with a GTR+I+Γ model of DNA evolution. Species names are followed by culture number, and status of the culture, where HT = ex-holotype, ET = epitype, NT = ex-neotype, IT = ex-isotype, AUT = authentic culture. Sequences from a number of non-validated cultures have been included in order to represent clades that have not yet been subject to revision based on multilocus data. These are indicated by the symbol §^*.

At a finer scale, several Colletotrichum species have been shown to exhibit substantial pathogenic variation at race level, although in most cases the precise phylogenetic position and diversity of the strains studied has not been established. In a large-scale project on strains identified as C. lindemuthianum from South, Central and North America, Balardin et al. (1997) characterised 41 races from a total of 138 isolates, based on virulence to 12 cultivars of Phaseolus vulgaris. No coevolutionary pattern between fungus and plant was detected, but greatest pathogen diversity occurred in Central America, which is the centre of origin of the host plant. In a similar study, 90 pathotypes were detected by Mahuku & Riascos (2004) from 200 isolates collected in Central and South America. Greater diversity was detected in the Mesoamerican region compared with Andean populations. Sharma et al. (2007) conducted a similar study in north-west India, detecting substantial further diversity with 29 pathogenic races from a pool of 90 isolates, of which 17 had not been reported by Mahuku & Riascos (2004). On a smaller scale, six different races of C. lindemuthianum were reported from two counties in the state of Minas Geraes, Brazil (Pinto et al. 2012), demonstrating complex population structure within a small area. Heterothallic mating and teleomorph formation were demonstrated for C. lindemuthianum by Rodriguez-Guerra et al. (2005). This body of research provides indications that the taxon concerned is undergoing rapid evolutionary change.

Variability and evolution at population level have been investigated for other species and species clusters in Colletotrichum including C. acutatum (e.g. Freeman et al. 2000, 2001, Denoyes-Rothan et al. 2003, Peres et al. 2008), C. cereale (Crouch et al. 2008, 2009d), C. coccodes (Ben-Daniel et al. 2010), C. gloeosporioides (Cisar et al. 1994, Cisar & TeBeest 1999), C. graminicola (e.g. Vaillancourt et al. 2000, Chen et al. 2002, Valèrio et al. 2005), C. sublineola (Rosewich et al. 1998), C. “truncatum” (actually a member of the C. destructivum clade; Menat et al. 2012). This is by no means a comprehensive list of research papers on this topic – a full assessment would justify a further major review.

HISTORY OF CLASSIFICATION

The generic name Colletotrichum was introduced by Corda (1831) for C. lineola, a species found associated with a member of the Apiaceae in the Czech Republic. Colletotrichum lineola was long considered a synonym of the older taxon C. dematium, but was recently re-established as an independent species (Damm et al. 2009). That work included the acquisition and culture of a recent collection of C. lineola from a similar host and locality, and designation of an epitype for the name.

The genus Vermicularia (Tode 1790) could be regarded as an earlier name for Colletotrichum according to some interpretations of the Code of Nomenclature for Algae, Fungi and Plants. The nomenclatural details have been outlined successively in the light of the then current rules by Duke (1928), Sutton (1992) and Damm et al. (2009), and will not be repeated here. Any move to establish Vermicularia as a replacement name for Colletotrichum would have disastrous consequences for scientific communication, and would certainly trigger a conservation proposal. Vermicularia was adopted quite widely for curved-spored species in the early years of Colletotrichum systematics, even though the type species of Colletotrichum also has curved conidia. The genus Gloeosporium (Montagne 1849) was also frequently confused with Colletotrichum in the late 19th and early 20th centuries. It was used for taxa of Colletotrichum without conidiomatal setae (their development in many species is variable) but also included quite unrelated fungi. The type of Gloeosporium, Gl. castagnei is not congeneric with Colletotrichum and is currently included in Marssonina, technically providing an earlier name for that genus (von Arx 1957a, 1970). A further 10 generic synonyms for Colletotrichum were listed by Sutton (1980); none has been in recent use.

Two further species (both currently of uncertain application) were added to Colletotrichum by Corda in the years following the original publication of the genus name (Corda 1837, 1840), but the group only came to prominence in the late 19th century with publication of Saccardo’s Sylloge Fungorum compilations. Fifty new taxa at species level or below were described between 1880 and 1900, and this trend of new species recognition accelerated well into the 20th century. At the time of the first formal monographic treatment of Colletotrichum, by von Arx (1957b), around 750 names were in existence. This explosion of what might now be regarded as largely futile taxonomic activity seems to have been driven largely by uncritical assumptions that Colletotrichum species are strongly host-specific. The result was that in many instances a new taxon was erected each time an infection caused by a Colletotrichum species was discovered on a plant genus for which no disease had previously been reported, even in the absence of unique morphological diagnostic characters.

The impact of von Arx’s monograph (von Arx 1957b) was considerable, and it set the stage for a new era in Colletotrichum taxonomy. His approach was based on morphological characteristics with little or no emphasis on placed on pathological features, which led to a reduction in accepted species from around 750 to 11 (within a total of 23 accepted specific and infraspecific taxa). Many taxa were evaluated based on descriptions from the literature rather than evaluation of type specimens. Such a drastic reduction in numbers of taxa provided a new foundation on which to develop subsequent systematic treatments, but it is clear that even von Arx himself regarded the 11 accepted species as broadly circumscribed aggregates rather than individual taxa. In particular, the account of C. gloeosporioides (itself with around 600 synonyms) incorporated a series of nine “abweichende Formen” [variant forms], including five taxa combined into Colletotrichum by von Arx in this work or the companion volume on Gloeosporium (von Arx 1957a). These variant forms were considered to be host-specific variants that could not reliably be distinguished on a morphological basis from the main bulk of C. gloeosporioides. Included were species now treated within the C. orbiculare, C. acutatum and C. gloeosporioides aggregates, as well as other taxa that are currently of uncertain affiliation. Von Arx’s approach to Colletotrichum classification now appears crude even in purely morphological terms, and as Sutton (1992) and Cannon et al. (2000) both noted, more attention to matters of typification would have been valuable. Nonetheless, this seminal work of von Arx laid the foundation for all subsequent morphological taxonomic work on the genus Colletotrichum.

Subsequent taxonomic treatments primarily focused on species groups, or taxa associated with particular crop plants. Important contributions were made in the 1960s by Simmonds (1965; recognition of Colletotrichum acutatum), and by Sutton (1966, 1968; taxonomy of the C. graminicola complex and the value of appressorial morphology in classification). The next comprehensive treatment of Colletotrichum was by Sutton (1980), who accepted 22 species, and a study of 11 South African species was contributed by Baxter et al. (1983). Both of these accounts focused primarily on morphological and cultural characteristics, and most of the taxa were considered to be plurivorous. Similar approaches were adopted by Smith & Black (1990) for species on strawberry, and Walker et al. (1991) for those associated with Xanthium, but with increased emphasis on integration of taxonomic and pathological data.

The first International Workshop on Colletotrichum was held in late 1990 at the University of Bath, UK (Bailey & Jeger 1992), bringing together experts on taxonomy, molecular biology, host/parasite interactions and pathology. This marked the advent of the wide-scale application of molecular methods in Colletotrichum studies, which has revolutionised research in that genus as with many other fungal groups. Initially, work focused on infraspecific variation; DNA polymorphisms were detected in C. gloeosporioides by Dale et al. (1988), Braithwaite & Manners (1989) and Braithwaite et al. (1990a, b), and strains of that species (as then circumscribed) were found to have variable numbers of chromosomes (Masel et al. 1990).

The first applications of DNA sequence data to distinguish between Colletotrichum species were published by Mills et al. (1992) and Sreenivasaprasad et al. (1992), who identified sequence variation in the ITS1 region of nrDNA between six species of Colletotrichum, as well as detecting polymorphisms in the same region between strains of C. gloeosporioides from different hosts. More comprehensive studies followed rapidly; Sherriff et al. (1994) presented the first bootstrapped NJ trees for Colletotrichum, using ITS2 and LSU sequences of 27 strains indicated as belonging to 13 species. This study recognised the C. orbiculare aggregate as a distinct taxonomic unit, and detected genetic congruence between the four curved-spored species studied. In a portent of things to come, Sherriff et al. showed that not all of the strains examined were correctly identified using morphological characteristics, with one strain each of C. gloeosporioides and C. lindemuthianum clustering separately from the others. A second phylogenetic study of the genus was published by Sreenivasaprasad et al. (1996) using parsimony analysis of ITS 1 and 2 sequences from 18 species of Colletotrichum, and the authors were able to identify six infrageneric groups. Sreenivasaprasad et al. also used infra- and interspecific nucleotide identity in the ITS region as indicators of the taxonomic rank at which strains should be differentiated, as an early forerunner of the DNA barcoding initiatives.

The number of papers using molecular methods to elucidate relationships in Colletotrichum increased rapidly after the early 1990s. Most of these studies focused on small groups within the genus, usually associated with a particular crop (see Table 1). More wide-ranging studies were presented by Johnston & Jones (1997), who used LSU rDNA sequences to analyse strains from diseased fruit crops in New Zealand, and Moriwaki et al. (2002) who studied ITS-2/LSU rDNA of Colletotrichum species from Japan. The first multilocus phylogenetic analyses of Colletotrichum species were published by Talhinhas et al. (2002), a study of the C. acutatum aggregate associated with lupins using ITS, TUB2 and HIS4 sequences, and Vinnere et al. (2002) using ITS, TUB2 and mtSSU in a study on the same species cluster associated with Rhododendron in Sweden and Latvia. Talhinhas et al. (2002) found that the three loci they studied displayed broadly similar levels of phylogenetic resolution. Guerber et al. (2003) used glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and glutamine synthetase (GS) nucleotide sequences in a further study of the C. acutatum group, and the HMG-box section of the mating-type genes MAT-1 was found to be a valuable evolutionary marker by Du et al. (2005). From around this time, multilocus analyses became the norm as sequencing costs reduced, with sequence data generated from loci such as actin (ACT), calmodulin (CAL), chitin synthase I (CHS-1), DNA lyase (APN2), manganese superoxide dismutase (SOD2), the large subunit of RNA polymerase II (RPB1) and the translation elongation factor 1-α (EF1α) (see Table 1 for references).

Table 1.

Summary of principal phylogenetic research papers on Colletotrichum species based on DNA sequence data.

