Reconsidering species boundaries in theCeratocystis paradoxacomplex, including a new species from oil palm and cacao in Cameroon

Michael Mbenoun; Z. Wilhelm de Beer; Michael J. Wingfield; Brenda D. Wingfield; Jolanda Roux

The Ceratocystis paradoxa complex accommodates a group of fungal pathogens that have become specialized to infect mostly monocotyledon-ous plants. Four species currently are recognized in this group, including C. paradoxa, which has a widespread distribution and broad host range. In this study, multigene phylogenetic analyses involving sequences of the ITS, b-tubulin and TEF-1a gene loci, in combination with phenotypic and mating studies, were used to characterize purported C. paradoxa isolates from Cameroon and to compare them with isolates from elsewhere, including proto-logs and type specimens of known species. We show that the C. paradoxa complex comprises substantially greater species diversity than previously recognized. One new species in this group is described from Cameroon as Ceratocystis cerberus, while C. paradoxa sensu stricto (s. str.) and four other species are redefined. Lectotypes are designated for C. ethacetica and Endoconidium fragrans (synonym of C. ethace-tica), while epitypes are designated for C. paradoxa s. str., C. ethacetica and C. musarum. A neotype is designated for Catenularia echinata (synonym of C. ethacetica) and two species, previously treated in Thielaviopsis, are transferred to Ceratocystis.

Mycologia, 106(4), 2014, pp. 757–784. DOI: 10.3852/13-298 2014 by The Mycological Society of America, Lawrence, KS 66044-8897 # Reconsidering species boundaries in the Ceratocystis paradoxa complex, including a new species from oil palm and cacao in Cameroon Michael Mbenoun Z. Wilhelm de Beer Michael J. Wingfield INTRODUCTION Ceratocystis paradoxa is a soilborne fungal plant pathogen, widely distributed across the globe and capable of infecting a broad range of hosts (MorganJones 1967, Anonymous 1987, Garofalo and McMillan 2004). First discovered associated with black rot of pineapple fruit in France by de Seynes (1886), the impact of this fungus is best known from tropical, subtropical and arid regions where it represents a serious constraint to the cultivation of monocotyledonous crops (Kile 1993, Garofalo and McMillan 2004). Some of the well known diseases caused by C. paradoxa include rots of pineapple, sett rot of sugarcane and bud and trunk rots affecting almost all species of palm (Garafalo and McMillan 2004, Girard and Rott 2004). In most previous reports identification of C. paradoxa was based on morphological characters of the fungus and relied primarily on two studies of approximately 100 y ago. In the first of these, Petch (1910) reviewed all earlier work and produced a comprehensive account of the asexual state of the fungus. Dade (1928) subsequently described the sexual state and provided the first insights into the mating system of C. paradoxa. The power of DNA sequencing technologies and molecular phylogenetics, which have deeply influenced fungal taxonomy during the past two decades, has not yet been fully explored in the recognition and delineation of C. paradoxa and related species. Ceratocystis belongs to the family Ceratocystidaceae, order Microascales in the Sordariomycetes (Réblová et al. 2011, de Beer et al. 2013a) and is characterized by ascomata with bulbous bases and long necks that give rise to ascospores in sticky masses (Hunt 1956, Upadhyay 1981, Wingfield et al. 1993, Seifert et al. 2013). All species have tubular, phialidic conidiophores with diverse forms of endoconidia. In addition, some species also produce thick-walled, pigmented conidia. Under the dual nomenclature system, the asexual states of Ceratocystis species were treated in the genus Chalara (Nag Raj and Kendrick 1975) and were more recently consolidated in Thielaviopsis (Paulin-Mahady et al. 2002). Based on DNA sequence analyses, it is now recognized that there are distinct phylogenetic lineages in Ceratocystis, defining complexes in which species have similar morphological and ecological characteristics (Harrington et al. 1996, Harrington and Wingfield 1998, Department of Microbiology and Plant Pathology, Forestry and Agricultural Biotechnology Institute (FABI), Private Bag X20, University of Pretoria, Pretoria, 0028, South Africa Brenda D. Wingfield Department of Genetics, Forestry and Agricultural Biotechnology Institute (FABI), Private Bag X20, University of Pretoria, Pretoria, 0028, South Africa Jolanda Roux1 Department of Microbiology and Plant Pathology, Forestry and Agricultural Biotechnology Institute (FABI), Private Bag X20, University of Pretoria, Pretoria, 0028, South Africa Abstract: The Ceratocystis paradoxa complex accommodates a group of fungal pathogens that have become specialized to infect mostly monocotyledonous plants. Four species currently are recognized in this group, including C. paradoxa, which has a widespread distribution and broad host range. In this study, multigene phylogenetic analyses involving sequences of the ITS, b-tubulin and TEF-1a gene loci, in combination with phenotypic and mating studies, were used to characterize purported C. paradoxa isolates from Cameroon and to compare them with isolates from elsewhere, including protologs and type specimens of known species. We show that the C. paradoxa complex comprises substantially greater species diversity than previously recognized. One new species in this group is described from Cameroon as Ceratocystis cerberus, while C. paradoxa sensu stricto (s. str.) and four other species are redefined. Lectotypes are designated for C. ethacetica and Endoconidium fragrans (synonym of C. ethacetica), while epitypes are designated for C. paradoxa s. str., C. ethacetica and C. musarum. A neotype is designated for Catenularia echinata (synonym of C. ethacetica) and two species, previously treated in Thielaviopsis, are transferred to Ceratocystis. Key words: Ceratocystidaceae, Microascales, phylogenetic species concept, Thielaviopsis Submitted 13 Sep 2013; accepted for publication 7 Jan 2014. 1 Corresponding author. Jolanda.Roux@up.ac.za 757 758 MYCOLOGIA Baker et al. 2003, Johnson et al. 2005, van Wyk et al. 2006). Indeed, it has been suggested that some of these lineages represent discrete genera (Wingfield et al. 2013). The species related to C. paradoxa, sometimes referred to as C. paradoxa sensu lato (s. lat.), most likely represent one of these generic entities within Ceratocystis. Four species currently are recognized in the C. paradoxa complex, but results from phylogenetic studies using sequences from single loci suggest that this group includes at least two undescribed species (Harrington 2009, Álvarez et al. 2012). The previously described species include C. paradoxa, C. radicicola (Bliss 1941), C. musarum (Riedl 1962) and Thielaviopsis euricoi (Batista and Vital 1956, Paulin-Mahady et al. 2002). The first two species are well known as soilborne and pathogenic to monocotyledonous plants (Garofalo and McMillan 2004, Abdullah et al. 2009). While C. paradoxa has been recognized as a cosmopolitan fungus recorded on various hosts and from many parts of the world (Morgan-Jones 1967, Anonymous 1987), little is known regarding the incidence of the other three species. For example, there are no reports of C. musarum or T. euricoi other than from their first descriptions. The species in the C. paradoxa complex, of which the sexual states are known, typically produce characteristic digitate ornamentations on their ascomatal bases (Dade 1928, Bliss 1941). Like many other Ceratocystis species, the asexual states of these species often produce a variety of conidial forms in culture. In his generic diagnosis of Thielaviopsis, Went (1893) recognized two major types of conidia, endoconidia (also referred to as phialoconidia, enteroblasticphialidic ameroconidia), as opposed to thick-walled, pigmented conidia produced in chains at the tips of specialized hyphae. In most of the earlier studies the various forms of endoconidia were simply described as a range or continuum (as ‘‘cylindrical to ellipsoid/ doliiform, hyaline to brown’’; Bliss 1941, Hunt 1956, Nag Raj and Kendrick 1975, Upadhyay 1981). However, in the taxonomic literature of Ceratocystis during the past