In Mexico, roselle (Hibiscus sabdariffa L.) is grown 20,061 ha yearly. Guerrero state produces >70% of this crop (SIAP 2016), where a leaf and calyx spot disease, caused by the ascomycete fungus Corynespora cassiicola (Berk. & M.A. Curtis) C.T. Wei, has been observed in recent years. The fungus affects the calyces that are commercialized. On leaves and calyces, up to 100% incidence has been reported (Ortega-Acosta et al. 2015). Corynespora cassiicola causes leaf spot in a large number of plants mainly distributed in tropical and subtropical areas (Dixon et al. 2009). In the United States and Brazil, C. cassiicola has caused serious problems in cotton (Gossypium hirsutum L.), soybean (Glycine max L.), and tomato (Solanum lycopersicum L.) (Galbieri et al. 2014; Sumabat et al. 2017). In Asia and Africa, the fungus is responsible for the main disease on Hevea brasiliensis (Willd. Ex A. Juss.) Müll. Arg. (Deon et al. 2012). In Mexico, C. cassiicola has been recorded affecting cucumber (Cucumis sativus L.), chili (Capsicum annum L.), papaya (Carica papaya L.), Hibiscus sabdariffa L., and Solanum lycopersicum L. (Castro 1979; Ale-Agha et al. 2008; Tun-Suárez et al. 2011; Ortega-Acosta et al. 2015; Farr and Rossman 2017). Several studies have shown that several weeds and cultivated species are hosts of C. cassiicola and they may be important sources of inoculum for seasonal epidemics (Onesirosan et al. 1974; Cutrim and Silva 2003). Knowledge of the host range for C. cassiicola is important for the regional management of disease (Smith et al. 2009). In Guerrero state of Mexico, roselle is cultivated along with maize (Zea mays L.), with few weeding labor that results in the presence of a large number of weeds growing inside and on the periphery of the parcels, some of them surviving all year-round. These plants may be reservoirs of C. cassiicola, the causal agent of spotting of leaves and calyces of roselle. The main objective of this research was to identify alternative host plants of C. cassiicola in the roselle producing area in Guerrero, Mexico, and assess the pathogenicity of the strains isolated from those hosts on roselle.

From May to June 2015, plants exhibiting leaf spot symptoms were collected from within and border of roselle field crops. The collected plants were placed in polyethylene bags, which were labeled and sent to laboratory for processing. Leaf pieces (0.2 cm) were taken from the border of diseased tissues, surface-disinfested using 1% sodium hypochlorite for 2 min, rinsed three times with autoclaved distilled water, dried with sterile absorbent paper, and transferred to Petri dishes with PDA. The plates were incubated for one week at 28 °C (Onesirosan et al. 1974). Whenever a fungal-growth was observed, the colonies were transferred to Petri dishes with fresh PDA. The colonies were purified by monosporic culture technique. From the fungi obtained from different hosts, an isolate was selected for further morphological and molecular identification. The plants which were hosts of Corynespora cassiicola were taxonomically identified and deposited in the Herbario Hortorio CHAPA of the Colegio de Postgraduados.

For each isolate, spores were transferred to glass slides and 50 conidia were measured (length x width). The color, contour and shape of conidia were recorded. Images were acquired with an AmScope® MU 1000 camera attached to a microscope (Nikon Eclipse Ci). Measurements were made with a micrometric ruler in ImageTool® v3.0 image analyzer (Hernández-Livera et al. 2005). Color and shape of colonies were recorded and both morphological and cultural characteristics were compared with literature (Wei 1950; Ellis 1971). One isolate was selected to represent a single host and molecularly identified. Mycelia of C. cassiicola of 12-days old cultures grown in Petri dishes with PDA were scraped from the surface of the agar with a sterile metallic tweezer. Genomic DNA was extracted using AP protocol described by Sambrook and Russel (2001). PCR was performed for the amplification of a segment of the Actin and EF-1α genes using ACT-512F/ACT-783R and EF1-728F/EF1-986R primers, respectively (Carbone and Kohn 1999). The PCR mixture consisted of: 2 μL buffer (1 X), 0.6 μL MgCl2 (2.0 μM), 0.2 μL dNTP’s (0.2 mM), Primer (10 μM) forward 0.6 μL and reverse 0.6 μL, 0.1 μL Taq DNA polymerase (0.5 U) (Promega®) and 10 ng of the DNA extracted from the fungus. The final mixture was adjusted with sterile ultra-pure water to a 10 μL volume. The Eppendorf tubes were placed in a thermocycler (Techne-TC-512®) with the following amplification program for both sets of primers: initial denaturation at 94 °C for 2 min; 40 cycles at 94 °C for 1 min, 58 °C for 1 min and 72 °C for 1 min; with a final extension of 72 °C for 7 min (Shimomoto et al. 2011). Amplified PCR products were purified with Wizard® Genomic DNA Purification Kit (Promega®) and they were sequenced. The sequences were assembled and edited using the SeqMan Pro module from the DNASTAR LaserGene® program. The consensus sequences were compared and deposited at the GenBank.

