Mycobiota associated with anthracnose and dieback symptoms on Theobroma cacao L. in Mérida State, Venezuela

ABSTRACT Diverse fungi collected from symptomatic fruit, stem and branch tissues of Theobroma cacao in five T. cacao-producing localities or municipalities of Mérida State, Venezuela, were identified using both morphological methods and sequencing of multiple loci (ITS, LSU, SSU, TEF1, BTUB, RPB2). Cophinforma atrovirens, Lasiodiplodia brasiliensis and Hypoxylon investiens are reported for the first time on T. cacao in Venezuela. Fungi found in association with fruit anthracnose included Cophinforma atrovirens, Fusarium solani and F. oxysporum, whereas species associated with dieback or sudden death symptoms include Cophinforma atrovirens, L. theobromae, L. brasiliensis and H. investiens. All of the aforementioned fungi are considered putative pathogens of T. cacao, which warrant further pathogenicity tests.

The name given by Linnaeus to the genus of Theobroma cacao L.
[theos (God) + broma (beverage) = beverage of the Gods] recognized the Mayan belief that the plant had divine origins (5).Cheesman (4) suggested that T. cacao has been cultivated in Mexico and Central America for over 2,000 years and that no truly wild populations were present in this region, indicating that T. cacao was introduced into Central America and Mexico.It has been hypothesized that T. cacao was naturally present throughout the Amazon Valley and may have subsequently been dispersed along two routes: one leading north and the other one heading west (42).Domestication of T. cacao began during this dispersal process in South America, before migrating indigenous human populations spread the seed to Central America and southern Mexico (42).Molecular analyses, combined with the hypotheses of Cheesman (4) and Schultes (42), uphold the theories that T. cacao originated in South America and was later introduced to Central America by human activities (26).
Historical records on the origin of T. cacao in Venezuela indicate wild T. cacao trees were first recorded in Aug 5, 1602, and highlight the discovery of 100,000 T. cacao trees close to Maracaibo Lake in western Venezuela (39).Subsequently, propagation of T. cacao occurred throughout Venezuela between 1600 and 1700, spreading the cultivar 'sweet cacao' or 'Criollo' (= 'Creole') (39).
The oldest recorded disease affecting T. cacao worldwide was documented in Venezuela in the mid-1630s on the "haciendas" upwind of La Guaira, where 'blight' (alhorra) had destroyed over half of all T. cacao on the coast within a decade (7,49).In 1824, the high-yielding T. cacao cultivar 'Forastero Trinitario' was introduced to Venezuela.The longer fermentation period required for beans of 'Forastero Trinitario' negatively influenced flavor, resulting in cacao of a lower quality compared to the 'Criollo' cultivar (39,49).The 'Forastero Trinitario' cultivar was rapidly distributed westward from the Paria peninsula to Barlovento and Tuy, where T. cacao 'Criollo' had practically disappeared because of disease pressures.Later, 'Forastero Trinitario' was transported into central Venezuela, but not to the western states of the country (34).
Information on diseases affecting T. cacao fruits and trees in Venezuela is typically sparse and dispersed in technical reports or meeting proceedings, and the fungi are mostly described using only morphology.Among the economically most important pathogens on T. cacao are Lasiodiplodia sp., causing pod rot and dieback in Ghana, Cameroon, Cuba, Ecuador and Venezuela (14,22,39,47), and Fusarium sp., as pathogens, endophytes and/or antagonists against other fungal pathogens of plants (12,22,39).
Though several fungal diseases have been reported attacking T. cacao in Venezuela, little research has been conducted to detect and identify the causal agents since the advent of molecular technologies.For closely related fungi, DNA-based analyses coupled with phylogenetic methods can be used to separate cryptic species.The aim of the present study is to use a combination of morphological and molecular techniques to determine the diversity of fungi associated with anthracnose and dieback of T. cacao in five important production localities of Mérida State, Venezuela.