Publication	Clade	Host taxa	Geographical limits	Loci used
Mills et al. (1992)	genus-wide	Tropical fruits		ITS
Sreenivasaprasad et al. (1992)	acutatum, gloeosporioides	Strawberry		ITS
Sreenivasaprasad et al. (1993)	gloeosporioides	Coffee		ITS
Sherriff et al. (1994)	genus-wide			ITS-2, LSU
Sherriff et al. (1995)	graminicola	Poaceae		LSU
Bailey et al. (1996)	orbiculare	Malvaceae		ITS, LSU
Sreenivasaprasad et al. (1996)	genus-wide			ITS
Johnston & Jones (1997)	genus-wide	Fruit crops	New Zealand	LSU
Munaut et al. (1998)	gloeosporioides	Stylosanthes	Africa, Australia	ITS
Balardin et al. (1999)	orbiculare	Phaseolus		ITS
Martin & García-Figueres (1999)	acutatum, gloeosporioides	Olive	Spain	ITS
Freeman et al. (2000)	acutatum, gloeosporioides	Almond, avocado, strawberry	Israel, USA	ITS, LSU
Freeman et al. (2001)	acutatum	Mostly fruit crops		ITS
Hsiang & Goodwin (2001)	graminicola	Poaceae		ITS
Abang et al. (2002)	gloeosporioides	Yam	Nigeria	ITS
Chen et al. (2002)	graminicola	Agrostis	Canada	MAT2
Moriwaki et al. (2002)	genus-wide		Japan	ITS-2, LSU
Munaut et al. (2002)	gloeosporioides	Stylosanthes	Mexico	ITS
Nirenberg et al. (2002)	acutatum	Lupin		ITS
Talhinhas et al. (2002)	acutatum	Lupin		ITS, TUB2, HIS4
Vinnere et al. (2002)	acutatum	Rhododendron	Sweden, Latvia	ITS, TUB2, mtSSU
Afanador-Kafuri et al. (2003)	acutatum, gloeosporioides	Mango, passion-fruit, tamarillo	Colombia	ITS
Denoyes-Rothan et al. (2003)	acutatum, gloeosporioides	Strawberry		ITS
Guerber et al. (2003)	acutatum		USA, New Zealand	GAPDH, GS
Martínez-Culebras et al. (2003)	acutatum, gloeosporioides	Strawberry		ITS
Moriwaki et al. (2003)	boninense		Japan	ITS
Sanders & Korsten (2003)	gloeosporioides	Avocado, mango	South Africa	ITS
Ford et al. (2004)	destructivum	Legumes		ITS
Lu et al. (2004)	boninense, gloeosporioides	Endophytes of tropical trees	Guyana	ITS
Lubbe et al. (2004)	Genus-wide	Proteaceae	primarily Africa	ITS, TUB2
O’Connell et al. (2004)	destructivum			ITS
Du et al. (2005)	acutatum, graminicola, gloeosporioides			ITS, MAT1-2 (HMG marker)
Lee et al. (2005)	boninense	Euonymus japonicus	Korea	ITS
Lotter & Berger (2005)	acutatum	Lupin	South Africa	ITS, TUB1, TUB2
Photita et al. (2005)	genus-wide		Thailand	ITS
Talhinhas et al. (2005)	acutatum, gloeosporioides	Olive	Portugal	ITS, TUB2
Chung et al. (2006)	acutatum, gloeosporioides	Fruit crops	Japan	ITS
Crouch et al. (2006)	graminicola	Poaceae	USA	ITS, MAT1-2 (HMG marker), SOD2
Farr et al. (2006)	genus-wide	Agavaceae		ITS, LSU
González et al. (2006)	acutatum, gloeosporioides	Apple	USA, Brazil	GAPDH
Ramos et al. (2006)	acutatum, gloeosporioides	Citrus	Portugal	ITS, TUB2
Latunde-Dada & Lucas (2007)	destructivum, truncatum, graminicola			ITS, LSU
Lee at al. (2007)	acutatum, gloeosporioides	Apple	Korea	ITS, TUB2
Liu et al. (2007a)	orbiculare			GAPDH, GS
Liu et al. (2007b)	dracaenophilum	Buxus	China	ITS
Shenoy et al. (2007)	truncatum	Solanaceae		ITS, TUB2
Whitelaw-Weckert et al. (2007)	acutatum	Grape	Australia	ITS, TUB2
Cannon et al. (2008)	gloeosporioides			ITS
Crouch et al. (2008)	graminicola	Poaceae		Ccret2
LoBuglio & Pfister (2008)	acutatum	Acer platanoides	USA	ITS, LSU
Marcelino et al. (2008)	acutatum	Insects	USA	ITS, LSU, TUB2, GAPDH, GS, MAT1-2
Peres et al. (2008)	acutatum	Citrus	N and S America	ITS, GAPDH
Than et al. (2008a)	acutatum, truncatum, gloeosporioides			ITS, TUB2
Than et al. (2008b)	acutatum			ITS, TUB2
Crouch et al. (2009c)	graminicola	Poaceae		ITS, APN2/IGS/MAT1-2, SOD2
Crouch et al. (2009d)	graminicola	Poaceae		ITS, APN2/IGS/MAT1-2, SOD2
Damm et al. (2009)	dematium, spaethianum, truncatum			ITS, ACT, GAPDH, CHS-1, TUB2, HIS3
Garrido et al. (2009)	acutatum	Strawberry	Spain	ITS
MacKenzie et al. (2009)	acutatum		USA, Costa Rica	ITS, GAPDH, GS
McKay et al. (2009)	acutatum, boninense, gloeosporioides	Almond	Australia	ITS
Moriwaki & Tsukiboshi (2009)	graminicola	Echinochloa	Japan	ITS, MAT1-2 (HMG marker), SOD2
Pileggi et al. (2009)	boninense, gloeosporioides	Maytenus ilicifolia	Brazil	ITS
Polashock et al. (2009)	acutatum, gloeosporioides	Cranberry	N America	ITS, LSU
Prihastuti et al. (2009)	gloeosporioides	Coffee	Thailand	ITS, ACT, TUB2, CAL, GS, GAPDH
Shivas & Tan (2009)	acutatum			ITS, TUB2
Sun & Zhang (2009)	destructivum			ITS
Talhinhas et al. (2009)	acutatum, gloeosporioides	Olive	Portugal	ITS, TUB2
Yang et al. (2009)	genus-wide	Amaryllidaceae	China, Thailand	ITS, ACT, TUB2, CAL, CHS- 1, GAPDH
Giaretta et al. (2010)	acutatum, gloeosporioides	Apple	Brazil	ITS
Hemelrijk et al. (2010)	acutatum	Strawberry	Belgium	ITS
Lopez & Lucas (2010)	gloeosporioides	Cashew	Brazil	LSU
Manuel et al. (2010)	gloeosporioides	Coffee	Angola	ITS
Nguyen et al. (2010)	genus-wide	Coffee	Vietnam	ITS, mtSSU
Phoulivong et al. (2010)	gloeosporioides	Tropical fruits	Laos, Thailand	ITS, TUB1, TUB2, ACT, GAPDH
Phuong et al. (2010)	genus-wide	Coffee	Vietnam	ITS, mtSSU
Prihastuti et al. (2010)	graminicola	Poaceae		ITS, APN2/IGS/MAT1
Rojas et al. (2010)	gloeosporioides	Cacao	S America, China	ITS, EF1α, TUB2, RPB1, APN2, MAT1-2
Weir & Johnston (2010)	gloeosporioides	Persimmon		ITS, GAPDH, EF1α
Wikee et al. 2010	gloeosporioides, truncatum	Jasmine	Vietnam	ITS, ACT, TUB2, CAL, GS, GAPDH
Xie et al. (2010)	acutatum, gloeosporioides	Strawberry	China	ITS
Choi et al. (2011)	destructivum		Korea	ITS, ACT, EF1α, GS
Faedda et al. (2011)	acutatum	Olive	Italy	ITS, TUB2
Gazis et al. (2011)	gloeosporioides	Hevea species	Peru	ITS, TEF, GPD
Liu et al. (2011)	coccodes	Potato		ITS, ACT, GAPDH, TUB2
Rampersad (2011)	gloeosporioides, truncatum	Papaya	Trinidad	ITS, TUB2
Silva-Rojas & Ávila-Quezada (2011)	acutatum, boninense, gloeosporioides	Avocado	Mexico	ITS, LSU
Yang et al. (2011)	genus-wide	Orchidaceae	China	ITS, ACT, TUB2, CAL, CHS- 1, GAPDH
Crouch & Tomaso-Peterson (2012)	graminicola	Centipedegrass, sorghum		ITS, APN2/IGS/MAT1-2, SOD2
Damm et al. (2012a)	acutatum			ITS, ACT, GAPDH, CHS-1, TUB2, HIS3
Damm et al. (2012b)	boninense			ITS, ACT, GAPDH, CHS-1, TUB2, HIS3, CAL
Silva et al. (2012a,b)	gloeosporioides	Coffee		ITS, ApMAT, Apn15L, MAT1-2, MAT5L, Apn1Ex3, Apn13L, TUB2, GS
Weir et al. (2012)	gloeosporioides			ITS, ACT, GAPDH, CHS-1, TUB2, CAL, GS, SOD2
Yang et al. (2012)	genus-wide	Hemerocallis	China	ITS, ACT, GAPDH, CHS-1, TUB2

A further milestone in Colletotrichum systematics was reached with publication of a special issue of the journal Fungal Diversity in late 2009, containing a group of papers presenting taxonomic revisions and review articles relevant to the genus. This includes an introductory paper focusing on the need for correct identification (Hyde et al. 2009b), a review of the cereal-inhabiting species (Crouch & Beirn 2009), a revision of the species with curved conidia from herbaceous hosts (Damm et al. 2009), a study of the species affecting coffee berries in Thailand (Prihastuti et al. 2009), a partial revision of the C. acutatum group (Shivas & Tan 2009) and research on the species associated with Amaryllidaceae (Yang et al. 2009). The issue concludes with a review of the status of Colletotrichum names in current use (Hyde et al. 2009a) and recommendations for polyphasic methods (Cai et al. 2009).

The list of Colletotrichum names in current use (Hyde et al. 2009a) accepted a total of 66 species, with an additional 20 recently used names considered as doubtful. This assessment represented a substantial increase in the number of recognised species compared with the 23 taxa recognised by von Arx (1957) and the 39 species accepted by Sutton (1992), and reflected the increasing reliance on molecular methods for species definition. With publication of the current volume of Studies in Mycology, a further 41 species are introduced, bringing the current number of accepted Colletotrichum species to over 100. It is likely that further Colletotrichum taxa remain to be recognised in the major clades that have not yet been the subject of comprehensive multilocus studies.

Colletotrichum species from non-cultivated plants in natural and semi-natural habitats are much less commonly studied than those associated with cultivated plant hosts, with most studies being of endophytic strains. A study on leaf endophytes of native forest trees by Lu et al. (2004) examined diversity within the C. gloeosporioides and C. boninense species clusters, and Xiao et al. (2004) and Mackenzie et al. (2007) compared strains of the C. gloeosporioides cluster from strawberry and non-crop species. Crouch et al. (2006, 2009d) distinguished clades within the C. cereale cluster that correlated with pathogenicity, with some causing disease of turfgrasses and others isolated from asymptomatic prairie grasses. Gazis et al. (2011) compared Amazonian populations of endophytic taxa belonging to the C. gloeosporioides cluster associated with two species of Hevea, the cultivated H. brasiliensis and the non-cultivated H. guianensis. Higgins et al. (2011) studied Colletotrichum endophytes from grass and non-grass hosts in tropical forest in Panama, recovering some genetically distinct taxa via direct sequence from surface-sterilised grass tissue that were not detected using cultural methods. They also observed that many taxa were detected from more than one grass host genus, corroborating observations by Lu et al. (2004) and Arnold & Lutzoni (2007) that the commonest tropical endophytes appear to be host generalists. However, the ITS sequences used to define OTUs in all these studies are too conservative to reflect all speciation events (Crouch et al. 2009b, Gazis et al. 2011). Several endophyte taxa isolated from cacao in Panama by Rojas et al. (2010) were thought to comprise part of the background endophytic community in the Panamanian forest ecosystem, but most strains studied came from crop plants and their status as native species needs further investigation.

All of the studies of Colletotrichum associated with non-crop plants detailed above demonstrate considerable diversity of taxa. Despite preliminary evidence that host specificity is less in native tropical forest ecosystems compared with managed environments, the sheer number of habitats (in the form of leaves, fruits etc.) that remain unsampled indicate the likelihood that overall species-level diversity of the genus is still significantly under-represented.

PHYLOGENETIC POSITION

Colletotrichum, as an asexual fungal genus, was included in morphological classifications of the Ascomycota as its sexual genus Glomerella. Successive editions of the Dictionary of the Fungi until edn 6 (Ainsworth, 1971) listed Glomerella as a member of the Phyllachoraceae in the order Sphaeriales. The Phyllachoraceae was originally described by Theissen & Sydow (1915) as part of the Dothideales. Petrak (1924) concluded that Phyllachora, Polystigma and Physalosporina (= Stigmatula; see Cannon 1996) constituted a natural family that did not belong to the Dothideales. Chadefaud (1960) introduced (but did not validly publish) the ordinal name Glomerellales, including Glomerella, Phyllachora and two other genera in a non-ranked group “Eu-Glomérellales”. Barr (1976) introduced (but again did not validly publish) the ordinal name Phyllachorales, in which was included a disparate set of families with the Phyllachoraceae subsumed into the Melogrammataceae. Glomerella was accepted as part of that assemblage. Seven years later, Barr (1983) validated the ordinal name Phyllachorales but did not explicitly alter its composition. The same year, Hawksworth et al. (1983) placed Glomerella in its traditional position in the Phyllachoraceae, but treated the family as the only representative of the Polystigmatales, yet another name that appears not to have been validly published. Edition 8 of the Dictionary of the Fungi (Hawksworth et al. 1995) adopted a similar classification, though the ordinal name Polystigmatales was replaced by Phyllachorales.

Glomerella had long been considered to be an outlier within the Phyllachoraceae due to its non-stromatic nature (Cannon 1991). The family name Glomerellaceae was first published (invalidly) by Locquin (1984), in a general account of the fungi in which no fewer than 278 new families were introduced. Locquin’s work was generally ignored, until preliminary sequence-based studies along with ontogenetic research (Uecker 1994) confirmed that Glomerella and Phyllachora did not belong to the same order of fungi. The Glomerellaceae was adopted in the 9^th edition of the Dictionary of the Fungi with an uncertain position within the Sordariomycetidae (Kirk et al. 2001), and in the 10^th edition as an unplaced taxon within the Hypocreomycetidae (Kirk et al. 2008).

The first attempts to place Glomerella/Colletotrichum within a molecular phylogenetic system were published by Illingworth et al. (1991) and Berbee & Taylor (1992), using 18S rDNA sequences. Although the number of taxa sampled was insufficient to provide reliable placement, the samples of C. gloeosporioides included in these studies were shown to cluster with members of the Hypocreales. Most subsequent phylogenetic studies included Glomerella/Colletotrichum only as outgroups, or to provide an overall framework for the phylogeny of unrelated groups (e.g. Zhang & Blackwell 2002, Castlebury et al. 2004, Huhndorf et al. 2004).

There is very little information available on sequences from the Phyllachoraceae sensu stricto. Winka & Eriksson (2000) found that two 18S rDNA sequences from Phyllachora species clustered in the Sordariomycetidae clade, while Glomerella cingulata was considered to be more closely related to the Hypocreomycetidae. Wanderlei-Silva et al. (2003) also published a study based on 18S rDNA, that claimed that the Phyllachoraceae was polyphyletic. In this work, core taxa clustered with the Sordariales, Ophiodothella vaccinii clustered within the Xylariales, and Glomerella/Colletotrichum was shown as a sister group to the Hypocreales.

Zhang et al. (2006) confirmed the phylogenetic position of Glomerella within the Hypocreomycetidae, and provided a Latin diagnosis for the Glomerellaceae. A sister taxon relationship with Verticillium was recovered (Zhang et al. 2006), but this clustering appears to be an artefact of limited taxa sampling. Subsequent investigations assigned Verticillium to the Plectosphaerellaceae (Zare et al. 2007, Cannon et al. 2012), following the conclusions of Zare et al. (2000). The phylogenetic position of the Glomerellaceae was further elucidated by Réblová et al. (2011) in a study using ITS, LSU, SSU and rpb2 genes. In this work, the Glomerellaceae occupied a common clade with two newly recognised families, the Australiascaceae and Reticulascaceae. They accordingly validated the order Glomerellales (first introduced by Chadefaud 1960 but without a Latin diagnosis) for the three families. Based on SSU data, Réblová et al. (2011) showed that the Glomerellales occupied a well-supported clade that included the Hypocreales, Microascales and the Plectosphaerellaceae, equivalent to the Hypocreomycetidae as delimited by Zhang et al. (2006). Similar results were obtained with LSU sequence data, although the separation of the Hypocreomycetidae was not supported by bootstrap analysis or posterior probability measures (Réblová et al. 2011). This is probably not the final word in elucidation of the phylogenetic position of Colletotrichum, but the Glomerellales clade is well supported despite significant morphological differences between the three families included.

SEXUAL MORPHS AND SEXUAL-ASEXUAL CONNECTIONS

In common with many other fungal pathogens, the Colletotrichum asexual morph is most commonly associated with disease symptoms, with the sexual morph tending to develop on moribund or dead host tissues (Sutton 1992). Colletotrichum sexual morphs are therefore under-studied in comparison with the asexual stages. This lack of attention to the sexual morphs is compounded by the need to identify species from cultures, the preparation of which may keep compatible strains separate. This makes it difficult to assess the prominence of the Glomerella stages in nature compared with their asexual morphs.

Colletotrichum sexual morphs were first described by Stoneman (1898) in the genus Gnomoniopsis Stoneman, in a comprehensive and well-illustrated account of the development of anthracnose diseases in the USA. Four species were described in full, all of which were linked to previously described asexual morphs; Gn. cingulata (anamorph Gloeosporium cingulatum, from Ligustrum vulgare), Gn. piperata (asexual Gl. piperatum, from Capsicum annuum), Gn. cincta (asexual Colletotrichum cinctum, from the orchids Maxillaria picta and Oncidium sp.) and Gn. rubicola (asexual C. rubicola, from Rubus strigosus). A fifth species, given the name Gnomoniopsis? vanillae (asexual Colletotrichum sp., from Vanilla) was also described in a preliminary manner. All of the species accepted were linked to their asexual morphs by cultural methods in the laboratory.

Von Schrenk & Spaulding (1903) pointed out that Stoneman’s genus was a later homonym of Gnomoniopsis Berl. (Berlese 1893; type Gn. chamaemori), which is not closely related to the anthracnose pathogens. Gnomoniopsis Berl. has recently been confirmed as a genus of the Gnomoniaceae (Diaporthales) rather than the Glomerellaceae (Sogonov et al. 2008). Von Schrenk and Spaulding (1903) accordingly proposed the name Glomerella for the anthracnose-causing species, making new combinations for the four species definitely accepted by Stoneman in her genus and adding a fifth, Glomerella rufomaculans, considered to be the causal agent of bitter rot of apple (see also Du et al. 2005). The type of Gnomoniopsis Stonem. was not originally specified, and nor was that of Glomerella. The earliest lectotypification of Glomerella appears to be by Clements & Shear (1931), who designated Ga. cingulata as type. This choice has been accepted by subsequent authors, most notably by von Arx & Müller (1954) and von Arx (1987).

