Phylogenetic analysis

Genomic DNA of the isolates was extracted using the method of Damm et al. (2008). The ITS, GAPDH, and partial sequences of the chitin synthase 1 (CHS-1), histone H3 (HIS3), actin (ACT) and beta-tubulin (TUB2) genes were amplified and sequenced using the primer pairs ITS-1F (Gardes and Bruns 1993) + ITS-4 (White et al. 1990), GDF1 + GDR1 (Guerber et al. 2003), CHS-354R + CHS-79F (Carbone & Kohn 1999), CYLH3F + CYLH3R (Crous et al. 2004b), ACT-512F + ACT-783R (Carbone & Kohn 1999) and T1 (O'Donnell & Cigelnik 1997) + Bt-2b (Glass & Donaldson 1995) or T1 + BT4R (Woudenberg et al. 2009), respectively. The PCRs were performed in a 2720 Thermal Cycler (Applied Biosystems, Foster City, California) in a total volume of 12.5 μL. The GAPDH, CHS-1, HIS3, ACT and TUB2 PCR mixture contained 1 μL 20× diluted genomic DNA, 0.2 μM of each primer, 1× PCR buffer (Bioline, Luckenwalde, Germany), 2 mM MgCl₂, 20 μM of each dNTP, 0.7 μL DMSO and 0.25 U Taq DNA polymerase (Bioline). Conditions for PCR of these genes constituted an initial denaturation step of 5 min at 94 °C, followed by 40 cycles of 30 s at 94 °C, 30 s at 52 °C and 30 s at 72 °C, and a final denaturation step of 7 min at 72 °C, while the ITS PCR was performed as described by Woudenberg et al. (2009). The DNA sequences generated with forward and reverse primers were used to obtain consensus sequences using Bionumerics v. 4.60 (Applied Maths, St-Marthens-Lathem, Belgium), and the alignment assembled and manually adjusted using Sequence Alignment Editor v. 2.0a11 (Rambaut 2002).

To determine whether the six sequence datasets were congruent and combinable, tree topologies of 70 % reciprocal Neighbour-Joining bootstrap with Maximum Likelihood distances (10 000 replicates) with substitution models determined separately for each partition using MrModeltest v. 2.3 (Nylander 2004) were compared visually (Mason-Gamer and Kellogg 1996). A maximum parsimony analysis was performed on the multilocus alignment (ITS, GAPDH, CHS-1, HIS3, ACT, TUB2) as well as for each gene separately with PAUP (Phylogenetic Analysis Using Parsimony) v. 4.0b10 (Swofford 2003) using the heuristic search option with 100 random sequence additions and tree bisection and reconstruction (TBR) as the branch-swapping algorithm. Alignment gaps were treated as new states and all characters were unordered and of equal weight. The robustness of the trees obtained was evaluated by 1 000 bootstrap replications using the same settings as for the parsimony analysis itself (Hillis & Bull 1993). Tree length, consistency index (CI), retention index (RI), rescaled consistency index (RC) and homoplasy index (HI) were calculated for the resulting trees. A Markov Chain Monte Carlo (MCMC) algorithm was used to generate phylogenetic trees with Bayesian probabilities using MrBayes v. 3.2.2 (Ronquist & Huelsenbeck 2003) for the combined sequence datasets. Models of nucleotide substitution for each gene determined by the AIC criterion as implemented in MrModeltest v. 2.3 were included for each gene partition. The analyses of two parallel Markov Chain Monte Carlo (MCMC) runs, each consisting of four chains, were run from random trees for 100 M generations and sampled every 1 000 generations until the runs converged with a split frequency of 0.01. The first 25 % of trees were discarded as the burn-in phase of the analysis and posterior probabilities determined from the remaining trees. For additional comparison, a Neighbour-Joining analysis was performed on the multilocus alignment using PAUP with 10 000 bootstrap replications. Sequences derived in this study have been lodged at GenBank, the alignment and tree in TreeBASE (www.treebase.org/treebase-web/home.html) (S16069), and taxonomic novelties in MycoBank (Crous et al. 2004a).

Taxonomy

Based on DNA sequence data and morphology, the 83 isolates studied (Table 1) are assigned to 16 species, including eight species that are considered to be new to science. All species studied in culture are characterised below.

Colletotrichum americae-borealis Damm, sp. nov. MycoBank MB809398. Fig. 2.

Open in a separate window

Fig. 2

Colletotrichum americae-borealis (A–N, U–V from ex-holotype strain CBS 136232. O–T from strain CBS 136855). A–B. Conidiomata. C, I. Tip of a seta. D, J. Base of a seta. E–H, K–N. Conidiophores. O–T. Appressoria. U–V. Conidia. A, C–H, U. from Anthriscus stem. B, I–T, V. from SNA. A–B. DM, C–V. DIC, Scale bars: A = 200 μm, G = 10 μm. Scale bar of A applies to A–B. Scale bar of G applies to C–V.

Etymology: The species epithet is derived from the region where the species was collected, North America.

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1–7.5 μm diam, hyaline, smooth-walled, septate, branched. Chlamydospores not observed. Conidiomata absent, conidiophores formed directly on hyphae or on a cushion of pale brown, angular cells, 3–6.5 μm diam. Setae medium brown, smooth-walled to finely verruculose, 55–230 μm long, 1–4-septate, base cylindrical to conical, 2.5–7.5 μm diam, tip ± acute to ± rounded. Conidiophores hyaline to pale brown, smooth-walled, septate, branched, to 40 μm long. Conidiogenous cells hyaline to pale brown, smooth-walled, cylindrical to ampulliform, sometimes intercalary (necks not separated from hyphae by septum), 9.5–24.5 × 3.5–5 μm, opening 1–1.5 μm diam, collarette 0.5–1 μm long, periclinal thickening observed. Conidia hyaline, smooth-walled, aseptate, cylindrical to fusoid, straight to slightly curved, both ends rounded, (13.5–)15.5–18(–19) × 3.5–4 μm, av. ± SD = 16.6 ± 1.3 × 3.7 ± 0.2 μm, L/W ratio = 4.5, conidia of strain ATCC 11869 shorter, measuring (9.5–)11.5–15.5(–17.5) × (3–)3.5–4(–4.5) μm, av. ± SD = 13.5 ± 2.2 × 3.8 ± 0.4 μm, L/W ratio = 3.5. Appressoria not observed, appressoria of strain CBS 136855 single or in loose groups, medium to dark brown, smooth-walled, ellipsoid, clavate or irregular outline, with an undulate to lobate margin, (4.5–)6–10.5(–13) × (3.5–)4–7(–10) μm, av. ± SD = 8.1 ± 2.2 × 5.4 ± 1.5 μm, L/W ratio = 1.5.

Asexual morph on Anthriscus stem. Conidiomata, conidiophores and setae formed on pale brown, angular cells, 4–7 μm diam. Setae medium brown, smooth-walled, 8–250 μm long, 1–6-septate, base cylindrical to conical, 5–10 μm diam, tip ± acute to ± rounded. Conidiophores hyaline to pale brown, smooth-walled, simple or septate and branched, to 30 μm long. Conidiogenous cells hyaline to pale brown, smooth-walled, cylindrical to ampulliform, 8.5–19 × 3.5–5.5 μm, opening 1–2 μm diam, collarette 0.5–1 μm long, periclinal thickening distinct. Conidia hyaline, smooth-walled, aseptate, cylindrical to fusoid, straight to slightly curved, both ends rounded, (14.5–)16.5–18.5(–19.5) × (3–)3.5(–4) μm, av. ± SD = 17.4 ± 1.0 × 3.5 ± 0.2 μm, L/W ratio = 5.0, conidia of strain ATCC 11869 shorter, measuring (8–)13–18(–19.5) × 3–4 μm, av. ± SD = 15.5 ± 2.4 × 3.8 ± 0.3 μm, L/W ratio = 4.5.

Culture characteristics: Colonies on SNA flat with entire margin, hyaline, pale cinnamon in the centre, agar medium, filter paper and Anthriscus stem partly covered with saffron to dark grey acervuli, medium and filter paper partly covered with sparse, whitish aerial mycelium, reverse same colours; growth 21.5–25 mm in 7 d (35–37 mm in 10 d). Colonies on OA flat with entire margin; buff, rosy buff to saffron, towards the centre saffron to dark grey acervuli, aerial mycelium lacking, reverse buff to rosy buff, growth 22.5–25 mm in 7 d (33.5–36 mm in 10 d). Conidia in mass saffron.

Materials examined: USA, Utah, Bluffdale (near Salt Lake City), from stems of Medicago sativa, 25 Aug. 2013, U. Damm (CBS H-21661 holotype, culture ex-holotype CBS 136232); Utah, Bluffdale (near Salt Lake City), from stems of Medicago sativa, 25 Aug. 2013, U. Damm, culture CBS 136855; Iowa, from Medicago sativa, collection date and collector unknown, (received from R. O'Connell, before from F. Uruburu, deposited in ATCC collection by L.H. Tiffany) culture ATCC 11869 = CPC 18946 = LARS 373.

Notes: The conidial shape of C. americae-borealis is similar to that of C. lini, but more complex appressoria were observed. In contrast to most of the other species in this complex, setae were very abundant. Several species have been described from Trifolium and Medicago that are discussed under C. destructivum.

The ITS and GAPDH sequences of C. americae-borealis are the same as those of C. lini. This species can be distinguished from other species in this complex by TUB2, CHS-1, HIS3 and ACT sequences.

Strain ATCC 11869 shows additional differences in CHS-1, HIS3 and ACT sequences to strains CBS 136232 and CBS 136855, the other two strains of this species studied. We prefer to treat this strain as C. americae-borealis for the present, because it has the same host and origin as the other two strains. This strain was hardly sporulating; appressoria resembled those of strains CBS 136232 and CBS 136855. Strain ATCC 11869 was deposited in the ATCC collection by L.H. Tiffany, and apparently belongs to the large collection of Colletotrichum isolates from legumes studied by Tiffany & Gilman (1954). It would be interesting to include more isolates related to ATCC 11869 in a future study to determine whether ATCC 11869 and additional isolates might form a distinct clade or reveal morphological or biological differences to the ex-type strain of C. americae-borealis.

The closest match in a blastn search with the TUB2 sequence of strain CBS 136232 was with 99 % identity (1 nucleotide difference) C. linicola (= C. lini) strain CBS 172.51 (GenBank JQ005849, O'Connell et al. 2012), which is included in this study. Blastn searches with the ITS sequence of strain CBS 136232 resulted in 100 % matches with sequences of C. destructivum (s. lat.) strains 1212, MP11 (GenBank KF181248, KF181247, Z. Wen & Z. Nan, an unpublished study on alfalfa root rot in Gansu, China), DAOM 179749 from an unknown host (GenBank EU400143, Chen et al. 2007) and strain Hamedan from clover in Iran (GenBank FJ185789, Zafari & Tarrah 2009), C. linicola (= C. lini) strain CBS 172.51 (GenBank JQ005765, O'Connell et al. 2012 and GenBank AB046609, Moriwaki et al. 2002), a C. linicola (= C. lini) isolate from Convolvulus in Turkey (GenBank EU000060, Tunali et al. 2008), unidentified fungus strains DJJ15 and DY20 from Oxytropis (GenBank JF461333, JF461335, J. Wang, unpubl. study), C. higginsianum strain IMI 391904 from Raphanus in Tunisia (GenBank JX499034, Naumann & Wicklow 2013) that is included in this study and re-identified as C. lini, Colletotrichum sp. isolates 2002 from Holcus (GenBank FN386304, Sánchez Márquez et al. 2012) and 842 and 865 from Arabidopsis (GenBank JX982460, JX982461, Garcia et al. 2013), both the latter reports concerning endophytes isolated in Spain.

Colletotrichum antirrhinicola Damm, sp. nov. MycoBank MB809399. Fig. 3.

Open in a separate window

Fig. 3

Colletotrichum antirrhinicola (from ex-holotype strain CBS 102189). A–B. Conidiomata. C, H. Tip of a seta. D, I. Base of a seta. E–G, J–M. Conidiophores. N–S. Appressoria. T–U. Conidia. A, C–G, T. from Anthriscus stem. B, H–S, U. from SNA. A–B. DM, C–U. DIC, Scale bars: A = 100 μm, E = 10 μm. Scale bar of A applies to A–B. Scale bar of E applies to C–U.

Etymology: The species epithet is derived from its host plant Antirrhinum.

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1–10 μm diam, hyaline, some are pale to medium brown, smooth-walled, septate, branched. Chlamydospores not observed. Conidiomata absent, conidiophores and setae formed directly on hyphae. Setae pale to medium brown, verruculose, 30–100 μm long, 1–4-septate, base cylindrical, conical to ± inflated, 5.5–6.5 μm diam, tip rounded. Conidiophores hyaline, smooth-walled, septate, branched, to 35 μm long. Conidiogenous cells hyaline, smooth-walled, cylindrical to ampulliform, sometimes intercalary (necks not separated from hyphae by septum), polyphialides observed, 8.5–25 × 3.5–5.5 μm, opening 1.5–2 μm diam, collarette 1 μm long, periclinal thickening distinct. Conidia hyaline, smooth-walled, aseptate, cylindrical, straight to slightly curved, with one end round and the other truncate, (14.5–)15.5–19(–23.5) × (3.5–)4–4.5(–5) μm, av. ± SD = 17.2 ± 1.7 × 4.3 ± 0.3 μm, L/W ratio = 4.0. Appressoria single, medium brown, smooth-walled, subglobose, ovate to broadly elliptical in outline, with an entire or undulate margin, (9–)9.5–12(–13.5) × (5–)6–8(–10) μm, av. ± SD = 10.9 ± 1.3 × 7.0 ± 1.0 μm, L/W ratio = 1.5.

Asexual morph on Anthriscus stem. Conidiomata, conidiophores and setae formed on pale brown, angular cells, 3–8 μm diam. Setae pale to medium brown, smooth-walled, 35–170 μm long, 1–3-septate, base cylindrical to conical, 5.5–6 μm diam, tip ± acute to ± rounded. Conidiophores hyaline, smooth-walled, septate, branched, to 25 μm long. Conidiogenous cells hyaline, smooth-walled, cylindrical to ampulliform, 11–15 × 4–5.5 μm, opening 1–1.5 μm diam, collarette 1 μm long, periclinal thickening distinct. Conidia hyaline, smooth-walled, aseptate, cylindrical, straight to slightly curved, with one end round and the other truncate, (13–)15.5–19(–20) × (3.5–)4–4.5(–5) μm, av. ± SD = 17.3 ± 1.6 × 4.2 ± 0.4 μm, L/W ratio = 4.1.

Culture characteristics: Colonies on SNA flat with entire margin, hyaline to pale rosy-buff, filter paper and agar medium in centre partly grey, agar medium, filter paper and Anthriscus stem partly covered with white aerial mycelium, reverse same colours; growth 21.5–23 mm in 7 d (33–34.5 mm in 10 d). Colonies on OA flat with entire margin; buff, partly covered with black acervuli and salmon conidial masses, aerial mycelium lacking, reverse buff, rosy-buff to pale olivaceous-grey, growth 20.5–22 mm in 7 d (30–31.5 mm in 10 d). Conidia in mass salmon.

Material examined: New Zealand, Auckland, Kingsland, from foliage of Antirrhinum majus, collection date unknown (deposited in CBS collection Sep. 1999 by C.F. Hill, isolated 22 Jul. 1999 by H.M. Dance, Agriquality N2, No. 017), HM Dance (CBS H-21647 holotype, culture ex-holotype CBS 102189).

Notes: Colletotrichum antirrhinicola is only known from snapdragon (Antirrhinum majus, Scrophulariaceae) in New Zealand. The species can be identified by its unique GAPDH and ITS sequences. The HIS3 sequence is the same as that of C. fuscum, while the ACT sequence is identical with C. fuscum and C. bryoniicola. Closest match in a blastn search with the ITS sequence of CBS 102189 with 99 % identity (1 nucleotide difference) was C. fuscum strain DAOM 216112 from an unknown host (GenBank EU400144, Chen et al. 2007), while the most similar GAPDH sequences on NCBI GenBank are 96 % identical to that of CBS 102189. In blastn searches, ACT and HIS3 sequences of CBS 102189 are identical with GenBank JQ005825 and GenBank JQ005804, respectively from C. fuscum CBS 130.57 (O'Connell et al. 2012) that are included here.

Tomioka et al. (2011) reported C. destructivum to cause a severe anthracnose disease on leaves of A. majus in Japan. ITS sequences of two strains (MAFF 239947, MAFF 239948) are available in NCBI GenBank (GenBank AB334521, AB334522); additionally, ACT, EF1-α, GAPDH, ITS and TUB2 sequences are available on NIAS GenBank. However, none of these sequences agree with those of strain CBS 102189 (95–99 % identity); the strains from Japan therefore probably represent a different species, most likely in the same species complex.

Stewart (1900a, b) reported a new anthracnose disease of cultivated snapdragon in the USA as C. antirrhini. The description by Stewart (1900b) indicates the species may belong to the C. destructivum complex with conidia measuring 16–21 × 4 μm. The species was regarded as synonym of C. gloeosporioides by von Arx (1957), and is listed as a synonym of C. coccodes in Index Fungorum (www.indexfungorum.org, retrieved 20 Aug. 2014). However, as there are apparently several Colletotrichum species causing anthracnose on this host we refrain from epitypifying this species with a strain from New Zealand instead of the USA, and rather describe it here as a new species.