Isolation of putative fungal pathogens
Stems, branches and fruits with symptoms of diseases, including dieback, sudden death and anthracnose, were collected from T. cacao from five municipalities in Mérida State (Table 1, Figure 1).Samples were transported to the laboratory and initially checked for the presence of fungal fruiting bodies; conidia and single spore isolations were performed as previously described (37).For additional fungal isolations, small pieces of stem, branch or fruit tissues (5 mm 2 ) were excised from the margins of necrotic lesions.Tissue pieces were surface disinfested for 30 s in 70% ethanol and 60 s in 0.5% sodium hypochlorite (NaOClbleach); then, they were washed in three changes of sterile water for 60 s each and dried on sterile filter paper.Surface-disinfested samples were plated onto 2% malt extract agar (MEA; Difco Laboratories, Detroit, MI, USA) and incubated at 25 °C for 7 days or until mycelia were observed growing from the plant tissues.Emerging fungal mycelium was sub-cultured to fresh MEA and hyphal tip methods were used to obtain pure cultures.
Mycelial plugs (5-mm) were taken from the periphery of actively growing cultures and transferred to 9-cm Petri dishes containing potato dextrose agar (PDA; Difco Laboratories, Detroit, MI, USA), MEA or synthetic nutrient-poor agar medium (SNA, 28).Following incubation at room temperature (25 °C) for 7 days, colony characteristics and pigment production were noted; colony diameters were measured after 7-to 10-day growth.Conidia produced on PDA and MEA were mounted in lacto-phenol for measurement of dimensions and morphological analyses.

DNA extraction and sequencing
Total DNA was extracted from 7-day-old cultures, grown on halfstrength PDA amended with 100 mg/L streptomycin sulphate (Sigma-Aldrich, USA).Isolates were sub-cultured to sterilized MF-47-mm Millipore membrane filters (Millipore Sigma, Burlington, MA, USA) overlaid on the surface of the half-strength PDA and incubated at 25 ˚C.DNA was extracted from cultured mycelia using the ZR Fungal/ Bacterial DNA MiniPrep (Zymo Research, Irvine, CA, USA), following the manufacturer's protocol.Polymerase chain reactions (PCR) were performed in a reaction mixture comprising 30 ng template genomic DNA, 2.5 μL10 standard Taq reaction buffer (New England BioLabs; NEB, Ipswich, MA, USA), 0.5 μL10 mM dNTP (Roche Applied Science, Penzberg, Germany), 1 μL of each 10 μM primer, 0.125 μL (0.6 units) Taq DNA polymerase (NEB), and sterile deionized water to a total volume of 25 μL.PCR products were cleaned with ExoSAP-IT ® PCR Product Cleanup (Thermo Fisher Scientific, Grand Island, NY, USA) and sequenced at Eurofins Scientific (www.eurofinsus.com).

Phylogenetic analyses
Sequence data from closely related species for phylogenetic analyses were obtained from GenBank (https://blast.ncbi.nlm.nih.gov/Blast.cgi).Sequence alignments, including manual adjustments and concatenation of loci, were conducted using MAFFT (17) within Geneious Pro v. 9.0.5.
For ML analyses within PhyML, phylogenies were run with 1,000 bootstrap pseudo-replicates.Bayesian Markov Chain Monte Carlo analyses were carried out using MrBayes (15).Six chains were run for 2,000,000 generations and trees sampled every 100 generations.The first 25,000 trees were discarded as burn-in and the remaining trees used for estimating posterior probabilities (PP) for the majority rule consensus tree.
Phylogenies for multiple loci datasets were estimated using Bayesian analyses implemented in Bayesian Evolutionary Analysis Sampling Trees (BEAST) (6).BEAST does not use concatenation but rather co-estimates the individual gene trees embedded inside the summary species tree.Bayesian Evolutionary Analysis Utility (BEAUti), version 1.7.5, was used to create XML-formatted input files for BEAST v1.7.5.In BEAST, a Markov Chain Monte Carlo algorithm was used to sample the posterior distribution of trees by conducting five independent runs of 100 million generations, each using a constant size tree prior, strict molecular clock, and uniform priors.Trees were sampled every 1,000 generations and the first 20% discarded as burnin.Post burn-in trees were combined with the program LogCombiner (BEAST v1.7.5); chains were assumed to converge when the average standard deviation of split frequencies was < 0.01.The maximum clade credibility tree with PP of each node was computed with Tree Annotator (BEAST v 1.7.5).Log files and tree files were visualized in Tracer v1.5 (http://tree.bio.ed.ac.uk/software/tracer/) and FigTree v1.3.1 (http:// tree.bio.ed.ac.uk/software/figtree/), respectively.