A comprehensive monograph for Glomerella has never been published. The broadest treatment to date is by von Arx & Müller (1954), at a similar level of detail to the revision of Colletotrichum three years later by von Arx (1957b). Von Arx & Müller recognised only five species, two of which are poorly known and cannot be confirmed as belonging to Glomerella.

Those excluded by us from von Arx & Müller’s concept of Glomerella include Ga. guevinae (syn. Chiloëlla guevinae), which has ascospores that are covered in a gelatinous sheath and are much smaller than those of typical Glomerella species. No asexual morph has been seen. Sydow (1928) suggested that Chiloëlla has affinities with Physalospora (Hyponectriaceae) or Plagiostoma (Gnomoniaceae). Type material has not been traced, and so Chiloëlla remains of uncertain affinity. Ga. montana (syn. Physalospora montana, Phyllachora montana) was considered by Parbery (1964) to have affinities with a small group of Phyllachora species on montane grasses with sexual morphs that mature on dead plant tissues. Authentic material of the species in K conforms with this interpretation. Von Arx & Müller (1954) did find the type material to be in association with old Colletotrichum fruit-bodies, but there is no demonstrated connection between the morphs.

The three species treated by von Arx & Müller (1954) that definitely belong to Glomerella are the type Ga. cingulata, Ga. tucumanensis and Ga. amenti. Glomerella tucumanensis is widely accepted as the sexual morph of Colletotrichum falcatum, the cause of red rot of sugarcane. Work by Sutton (1968) and Crouch et al. (2009c) confirm this species as a distinct and apparently host-specific pathogen using both morphological and molecular criteria. Glomerella amenti (syn. Phyllachora amenti, Haplothecium amenti) was described from flower stalks and bracts of the arctic-alpine species Salix reticulata, an unexpected habitat for a species of Glomerella, but its phylogenetic position has been reassessed (Damm et al. 2012a), and confirmed as a synonym of C. salicis, a member of the C. acutatum clade.

Glomerella cingulata is now widely recognised as a species aggregate and the sexual counterpart to the C. gloeosporioides aggregate, although the connection has not been explicitly proved, and the link at species level may well be incorrect. As far as we are aware, type material of Ga. cingulata has not been examined in modern times (though a possible authentic specimen is preserved in BPI). Similarly, the identity of Gloeosporium cingulatum Atk., with which Ga. cingulata was linked by Stoneman (1898), has not been critically reassessed, and the conidia of Gloeo. cingulatum as illustrated by Stoneman could also belong to the C. acutatum clade.

Shear & Wood (1907) and Edgerton (1908) considered that at least several of the putatively host-specific taxa described by Stoneman (1898) as species of Gnomoniopsis were conspecific, although they did not include material ascribed to Ga. cingulata in their studies. The equation of the name Ga. cingulata with the species aggregate rather than the fungus causing disease of Ligustrum was further established in works by Dastur (1920) and Small (1921, 1926), which focused on cross-inoculation experiments.

Since the name Glomerella cingulata was originally published, unnecessary or poorly justified taxa proliferated for the same reason as did those for Colletotrichum gloeosporioides, i.e. assumed host specificity. Von Arx & Müller provided a long list of 117 synonyms belonging to at least 42 independent taxa (they did not distinguish between homotypic synonyms and taxa in different genera with the same epithet). As with previous work on C. gloeosporioides, the contribution of Von Arx & Müller provided a valuable foundation for later investigations. Subsequent research has identified further distinct Glomerella taxa, and currently around 30 species of Colletotrichum are known to have (or have at least been claimed to have) Glomerella sexual morphs. They are listed in Table 2.

Table 2.

Colletotrichum species with reported Glomerella sexual morphs.

Colletotrichum species	Glomerella species	Reference	Method	Teleomorph placement (von Arx & Müller 1954)	Currnet clade	Notes
C. “acutatum”	Ga. acutata	Guerber & Correll (2001), Damm et al. (2012a)	Laboratory crossing	NA	acutatum	Teleomorph type a hybrid between C. acutatum and C. fioriniae
C. annellatum	Unnamed	Damm et al. (2012b)	Developed on SNA medium and sterile plant stem in culture	NA	boninense
C. boninense	Unnamed	Damm et al. (2012b)	Developed on SNA and OA medium	NA	boninense
C. brassicicola	Unnamed	Damm et al. (2012b)	Developed on sterile plant stem in culture	NA	boninense
C. cinctum	Ga. cincta	Stoneman (1898)	Laboratory culture	Ga. cingulata		Connection doubtful (see Damm et al. 2012b), modern revision needed
Gloeosporium cingulatum	Ga. cingulata	Stoneman (1898)	Laboratory culture of sterilised bean stem, single-ascospore cultures	Ga. cingulata	gloeosporioides ?	Identity and placement uncertain, modern revision needed
C. cliviae	Unnamed	Yang et al. (2011)	Developed on PDA medium	NA		Not closely related to any established clade
C. constrictum	Unnamed	Damm et al. (2012b)	Developed on SNA medium and sterile plant stem in culture	NA	boninense
C. cymbidiicola	Unnamed	Damm et al. (2012b)	Developed on SNA medium and sterile plant stem in culture	NA	boninense
C. destructivum	Ga. glycines	Manandhar et al. (1986)	Laboratory culture	Ga. cingulata	destructivum	Identification of both morphs doubtful, modern revision needed
C. falcatum	Ga. tucumanensis	Carvajal & Edgerton (1944), Politis (1975)	Laboratory culture	Ga. tucumanensis	graminicola
C. fioriniae	Ga. fioriniae	Marcelino et al. (2008), Shivas & Tan (2009)	Laboratory mating study	NA	acutatum
C. fructicola	Unnamed	Prihastuti et al. (2009)	Laboratory culture	NA	gloeosporioides
C. gloeosporioides	Ga. cingulata	e.g. Cisar et al. (1994), Cisar & TeBeest (1999)	Co-occurrence on host, laboratory mating study	Ga. cingulata	gloeosporioides	Connection unlikely to be correct, placement uncertain
C. glycines	Ga. glycines	Lehman & Wolf (1926)	Culture of both morphs	Ga. cingulata	truncatum	Treated as an independent species by von Arx (1987). Connection doubtful, modern revision needed
C. gossypii	Ga. gossypii	Edgerton (1909)	Laboratory culture	Ga. cingulata	gloeosporioides	Modern revision needed
C. graminicola	Ga. graminicola	Politis (1975), Vaillancourt & Hanau (1991, 1992)	Laboratory mating study	NA	graminicola
C. “heveae”	Ga. phyllanthi	Pai et al. (1970)	Developed on PDA medium	NA	boninense	Connection based on wrong identification of the anamorph, see C. phyllanthi
C. ignotum	Unnamed	Rojas et al. (2010)	Laboratory culture	NA	gloeosporioides
C. karstii	Unnamed	Yang et al. (2011), Damm et al. (2012b)	Developed on SNA and PDA medium	NA	boninense
C. lagenarium	Ga. lagenaria	Stevens (1931)	CMA culture with UV irradiation	Ga. cingulata	orbiculare ?	Modern revision needed
C. lindemuthianum	Ga. lindemuthiana	Shear & Wood (1913), Rodríguez-Guerra et al. (2005)	Laboratory culture or laboratory crossing	Ga. cingulata	orbiculare	Modern revision needed
Gloeosporium lycopersici	Ga. lycopersici	Krüger (1913)	Laboratory culture, inoculated tomato fruits	Ga. cingulata	acutatum	Synonym of C. salicis
C. mume	Ga. mume	Hemmi (1920)	Laboratory culture	Ga. cingulata		Modern revision needed
C. musae	Ga. musarum	Petch (1917)	present on same piece of host tissue	Ga. cingulata	gloeosporioides	Connection needs further research: see Weir et al. (2012)
C. orchidearum	Unnamed	Yang et al. (2011)	Developed on PDA medium	Ga. cingulata	Not closely related to any established clade	Identity of this fungus is not completely clarified
C. parsonsii	Unnamed	Damm et al. (2012b)	Developed on SNA medium	NA	boninense
C. petchii	Unnamed	Damm et al. (2012b)	Developed on sterile plant stem in culture	NA	boninense
C. phomoides	Ga. phomoides	Swank (1953)	both morphs developing from singleconidium isolate	Ga. cingulata	dematium ?	Modern revision needed
C. phormii	Ga. phormii	Hennings (1898), Farr et al. (2006), Damm et al. (2012a)	Developed on leaves	Ga. phacidiomorpha and Ga. cingulata	acutatum	Also see Kinghorn (1936) and von Arx (1987), misapplied as Ga. phacidiomorpha
C. phyllanthi	Ga. phyllanthi	Pai et al. (1970), Damm et al. (2012b)	based on type specimen (dried culture) and description (living culture sterile)	NA	boninense	Anamorph and teleomorph based on same type
C. piperatum	Ga. piperata	Stoneman (1898)	Laboratory culture	Ga. cingulata	gloeosporioides ?	Modern revision needed
C. rhodocyclum	Ga. phacidiomorpha	Kinghorn (1936)	Developed on the surface of living leaves, not in culture	Ga. cingulata	acutatum	Synonym of C. phormii, name Ga. phacidiomorpha misapplied (Farr et al. 2006)
C. rhombiforme	Unnamed	Damm et al. (2012a)	Developed on sterile plant stem in culture	NA	acutatum
C. rubicola	Ga. rubicola	Stoneman (1898)	Single-conidium isolations produced both morphs	Ga. cingulata	acutatum ?	Modern revision needed
C. salicis	Ga. salicis	Damm et al. (2012a)	Developed on sterile plant stem in culture	Ga. amenti, Ga. cingulata	acutata	Ga. amenti forms no anamorph according to Arx and Müller (1954)
C. sublineolum	Unnamed	Vaillancourt & Hanau (1992)	Laboratory mating study	NA	graminicola
C. taiwanense	Ga. septospora	Sivanesan & Hsieh (1993)	Single-ascospore isolations produced both morphs	NA		Perhaps does not belong to Colletotrichum, modern revision needed
Unnamed	Ga. magna	Jenkins & Winstead (1964)	Laboratory crossing	NA	Not closely related to any established clade	The anamorph has been referred to as “C. magna” (e.g. Redman et al. 1999) but the name does not appear to have been formally published. Modern revision needed
Unnamed	Ga. miyabeana	Fukushi (1921), Johnston & Jones (1997)	Found on stems and leaves of Salix purpurea var. angustifolia and on sterilised pieces of willow stem in culture	Ga. cingulata	acutatum	Synonym of C. salicis, treated as Ga. miyabeana by von Arx (1957b, 1987)
Unnamed	Ga. truncata	Armstrong-Cho & Banniza (2006)	Pairing of anamorph isolates	NA	destructivum	Anamorph misidentified as C. truncatum (Latunde-Dada & Lucas 2007, Damm et al. 2009)

There has been little morphology-based comparison of the sexual taxa, and differential characters cited by researchers seem restricted to ascospore shape and size, with individual taxa showing wide variation and exhibiting overlapping ranges. For example, Lehman & Wolf (1926) described the ascospores of Glomerella glycines as ranging between 13 and 43 μm (though chiefly 19–28 μm) in length. Elsewhere, von Arx & Müller (1954) gave measurements for Ga. cingulata of 9–30 × 3–8 μm (mostly 12–24 × 4–7 μm). Comparative study has certainly been compromised by the excessively wide species concept for Ga. cingulata. However, the ascospores of Ga. tucumanensis were described as larger than the norm for Ga. cingulata by von Arx & Müller (1954). Guerber & Correll (2001) established that ascospores of Ga. acutata were smaller and somewhat less strongly pointed than those of Ga. cingulata, but qualified their conclusions as the strains studied of the latter species were too few to establish clear boundaries between the two taxa based on these criteria. Future study may identify further diagnostic morphology-based characters for the sexual morph of Colletotrichum, particularly when viewed in light of modern phylogenetic species concepts.

Assessment of historical asexual-sexual connections in Colletotrichum is very problematic. Many of the claimed links are not based on authentic material, thereby casting doubt on the identities of both morphs. Some are based on little more than juxtaposition on diseased plant samples. Even when the connections are well-researched and use correctly identified material (for the time), the identity of the holomorph may not be easy to establish using modern phylogenetic methods. Some of the information in Table 2 must therefore be considered as more of historic than scientific value.

The substantial changes in Colletotrichum species delimitation made possible by molecular systematic analysis mean that many asexual-sexual connections need further study, and in most cases the sexual names are not typified according to modern practice. From a nomenclatural perspective, the need for this work is now less critical as the requirement for separate naming of asexual and sexual taxa has been abolished (Hawksworth 2011). Nevertheless, the need to understand sexual recombination and production in terms of biological strategy (and potentially also economic significance) at species and population level remains clear.

Although currently available data are incomplete, it does appear that some Colletotrichum clades have species that form sexual morphs more readily than others. Those where sexual morphs are generated frequently, measured in terms of the proportion of consituent species with known meiotic morphs, include the gloeosporioides and boninense clades. To our knowledge, in contrast, there are no reliable reports of a sexual morph from any taxon within the truncatum clade. In other groups, such as the graminicola clade, individual species are well known to produce sexual morphs (e.g. C. falcatum, C. graminicola), but others seem to form them rarely or not at all (Crouch and Beirn 2009). Mating seems to be rare in the orbiculare clade, with only a small proportion of crosses between C. lindemuthianum strains producing fertile progeny (Rodríguez-Guerra et al. 2005).

The mechanisms of recombination and sexual production in Colletotrichum are still inadequately understood. Classical genetic research on mating systems in strains identified as Glomerella cingulata (e.g. Olive 1951, Wheeler 1954) indicated that both homothallic and heterothallic isolates exist, although their modern taxonomic placement within the gloeosporioides clade is not known. Despite documented heterothallic behaviour, only one mating type idiomorph has been recovered from population-level screening in a number of studies (e.g. Chen et al. 2002, Du et al. 2005, Crouch et al. 2008).