decade, the endoconidia have been described consistently as of two forms. These have been referred to as cylindrical versus doliform endoconidia (Baker-Engelbrecht and Harrington 2005, Johnson et al. 2005) or by other authors as primary versus secondary conidia (van Wyk et al. 2006, 2007, 2009, 2011; Heath et al. 2009). For the purpose of the present study we define the endoconidia produced in phialides as primary conidia when they are aseptate, hyaline and cylindrical. In contrast, secondary conidia are cylindrical to oblong, often doliiform (barrel-shaped), initially hyaline, turning brown and thick-walled with age. In addition to the two forms of endoconidia, species in the C. paradoxa complex also form what have been referred to as chlamydospores or aleurioconidia. Went (1893) described and illustrated pigmented conidia in short chains on terminal ends of hyphae of Thielaviopsis, while Peyronel (1918) established Chalaropsis for species with aleuriospores borne solitarily at the end of a branched conidiophore (Hennebert 1967). These structures often have been called chlamydospores (Nag Raj and Kendrick 1975; Upadhyay 1981; van Wyk et al. 2007, 2009, 2011; Kamgan et al. 2008; Heath et al. 2009), but the term ‘‘aleurioconidia’’ was reintroduced to taxonomic studies in Ceratocystis by Paulin-Mahady et al. (2002). Seifert et al. (2011) use the term aleurioconidia for ‘‘solitary, thallic conidia that secede rhexolytically, usually with thickened and sometimes darkened walls’’ and distinguishes it from chlamydospores, described as ‘‘thick-walled resting spores, intercalary or terminal on hyphae … usually not readily liberated’’. We concur with the definition of aleurioconidia by Paulin-Mahady et al. (2002) and Harrington (2009) as thick-walled, survival spores produced singly or in chains from the tips of specialized hyphae, which also corresponds to the definition by Seifert et al. (2011). The aim of this study is to characterize recent isolates resembling C. paradoxa s. lat. from oil palm (Elaeis guineensis), cacao (Theobroma cacao) and pineapple (Ananas comosus) in Cameroon, with multigene DNA phylogenies, morphology and mating tests. Species boundaries and previously suggested synonymies of all species in the complex also are reconsidered. For this purpose, the Cameroonian isolates were compared with herbarium specimens and representative isolates of known species. MATERIALS AND METHODS Isolates.—Those forming the core material of this study were collected in Cameroon in 2009 and 2010, during field surveys for Ceratocystis species occurring on pineapple, oil palm and cacao. The surveys were conducted at various localities of the central, littoral and southwestern regions of the country. On cacao, fungi resembling C. paradoxa s. lat. were isolated mostly from discarded pod husks (FIG. 1), but a few isolates also were obtained from artificially induced wounds on the stems of cacao trees. Oil palm isolates were obtained from trunks and stumps of felled trees, cut basal ends of leaf fronds and rotting palm-nut bundles (FIG. 1). Pineapple isolates were obtained from rotting fruits, fruit peduncle sections and damaged leaves. Fungal structures, including mycelium and less commonly ascomata (only from cacao pod husks), generally were observed directly on plant tissue in the field. When they were not present, incubation of tissue samples in moist chambers for a few days was useful for inducing growth and sporulation. MBENOUN ET AL.: THE CERATOCYSTIS PARADOXA COMPLEX 759 FIG. 1. Symptoms of infection by Ceratocystis paradoxa s. lat. a. Basal section of felled Elaeis guineensis tree. b. External surface of Theobroma cacao pod husk. Isolations were made by aseptically lifting a few strands of aerial mycelium or single ascospore droplets from the surfaces of infected plant material with a sterile needle and transferring these to sterile 2% malt extract agar (MEA) (Biolab, Midrand, South Africa), amended with , 0.01 g/L streptomycin sulphate (Sigma, Steinheim, Germany). Isolates were purified further by subculturing from single hyphal tips, and they were maintained on MEA. The Ceratocystis collections from Cameroon were supplemented with cultures obtained from the culture collection (CMW) of the Forestry and Agricultural Biotechnology Institute (FABI), University of Pretoria, South Africa, the Centraalbureau voor Schimmelcultures (CBS), Utrecht, the Netherlands, and the Commonwealth Agricultural Bureau International Bioscience (CABI), UK. Additional isolates were sourced from international fungal culture collections and specifically chosen to represent different hosts and geographic regions. All fungal isolates included in this study are maintained in the culture collection (CMW) at FABI (TABLE I). Herbarium specimens of types of Ceratocystis musarum, Ceratostomella paradoxa and Stilbochalara dimorpha also were investigated (see TAXONOMY). DNA extraction, PCR and sequencing.—DNA was extracted from 7 d old cultures maintained on MEA at 25 C. Mycelium was scraped from the surfaces of cultures and transferred to Eppendorf tubes for freeze-drying. Dried mycelium was crushed into a fine powder in a Retsch cell disrupter (Retsch Gmbh, Germany); thereafter total genomic DNA was isolated with the CTAB (cetyl trimethyl ammonium bromide) protocol described by Möller et al. (1992). Final DNA working concentrations were adjusted to , 75 gg mL21, with a Thermo Scientific NanoDropH ND-1000 Spectrophotometer (NanoDrop Technologies, Wilmington, Delaware). For primary identification of Ceratocystis isolates, the internal transcribed spacer regions, ITS1 and ITS2 and intervening 5.8S rDNA of the ribosomal RNA gene cluster (ITS) were amplified with the polymerase chain reaction (PCR) and sequenced for all isolates. Two additional gene regions, including part of the beta tubulin gene (b-tubulin) and part of the translation elongation factor 1-alpha gene (TEF-1a), were sequenced for selected isolates and used in phylogenetic reconstructions. The oligonucleotide primer pairs included in the reactions were respectively the ITS1 (59–TCCGTAGGTGAACCTGCGG–39) and ITS4 (59– TCCTCCGCTTATTGATATGC–39) for the ITS region (White et al. 1990), bt1a (59–TTCCCCCGTCTCCACTTCTTCATG–39) and bt1b (59–GACGAGATCGTTCATGTTGAACTC–39) for the b-tubulin region (Glass and Donaldson 1995), and EF1F (59–TGCGGTGGTATCGACAAGCGT–39) and EF2R (59–AGCATGTTGTCGCCGTTGAAG–39) for the TEF-1a region (Jacobs et al. 2004). PCR reaction mixtures contained 1 mL DNA template working solution, 0.5 mL each oligonucleotide primer (10 mM), 0.5 mL MyTaqTM DNA polymerase (Bioline) and 5 mL 53 MyTaqTM reaction buffer (supplied with the enzyme). The final reaction volumes were adjusted to 25 mL with sterile distilled water (SABAX water, Adcock Ingram Ltd, Bryanston, South Africa). PCR reactions were performed on a Biorad thermo-cycler (BIO-RAD, Hercules, California). The same cycling conditions were applied for both the ITS and b-tubulin regions, which comprised an initial denaturation step at 96 C for 2 min, followed by 35 cycles of 30 s at 94 C (denaturation), 60 s at 54 C (annealing), 90 s at 72 C 760 TABLE I. Isolates of C. paradoxa s. lat. included in this study GenBank accession Nos. CMW No. Other ITS b-tubulin TEF-1a Host Origin Collector (supplier) C. cerberus sp. nov. 35021b 35024 36641 36653b 36668b 14790 28533 28534 36644a 36662a,b 36671a,b 36691b 36725a,b 36735a 36741a,b 36743a 36745a 36747a 36771 37450 37775 37777 37778 CBS 130763 JX518355 JX518356 JX518345 JX518349 JX518348 JX518329 JX518331 JX518332 JX518387 JX518388 JX518377 JX518381 JX518380 JX518361 JX518363 JX518364 JX518323 JX518324 JX518313 JX518317 JX518316 JX518297 JX518299 JX518300 CBS 130758 JX518353 JX518351 JX518343 JX518347 JX518385 JX518383 JX518375 JX518379 JX518321 JX518319 JX518311 JX518315 CBS 130759 JX518354 JX518386 JX518322 CBS 128.32 IMI 50560 IMI 344082 IMI 378943 JX518330 JX518337 JX518341 JX518339 JX518340 JX518362 JX518369 JX518373 JX518371 JX518372 JX518298 JX518305 JX518309 JX518307 JX518308 Theobroma cacao Theobroma cacao Elaeis guineensis Elaeis guineensis Elaeis guineensis Phoenix canariensis Ananas comosus Cocos nucifera Elaeis guineensis Elaeis guineensis Elaeis guineensis Theobroma cacao Ananas comosus Ananas comosus Theobroma cacao Theobroma cacao Theobroma