Strain CBS 102189 was previously identified as C. fuscum by C.F. Hill. A strain from the same host and country (IMI 197877), also identified by C.F. Hill but apparently much earlier, as well as collections from UK were listed by Sutton (1980) as C. fuscum, which is closely related to C. antirrhinicola.

Colletotrichum bryoniicola Damm, sp. nov. MycoBank MB809400. Fig. 4.

Open in a separate window

Fig. 4

Colletotrichum bryoniicola (from ex-holotype strain CBS 109849). A–B. Conidiomata. C–I. Conidiophores. J–O. Appressorium-like structures. P–Q. Conidia. A, C–F, P. from Anthriscus stem. B, G–O, Q. from SNA. A–B. DM, C–Q. DIC, Scale bars: A = 100 μm, C = 10 μm. Scale bar of A applies to A–B. Scale bar of C applies to C–Q.

Etymology: The species epithet is derived from its host plant Bryonia.

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1–10.5 μm diam, hyaline, smooth-walled, septate, branched. Chlamydospores not observed. Conidiomata absent, conidiophores formed directly on hyphae. Additionally, structures formed resembling the basal cushions of acervuli, but lacking conidiophores, cells angular to roundish, pale brown, 3.5–10 μm diam. Setae not observed. Conidiophores hyaline to pale brown, smooth-walled, sometimes septate and branched, to 15 μm long. Conidiogenous cells rarely observed, hyaline to pale brown, smooth-walled, cylindrical to conical, 5.5–10 × 3.5–4 μm, opening 1–1.5 μm diam, collarette 0.5 μm long, periclinal thickening not observed. Conidia hyaline, smooth-walled, aseptate, cylindrical, straight to slightly curved, with one end round and the other truncate, (13.5–)15–18.5(–22) × 4–5(–5.5) μm, av. ± SD = 16.8 ± 1.6 × 4.6 ± 0.4 μm, L/W ratio = 3.6. Appressoria not observed on the undersurface of the medium. Appressoria-like structures that possibly function as chlamydospores were observed within the medium. These are single or in loose groups, pale brown, smooth-walled, subglobose to elliptical in outline, with an entire or slightly undulate margin, (3.5–)4–10(–18) × (2.5–)3.5–6.5(–7.5) μm, av. ± SD = 7.1 ± 3.0 × 4.9 ± 1.5 μm, L/W ratio = 1.4.

Asexual morph on Anthriscus stem. Conidiomata, conidiophores formed on pale brown, angular cells, 3–9 μm diam. Setae not observed. Conidiophores rarely seen, pale brown, smooth-walled. Conidiogenous cells hyaline to pale brown, smooth-walled, doliiform, ampulliform to cylindrical, 7–13.5 × 3–5 μm, opening 1–1.5 μm diam, collarette 0.5 μm long, periclinal thickening not observed. Conidia hyaline, smooth-walled, aseptate, cylindrical, straight to slightly curved, with one end round and the other truncate, (16–)17–19.5(–21) × 4.5(–5) μm, av. ± SD = 18.2 ± 1.1 × 4.5 ± 0.1 μm, L/W ratio = 4.0.

Culture characteristics: Colonies on SNA flat with entire margin, hyaline, agar medium, filter paper and Anthriscus stem partly covered with white aerial mycelium, reverse same colours; growth 20.5–21.5 mm in 7 d (29.5–31.5 mm in 10 d). Colonies on OA flat with entire margin; buff, with cinnamon, olivaceous-grey to iron-grey sectors, aerial mycelium lacking, reverse same colours, growth 17.5–18.5 mm in 7 d (27.5–29 mm in 10 d). Conidia in mass whitish to pale salmon.

Material examined: Netherlands, Wissenkerke, Camperduin, coord. 35.5/401.6, from decaying leaves of Bryonia dioica, 27 Aug. 2001, G. Verkley, No. V1114 (CBS H-21663 holotype, culture ex-holotype CBS 109849).

Notes: Colletotrichum bryoniicola differs from closely related species in ITS, GAPDH, HIS3 and TUB2 sequences by a single nucleotide in each locus. The ACT sequence is the same as that of C. fuscum and C. antirrhinicola, the CHS-1 sequence is identical with that of C. tanaceti. There are no previously accessioned sequences of a Colletotrichum species from Bryonia in GenBank. With the exception of the ACT and CHS-1 sequences, there are no sequences in GenBank that are identical to those of C. bryoniicola.

Conidia of C. bryoniicola are broader (≥ 4 μm on SNA, ≥ 4.5 μm on Anthriscus stem) than the other species in the C. destructivum complex, no setae were observed and the conidiogenous cells are very indistinct.

A species from Bryonia dioica (Cucurbitaceae) was previously described ad interim by Maire (1917), as C. bryoniae. Although Maire (1917) did not mention any connection of the new species from Alger (= Algiers), Mauretania (today Algeria) with C. oligochaetum f. bryoniae Ferraris from B. dioica in Italy (Ferraris & Massa 1912), that taxon was accepted as an independent species by Saccardo et al. (1931) and cited incorrectly as C. bryoniae (Ferraris) Maire (1917). As it is based on a forma of C. oligochaetum that was considered as a synonym of C. orbiculare by von Arx (1957), C. bryoniae was regarded as a synonym of C. orbiculare as well. The conidial size is similar to the strain from the Netherlands, measuring 18–22 × 4–5 μm (Maire 1917). However, as there are often several Colletotrichum species within this species complex causing anthracnose on the same host plants and we have no proof that this species belongs to this complex, we refrain from epitypifying this species with a strain from the Netherlands instead of Algeria, and rather describe it here as a new species. Material of the species from Algeria has not been examined and living cultures derived from its type are not available.

Colletotrichum destructivum O'Gara, Mycologia 7: 38. 1915. Fig. 5.

Open in a separate window

Fig. 5

Colletotrichum destructivum (from ex-epitype strain CBS 136228). A–B. Conidiomata. C, I. Tip of a seta. D, J. Base of a seta. E–H, K–N. Conidiophores. O–T. Appressoria. U–V. Conidia. A, C–H, U. from Anthriscus stem. B, I–T, V. from SNA. A–B. DM, C–V. DIC, Scale bars: A = 100 μm, E = 10 μm. Scale bar of A applies to A–B. Scale bar of E applies to C–V.

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1–9 μm diam, hyaline, smooth-walled, septate, branched. Chlamydospores not observed. Conidiomata absent, conidiophores and setae formed directly on hyphae or on pale to medium brown, angular cells, 3–10 μm diam, sometimes developing to dark brown, round structures on which many setae and conidiophore-like structures are formed, conidia released as well, however, most conidiophore-like structures without visible conidiogenous openings, thick-walled, septate, branched at the base, up to 70 μm long, very broad and usually broadest at the tip, cells at the tip measure 6.5–22 × 4–6 μm, surrounded by a slime sheath. Setae medium brown, smooth-walled to finely verruculose, sometimes verrucose, 50–180 μm long, 1–3-septate, base cylindrical, conical, sometimes ± inflated, 3.5–6 μm diam, tip ± rounded to ± acute. Conidiophores pale to medium brown, smooth-walled, septate, branched, to 85 μm long. Conidiogenous cells pale to medium brown, smooth-walled, elongate-ampulliform to cylindrical, 9.5–17 × 3.5–5 μm, opening 1–1.5 μm diam, collarette 0.5–1 μm long, periclinal thickening visible. Conidia hyaline, smooth-walled, aseptate, cylindrical, straight to slightly curved, with both ends ± rounded, (14–)14.5–16.5(–18) × 3.5–4(–4.5) μm, av. ± SD = 15.4 ± 0.8 × 3.7 ± 0.2 μm, L/W ratio = 4.2. Appressoria single, pale brown, smooth-walled, clavate, fusiform to ellipsoidal outline, with a lobate, undulate or crenate margin, (6.5–)10–15.5(–20.5) × (4.5–)5–8(–10.5) μm, av. ± SD = 12.5 ± 2.7 × 6.7 ± 1.5 μm, L/W ratio = 1.9, Appressoria of strain CBS 149.34 smaller, measuring (4–)6–14(–25) × (3.5–)4.5–7.5(–10) μm, av. ± SD = 9.8 ± 4.1 × 5.9 ± 1.5 μm, L/W ratio = 1.7, strain CBS 149.34 also forms appressorium-like structures inside the medium, single, medium brown, smooth-walled, subglobose, ovate to broadly elliptical in outline, with an entire or undulate margin, (3.5–)5–10(–13) × (3–)3.5–7(–8.5) μm, av. ± SD = 7.6 ± 2.5 × 5.2 ± 1.7 μm, L/W ratio = 1.4.

Asexual morph on Anthriscus stem. Conidiomata, conidiophores and setae formed on a cushion of pale to medium brown, angular cells, 3–8.5 μm diam that are intermingled and surrounded by medium brown, thick-walled hyphae, with ± inflated cells, up to 8.5 μm diam. Setae medium brown, smooth-walled, towards the tip often verruculose to verrucose, constricted and slightly wavy, 65–110 μm long, 1–4-septate, base conical to ± inflated, 4.5–12 μm diam, tip ± rounded to ± acute. Conidiophores hyaline to pale brown, smooth-walled, simple or septate and branched, to 25 μm long. Conidiogenous cells hyaline to pale brown, smooth-walled, cylindrical, doliiform to ampulliform, 5–14 × 2.5–5.5 μm, opening 1–2 μm diam, collarette 0.5–1.5 μm long, periclinal thickening distinct. Conidia hyaline, smooth-walled, aseptate, cylindrical, straight to slightly curved, with both ends ± rounded, (15–)16–18(–19) × (3–)3.5–4 μm, av. ± SD = 16.9 ± 1.0 × 3.6 ± 0.2 μm, L/W ratio = 4.7.

Culture characteristics: Colonies on SNA flat with entire margin, hyaline to honey, agar medium and filter paper partly covered by sparse filty white aerial mycelium; agar medium, filter paper and Anthriscus stem partly covered with grey to black acervuli, reverse same colours; growth 23–25 mm in 7 d (36–37.5 mm in 10 d). Colonies on OA flat with entire margin; buff to honey, almost entirely covered with grey to black acervuli; partly covered with whitish to grey aerial mycelium, reverse buff to olivaceous-grey, growth 25–27.5 mm in 7 d (37.5–40 mm in 10 d). Conidia in mass whitish to rosy-buff.

Materials examined: Canada, south-western Ontario, from anthracnose on stems of Medicago sativa, 1985/86 collector unknown (deposited by R. O'Connell, before from G.J. Boland, A 22), culture CBS 128509 = LARS 320. Korea, from Rumex sp., collection date and collector unknown (CBS H-21655, culture ex IMI 387103 = CPC 18082). Netherlands, experimental plot of P.M.L.Tammes, from stem burn of Trifolium sp., 1 Oct. 1934, P.M.L.Tammes, culture CBS 149.34. USA, Utah, probably Salt Lake City, on stems and leaves of Trifolium pratense, 30 Jun. 1914, P. J. O'Gara (BPI 397373 holotype); Utah, Salt Lake City, cemetery, from small black spots on petioles of T. hybridum, 24 Aug. 2013, U. Damm (CBS H-21652 epitype, here designated, MBT178515, culture ex-epitype CBS 136228); Utah, Syracuse (close to Salt Lake City), pasture, from small black spots on petioles of T. hybridum, 24 Aug. 2013, U. Damm, CBS H-21653, culture CBS 136229; from Phragmites sp., collection date and collector unknown, CBS H-21654, culture CBS 130238.

Notes: Colletotrichum destructivum was described by O'Gara (1915) from stems and petioles of red clover (Trifolium pratense) and alsike clover (T. hybridum) in clover fields in the Salt Lake Valley, Utah, USA. The species forms minute acervuli, 25–70 μm diam, hyaline conidia with 1–4 guttules that are straight to slightly curved with rounded apices and bases, measuring 14–22 × 3.5–5 μm, few to numerous setae that are straight, curved to flexuous, often nodulose, aseptate to obscure 1-septate, subacute to rounded, constricted at the apex, 38–205 μm long and 4.5–7 μm diam at the basis (O'Gara 1915). Conidia from small acervuli were observed on stems, petioles and leaves of the holotype specimen (BPI 397373) and measure (14–)15–19.5(–23.5) × (3–)4–4.5 μm, av. ± SD = 17.2 ± 2.1 × 3.8 ± 0.4 μm, L/W ratio = 4.5; setae were 50–110 μm long with a cylindrical to ± inflated base, 3.5–8 μm diam and a ± rounded tip.

The specimen BPI 397373 was collected on 30 June 1914 by P.J. O'Gara and designated no. 20. The package has a type stamp and origins from Fungi Utahensis, Herbarium of Department of Agricultural Investigation, American Smelting & Refining Co., Salt Lake City, Utah, which was the institute where P.J. O'Gara used to work as stated in the publication. There is an isotype of this fungus in herbarium NY with the information “Incorporated herbarium: Garrett Herbarium, University of Utah”. There is no specimen of C. destructivum available in the Garrett Herbarium any more; the specimen was apparently sent away together with many specimens from that herbarium (M. Power, in lit.). This agrees with a note on the specimen packet stating “Rec. by Path. Coll. Apr. 8, 1921”; BPI 397373 must therefore be the holotype.

In order to epitypify C. destructivum, collections of Trifolium hybridum and Medicago sativa were made in August 2014 from field as well as urban locations in and around Salt Lake City; T. pratense was not found in the area. Colletotrichum spp. were isolated from small black spots on stems, petioles and leaves of both host plants. Isolates of the C. destructivum species complex were identified based on morphology. Some of the isolates grouped with a species that had often been collected from both host genera worldwide, for which the name C. destructivum is usually applied. The other isolates belonged to C. lini or a species closely related to C. lini. Colletotrichum lini contained both Trifolium and Medicago isolates, the other clade only Medicago isolates (see C. americae-borealis). It is possible that O'Gara (1915) collected more than one species as well. However, there is not enough material available of the holotype to extract DNA for molecular examination. The two species collected from clover are morphologically very similar. However, setae of the ex-epitype strain CBS 136228 grown on Anthriscus stem were often constricted towards the tip, which is in accordance with the observations made by O'Gara (1915). This was not observed in C. lini.

Several Colletotrichum and Gloeosporium species have been described from Trifolium and Medicago as already discussed in Damm et al. (2013). Except for Gm. trifolii and Gm. medicaginis, these species were described later than C. destructivum and cannot be considered as possible synonyms of C. destructivum. Except for C. destructivum and C. trifolii, these names have not been used since their description. Colletotrichum trifolii was originally described from T. pratense in the USA (Bain & Essary 1906); the species was epitypified recently and revealed to belong to the C. orbiculare species complex (Damm et al. 2013) that is distinct from the species complex studied here (Cannon et al. 2012). Strain CBS 149.34 was previously identified as C. trifolii (Nirenberg et al. 2002), but re-identified here as C. destructivum.

Gloeosporium trifolii described from T. pratense in Albany, NY, USA (Peck 1879, publ. 1883) forms conidia that measure 15–23 × 4–6.3 μm, which are slightly larger than C. destructivum. Gloeosporium medicaginis forms acervuli on Medicago sativa in Kansas, USA, with cylindrical conidia that are subhyaline and mostly narrowed in the middle, measuring 15–20 × 3–4 μm (Ellis & Kellerman 1887). Conidia that are constricted in the middle have not been observed in this species complex. We have not studied the type specimens of these species. However, even if either of these “forgotten” species belong to the C. destructivum species complex, it would be difficult to link recent collections to one of them based on morphology alone.

Colletotrichum sativum, a species described from Vicia sativa in Louisiana, USA (Horn 1952), was listed as a synonym of C. destructivum by von Arx (1957). We cannot confirm this synonymy as no strain from Vicia sativa belonging to this species complex was studied.

The description and drawing of C. rumicis-crispi from Rumex crispus in Taiwan described by Sawada (1959) are similar to our observations made of C. destructivum that includes a strain from Rumex in Korea (IMI 387103). Colletotrichum rumicis-crispi is probably a synonym of C. destructivum. However, as we have not seen the type specimen, we cannot confirm this.

Strain IMI 387103 differs from the other C. destructivum sequences in ACT and TUB2 sequences in one nucleotide each, but the other loci are identical. In contrast, the ITS sequence of C. destructivum strain RGT-S12 from R. gmelinii from China (GenBank HQ674658, Hu et al. 2012) was 99 % identical (2 nucleotides different) with the ITS sequence of the ex-type strain of C. destructivum (CBS 136228), and strain IMI 387103 from Rumex, but identical with those of C. higginsianum. This indicates the occurrence of at least two Colletotrichum species on Rumex.

Manandhar et al. (1986) claimed C. destructivum to be the asexual morph of Ga. glycines based on morphological comparison of isolates from Glycine that formed a sexual morph resembling Ga. glycines (Lehman & Wolf 1926) and an isolate from Medicago that was initially identified as C. destructivum. The authors apparently assumed that Colletotrichum strains from legumes are all C. destructivum unless the conidia are curved. Isolates from both hosts that were included in the study of Manandhar et al. (1986) were sequenced and confirmed to belong to the same species that is, however, not closely related to C. destructivum and belongs in a different species complex (U. Damm, unpubl. data). Consequently, there is no evidence that C. destructivum forms a sexual morph.