In a number of species, sexual production has only been documented in laboratory crosses (see Table 2), and the role of mating in natural populations is unclear. Fertile sexual morphs were produced resulting from what is now considered to be interspecific hybridisation of strains within the C. acutatum clade (Guerber & Correll 2001, Damm et al. 2012a), and this phenomenon may be widespread. Hybridisation between taxa within infrageneric clades of fungi has been demonstrated before, e.g. by O’Donnell et al. (2000) in the Fusarium graminearum complex, by Stukenbrock et al. (2012) in Zymoseptoria and by Turner et al. (2010, 2011) in Neurospora. In the Neurospora example, fertile progeny were produced from geographically isolated strains but not from sympatric isolates, suggesting that reproductive barriers evolve at a local level and can be overcome following long-distance dispersal of conidia. Not all of the strains used to produce sexual morphs in the acutatum clade (Guerber & Correll 2001) have been analysed using multilocus sequence technology, so we cannot say whether similar mechanisms are operating in Colletotrichum.

Mating-type gene sequences have been shown to be good markers for phylogenetic analysis. To date, they have been studied in the acutatum, graminicola, gloeosporioides and orbiculare clades (e.g. Du et al. 2005, García-Serrano et al. 2008, Marcelino et al. 2008, Crouch et al. 2009, Moriwaki & Tsukiboshi 2009, Rojas et al. 2010).

TYPIFICATION

Communication of information relating to Colletotrichum species has been seriously compromised in the past by misidentification, misapplication of names and grossly differing species concepts. Many of these problems were caused by uncritical use of species names on the assumptions that (a) all species are host-specific and (b) that only one species of Colletotrichum (or at least only one species with similar gross morphology) parasitises each host genus. Many older Colletotrichum names lack type specimens that are suitable for molecular analysis, and do not have authentic living strains preserved in culture collections. Because the nomenclatural Code (now entitled the International Code of Nomenclature for Algae, Fungi and Plants; Hawksworth 2011) now allows for the designation of epitypes, modern sequenceable collections can be used as substitutes for the original material. An epitype should have morphological, cultural and pathological characteristics similar to those described in the original publication, originate from the same geographical region and host, and preserve (where at all possible) application of the name in concord with modern usage (Cannon et al. 2008). Many currently used names of Colletotrichum now have epitypes designated (e.g. Cannon et al. 2008, Than et al. 2008, Damm et al. 2009, 2012a, b, Su et al. 2011, Weir et al. 2012).

Table 3 summarises the nucleotide sequences associated with type or other representative strains of Colletotrichum species, which we recommend as reference data to aid researchers and plant health practitioners in species identification. Some widely used species names included in Table 3 are of uncertain taxonomic application, as they have not been recently revised or their typification is in doubt. In some of these cases, strains and/or sequences are included in Table 3 that represent the species as generally accepted by modern authors (not necessarily taxonomists), and might thus be appropriate material on which to base epitypes or neotypes in order to preserve current application of the names. We cite these also in Table 3, but stress strongly that they do not have formal nomenclatural status and they should not be taken to be endorsed as authentic. These exceptions are indicated by the marker ”none” in the column labelled “status of source material”.

Table 3.

Authentic sequences for accepted Colletotrichum species.