Morphological characterization
Morphological characteristics were evaluated to separate and group the fungi at the genus level.Three genera were separated and identified on the basis of morphology: 1) Cophinforma sp., conidiophores and paraphyses absent; conidia hyaline, unicellular, rarely becoming septate, mostly fusoid to ellipsoidal, most conidia longer than 30 μm; 2) Lasiodiplodia sp., conidia hyaline when young, later becoming 1-septate, dark brown with longitudinal striations, thick-walled, oblong to ellipsoid, straight, broadly rounded at the apex, base truncate; and 3) Fusarium sp., chlamydospores abundant, macroconidia septate, apical cell curved and tapering to a point, basal cell foot-shaped.Hypoxylon sp. could not be identified by morphology and was identified with DNA sequence-based analyses (Figure 2).

Phylogenetic analyses
ML and BI analyses of all examined loci produced similar results.Representative sequences for fungal isolates from Venezuela are highlighted in bold type in Figures 3-6.GenBank numbers are listed next to each sequence for reference.Posterior probability support for each clade ranged from 1 to 0.80.For internal nodes, however, PP was often lower.Bootstrap (BS) values below 50% are marked with a hyphen (-) on the phylogenetic trees.Fungal genera are presented based on their potential pathogenic importance as species newly reported on T. cacao, based on field observations.A total of 28 taxa and 56 isolates were included in the phylogenetic analyses; Macrophomina phaseolina was included as the outgroup (Table 2).A concatenation of three gene regions (ITS+TEF1+ BTUB) was used to separate Lasiodiplodia isolates from T. cacao from other species of Lasiodiplodia.Approximately 570 base pairs were sequenced at the ITS, 540 bp at the TEF1, and 430 bp at the BTUB, for a total of 1,540 bp (Figure 3).The concatenated phylogeny highlighted the   16), while L. theobromae was isolated from rotting T. cacao fruits in Ghana and Cuba (22,46) and from branches and twigs of healthy T. cacao in Brazil (12).In addition, microsatellite markers were used to identify L. theobromae and L. pseudotheobromae associated with dieback symptoms on T. cacao in Cameroon (2).
In Venezuela, L. theobromae was identified morphologically in association with T. cacao showing various symptoms, including fruit rot, dieback and cankers on branches, twigs and roots (35,39).Lasiodiplodia theobromae, identified with ITS sequencing, was also reported as an isolate from T. cacao with cushion galls (45); however, ITS sequences appear unreliable as a sole method for identifying currently recognized species of Lasiodiplodia.Several well-established cryptic species are recognized in the Lasiodiplodia complex, all with nearly identical ITS sequences and/or similar morphology (44).For this reason, multiple loci, including ITS, TEF1, BTUB and RPB2, are necessary for Lasiodiplodia species identification, and sequences of these loci are well-represented in databases.Importantly, these loci provide sufficiently stringent phylogenetic signals to distinguish most known cryptic species within Lasiodiplodia (43).Lasiodiplodia brasiliensis, first described in Brazil as causing stem-end rot of papaya (27), was also reported causing dieback of mango in Peru (40) and postharvest fruit rot of custard apple (Annona squamosa) in Brazil (3).Similar methodology was used to isolate Lasiodiplodia brasiliensis from branches and twigs of T. cacao with dieback or sudden death symptoms in Mérida State.This paper is the first to report L. brasiliensis potentially associated with dieback or sudden death on T. cacao in Venezuela; however, further studies are needed to confirm pathogenicity of L. brasiliensis on T. cacao.3).Phylogenies of the ITS and TEF1 showed that isolates of Cophinforma spp.from T. cacao fruits and stem grouped with known isolates of C. atrovirens in well-supported clades (BS/ PP = -/0.98 and 100%/1.00);phylogenies of the LSU and BTUB loci had moderate levels of support (BS/PP = 82%/0.65and 80%/0.57);however, no support was identified at the SSU locus (BS/PP = -/-) (Figure 4).
The genus Cophinforma was introduced within Botryosphaeriaceae by Liu et al. (20) as a monotypic genus that comprised C. eucalypti.Phillips et al. (37), however, showed that two species previously included in Botryosphaeria were better accommodated in Cophinforma: Botryosphaeria mamane as Cophinforma mamane, and Cophinforma eucalypti renamed as Cophinforma atrovirens.The conidia produced by Cophinforma are longer than those of any known species of Botryosphaeria.Although Botryosphaeria and Cophinforma share other morphological similarities, these genera are phylogenetically distinct.Based on phylogenetic relationships, C. mamane and C. atrovirens are currently recognized within Cophinforma (37).
Little published information is available for Cophinforma spp.Cophinforma atrovirens (= Cophinforma eucalypti = Fusicoccum atrovirens) was reported as a saprophyte on a dead branch of Eucalyptus sp. in Thailand (20), whereas C. atrovirens (as F. atrovirens) was isolated from asymptomatic branches and twigs of Pterocarpus angolensis in South Africa (23).In Venezuela, C. mamane (as B. mamane) was isolated from stems and branches of Acacia mangium and Eucalyptus hybrids (25), and it was also isolated from Sophora chrysophylla in Hawaii (9).In the present study, C. atrovirens was isolated from fruit anthracnose-and stem/branch dieback-associated tissues of T. cacao, and it was accompanied by L. theobromae and L. brasiliensis in Mérida State.Despite these findings, the capacity of C. atrovirens to cause disease is not verified.This report is the first showing that C. atrovirens is associated with fruit anthracnose, dieback or sudden death on T. cacao in Venezuela; however, inoculation work is required to conclusively determine if C. atrovirens is a pathogen of T. cacao.A total of 41 isolates comprised within 26 taxa of Fusarium spp.were used in the phylogenetic analyses with F. lyarnte RBG 5331 as the outgroup (Table 4).A total of 1,680 bp were sequenced at the TEF1 (710 bp) and RPB2 (970 bp) loci to determine species identity of 13 isolates of Fusarium from T. cacao.Phylogenetic analyses of the TEF1 and RPB2 showed that 11 isolates belonged to Neocosmospora solani, while two isolates belonged to Fusarium oxysporum; both species clades possessed well-supported nodes (BS/PP = 100%/1.00;Figure 5).