Isolates were grown on PDA at 28 °C during 7 days. Conidia were collected in a glass objects with autoclaved distilled water and adjusted to 2 × 105 spores/mL (Jin-Hyeuk and Chang-Seuk 2003; Ortega-Acosta et al. 2015). One C. cassiicola isolated, named CCHFR (Access No. KM207768) was utilized as positive control. Plants of roselle cultivar “Criollo de Guerrero” were grown in 15 cm diameter pots during 4.5 months. The inoculum suspension of each isolate considered as a treatment was sprayed with a hand atomizer on ten roselle plants (100 mL/plant) (Jin-Hyeuk and Chang-Seuk 2003). As a negative control, autoclaved distilled water was sprayed on five roselle plants. Isolates were not tested against their respective host of origin. The plants were covered with polyethylene bags during the first 48 h, and then kept under greenhouse conditions. Temperature and relative humidity were recorded with a digital hygrothermograph (Extech®-Rht10). To increase relative humidity an ultrasonic humidifier was placed on the center of the experiment. The average temperature and relative humidity recorded during the study period was 24 °C and 75%, respectively. Twenty days after inoculation the absence or presence of symptoms were registered. Ten fragments of plant symptom tissues from the leaves and calyx were collected from each inoculated isolate (treatment). Re-isolation was carried out using the methodology described above. Fragments of asymptomatic tissue from the control treatment were also inspected.

A total of fifty-four plant species were collected from fields cultivated with roselle and Corynespora cassiicola was isolated from nine of them: S. lycopersicum L., Chromolaena odorata (L.) R.M. King & H. Rob, Senna alata (L.) Roxb, Hyptis suaveolens (L.) Poit, Eugenia oerstediana O. Berg, Passiflora viridiflora Cav., Momordica charantia L., Ricinus communis L. and Gossypium hirsutum L. (Tables 1 and 2). Symptoms observed in these plants differed in leaf spot size and shape (Table 1, Fig. 1).

Table 1 Host plants of Corynespora cassiicola in Guerrero, Mexico
Full size table
Table 2 Host plants of Corynespora cassiicola reported worldwide
Full size table
Fig. 1
figure 1

Leaf spots in different plant species from which C. cassiicola were isolated. A = S. lycopersicum; B = C. odorata; C = S. alata; D = H. suaveolens; E = E. oerstediana; F = P. viridiflora; G = M. charantia; H = R. communis; I = G. hirsutum

Full size image

The fungus growth on PDA showed a wide range of morphological diversity according to the host from where they were isolated. The main characteristics of the colonies were round or in polygon, gray, green or cream color, with abundant aerial mycelium. The conidia were olivaceous to brown, with straight or slightly curved contours; oblong, cylindrical or oval, from 284.62–11.17 μm × 10.52–4.0 μm (length x width), and from 0 to 15 pseudosepta, typical characteristics of C. cassiicola (Wei 1950; Ellis 1971). Based on the GenBank search with the BLAST tool (Zhang et al. 2000; Morgulis et al. 2008), the identity of the sequences of the PCR-amplified fragments (Actin and EF-1α) of the different fungi isolates varied from 99 to 100% with C. cassiicola (Table 3).