Species	Clade	Source material1	Status of source material	GenBank accession number(s)	Reference
C. acerbum	acutatum	CBS 128530, ICMP 12921	Culture from holotype	ITS: JQ948459; TUB2: JQ950110; ACT: JQ949780; CHS-1: JQ949120; GAPDH: JQ948790; HIS3: JQ949450	Damm et al. (2012a)
C. acutatum	acutatum	IMI 117617	Holotype	ITS: AF411700	Vinnere et al. (2002)
		CBS 112996, ATCC 56816	Culture from epitype	ITS: JQ005776; TUB2: JQ005860; ACT: JQ005839; CHS-1: JQ005797; GAPDH: JQ948677; HIS3: JQ005818	Damm et al. (2012a)
C. aenigma	gloeosporioides	ICMP 18608	Culture from holotype	ITS: JX010244; TUB2: JX010389; ACT: JX009443; CHS-1: JX009774; GAPDH: JX010044; CAL: JX009683; GS: JX010078; SOD2: JX010311	Weir et al. (2012)
C. aeschynomenes	gloeosporioides	ICMP 17673, ATCC 201874	Culture from holotype	ITS: JX010176; TUB2: JX010392; ACT: JX009483; CHS-1: JX009799; GAPDH: JX009930; CAL: JX009721; GS: JX010081; SOD2: JX010314	Weir et al. (2012)
C. agaves		CBS 118190	Morphology congruent with the type	ITS: DQ286221; LSU: DQ286222	Farr et al. (2006)
C. alatae	gloeosporioides	CBS 304.67, ICMP 17919	Culture from holotype	ITS: JX010190; TUB2: JX010383; ACT: JX009471; CHS-1: JX009837; GAPDH: JX009990; CAL: JX009738; GS: JX010065; SOD2: JX010305	Weir et al. (2012)
C. alienum	gloeosporioides	ICMP 12071	Culture from holotype	ITS: JX010251; TUB2: JX010411; ACT: JX009572; CHS-1: JX009882; GAPDH: JX010028; CAL: JX009654; GS: JX010101; SOD2: JX010333	Weir et al. (2012)
C. annellatum	boninense	CBS 129826	Culture from holotype	ITS: JQ005222; TUB2: JQ005656; ACT: JQ005570; CHS-1: JQ005396; GAPDH: JQ005309; HIS3: JQ005483; CAL: JQ005743	Damm et al. (2012b)
C. anthrisci	dematium	CBS 125334	Culture from holotype	ITS: GU227845; TUB2: GU228139; ACT: GU227943; CHS-1: GU228335; GAPDH: GU228237; HIS3: GU228041	Damm et al. (2009)
C. aotearoa	gloeosporioides	ICMP 18537	Culture from holotype	ITS: JX010205; TUB2: JX010420; ACT: JX009564; CHS-1: JX009853; GAPDH: JX010005; CAL: JX009611; GS: JX010113; SOD2: JX010345	Weir et al. (2012)
C. asianum	gloeosporioides	MFU 090233, ICMP 18580, CBS 130418	Culture from holotype	ITS: FJ972612; TUB2: JX010406; ACT: JX009584; CHS-1: JX009867; GAPDH: JX010053; CAL: FJ917506; GS: JX010096; SOD2: JX010328	Prihastuti et al. (2009), Weir et al. (2012)
C. australe	acutatum	CBS 116478, HKUCC 2616	Culture from holotype	ITS: JQ948455; TUB2: JQ950106; ACT: JQ949776; CHS-1: JQ949116; GAPDH: JQ948786; HIS3: JQ949446	Damm et al. (2012a)
C. axonopodi	graminicola?	IMI 279189	Culture from holotype	ITS: EU554086; Mat1/APN2: FJ377907; APN2: EU364993	Crouch et al. (2009c, d)
C. beeveri	boninense	CBS 128527, ICMP 18594	Culture from holotype	ITS: JQ005171; TUB2: JQ005605; ACT: JQ005519; CHS-1: JQ005345; GAPDH: JQ005258; HIS3: JQ005432; CAL: JQ005692	Damm et al. (2012b)
C. boninense	boninense	MAFF 305972, CBS 123755	Culture from holotype	ITS: AB051400, JQ005153; TUB2: JQ005588; ACT: JQ005501; CHS-1: JQ005327; GAPDH: GQ221769, JQ005240; HIS3: JQ005414; CAL: JQ005674	Moriwaki et al. (2003), Damm et al. (2012b)
C. brasiliense	boninense	CBS 128501, ICMP 18607	Culture from holotype	ITS: JQ005235; TUB2: JQ005669; ACT: JQ005583; CHS-1: JQ005409; GAPDH: JQ005322; HIS3: JQ005496; CAL: JQ005756	Damm et al. (2012b)
C. brassicicola	boninense	CBS 101059	Culture from holotype	ITS: JQ005172; TUB2: JQ005606; ACT: JQ005520; CHS-1: JQ005346; GAPDH: JQ005259; HIS3: JQ005433; CAL: JQ005693	Damm et al. (2012b)
C. brisbaniense	acutatum	CBS 292.67	Culture from holotype	ITS: JQ948291; TUB2: JQ949942; ACT: JQ949612; CHS-1: JQ948952; GAPDH: JQ948621; HIS3: JQ949282	Damm et al. (2012a)
C. carthami	acutatum	SAPA100011	Epitype	ITS: AB696998; TUB2: AB696992	Uematsu et al. (2012)
C. cereale[2]	graminicola	CBS 129663, KS20BIG	None	ITS: DQ126177, JQ005774; TUB2: JQ005858; ACT: JQ005837; CHS-1: JQ005795; HIS3: JQ005816; SOD2: DQ133277; MAT1-2: DQ131946	Crouch et al. (2006). O’Connell et al. (2012)
C. chlorophyti		IMI 103806	Culture from holotype	ITS: GU227894; TUB2: GU228188; ACT: GU227992; CHS-1: GU228384; GAPDH: GU228286; HIS3: GU228090	Damm et al. (2009)
C. chrysanthemi[3]	acutatum	SAPA 100010	Authentic specimen	ITS: AB696999; TUB2: AB696993	Uematsu et al. (2012)
		IMI 364540	None	ITS: JQ948273; TUB2: JQ949924; ACT: JQ949594; CHS-1: JQ948934; GAPDH: JQ948603; HIS3: JQ949264	Damm et al. (2012a)
C. circinans	dematium	CBS 221.81	Culture from epitype	ITS: GU227855; TUB2: GU228149; ACT: GU227953; CHS-1: GU228345; GAPDH: GU228247; HIS3: GU228051; LSU: JN940807	Damm et al. (2009), Schoch et al. (2012)
C. clidemiae	gloeosporioides	ICMP 18658	Culture from holotype	ITS: JX010265; TUB2: JX010438; ACT: JX009537; CHS-1: JX009877; GAPDH: JX009989; CAL: JX009645; GS: JX010129; SOD2: JX010356	Weir et al. (2012)
C. cliviae		CBS 125375	Culture from holotype	ITS: GQ485607, JX519223; TUB2: GQ849440, JX519249; ACT: GQ856777, JX519240; CHS-1: GQ856722, JX519232; GAPDH: GQ856756; CAL: GQ849464	Yang et al. (2009), this study
C. coccodes		CBS 369.75	Culture from neotype	ITS: HM171679, JQ005775; TUB2: JQ005859; ACT: HM171667, JQ005838; CHS-1: JQ005796; GAPDH: HM171673; HIS3: JQ005817; CAL: HM171670; GS: HM171676	Liu et al. (2011), O’Connell et al. (2012)
C. colombiense	boninense	CBS 129818	Culture from holotype	ITS: JQ005174; TUB2: JQ005608; ACT: JQ005522; CHS-1: JQ005348; GAPDH: JQ005261; HIS3: JQ005435; CAL: JQ005695	Damm et al. (2012b)
C. constrictum	boninense	CBS 128504, ICMP 12941	Culture from holotype	ITS: JQ005238; TUB2: JQ005672; ACT: JQ005586; CHS-1: JQ005412; GAPDH: JQ005325; HIS3: JQ005499; CAL: JQ005759	Damm et al. (2012b)
C. cordylinicola	gloeosporioides	MFU090551, ICMP 18579	Culture from holotype	ITS: HM470246, JX010226; TUB2: HM470249, JX010440; ACT: HM470234; CHS-1: JX009864; GAPDH: HM470240, JX009975; CAL: HM470237; GS: HM470243, JX010122; SOD2: JX010361	Phoulivong et al. (2010), Weir et al. (2012)
C. cosmi	acutatum	CBS 853.73	Culture from holotype	ITS: JQ948274; TUB2: JQ949925; ACT: JQ949595; CHS-1: JQ948935; GAPDH: JQ948604; HIS3: JQ949265	Damm et al. (2012a)
C. costaricense	acutatum	CBS 330.75	Culture from holotype	ITS: JQ948180; TUB2: JQ949831; ACT: JQ949501; CHS-1: JQ948841; GAPDH: JQ948510; HIS3: JQ949171	Damm et al. (2012a)
C. curcumae	truncatum	IMI 288937	Culture from epitype	ITS: GU227893; TUB2: GU228187; ACT: GU227991; CHS-1: GU228383; GAPDH: GU228285; HIS3: GU228089	Damm et al. (2009)
C. cuscutae	acutatum	IMI 304802	Culture from holotype	ITS: JQ948195; TUB2: JQ949846; ACT: JQ949516; CHS-1: JQ948856; GAPDH: JQ948525; HIS3: JQ949186	Damm et al. (2012a)
C. cymbidiicola	boninense	IMI 347923	Culture from holotype	ITS: JQ005166; TUB2: JQ005600; ACT: JQ005514; CHS-1: JQ005340; GAPDH: JQ005253; HIS3: JQ005427; CAL: JQ005687	Damm et al. (2012b)
C. dacrycarpi	boninense	CBS 130241, ICMP 19107	Culture from holotype	ITS: JQ005236; TUB2: JQ005670; ACT: JQ005584; CHS-1: JQ005410; GAPDH: JQ005323; HIS3: JQ005497; CAL: JQ005757	Damm et al. (2012b)
C. dematium	dematium	CBS 125.25	Culture from epitype	ITS: GU227819; TUB2: GU228113; ACT: GU227917; CHS-1: GU228309; GAPDH: GU228211; HIS3: GU228015; LSU: JN940809	Damm et al. (2009), Schoch et al. (2012)
C. destructivum	destructivum	CBS 149.34	None	ITS: AJ301942; TUB2: JQ005848; ACT: JQ005827; CHS-1: JQ005785; HIS3: JQ005806	O’Connell et al. (2012)
C. dracaenophilum		CBS 118199	Culture from holotype	ITS: DQ286209, JX519222; TUB2: JX519247; ACT: JX519238; CHS-1: JX519230; LSU: DQ286210	Farr et al. (2006), this study
C. echinochloae	graminicola	MAFF 511473	Culture from holotype	ITS: AB439811; SOD2: AB440153; MAT1-2: AB439820	Moriwaki & Tsukiboshi (2009), Crouch et al. (2009c, d)
C. eleusines	graminicola	MAFF 511155	Culture from epitype	ITS: EU554131, JX519218; TUB2: JX519243; ACT: JX519234; CHS-1: JX519226; SOD2: EU554234; APN2: EU365038	Crouch et al. (2009c, d), this study
C. eremochloae	graminicola	CBS 129661	Culture from holotype	ITS: JQ478447, JX519220; TUB2: JX519245; ACT: JX519236; CHS-1: JX519228; SOD2: JQ478449; Mat1/APN2: JQ478462; APN2: JQ478476	Crouch & Tomaso-Peterson (2012), this study
C. falcatum	graminicola	CGMCC 3.14187, CBS 147945	Culture from neotype	ITS: HM171677, JQ005772; TUB2: JQ005856; ACT: JQ005835; CHS-1: JQ005793; HIS3: JQ005814; Mat1/APN2: HM569769; APN2: HM569770	Prihastuti et al. 2010, O’Connell et al. (2012)
C. fioriniae	acutatum	EHS 58, CBS 128517, ARSEF 10222	Culture from holotype	ITS: EF464594, JQ948292; TUB2: EF593325, JQ949943; ACT: JQ949613; CHS-1: JQ948953; GAPDH: EF593344, JQ948622; HIS3: JQ949283; GS: EF593353; MAT1-2: EF593362; LSU: EF464581	Marcelino et al. (2008), Shivas & Tan (2009), Damm et al. (2012a)
C. fructi	dematium	CBS 346.37	Culture from epitype	ITS: GU227844; TUB2: GU228138; ACT: GU227942; CHS-1: GU228334; GAPDH: GU228236; HIS3: GU228040	Damm et al. (2009)
C. fructicola	gloeosporioides	MFU090228, ICMP 18581*, CBS 130416	Culture from holotype	ITS: FJ972603, JX010165; TUB2: FJ907441, JX010405; ACT: FJ907426; CHS-1: JX009866; GAPDH: FJ972578, JX010033; CAL: FJ917508; GS: FJ972593, JX010095; SOD2: JX010327	Prihastuti et al. (2009), Weir et al. (2012)
C. fuscum	destructivum	CBS 130.57	None	ITS: JQ005762; TUB2: JQ005846; ACT: JQ005825; CHS-1: JQ005783; HIS3: JQ005804	O’Connell et al. (2012)
C. gloeosporioides	gloeosporioides	IMI 356878, CBS 112999, ICMP17821	Culture from epitype	ITS: EU371022, JQ005152, JX010152; TUB2: FJ907445, JQ005587, JX010445; ACT: FJ907430, JQ005500, JX009531; CHS-1: JQ005326, JX009818; GAPDH: FJ972582, JQ005239, JX010056; HIS3: JQ005413; CAL: FJ917512, JQ005673, JX009731; GS: FJ972589, JX010085; SOD2: JX010365	Damm et al. (2012b), Weir et al. (2012)
C. godetiae	acutatum	CBS 133.44	Culture from holotype	ITS: JQ948402; TUB2: JQ950053; ACT: JQ949723; CHS-1: JQ949063; GAPDH: JQ948733; HIS3: JQ949393	Damm et al. (2012a)
C. graminicola	graminicola	CBS 130836, M 1.001	Culture from epitype	ITS: DQ003110, JQ005767; TUB2: JQ005851; ACT: JQ005830; CHS-1: JQ005788; HIS3: HQ005809; Mat1/APN2: FJ377994; MAT1-2: EU365081	Du et al. (2005), Crouch et al. (2009d), O’Connell et al. (2012)
C. guajavae	acutatum	IMI 350839	Culture from holotype	ITS: JQ948270; TUB2: JQ949921; ACT: JQ949591; CHS-1: JQ948931; GAPDH: JQ948600; HIS3: JQ949261	Damm et al. (2012a)
C. hanaui	graminicola	MAFF 305404	Culture from holotype	ITS: EU554101, JX519217; TUB2: JX519242; CHS-1: JX519225; SOD2: EU554205; Mat1/APN2: FJ377922; APN2: EU365008	Crouch et al. (2009c, d), this study
C. hemerocallidis	dematium	CDLG5	Culture from holotype	ITS: JQ400005; TUB2: JQ400019; ACT: JQ399991; CHS-1: Q399998; GAPDH: JQ400012	Yang et al. 2012
C. higginsianum	destructivum	IMI 349063	None	ITS: JQ005760; TUB2: JQ005844; ACT: JQ005823; CHS-1: JQ005781; HIS3: JQ005802	O’Connell et al. (2012)
C. hippeastri	boninense	CBS 125376	Culture from holotype	ITS: GQ485599, JQ005231; TUB2: GQ849446, JQ005665; ACT: GQ856788, JQ005579; CHS-1: GQ856725, JQ005405; GAPDH: GQ856764, JQ005318; HIS3: JQ005492; CAL: GQ849469, JQ005752	Yang et al. (2009), Damm et al. (2012b)
C. horii	gloeosporioides	NBRC 7478, ICMP 10492	Culture from neotype	ITS: GQ329690; TUB2: JX010450; ACT: JX009438; CHS-1: JX009752; GAPDH: GQ329681; CAL: JX009604; GS: JX010137; SOD2: JX010370; TEF1: GQ329693	Weir & Johnston (2010), Weir et al. (2012)
C. indonesiense	acutatum	CBS 127551	Culture from holotype	ITS: JQ948288; TUB2: JQ949939; ACT: JQ949609; CHS-1: JQ948949; GAPDH: JQ948618; HIS3: JQ949279	Damm et al. (2012a)
C. jacksonii	graminicola	MAFF 305460	Culture from holotype	ITS: EU554108, JX519216; TUB2: JX519241; ACT: JX519233; CHS-1: JX519224; SOD2: EU554212	Crouch et al. (2009c, d), this study
C. jasminigenum	truncatum	CGMCC LLTX–01, MFU 10–0273	Culture from type	ITS: HM131513; TUB2: HM153770; ACT: HM131508; GAPDH: HM131499; CAL: HM131494; GS: HM131504	Wikee et al. 2010
C. johnstonii	acutatum	CBS 128532, ICMP 12926	Culture from holotype	ITS: JQ948444; TUB2: JQ950095; ACT: JQ949765; CHS-1: JQ949105; GAPDH: JQ948775; HIS3: JQ949435	Damm et al. (2012a)
C. kahawae subsp. ciggaro	gloeosporioides	ICMP 18539	Culture from holotype	ITS: JX010230; TUB2: JX010434; ACT: JX009523; CHS-1: JX009800; GAPDH: JX009966; CAL: JX009635; GS: JX010132; SOD2: JX010346	Weir et al. (2012)
C. kahawae subsp. kahawae	gloeosporioides	IMI 319418, ICMP17816	Culture from holotype	ITS: GU174550, JX010231; TUB2: JX010444; ACT: JX009452; CHS-1: JX009813; GAPDH: GU174562, JX010012; CAL: JX009642; GS: JX010130; SOD2: JX010130	Weir et al. (2012)
C. karstii	boninense	CBS 132134, CORCG6, CGMCC3.14194	Culture from holotype	ITS: HM585409; TUB2: HM585428; ACT: HM581995; CHS-1: HM582023; GAPDH: HM585391; CAL: HM582013	Yang et al. (2011)
C. kinghornii	acutatum	CBS 198.35	Culture from holotype	ITS: JQ948454; TUB2: JQ950105; ACT: JQ949775; CHS-1: JQ949115; GAPDH: JQ948785; HIS3: JQ949445	Damm et al. (2012a)
C. laticiphilum	acutatum	CBS 112989, IMI 383015, STE-U 5303	Culture from holotype	ITS: JQ948289; TUB2: JQ949940; ACT: JQ949610; CHS-1: JQ948950; GAPDH: JQ948619; HIS3: JQ949280	Damm et al. (2012a)
C. lilii	spaethianum	CBS 109214	Morphology congruent with original description	ITS: GU227810; TUB2: GU228104; ACT: GU227908; CHS-1: GU228300; GAPDH: GU228202; HIS3: GU228006	Damm et al. (2009)
C. limetticola	acutatum	CBS 114.14	Culture from epitype	ITS: JQ948193; TUB2: JQ949844; ACT: JQ949514; CHS-1: JQ948854; GAPDH: JQ948523; HIS3: JQ949184	Damm et al. (2012a)
C. lindemuthianum	orbiculare	CBS 144.31	None	ITS: JQ005779; TUB2: JQ005863; ACT: JQ005842; CHS-1: JQ005800; HIS3: JQ005821	O’Connell et al. (2012)
C. lineola	dematium	CBS 125337	Culture from epitype	ITS: GU227829; TUB2: GU228123; ACT: GU227927; CHS-1: GU228319; GAPDH: GU228221; HIS3: GU228025	Damm et al. (2009)
C. linicola	destructivum	CBS 172.51	None	ITS: JQ005765; TUB2: JQ005849; ACT: JQ005828; CHS-1: JQ005786; HIS3: JQ005807	O’Connell et al. (2012)
C. liriopes	spaethianum	CBS 119444	Culture from holotype	ITS: GU227804; TUB2: GU228098; ACT: GU227902; CHS-1: GU228294; GAPDH: GU228196; HIS3: GU228000	Damm et al. (2009)
C. lupini	acutatum	BBA 70884, CBS 109225	Culture from neotype	ITS: DQ286119, JQ948155; TUB2: JQ949806; ACT: JQ949476; CHS-1: JQ948816; GAPDH: JQ948485; HIS3: JQ949146; Mat1/APN2: DQ174704; TUB1: AJ301948	Nirenberg et al. (2002), Damm et al. (2012a)
C. malvarum	orbiculare	LW1	None	GAPDH: DQ792860	Liu et al. (2007a)
C. melonis	acutatum	CBS 159.84	Culture from holotype	ITS: JQ948194; TUB2: JQ949845; ACT: JQ949515; CHS-1: JQ948855; GAPDH: JQ948524; HIS3: JQ949185	Damm et al. (2012a)
C. miscanthi	graminicola	MAFF 510857	Culture from holotype	ITS: EU554121, JX519221; TUB2: JX519246; ACT: JX519237; CHS-1: JX519229; SOD2: EU554224; APN2: EU365028	Crouch et al. (2009c, d), this study
C. musae	gloeosporioides	CBS 116870, ICMP19119	Culture from epitype	ITS: HQ596292, JX010146; TUB2: HQ596280; ACT: HQ596284, JX009433; CHS-1: JX009896; GAPDH: HQ596299, JX010050; CAL: JX009742; GS: HQ596288, JX010103; SOD2: JX010335	Su et al. (2011), Weir et al. (2012)
C. navitas	graminicola	CBS 125086	Culture from holotype	ITS: GQ919067, JQ005769; TUB2: JQ005853; ACT: JQ005832; CHS-1: JQ005790; HIS3: JQ005811; SOD2: GQ919073; Mat1/APN2: GQ919071; APN2: GQ919069	Crouch et al. (2009a), O’Connell et al. (2012)
C. nicholsonii	graminicola	MAFF 511115	Culture from holotype	ITS: EU554126, JQ005770; TUB2: JQ005854; ACT: JQ005833; CHS-1: JQ005791; HIS3: JQ005812; SOD2: EU554229; Mat1/APN2: FJ377946; APN2: EU365033	Crouch et al. (2009c, d), O’Connell et al. (2012)
C. novae-zelandiae	boninense	CBS 128505, ICMP 12944	Culture from holotype	ITS: JQ005228; TUB2: JQ005662; ACT: JQ005576; CHS-1: JQ005402; GAPDH: JQ005315; HIS3: JQ005489; CAL: JQ005749	Damm et al. (2012b)
C. nupharicola	gloeosporioides	CBS 470.96, ICMP 18187	Culture from holotype	ITS: JX010187; TUB2: JX010398; ACT: JX009437; CHS-1: JX009835; GAPDH: JX009972; CAL: JX009663; GS: JX010088; SOD2: JX010320	Weir et al. (2012)
C. nymphaeae	acutatum	CBS 515.78	Culture from epitype	ITS: JQ948197; TUB2: JQ949848; ACT: JQ949518; CHS-1: JQ948858; GAPDH: JQ948527; HIS3: JQ949188	Damm et al. (2012a)
C. oncidii	boninense	CBS 129828	Culture from holotype	ITS: JQ005169; TUB2: JQ005603; ACT: JQ005517; CHS-1: JQ005343; GAPDH: JQ005256; HIS3: JQ005430; CAL: JQ005690	Damm et al. (2012b)
C. orbiculare	orbiculare	LARS 414, 104T, CBS 514.97	None	ITS: JQ005778; TUB2: JQ005862; ACT: JQ005841; CHS-1: JQ005799; HIS3: JQ005820	O’Connell et al. (2012)
C. orchidophilum		CBS 632.80	Culture from holotype	ITS: JQ948151; TUB2: JQ949802; ACT: JQ949472; CHS-1: JQ948812; GAPDH: JQ948481; HIS3: JQ949142	Damm et al. (2012a)
C. parsonsiae	boninense	CBS 128525, ICMP 18590	Culture from holotype	ITS: JQ005233; TUB2: JQ005667; ACT: JQ005581; CHS-1: JQ005407; GAPDH: JQ005320; HIS3: JQ005494; CAL: JQ005754	Damm et al. (2012b)
C. paspali	graminicola	MAFF 305403	Culture from holotype	ITS: EU554100, JX519219; TUB2: JX519244; ACT: JX519235; CHS-1: JX519227; SOD2: EU554204; Mat1/APN2: FJ377921; APN2: EU365007	Crouch et al. (2009c, d), this study
C. paxtonii	acutatum	IMI 165753	Culture from holotype	ITS: JQ948285; TUB2: JQ949936; ACT: JQ949606; CHS-1: JQ948946; GAPDH: JQ948615; HIS3: JQ949276	Damm et al. (2012a)
C. petchii	boninense	CBS 378.94	Culture from epitype	ITS: JQ005223; TUB2: JQ005657; ACT: JQ005571; CHS-1: JQ005397; GAPDH: JQ005310; HIS3: JQ005484; CAL: JQ005744	Damm et al. (2012b)
C. phaseolorum[4]	dematium	CBS 157.36	Authentic strain	ITS: GU227896; TUB2: GU228190; ACT: GU227994; CHS-1: GU228386; GAPDH: GU228288; HIS3: GU228092	Damm et al. (2009)
C. phormii	acutatum	CBS 118194	Culture from epitype	ITS: DQ286136, JQ948446; TUB2: JQ950097; ACT: JQ949767; CHS-1: JQ949107; GAPDH: JQ948777; HIS3: JQ949437; LSU: DQ286137	Farr et al. (2006), Damm et al. (2012a)
C. phyllanthi	boninense	CBS 175.67	Culture from holotype	ITS: JQ005221; TUB2: JQ005655; ACT: JQ005569; CHS-1: JQ005395; GAPDH: JQ005308; HIS3: JQ005482; CAL: JQ005742	Damm et al. (2012b)
C. pseudoacutatum		CBS 436.77	Culture from holotype	ITS: JQ948480; TUB2: JQ950131; ACT: JQ949801; CHS-1: JQ949141; GAPDH: JQ948811; HIS3: JQ949471	Damm et al. (2012a)
C. psidii	gloeosporioides	CBS 145.29*, ICMP 19120	Authentic strain	ITS: JX010219; TUB2: JX010443; ACT: JX009515; CHS-1: JX009901; GAPDH: JX009967; CAL: JX009743; GS: JX010133; SOD2: JX010366	Weir et al. (2012)
C. pyricola	acutatum	CBS 128531, ICMP 12924	Culture from holotype	ITS: JQ948445; TUB2: JQ950096; ACT: JQ949766; CHS-1: JQ949106; GAPDH: JQ948776; HIS3: JQ949436	Damm et al. (2012a)
C. queenslandicum	gloeosporioides	ICMP 1778	Culture from epitype	ITS: JX010276; TUB2: JX010414; ACT: JX009447; CHS-1: JX009899; GAPDH: JX009934; CAL: JX009691; GS: JX010104; SOD2: JX010336	Weir et al. (2012)
C. rhombiforme	acutatum	CBS 129953	Culture from holotype	ITS: JQ948457; TUB2: JQ950108; ACT: JQ949778; CHS-1: JQ949118; GAPDH: JQ948788; HIS3: JQ949448	Damm et al. (2012a)
C. rusci		CBS 119206	Culture from holotype	ITS: GU227818; TUB2: GU228112; ACT: GU227916; CHS-1: GU228308; GAPDH: GU228210; HIS3: GU228014	Damm et al. (2009)
C. salicis	acutatum	CBS 607.94	Culture from epitype	ITS: JQ948460; TUB2: JQ950111; ACT: JQ949781; CHS-1: JQ949121; GAPDH: JQ948791; HIS3: JQ949451	Damm et al. (2012a)
C. salsolae	gloeosporioides	ICMP 19051	Culture from holotype	ITS: JX010242; TUB2: JX010403; ACT: JX009562; CHS-1: JX009863; GAPDH: JX009916; CAL: JX009696; GS: JX010093; SOD2: JX010325	Weir et al. (2012)
C. scovillei	acutatum	CBS 126529, BBA 70349	Culture from holotype	ITS: JQ948267; TUB2: JQ949918; ACT: JQ949588; CHS-1: JQ948928; GAPDH: JQ948597; HIS3: JQ949258	Damm et al. (2012a)
C. sansevieriae		MAFF 239721	Culture from holotype	ITS: AB212991	Nakamura et al. (2006)
C. siamense	gloeosporioides	MFU 090230, ICMP 18578, CBS 130417	Culture from holotype	ITS: FJ972613, JX010171; TUB2: FJ907438, JX010404; ACT: FJ907423; CHS-1: JX009865; GAPDH: FJ972575, JX009924; CAL: FJ917505; GS: FJ972596, JX010094; SOD2: JX010326	Prihastuti et al. (2009), Weir et al. (2012)
C. simmondsii	acutatum	BRIP 28519, CBS 122122	Culture from holotype	ITS: FJ972601, JQ948276; TUB2: FJ907443, JQ949927; ACT: FJ907428, JQ949597; CHS-1: JQ948937; GAPDH: FJ972580, JQ948606; HIS3: JQ949267; CAL: FJ917510; GS: FJ972591	Shivas & Tan (2009), Damm et al. (2012a)
C. sloanei	acutatum	IMI 364297	Culture from holotype	ITS: JQ948287; TUB2: JQ949938; ACT: JQ949608; CHS-1: JQ948948; GAPDH: JQ948617; HIS3: JQ949278	Damm et al. (2012a)
C. spaethianum	spaethianum	CBS 167.49	Culture from epitype	ITS: GU227807; TUB2: GU228101; ACT: GU227905; CHS-1: GU228297; GAPDH: GU228199; HIS3: GU228003; LSU: JN940813	Damm et al. (2009), Schoch et al. (2012)
C. spinaciae	dematium	CBS 128.57	Morphology congruent with original description	ITS: GU227847; TUB2: GU228141; ACT: GU227945; CHS-1: GU228337; GAPDH: GU228239; HIS3: GU228043	Damm et al. (2009),
C. sublineola[5]	graminicola	BPI399463	Lectotype	ITS: JQ478437; HIS3: JQ005813; SOD2: JQ478453; Mat1/APN2: JQ478466; APN2: JQ478477	Crouch & Tomaso-Peterson (2012),
		CBS 131301, S3.001	Culture from epitype	ITS: DQ003114, JQ005771; TUB2: JQ005855; ACT: JQ005834; CHS-1: JQ005792; HIS3: JQ005813; SOD2: DO132051; Mat1/APN2: FJ378029; APN2: EU365121; MAT1-2: DQ002865	Crouch & Tomaso-Peterson (2012), Crouch et al. (2006), O’Connell et al. (2012)
C. tabacum	destructivum	CBS 161.53	None	ITS: JQ005763; TUB2: JQ005847; ACT: JQ005826; CHS-1: JQ005784; HIS3: JQ005805	O’Connell et al. (2012)
C. tamarilloi	acutatum	CBS 129814	Culture from holotype	ITS: JQ948184; TUB2: JQ949835; ACT: JQ949505; CHS-1: JQ948845; GAPDH: JQ948514; HIS3: JQ949175	Damm et al. (2012a)
C. theobromicola	gloeosporioides	ICMP 18649, CBS 124945	Culture from neotype	ITS: GU994360, JX010294; TUB2: GU994477, JX010447; ACT: JX009444; CHS-1: JX009869; GAPDH: JX010006; CAL: JX009591; GS: JX010139; SOD2: JX010372; Mat1/APN2: GU994448; APN2: GU994419; TEF1: GU994506	Rojas et al. (2010), Weir et al. (2012)
C. ti	gloeosporioides	ICMP 4832	Culture from holotype	ITS: JX010269; TUB2: JX010442; ACT: JX009520; CHS-1: JX009898; GAPDH: JX009952; CAL: JX009649; GS: JX010123; SOD2: JX010362	Weir et al. (2012)
C. tofieldiae	spaethianum	CBS 495.85	Morphology congruent with original description	ITS: GU227801; TUB2: GU228095; ACT: GU227899; CHS-1: GU228291; GAPDH: GU228193; HIS3: GU227997; LSU: JN940815	Damm et al. (2009), Schoch et al. (2012)
C. torulosum	boninense	CBS 128544, ICMP 18586	Culture from holotype	ITS: JQ005164; TUB2: JQ005598; ACT: JQ005512; CHS-1: JQ005338; GAPDH: JQ005251; HIS3: JQ005425; CAL: JQ005685	Damm et al. (2012b)
C. trichellum		CBS 217.64	Morphology congruent with original description	ITS: GU227812; TUB2: GU228106; ACT: GU227910; CHS-1: GU228302; GAPDH: GU228204; HIS3: GU228008	Damm et al. (2009)
C. tropicale	gloeosporioides	CBS 124949, ICMP18653	Culture from holotype	ITS: GU994331, JX010264; TUB2: GU994454, JX010407; ACT: JX009489; CHS-1: JX009870; GAPDH: JX010007; CAL: JX009719; GS: JX010097; SOD2: JX010329; Mat1/APN2: GU994425; APN2: GU994396; TEF1: GU994483	Rojas et al. (2010), Weir et al. (2012)
C. truncatum	truncatum	CBS 151.35	Culture from epitype	ITS: GU227862; TUB2: GU228156; ACT: GU227960; CHS-1: GU228352; GAPDH: GU228254; HIS3: GU228058; LSU: JN940819	Damm et al. (2009), Schoch et al. (2012)
C. verruculosum	spaethianum	IMI 45525	Culture from holotype	ITS: GU227806; TUB2: GU228100; ACT: GU227904; CHS-1: GU228296; GAPDH: GU228198; HIS3: GU228002	Damm et al. (2009)
C. walleri	acutatum	CBS 125472	Culture from holotype	ITS: JQ948275; TUB2: JQ949926; ACT: JQ949596; CHS-1: JQ948936; GAPDH: JQ948605; HIS3: JQ949266	Damm et al. (2012a)
C. xanthorrhoeae	gloeosporioides	BRIP 45094, ICMP 17903, CBS127831	Culture from holotype	ITS: GU048667, GU174551, JX010261; TUB2: JX010448; ACT: JX009478; CHS-1: JX009823; GAPDH: GU174563, JX009927; CAL: JX009653; GS: JX010138; SOD2: JX010369; TEF1: GU174575	Hyde et al. (2009), Weir & Johnston (2010), Weir et al. (2012)
C. yunnanense		CGMCC AS3.9167, CBS 132135	Culture from holotype	ITS: EF369490 ; TUB2: JX519248; ACT: JX519239; CHS-1: JX519231	Liu et al. (2007b), this study