Table 4. Continuation
Taxa within the genus Fusarium are known as pathogens, endophytes and/or antagonists against other fungal pathogens of plants (12).Fusarium decemcellulare, the cause of cushion disease, is a weak pathogen that requires injuries to infect T. cacao (14).Cushion disease occurs in Sri Lanka, West Africa and the Western Hemisphere, and it causes economically important damage in Colombia, Costa Rica and Nicaragua (13).Fusarium decemcellulare has also been reported in Cuba and Venezuela, where it occurs with other diseases and pathogens of T. cacao, including fruit rots and stem/branch/root cankers associated with Phytophthora spp., Lasiodiplodia spp.and M. perniciosa (22,36,38,39).
In the present study, isolates of F. oxysporum and F. solani were obtained from T. cacao fruits with anthracnose symptoms.In Mérida State, these same Fusarium species were also isolated from T. cacao tissues with dieback or sudden death symptoms in association with L. brasiliensis and L. theobromae.
Fusarium oxysporum, F. solani and other Fusarium species, such as F. incarnatum, F. equiseti, F. camptoceras, F. crokwellense, F. moniliforme, F. proliferatum and F. decemcellulare, were previously identified on T. cacao in Venezuela, based on morphological and ITS sequence data (39,45,47).The use of the ITS region alone, however, does not provide sufficient resolution to distinguish some species of Fusarium that diverged relatively recently from common ancestors (32), and ITS is less informative than TEF1, RPB2 and BTUB (33) for phylogenetic analyses.The current paper confirmed the presence of F. oxysporum and F. solani on T. cacao in Venezuela based on multilocus sequencing.

Figure 1 .
Figure 1.(A) Sites of sampling (red stars) for Theobroma cacao in the five municipalities in Mérida State, Venezuela; map in grey scale shows additional T. cacao-producing states in Venezuela (55); (B, C) Dieback or sudden death symptoms; (D) Anthracnose on T. cacao fruits.