The hemibiotrophic infection of Medicago sativa by C. destructivum was observed by Latunde-Dada et al. (1997). The fungus initially produced large, prominently multilobed infection structures that were localised within single epidermal cells of the infected host. Two of the three isolates studied, LARS 202 (= CBS 511.97) and LARS 709 (= CBS 520.97), were confirmed as C. destructivum s. str. in this study. The third isolate LARS 319 originated from the same collection (A22) from Canada as C. destructivum s. str. strain LARS 320 (= CBS 128509), and is also included here. All isolates originated from a pathogenicity study by Boland & Brochu (1989).

Conidia of C. destructivum are very slightly curved and appear almost straight, similar to those of C. tabacum, especially on SNA; however, no dark halo around the penetration pores of appressoria was observed.

Colletotrichum destructivum can be distinguished by its ITS, HIS3, ACT and TUB2 sequences, while the GAPDH sequence is identical to that of C. ocimi that is described as a new species in this study. Further intraspecific grouping was observed with sequences of all loci studied. Strain IMI 387103 from Rumex in Korea was the most distant strain from the rest of the C. destructivum strains that form a cluster with a bootstrap support value of 84 %. However, we refrained from considering strain IMI 387103 as a separate species as it forms a strong cluster with C. destructivum (95/0.99) and only differs in a few nucleotides from the majority of the C. destructivum strains. In order to investigate if this represents a distinct species, additional collections from Rumex should be studied. The three subclades with a bootstrap support ≤69 % did not show clear host preferences and did not suggest any further splitting of the species.

ITS sequences of a large number of isolates detected in blastn searches were identical to the ITS sequence of the ex-epitype strain (CBS 136228) of C. destructivum: C. destructivum strains CBS 149.34 from Trifolium in the Netherlands (GenBank JQ005764, O'Connell et al. 2012) that is included in this study, Coll-48, Coll-68, Coll-75 and CC 657 from Medicago in Serbia (GenBank JX908362, JX908363, JX908361, Vasić, unpubl. data), MAFF 239947 and MAFF 239948 from Antirrhinum in Japan (GenBank AB334521, AB334522, Tomioka et al. 2011), MAFF 410037 from Robinia in Japan (GenBank AB105961, Moriwaki et al. 2002), CGMCC 3.15129 from Bletilla in China (GenBank JX625174, Tao et al. 2013), uncultured fungus clone CMH309 from house dust in the USA (GenBank KF800400, Rittenour et al. 2013), DAOM 196849 from an unknown host (GenBank EU400156, Chen et al. 2007), C. trifolii isolate UQ349 from Medicago in Australia (GenBank AF451909, Ford et al. 2004) and CBS 149.34 (GenBank AJ301942, Nirenberg et al. 2002) and C. cf. gloeosporioides strain AR 4031 (= CBS 119187) from Crupina in Greece (GenBank AY539806, Berner et al. 2004). Colletotrichum trifolii strain CBS 149.34 and C. cf. gloeosporioides CBS 119187 are included in this study and re-identified as C. destructivum s. str.

The TUB2 sequence of CBS 136228 is identical with GenBank JQ005848 from C. destructivum strain CBS 149.43 (O'Connell et al. 2012), 99 % (1 nucleotide difference) identical with GenBank JX625198 and JX625200 from isolates CGMCC 3.15127 and CGMCC 3.15128 and 99 % identical (3 nucleotides difference) with GenBank JX625203 from isolate CGMCC 3.15129 from Bletilla in China (Tao et al. 2013).

Colletotrichum fuscum Laubert, Gartenwelt 31: 675. 1927. Fig. 6.

Open in a separate window

Fig. 6

Colletotrichum fuscum (from ex-epitype strain CBS 133701). A–B. Conidiomata. C, H. Tip of a seta. D, I. Base of a seta. E–G, J–L. Conidiophores. M–Q. Appressoria. R–S. Conidia. A, C–G, R. from Anthriscus stem. B, H–Q, S. from SNA. A–B. DM, C–S. DIC, Scale bars: A = 100 μm, E = 10 μm. Scale bar of A applies to A–B. Scale bar of E applies to C–S.

= Colletotrichum digitalis Unamuno, Revista Real Acad. Ci. Madrid. 30: 503. 1933 – Nom. illegit., Art. 53.1

≡ Colletotrichum unamunoi Cash, Syll. fung. (Abellini) 26: 1222. 1972.

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1–9.5 μm diam, hyaline to pale brown, smooth-walled, septate, branched. Chlamydospores not observed. Conidiomata absent, conidiophores aggregated directly on pale to medium brown hyphae or on clusters of irregularly arranged medium brown hyphae. Setae (one observed) medium brown, smooth-walled, 71 μm long, 3-septate, base conical, 5.5 μm diam, tip rounded. Conidiophores hyaline to medium brown, smooth-walled, simple or septate and branched, to 20 μm long. Conidiogenous cells hyaline to pale brown, smooth-walled, ampulliform, doliiform to cylindrical, 7–18.5 × 3.5–6 μm, sometimes not separated from hyphae by a septum (intercalary) or opening with collarette formed directly on hyphae, opening 1–1.5 μm diam, collarette 0.5–1 μm long, periclinal thickening visible. Conidia hyaline, smooth-walled, aseptate, cylindrical, slightly curved to straight, with one end round and the other truncate, (16–)16.5–20(–34) × (3.5–)4–4.5(–5.5) μm, av. ± SD = 18.3 ± 1.9 × 4.1 ± 0.3 μm, L/W ratio = 4.5. Appressoria single, medium brown, smooth-walled, roundish, ellipsoidal to clavate in outline, with an lobate (to undulate) margin, (6–)8.5–14.5(–19) × (6.5–)7–10(–11.5) μm, av. ± SD = 11.5 ± 2.8 × 8.6 ± 1.5 μm, L/W ratio = 1.3.

Asexual morph on Anthriscus stem. Conidiomata, conidiophores and setae formed on a cushion of medium brown, angular to roundish cells, 4.5–12.5 μm diam. Setae medium to dark brown, smooth-walled, 3–160 μm long, 1–3-septate, base conical, 5–7.5 μm diam, tip ± rounded to ± acute. Conidiophores hyaline to medium brown, smooth-walled, simple or septate and branched, to 20 μm long. Conidiogenous cells hyaline to medium brown, smooth-walled, ampulliform to conical, 5.5–17.5 × 3–5.5 μm, opening 1–1.5 μm diam, collarette 0.5–1 μm long, periclinal thickening visible. Conidia hyaline, smooth-walled, aseptate, cylindrical, slightly curved to straight, with one end round and the other truncate, (16–)17–19.5(–20.5) × (3.5–)4–4.5(–5) μm, av. ± SD = 18.1 ± 1.4 × 4.1 ± 0.3 μm, L/W ratio = 4.4.

Culture characteristics: Colonies on SNA flat with entire margin, hyaline to pale honey, agar medium, filter paper and Anthriscus stem partly covered with appressed whitish aerial mycelium, Anthriscus stem partly covered with black acervuli, reverse same colours; growth 21–24 mm in 7 d (35–40 mm in 10 d). Colonies on OA flat with entire margin; buff to rosy-buff with small black spots (not clearly recognisable as conidiomata) towards the centre, aerial mycelium lacking, reverse buff to rosy-buff, vinaceous buff towards the centre, growth 21–24 mm in 7 d (31–34 mm in 10 d). Conidia in mass rosy-buff to pale salmon.

Materials examined: Germany, Berlin, Zehlendorf, garden, from leaves of Digitalis purpurea, 1927, R. Laubert [B 70 0021851 (ex BBA acc. 9.1.1980) lectotype, here designated, MBT178720]; Berlin, Zehlendorf, garden, from leaves of Digitalis purpurea, 1927–1933, R. Laubert (B 70 0021852); Berlin, garden, from leaf of Digitalis lutea, 2 Aug. 2012, U. Damm (CBS H-21651 epitype, here designated, MBT178517, culture ex-epitype CBS 133701). Netherlands, Utrecht, garden, from leaf of Digitalis obscura, 29 Aug. 2012, U. Damm, culture CBS 133703; Baarn, garden Eemnesserweg 90, from living leaves of Digitalis purpurea, Nov. 1968, H.A. van der Aa, 944, CBS H-10616, culture CBS 825.68; Utrecht, from dead stem of Heracleum sp., 12 Aug. 2009, U. Damm, CBS H-21666, culture CBS 125336; Utrecht, from dead stem of Heracleum sp., 12 Aug. 2009, U. Damm, CBS H-20404, culture CBS 126510; Utrecht, from dead stem of Heracleum sp., 12 Aug. 2009, U. Damm, CBS H-20405, culture CPC 18076.

Notes: Colletotrichum fuscum causes anthracnose on some Digitalis spp. (foxglove) and was reported from the USA (Connecticut, Maryland, Oregon, Pennsylvania, South Dakota), Poland, Australia, Canada, Germany, England, New Zealand, Portugal and Czechoslovakia (Farr & Rossman 2014). According to Sutton (1980), C. fuscum also attacks Antirrhinum majus in New Zealand and the UK. However, the IMI strain from Antirrhinum majus in New Zealand listed by Sutton (IMI 197877) belongs to a different species (see C. antirrhinicola). Tomioka et al. (2001) showed C. fuscum caused anthracnose of Nemesia strumosa in Japan. Based on morphology (and host), this Japanese collection may belong to the C. destructivum complex, but its identification needs to be confirmed based on molecular data. Thomas (1951) reports serious damage of Digitalis lanata in commercial plantings by C. fuscum.

Laubert (1927) described C. fuscum from diseased leaves of Digitalis purpurea in Berlin with conidia that are 12–24 μm long and 2–4 μm wide, straight or slightly clavate and then slightly curved at the narrow end, formed from short crowded conidiophores, setae 8–10 × 45–100 μm with a slightly inflated base up to 9 μm diam. Two authentic specimens were located in the fungarium B, both without type designation, of which B 70 0021851 collected by R. Laubert in 1927, was selected as lectotype of C. fuscum. The collection date of the second specimen (B 70 0021852) was imprecise (1927–1933); the specimen might have been collected after the publication of Laubert's description. Conidia observed on the lectotype specimen measured (15–)17–21.5(–23) × 3.5–5(–5.5) μm, av. ± SD = 19.3 ± 2.2 × 4.2 ± 0.6 μm, L/W ratio = 4.6 and resembled those seen in culture.

Several other species have been described on Digitalis. Gloeosporium digitalis, which was described from leaves of Digitalis purpurea in Landbohøjskolens Have, Frederiksberg, Denmark, forms smaller conidia than C. fuscum, measuring 8–10 × 3–4 μm and apparently lacks setae (Rostrup 1899). Goto (1938) concluded Gm. digitalis to be a different species. Von Arx (1957) regarded Gm. digitalis as a synonym of Ascochyta digitalis Fuckel.

However, Moesz (1931) combined Gm. digitalis in Colletotrichum on the basis of observations of a fungus on Digitalis ferruginea from Hungary that more closely resembled C. fuscum than Gm. digitalis. Goto (1938) regarded this fungus as a form of C. fuscum, while von Arx (1957) listed this species as a synonym of C. fuscum and called it C. digitalis Moesz. Colletotrichum digitalis could also be a different species based on shape and size of the conidia (cylindrical with blunt ends, measuring 10–15 × 3 μm) and the long conidiophores that are illustrated (Moesz 1931).

Unamuno (1933) described a Colletotrichum species from leaves of Digitalis purpurea in Spain and, apparently unaware of Moesz's combination, called it C. digitalis Unamuno. As this name is illegitimate (Art. 53.1), Trotter & Cash (1972) gave the species a new name, C. unamunoi Cash. Based on the morphological features (conidia 16–22 × 3–3.5 μm, hyaline, cylindrical, usually straight, sometimes slightly curved, rounded at both ends, setae 63 × 3.5–4 μm, brown, septate, straight, curved or flexuous, often nodular; Trotter & Cash 1972), both Goto (1938) and von Arx (1957) regarded C. digitalis Unamuno as a synonym of C. fuscum. We have not seen the type of C. digitalis Unamuno, but agree that this species is most likely a synonym of C. fuscum that seems to be the common anthracnose pathogen of several Digitalis spp., at least in Europe.

Colletotrichum dematium was reported from Digitalis atropurpurea in UK and Scotland (Kirk & Spooner 1984). We do not know whether this report refers to C. dematium s. str.; all species called C. dematium (s. lat.) usually have distinctly curved conidia and are not closely related with C. fuscum (Damm et al. 2009, Cannon et al. 2012).

Goodman (1960) discovered the phytotoxin colletotin in three strains of C. fuscum, one of which was obtained from the CBS collection and called the “von Arx strain”. This strain is probably identical to strain CBS 130.57 that was deposited in the CBS collection in Sep. 1957 by von Arx, listed as forming colletotin with a reference to R.N. Goodman in the CBS strain database, and is included in this study.

Moriwaki et al. (2002) noticed the similarity and close relationship of C. fuscum to C. destructivum and assumed them to be conspecific. Preliminary multilocus phylogenies (O'Connell et al. 2012, Cannon et al. 2012) recently indicated C. fuscum to be a distinct species, which is confirmed in this study.

The complex appressoria and the conidiogenous cells that are often ampulliform on the two media tested are diagnostic for this species. Colletotrichum fuscum is distinguishable by GAPDH, but has only one nucleotide difference from C. bryoniicola. The ITS sequence is variable; isolates do cluster, but one strain (CBS 825.68) sits separately. The CHS-1 sequence is the same as that of C. antirrhinicola, the ACT sequence the same as that of C. antirrhinicola and C. bryoniicola. Additionally, unnamed isolates from Heracleum are basal to C. fuscum, C. antirrhinicola, C. bryoniicola and C. vignae in our phylogeny, and could represent an additional, currently unidentified species (Fig. 1).

The closest matches with the GAPDH sequence of strain CBS 133701, with 98 % identity (3 nucleotides different), are GenBank GU935850 and GU935851 from C. higginsianum isolates C97027 and C97031 (Choi et al. 2011). The closest matches with the ITS sequence, with 99 % identity (2 nucleotides different) were C. fuscum strains CBS 130.57 from Digitalis (GenBank JQ005762, O'Connell et al. 2012), DAOM 216112 (GenBank EU400144, Chen et al. 2007) and BBA 70535 from Digitalis in Germany (GenBank AJ301938, Nirenberg et al. 2002).

Colletotrichum higginsianum Sacc., J. Agric. Res., Washington 10: 161. 1917. Fig. 7.

Open in a separate window

Fig. 7

Colletotrichum higginsianum (from ex-epitype strain IMI 349061). A–B. Conidiomata. C, H. Tip of a seta. D, I. Base of a seta. E–F, J–L. Conidiophores. G. Bases of a seta and conidiophores. M–R. Appressoria. S–T. Conidia. A, C–G, S. from Anthriscus stem. B, H–R, T. from SNA. A–B. DM, C–T. DIC, Scale bars: A = 100 μm, G = 10 μm. Scale bar of A applies to A–B. Scale bar of G applies to C–T.

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1–8.5 μm diam, hyaline, smooth-walled, septate, branched. Chlamydospores not observed. Conidiomata conidiophores and setae on pale brown, angular cells, 3–9 μm diam. Setae medium brown, smooth-walled to finely verruculose, 60–185 μm long, 1–5-septate, base cylindrical to conical, 3.5–6 μm diam, tip rounded to ± acute. Conidiophores hyaline, smooth-walled, septate, branched, to 35 μm long. Conidiogenous cells hyaline, smooth-walled, cylindrical, 8–27 × 3.5–4.5 μm, sometimes intercalary (necks not separated from hyphae by septum), opening 1–2 μm diam, collarette 1–2 μm long, periclinal thickening distinct. Conidia hyaline, smooth-walled, aseptate, cylindrical, straight to very slightly curved, with one end rounded and the other truncate, (17–)19–20.5(–21) × (3–)3.5–4(–4.5) μm, av. ± SD = 19.6 ± 0.9 × 3.7 ± 0.2 μm, L/W ratio = 5.3; conidia of strain IMI 349063 shorter, measuring (13.5–)15–19(–21.5) × 3.5–4(–4.5) μm, av. ± SD = 17.0 ± 1.8 × 3.7 ± 0.3 μm, L/W ratio = 4.6.

Appressoria in loose groups, medium brown, smooth-walled, fusiform, clavate, elliptical or irregular outline, with an entire, crenate or lobate margin, (5.5–)10–20(–28.5) × (3.5–)5–9(–12) μm, av. ± SD = 15.0 ± 5.1 × 6.8 ± 2.0 μm, L/W ratio = 2.2; appressoria of strain MAFF 305635 smaller, measuring (7.5–)9–13(–15) × (3.5–)4.5–6.5(–8) μm, av. ± SD = 11.0 ± 1.9 × 5.4 ± 0.9 μm, L/W ratio = 2.0.