ARSEF: ARS Collection of Entomopathogenic Fungal Cultures, Ithaca, NY, USA.

BBA: Culture collection of the Biologische Bundesanstalt für Land- und Forstwirtschaft, Berlin, Germany.

BPI: Systematic Mycology and Microbiology Laboratory, USDA Agricultural Research Service, Beltsville, MD, USA.

BRIP: Culture Collection of the DPI&F Plant Pathology Herbarium, Indooroopilly, Queensland, Australia.

CBS: Culture collection of the Centraalbureau voor Schimmelcultures, Fungal Biodiversity Centre, Utrecht, The Netherlands.

CGMCC: China General Microbiological Culture Collection Center, Beijing, China.

ICMP: International Collection of Microorganisms from Plants, Landcare Research, Auckland, New Zealand.

IMI: Culture collection of CABI Europe UK Centre, Egham, UK.

M1.001: sourced from Lisa Vaillancourt, University of Kentucky.

MAFF: NIAS Genebank, Microorganism Section, Tsukuba, Japan.

MFU: fungarium of Mae Fah Luang University, Thailand (cultures in BCC (BIOTEC Culture Collection, Thailand).

NBRC: NITE Biological Resource Center, Chiba, Japan.

STE-U: Culture collection of the Department of Plant Pathology, University of Stellenbosch, South Africa.

^[1]Where possible, all taxa are represented by sequences from type or other authentic material. For some however, the necessary research to identify such cultures and/or to designate epitype material is not complete, especially for species within the destructivum and orbiculare clades. To be able to generate robust phylogenetic trees for the entire genus (Figs 2, 3) that include all of the major clades, we have used sequences from some strains that have been used to represent the relevant species (mostly in recent literature) but which do not currently have any special nomenclatural status. Their details are included in Table 3 for reference, and can be recognised with “none” in the type status column. It may be that some or all of these strains will be designated as epitypes in the future, but for the present it should not be assumed that they represent the species as originally circumscribed.

^[2]KS20BIG was one of four epitypes designated by Crouch et al. (2006) for C. cereale; the application of the name needs to be more precisely established.

^[3]Preliminary multilocus analysis suggests that C. chrysanthemi may not be a synonym of C. carthami as stated by Uematsu et al. (2012).

^[4]These sequences derive from one of two authentic but not genetically identical strains; the species was not epitypified as neither of them are now fertile.

^[5]A further collection from which a culture was obtained (CBS 131301) was designated as an epitype by Crouch et al. (2006) and recognised as representative of the species also by Du et al. (2005) and Crouch et al. (2009d). It was subsequently confirmed as closely similar to the lectotype based on multilocus DNA sequence analysis (Crouch & Tomaso-Peterson 2012).

These data form the framework for an online identification system for Colletotrichum species, hosted by the Centraalbureau voor Schimmelcultures but administered by the recently formed Colletotrichum subcommission of the International Commission on Taxonomy of Fungi (ICTF; http://www.fungaltaxonomy.org/), which is in turn a body under the auspices of both the International Mycological Association (http://www.ima-mycology.org/) and the International Union of Microbiological Societies (http://www.iums.org/). This database can be accessed at http://www.cbs.knaw.nl/Colletotrichum/. The database will be updated periodically to include reference sequences for novel taxa and for species that have been subjected to modern phylogeny-based revision.

SPECIES CONCEPTS AND BARCODING

Our understanding of Colletotrichum species and the processes by which they have evolved has undergone several step changes over the years. The first part of this review focuses on the unreliability of host-based diagnosis, and the lack of resolution of taxonomic systems based firstly on morphological features, and latterly by ITS rDNA sequences. Here, we concentrate on the changes of the last 10 years, with rapid moves to species definition based on multilocus analysis, knowledge gains from molecular plant/fungus interaction studies, and the synergies with wider genetic research.

At the beginning of the century, concern was expressed at the wide constituent genetic variation between taxa of Colletotrichum recognised at the species level, and the varying utility of species concepts in the eyes of pathologists (Cannon et al. 2000). Some species, such as C. gloeosporioides, were defined partially by ITS sequence, but were primarily considered to represent morphological taxa. These were known to encompass extensive genetic variation, but were maintained for utilitarian reasons. Colletotrichum kahawae on the other hand was thought at the time to represent a single clonal population causing a specific, devastating disease of coffee berries. That species has recently been redefined with a broader circumscription (Weir et al. 2012).

In Colletotrichum, species definition based on ITS sequence has proved unsatisfactory, that gene fragment being too evolutionarily conservative to distinguish between taxa that can be recognised using other genes and gene combinations (e.g. Du et al. 2005, Crouch et al. 2009b, Gazis et al. 2009). This is of some concern, as the ITS region is widely used for species definition in the Fungi (e.g. Begerow et al. 2010, Druzhinina et al. 2005, Eberhardt 2010, Kelly et al. 2011), and has recently been proposed as a universal barcode sequence (Schoch et al. 2011, 2012).

ITS was proposed as the primary fungal barcode marker for various reasons, including pragmatism – the number of existing fungal ITS sequences is far greater than that for any other gene. Many other genes/gene fragments have been used for diagnostic purposes in the Fungi, especially beta-tubulin (TUB2) and calmodulin (e.g. for Aspergillus and Penicillium; Samson et al. 2007, Peterson 2008, Houbraken et al. 2011), TEF1 (for Fusarium; Geiser et al. 2004, O’Donnell et al. 2009) and COX1 (for Penicillium; Seifert et al. 2007).

Many other molecular markers have wide diagnostic potential for the Fungi, including most of those currently used for phylogenetic analysis in Colletotrichum (see Table 3). Further candidates are being considered. Aguileta et al. (2008) identified no fewer than 246 single-copy orthologous gene clusters in an optimally performing gene set, from analysis of 21 fungal genomes. Several widely used markers, including TUB2 and TEF1, were not included within their list of best-performing genes, and are probably unsuitable as universal fungal markers due to the presence of paralogs (James et al. 2006, Walker et al. 2012). Building on this work, Schmitt et al. (2009) developed primer sets for MCM7 and Tsr1, two of the most phylogenetically informative sequences identified by Aguileta et al. (2008). MCM7 has been shown to work effectively in widely divergent fungal groups within the Ascomycota (Schmitt et al. 2009, Raja et al. 2011). Walker et al. (2012) evaluated two further single-copy protein-encoding genes, FG1093 and MS204 that also have potential in fungal diagnostics.