Figure 3 .
Figure 3. Phylogeny of Lasiodiplodia species generated from Bayesian analysis based on the concatenated ITS, BTUB, and TEF1 sequences.The phylogeny was rooted with Macrophomina phaseolina (CMM 3615 and PD 112).Node support (bootstrap ≥ 60%:1,000 replicates and Bayesian posterior probabilities ≥ 0.80, BS/PP) are highlighted by arrows.The new isolates from this study are in bold italics.

Figure 6 .
Figure 6.Phylogeny of Hypoxylon species generated from Bayesian analysis based on combined ITS and BTUB sequences.The phylogeny was rooted with Nemania serpens (235).Node support (bootstrap ≥ 90%:1,000 replicates and Bayesian posterior probabilities ≥ 0.90, BS/PP) are highlighted by arrows.The new isolates from this study are in bold italics.

Table 1 .
Municipalities and localities within Mérida State, Venezuela, where fungal samples were collected from Theobroma cacao showing different disease symptoms.
Obispo Ramos de Lora Santa Elena de Arenales Eloy Paredes-Las Honduras Sector Dieback or sudden death Sucre Lagunillas Estanques-Santo Domingo Sector San juan de Lagunillas-Instituto Nacional de Investigaciones Agrícolas (INIA) Sector Anthracnose on fruits Dieback or sudden death

Table 2 .
GenBank and culture collection accession numbers for Lasiodiplodia spp.isolates included in this study.Sequences that represented the isolates from Venezuela in this study are highlighted in bold.
a CBS: Centraalbureau voor Schimmelcultures, Fungal Biodiversity Centre, Utrecht, The Netherlands; IBL: Independent Biological Laboratories Israel.KEFAR MALAL; CMW: Tree Patholgy Co-operative Program, Forestry and Agricultural Biotechnology Institute, University of Pretoria, South Africa; IRAN: Iranian Fungal Culture Collection, Iranian Research Institute of Plant Protection, Iran; WAC: Department of Agriculture, Western Australia Plant Pathogen Collection, South Perth, Western Australia; STE-U: Culture Collection of the Department of Plant Pathology, University of Stellenbosch, South Africa; CMM: Bradford Art Galleries and Museums U.K. England.BRADFORD; PD: Culture Collection, University of California, Davis, USA; CERC, Culture Collection of China Eucalypt Research Centre, Chinese Academy of Forestry, ZhanJiang, GuangDong, China; CSM: Personal culture collection deposited in the Department of Bioagricultural Sciences & Pest Management, Colorado State University, USA.b See Table2for full locus name.

Table 2 .
Continuation separation of L. brasiliensis and L. theobromae (BS/PP = 86%/1.00and65%/0.82).Species of Lasiodiplodia are known to cause diseases on T. cacao in plantations across diverse global regions.Based on morphological descriptions, L. theobromae and L. citricola were associated with branch and trunk cankers on T. cacao in Nigeria (

Table 3 .
GenBank and culture collection accession numbers of Cophinforma spp.isolates included in this study.Sequences that represented the isolates from Venezuela in this study are highlighted in bold.
a CBS: Centraalbureau voor Schimmelcultures, Fungal Biodiversity Centre, Utrecht, The Netherlands; MFLUCC: Mae Fah Luang University Culture Collection, Chiang Rai, Thailand; CSM: Personal culture collection deposited in the Department of Bioagricultural Sciences & Pest Management, Colorado State University, USA.b See Table 2 for full locus name.

Table 4 .
GenBank and culture collection accession numbers of Fusarium spp.included in this study.Sequences that represented the isolates from Venezuela in this study are highlighted in bold.University of Sydney, Sydney, New South Wales, Australia; CBS: Centraalbureau voor Schimmelcultures, Fungal Biodiversity Centre, Utrecht, The Netherlands; CSM: Personal culture collection deposited in the Department of Bioagricultural Sciences & Pest Management, Colorado State University, USA.b See Table 2 for full locus name.
a RBG: Royal Botanic Gardens Trust, Sydney, New South Wales, Australia; NRRL: Agricultural Research Service Culture Collection, Peoria, Illinois, USA; F: -: Data not found.