Other specimen examined: Nigeria, Kpite, on Dioscorea alata leaf, coll. M.M. Abang Cg25, 2001 (ICMP 18122).

Colonies grown from single conidia on Difco PDA 30–40 mm diam after 10 d. Ex-holotype culture looks “stale”, with low, felted, dense, pale grey aerial mycelium, orange agar surface showing through near the margin, scattered dark based acervuli with orange conidial masses near centre; in reverse deep pinkish orange with patches of grey pigment near centre. ICMP 18122 with aerial mycelium sparse, colony surface with numerous discrete, dark-based acervuli with bright orange conidial ooze, margin of colony feathery; in reverse irregular sectors with pale grey pigment within the grey, otherwise colourless apart from the colour of the acervuli and conidial masses. Conidia (14.5−)18–19.5(−23.5) × (4.5−)5−5.5(−6.5) μm (av. 18.9 × 5.2 μm, n = 40), cylindric, straight, ends rounded, a few tapering towards the basal end. Appressoria mostly simple, elliptic to fusoid in shape, sometime developing broad, irregular lobes, about 7–13.5 × 5–10.5 μm. perithecia not seen.

Geographic distribution and host range: Known only from yam (Dioscorea alata), from Nigeria, Barbardos, India, Guadeloupe.

Genetic identification: ITS sequences distinguish C. alatae from all other taxa. Notes: Anthracnose diseases of yam are found throughout the regions where the host is grown (e.g. Winch et al. 1984, Prasad & Singh 1960, Singh et al. 1966, Abang et al. 2002, 2003). Isolates from diseased yam leaves are morphologically (Winch et al. 1984) and genetically (Abang et al. 2002) diverse. Both of these authors used a broad species concept, grouping all isolates sourced from yam under the single name C. gloeosporioides. In this paper we accept part of that diversity to represent a distinct species, newly described here as C. alatae. The type specimen of C. alatae matches the SGG (slow growing grey) group of Abang et al. (2002), the group that these authors found to be more pathogenic to yam than the other morphological and genetic groups they recognised within C. gloeosporioides. In addition to the Nigerian isolates of Abang et al. (2002), isolates from yam from Barbados (isolates SAS8 and SAS9 from Sreenivasaprasad et al. 1996), Guadeloupe (GenBank accession GQ495617) and India (CBS 304.67 and GenBank accession FJ940734) belong in this clade, while no isolates from other hosts have been found. Other isolates from yam that we sequenced included some representing the Abang et al. (2002) FGS group (Abang Cg22 = ICMP 18120, Abang Cg13 = ICMP 18125, Abang CgS6 = ICMP 18117, Abang CgS2 = ICMP 18121), a group distinguished from the highly pathogenic SGG isolates by faster growth in culture and shorter conidia (Abang et al. 2002). Two of these isolates (ICMP 18120, 18125) genetically match C. fructicola, the others match C. siamense. Several names have been applied to Colletotrichum specimens from anthracnose of yam stems and leaves, including Gloeosporium pestis Massee, G. “dioscoreae” Sawada (nom. inval.; no Latin diagnosis), Colletotrichum dioscoreae Av.-Saccá 1917, and C. dioscoreae Tehon 1933. In addition, Gloeosporium bomplandii Speg. was described from a host doubtfully identified as Dioscorea. Because of the broad genetic diversity of Colletotrichum spp. associated with diseased yam, the lack of cultures from any of these early type specimens, and the uncertainty to which part of the yam-associated diversity they correspond, we have chosen not to use these names for our newly recognised, yam-specialised pathogen. Whether the postharvest tuber rot referred to as dead skin disease of yam (Abang et al. 2003, Green & Simmons 1994) is caused by the same Colletotrichum population as associated with diseased foliage is not known.

Holotype: India, Rajasthan, Udaipur, on Dioscorea alata leaves and stems, coll. K.L. Kothari & J. Abramham, 1959, CBS H-6939; extype culture and putatively authentic isolate of C. gloeosporioides f. alatae CBS 304.67 = ICMP 17919.