Asexual morph on Anthriscus stem. Conidiomata, conidiophores and setae formed on pale brown, angular cells, 3–7.5 μm diam. Setae medium brown, smooth-walled, 50–170 μm long, 2–5-septate, base cylindrical to conical, 5–12 μm diam, tip ± rounded to ± acute. Conidiophores hyaline to pale brown, smooth-walled, simple or septate and branched, to 15 μm long. Conidiogenous cells hyaline to pale brown, smooth-walled, cylindrical to ampulliform, 8–14 × 3–3.5 μm, opening 1–1.5 μm diam, collarette 0.5–1 μm long, periclinal thickening distinct. Conidia hyaline, smooth-walled, aseptate, cylindrical, straight to very slightly curved, with one end rounded and the other truncate, (17.5–)18–20(–22) × (3–)3.5–4 μm, av. ± SD = 19.0 ± 0.9 × 3.6 ± 0.2 μm, L/W ratio = 5.2; conidia of strain IMI 349063 shorter, measuring (12.5–)15–18(–18.5) × 3.5–4.5 μm, av. ± SD = 16.5 ± 1.7 × 4.0 ± 0.3 μm, L/W ratio = 4.1.

Culture characteristics: Colonies on SNA flat with entire margin, hyaline, agar medium, filter paper and Anthriscus stem partly covered with salmon to grey acervuli and floccose white aerial mycelium, filter paper partly pale luteous to pale orange, reverse same colours; growth 23–24 mm in 7 d (35–37.5 mm in 10 d); with strain IMI 349063 aerial mycelium lacking and filter paper partly pale luteous. Colonies on OA flat with entire margin; buff, saffron to pale orange, partly covered with pale orange, grey to black acervuli and floccose white aerial mycelium, reverse buff, salmon to olivaceous-grey, growth 23.5–29 mm in 7 d (36–39.5 mm in 10 d); with strain IMI 349063 aerial mycelium lacking and slower growth 20.5–22.5 mm in 7 d (30–33.5 mm in 10 d). Conidia in mass saffron to orange.

Materials examined: Japan, Edogawa, Tokyo, from Brassica rapa var. komatsuna, collection date and collector unknown (isolated 6 Oct. 1980 by H. Horie), culture MAFF 305635 = Abr 1-5 = CPC 19366, CPC 18943; Edogawa, Tokyo, from Raphanus sativus, collection date and collector unknown (isolated 21 Oct. 1980), culture AR 3-1 = CPC 19394; Tateyama, Chiba, from Matthiola incana, collection date and collector unknown (isolated Oct. 1990), CBS H-21665, culture CH90-M1 = CPC 19361; Chikura, Chiba from Matthiola incana, collection date and collector unknown (isolated Oct. 1990), culture CH93-M1 = CPC 19362. Romania, Târgu Neam, garden near Văratec, on leaves of Matthiola incana, 20 Jul. 1952, C. Sandu-Ville, (GLM-F102751 holotype of C. mathiolae Sandu ex Herbarul Micologic “C. Sandu-Ville”). Trinidad and Tobago, Trinidad, Wallerfield, from leaf spot on living leaf of Brassica rapa subsp. chinensis, collection date and collector unknown (IMI 349061 epitype of C. higginsianum, here designated, MBT178519, CBS H-21664 isoepitype, culture ex-epitype IMI 349061 = CPC 18941, CPC 19379); Trinidad, from leaf spot on living leaf of Brassica rapa subsp. chinensis, collection date and collector unknown, culture IMI 349063 = CPC 18942, CPC 19380. USA, Georgia, experiment, on leaf spots of Brassica rapa, 24 Jul. 1916, B. B. Higgins (no. 340), (BPI 398582 holotype of C. higginsianum).

Notes: Colletotrichum higginsianum is known as the causal organism of anthracnose disease of a wide range of cruciferous plants (Brassicaceae) and causes mainly leaf spots but also attacks stems, petioles, seed pods and even roots, and is especially destructive in the south Atlantic and Gulf states of the USA (Higgins 1917, Rimmer 2007), but also occurs in the Carribean and south-east Asia (Birker et al. 2009).

Higgins (1917) noted that this species was associated with a leaf spot disease of turnips (Brassica rapa) in various localities in Georgia, USA and tentatively called it C. brassicae Schulzer & Sacc. However, Higgins had doubts about this identification and sent specimens to P.A. Saccardo. In a footnote, Higgins explained that Saccardo considered that the fungus was a new species and added Saccardo's species description from the note he received after his paper was ready for publication.

A specimen of C. higginsianum was located at BPI that was collected by B.B. Higgins prior to the publication (BPI 398582), and is therefore considered as the holotype. The specimen comprises two leaves with leaf spots that agree with the description and figures in the publication. Conidia of the holotype are nearly straight, sometimes very slightly curved, measuring (14–)16–20(–22) × (3–)3.5–4.5(–5) μm, av. ± SD = 18.1 ± 1.9 × 4.1 ± 0.6 μm, L/W ratio = 4.5. This agrees with the shape and measurements of the isolates studied here.

The only species that was described on Brassica prior to Higgins (1917) is C. brassicae Schulzer & Sacc. (1884), on Brassica oleracea v. caulocarpa from Vinkovce, Slovenia that forms curved conidia 19–24 μm long (Schultzer von Mueggenburg & Saccardo 1884). The ITS sequence (GenBank EU400155) of strain DAOM 116226 identified as C. brassicae (Chen et al. 2007) is identical to that of the ex-type strain of C. spaethianum (CBS 167.49) from a study on Colletotrichum species with curved conidia (Damm et al. 2009). Another strain from a stump of Brassica sp. in the Netherlands included in the study of Damm et al. (2009) was identified as C. truncatum, based on sequence similarities with the ex-epitype strain of that species. It is possible that C. brassicae is synonymous with either C. spaethianum or C. truncatum. We have not studied the type specimen as we do not consider this species to be part of the C. destructivum species complex. Colletotrichum brassicicola was described recently from Brassica; it forms straight conidia and belongs to the C. boninense species complex (Damm et al. 2012).

A species described from leaf spots on Matthiola incana in Romania by Sandu-Ville (1959), C. mathiolae, also resembles C. higginsianum and could be a synonym of this species or closely related based on similar conidia shape and size. Colletotrichum mathiolae also forms conidia that are straight to slightly curved, measuring 12–21 × 3–4 μm. The two isolates from Matthiola in Japan are closely related to C. higginsianum, but do not have the same GAPDH and HIS3 sequences, which may explain why they do not form a stable clade in our phylogeny. Additional isolates from this host, especially from Romania, are required to determine species boundaries and its affinity to C. mathiolae.

Colletotrichum higginsianum was regarded as a synonym of C. gloeosporioides by von Arx (1957), but Sutton (1980, 1992) considered it as a distinct species based on its conidial morphology and consistent association with cruciferous hosts. O'Connell et al. (2004) recognised the similarity and relatedness with C. destructivum and regarded C. higginsianum as a synonym of C. destructivum based on ITS sequences. Colletotrichum higginsianum is confirmed as a distinct species in the present study.

Two isolates included in this study, strain CBS 124249 from Centella asiatica in Madagascar (Rakotoniriana et al. 2008) and strain IMI 391904 from Raphanus raphanistrum in Tunisia (Djebali et al. 2009), which were previously identified as C. higginsianum, were re-identified as C. tabacum and C. lini, respectively.

O'Connell et al. (2004) observed the two-stage hemibiotrophic infection process of C. destructivum (re-identified here as C. higginsianum) from Brassica rapa subsp. chinensis on Arabidopsis thaliana. They also established an Agrobacterium-mediated DNA-transformation system for this fungus. The Arabidopsis-Colletotrichum pathosystem provides a model for molecular analysis of plant-fungal interactions in which both partners can be genetically manipulated. This pathosystem has been intensively studied in recent years (Huser et al. 2009, Ushimaru et al. 2010, Kleemann et al. 2012). The genome and in planta transcriptome of C. higginsianum strain IMI 349063 were sequenced (O'Connell et al. 2012), and this is one of the strains included in the present study.

Colletotrichum higginsianum can be identified by TUB2 and ITS sequences. However, there is only one nucleotide difference to the TUB2 sequences of the two unnamed strains from Matthiola (CM90-M1 and CM93-M1) and one nucleotide difference to the ITS sequences of C. tabacum, respectively. Three of the strains diverge with a further single nucleotide difference to the other C. higginsianum strains.

The ITS of strain IMI 349061 was identical with those of C. higginsianum isolates 05131 from Eruca in the USA (GenBank KF550281, Patel et al. 2014), 12-223 (GenBank JX997428, K.S. Han et al., unpubl. data), C97027 and C00112 (GenBank GU935870, GU935872, Choi et al. 2011) from Brassica probably in Korea, IMI 349063 and MAFF 305635 (GenBank JQ005760, JQ005761 O'Connell et al. 2012, Naumann & Wicklow 2013), MAFF 305635, MAFF 238563, MAFF 305970, IFO6182 (GenBank AB042302, AB042303, AB105955, AB105957, Moriwaki et al. 2002) from Brassica, Matthiola and an unknown host, and except for the last, included in this study, and C. destructivum isolates RGT-S12, endophyte of Rumex probably in China (GenBank HQ674658, Hu et al. 2012) and CD-hz 01–CD-hz 03 from Vigna in China (GenBank EU070911–EU070913, Sun & Zhang 2009).

Colletotrichum lentis Damm, sp. nov. MycoBank MB809921. Fig. 8.

Open in a separate window

Fig. 8

Colletotrichum lentis (from ex-holotype strain CBS 127604). A–B. Conidiomata. C, G. Seta. D–F, H–K. Conidiophores. L–Q. Appressoria. R–S. Conidia. A, C–F, R. from Anthriscus stem. B, G–Q, S. from SNA. A–B. DM, C–S. DIC, Scale bars: A = 100 μm, F = 10 μm. Scale bar of A applies to A–B. Scale bar of F applies to C–S.

≠ Colletotrichum truncatum (Schwein.) Andrus & W.D. Moore, Phytopathology 25: 121. 1935.

Basionym: Vermicularia truncata Schwein., Trans. Amer. Philos. Soc. 4(2): 230. 1832.

≡ Glomerella truncata (Schwein.) C.L. Armstrong & Banniza, Mycol. Res. 110: 953. 2006.

Etymology: The species epithet is derived from the host genus Lens.

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1.5–11 μm diam, hyaline, smooth-walled, septate, branched. Chlamydospores not observed. Conidiomata absent, conidiophores and setae formed directly on hyphae or on or close to chains or clusters of pale to dark brown, verruculose, cylindrical to subglobose, cells. Setae pale to medium brown, smooth-walled, 40–85 μm long, 1–3-septate, base ± inflated, sometimes constricted at the basal septum, 5–6 μm diam, tip round. Conidiophores hyaline, smooth-walled, septate, branched, to 30 μm long. Conidiogenous cells hyaline, smooth-walled, cylindrical to ampulliform, 9–28 × 3.5–5 μm, sometimes intercalary (necks not separated from hyphae by septum) and sometimes polyphialides observed, opening 1–2 μm diam, collarette 0.5–1 μm long, periclinal thickening observed. Conidia hyaline, smooth-walled, aseptate straight to slightly curved, fusiform with ± acute ends, (13–)16–20(–26) × 3–4(–5) μm, av. ± SD = 18.1 ± 2.0 × 3.5 ± 0.4 μm, L/W ratio = 5.1, conidia of strain CBS 127605 shorter, measuring (13–)15–17.5(–19.5) × 3–3.5(–4) μm, av. ± SD = 16.3 ± 1.4 × 3.4 ± 0.2 μm, L/W ratio = 4.8. Appressoria single or in loose groups, medium brown, smooth-walled, globose, subglobose to elliptical in outline, with an entire margin, (5–)5.5–7.5(–9) × (3.5–)4.5–6(–6.5) μm, av. ± SD = 6.4 ± 0.8 × 5.2 ± 0.6 μm, L/W ratio = 1.2.

Asexual morph on Anthriscus stem. Conidiomata, conidiophores and setae formed on hyaline to pale brown, angular cells, 3.5–9 μm diam. Setae pale brown, smooth-walled, 30–120 μm long, 1–3-septate, base ± inflated, 5–6 μm diam, tip round. Conidiophores hyaline to pale brown, smooth-walled, septate, branched, to 20 μm long. Conidiogenous cells hyaline, smooth-walled, cylindrical to ampulliform, 11–22 × 3.5–5 μm, opening 1.5–2 μm diam, collarette 0.5–1 μm long, periclinal thickening distinct. Conidia hyaline, smooth-walled, aseptate, straight to slightly curved, fusiform with ± acute ends, (15.5–)17–20(–21.5) × 3–3.5(–4 μm, av. ± SD = 18.6 ± 1.6 × 3.4 ± 0.3 μm, L/W ratio = 5.5.

Culture characteristics: Colonies on SNA flat with entire margin, hyaline, partly pale rosy buff to pale olivaceous grey, agar medium, filter paper and Anthriscus stem partly covered with floccose white aerial mycelium or aerial mycelium lacking, reverse same colours; growth 15–18.5 mm in 7 d (23.5–25 mm in 10 d). Colonies on OA flat with entire margin; surface straw, pale luteous to amber, partly covered with very short aerial mycelium and partly covert with black to salmon acervuli, aerial mycelium lacking, reverse same colours; growth 21–22.5 mm in 7 d (30–34 mm in 10 d). Conidia in mass whitish to salmon.

Materials examined: Canada, Saskatchewan, North Battlefield, from seed, 2001 crop, sample 90812, of Lens culinaris cv. ‘CDA Grandora’, 2001, R.A.A. Morrall and Discovery Seed Labs (CBS H-21649 holotype of C. lentis, culture ex-holotype CBS 127604 = DAOM 235316 = CT21, Race Ct1); Saskatchewan, Moose Jaw, from seed, 2001 crop, sample 91639, of Lens culinaris cv. ‘CDA Grandora’, 2001, R.A.A. Morrall and Discovery Seed Labs, CBS H-21650, culture CBS 127605 = DAOM 235317 = CT26, Race Ct0. Romania, Iaşi, on pods and leaves of Lens culinaris, 30 Jun. 1950, C. Sandu-Ville (GLM-F102752 holotype of C. savulescui Sandu ex Herbarul Micologic “C. Sandu-Ville”).

Notes: In 1986 and 1987, an anthracnose disease of lentil (Lens culinaris) was observed in Manitoba, Canada, and identified as C. truncatum by Morrall (1988). Armstrong-Cho & Banniza (2006) induced the formation of perithecia by crossing single conidial isolates of the lentil pathogen in the laboratory. Consequently, they considered these crosses as the sexual morph of C. truncatum and with the dual nomenclature still in place, named this sexual morph Glomerella truncata, although morphological as well as molecular studies (Ford et al. 2004) comparing lentil isolates with “C. truncatum” isolates from soybean, clover, peanut and cocklebur indicated different species were involved. Latunde-Dada & Lucas (2007) and Gossen et al. (2009) found isolates from anthracnose of lentil in Canada to be closely related to C. destructivum. Damm et al. (2009) epitypified C. truncatum and revealed the lentil pathogen from Canada to be a different species. In contrast to that species, C. truncatum forms strongly curved conidia and does not belong to the C. destructivum complex (Damm et al. 2009). The phylogenetic relationship between the two species was demonstrated by O'Connell et al. (2012) and Cannon et al. (2012).

With the adoption of the new International Code of Nomenclature for algae, fungi and plants concerning species names for morphs with the same epithet introduced prior to 1 January 2013, the name Ga. truncata will be considered as a new combination of the previously described C. truncatum and not as a new species, although it is based on a different type and the two types are not conspecific (McNeill et al. 2012, Hawksworth et al. 2013). Consequently, the lentil pathogen from Canada is described as a new species in this study, C. lentis. As suggested by Hawksworth et al. (2013), the name Ga. truncata was treated as a formal error for a new combination and corrected accordingly to Glomerella truncata (Schwein.) C.L. Armstrong & Banniza.

The two strains studied here, CBS 127604 and CBS 127605, were crossed to produce the original holotype specimen of “Ga. truncata”, DAOM 235318, on inoculated sterilised stems of Lens culinaris, but the latter will have no nomenclatural status under the new changes to the Code (Hawksworth et al. 2013). As there is no strain derived from this specimen and an epitypification would be needed in order to have an ex-type strain, it was not used as holotype of C. lentis. Instead, CBS 127604 was chosen for the holotype of C. lentis.

A second species from Lens culinaris, C. savulescui, was described in Romania by Sandu-Ville (1959). As specimen GLM-F102752 is apparently the only specimen of this species from the Herbarium Mycologicum Moldavicum “Constantin Sandu-Ville” (www.uaiasi.ro/agricultura/index.php?lang=en&pagina=pagini/herbarium.html) and no specimen was located elsewhere, it was considered as the holotype of C. savulescui, although its collection date was prior to that listed in the publication of Sandu-Ville (1959). Conidia of C. savulescui are hyaline, cylindrical with both ends rounded, straight or slightly curved, measuring 7.5–18 × 3–4.5 μm, mostly 12–18 × 4 μm. The size of the conidia found on the holotype is similar to that of C. lentis; the conidia are also described as straight to slightly curved, which indicates this species might belong to the C. destructivum species complex. However, C. lentis has conidia that are fusiform with ± acute ends. The name of this species, as far as we know, has not been used since its original description.

Additionally, isolates from lentil were also included in the study of Liu et al. (2013a), identified as C. nigrum. Colletotrichum nigrum is not closely related to the species treated here and forms entirely straight conidia.