cacao Elaeis guineensis Saccharum sp. Elaeis guineensis Ananas comosus Cocos nucifera Elaeis guineensis M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux (CBS) M. Barreto Figueiredo (CBS) O.A. Akinrefon (CBS) M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux N. van Wyk (FABI) (CBS) (CABI) (CABI) (CABI) CBS 893.70 CBS 107.22 CBS 130760 JX518326 JX518327 JX518328 JX518335 JX518336 JX518346 JX518358 JX518359 JX518360 JX518367 JX518368 JX518378 JX518294 JX518295 JX518296 JX518303 JX518304 JX518314 JX518350 JX518382 JX518318 JX518352 JX518342 JX518344 JX518384 JX518374 JX518376 JX518320 JX518310 JX518312 Cameroon Cameroon Cameroon Cameroon Cameroon Spain Brazil Nigeria Cameroon Cameroon Cameroon Cameroon Cameroon Cameroon Cameroon Cameroon Cameroon Cameroon South Africa NA Malaysia Tanzania Papua New Guinea Indonesia Indonesia Indonesia Brazil NA Cameroon Cameroon Cameroon Cameroon Cameroon Cameroon Cameroon Cameroon Cameroon Cameroon C. ethacetica C. euricoi C. paradoxa s. str. 8788 8790 8799b 28537b 28538b 36642a,b 36650a 36654a,b 36655a,b 36674a 36683a 36686a,b 36689a,b 36754b 37755a CBS CBS CBS CBS CBS 130764 130765 374.83 601.70 453.66 CBS 130757 CBS 130762 CBS 130761 Cocos nucifera Cocos nucifera Cocos nucifera NA Cocos nucifera Elaeis guineensis Elaeis guineensis Elaeis guineensis Elaeis guineensis Elaeis guineensis Elaeis guineensis Elaeis guineensis Theobroma cacao Elaeis guineensis Elaeis guineensis M.J. Wingfield (FABI) M.J. Wingfield (FABI) M.J. Wingfield (FABI) E. de Matta (CBS) (CBS) M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux M. Mbenoun & J. Roux MYCOLOGIA Species name CBS 114.47 CBS 167.67 IM I 316225 CABI: Commonwealth Agricultural Bureaux International Bioscience, formerly International Mycological Institute (IMI). CBS: Centraalbureau voor Schimmelcultures. CMW: Culture collection of the Forestry and Agricultural Biotechnology Institute (FABI), University of Pretoria. a Isolates used in mating tests. b Isolate used for microscopic and/or growth studies. NA: Data not available. J.W. Veenbaas- Rijks (CBS) M.C. Aime (CBS) T.W. Canter-Visscher (FABI) D.E. Bliss G.L. Hennebert (CABI) Netherlands Ecuador New Zealand USA Mauritania Iraq Rosa sp. Palm species Musa sp. Phoenix dactylifera Lawsonia inermis Phoenix dactylifera JX518301 JX518302 JX518293 KF612024 KF917202 JX518306 JX518365 JX518366 JX518357 KF612025 KF953931 JX518370 28535 28536 1546a 1032 26389 37776b C. C. C. C. paradoxa s. lat. 1 paradoxa s. lat. 2 musarum radicicola CBS 101054 CBS 116770 JX518333 JX518334 JX518325 KF612023 KF953932 JX518338 TEF-1a b-tubulin ITS CMW No. Species name TABLE I. Continued Other GenBank accession Nos. Host Origin Collector (supplier) MBENOUN ET AL.: THE CERATOCYSTIS PARADOXA COMPLEX 761 (extension) and a final extension step at 72 C for 10 min to complete the reaction. For the TEF-1a gene region, the cycling conditions included 4 min initial denaturation at 96 C, 10 cycles of 40 s at 94 C, 40 s at 55 C (annealing step) and 45 s at 72 C (extension), followed by 30 additional cycles of the same sequence, with a 5 s increase in the annealing step per cycle. Reactions were completed by a final extension step at 72 C for 10 min. To confirm amplification of targeted gene regions, 4 mL aliquots reaction products were mixed with 1.5 mL GelRedTM (Biotium Inc., USA) dye and electrophoresed on 2% agarose gels with a DNA molecular weight marker (100 bp ladder) (Fermentas O’Gene RulerTM) and visualized under UV light. PCR products were purified with 6% Sephadex G-50 (50– 150 mm diam beads) columns (Sigma, Steinheim, Germany), following the manufacturer’s instructions. Aliquots (4 mL) purified PCR products were used in the forward and reverse sequencing reactions, with the ABI PRISMTM BIG DYE Terminator Cycle Sequencing Ready Reaction kit (Applied BioSystems, 142 Foster City, California). In addition to the PCR products, sequencing mixtures contained 2.5 mL sequencing buffer, 0.5 mL Big Dye ready reaction mixture and 1 mL selected primer (10 mM). The final reaction volume was adjusted to 12 mL with sterile Sabax water. The sequencing PCR conditions comprised 25 cycles of a 10 s denaturation step at 96 C, 5 s annealing step at 52 C and a primer extension step at 60 C for 4 min. Sequencing products were cleaned with Sephadex G-50 columns and concentrated in an Eppendorf 5301 vacuum concentrator at 45 C. They were run on an ABI PRISMTM 3100 DNA Analyzer (Applied BioSystems, 142 Foster City, California). Bidirectional reads were assembled into consensus sequences and edited with CLC Main Workbench 6.1 software package (CLC Bio, Aarhus, Denmark). Phylogenetic analyses.—As a means of primary identification and sorting of Cameroonian isolates, their ITS sequences were aligned with MUSCLE (Edgar 2004) as implemented in MEGA 5 (Tamura et al. 2011), and a neighbor-joining (NJ) phylogenetic tree was generated with MEGA. Representative sequences from each resulting group were submitted to BLASTN query in GenBank on NCBI (http://www.ncbi.nlm.nih.gov). The same BLASTN procedure was applied to the ITS sequences of all isolates obtained from the culture collections to confirm their identities. The NJ tree resolved the isolates from Cameroon into three groups, each showing 97–100% sequence similarity with C. paradoxa accessions in GenBank. Five representatives of each NJ group, originating from different localities and hosts in Cameroon, were selected for phylogenetic analyses together with related CMW, CBS and CABI isolates (TABLE I). Three datasets, including sequences generated in this study and relevant GenBank accessions, were constructed for the three gene regions, as well as one dataset combining all gene regions. Ceratocystis virescens, isolate CMW 11164, (van Wyk et al. 2007) was used as the outgroup taxon in the analyses. Alignments were constructed with MAFFT 6 (http://www.align.bmr.kyushu-u.ac.jp/ mafft/online/server/) (Katoh et al. 2005) and trimmed in 762 MYCOLOGIA MEGA. The three gene regions were considered separately and in a three-partition combination, applying three different approaches of phylogenetic inference, that is Bayesian inference (BI), maximum likelihood (ML) and maximum parsimony (MP). Where applicable, models of nucleotide substitution were selected with jModelTest 2.2 (Posada 2008). ML analyses were performed with the PhyML 3.0 online interface (Guindon et al. 2010 http://www.atgc-montpellier. fr/phyml/). Confidence support for branch nodes were estimated with 1000 bootstrap replications. BI analyses based on Markov chain Monte Carlo (MCMC) algorithms were performed with MrBayes 3.1.2 (Ronquist and Huelsenbeck 2003). The MCMC procedure was initiated from a random tree topology and involved 1 000 000 random tree generations with four parallel chains and tree sampling every 100th generation. The software Tracer 1.5 (Rambaut and Drummond 2007) was used to examine the convergence of the chains from the distribution of posterior probabilities. The default burn-in of 10% of the generations (i.e. the first 1000 trees sampled), which fell beyond the point of convergence in all cases, was enforced during the construction of 50% majority rule consensus trees. MP analyses were performed with PAUP 4.0b10* (Swofford 2002). Uninformative characters were excluded and gaps coded as a fifth character state. All characters were unordered and of equal weight. Trees were generated via a heuristic search, with random stepwise addition of 1000 replicates and tree bisection and reconstruction (TBR) branch-swapping algorithms. Branches of zero length were collapsed, and all multiple, equally parsimonious trees were saved. Statistical support for branch nodes in the most parsimonious trees (MPTs) was assessed with 1000 bootstrap replicates with the full heuristic search. Tree length (TL), consistency index (CI), retention index (RI), homoplasy index (HI) and rescaled consistency index (RC) also were calculated for MPTs. Finally, the genealogical concordance of the three gene loci was assessed with partition homogeneity tests (PHT) with 1000 heuristic search replicates (Swofford 2002). Mating studies.