The prospect of a single short universally amplifiable DNA sequence being diagnostic for all organisms (or even all species within a major taxonomic group) is enticing, but unrealistic. This does not mean that data from single loci such as ITS do not have wide application, for example in environmental sequencing (e.g. Buée et al. 2009) or analysis of historical specimens (e.g. Brock et al. 2009, Dentinger et al. 2010b). There is also evidence that ITS sequences alone can constitute useful barcode markers for some groups of the Basidiomycota (e.g. Kõljalg et al. 2005, Dentinger et al. 2011). It is not clear whether this apparent difference in utility of ITS-based diagnostics between ascomycetous and basidiomycetous fungi reflects different speciation patterns or variation in species concepts.

Comparison of a phylogenetic tree of Colletotrichum species derived from ITS sequences alone and one generated from multilocus data (Figs 2, 3) confirms that ITS resolves major clades well, though does not reflect their higher-order topology accurately in all cases. However, posterior probability support is lacking within many of the major clades, especially those containing C. acutatum and C. gloeosporioides and their respective relatives. A robust sequence-based identification system for Colletotrichum species must therefore use an alternative molecular marker, or a combination of markers.

Fig. 3.

Phylogenetic tree derived from a Bayesian analysis of a partitioned, concatenated alignment of CHS-1 (251 bp), ACT (305 bp), TUB2 (545 bp) and ITS (599 bp) sequences, run for 1×10⁷ generations with a GTR+I+Γ model of DNA evolution for each partition. The major clades recognised in this paper are indicated. Other details as per Fig. 2.

Performance analysis of the genes used in a multilocus analysis of the C. acutatum clade (Damm et al. 2012a) indicates that the two most diagnostic markers are TUB2 and GAPDH, which resolved all 29 subclades. These were equated by those authors to species. In contrast, ITS sequences could only resolve 11 of the 29 taxa within the clade. TUB2 performed marginally better than GAPDH due to a larger overall number of bp differences, but even so, some clades differed only by one bp in the TUB2 sequence. An identification system based on this gene alone would therefore be vulnerable to sequencing error, suggesting that data from multiple loci should be used.

Multilocus phylogenies are now typically used as the primary basis on which to describe new species of Colletotrichum (see Table 1) and the trend is to include more and more sequences into the analyses. One might conclude that phylogenetic signal is strongly correlated with the number of characters (in this case base pairs) included in the analysis, a position first advanced nearly 250 years ago (Adanson 1763), but genes are differential at varying positions in the hierachy of taxa. Inclusion of multiple genes that resolve at similar positions in the hierarchy can therefore increase the size (not to mention the cost) of the data set without clarifying the phylogenetic signal. This is highly relevant to species diagnosis, as was observed by Min & Hickey (2007) in a study of mitochondrial genes from 31 fungi of widely varying taxonomic position to determine the optimum sequence length for robust identification. Research by Dentinger et al. (2010a) showed that both bootstrap support and Bayesian posterior probability values were eroded in a multilocus ATP6/LSU/RPB1 analysis of Boletus species compared with an analysis based on RPB1 alone. Similar results were obtained by Walker et al. (2012) in a study on two genera of the Diaporthales. They found in an analysis of Ophiognomonia species that adding TEF1 sequence data to any combination of three of the other loci used (ITS, Tub2, FG1093 and MS204) decreased support and increased the number of tree topologies recovered. Our own preliminary studies on Colletotrichum (data not shown) also indicate that in some circumstances, increasing the number of loci may decrease phylogenetic performance, although the effect is minor. Taken together, these data suggest that the recent fungal phylogenetic “arms race”, whereby a steadily increasing number of loci are analysed in concert, may add complexity but not improve insight.

MAJOR CLADES

Phylogenetic analysis of the genus Colletotrichum reveals that it comprises nine major clades, as well as a number of small clusters and isolated species (Figs 2, 3).

There is currently no universally accepted process for naming clades and reconciling them with the traditional taxonomic categories of the International Code of Nomenclature for Algae, Fungi and Plants (ICNAFP), although the draft PhyloCode (http://www.ohio.edu/phylocode/) represents a major step in this direction. Formal recognition of infrageneric categories within Colletotrichum is highly desirable. This is for phylogenetic reasons, in that the genus contains many monophyletic subunits with common characteristics (not least in spore morphology). There are also pragmatic reasons for defining such categories, for example to allow linkage to the immense historical body of pathological literature in which the fungal subjects are not assignable to currently accepted species.

Use of the strictly hierarchical infrageneric nomenclature system in the ICNAFP is a possible way to assign formal names to species groups within Colletotrichum. However, although the Code allows for extra categories to be interspersed between the three formal ranks (subgenus, section and series), their adoption implies an equality of taxa at the same rank that is not reconcilable with evolutionary processes. We therefore favour a formal (or at least semi-formal) clade-based nomenclature system.

In this paper, we refer to 119 Colletotrichum species (Table 3) that collectively encompass almost all of the known phylogenetic variety within the genus, most of them belonging to one of the nine major clades. Additionally, there is a number of small clusters and isolated species, which we believe to represent independent evolutionary units, but which are insufficiently well known to justify formal nomenclatural recognition. Throughout this paper, we refer to these clades using the specific epithet of the first-recognised (or historically most prominent) of their constituent species – for example the acutatum clade is the monophyletic unit containing C. acutatum and its close relatives (see Fig. 3). An obvious shortcoming of this system is that there is no objective method of deciding which is the basal node of the named clade. In the case of the acutatum clade, we have decided that the clade has C. orchidophilum as its sister taxon, because the ingroup taxa are much more closely related to each other than to C. orchidophilum or C. pseudoacutatum, but there are arguments for extending the clade to include this species, and indeed also C. pseudoacutatum. The species in the graminicola clade are much less closely related than those of the C. acutatum clade; the decision for combining them was made rather on the basis of common morphology and host family. The process is to some extent subjective, so while we commend adoption of the nine clades detailed below as formal entities, we hope that clade definition and recognition will be taken on as a task by the new ICTF Subcommission on Colletotrichum.

In this paper, reference to the term clade indicates that we are confident that the associated information can be referred to our formal clades (or to species within the clades). We also refer on occasion to informal groupings of taxa, generally as species clusters. In these circumstances, we may know that the knowledge is associated with a particular species group, but are unsure as to its constituent taxa, or to the phylogenetic extent it represents. This frequently occurs when attempting to relate information from pathology papers to our new phylogeny.

Several of the clades indicated in Fig. 3 represent the species complexes as defined by Crouch et al. (2009c, d), Damm et al. (2012a, b, this issue), and Weir et al. (2012, this issue). While these four complexes can be confirmed as monophyletic, the assemblage of curved-spored species from herbaceous hosts studied by Damm et al. (2009) can be seen to be polyphyletic; the species included in that research are placed in three of the formal clades we recognise here, with additional outliers.

In this section, we provide an overview of the nine Colletotrichum clades that we recognise. Several additional individual species and small clusters are recognised that do not fall into clear clades (see Fig. 3). The phylogenetic tree presented as Fig. 3 provides a comprehensive visual overview of phylogenetic diversity within Colletotrichum as treated in the current literature, but it seems likely that there are further outlying taxa that have not yet been sampled, or for which phylogenetic positions have not been fixed effectively. For example, the tea pathogen C. theae-sinensis (Moriwaki et al. 2002, Yoshida & Takeda 2006) has unusually small conidia and may well fall outside of Colletotrichum as outlined in Fig. 3. Although Moriwaki et al. (2002) included a strain of C. theae-sinensis in phylogenies derived from rDNA datasets, the relevant sequence data from that study are not found in public sequence repositories. Based on these rDNA sequence data, Moriwaki and colleagues suggested that C. theae-sinensis might constitute a sister group to the genus, a prediction that needs to be tested further.

Acutatum clade

The acutatum clade is defined as a collective of Colletotrichum acutatum and 29 closely related species (see Fig. 3), with C. orchidophilum as sister taxon. The clade, along with a small number of outlying taxa, forms a sister taxon to a combination of the destructivum, graminicola and spaethianum clades and C. coccodes. Two principal subclades may be detected within the acutatum clade, containing 19 and nine species respectively, and C. acutatum sensu stricto is resolved as an outlier of a clade consisting of the larger of the two subclades along with C. fioriniae. The acutatum clade can be effectively resolved using ITS sequence data alone (Fig. 2). The major subclades are also distinguishable using ITS alone, but the analysis reveals little or no internal structure within the subclades. A comprehensive account of its constituent species can be found as Damm et al. (2012a).

Boninense clade

The boninense clade contains 17 species as defined here (see Fig. 3). It forms a sister taxon to the gloeosporioides clade, and our multilocus analysis reveals three subclades containing 12, three and two species respectively. Colletotrichum boninense sensu stricto falls within the largest subclade. The nodal structure is complex and we do not see good reason to name the subclades formally. The ITS tree (Fig. 2) shows that the boninense clade can be detected effectively using this single locus, but it is resolved as a sister clade to the truncatum rather than the gloeosporioides clade. The clade has been revised in detail by Damm et al. (2012b).

Dematium clade

The dematium clade contains the type species of the genus, Colletotrichum lineola, and was investigated by Damm et al. (2009), as part of a study of Colletotrichum species with curved conidia. As defined by ourselves, the dematium clade contains six species (Fig. 3) and forms a sister clade to a superclade consisting of the acutatum, destructivum, graminicola and spaethianum clades, along with five further outlying taxa. In the ITS tree (Fig. 2) the clade is fairly well resolved with a Bayesian posterior probability value of 0.89, but the structure of the superclade referred to above is less well defined. An additional species, C. hemerocallidis, closely related to C. dematium, was described just before finishing this review (Yang et al. 2012).

Colletotrichum dematium and C. truncatum (often referred to under its synonym C. capsici) have been confused historically (Sutton 1981), but are found to occupy distinct clades, with the latter species belonging to a small clade near the base of the multilocus phylogeny (Fig. 3). Strains of the six species included in the dematium clade appear to be characteristic of temperate environments, though the sample size for several of the species is inadequate to allow definite conclusions as to their climatic range. In general, members of the dematium clade are not significant in economic terms, but C. spinaciae (a pathogen of Beta and Spinacia; Gourley 1966, Washington et al. 2006) and C. circinans (attacking Allium species; Hall et al. 2007, Kim et al. 2008) both cause substantial crop losses under some circumstances. These two plant pathogenic species occupy a well-defined subclade distinct from a separate subclade made up of the putatively saprobic species C. dematium, C. lineola, C. fructi and C. anthrisci (Damm et al. 2009; Fig. 3). The type species of Colletotrichum, C. lineola, belongs to the dematium clade; it was described by Corda (1831) but treated as a synonym of C. dematium by von Arx (1957) and Sutton (1981). However, research based on newly collected strains from the region of the original collection showed that C. lineola and C. dematium are separable based on DNA sequence data (Damm et al. 2009).

Destructivum clade

The destructivum clade contains several important plant pathogens, but to date has not been studied in depth using molecular methods. Economically significant constituent taxa include Colletotrichum destructivum, C. fuscum, C. higginsianum and C. linicola. Colletotrichum destructivum is considered to be pathogenic on lucerne (alfalfa; Medicago sativa) and soybean (Glycine max) (Manandhar et al. 1986, Latunde-Dada et al. 1999), and has also been reported to parasitise a range of unrelated plants including species in the Brassicaceae, Cuscutaceae, Lamiaceae and Solanaceae (reviewed in Hyde et al. 2009a). Colletotrichum higginsianum is known as a pathogen of Brassicaceae (Huser et al. 2009) that is responsible for crop losses in northern temperate climates, and was found to be related to C. destructivum by O’Connell et al. (2004). The fungus is of particular significance as the subject of a whole-genome analysis project, and is increasingly studied as a model for host/pathogen interactions because of its pathogenicity to the model plant Arabidopsis thaliana (Birker et al. 2009, Huser et al. 2009, Kleeman et al. 2012, O’Connell et al. 2012). Colletotrichum higginsianum was reported to be synonymous with C. destructivum by Sun & Zhang (2009) based on ITS sequence similarity, but multilocus phylogenies of strains provisionally accepted as representative of C. higginsianum and C. destructivum indicate that these two species are distinct entities (O’Connell et al. 2012 and Fig. 3 of this study). Thus, although formal taxonomic work with authentic types is still pending, it appears that as with other Colletotrichum groups, the ITS sequence is not sufficiently differential within the destructivum clade to act as a species-level marker in isolation.

Colletotrichum fuscum is a pathogen of Digitalis and Nemesia (Scrophulariaceae; Tomioka et al. 2001). ITS and multilocus data place this species within the destructivum clade (Moriwaki et al. 2002, Cannon et al. 2008; Figs 2, 3), but more detailed information on its taxonomy and phylogenetic relationships is needed. Similarly, C. linicola was shown to belong in this clade based on ITS2/D2 rDNA sequences (Latunde-Dada & Lucas 2007), and preliminary multilocus studies indicate that the species is clearly distinct from others belonging to the destructivum clade (O’Connell et al. 2012; Fig 2).

Glomerella truncata was described as the teleomorph of C. truncatum (Armstrong-Cho & Banniza 2006, Menat et al. 2012), but the strains studied (from lentil (Lens culinaris) in Canada) belong to the destructivum rather than the truncatum clade (Damm et al. 2009; O’Connell et al. 2012; Figs 2, 3). The name G. truncata remains valid and legitimate to represent a taxon within the destructivum clade despite the misidentification of its anamorph, but assuming that no earlier synonyms are discovered, it will require a new name now that separate binomials for teleomorph and anamorph are prohibited (Hawksworth 2011) to avoid homonymy with C. truncatum.

An outline whole-genus multilocus phylogeny (O’Connell et al. 2012) shows that the destructivum clade is monophyletic and distinct from other clades within Colletotrichum. This is confirmed by our present multilocus study (Fig. 3), with the destructivum clade being resolved as a sister taxon to the combined graminicola and spaethianum clades, and it is also clearly resolved using ITS data alone (Fig. 2). However, none of the strains sequenced in these studies is derived from type or authentic material for the names used, and further research is required to elucidate species concepts and correct nomenclature.

Gloeosporioides clade

The C. gloeosporioides species complex has been studied by Weir et al. (2012, this issue). It is a well-supported clade (Bayesian posterior probability value 1) on a very long branch and shows few differences in the gene loci studied between most of the 22 species included. However it is a diverse clade in terms of morphology and includes a number of important plant pathogens. Weir et al. (2012) recognised two subclades within the species complex based on an eight-locus analysis, both of which were supported by Bayesian posterior probability values of 1. They were named as the kahawae and musae clades. Only one of these, the kahawae clade, can be detected unequivocally in our multigene phylogeny (Fig. 3), while the musae clade as recognised by Weir et al. (2012) has a Bayesian posterior probability value of only 0.59. This is a result of the limited number of loci that could be included in the genus-wide alignment. The subclades cannot be effectively distinguished using ITS sequence data alone (see Fig. 2).