Armstrong-Cho & Banniza (2006) observed self-sterility of all isolates tested, while many pairings produced perithecia and concluded the species to have a homothallic mating system. The study by Menat et al. (2012) confirmed a bipolar mating system, however an atypical one, with the HMG box that is part of the MAT1-2 idiomorph being present in both incompatibility groups. The sexual morph was described by Armstrong-Cho & Banniza (2006) as follows “Perithecia were brown-black, superficial, solitary or in small groups, obpyriform to ovate or ampulliform, 200–520 × 110–320 μm (mean: 350 × 200 μm). Asci were cylindrical, narrowing slightly at the apex, unitunicate, evanescent, 53–142 × 5–14 μm (mean: 90 × 8 μm), and contained eight ascospores. Ascospores were hyaline, aseptate, oblong, 12–20 × 5–8 mm (mean: 15.7–6.7 μm).”

Buchwaldt et al. (2004) identified two physiological races of “C. truncatum” from lentil on the basis of their pathogenicity on a number of lentil cultivars and germplasm lines in western Canada, designating them Ct0 and Ct1.

The intracellular hemibiotrophic infection of the lentil pathogen was studied by Latunde-Dada & Lucas (2007) and Armstrong-Cho et al. (2012). This pathosystem was used to identify secreted effector proteins expressed at the switch from biotrophy to necrotrophy (Bhadauria et al. 2011) and functional analysis of a nudix hydrolase effector eliciting plant cell death (Bhadauria et al. 2012).

Strains that are morphologically similar and molecularly closely related (based on ITS) to C. lentis were isolated from the noxious weed scentless chamomile (Tripleurospermum inodorum) in Canada (Forseille 2007). The potential of this fungus for biocontrol of scentless chamomile was tested (Peng et al. 2005, Forseille et al. 2009). In the field, chamomile isolates caused symptoms on its original host but not on lentil or pea. Forseille et al. (2009) also observed the hemibiotrophic infection process of this fungus, which might represent a further species of the C. destructivum complex.

Colletotrichum lentis is characterised by its slightly curved, fusoid conidia that are gradually tapering to the ± acute ends and by the ± globose appressoria with an entire margin. It can be identified by all loci included in this study.

The ITS sequence of strain CBS 127604 matched in a blastn search with the same sequence (GenBank JQ005766, O'Connell et al. 2012) and that of “C. truncatum” isolate 9969473 (GenBank AF451902, Ford et al. 2004) and with 99 % identity (1–3 nucleotides difference) with “C. truncatum” isolates 95S25, 9971646, 95A8, 9970034 from lentil in Canada (GenBank AF451901, AF451904, AF451900, AF451903, Ford et al. 2004) and “Ga. glycines” isolate IFO7384 from an unknown host (GenBank AB057435, Moriwaki et al. 2002). The only matching TUB2 sequence found in GenBank is that of the same strain (GenBank JQ005850, O'Connell et al. 2012); all other TUB2 sequences are ≤95 % identical.

Colletotrichum lini (Westerd.) Tochinai, J. Coll. Agric. Hokkaido Imp. Univ. 14: 176. 1926. Fig. 9.

Open in a separate window

Fig. 9

Colletotrichum lini (from ex-epitype strain CBS 172.51). A–B. Conidiomata. C–J. Conidiophores. K–P. Appressoria. Q–R. Conidia. A, C–E, Q. from Anthriscus stem. B, F–P, R. from SNA. A–B. DM, C–R. DIC, Scale bars: A = 100 μm, C = 10 μm. Scale bar of A applies to A–B. Scale bar of C applies to C–R.

Basionym: Gloeosporium lini Westerd., Jaarversl. Phytopathol. Lab. “Willie Commelin Scholten” 6. 1916 [1915].

= Colletotrichum linicola Pethybr. & Laff. [as ‘linicolum’], Sci. Proc. Roy. Dublin Soc. 15: 368. 1918.

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1.5–6 μm diam, hyaline, smooth-walled, septate, branched. Chlamydospores not observed. Conidiomata absent, conidiophores formed directly on hyphae. Setae not observed. Setae of strain IMI 391904 medium brown, smooth-walled to verruculose, 52–94 μm long, 1–3-septate, base cylindrical to conical, 3.5–6.5 μm diam, tip rounded. Conidiophores hyaline, smooth-walled, septate, branched, to 40 μm long. Conidiogenous cells hyaline, smooth-walled, cylindrical, 9–32 × 2.5–4.5 μm, opening 1–1.5 μm diam, collarette 0.5 μm long, periclinal thickening rarely observed. Conidia hyaline, smooth-walled, aseptate, fusiform, slightly curved to straight, tapering to the slightly rounded to acute ends, (13–)15–18(–22.5) × (3–)3.5–4(–4.5) μm, av. ± SD = 16.6 ± 1.6 × 3.8 ± 0.3 μm, L/W ratio = 4.4, conidia of strain CBS 112.21 are smaller, measuring (12–)13.5–16.5(–18.5) × (3–)3.5–4.5(–5) μm, av. ± SD = 15.0 ± 1.4 × 4.0 ± 0.4 μm, L/W ratio = 3.7, conidia of strain CBS 117156 are longer, measuring (18–)18.5–20(–21) × 3.5–4(–4.5) μm, av. ± SD = 19.3 ± 0.8 × 3.9 ± 0.2 μm, L/W ratio = 5.0, the ex-epitype strain CBS 172.51 and of strain CBS 112.21 formed inside SNA agar medium are larger conidia than on the surface of the medium, those of strain CBS 172.51 measure (23.5–)24–33(–52.5) × 4–4.5(–5) μm, av. ± SD = 28.6 ± 4.3 × 4.3 ± 0.3 μm, L/W ratio = 6.7. Appressoria single or in loose groups, pale brown, smooth-walled, ellipsoidal to subglobose outline, with an entire or undulate margin, (5–)6.5–10(–12.5) × (4–)4.5–6(–7) μm, av. ± SD = 8.3 ± 1.9 × 5.3 ± 0.9 μm, L/W ratio = 1.6, strain IMI 391904 additionally formed appressoria-like structures within the mycelium, measuring (3.5–)5–7.5(–8) × (2.5–)3.5–5.5(–6) μm, av. ± SD = 6.3 ± 1.2 × 4.5 ± 0.8 μm, L/W ratio = 1.4.

Asexual morph on Anthriscus stem. Conidiomata, conidiophores and setae formed on pale brown, angular cells, 3–8.5 μm diam. Setae not observed. Setae of strain IMI 391904 medium brown, smooth-walled to finely verruculose, 55–210 μm long, 1–5(–6)-septate, base cylindrical to conical, 3.5–7 μm diam, tip slightly rounded. Conidiophores hyaline, smooth-walled, septate, branched, to 40 μm long. Conidiogenous cells hyaline, smooth-walled, cylindrical to elongate ampulliform, 8–22 × 2.5–4 μm, opening 1–2 μm diam, collarette 0.5–1 μm long, periclinal thickening observed. Conidia hyaline, smooth-walled, aseptate, fusiform, slightly curved to straight, tapering to the slightly rounded to acute ends, (14.5–)16.5–19.5(–21.5) × 3.5–4 μm, av. ± SD = 18.0 ± 1.5 × 3.8 ± 0.2 μm, L/W ratio = 4.7, conidia of strain CBS 112.21 are smaller, measuring (13.5–)15–17.5(–19.5) × 4–4.5 μm, av. ± SD = 16.3 ± 1.4 × 4.3 ± 0.2 μm, L/W ratio = 3.8, conidia of strain CBS 117156 are longer (17.5–)19.5–22.5(–23.5) × (3–)3.5–4(–4.5) μm, av. ± SD = 21.1 ± 1.4 × 3.8 ± 0.2 μm, L/W ratio = 5.5.

Culture characteristics: Colonies on SNA flat with entire margin, hyaline to pale luteous, filter paper partly pale luteous, agar medium and Anthriscus stem partly covered with floccose white aerial mycelium, reverse same colours; growth 27.5–30 mm in 7 d (>40 mm in 10 d). Colonies on OA flat with entire margin; buff to rosy-buff, aerial mycelium lacking, reverse buff, growth 26–29 mm in 7 d (37.5–>40 mm in 10 d). Conidia in mass not observed.

Materials examined: Germany, Dierhagen-Neuhaus, walkway, from stems and leaves with black spots of Trifolium repens, 4 Aug. 2010, U. Damm, CBS H-21660, culture CBS 130828. Ireland, from Linum usitatissimum, collection date unknown (isolated by P. Mercer and deposited in CBS collection Feb. 1997 by J.A. Bailey), P. Mercer, culture CBS 505.97 = LARS 77. Netherlands, from leaves and stems of Linum sp., collection date and collector unknown (IMI 194722 ex coll. Prof. J. van Westerdijk, lectotype of Gm. lini, here designated, MBT178721); from seed plants of Linum sp., collection date and collector unknown (IMI 194721 ex coll. Prof. J. van Westerdijk); from seedling disease of Linum usitatissimum, collection date and collector unknown (deposited in CBS collection Sep. 1951 by Plantenziektenkundige Dienst Wageningen, Nederland, identified by A.C. Stolk) (CBS H-21657 epitype of Gm. lini, here designated, MBT178521, culture ex-epitype CBS 172.51); Province Gelderland, Malden, closed railway nearby gliding-club, from leaf spots of Teucrium scorodonia, 23 Aug. 2004, G. Verkley and M. Starink, V3037, culture CBS 117156. New Zealand, from Trifolium sp., collection date and collector unknown, (history: I.D. Blair, 1957 CABI), culture IMI 69991 = CPC 20242. Tunisia, site Barragage Jjoumine, from symptoms on a living leaf of Raphanus raphanistrum, collection date and collector unknown (deposited in IMI collection by Dr. M. Jourdan), CBS H-21658, culture IMI 391904 = CPC 19382 = IS320. UK, from seedling disease of Linum usitatissimum, collection date and collector unknown (isolated by G.H. Pethybridge, deposited in CBS collection Aug. 1921 by G.H. Pethybridge), CBS H-21656, culture CBS 112.21 = LCP 46.621. USA, Utah, Salt Lake City, cemetery, from small black spots on petioles of Trifolium hybridum, 24 Aug. 2013, U. Damm, CBS H-21659, culture CBS 136850; Utah, Bluffdale near Salt Lake City, stems of Medicago sativa, 25 Aug. 2013, U. Damm, culture CBS 136856.

Notes: Anthracnose has a serious impact on yield and fibre quality of flax (Linum usitatissimum) and is well-known in Europe, Asia and America. Flax anthracnose increased in Germany when flax production was expanding in the 1930s (Rost 1938). The anthracnose pathogen is seed- and soilborne, causes damping off of flax seedlings (Rost 1938), and is one of the causal organisms of so-called flax-sick soils (Bolley & Manns 1932).

Van Westerdijk (1916) described the flax anthracnose pathogen in the Netherlands as Gloeosporium lini, citing the genus as Gloeosporium (Colletotrichum). This name was combined into Colletotrichum by Tochinai (1926), following the study of several Japanese collections. Neither the location of the fungus nor a type was listed by van Westerdijk. Unfortunately, no strain was preserved in the CBS culture collection. However, two specimens from Van Westerdijk's Gm. lini collections were sent to the IMI fungarium by von Arx, and the one containing a larger amount of diseased plant material (IMI 194722) is designated as lectotype. The specimen includes fusiform, slightly curved to straight conidia with slightly rounded to acute ends that measure (14–)16–21(–24) × (3–)3.5–5 μm, av. ± SD = 18.4 ± 2.3 × 4.2 ± 0.7 μm, L/W ratio = 4.4. Sutton (1980) listed the species from flax as C. lini (Westerd.) Tochinai, but later (Sutton 1992) followed Dickson's (1956) opinion that the basionym, Gm. lini Westerd. was probably synonymous with Polyspora lini Laff. (current name in Species Fungorum: Kabatiella lini) and not a Colletotrichum. However, the conidia on the lectotype specimen of Gm. lini agreed both in shape and size with the Colletotrichum species from Linum we treat in this study.

A Colletotrichum species on Linum in North Dakota, USA, was studied between 1901 and 1903 by T.F. Manns and also called C. lini; however, his thesis was never published (Manns & Bolley 1932). The name was taken up by Bolley (1910); however, it is illegitimate as it is a “nomen nudum”. Bolley & Manns later (1932) treated the fungus as C. lini Manns et Bolley. Conidia of this species measure 15–20 × 2–4.5 μm, setae are 70–130 μm long and 2–4-septate, and the olive brown “chlamydospores” measure 10–15 × 10–12 μm (Bolley & Manns 1932). This agrees with the observations of the Colletotrichum from Linum in this study and is probably a synonym. We have not seen the type of this species and no isolates from Linum in the USA were available to us.

Pethybridge & Lafferty (1918) described C. linicola as the causal agent of damping off of flax seedlings in Ireland with conidia measuring 17 × 4 μm and 3-septate setae measuring 150 × 4 μm. This species is most probably a synonym of C. lini (Westerd.) Tochinai. Both an authentic strain from the UK isolated by G.H. Pethybridge (CBS 112.21) and a strain from Ireland (CBS 505.97) are included in our study.

Rost (1938) lists C. atramentarium that formed straight conidia on flax in Germany and which is probably a synonym of C. coccodes (Liu et al. 2011). Wollenweber & Hochapfel (1949) also identified a collection from stems of Linum from Silesia as C. atramentarium.

Hahn (1952) examined the infection process of C. lini on resistant and susceptible flax lines and provided the first description of bulbous primary hyphae colonising single epidermal cells. These were subsequently found to be the characteristic biotrophic infection structures formed by all members of the C. destructivum species complex examined to date.

Conidia of C. lini strains from Linum are similar to those of C. lentis. They are both slightly curved and fusiform, but conidia of C. lini are more abruptly tapering to the slightly acute ends; this shape was noticed in the type material (not shown), and very long conidia were found within the agar medium. In accordance with the original description and the observations on the type, no setae were observed on the strains from flax, but none of the isolates was recently collected.

In contrast to the C. lini strains from Linum, the strains from Trifolium hybridum, T. repens, Medicago sativa and Taraxacum sp. formed setae and rather cylindrical conidia with rounded ends. These strains formed a subclade within C. lini. However, we refrained from describing these strains as a new species, because there was only one nucleotide difference in the TUB2 sequence to separate them from the remaining C. linum strains; the overall sequence variability within C. lini was higher. Moreover, their morphology was similar to strains from Nigella, Raphanus and Teucrium, which belong to the same subclade as the strains from Linum. Both subclades contain strains from multiple hosts.

Colletotrichum lini is distinguishable by CHS-1, HIS3, ACT and TUB2. The ITS and GAPDH sequences are the same as those of C. americae-borealis. The sequences of all genes in strains from Linum, Nigella and Teucrium are identical. The strain from Raphanus in Tunisia (IMI 391904) with the longest branch differs only in its GAPDH sequence.

As the sequences are the same, blastn searches with the ITS sequence of C. lini strain CBS 172.51 resulted in the same matching sequences as those with the ITS of C. americae-borealis, including isolates from alfalfa, clover, Oxytropis, endophytes from Holcus and Arabidopsis as well as strain IMI 391904 that is included in our study and a strain from Convolvulus in Turkey. Strain IMI 391904 originated from a study on pathogenic fungi on wild radish (Raphanus raphanistrum) in northern Tunisia in order to screen for potential biocontrol agents against this weed (Djebali et al. 2009). It was previously identified as C. higginsianum and re-identified as C. lini in this study. The identification of the strain from field bindweed (Convolvulus arvensis) in Turkey as C. linicola is based on the ITS sequence only (Tunali et al. 2008); it was tested to be effective as a potential biocontrol agent against that plant (Tunali et al. 2009). However, the identity of this strain needs to be confirmed with sequences of additional loci.

The ITS sequence of C. lini strain Coll-44 from a recent disease report of anthracnose on Medicago in Serbia (GenBank JX908364, Vasić et al. 2014) is identical to strains from C. americae-borealis and C. lini. As the TUB2 sequence (GenBank KJ556347, kindly provided by Tanja Vasić) was identical to that of C. lini strain CBS 136850 (from Trifolium hybridum, USA), strain Coll-44 is confirmed as C. lini. Sequences of an isolate from our study (CBS 157.83) and several ITS sequences detected in GenBank (GenBank JX908362, JX908363, JX908361, Vasić, unpubl. data) are identical to those of C. destructivum s. str., indicating the occurrence of at least two species on Medicago in Serbia.

The performance of C. lini strain CBS 112.21 in comparison with Botryodiplodia malorum in steroid hydroxylations, to improve the biotransformation of steroids for the pharmaceutical industry, was studied by Romano et al. (2006).

Colletotrichum ocimi Damm, sp. nov. MycoBank MB809401. Fig. 10.

Open in a separate window

Fig. 10

Colletotrichum ocimi (from ex-holotype strain CBS 298.94). A–B. Conidiomata. C, G. Tip of a seta. D, H. Base of a seta. E–F, I–K. Conidiophores. L–Q. Appressoria. R–S. Conidia. A, C–F, R. from Anthriscus stem. B, G–Q, S. from SNA. A–B. DM, C–S. DIC, Scale bars: A = 100 μm, E = 10 μm. Scale bar of A applies to A–B. Scale bar of E applies to C–S.

Etymology: The species epithet is derived from the host genus name Ocimum.