—A trial was set to characterize the mating system of the putative C. paradoxa isolates from Cameroon and to produce possible sexual structures for morphological descriptions. This involved 10 randomly selected isolates from each discrete phylogenetic group, which produced only asexual structures. Single hyphal tip cultures were paired against themselves as well as against each other on 2% MEA and water agar (WA) plates in the presence of either sterile pineapple or sugarcane chips. This was done by plating the two isolates, opposite each other, approximately 5 cm apart, and separated by a line of plant material. Plates were incubated 2–4 wk in the dark at 25 C. The experiment was done in duplicate by preparing two plates for each pairing combination. Four isolates shown to represent opposite mating types from each group also were used in reciprocal crosses to assess the presence of reproductive barriers between isolates in different groups. Morphology and taxonomy.—We studied growth and morphology of three isolates for each of the Ceratocystis species identified from Cameroon. For each isolate, five replicate plates were prepared by cutting out single agar plugs, with an 8 mm diam cork borer, from the margin of an actively growing culture on 2% MEA. These were transferred, mycelium side down, to the centers of 90 mm Petri dishes containing fresh, sterile 2% MEA. Plates were incubated 3 d in the dark at 10–35 C at 5 C intervals. Growth diameter measurements were taken for each colony on two axes at right angles, and averages were computed. Morphological characteristics were determined for 10 d old cultures maintained on 2% MEA at their optimum growth temperature and herbarium specimens of type material when available. The color for fungal colonies and structures was determined with the mycological color charts of Rayner (1970). Microscope slides were prepared for fungal structures by mounting these in 85% lactic acid. Slides were examined under a Zeiss Axioskop microscope fitted with HRc Axiocam digital camera and Axiovision 3.1 software (Carl Zeiss Ltd, Germany) used for image capture and to determine sizes of structures. Where possible, 50 measurements were made for each taxonomically informative morphological character for isolates used as holotypes, with an additional 10 measurements for each isolate designated as a paratype. Mean and standard deviation values were computed for each character. These measurements are presented as the extremes in brackets and the range calculated as the mean of the overall measurements plus or minus the standard deviation. Morphometric data and other phenotypic information for all previously known species related to C. paradoxa s. lat. were obtained from the literature. Additional information was generated through microscopic examination of isolates representing these species from isolates obtained from culture collections. A comparison of phenotypic features across all species in the C. paradoxa complex was made to identify possible diagnostic characters for each species considered in this study. All isolates used as type cultures in morphological descriptions were deposited with the CBS culture collection. Dried herbarium specimens, including paired cultures of sexually compatible isolates, also have been deposited in the South African National Herbarium at the Plant Protection Research Institute (PREM), Pretoria, South Africa. RESULTS Isolates.—A total of 143 fungal isolates representing general growth and morphological characteristics of C. paradoxa s. lat. were collected in Cameroon. Of these, 22 were isolated from pineapple, 69 from oil palm and 52 from cacao (SUPPLEMENTARY TABLE I). Two additional isolates were collected from a fallen Erytrophleum ivorense tree alongside a cacao plantation after a windstorm. Based on NJ sequence comparisons of ITS sequence data, the Cameroonian isolates represented three distinct groups that we designated CMR1, CMR2 and CMR3. Group CMR1 included , 80% of isolates (SUPPLEMENTARY FIG. 1), MBENOUN ET AL.: THE CERATOCYSTIS PARADOXA TABLE II. Summary of the phylogenetic information for the individual and combined nuclear regions used in this study Taxa (Nr) Character (Nr) MP BI/ML 763 COMPLEX Dataset ITS b-tubulin TEF-1a Combined Total Variable constant PIC (Nr) MPT (Nr) TL CI HI RI RC Model Gamma shape P-inv ti/tv 49 408 70 338 40 1 60 0.817 0.183 0.959 0.783 HKY+I — 0.7840 2.4725 36 508 134 374 88 2 164 0.787 0.213 0.959 0.754 TIM+G 0.1670 — — 41 771 367 404 269 10 770 0.666 0.334 0.902 0.601 TVM+I+G 0.9030 0.4090 — 35 1684 561 1123 382 70 911 0.720 0.280 0.917 0.660 GTR+I+G 0.5730 0.4650 — with all the isolates from pineapple, most (, 90%) isolates from cacao as well as some (, 57%) from oil palm. Group CMR2 was the second most prevalent and represented , 16% of the isolates, mostly from oil palm with a single isolate from cacao. Group CMR3 was the least prevalent of the three groups, representing less than 5% of the collections. It included isolates from oil palm and cacao, with cacao isolates originating from artificially induced trunk wounds in one orchard. Phylogenetic analyses.—For the 32 putative C. paradoxa s. lat. isolates, DNA sequence data of approximately 500, 500 and 800 bp were generated respectively for the ITS, b-tubulin and TEF1-a loci (see TABLE I for GenBank accession numbers). A summary of the phylogenetic information for each of the three loci, as well as the combined dataset, is provided (TABLE II). There was a general concordance between the BI, ML and MP phylogenies in tree topology and phylogenetic relationship among taxa. The ITS constituted the largest dataset but had the least resolution (SUPPLEMENTARY FIG. 2). Cameroonian isolates separated into three clades. Isolates in groups CMR1 and CMR2 formed two separate but monophyletic clades, forming a cluster with high statistical support, and each was composed of several GenBank accessions as well as isolates obtained from culture collections. Clade CMR1 had the greatest diversity in terms of hosts and geographic origin, with taxa originating from Cameroon and other countries in Africa, Asia, Europe, North and South America and Oceania, collected from banana (Musa sp.), Butia sp., cacao, date palm (Phoenix sp.), coconut palm, pineapple, oil palm and sugarcane (Saccharum sp.). Clade CMR2 included, apart from the isolates from Cameroon, the ex-type (CMW 28537 5 CBS 893.70) of Thielaviopsis euricoi, which originally was collected from air in Brazil, isolate CMW 28535 5 CBS 101054 from Rosa sp. in the Netherlands, five isolates collected from coconut palm (Cocos nucifera) in Brazil and Indonesia and one GenBank accession from Jamaica, also associated with coconut palm. Isolates in group CMR3 formed a phylogenetic clade, separate from CMR1 and CMR2 isolates, supported by strong bootstrap and posterior probability values. In addition to the isolates from Cameroon, this clade included one GenBank accession representing an isolate from a palm species in Colombia. Apart from the three clades containing the Cameroon isolates, three other well supported clades were present in the ITS tree. One of these had a basal position and included an ex-type isolate (CMW 1032 5 CBS 114.47) of C. radicicola and two additional taxa from date palm in three countries (Iraq, Oman, USA). The second clade was sister to group CMR3 isolates from Cameroon, including an isolate (CMW 28536 5 CBS 116770) from a palm species in Ecuador and one GenBank accession. The third clade was monophyletic with group CMR1 and included one taxon (CMW 1546) from banana in New Zealand and one from date palm in Saudi Arabia. Sequence analyses of the b-tubulin and TEF-1a loci had better resolution than those for the ITS region. All the clades delineated in the ITS tree were recovered in the b-tubulin (SUPPLEMENTARY FIG. 3) and TEF-1a (SUPPLEMENTARY FIG. 4) trees, with stronger statistical support. Furthermore, clade CMR2 including T. euricoi was separated into two well supported subclades (CMR2, T. euricoi) and one nonaligned isolate, CMW28535 5 CBS 101054. 