Mexico is the 12th largest producer of banana fruit worldwide, providing 2.3 million t annually, of which approximately 25% is exported (FAOSTAT 2018). However, fungal diseases such as Fusarium wilt (Fusarium oxysporum f. sp. cubense), black sigatoka (Pseudocercospora fijiensis), and anthracnose (Colletotrichum spp.) are the main threats to banana production, causing yield losses and reduced fruit marketability.

Anthracnose caused by Colletotrichum musae (Berk. & M.A. Curtis) Arx, a host-specific pathogen, is the most worldwide dangerous disease in the postharvest stage due to the quiescent infections caused by the pathogen during fruit ripening (Bellaire et al. 2007; Su et al. 2011). Nevertheless, other species have also been associated with anthracnose symptoms: Colletotrichum aotearoa B. Weir & P.R. Johnst in Japan and India (Sharma et al. 2015); Colletotrichum chrysophilum W.A.S. Vieira, W.G. Lima, M.P.S. Cãmara & V.P. Doyle, Colletotrichum tropicale Rojas, Rehner & Samuels, Colletotrichum theobromicola Delac., and Colletotrichum siamense Prihast., L. Cai & K.D. Hyde in Brazil (Vieira et al. 2017); Colletotrichum gloeosporioides (Penz.) Penz. & Sacc. sensu stricto in Malaysia and Ecuador (Intan et al. 2013; Riera et al. 2019); Colletotrichum paxtonii Damm, P.F. Cannon & Crous in Saint Lucia (Damm et al. 2012); Colletotrichum plurivorum Damm, Alizadeh & Toy. Sato in Japan (Damm et al. 2019); and Colletotrichum scovillei Damm, P.F. Cannon & Crous in China (Zhou et al. 2017). Furthermore, other species, such as Colletotrichum karsti You L. Yang, Zuo Y. Liu, K.D. Hyde & L. Cai, Colletotrichum gigasporum Rakotonir. & Munaut, and Colletotrichum musicola Damm have been reported in Mexico (Damm et al. 2012, 2019; Liu et al. 2014).

From June to September 2017, banana fruits cv. Tabasco from two orchards in Teapa, Tabasco state, Mexico, with anthracnose symptoms, were sampled in an advanced stage of ripeness. The disease incidence was estimated at around 10%. The sunken necrotic tissue was washed thoroughly with tap water. Small pieces (0.5 cm in length) were cut from the outer margin of the lesions, disinfested in 2% NaClO for 2 min, rinsed three times with distilled sterilized water, and then dried in a biosafety chamber. The tissues were placed on potato dextrose agar (PDA, Bioxon, USA) plates and subsequently incubated in the dark at ~ 25 °C for 8 days. Twenty-seven isolates were obtained according to the monoconidial method (Lim et al., 2002).

Morphological characterization was carried out on PDA media and synthetic nutrient-poor agar (SNA, Nirenberg et al. 1976) amended with pine needles (Liu et al. 2013) previously autoclaved two times to 15 psi for 10 min each. The SNA and PDA plates were incubated at ~ 25 °C under a 12 h fluorescent light/dark photoperiod for 7 and 15 days, respectively. The development of appressoria was induced via the slide culture technique where the size and shape were recorded (Cai et al. 2009). Morphological structures were measured using ImageJ software (https://imagej.nih.gov/ij/index.html) from pictures taken with an Infinity 1–2 camera mounted on a BX51 microscope (Olympus, Japan). On PDA, colonies were a pale brown colour on the upper side, a dark brown in the centre, and a pale brown in the periphery on the reverse side. The growth rate was 5.9 mm day −1 (n = 5). The conidia were hyaline and aseptate with a smooth-walled shape, measuring 17.2–11.4 µm (length) × 5.8–2.9 µm (width) (n = 30), and some of them had a constricted centre. On SNA, cylindrical conidia were hyaline and aseptate, measuring 17.0–11.1 µm × 4.1–2.5 µm (n = 30). The appressoria were solitary, dark brown, and ovoid or pyriform in shape, measuring 11.2–7.2 µm × 6.8–3.9 µm (n = 20) (Fig. 1). The sexual morph was absent, and setae and conidia masses occurred on both media. The size and growth rate of the CPO 27.222, CPO 27.223, and CPO 27.224 strains were similar to those of C. chrysophilum holotype URM89949 in PDA and SNA media but different in conidia shape and cultural characteristics. This difference may be due to the culture medium used (Vieira et al. 2017); or the high intraspecific genetic changes resulting in different morphotypes, as demonstrated in other Colletotrichum species (Damm et al. 2012).

Fig. 1
figure 1

Cultural and morphological characteristics of Colletotrichum chrysophilum. a Upper side of colony growing on potato dextrose agar (PDA) medium, b reverse side of colony growing on PDA, c conidia on PDA, d conidia on synthetic nutrient-poor agar (SNA), e, f, g solitary appressorium. Scale bars = 10 µm

Full size image

For molecular identification, DNA extraction was performed according to the 2% cetyltrimethylammonium bromide (CTAB) method (Doyle and Doyle 1990). PCR mixtures and thermocycler programmes of the annealing temperatures were previously standardized for the six loci amplified (Fuentes-Aragón et al. 2018). The glyceraldehyde-3-phosphate dehydrogenase (GAPDH), chitin synthase (CHS-1), actin (ACT), β-tubulin (TUB2), and glutamine synthetase (GS), as well as the Apn2-MAT1-2 intergenic spacer and mating type MAT1-2 (APMAT) partial genes were amplified and sequenced using the primers GDF1/GDR1 (Guerber et al. 2003), CHS-783F/CHS-248R (Carbone and Kohn 1999), ACT-512F/ACT-783R (Carbone and Kohn 1999), T1/Bt2b (O’Donnell and Cigelnik 1997; Glass and Donaldson 1995), GSF1/GSR1 (Guerber et al. 2003), and AMF1/AMR1 (Silva et al. 2012), respectively. The PCR products were purified with ExoSAP-IT (Affymetrix, USA) and directly sequenced in a 3130 Genetic Analyser (Applied Biosystems, USA) at the facilities of the Postgraduates College in Mexico, following the procedures described by Juárez-Vázquez et al. (2019).