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1–7 μm diam, hyaline, smooth-walled, septate, branched. Chlamydospores not observed. Conidiomata conidiophores and setae formed on pale brown, roundish cells, 5–22 μm diam. Setae medium brown, smooth-walled to verruculose, 43–103 μm long, 1–2-septate, base cylindrical to ± inflated, 4.5–9.5 μm diam, tip ± rounded to ± acute. Conidiophores hyaline, smooth-walled, septate, branched, to 60 μm long. Conidiogenous cells hyaline to pale brown, smooth-walled to verruculose, cylindrical to clavate, sometimes intercalary (necks not separated from hyphae by septum), often with slime sheaths, 10.5–24 × 3.5–5.5 μm, opening 1–1.5 μm diam, collarette 0.5–1.5 μm long, periclinal thickening distinct. Conidia hyaline, smooth-walled, aseptate, straight, cylindrical, with both ends rounded or one end round and the other truncate, (13.5–)14.5–15.5(–16.5) × (3.5–)4–4.5 μm, av. ± SD = 15.0 ± 0.7 × 4.1 ± 0.2 μm, L/W ratio = 3.7. Appressoria very few, single, scattered, pale brown, smooth-walled, ellipsoidal, clavate, subglobose or irregular outline, with a lobate or entire margin, (6.5–)7–13(–15.5) × (4–)4.5–7.5(–9) μm, av. ± SD = 9.9 ± 2.9 × 6.0 ± 1.3 μm, L/W ratio = 1.6.

Asexual morph on Anthriscus stem. Conidiomata, conidiophores and setae formed on pale brown, verruculose, roundish cells, 4–17 μm diam. Setae medium brown, verruculose, 30–145 μm long, 1–4-septate, base cylindrical, conical to ± inflated, 4–7.5 μm diam, tip ± rounded to ± acute. Conidiophores hyaline to pale brown, smooth-walled, septate, branched, to 25 μm long. Conidiogenous cells hyaline to pale brown, smooth-walled, cylindrical to ampulliform, 8–21.5 × 3.5–5 μm, opening 1–1.5 μm diam, collarette 0.5 μm long, periclinal thickening visible. Conidia hyaline, smooth-walled, aseptate, straight, cylindrical, with both ends rounded or one end round and the other truncate, (11–)14–16(–16.5) × (3.5–)4(–4.5) μm, av. ± SD = 14.8 ± 1.0 × 4.0 ± 0.2 μm, L/W ratio = 3.7.

Culture characteristics: Colonies on SNA flat with entire margin, hyaline to cinnamon, agar medium, filter paper and Anthriscus stem partly covered with grey acervuli, aerial mycelium lacking, reverse same colours; growth 18–19.5 mm in 7 d (28–29.5 mm in 10 d). Colonies on OA flat with entire margin; buff to honey, almost entirely covert with dark grey to black acervuli and salmon conidial masses, aerial mycelium lacking, reverse rosy buff, vinaceous buff to pale olivaceous-grey, growth 20.5–22 mm in 7 d (30–31.5 mm in 10 d). Conidia in mass salmon.

Material examined: Italy, Riviera Ligure, from a black spot on leaf of Ocimum basilicum, collection date and collector unknown (deposited in CBS collection May 1994 by A. Garibaldi, Inst. degli studi di Torino, Depart. di Valorizzazione e Protezione delle Risore agroforestiali) (CBS H-21646 holotype, culture ex-holotype CBS 298.94).

Notes: Basil (Ocimum basilicum) is an aromatic culinary herb, for which flawless leaves are of special importance. Gullino et al. (1995) reported an outbreak of a new foliar disease of basil cultivated in greenhouses in northern Italy and consistently isolated a Colletotrichum species. The fungus caused black spots on stems and leaves of basil; lesions on stems often resulted in girdling and plant death. One strain (CBS 298.94) was sent to CBS and identified as Glomerella cingulata var. cingulata (until recently regarded as the sexual stage of C. gloeosporioides) by H.A. van der Aa (HA 11925) as indicated in the database of the CBS culture collection.

This species forms cylindrical, straight conidia with round ends, reminiscent of species in the C. gloeosporioides complex (Weir et al. 2012). However, we found that C. ocimi belongs to the C. destructivum species complex. Gullino et al. (1995) did not observe a sexual stage of the basil fungus. Apart from the conidia, C. ocimi differs from the other species in the C. destructivum complex by its conidiogenous cells that are often covered by mucoid sheaths.

No species were previously described on Ocimum. Additionally to Gullino et al. (1995), Farr & Rossman (2014) list a few further reports of Colletotrichum species on basil: C. capsici in India, C. gloeosporioides in Cambodia and Colletotrichum sp. in Florida, USA. It is possible that the latter two reports refer to C. ocimi as well. However, the only sequence of a Colletotrichum strain from basil in GenBank, which is an ITS sequence of strain EGJMP 40 probably from India (GenBank KF234012) identified as C. aotearoa (E.G. Jagan et al., unpubl. data), indeed refers to a species belonging to the C. gloeosporioides species complex.

This species can be identified by its unique ITS, CHS-1, HIS3, ACT, and TUB2 sequences. The closest match in a blastn search with the ITS sequence of strain CBS 298.94 is GenBank EU400148 from C. lini strain DAOM 183091 (Chen et al. 2007). No TUB2 sequences were detected in GenBank with >97 % identity. The GAPDH sequence of C. ocimi is the same as that of C. destructivum (s. str.).

Colletotrichum panacicola Uyeda & S. Takim., Bull. Korean Agric. Soc. 14: 24. 1919. MycoBank MB809665.

≡ Colletotrichum panacicola Uyeda & S. Takim., Bull. Agric. Experiment Stat. Chosen (Korea) 5: 16. 1922. Nom. illegit., Art. 53.1.

Notes: Colletotrichum panacicola, originally described from Panax ginseng in Korea, has also been reported from China, eastern Russia and Japan, while anthracnose of American ginseng (P. quinquefolius) is caused by C. dematium (s. lat.) and C. coccodes (McPartland & Hosoya 1997). McPartland & Hosoya (1997) corrected the author citation of the species that had been described already by Takimoto (1919), but that would have to be cited as C. panacicola Uyeda & S. Takim. The confusion was caused by Nakata & Takimoto (1922) who described the same species again as a new species. Petrak (1953) cited the species wrongly as C. panacicola Nakata & S. Takim., which was subsequently taken up by Index Fungorum.

The species was characterised with aseptate, cylindrical, straight or slightly curved conidia with rounded ends, measuring 17.0–22.1 × 3.4–5.1 μm, pyriform olive coloured appressoria, measuring 14–8 μm and dark olive 1–3-septate setae with acute paler apices that measure 31–144 × 2.4–8.4 μm (Takimoto 1919, Nakata & Takimoto 1922, both cited by McPartland & Hosoya 1997). McPartland & Hosoya (1997) were unable to locate either type or authentic specimens. As the illustration in Nakata & Takimoto (1922) is not sufficiently diagnostic to act as a lectotype, the species needs to be neotypified.

Fresh cultures are available as Choi et al. (2011) recently studied isolates of this species and observed similarity with C. higginsianum, C. destructivum and C. coccodes. ITS sequences did not distinguish the species from C. higginsianum and C. destructivum. The inclusion of more genes (ACT, translation elongation factor 1-α, glutamine synthase) clearly showed this species to be different from the other two (Choi et al. 2011). Choi et al. (2011) who also showed this species to only infect Korean ginseng, suggesting it was a distinct taxon.

As there were no isolates available to us, we could not directly compare the morphology of C. panacicola. However, we included DNA sequences of three isolates from the study of Choi et al. (2011) that were retrieved from GenBank in our molecular analyses (only with ITS, GAPDH and ACT), which confirmed C. panacicola to belong to the C. destructivum complex and to be a distinct species, although closely related to the newly described C. utrechtense, which has the same ACT sequence. Colletotrichum panacicola can be identified by ITS and GAPDH sequences; 100 % sequence identities on GenBank were only found with the C. panacicola sequences from the study of Choi et al. (2011). Unfortunately, the TUB2 region sequenced by Choi et al. (2011) was different from the region we studied and could not be compared to our dataset; the CHS-1 and HIS sequences of this species were not available for comparison.

Colletotrichum pisicola Damm, sp. nov. MycoBank MB809403. Fig. 11.

Open in a separate window

Fig. 11

Colletotrichum pisicola (from ex-holotype strain CBS 724.97). A–B. Conidiomata. C, G. Tip of a seta. D, H. Base of a seta. E–F, I–M. Conidiophores. N–S. Appressoria. T–U. Conidia. A, C–F, T. from Anthriscus stem. B, G–S, U. from SNA. A–B. DM, C–S. DIC, Scale bars: A = 100 μm, E = 10 μm. Scale bar of A applies to A–B. Scale bar of E applies to C–S.

Etymology: The species epithet is derived from the host plant genus, Pisum.

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1–7.5 μm diam, hyaline, smooth-walled, septate, branched, at some parts pale to medium brown. Chlamydospores not observed. Conidiomata absent, conidiophores and setae formed directly on hyphae or aggregated on clusters of pale to medium brown, roundish cells, 3.5–11 μm diam. Setae few observed, pale brown, smooth-walled to verrucose, 30–40 μm long, 1–2-septate, base cylindrical to conical, 4–5 μm diam, tip round or with a conidiogenous locus. Conidiophores hyaline, smooth-walled, septate, branched, to 45 μm long. Conidiogenous cells hyaline, smooth-walled, cylindrical to ampulliform, sometimes intercalary (necks not separated from hyphae by septum), 11–30 × 2.5–5 μm, opening 1–1,5 μm diam, collarette 1–2,5 μm long, periclinal thickening visible, sometimes distinct. Conidia hyaline, smooth-walled, aseptate, fusiform, distinctly curved gradually tapering to the ± acute ends, (11–)15–21(–29.5) × (3–)3.5–4 μm, av. ± SD = 18.1 ± 2.9 × 3.5 ± 0.2 μm, L/W ratio = 5.2. Appressoria single, pale brown, smooth-walled, elliptical, clavate to irregular outline, with an entire or undulate margin, (5.5–)7–11.5(–13.5) × (4–)4.5–6(–6.5) μm, av. ± SD = 9.3 ± 2.2 × 5.1 ± 0.7 μm, L/W ratio = 1.8.

Asexual morph on Anthriscus stem. Conidiomata, conidiophores and setae formed on pale to medium brown, roundish to angular cells, 3.5–12 μm diam. Setae pale brown, smooth-walled to verrucose, 40–55 μm long, 1–2-septate, base conical, 4–6.5 μm diam, tip rounded. Conidiophores pale brown, smooth-walled, sometimes septate and branched, to 30 μm long. Conidiogenous cells pale brown, smooth-walled, ampulliform to cylindrical, 10.5–24 × 3.5–5.5 μm, opening 1–1.5 μm diam, collarette 0.5–1 μm long, periclinal thickening observed. Conidia hyaline, smooth-walled, aseptate, fusoid, distinctly curved, gradually tapering to the ± acute ends, (12–)15–20.5(–23.5) × (3–)3.5–4 μm, av. ± SD = 17.8 ± 2.8 × 3.7 ± 0.3 μm, L/W ratio = 4.7.

Culture characteristics: Colonies on SNA flat with entire margin, pale straw, covered short filty whitish aerial mycelium, reverse pale luteous; growth 6.5–7.5 mm in 7 d (8.5–10 mm in 10 d). Colonies on OA flat with entire margin; pure yellow to luteous, with a buff margin, covered with very short aerial mycelium, reverse pale luteous to luteous, growth 12–15 mm in 7 d (17.5–20 mm in 10 d). Conidia in mass not visible.

Materials examined: Ecuador, Quito, from anthracnose symptoms on pods of Pisum sp., Jan. 1891, G. Lagerheim (BPI 797146 (ex herbarium N. Patouillard) lectotype of C. pisi, here designated, MBT178523); Quito, from anthracnose symptoms on pods of Pisum sativum, Jan. 1892, G. Lagerheim, BPI 399530, includes slide; Quito, from anthracnose symptoms on pods of Pisum sativum, Feb. 1892, G. Lagerheim, BPI 399531, includes slide; Quito, from pods of Pisum sativum, Feb. 1892, G. Lagerheim (No. 2944), BPI 399532, includes slide. Mexico, intercepted at El Paso, Texas, USA, from anthracnose symptoms on pods of Pisum sativum, 17 Dec. 1952, J.A. Baker (No. 53856), BPI 399536; intercepted at Laredo, Texas, USA, from anthracnose symptoms on pods of Pisum sativum, 19 Nov. 1954, Ragsdale (No. 55077), BPI 399534. USA, Wisconsin, from Pisum sativum, collection date unknown (isolated by H.D. van Etten, deposited in LARS collection by D.O. TeBeest, No. 403, deposited in CBS collection Apr. 1997 by J.A. Bailey), H.D. van Etten (CBS H-21644 holotype of C. pisicola, culture ex-holotype CBS 724.97 = LARS 60 = ATCC 64197 = IMI 317934).

Notes: Patouillard & Lagerheim (1891) described C. pisi from Pisum sativum in Quito, Ecuador with hyaline, fusoid conidia with acute ends, straight to curved, measuring 11–13 × 3–4 μm and setae measuring 60–90 × 6 μm. Three specimens were located in the BPI fungarium that were collected by G. Lagerheim, but only one was collected in 1891. This specimen, BPI 797146 that also originated from the collection of N. Patouillard is designated as the lectotype of C. pisi in our study. Conidia found on the material are fusiform and mostly ± curved and agree with the original description of the species: (10–)11.5–15(–16.5) × (3–)3.5–4(–4.5) μm, av. ± SD = 13.2 ± 1.8 × 3.7 ± 0.4 μm, L/W ratio = 3.5. Conidia found on BPI 399534 were larger, measuring (10–)14–17.5(–20) × (3–)3.5–4.5 μm, av. ± SD = 15.6 ± 1.8 × 3.8 ± 0.5 μm, L/W ratio = 4.1. Whether the specimens collected by G. Lagerheim represent the same species is doubtful.

Among other C. pisi specimens in BPI were two (BPI 399534, BPI 399536) that originated from Mexico and were intercepted at the border with the USA. The species on the pea pods and seeds of these specimens were identified as C. pisi, although considerably longer conidia were noted; for BPI 399536 the measurements were included in the note on the package as 16–22 × 3–5 μm, which agrees with the observations on strain CBS 724.97 studied here. Hemmi (1921) also observed larger conidia in material on P. sativum in Japan compared to those of the original description. He also considered the fungus as C. pisi, with straight to slightly curved, fusiform conidia with slightly acute ends. This indicates the presence of at least two Colletotrichum species with curved conidia on Pisum sativum.

Conidia of strain CBS 724.97 are larger than those of the lectotype of C. pisi, measuring (11–)15–21(–29.5) × (3–)3.5–4 μm on SNA compared to (10–)11.5–15(–16.5) × (3–)3.5–4(–4.5) μm of C. pisi. The species represented by strain CBS 724.97 is described as new. Regarding conidial size, the specimens from Mexico and Hemmi's Japanese collection resemble C. pisicola rather than C. pisi.

The existence of at least two Colletotrichum species is further supported by the second strain from Pisum sativum included in this study, strain CBS 107.40 from Russia. This strain is deposited in CBS as C. pisi and belongs to a species closely related to C. pisicola (see below Colletotrichum sp. CBS 107.40).

Farr & Rossman (2014) report C. pisi on Pisum sativum in Brazil, China, Canada, USA (Connecticut, Florida, Georgia, Hawaii, Iowa, Idaho, Louisiana, Maine, Minnesota, Texas, Wisconsin), USSR, Guatemala, India and the Malay Peninsula. Hemmi (1921) also reported the species to be common on P. sativum in Japan. Further species reported on P. sativum (or P. arvense) include C. dematium from Barbados and Mexico, C. falcatum from Hawaii, C. gloeosporioides from China, India and USA (North Carolina), C. lindemuthianum from Chile, China and Poland, C. truncatum from Pakistan and the USA and Colletotrichum sp. from Brazil, Malaysia and the USA (Oregon). Hagedorn (1974) reports widespread and serious local damage by pea anthracnose in Wisconsin, USA. We cannot prove which of these reports actually refer to C. pisicola as there are no isolates available.

Strain CBS 724.97 was regarded as C. truncatum e.g. by Sherriff et al. (1994), Shen et al. (2001) and Latunde-Dada & Lucas (2007) and is included in the ATCC collection as C. dematium f. truncatum. Based on information in the CBS strain database, this strain was also previously identified as C. destructivum and as C. pisi.

The two Colletotrichum strains from Pisum represent basal species in the C. destructivum complex. This was also observed in preliminary LSU and ITS phylogenies of the genus Colletotrichum, in which they formed a sister clade to the other species in this complex (U. Damm, unpubl. data). Consequently, the two species were chosen as outgroup in the phylogeny of the species complex in this study. Their morphological features are not typical for this complex: conidia at least of C. pisicola are curved. However, O'Connell et al. (1993) investigated the hemibiotrophic infection of Pisum sativum by strain LARS 60 (= CBS 724.97, C. pisicola) with light and electron microscopy. Both the biotrophic phase and primary hyphae of this fungus were confined to the first infected epidermal cell, but these hyphae were less bulbous and more convoluted than those reported for other members of the C. destructivum species complex.

The identification of a strain from roots of Salix as C. pisi from Corredor et al. (2012) based on a blastn search on GenBank with its ITS sequence (Genbank GU934514) is based on another apparently wrongly identified strain, DAOM 196850 (Chen et al. 2007), that is not a Colletotrichum species; its ITS sequence (GenBank EU400150) is identical to several Plectosphaerella cucumerina strains.