764 MYCOLOGIA Polymorphisms also were observed within clade CMR1 in the TEF-1a tree where at least three subclades were identified. There was some conflict among the individual gene trees in the deep branches, especially regarding the relationship among the three CMR groups. While the CMR1 and CMR2 clades in the ITS tree were monophyletic with strong statistical support, the CMR1 clade in the TEF-1a tree was monophyletic with the CMR3 clade, also with strong statistical support. In the b-tubulin tree, a common ancestry was shared between the CMR2 and CMR3 clades, although this was less well supported. This incongruence was reflected in the PHT, which resulted in a P value of 0.001. Overall, only two well supported deep branches were identified in the combined species tree (FIG. 2). The one branch linked C. radicicola and group CMR3 as well as isolate CMW 28536 5 CBS 116770, each representing a distinct taxon. The second branch also linked three taxa, T. euricoi, group CMR2 and isolate CMW 28535 5 CBS 101054. A third deep branch connected C. musarum and group CMR1. However, this link was supported statistically only in the MP tree. Mating studies.—Isolates in the CMR3 group from Cameroon produced ascomata in single-spore or hyphal-tip cultures, and this appears to be a homothallic taxon. In contrast, none of the single-spore or hyphal-tip cultures representing groups CMR1 and CMR2 produced ascomata. Ten randomly selected isolates representing each of these two groups were used in mating trials. Of the 55 paring combinations used for each group, we obtained 12 and six fertile combinations respectively for CMR1 and CMR2, with ascomata forming along the lines of interaction between the two colonies. The pattern of fertile combinations clearly indicated the existence of two mating types in each group and thus a heterothallic mating system (FIG. 3). However, reproductive compatibility between pairs of isolates was variable. This was illustrated in the relative abundance of ascomata produced by the various pairs and the fact that some isolates did not produce ascomata in any combination. These infertile isolates could not be assigned a mating type (FIG. 3). Mating was never observed between isolates representing the two different phylogenetic groups. TAXONOMY Phylogenetic analyses and careful comparisons of material collected in Cameroon with protologs, type specimens and isolates from elsewhere made it possible to delineate C. paradoxa s. str. and other species in the C. paradoxa complex. All described species included in the complex at present are listed below, and issues related to typification, synonymies, nomenclature and morphology are discussed for each species. Morphological differences among the species are summarized (TABLE III). Ceratocystis paradoxa (de Seynes) C. Moreau, Rev. Mycol. (Paris) Suppl. Col. 17:22. 1952. FIGS. 4–6 MycoBank MB294224 5 Sporoschisma paradoxum de Seynes, Recherches pour Servir à l’Histoire Naturelle des Végétaux Inférieurs 3:30. 1886. (basionym) 5 Chalara paradoxa (de Seynes) Sacc., Syll. Fung. 10:595. 1892. 5 Thielaviopsis paradoxa (de Seynes) Höhn., Hedwigia 43:295. 1904. 5 Ceratostomella paradoxa (de Seynes) Dade, Trans. Br. Mycol. Soc. 13:191. 1928. 5 Ophiostoma paradoxum (de Seynes) Nannf., In Melin & Nannf., Svenska SkogsvFör. Tidskr. 32:408. 1934. 5 Endoconidiophora paradoxa (de Seynes) R.W. Davidson, J. Agric. Res. 50:802. 1935. 5 Stilbochalara dimorpha Ferd. & Winge, Bot. Tidsskr. 30:220. 1910. Ascomata perithecial, forming on sugarcane chips on WA and on agar in paired cultures of compatible strains, absent in single colonies, but observed in nature on endosperma of cacao pod husks. Ascomatal bases fully or partially submerged in substrata, mostly globose, (201–)237–317(–392) mm high 3 (249–) 279–348(–382) mm wide, straw (21f ), partially or completely covered by aleurioconidia and ascomatal appendages. Ascomatal appendages digitate, (29–) 31–45(–51) 3 (10–)11–25(–26), mostly on exposed areas of ascomatal bases. Ascomatal necks umber (9), erect, (961–)1063–1367(–1538) mm long, (20–)26– 37(–45) mm wide at apices; bases of the necks occasionally swollen, forming collar-like structures, (53–)72–96(–110) mm wide. Ostiolar hyphae hyaline, divergent, (62–)102–150(–158) mm long. Asci not observed. Ascospores invested in mucous sheaths, hyaline, aseptate, ellipsoidal, (9–)9–11(–12) 3 (3–)3– 4(–4) mm, accumulating in mucilaginous, buff (19’’f ) droplets at tips of ascomatal necks. Conidiophores hyaline to grayish sepia (17’’’i), 1–4 septa, phialidic, lageniform, (73–)93–179(–301) mm long, (4–)5–8 (–9) mm wide at bases and (4–)4–6(–6) mm wide at apices, mononematous with enteroblastic conidium ontogeny, commonly solitary, but occasionally aggregated in synnemata, variable in size, (406–)488– 1312(–1640) mm long 3 (43–)47–65(–69) mm wide. Primary conidia hyaline, aseptate, cylindrical, (8.1–) 10–14(–20) 3 (4–)4–5(–6) mm. Secondary conidia MBENOUN ET AL.: THE CERATOCYSTIS PARADOXA COMPLEX 765 FIG. 2. Species tree of Ceratocystis paradoxa s. lat. derived from ML analysis of combined ITS, b-tubulin and TEF1-a sequences Bootstrap support values $ 70% from 1000 heuristic search replicates are indicated next to branch nodes as ML (MP). Thick branches are those supported by $ 95% posterior probability in Bayesian analyses. The phylogeny is subdivided in three major clades highlighted by blue circles. 766 MYCOLOGIA FIG. 3. Summary of mating tests realized with 10 isolates each of group CMR1 (Ceratocystis ethacetica) and group CMR2 (Ceratocystis paradoxa s. str.) groups. The number of stars is indicative of the relative abundance of ascomata obtained with the various mating combinations. TABLE III. Morphological comparisons between species of the Ceratocystis paradoxa complex C. cerberus C. ethacetica Homothallic Heterothallic Ascomatal bases Rooting Shape Ornamentations Length Width Superficial Globose Digitate–stellar (137–)260–340(–370) (148–)268–348(–368) Submerged Globose Digitate (107–)154–215(–260) (125–)156–216(–251) Acomatal necks Shape Length C. radicicolab C. paradoxa s. str. Heterothallic Heterothallic Submerged Spherical–ellipsoid Absent 300 350 Submerged Globose Digitate (201–)237–317(–392) (249–)279–348(–382) Submerged Spherical Digitate 180–320 NA Erect/curled Erect (280–)650–984(–1244) (451–)672–862(–983) NA 1100–1200 NA 440–980 (24–)29–43(–56) (43–)61–83(–101) (20–)22–29(–35) (33–)44–64(–79) 17–18 50 Erect (961–)1063–1367(– 1538) (20.2–)26–37(–45) (53–)72–96(–110) Ostiolar hyphae Shape Length Divergent (47–)70–108(–142) Divergent (71–)92–112(–121) Convergent ,100 Divergent (62–)102–150(–158) Fimbriate 56–112 Ascospores Shape Length Width Ellipsoidal (6–)7–9(–13) (3–)3–4(–6) Ellipsoidal (7–)7–9(–10) (2–)3–4(–4) Cylindrical/ellipsoidal 6–11 2–3.5 Ellipsoidal (9–)9–11(–12) (3–)3–4(–4) Ellipsoidal 8–15 3–4 Phialides Shape Length Width (apice) Width (base) Synnemata Lageniform (63–)92–150(–215) (3–)4–5(–6) (5–)6–7(–8) Absent Lageniform (75–)87–148(–257) (3–)4–5(–6) (4–)6–8(–10) Absent Lageniform (77–)101–144(–171) (3–)4–6(–7) (6–)7–10(–12) Present Lageniform (174–)206–252(–295) (4–)4–5(–6) (5–)7–9(–10) Absent Lageniform (73–)93–179(–301) (4–)4–6(–6) (4–)5–8(–9) Present Lageniform (108–)141–207(–252) (4–)4–5(–6) (5–)7–8(–10) Absent Cylindrical (7–)7–12(–19) (3–)4–5(–6) Not observed Cylindrical (5–)7–8(–10) (2–)4–5(–6) Cylindrical (8–)9–11(–12) (3–)4–5(–5) Cylindrical (9–)10–12(–13) (3–)4–5(–5) Cylindrical (8–)10–14(–20) (4–)4–5(–6) Cylindrical (6–)8–14(–16) (3–)4–5(–5) Cylindrical/oblong (7–)8–12(–16) (4–)6–7(–7) Cylindrical/oblong (10–)11–16(–25) (5–)6–7(–8) Cylindrical/oblong (9–)10–13(–15) (5–)6–7(–8) Cylindrical/oblong (4–)6–7(–8) (8–)9–13(–16) Cylindrical/oblong (3–)4–6(–6) (6–)10–14(–16) Primary conidia Shape Length Width Second. conidia Shape Length Width 24 71 COMPLEX Undetermined Width (apice) Width (base) Sexual state not observed C. musaruma MBENOUN ET AL.: THE CERATOCYSTIS PARADOXA Mating system C. euricoi 767 MYCOLOGIA Singly/in chain Obovoid-subglobose (8–)10–16(–20) (7–)8–12(–19) Singly Subglobose (10–)13–16(–18) (9–)10–12(–13) aseptate, initially hyaline, turning grayish sepia (17’’’i) to umber (9), thick-walled when mature, cylindrical to oblong, (8–)9–13(–16) 3 (4–)6–7(–8) mm. Aleurioconidia produced holoblastically, singly or in short chains, dark mouse umber (9), granulated, thick-walled, mostly oblong to subglobose, (8–)10– 16(–20) 3 (7–)8–12(–19) mm. Colonies on MEA initially hyaline to white, becoming green-olivaceous (23’’’i) or gray-olivaceous (23’’’’i) after 10 d, reverse gray-olivaceous (23’’’’i). Mycelium aerial, submerged, hyphae hyaline, smooth, often terminating as conidiophores, septate, no