Graminicola clade

The Colletotrichum species associated with grasses form a well-defined monophyletic clade, the species of which possess characteristic widely falcate conidia. It is the only major clade that appears to be composed (at least largely) of host-specific taxa (Crouch & Beirn 2009), although further research may confirm that the orbiculare clade shares this characteristic. Multilocus analyses (Fig. 3) revealed two major subclades within the graminicola clade, in agreement with studies published by Crouch et al. (2009c, d). One, represented only by a single strain in Fig. 3, contains the plurivorous taxon Colletotrichum cereale. This is a diverse taxon in phylogenetic terms and there is evidence of significant gene flow between the various constituent populations (Crouch, in litt. Aug. 2012). Colletotrichum cereale is associated with grasses with C3 (cool-season) photosynthetic pathways as either pathogens or endophytes (Crouch et al. 2009d). The second subclade affects C4 (warm-season) grasses including several economically important cereal crops (Crouch et al. 2009a) and comprises a number of apparently host-specific species, not all of which have been described to date (Crouch et al. 2009c, Prihastuti et al. 2010). Several of the species included in the graminicola clade are of major importance, including C. falcatum on sugarcane (Saccharum), C. graminicola on maize (Zea) and C. sublineola on Sorghum species. Colletotrichum cereale and C. eremochloae are pathogens of cultivated turfgrasses (Crouch & Beirn 2009). Research has demonstrated the inadequacy of ITS sequences to differentiate between species within this group (Crouch et al. 2009b), and multigene analyses to date do not clearly resolve relationships within the major subclade (Crouch et al. 2009c, Fig. 3). The biology and evolution of the clade was reviewed by Crouch & Beirn (2009), focusing on the genetics, biology and epidemiology of the three best-researched species, C. falcatum, C. graminicola and C. sublineola. The first two of these species are essentially homothallic, while C. sublineola may be strictly heterothallic (Vaillancourt & Hanau 1992, Vaillancourt et al. 2000). With the exception of C. falcatum, the teleomorphs of these species have never been encountered in nature (Crouch & Beirn 2009). A whole-genome analysis of a strain of C. graminicola has recently been completed (O’Connell et al. 2012) and this work is now being extended to include further strains from grass hosts (http://www.ars.usda.gov/pandp/docs.htm?docid=22211).

Orbiculare clade

The orbiculare clade contains several important pathogen assemblages. It has been studied in a preliminary fashion from a molecular phylogenetic perspective, but has not been the subject of a recent formal revision. The orbiculare clade is thought to include the species Colletotrichum lindemuthianum, C. malvarum, C. orbiculare and C. trifolii (Liu et al. 2007). Multilocus phylogenies using provisionally identified strains of C. lindemuthianum and C. orbiculare (Fig. 3) show that the orbiculare group occupies a basal clade of Colletotrichum, and that separation of these taxa from Colletotrichum at generic level cannot at present be ruled out. Members of the orbiculare clade as it is currently understood share some morphological features including conidia that are not curved and are relatively short and broad, and small appressoria with simple outlines (Sutton 1980). It must be pointed out that none of these taxa has been adequately typified and linked to authentic sequences. There are in fact separate concepts in the literature for three of the species currently placed within the orbiculare clade (see below), which contributes in no small way to confusion over their identity.

As pointed out by Cannon et al. (2000), Mordue (1971) considered C. lindemuthianum to have relatively long narrow conidia with a very large size range. Mordue’s illustration shows a species that would be placed in the gloeosporioides cluster based on morphological data by most authors. Sutton (1980) described and illustrated C. lindemuthianum with short, broad and rounded conidia – typical of those here included in the orbiculare clade (Fig. 3). The confusion presumably arose due to the frequent occurrence of fungi from the gloeosporioides cluster on host plants belonging to the Fabaceae. A similar confusion seems to exist for C. orbiculare; the species as described and illustrated by Baxter et al. (1983) has much longer conidia than those of the taxon as defined by other authors, and again it seems possible that strains of the gloeosporioides clade parasitising cucurbits were misidentified. Until both species names are properly typified using modern methods, confusion is likely to continue. As far as we can tell, all of the sequence-based research (bar a single sequence derived from a Taiwanese strain that is certainly misidentified; see Fig. 4) and probably a large majority of pathology reports using the names C. lindemuthianum and C. orbiculare refer to the short-spored taxa belonging to the orbiculare clade. As such, it would be highly appropriate to fix application of these species names to allow their continued use in this manner. Approximately half of the ITS sequences of strains identified as C. trifolii are placed in the destructivum rather than the orbiculare clade (see Fig. 4). Further research is needed before the most appropriate typification can be made; however the original description (Bain & Essary 1906) gives conidial dimensions and shape that are typical of the orbiculare clade.

Fig. 4.

. Maximum likelihood phylogeny based on an ITS alignment of GenBank accessions of the C. lindemuthianum, C. orbiculare, C. trifolii and C. destructivum species complexes (alignment with ClustalX, 1000 bootstrap replicates, PhyML package).

The orbiculare clade was recognised as a monophyletic unit by Sherriff et al. (1994) and Johnston & Jones (1997) using LSU sequence analysis, Sreenivasaprasad et al. (1996) using ITS data, and Farr et al. (2006) using both gene sequences. A preliminary phylogenetic analysis based only on existing ITS sequences curated by GenBank (Fig. 4) demonstrates that the orbiculare clade is a sister taxon to the whole of the rest of the genus Colletotrichum. This result is consistent with previous research findings. For example, an ITS tree constructed by Yang et al. (2009) showed the orbiculare clade as a sister to C. cliviae, with the combined clade sister to C. yunnanense and C. dracaenophilum, but the clade comprising all three taxa was supported by bootstrap values below 50. Liu et al. (2007) published a phylogenetic analysis of the orbiculare clade, based on GAPDH and GS sequences; this also indicated that the orbiculare group is monophyletic, and that C. lindemuthianum, C. malvarum and C. trifolii form separate clades from a paraphyletic C. orbiculare.

As with other Colletotrichum clades, ITS data do not appear to be sufficiently variable for species level diagnostics within the orbiculare assemblage. However, ITS data do indicate (Fig. 4) that C. lindemuthianum is a separate lineage from C. orbiculare and C. trifolii, and that it might comprise more than one taxon. An analysis of C. lindemuthianum rDNA data by Balardin et al. (1999) showed that Phaseolus pathogens may occur in numerous subordinate clades within the lindemuthianum subclade. The number of sequences available is too small for confidence, but it does appear that C. lindemuthianum is specific to Phaseolus. However, none of the sequence data or strains used by Balardin et al. (1999) is available through public databases or collections; therefore these conclusions require further evaluation. There are no full ITS sequences from Colletotrichum malvarum available from public databanks, but a study using ITS2/LSU (Bailey et al. 1996) indicated that Colletotrichum species from Malvaceae occupy at least three subclades within the overall orbiculare clade.

Spaethianum clade

The spaethianum clade receives strong support in both the multilocus and ITS-only analyses (Figs 2, 3). It contains only five species as currently circumscribed, four of which are associated with petaloid monocot plants, and none appears to have economic importance. Its phylogenetic significance is as a sister group to the graminicola clade. The spaethianum clade was recognised as a distinct assemblage by Damm et al. (2009) in their work on the non-grass associated species of Colletotrichum with curved conidia. Four of the five species in this assemblage have complex appressoria, but the clade does not otherwise have diagnostic characteristics in morphological terms.

Truncatum clade

The truncatum clade includes only one major species, C. truncatum (also frequently referred to as C. capsici; Damm et al. 2009), which is reported as an economically destructive pathogen of many tropical crops including legumes and solanaceous plants. The truncatum clade occupies a sister position to the combined C. gloeosporioides and C. boninense clade according to our multilocus analysis (Fig. 3), but to the boninense clade only in the ITS-only analysis (Fig. 2). Conidial morphology in the truncatum group is quite different to that found in the gloeosporioides and boninense clades (Damm et al. 2012b, Weir et al. 2012), providing evidence to support the old hypothesis (Sreenivasaprasad et al. 1996) that the evolution of conidial form followed a complex pattern in Colletotrichum.

Colletotrichum curcumae also belongs to this clade, a poorly-known species considered to be the causal agent of turmeric leaf spot disease (Curcuma longa, Zingiberaceae; Palarpawar & Ghurde 1988). The third member of the clade is C. jasminigenum, which was described as a new species causing leaf and blossom anthracnose disease on Jasminum sambac in Vietnam (Wikee et al. 2011).

Other taxa

Our multilocus tree (Fig. 3) includes various species that are isolated in phylogenetic terms, or form small clusters that do not justify recognition as major clades.

The most important of these species in economic terms is Colletotrichum coccodes. This is primarily a pathogen of Solanaceae (potato and tomato), but also survives well in soil and is reported as an associate of a wide range of crops including strawberry (Buddie et al. 1999, Heilmann et al. 2006). Colletotrichum coccodes was recently epitypified (Liu et al. 2011). The species is known to be variable in genetic terms (Ben-Daniel et al. 2010). It has been researched into as a potential biocontrol agent for Abutilon theophrasti (Dauch et al. 2006). Colletotrichum coccodes has distinctive conidia that are straight, have acute ends and are often slightly constricted in the mid portion. Our multilocus analysis (Fig. 3) places it as a sister taxon to the destructivum/spaethianum/graminicola clade. In our ITS-only tree (Fig. 2) it occupies the same position, although the posterior probability values are inadequate to confirm its phylogeny from this gene fragment alone.

Colletotrichum trichellum was placed into synonymy with C. dematium by von Arx (1957), though it was treated as a separate, apparently host-limited species by Sutton (1962, 1981) based on the degree of curvature of the conidia. ITS-only and multilocus phylogenetic analyses (Figs 2, 3) indicate that this species does not belong to the dematium clade, but forms a sister clade (along with C. rusci) with the acutatum clade.

Three poorly-known species occupy basal positions in the ITS-only and multilocus phylogenetic trees (Figs 2, 3). Colletotrichum cliviae (from anthracnose of Clivia miniata, Amaryllidaceae; Yang et al. 2009) appears to constitute a monophyletic lineage that is a sister clade to the entire genus apart from the orbiculare clade. Colletotrichum yunnanense and C. dracaenophilum together form a small clade that is basal to the entire genus apart from the combined orbiculare and C. cliviae clade. Colletotrichum dracaenophilum is a stem pathogen of Dracaena species (Asparagaceae; Farr et al. 2006), while C. yunnanense was isolated as an endophyte of Buxus (Buxaceae; Liu et al. 2007b). According to their publishing authors, all three species have unusually large conidia. Colletotrichum yunnanense and C. cliviae have complex appressoria; those of C. dracaenophilum were not recorded by the describing authors.

WHERE DO WE GO FROM HERE?

What more can we learn about Colletotrichum systematics? Several of the major clades have not yet been analysed comprehensively using multilocus technologies. The phylogenetic position of a large part of the species described is still unknown; these species would have to be recollected and epitypified. However, linking new strains to old species is difficult and there are hundreds of “forgotten species” with little information among them. We should therefore focus on clarifying the identity of well-known species that are commonly used and of Glomerella species in order to synonymise them in Colletotrichum. New species have been discovered regularly over the last five years (including some that are highly distinct in phylogenetic terms) and novel taxa will doubtless continue to appear. Studies of Colletotrichum from wild plants would be likely to be particularly fruitful, and provide insights into the taxa currently known from crops and ornamentals. It would be presumptuous even to speculate that the overall systematic framework for the genus cannot be improved.

Future innovations are likely to focus increasingly on understanding populations and host/parasite relationships, and on using increasingly sophisticated analyses of whole genomes. It is only then that we are likely to begin to understand Colletotrichum species in their evolutionary context, rather than as cultures in collections. The first major output in this new era of Colletotrichum research has now been published (O’Connell et al. 2012), devoted to a comparison of the genomes and transcriptomes of two individual strains of Colletotrichum, one each from C. graminicola (from Zea mays, Poaceae) and C. higginsianum (from Brassica capestris, Brassicaceae). Overall genome size and chromosome number was found to be broadly similar, but substantial differences were noted between the two taxa in intrachromosomal organisation and in their suites of pathogenicity-related genes, These last were shown to be a reflection of differing host cell wall characteristics; cell walls of Poaceae contain higher quantities of hemicellulose and phenolic compounds, while those of Brassicaceae are richer in pectins. The two species were estimated as diverging aound 47 M years ago, well after the divergence of their host clades.

Recent changes to the newly renamed International Code of Nomenclature for Algae, Fungi and Plants (Hawksworth 2011), especially those Articles relating to registration of names and the abolition of the dual nomenclature system for Fungi, mark a further step away from the inflexible application of the rules of date priority towards a consensus approach for choosing between competing names. In response to these historic changes, the International Subcommission on Colletotrichum Taxonomy has been set up within the framework of the International Commission on the Taxonomy of Fungi (http://www.fungaltaxonomy.org/). Its remit will be to promote nomenclatural stability for the genus, develop consensus phylogenies, and develop a list of protected names for key taxa that cannot be overturned by the rediscovery of obscure earlier names within the historical literature. An important part of this work is to ensure that all currently accepted species of Colletotrichum are adequately typified, with epitypes or neotypes linked to cultures where original type material is lost or inadequate for modern phylogenetic placement, or where no authentic original cultures have been preserved.

In the context of moving to a single name system for these fungi, probably few would argue for the retention of Glomerella (the later, sexual genus name with priority until the Melbourne nomenclatural congress in 2011) over Colletotrichum (the earlier, asexual name), but it will be the responsibility of the Subcommission to weigh the arguments for each and to recommend one or the other. Technically, we are aware that our publication prejudges this issue, but the transfer of such a large number of the names of multiple well-known economically important species currently accepted as Colletotrichum to Glomerella would cause chaos amongst the user community. The issues of synonymy between anamorph and teleomorph at the species level are complex (as exemplified by our knowledge of the identities of Glomerella acutata (Damm et al. 2012a) and Ga. cingulata (Weir et al. 2012), and it will in most cases be more practical to assign protected status to the asexual species names rather than go through the formal nomenclatural conservation procedures.

A further important activity for the Colletotrichum community is to establish a robust phylogeny-based online identification system with barcode reference sequences from ex-type or other verified material that can be queried using Blast tools, as a rigorous alternative to the uncurated data set accessible in GenBank. A preliminary system has already been set up by the CBS-KNAW Fungal Biodiversity Centre based on the multilocus sequence data listed in Table 3 (http://www.cbs.knaw.nl/colletotrichum). In addition to sequences, it will also include morphological and cultural characters and pictures of each species facilitating comprehensive polyphasic identification. Methods used to collect the data are explained, and cultures are listed along with ecological data available. This database will be updated as new taxa are discovered and typifications completed by members of the Subcommission on Colletotrichum.