Table 3 Comparison of Corynespora cassiicola sequences in GenBank
Full size table

Under greenhouse conditions, all C. cassiicola isolates (including the positive control) caused necrosis in leaves and calyces. The plants sprayed only with sterile distilled water (negative control) showed no symptoms (Figs. 2 and 3, Table 4). The re-isolated fungi of leaves and calyces of inoculated plants that showed symptoms were identified as C. cassiicola based on their cultural and morphological characteristics (Wei 1950; Ellis 1971). No mycelial growth was observed in plant tissues sprayed only with autoclaved distilled water (negative control).

Fig. 2
figure 2

Symptoms on leaves of Hibiscus sabdariffa caused by C. cassiicola isolated from different plant species. Twenty days after inoculation. A = JIT; B = M26; C = GUG; D = CHI; E = CAP; F=PAS; G = MAT; H=HGA; I = ALG; J = CCHFR (positive control); K = Autoclaved distilled water (negative control)

Full size image
Fig. 3
figure 3

Symptoms on calyces of Hibiscus sabdariffa caused by C. cassiicola isolated from different plant species. Twenty days after inoculation. A = JIT; B = M26; C = GUG; D = CHI; E = CAP; F=PAS; G = MAT; H=HGA; I = ALG; J = CCHFR (positive control); K = Autoclaved distilled water (negative control).

Full size image
Table 4 Pathogenicity tests on Hibiscus sabdariffa with ten isolates of Corynespora cassiicola
Full size table

In this study, we found a range of hosts of Corynespora cassiicola, which were able to cause the leaf and calyx spot in roselle. The differences in symptoms in the original host may be due to differences in plant age, infection timing or environmental conditions (Table 1, Fig. 1). In Mexico, C. cassiicola has been reported as a pathogen of cucumber (Cucumis sativus L.) (Castro 1979), chili (Capsium annum L.) (Tun-Suárez et al. 2011), roselle (Hibiscus sabdariffa L.) (Ortega-Acosta et al. 2015) and tomato (Solanum lycopesicum L.) (Farr and Rossman 2017). In our study, all but S. lycopersicum are new reports of hosts of C. cassiicola in Mexico (Table 2). The fungus has been reported affecting Passiflora foetida L. and Passiflora sp., in Guam and Florida, respectively; and on Eugenia uniflora L. in Guam (Alfieri et al. 1984; Dixon et al. 2009). No reports were found on E. oerstediana and P. viridiflora (Table 2). Therefore, these two species of plants are new hosts for C. cassiicola.

In pathogenicity tests, symptoms caused by the different isolates inoculated on roselle plants were similar to previously reported (Ortega-Acosta et al. 2015). Besides, slight variations of symptoms were observed across the isolates, which agrees with reports by Kingsland (1986) who cross-inoculated C. cassiicola isolates obtained from tomato and cucumber. The authors suggested that there is some degree of specificity towards the host of origin. A similar pattern was observed in the present study. With the exception of cotton, all hosts are non-cultivated weeds in the roselle producing area, which grow on the edge or within the field. Cotton is generally grown in family orchards close to roselle fields. Conidia of C. cassiicola are able to disperse by the wind (Jayasinghe 2000; Smith et al. 2009), hence these hosts are potential sources of inoculum for epidemics in roselle crop. Weeds and cultivated plants that act as C. cassiicola reservoirs may favor the emergence of the disease (Onesirosan et al. 1974; Cutrim and Silva 2003). In Nigeria, Onesirosan et al. (1974) conducted studies with C. cassiicola isolates from Aspilia africana (P. Beauv. ex Pers.) C.D. Adams and Lepistemos sp. Weeds. When the pathogen was inoculated on cotton, it caused typical severe symptoms. In Brazil, Cutrim and Silva (2003) isolated C. cassiicola from tomato which was inoculated on two weeds, Commelina benghalensis L., and Vernonia cinerea (L.) Less., where it caused typical lesions. The authors suggested eradication of these weeds as a measure to combat the disease. Qi et al. (2011) inoculated C. cassiicola isolates from different hosts on Hevea brasiliensis and found results similar to our study. Further surveys of C. cassiicola on other weeds and cultivated plant species in the roselle producing areas of Guerrero may be important to expand knowledge of the potential hosts, which should not be grown or eliminated in the vicinities of roselle crops.