constriction at septa. Optimal temperature for growth 25–30 C, fast growing, 60– 80 mm diam after 36 h at 30 C, marginal growth at 35 C, no growth at 10 C after 10 d. Mating system: Heterothallic. Types: Lectotype of Sporoschisma paradoxum (designated by Nag Raj and Kendrick [1975], p. 129, MBT 178353): FRANCE, exact origin unknown, on fruit of Ananas comosus, 1886, coll. J. de Seynes, represented by illustrations (Plate I, Figs. 22–24) from de Seynes (1886). Previously considered lectotype for Ceratostomella paradoxa (designated by Hunt 1956 p 19, but see argument below): GHANA, Anyinam, on discarded Theobroma cacao pod husks, 1927, coll. H.A. Dade, IMI 41297 5 CB 449. Epitype of Ceratocystis paradoxa (designated here, MBT 178352): CAMEROON, southwest region, Kumba (N4 35.271 E9 28.112), on endosperm of Theobroma cacao pod husk, 12 Oct 2010, coll. M. Mbenoun & J. Roux, dried culture PREM 60766, living ex-epitype culture CMW 36689 5 CBS 130761. Holotype of Stilbochalara dimorpha (MBT 178354): VENEZUELA, Las Trincheras, on rotten Theobroma cacao pod husks, 25 Dec. 1891, coll. H. Lassen, unnumbered specimen from Museum Botanicum Hauniense. Sexual state data from Riedl (1961). Sexual state data from Bliss (1941). b a Singly/in chain Obovoid-subglobose (9–)11–14(–17) (7–)8–11(–13) Singly/in chain Obovoid-subglobose (9–)14–18(–19) (6–)8–11(–12) Singly/in chain Obovoid-subglobose (7–)9–12(–16) (4–)6–8(–10) Singly/in chain Obovoid-subglobose (10–)12–16(–18) (4–)6–9(–11) C. euricoi C. ethacetica C. cerberus Aleurioconidia Aggregation Shape Length Width TABLE III. Continued C. musaruma C. paradoxa s. str. C. radicicolab 768 Additional material examined (MBT 178354): CAMEROON, Littoral Region, Dibamba (N3 54.510 E9 50.177), on cut end of Elaeis guineensis leaf, 15 Oct 2010, coll. M. Mbenoun & J. Roux, living culture CMW 36642 5 CBS 130760, herbarium specimen of dried culture PREM 60765; South West Region, Kumba (N4 35.271 E9 28.112), on cut end of Elaeis guineensis leaf, 12 Oct. 2010, coll. M. Mbenoun & J. Roux, living culture CMW 36686 5 CBS 130762, herbarium specimen of dried culture PREM 60767. Sexual state: CAMEROON, herbarium specimens with sexual structures obtained from crosses, PREM 60777 (CMW 36642 3 CMW 36655), PREM 60775 (CMW 36654 3 CMW 36686), PREM 60776 (CMW 36654 3 CMW 36655). Notes: The asexual state of Ceratocystis paradoxa first was described as Sporoschisma paradoxum de Seynes from pineapple in France, although the origin of the pineapple collection remains unknown (de Seynes 1886). De Seynes (1886) did not designate a type MBENOUN ET AL.: THE CERATOCYSTIS PARADOXA COMPLEX 769 FIG. 4. Sexual and asexual structures from fresh cultures of Ceratocystis paradoxa s. str. (isolates CMW 36642, CMW 36686 and CMW 36689). a. Ascomata with globose base and extended neck. b. Divergent ostiolar hyphae. c. Ellipsoidal ascospores in mucous sheaths. d. Digitate ascomatal ornamentations. e. Flasked-shaped phialidic conidiophore. f. Thick-walled aleurioconidia. g. Cylindrical primary conidia. h. Obovoid secondary conidia. Bars: a 5 100 mm, b–h 5 10 mm. 770 MYCOLOGIA FIG. 5. Sexual and asexual structures from the herbarium specimen (IMI 41297) previously treated as lectotype of Ceratostomella paradoxa. a. Perithecium. b, d. Synnemata. c. Ascomatal appendages. e. Ascospores. f. Primary conidia. g. Secondary condia. h. Phialides extruding chains of primary conidia. Bars: a 5 100 mm, b–h 5 10 mm. specimen, and no herbarium specimens or cultures from his study are available. Nag Raj and Kendrick (1975) designated the illustrations from the original publication as lectotype, which is legitimate under the current code (Articles 8.1, 40.4; McNeill et al. 2012). Saccardo (1892) treated the species in Chalara and Höhnel (1904) placed it in Thielaviopsis. Dade (1928) described the sexual state of this fungus as Ceratostomella paradoxa Dade, mentioning several specimens but not designating one as type. Hunt (1956) designated one of Dade’s original specimens as lectotype (IMI 41301), but Nag Raj and Kendrick (1975) listed IMI 41297 as holotype. The latter specimen also came from the original collection of Dade and is the only specimen presently available in IMI (CABI). The specimen (IMI 41297) currently is labeled as lectotype and clearly corresponds in all respects with the description of Dade (1928). However, considering the abolishment of the dual nomenclature system (Hawksworth 2011, Hawksworth et al. 2011), and based on the suggestions by Hawksworth et al. (2013), the sexual state of C. paradoxa, described subsequently to the asexual state, should not be treated as a new species (i.e. Ceratostomella paradoxa Dade), but as a formal error for a new combination (i.e. Ceratostomella paradoxa [de Seynes] Dade). We thus replaced Dade with de Seynes in the authors of the homotypic synonyms Ceratocystis paradoxa, Ophiostoma paradoxum and Endoconidiophora paradoxa. Following this interpretation, the type of the sexual morph (IMI 41297) no longer has nomenclatural status because the type for the species is that of the basionym (i.e. the illustrations from de Seynes 1886). Although the option was available to designate IMI 41297 (a herbarium specimen) from cacao husks in Ghana as epitype, we chose to designate as epitype a dried specimen of a living culture from which DNA could be obtained, originating also from cacao husks but from Cameroon. Ainsworth and Bisby (1943) listed Stilbochalara as a synonym of Thielaviopsis, based on study of the MBENOUN ET AL.: THE CERATOCYSTIS PARADOXA COMPLEX 771 FIG. 6. Synnemata from a fresh culture of Ceratocystis paradoxa s. str. (isolate CMW 36642). a. Mature synnema with round, dark head. b, c. Young synnema producing hyaline conidia. Bars: a, c 5 100 mm; b 5 10 mm. holotype of its type species, Stilbochalara dimorpha (Ferdinandsen and Winge 1910). This specimen, previously stored in Museum Botanicum Hauniense, was obtained for the present study from the Natural History Museum of Denmark. It also was studied by Nag Raj and Kendrick (1975) who listed the species as a synonym of C. paradoxa, as did Paulin-Mahady et al. (2002). In the original account of Sporochisma paradoxum, de Seynes (1886) did not mention synnemata. However, shortly thereafter he expanded his description of the same material and described synnemata as fructifications reminiscent of those of species in the genera Isaria, Stysanus or Sporocybe (de Seynes 1888). These also were mentioned by several other early authors (Höhnel 1904, 1909; Petch 1910; Dade 1928) and were observed in this study on IMI 41297, previously treated as lectotype of Ceratostomella paradoxa (FIG. 4). Confusion emerged when Höhnel (1904) reduced T. ethacetica to synonymy with Sporochisma paradoxum and treated the latter species in Thielaviopsis. Thielaviopsis ethacetica was described earlier without synnemata from sugarcane in Java (Went 1893). Although Nag Raj and Kendrick (1975) did not mention synnemata in their diagnosis of the species (p 128), they stated on p 55 that ‘‘Chalara paradoxa … is also known to form occasional lax coremia’’ (5 synnemata). Based on the descriptions of synnemata in Sporochisma paradoxum by de Seynes (1888) and the obvious synnemata observed in the original material of Ceratostomella paradoxa in this study, we conclude that Sporochisma paradoxum and Ceratostomella paradoxa represent the same taxon and we define C. paradoxa s. str. as a species forming synnemata. The shape and size of these structures are highly variable even within the same culture. In the present study isolates of C. paradoxa s. str. produced synnemata under the same conditions as, and alongside ascomata, in paired cultures on water agar supplemented with chips of sugarcane. They consisted of bundles of erect phialides, held together but individually extruding slimy conidia at their apices (FIG. 5). Conidia are agglutinated in round spore drops at the tips of synnemata. Initially slimy and cream-colored, the spore drops turned black when old, consisting essentially of matured, dry 772 MYCOLOGIA secondary conidia (FIG. 5). We thus designated a dried specimen of a fresh, synnema-producing culture as epitype based on its similarity to IMI 41297. The similar structures observed on the holotype of Stilbochalara dimorpha, although fragmented (SUPPLEMENTARY FIG. 6), confirmed that this species represents the same fungus and is best treated as synonym of C. paradoxa s. str. The forms of C. paradoxa that do not produce synnemata are discussed under C. ethacetica. Ceratocystis ethacetica (Went) Mbenoun & Z.W. de Beer, comb. nov. FIG. 7 MycoBank MB805506 ; Thielaviopsis ethacetica Went, Meded. Proefstn W. Java ‘Kagok’ 5:4. 