Maximum likelihood (ML) and Bayesian inference (BI) analyses were carried out using the raxmlGUI (Silvestro and Michalak 2012) and MrBayes v.3.2 (Ronquist et al. 2012), respectively. Consensus sequences were obtained using Geneious v.9.1. (Kearse et al. 2012), and alignments were performed via MAFFT (Katoh et al. 2002). A concatenated alignment was achieved with Mesquite v3.6 (Maddison and Maddison 2018). For the ML method, the GTR + G + I nucleotide substitution model was implemented with 1,000 bootstrap repetitions to determine a posterior probability. For BI, two simultaneous runs were executed based on the four Markov chain Monte Carlo methods with 450,000 generations (standard deviation of split frequencies = 0.009120) that were sampled every 1,000 generations. The 25% of the resultant trees were discarded as ‘burn-in’ phase option. Nucleotide substitution models were previously selected for each partition via Jmodel Test (Posada 2008) and implemented in the BI analyses. C. theobromicola CMM4242 was used as an outgroup for ML and BI methods. The trees were visualised in FigTree v1.4.4 (https://tree.bio.ed.ac.uk/software/figtree/).

The concatenated multilocus analyses, performed with GAPDH, CHS-1, ACT, TUB2, GS, and APMAT datasets contained 30 taxa and 3,023 characters composing the matrix. It includes sequences of Colletotrichum ex-type strains retrieved from GenBank, the representative strains generated in this study (CPO 27.222, CPO 27.223, and CPO 27.224), and the Colletotrichum species belonging to the Musae clade (Table 1). For BI analyses, the substitution models implemented were the HKY + I with an invariable site for GAPDH, the SYM + G with gamma distribution rates for CHS-1, the HKY for ACT, the GTR for TUB2, the GTR + G for GS, and the HKY + G for APMAT. This analyses generated 902 trees, of which 25% were discarded as ‘burn-in’ phase, and the posterior probabilities were calculated with the remaining trees (678 trees). The sequences generated in this study clustered into the C. chrysophilum clade, whose posterior probability and bootstrap support values are high (1/100, respectively) (Fig. 2).

Table 1 Colletotrichum species and their gene accession numbers used for phylogenetic analyses in this study
Full size table
Fig. 2
figure 2

Bayesian phylogenetic tree of the Musae clade of the Colletotrichum gloeosporioides species complex generated via concatenate GAPDH, ACT, CHS-1, TUB2, GS, and APMAT sequences. Colletotrichum theobromicola strain CMM4242 isolated from Musa sp. was used as an outgroup. Posterior probability and bootstrap support values (PP/BS) are shown at the nodes. The banana strains from Mexico sequenced in this study are indicated in blue. The scale bar indicates the expected changes per site

Full size image

Koch’s postulate was achieved to evaluate the pathogenicity of the CPO 27.222, CPO 27.223, and CPO 27.224 strains. Five unripe detached banana fruit cv. Tabasco per strain were washed with tap water, disinfested with 70% ethanol for 1 min, followed by 1.5% NaClO for 3 min, and then rinsed three times with distilled sterilized water. The inoculation method consisted of making a wound into the pericarp with a diameter of ~ 3 mm and a depth of 2 mm using a sterile dissection needle. The three strains in the present study were reactivated on fresh PDA plates, which were incubated at ~ 25 °C under a 12 h light/dark photoperiod for 4 days. For inoculation, a mycelial plug 0.5 cm in diameter was taken from the margin of an actively growing culture and placed on the disinfested fruit surface. A PDA plug without mycelium was used as control. The five inoculated fruit and controls were placed in a growth chamber at ~ 25 °C in the dark for 8 days. All inoculated strains were pathogenic on unripe detached banana fruit cv. Tabasco. The inoculated fruit displayed initial necrotic lesions at 4 days after inoculation (dai), and typical black lesions with acervuli and conidia on the fruit surface at 8 dai. The control fruit did not display symptoms (Fig. 3). The strains were reisolated, and the GAPDH partial gene was amplified and sequenced for each strain. The sequences of the inoculated strains were identical to those of the original inoculated.

Fig. 3
figure 3

Symptoms of anthracnose on banana caused by Colletotrichum chrysophilum strain CPO 27.222 8 days after inoculation

Full size image

Colletotrichum chrysophilum was recently considered a new species in Brazil, causing anthracnose on banana (Vieira et al. 2017). Later, it was reported on Anacardium occidentale and Anacardium humble in the same country (Veloso et al. 2018), and on Mangifera indica in Mexico (Fuentes-Aragón et al. 2020). In conclusion, this species was demonstrated for the first time to cause anthracnose on banana fruit in Mexico. These results provide a basis for studies on the epidemiology and management of this species, contributing to the knowledge of the host range of C. chrysophilum.