Colletotrichum pisicola is characterised by distinctly curved conidia that gradually taper to the ± acute ends, short and few pale brown setae with rounded tips. Strain CBS 724.97 is the slowest growing culture in the species complex studied.

The sequences of all loci studied of C. pisicola strain CBS 724.97 are unique; there is on CHS-1 only a single nucleotide difference to Colletotrichum sp. strain CBS 107.40 from Pisum in Russia (see Colletotrichum sp. CBS 107.40). No ITS sequences with >98 % identity (9 nucleotides different) and no TUB2 sequence with >91% identity were found in GenBank.

Colletotrichum sp. CBS 107.40

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1–8 μm diam, hyaline, smooth-walled, septate, branched. Chlamydospores not observed. Conidiomata, conidiophores, conidiogenous cells and Setae not observed. No sporulation. Appressoria single, scattered, pale brown, smooth-walled, ellipsoidal, clavate to navicular outline, with an entire or undulate margin, (4.5–)6.5–12(–15) × (3.5–)4.5–7.5(–9.5) μm, av. ± SD = 9.2 ± 2.7 × 5.9 ± 1.5 μm, L/W ratio = 1.6.

Asexual morph on Anthriscus stem. Conidiomata, conidiophores and setae formed on pale to medium brown, roundish to angular cells, 4–11.5 μm diam. Chlamydospores not observed. Conidiomata absent, conidiophores and setae formed directly on hyphae. Setae not observed. Conidiophores and conidiogenous cells not observed. Conidia only few observed, hyaline, smooth-walled, aseptate, ± curved, with slightly acute ends, 13–16.5 × 3.5–4 μm, mean = 14.9 × 3.8 μm, L/W ratio = 4.0.

Culture characteristics: Colonies on SNA flat with entire margin, hyaline, agar medium, filter paper and Anthriscus stem partly covered with sparse aerial mycelium, reverse same colours; growth 12.5–14.5 mm in 7 d (18.5–20.5 mm in 10 d). Colonies on OA flat with entire margin; greenish olivaceous to citrine, with a straw margin, aerial mycelium lacking, reverse straw to greenish olivaceous, growth 17.5–19 mm in 7 d (25–28.5 mm in 10 d). Conidia in mass not visible.

Material examined: Russia, Omsk, from Pisum sativum, collection date and collector unknown (deposited in CBS collection Feb. 1940 by K. Murashkinsky), CBS H-21645, culture CBS 107.40.

Notes: Stain CBS 107.40 from peas in Russia was deposited as Macrophoma sheldonii in CBS by K. Murashkinsky. This species was described by Rodigin (1928) from seeds of Pisum sativum in Russia as forming cylindrical-ovate, thick-walled conidia, measuring 10–18 × 5–6 μm that are mass pink and formed in spherical to flattened pycnidia. This species, if a Colletotrichum species at all, is not the same species as strain CBS 107.40, as conidial shapes and sizes are different. The spherical “pycnidia” could refer to the closed conidiomata that have been observed in species of the C. boninense species complex, e.g. C. dacrycarpi and C. karstii (Damm et al. 2012). Macrophoma sheldonii was regarded as a synonym of C. lagenarium by Vassiljevski & Karakulin (1950) and of C. orbiculare by von Arx (1957). Since we have not seen type material of this fungus, we cannot confirm this species as a Colletotrichum sp.

After the strain was deposited in CBS, it was re-identified as C. pisi. The strain was also treated as C. pisi by Nirenberg et al. (2002), who submitted an ITS sequence to GenBank (GenBank AJ301940). The conidia of strain CBS 107.40 are shorter than those of C. pisicola strain CBS 724.97, and more similar to C. pisi than those of C. pisicola (newly described in this study). However, we refrain from using this strain to epitypify C. pisi, because the strain is degenerated, the sporulation almost suppressed, and only four conidia were observed that might not be typical of the species. Moreover, the strain was from a different continent than C. pisi.

The sequences of all loci studied are unique for this species, and different from those of C. pisicola strain CBS 724.97; however, the CHS-1 sequence differs in only one nucleotide from that of C. pisicola.

There is no match with sequences >97 % identical to our ITS sequence and no match with sequences >89 % identical to our TUB2 sequence in GenBank.

Colletotrichum tabacum Böning, Prakt. Blätt. Pflanzenbau Pflanzenschutz 10: 89. 1932. Fig. 12.

Open in a separate window

Fig. 12

Colletotrichum tabacum (from ex-neotype strain N150). A–B. Conidiomata. C, H. Tip of a seta. D, I. Base of a seta. E–G, J–M. Conidiophores. N–S. Appressoria. T–U. Conidia. A, C–G, T. from Anthriscus stem. B, H–S, U. from SNA. A–B. DM, C–U. DIC, Scale bars: A = 100 μm, E = 10 μm. Scale bar of A applies to A–B. Scale bar of E applies to C–U.

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1–7 μm diam, hyaline, smooth-walled, septate, branched. Chlamydospores not observed. Conidiomata absent, conidiophores and setae formed directly on hyphae, sometimes also on few pale to medium brown, roundish cells, 5–10.5 μm diam. Setae pale to medium brown, smooth-walled, 65–150 μm long, 1–4-septate, base cylindrical to conical, 4–5.5 μm diam, tip ± rounded to ± acute. Conidiophores hyaline to pale brown, smooth-walled, septate, branched, to 50 μm long. Conidiogenous cells hyaline to pale brown, smooth-walled, cylindrical to ampulliform, 9–22.5 × 3–4.5 μm, opening 1–1.5 μm diam, collarette 0.5–1 μm long, periclinal thickening observed. Conidia hyaline, smooth-walled, aseptate, narrowly cylindrical, mostly straight, with round ends, one of the ends sometimes very slightly bent to one side, (13.5–)15.5–18.5(–20) × 3–3.5(–4) μm, av. ± SD = 17.0 ± 1.4 × 3.4 ± 0.2 μm, L/W ratio = 5.0, conidia of strain CBS 124249 longer, measuring (16–)17–20(–23.5) × 3–3.5. Appressoria single or in loose groups, medium brown, smooth-walled, clavate, ellipsoidal or irregular outline, with a lobate to undulate margin, with a distinct penetration pore with a dark halo, (7–)8–13(–19) × (4.5–)5.5–8(–10) μm, av. ± SD = 10.4 ± 2.3 × 6.6 ± 1.3 μm, L/W ratio = 1.6.

Asexual morph on Anthriscus stem. Conidiomata, conidiophores and setae formed on a small cushion of hyaline to pale brown, angular cells, 3–6 μm diam. Setae medium brown, smooth-walled to finely verruculose, 55–170 μm long, 1–5-septate, base cylindrical, 3.5–8.5 μm diam, tip ± rounded to ± acute. Conidiophores hyaline to pale brown, single or smooth-walled and septate, branched, to 30 μm long. Conidiogenous cells hyaline to pale brown, smooth-walled, cylindrical to doliiform, 7–16.5 × 3.5–5 μm, opening 1–1.5 μm diam, collarette 0.5–1.5 μm long, periclinal thickening distinct. Conidia hyaline, smooth-walled, aseptate, narrowly cylindrical, mostly straight, with round ends, one of the ends sometimes very slightly bent to one side, (16.5–)17–19(–21) × (3–)3.5(–4) μm, av. ± SD = 17.8 ± 1.0 × 3.5 ± 0.2 μm, L/W ratio = 5.1.

Culture characteristics: Colonies on SNA flat with entire margin, hyaline to cinnamon, aerial mycelium lacking, reverse same colours; growth 25–26.5 mm in 7 d (>40 mm in 10 d). Colonies on OA flat with entire margin; isabelline, honey, buff to rosy buff, aerial mycelium lacking, reverse buff, vinaceous buff, hazel to pale olivaceous-grey, colonies of strain CBS 124249 differed slightly on OA: colonies buff, almost entirely covered with honey, grey to black acervuli and partly covert with white short filty aerial mycelium, reverse buff, honey to olivaceous-grey, growth 27.5–29 mm in 7 d (>40 mm in 10 d). Conidia in mass salmon, conidia of strain CBS 161.53 in mass whitish.

Materials examined: France, from Nicotiana tabacum, collection date and collector unknown (received from R. O'Connell, before from P. Goodwin, before from M. Maurhofer Bringolf, originally from Novartis as Novartis Isolate 150) (CBS H-21669 neotype here designated, MBT178524, culture ex-neotype N150 = CPC 18945). Germany, Middle Franconia, from leaves of Nicotiana rustica, holotype, presumably lost. India, Rajahnundry, from Nicotiana tabacum, collection date unknown, B. S. Kadam, culture IMI 50187 = CPC 16820. Madagascar, from Centella asiatica, collection date and collector unknown (isolated by Rakotoniriana F. 2003), CBS H-21668, culture CBS 124249 = MUCL 44942. Zambia, from Nicotiana tabacum, collection date and collector unknown (send to CBS collection Nov. 1953 from Mt. Makulu Research St., Zambia), CBS H-21667, culture CBS 161.53.

Notes: In the late 1920s anthracnose of tobacco, especially Nicotiana rustica, was observed in Middle Franconia, Germany. The pathogen, C. tabacum, differed morphologically from the previously described C. nicotianae Averna (Böning 1929, 1932). The fungus formed conidia that measured 15–22 × 4–5 μm in small open clusters and setae that were 60–90 μm long (Böning 1929). In contrast, C. nicotianae that was described from stems of N. tabacum in Sao Paulo, Brazil, formed straight to curved conidia that were larger than those of C. tabacum, measuring 19–32.5 × 8–8.6 μm and turn yellow with age, and setae that were 60–175 × 8.5 μm long and 3–5-septate (Averna-Saccá 1922). Colletotrichum tabacum forms distinct spots with necrotic centres on leaves, stems, flowers and seeds and also causes a seedling disease of tobacco (Böning 1929). The microscopical features of the isolates studied here agree with C. tabacum, although the conidia are slightly smaller than those observed by Böning (1929). Böning (1929, 1932) did not designate a type, and no type or authentic material could be located in any fungarium.

Shortly after, an additional species was described by Böning (1933), Gloeosporium nicotianae that caused blisters and diffuse browning on leaf surfaces of N. rustica in Königsberg, East Prussia (today Kaliningrad, Russia), consistently lacked setae and also exhibited different cultural characteristics. Colletotrichum tabacum formed greenish black cultures with a uniform grey aerial mycelium vs. Gm. nicotianae with slightly brownish cultures and floccose aerial mycelium. Conidia of Gm. nicotianae are on average smaller than those of C. tabacum, measuring 8–18 × 2–5 μm, depending on the substrate and formed swollen, 12 μm diam cells in chains in the mycelium as well as sterile pycnidia- or perithecia-like structures (Böning 1933). Based on the description alone it is difficult to confirm whether C. tabacum and Gm. nicotianae are different species.

Lucas & Shew (1991) concluded C. nicotiniae and C. tabacum were synonyms of C. gloeosporioides. This was probably based on von Arx (1957), who listed C. tabacum as synonym of C. gloeosporioides. Farr & Rossman (2014) cited various reports of C. nicotianae, C. tabacum, C. destructivum, C. coccodes, C. gloeosporioides and Colletotrichum sp. from tobacco around the world. One of the studies cited (Barksdale 1972) includes a picture and measurements of conidia of C. destructivum from tobacco that resemble those of C. tabacum. Isolates from Nicotiana used in molecular studies of pathogen-host-interactions are either called C. nicotianae, or C. destructivum (e.g. Chen et al. 2003; Yang et al. 2010). Based on rDNA ITS sequences and morphology, Shen et al. (2001) identified strain N150 (here re-identified as C. tabacum) as C. destructivum. As the isolates studied here were previously identified as C. destructivum, C. higginsianum, C. gloeosporioides or C. tabaci, many of the reports listed by Farr & Rossman (2014) might actually refer to C. tabacum. The few isolates of C. tabacum included in this study already represent the occurrence of the species on three continents. But to our knowledge, there is no report listed from Germany since Böning (1933).

Shen et al. (2001) discovered the intracellular hemibiotrophic infection process of C. destructivum (here re-identified as C. tabacum) strain N150 on tobacco. Shan & Goodwin (2004, 2005) used a GFP-expressing transgenic strain of this fungus to study rearrangement of host actin microfilaments and nuclei around biotrophic hyphae. Secondary metabolite production by C. tabacum (ATCC 11995) was extensively studied by Gohbara and co-workers during the 1970s, leading to the identification and structural characterisation of two novel terpenoid phytotoxins, colletotrichin and colletopyrone (Gohbara et al. 1976, 1978).

One of the strains included in this study, CBS 124249 (= MUCL 44942) was isolated by F. Rakotoniriana from Centella asiatica in Madagascar and identified as C. higginsianum (Rakotoniriana et al. 2008). It is re-identified as C. tabacum in this study. Rakotoniriana et al. (2013) recently described a species from Centella asiatica in Madagascar, C. gigasporum that forms larger conidia than C. tabacum and belongs to the C. gigasporum complex (Liu et al. 2014), confirming that more Colletotrichum species occur on this host in Madagascar.

Conidia of C. tabacum are narrowly cylindrical with round ends, one of the ends sometimes slightly bent to one side; the conidia still appearing straight. Appressoria with a distinct penetration pore with a dark halo were observed.

Colletotrichum tabacum is distinguished from the other species in the C. destructivum complex by all loci studied, but sequences of some loci only differ with a single nucleotide from its closest relative. Strain CBS 124249 from Centella differs additionally in CHS-1 and TUB2 sequences from the other three strains, but intraspecific variability was also observed with ITS, GAPDH and ACT.

The closest match in a blastn search with the TUB2 sequence of strain N150 with 100 % identity was C. tabacum strain CBS 161.53 (GenBank JQ005847, O'Connell et al. 2012). No GAPDH sequence with <93 % identity was found in GenBank.

Colletotrichum tanaceti M. Barimani, et al., Plant Pathol. 62: 1252. 2013. Fig. 13.

Open in a separate window

Fig. 13

Colletotrichum tanaceti (from ex-holotype strain CBS 132693). A–B. Conidiomata. C. Tip of a seta. D, G–I. Conidiophores. E. Base of a seta and conidiophores. F. Seta. J–N. Appressoria. O–P. Conidia. A, C–E, O. from Anthriscus stem. B, F–N, P. from SNA. A–B. DM, C–P. DIC, Scale bars: A = 100 μm, D = 10 μm. Scale bar of A applies to A–B. Scale bar of D applies to C–P.

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1–10 μm diam, hyaline, smooth-walled, septate, branched. Chlamydospores not observed. Conidiomata absent, conidiophores and setae formed directly on hyphae. Setae medium brown, smooth-walled to verruculose, 40–140 μm long, 2–4-septate, base cylindrical to conical, 4–5.5 μm diam, tip rounded to slightly acute. Conidiophores, smooth-walled, septate, branched, to 75 μm long. Conidiogenous cells hyaline, smooth-walled, sometimes extending to form new conidiogenous loci, 15–28 × 3.5–4.5 μm, opening 1–2 μm diam, collarette 1 μm long, periclinal thickening distinct. Conidia hyaline, smooth-walled, aseptate, cylindrical to slightly clavate, slightly but distinctly curved with both ends ± rounded, (13–)14.5–17.5(–19) × (3–)3.5–4(–4.5) μm, av. ± SD = 16.0 ± 1.5 × 3.8 ± 0.3 μm, L/W ratio = 4.2. Appressoria single or in loose groups, medium brown, smooth-walled, subglobose, to elliptical in outline, with an entire or undulate margin, (5–)6.5–12(–14.6) × (3.5–)4.5–7(–10) μm, av. ± SD = 9.1 ± 2.7 × 5.7 ± 1.4 μm, L/W ratio = 1.6, appressoria of stem CBS 132818 are slightly larger, measuring (7.5–)8.5–13.5(–16) × (5–)5.5–9(–12) μm, av. ± SD = 11.0 ± 2.5 × 7.4 ± 1.8 μm, L/W ratio = 1.5.

Asexual morph on Anthriscus stem. Conidiomata, conidiophores and setae formed on hyaline to pale brown, angular cells, 3–7.5 μm diam. Setae medium brown, smooth-walled to finely verruculose, 30–165 μm long, 1–4-septate, base cylindrical to conical, 4–7 μm diam, tip rounded to slightly acute. Conidiophores hyaline to pale brown, smooth-walled, septate, branched, to 50 μm long. Conidiogenous cells hyaline to pale brown, smooth-walled, cylindrical, sometimes extending to form new conidiogenous loci, 18–28 × 4–5 μm, opening 1–1.5 μm diam, collarette 0.5 μm long, periclinal thickening distinct. Conidia hyaline, smooth-walled, aseptate, cylindrical, slightly but distinctly curved with both ends ± rounded or one end ± acute, (12–)16–20.5(–22) × (3–)3.5–4(–4.5) μm, av. ± SD = 18.1 ± 2.1 × 3.7 ± 0.3 μm, L/W ratio = 4.9.

Culture characteristics: Colonies on SNA flat with entire margin, hyaline to pale isabelline, filter paper partly yellow, aerial mycelium lacking, reverse same colours; growth 14–16 mm in 7 d (22.5–25 mm in 10 d). Colonies on OA flat with entire margin, buff to straw, partly covered with tiny grey to black acervuli, aerial mycelium lacking, reverse olivaceous-grey, growth 14.5–17 mm in 7 d (21.5–24.5 mm in 10 d). Conidia in mass whitish to rosy-buff.