1893. [as ‘ethaceticus’] (basionym) 5 Endoconidium fragrans Delacr., Bull. Soc. Mycol. Fr. 9:184. 1893. 5 Catenularia echinata Wakker in Wakker & Went, de Ziekten van het Suikerriet op Java, EJ Brill, Leiden p 196. 1898. Ascomata perithecial, formed in patches along the lines of interaction between mating-compatible colonies in paired cultures on agar (MEA and WA) and on plant chips (sugarcane and pineapple), absent in single colonies but observed in nature on endosperma of cacao pod husks. Ascomatal bases fully or partially submerged in substrata, mostly globose, (107–)154–215(–260) mm high 3 (125–)156–216 (–251) mm wide, originally straw (21f), appearing dark in old cultures when surrounded with aleurioconidia and ascomatal appendages. Ascomatal appendages stellate or digitate, (22–)22–41(–51) 3 (12–) 14–23(–27), mostly restricted to aerial parts of partially submerged ascomatal bases. Ascomatal necks dark mouse gray (14’’’’’k), erect, (451–)672–862(–983) mm long, (20–)22–29(–35) mm wide at apices, (33–)44– 64(–79) mm wide at bases. Ostiolar hyphae hyaline, divergent, (71–)92–112(–121) mm long. Asci not observed. Ascospores in sheaths, hyaline, aseptate, ellipsoidal, (7–)7–9(–10) 3 (2–)3–4(–4) mm, accumulating in mucilaginous droplets, buff (19’’f ) at tips of ascomatal necks with strong adherence to ostiola hyphae when old and dry. Conidiophores mostly hyaline, phialidic, lageniform, (75–)87–148(–257) mm long, (4–)6–8(–10) mm wide at bases and (3–)4–5 (–6) mm wide at apices, mononematous with enteroblastic conidium ontogeny, solitary. Primary conidia hyaline, aseptate, cylindrical, (5–)7–8(–10) 3 (2–)4– 5 (–6) mm. Secondary conidia aseptate, initially hyaline, turning grayish sepia (17’’’i), thick-walled at maturity, cylindrical to oblong, (7–)8–11(–16) 3 (4–)6–7(–7) mm. Aleurioconidia produced holoblastically, singly or in chains of 2–10 units, grayish sepia (17’’’i) to umber (9), granulated, thick-walled, subglobose, oblong or ovoid, (9–)14–18(–19) 3 (6–)8–11(–12) mm. Colonies on MEA initially hyaline to white, progressively darkening, turning citrine-green (25’’b), green-olivaceous (23’’’i) or gray-olivaceous (23’’’’i) after 10 d, reverse gray-olivaceous (23’’’’i). Mycelium aerial and submerged, hyphae hyaline, smooth, often terminating as conidiophores, septate, no constriction at septa. Optimal temperature around 30 C, fast growing, reaching , 75 mm in 36 h at 30 C, marginal growth at 35 C, no growth at 10 C after 10 d. Mating system: Heterothallic. Types: Lectotype of Thielaviopsis ethacetica (designated here, MBT 178355): INDONESIA, Java, Tegal, on stems, fruit, and leaves of Saccharum sp., 1893, coll. F.A.F.C. Went, represented by line drawings (Plate III, Figs. 1–6) from Went (1893). Epitype of Thielaviopsis ethacetica (designated here, MBT 178356): MALAYSIA, Western Malaysia, on fruit of Ananas comosus, 8 Jul 1952, coll. A. Johnson, dried culture PREM 60961, living ex-epitype culture IMI 50560 (CABI) 5 MUCL 2170 5 CMW 37775. Lectotype of Endoconidium fragrans (designated here, MBT 178357): FRANCE, Paris, in fermented pineapple juice, 1893, coll. G. Delacroix, represented by line drawings (Plate XI, Fig. IIa, b) from Delacroix (1893). Neotype of Catenularia echinata (designated here, MBT 178358): SOUTH AFRICA, KwaZulu-Natal Province, Saccharum sp., 2010, coll. N. van Wyk, dried culture PREM 60963, living ex-neotype culture CMW 36771. Additional specimens examined: CAMEROON, southwest region, near Tiko (N4 14.488 E9 21.698), on stump of felled Elaeis guineensis tree, 13 Oct 2010, coll. M. Mbenoun & J. Roux, culture CMW 36671 5 CBS 130757, herbarium specimen PREM 60762; littoral region, Njombe (N4 34.032 E9 37.204), on damaged leaf of Ananas comosus, 14 Oct 2010, coll. M. Mbenoun & J. Roux, culture CMW 36725 5 CBS 130758, herbarium specimen PREM 60763; center region, Ngomedzap (N316.107 E11 14.498), on endosperm of Theobroma cacao pod husk, Dec 2010, coll. M. Mbenoun, culture CMW 36741 5 CBS 130759, herbarium specimen PREM 60764; herbarium specimens with sexual structures obtained from crosses, PREM 60774 (CMW 36671 3 CMW 36691), PREM 60771 (CMW 36671 3 CMW 36741), PREM 60772 (CMW 36671 3 CMW 36644), PREM 60773 (CMW 36671 3 CMW 36745). Notes: Thielaviopsis ethacetica was described originally as the type species for a new genus, Thielaviopsis, from sugarcane in Java (Went 1893). In his description of the species, Went (1893) did not mention or illustrate synnemata or any similar structures. Höhnel (1904) found a similar fungus on coconut in Vienna, Austria, but it produced synnemata in addition to other conidiogenous structures. He thought his MBENOUN ET AL.: THE CERATOCYSTIS PARADOXA COMPLEX 773 FIG. 7. Sexual and asexual structures from fresh cultures of Ceratocystis ethacetica (isolates CMW 36671, CMW 36725, CMW 36741, CMW 37775). a. Perithecium with globose base. b. Divergent ostiolar hyphae. c. Ellipsoidal ascospores invested in mucous sheaths. d. Digitate ornamentations on perithecial base. e. Oblong secondary conidia. f. Cylindrical primary conidia. g. Thick-walled aleurioconidia in short chains. h. Flask-shaped phialidic conidiophore. Bars: a 5 100 mm, b–e 5 10 mm. 774 MYCOLOGIA material represented Sporoschisma paradoxum but sent it to Went for confirmation. Went suggested that the fungus from Vienna was conspecific with T. ethacetica (Höhnel 1904), despite the presence of synnemata. Höhnel (1904) transferred Sporoschisma paradoxum to Thielaviopsis, which he considered a more appropriate genus, and treated T. ethacetica as synonym of T. paradoxa, based on the older epithet. Höhnel’s (1904) synonymy was accepted in all major treatments of the species (Petch 1910, Dade 1928, Davidson 1935, Hunt 1956, Nag Raj and Kendrick 1975, Upadhyay 1981, Paulin-Mahady et al. 2002). Of note, in the majority of descriptions of C. paradoxa following Dade (1928) no mention was made of synnemata (Davidson 1935, Hunt 1965, Morgan-Jones 1967, Upadhyay 1981, Paulin-Mahady et al. 2002). As explained under C. paradoxa, results of this study show that C. paradoxa s. str. is characterized by synnemata and that isolates with and without synnemata, but otherwise morphologically similar, are phylogenetically distinct. The name T. ethacetica therefore is reinstated for species without synnemata. In view of the lack of type material for T. ethacetica from sugarcane in Indonesia, the illustrations from Went (1893) are designated as lectotype. To ensure stability, a dried specimen of a morphologically similar living culture from pineapple in Malaysia, which lacks synnemata and clustered separately from C. paradoxa s. str. in phylogenetic analyses, is designated as an epitype for this species. Delacroix (1893) described and illustrated a fungus with hyaline conidia and without synnemata from fermented pineapple juice in France as Endoconidium fragrans. Höhnel (1909) thought the species represented an immature stage of T. paradoxa and treated it as a synonym of the latter species. The synonymy was accepted in most subsequent taxonomic treatments (Saccardo 1913, Nag Raj and Kendrick 1975, Paulin-Mahady et al. 2002). Nag Raj and Kendrick (1975) reported that the original specimen of Endoconidium fragrans had been lost. Because no synnemata were illustrated or mentioned in the cryptic description by Delacroix (1893), we suggest that the name should be treated as a synonym of C. ethacetica. However, to ensure taxonomic stability, we designated the illustrations of Delacroix (1893) as lectotype for Endoconidium fragrans. In an extensive treatment of diseases of sugarcane in Java (Wakker and Went 1898), Wakker described Catenularia echinata with pigmented macroconidia and hyaline, enteroblastic conidia. In the same paper T. ethacetica was treated as distinct, but the descriptions of the two species largely overlap and no explanation was provided as to why these fungi should be treated separately. Based on their similarities, Höhnel (1909) suggested Catenularia echinata as a possible synonym of T. paradoxa. Apart from Saccardo (1899), who initially listed it as a distinct species but later followed Höhnel’s synonymy (Saccardo 1913), the species was ignored in all subsequent treatments of the species complex. In the absence of synnemata and based on its similarities with C. ethacetica, Catenularia echinata is best treated as a synonym of the latter species. With no material or illustrations available for Catenularia echinata, lectotypification is not possible, but we designated a neotype for the name from a morphologically similar isolate from sugarcane in South Africa. Ceratocystis euricoi (Bat. & A.F. Vital) Mbenoun, Z.W. de Beer, comb. nov. FIG. 8a–d, FIG. 9 MycoBank MB805507 ; Hughesiella euricoi Bat. & A.F. Vital, Anais Soc. Biol. Pernambuco 14:142. 