Materials examined: Australia, northern Tasmania, Scottsdale, from anthracnose on leaves of Tanacetum cinerariifolium, Aug. 2010, S.J. Pethybridge, culture ex-holotype CBS 132693 = BRIP 57314 = UM01; Australia, northern Tasmania, Ulverstone, from Tanacetum cinerariifolium, collection date unknown, S.J. Pethybridge, living strain CBS 132818 = BRIP 57315 = TAS060-0003.

Notes: Pyrethrum (Tanacetum cinerariifolium, Asteraceae) is a perennial plant grown for the extraction of pyrethrin insecticides in Australia, mainly in Tasmania, one of the largest producers of pyrethrin worldwide (Greenhill 2007). Colletotrichum tanaceti was recently described as an anthracnose pathogen of pyrethrum in Tasmania and revealed to be closely related to C. destructivum, C. higginsianum and C. panacicola (Barimani et al. 2013). This species can be confirmed as distinct in this study, and can be identified with all loci studied.

Additionally, C. tanaceti is one of the two species in this complex with distinctly curved conidia. In contrast to C. pisicola, the conidia are more abruptly tapered towards mostly rounded ends. In both media, conidiogenous cells were observed that extended to form new conidiogenous loci (Fig. 14E, H), a feature common for species in the C. boninense species complex (Damm et al. 2012) but not elsewhere in the C. destructivum complex.

Open in a separate window

Fig. 14

Colletotrichum utrechtense (from ex-holotype strain CBS 130243). A–B. Conidiomata. C, F. Tip of a seta. D, H–J. Conidiophores. E. Bases of setae and conidiophores. G. Base of a seta. K–P. Appressoria. Q–R. Conidia. A, C–E, Q. from Anthriscus stem. B, F–P, R. from SNA. A–B. DM, C–R. DIC, Scale bars: A = 100 μm, D = 10 μm. Scale bar of A applies to A–B. Scale bar of D applies to C–R.

Our conidia measurements differ from those given in the study of Barimani et al. (2013). In that study, conidia on pyrethrum tissue measured on average 30.9 × 5.6 μm, and those on SNA, 22.5 × 4.1 μm. In contrast, conidia of the same strain on SNA measured in our study on average 16.0 × 3.8 μm.

This fungus formed perithecia in a mating experiment and is apparently heterothallic (Barimani et al. 2013). The sexual morph was described by Barimani et al. (2013) as follows “Perithecia dark brown, ampulliform with setaceous hairs in ostiole, becoming erumpent through the epidermis, perithecia ostiolate measuring 33 × 31 μm in diameter, individual locules measuring 200 × 380 μm (length × width), thick-walled texture. Asci 89.6 ± 2.9 × 10.9 ± 0.4 μm (n = 30), unitunicate, thin-walled, clavate or cymbiform, stipitate, 8–10 spored. Ascospores (18–)21.5–22.5(–26.5) × (4–)5.5–6(–7) μm (n > 50), av. ± SD = 22 ± 1.7 × 5.8 ± 0.7 μm, one-celled, hyaline, smooth, becoming septate through germination, fusiform and blunt at both ends (widest at middle and narrower at the ends) or widest at middle and upper third, many formed within 2 months.”

Barimani et al. (2013) also studied the infection strategy, which they suggested to be intracellular hemibiotrophic, similar to that of C. destructivum and C. higginsianum.

Colletotrichum utrechtense Damm, sp. nov. MycoBank MB809404. Fig. 14.

Etymology: The species epithet is derived from the place where it was collected, Utrecht, the Netherlands.

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1–7.5 μm diam, hyaline to pale brown, smooth-walled, septate, branched. Chlamydospores not observed. Conidiomata absent, conidiophores and setae formed directly on hyphae. Setae medium brown, smooth-walled to finely verruculose, 95–180 μm long, 2–5-septate, base cylindrical to ± inflated, 3–6.5 μm diam, tip ± rounded to slightly acute. Conidiophores hyaline to pale brown, smooth-walled, septate, branched, to 70 μm long. Conidiogenous cells hyaline to pale brown, smooth-walled, cylindrical to ± inflated, 13–26 × 3–4.5 μm, opening 1–1.5 μm diam, collarette 1–1.5 μm long, periclinal thickening distinct. Conidia hyaline, smooth-walled, aseptate, straight to slightly curved, with both ends ± rounded, 17.5–20.5(–23) × 3.5–4(–4.5) μm, av. ± SD = 19.0 ± 1.4 × 4.0 ± 0.2 μm, L/W ratio = 4.8. Appressoria single, sometimes in clusters of two, medium brown, smooth-walled, navicular, ellipsoidal or irregular in outline, with an lobate or undulate margin, (7–)10–14.5(–15) × (5–)6.5–9.5(–10) μm, av. ± SD = 12.2 ± 2.1 × 8.0 ± 1.5 μm, L/W ratio = 1.5, appressoria of strain CBS 135827 smaller, measuring (6.5–)7.5–13.5(–19) × (3.5–)4.5–7(–9) μm, av. ± SD = 10.5 ± 3.0 × 5.7 ± 1.3 L/W μm, ratio = 1.8.

Asexual morph on Anthriscus stem. Conidiomata absent, conidiophores and setae formed directly on hyphae, or rarely on pale brown, angular cells, 3.5–8 μm diam. Setae medium brown, basal cell pale brown, smooth-walled to finely verruculose, 75–255 μm long, 2–4-septate, base ± inflated or cylindrical, 3.5–8.5 μm diam, tip slightly rounded to slightly acute. Conidiophores hyaline to pale brown, smooth-walled, simple or septate and branched, to 20 μm long. Conidiogenous cells hyaline to pale brown, smooth-walled, ellipsoidal to cylindrical, 9–17 × 4–5.5 μm, opening 1–2 μm diam, collarette 1–2 μm long, periclinal thickening distinct. Conidia hyaline, smooth-walled, aseptate, straight to slightly curved, with both ends ± rounded, (16.5–)18–20(–21.5) × 3.5–4 μm, av. ± SD = 19.0 ± 1.0 × 3.7 ± 0.2 μm, L/W ratio = 5.2.

Culture characteristics: Colonies on SNA flat with entire margin, hyaline, pale cinnamon in the centre, filter paper partly pale olivaceous grey, Anthriscus stem partly covert with filty white aerial mycelium, reverse same colours; 20–24 mm in 7 d (35–36.5 mm in 10 d). Colonies on OA flat with entire margin; buff, pale cinnamon, pale olivaceous grey to olivaceous grey, with few patches of floccose, whitish, aerial mycelium, reverse same colours, 22.5–25 mm in 7 d (33.5–39 mm in 10 d). Conidia in mass whitish to very pale salmon.

Materials examined: Netherlands, Utrecht, from a leaf of Trifolium pratense, 13 Jun. 2011, U. Damm (CBS H-21662 holotype, culture ex-holotype CBS 130243); Utrecht, from a leaf of T. pratense, 13 Jun. 2011, U. Damm, culture CBS 135827; Utrecht, from a leaf of T. pratense, 13 Jun. 2011, U. Damm, culture CBS 135828.

Notes: This species is only known from Trifolium pratense in the Netherlands. Other Colletotrichum species described from this host are reviewed in the notes under C. destructivum.

The CHS-1, HIS3 and TUB2 sequences are different from all species included. The ACT sequences are the same as that of C. panacicola; ITS and GAPDH distinguishes the species from C. panacicola but the ITS is identical with the unnamed isolates from Heracleum, while the GAPDH sequence is the same as that of C. higginsianum and the isolates from Heracleum and Matthiola.

In blastn searches the ITS and GAPDH sequences of strain CBS 130243 were found to be identical to the ITS sequence of “C. coccodes” strain BBA 71527 from Lupinus in Germany (GenBank AJ301984, Nirenberg et al. 2002) and the GAPDH sequences of C. higginsianum isolates C97027 and C97031 from Brassica and Raphanus probably from Korea (GenBank GU935850, GU935851, Choi et al. 2011). Closest matches in blastn searches with the TUB2 sequences of strain CBS 130243 with 99 % identity (3 nucleotides different) were C. fuscum CBS 130.57 (GenBank JQ005846, O'Connell et al. 2012) and Colletotrichum isolates from a study on ramie (Boehmeria nivea) anthracnose in China (GenBank JF811024–JF811028, W.X. Xia, unpubl. data).

Colletotrichum vignae Damm, sp. nov. MycoBank MB809405. Fig. 15.

Open in a separate window

Fig. 15

Colletotrichum vignae (from ex-holotype strain CBS 501.97). A–B. Conidiomata. C, G. Tip of a seta. D–E. Conidiophores. F. Base of a seta and conidiophores. H. Base of a seta. I–L. Conidiophores. M–R. Appressorium-like structures. S–T. Conidia. A, C–F, S. from Anthriscus stem. B, G–R, T. from SNA. A–B. DM, C–T. DIC, Scale bars: A = 100 μm, F = 10 μm. Scale bar of A applies to A–B. Scale bar of F applies to C–T.

Etymology: The species epithet is derived from the host genus name Vigna.

Sexual morph not observed. Asexual morph on SNA. Vegetative hyphae 1–8 μm diam, hyaline, smooth-walled, septate, branched. Chlamydospores not observed. Conidiomata absent, conidiophores and setae formed directly on hyphae. Setae hyaline to very pale brown, smooth-walled, wall up to 0.8 μm wide, 30–90 μm long, 1–3-septate, base cylindrical to conical, 3–4.5 μm diam, tip rounded to ± acute. Conidiophores hyaline, sometimes pale brown, smooth-walled, septate, branched, to 35 μm long. Conidiogenous cells hyaline, sometimes pale brown, smooth-walled, cylindrical, 12–25 × 3–5 μm, polyphialides observed, opening 1–1.5 μm diam, collarette 0.5–2 μm long, periclinal thickening sometimes observed. Conidia hyaline, smooth-walled, aseptate, old conidia sometimes septate, cylindrical, straight to slightly curved, with one end round and the other truncate, (12–)14–17.5(–18.5) × (3–)3.5–4(–4.5) μm, av. ± SD = 15.8 ± 1.6 × 3.8 ± 0.3 μm, L/W ratio = 4.2. Appressoria not observed on the undersurface of the medium. Appressoria-like structures that possibly function as chlamydospores were observed within the medium. These are single or in dense clusters, medium brown, smooth-walled, ellipsoidal, subglobose to clavate outline, with an entire or undulate margin, because not attached to any surface (4–)4.5–8.5(–12.4) × (3.5–)4–5(–6.5) μm, av. ± SD = 6.6 ± 2.0 × 4.6 ± 0.6 μm, L/W ratio = 1.4.

Asexual morph on Anthriscus stem. Conidiomata, conidiophores and setae formed on pale brown, angular cells, 2.5–8 μm diam. Setae pale to medium brown, smooth-walled to verruculose, very thick-walled (up to 1.5 μm wide), 40–120 μm long, 1–3-septate, base conical to cylindrical, 5–8.5 μm diam, tip rounded to slightly acute. Conidiophores hyaline, smooth-walled, septate, branched, to 60 μm long. Conidiogenous cells hyaline to pale brown, smooth-walled, cylindrical, 8–35 × 3.5–4(–6.5) μm, opening 1–1.5 μm diam, collarette 0.5–1 μm long, periclinal thickening visible. Conidia hyaline, smooth-walled, aseptate, cylindrical, straight to slightly curved, with one end round to slightly acute and the other truncate, (10–)12–16.5(–21.5) × 3.5–4 μm, av. ± SD = 14.2 ± 2.2 × 3.8 ± 0.2 μm, L/W ratio = 3.8. conidia of strain IMI 334960 are shorter, measuring (8–)9–13(–15.5) × (3.5–)4–4.5(–5) μm, av. ± SD = 11.0 ± 1.8 × 4.3 ± 0.3 μm, L/W ratio = 2.6.

Culture characteristics: Colonies on SNA flat with entire margin, hyaline, agar medium, filter paper and Anthriscus stem partly covered with saffron to cinnamon acervuli, aerial mycelium lacking, reverse same colours; growth 12.5–15 mm in 7 d (19–22.5 mm in 10 d). Colonies on OA flat with entire margin; honey to cinnamon, with a buff margin, aerial mycelium lacking, reverse same colours; growth 13.5–15 mm in 7 d (20–21.5 mm in 10 d). Colours and growth rate of strain IMI 334960 differed on OA by being dark grey olivaceous, partly covered with short white aerial mycelium, reveres dark grey olivaceous to olivaceous grey; growth 17–18.5 mm in 7 d (24.5–26 mm in 10 d). Conidia in mass saffron.

Materials examined: Nigeria, from Vigna unguiculata, collection date unknown (deposited in CBS collection Feb. 1997 by J.A. Bailey, isolated by R.A. Skipp, No. I 57), R.A. Skipp (CBS H-21648 holotype, culture ex-type CBS 501.97 = LARS 56); from Vigna unguiculata, collection date and collector unknown, culture IMI 334960 = CPC 19383.

Notes: The isolates studied here apparently originate from a study on cowpea diseases in Nigeria by Williams (1975), who sent isolates to IMI where they were identified as C. lindemuthianum. In contrast to the species studied here, C. lindemuthianum belongs to the C. orbiculare species complex (Damm et al. 2013, Liu et al. 2013a). Judging from information retrieved from Bailey et al. (1990) these two strains originated from the same isolate.

Bailey et al. (1990) observed the single-cell hemibiotrophic infection of cowpea by C. lindemuthianum from cowpea (= C. vignae) for the first time. They also revealed the morphology, pathogenicity and host specificity of strain I57 (= LARS 56 = CBS 501.97) to be different from C. lindemuthianum isolates from Phaseolus vulgaris. Latunde-Dada et al. (1996, 1999) studied the infection of cowpea by the same strain and by strain LARS 860, another strain from cowpea in Nigeria. They identified the species as C. destructivum based on similarity of morphological features and the ITS2-D2 sequences with isolates from Medicago that were confirmed as C. destructivum s. str. in our study. However, based on the ITS2-D2 phylogeny of the second study strain, LARS 860 is a different species to LARS 56 (C. vignae), not belonging to the C. destructivum complex and more closely related to C. gloeosporioides (s. lat.) strains; the infection process differs considerably and is not hemibiotrophic.

Takimoto (1934) described C. phaseolorum from Vigna angularis and V. sinensis in Japan. The type of this species was not designated. Authentic isolates from the two hosts studied by Damm et al. (2009) are not conspecific, but neither is closely related to C. vignae. In contrast to C. vignae, this species forms distinctly curved conidia. Further isolates from Vigna, from V. unguiculata in Burkina Faso and V. sinensis in Pakistan, respectively, were recently identified as C. truncatum in the same multilocus analyses (Damm et al. 2009). Colletotrichum phaseolorum was treated as a synonym of C. gloeosporioides by von Arx (1957, wrongly cited as C. phascorum). Shen et al. (2010) reported anthracnose of mung bean (V. radiata) sprouts to be caused by C. acutatum (s. lat.) in Taiwan.

Glomerella vignicaulis was described by Tehon (1937) on Vigna sinensis in Illinois, USA. Tehon never found an asexual Colletotrichum morph on the host; however, a Cercospora stage accompanying the perithecia that appeared to arise from the same mycelium was always observed. If Ga. vignicaulis is a Colletotrichum species at all, it is unlikely to belong to the C. destructivum species complex, as in this complex conidia are dominating; none of the species is known to form a sexual morph in nature. The two species that are known to form a sexual stage, C. lentis (= Ga. truncata) and C. tanaceti, are apparently heterothallic and sexual morphs were only observed by crossing experiments in the laboratory (Armstrong-Cho & Banniza 2006, Barimani et al. 2013).

Colletotrichum vignae was one of the slowest growing species studied in the C. destructivum complex and the slowest growing species within the first main clade (Fig. 1). Conidia of C. vignae are highly variable in length; the few setae observed were pale brown and thick-walled.

This species can be identified by its ITS, GAPDH, HIS3 and ACT sequences. Blastn searches with the respective sequences of strain CBS 501.97 resulted in 99 % identity (a single nucleotide difference) with the ITS sequence of C. fuscum strain DAOM 216112 (GenBank EU400144; Chen et al. 2007), and 98 % identity with the GAPDH sequences of C. higginsianum isolates C97027 and C97031 from Brassica and Raphanus, respectively, probably from Korea (GenBank GU935870 and GU935873; Choi et al. 2011), and 99 % identity (2 nucleotides difference) with the HIS3 sequence of C. higginsianum isolate MAFF 305635 (GenBank JQ005803; O'Connell et al. 2012, included in this study) and 99 % identity (a single nucleotide difference) with the ACT sequences of C. fuscum CBS 130.57 (GenBank JQ005825; O'Connell et al. 2012, included in this study), respectively. The CHS-1 sequences are the same as those of C. fuscum, C. higginsianum, C. antirrhinicola and the unnamed isolates from Heracleum and Matthiola.

Sun & Zhang (2009) isolated Colletotrichum from anthracnose lesions on leaves of cowpea in China that they identified as C. destructivum based on morphology. As the ITS sequences were the same as those from cruciferous hosts, they concluded C. higginsianum to be a synonym of C. destructivum. The ITS sequence from those strains, however, differed in 4 nucleotides from those of C. vignae.