1956. (basionym) ; Thielaviopsis euricoi (Bat. & A.F. Vital) A.E. Paulin, T.C. Harr. & McNew, in Paulin-Mahady et al., Mycologia 94:70. 2002. Ascomata not observed. Conidiophores hyaline, 1– 4 septa, phialidic, lageniform, (77–)101–144(–171) mm long, (6–)7–10(–12) mm wide at bases and (3–)4–6 (–7) mm wide at apices, mononematous with enteroblastic conidium ontogeny, commonly solitary, occasionally aggregated in synnemata, variable in size, (853–)946–1295(–1380) mm long 3 (54–)56–82(–87) mm wide. Primary conidia hyaline, aseptate, cylindrical, (8–)9–11(–12) 3 (3–)4–5(–5) mm. Secondary conidia aseptate, initially hyaline, turning mouse gray (15’’’’’) and thick-walled at maturity, mostly oblong, (10–)11–16(–25) 3 (5–)6–7(–8) mm. Aleurioconidia produced holoblastically, singly or in short chains, dark mouse gray (15’’’’’k), granulated, thick-walled, mostly subglobose to globose, (9–)11–14(–17) 3 (7–) 8–11(–13) mm. Colonies on MEA initially hyaline to white, becoming gray-olivaceous (23’’’’i) with age. Mating system: Undetermined. Type: Holotype of Hughesiella euricoi (MBT 178359): BRAZIL, Bahia, Salvador, atmospheric air, 1956, coll. E.A.F. da Matta, holotype (not seen) URM 640, isotype (not seen) DAOM 75211, culture exholotype CBS 893.70 5 CMW 28537 5 MUCL 1887 5 UAMH 1382. Additional specimens examined: INDONESIA, Sulawesi, on trunk of Cocos nucifera, 29 Nov. 2002, M.J. Wingfield, CMW 8799. UNKNOWN ORIGIN, on endosperm of Cocos nucifera, Mar 1922, coll. H.G. Derx, CBS 107.22 5 CMW 28538 5 MUCL 8358. Notes: Hughesiella euricoi was described originally from an air sample in Brazil as type species of the genus Hughesiella (Batista and Vital 1956). Nag Raj and Kendrick (1975) considered the species a synonym of Chalara paradoxa based on morpholog- MBENOUN ET AL.: THE CERATOCYSTIS PARADOXA COMPLEX 775 FIG. 8. Asexual structures from fresh cultures of Ceratocystis euricoi (ex holotype CMW 28537, CMW 28538, CMW 8799). a. Phialide. b. Primary conidia. c. Secondary conidia. d. Aleurioconidia. Ceratocystis musarum (isolate CMW 1546). e. Phialides. f. Primary conidia. g. Secondary conidia. h. Aleurioconidia. Ceratocystis radicicola (isolate CMW 37776). i. Phialides. j. Primary conidia. k. Secondary conidia. l. Aleurioconidia. Bars: a, e, i 5 20 mm; b, c, d, f, g, h, j, k, l 5 10 mm. FIG. 9. Synnemata produced by Ceratocystis euricoi ex holotype (CMW 28537). Bars 5 1000 mm. ical similarities and thus listed the genus Hughesiella as synonym of Chalara. Paulin-Mahady et al. (2002) did not include any material of the species in their study but treated T. euricoi as distinct from T. paradoxa and listed Hughesiella as synonym of Thielaviopsis. In the present study, the ex-type culture (CMW 28537 5 CBS 893.70) of the fungus studied by Batista and Vital (1956) appeared to have similar asexual morphology as C. paradoxa s. str., including the formation of synnemata (FIG. 11), not mentioned in the original description. However, DNA sequence analyses confirmed that T. euricoi belongs in the C. paradoxa complex but that this species is distinct from C. paradoxa s. str. and other species in the complex. Several isolates from coconut palm in Indonesia, as well as one (CMW 28538 5 CBS 107.22) isolated by a Dutch mycologist in the 1920s from the same host but of an unknown origin, previously thought to be C. 776 MYCOLOGIA FIG. 10. Sexual and asexual structures from the holotype of Ceratocystis musarum. a. Ascomata. b. Primary conidia. c. Secondary conidia. Bars: a 5 100 mm; b, c 5 10 mm. paradoxa, were shown to be conspecific with the Brazilian isolate. Although no sexual state has been observed for this species, the one fungus one name principles adopted in the Melbourne Code require that the species be treated in Ceratocystis (Hawksworth 2011, Hawksworth et al. 2011). Ceratocystis musarum Riedl, Sydowia 15:248. 1962. FIGS. 8e–h, 10 MycoBank MB327636 5 Thielaviopsis musarum (R.S. Mitchell) Riedl, Sydowia 15:249. 1962. [5 Thielaviopsis paradoxa var. musarum R.S. Mitchell, J. Coun. Sci. Ind. Res. Australia, 10:130. 1937. nom. inval., Art. 39.1] ‘‘Ascomata perithecial. Ascomatal bases partially submerged or on the surface of substrate, sienna (14i), globose to subglobose, 300 mm high 3350 mm wide, no ornamentations observed. Ascomatal necks sienna (14i) to orange (13), 1100–1200 mm long, 17– 18 mm wide at apices and 50 mm wide at bases. Ostiolar hyphae hyaline, convergent, , 100 mm long. Ascospores hyaline, cylindrical, 6–11 mm long, 2–3.5 mm thick’’ (Riedl 1962). Conidiophores hyaline, 1–5 septa, phialidic, lageniform, (174–)206–252(–295) mm long, (5–)7–9(–10) mm wide at bases and (4–)4– 5(–6) m m wide at apices, mononematous with enteroblastic conidium ontogeny, solitary, emerging laterally or terminal on hyphae. Primary conidia hyaline, aseptate, cylindrical, (9–)10–12(–13) 3 (3–) MBENOUN ET AL.: THE CERATOCYSTIS PARADOXA 4–5(–5) mm. Secondary conidia aseptate, initially hyaline, turning umber (9) and thick-walled at maturity, mostly oblong, (9–)10–13(–15) 3 (5–)6–7 (–8) mm wide. Aleurioconidia produced holoblastically, singly or in short chains, dark mouse gray (14’’’’’k), granulated, thick-walled, mostly ovoid to subglobose, (10–)12–16(–18) 3 (4–)6–9(–11) mm wide. Colonies on MEA initially hyaline to white, becoming smoke gray (21’’’’f ) with age. Mating system: Undetermined. Types: Holotype of Ceratocystis musarum (MBT 178360): AUSTRIA, Vienna, peduncles of Musa speciosa (5 M. ornata), 20 Feb 1962, coll. H. Riedl, W 28259 (Naturhistorisches Museum Wien). Epitype of Ceratocystis musarum (designated here, MBT 178360): NEW ZEALAND, on Musa sp., coll. T. W. Canter-Visscher, dried culture PREM 60962, living ex-epitype culture CMW 1546 5 C 907. Notes: Mitchell (1937) described a new variety of T. paradoxa associated with stem-end rot of banana in Australia. This description was invalid however in that it lacked a Latin diagnosis. Riedl (1962) isolated a similar fungus from banana stems in Vienna, although the source of the bananas was probably not Vienna, and described it as a new species, distinct from T. paradoxa. Following the practice of dual nomenclature, he named the sexual and asexual states of the fungus separately, altering the rank of the asexual state described by Mitchell (1937) from variety to species level. Both de Hoog (1974) and Nag Raj and Kendrick (1975) accepted the species from banana as distinct from C. paradoxa, while Upadhyay (1981) listed C. musarum as synonym of C. paradoxa based on a single specimen originating from banana in Canada. The material of Mitchell and that of Riedl was not included in the latter three studies. Paulin-Mahady et al. (2002) did not mention T. musarum in their treatment of Thielaviopsis. In a subsequent study (Harrington 2009) the sequence of an isolate (C 1480) presumably from banana, but of unknown origin, grouped separately from C. paradoxa s. str. and was labeled C. musarum in a phylogenetic tree. Based on this sequence, de Beer et al. (2013b) treated C. musarum as a distinct taxon. In this study we compared the holotype of C. musarum and the original description of its asexual state (Riedl 1962) with an isolate (CMW 1546) from banana in New Zealand and found that they correspond morphologically. The dried specimen of the latter isolate is designated here as epitype for C. musarum. The ITS sequence for this isolate corresponded with the ITS sequence produced for the same isolate by Witthuhn et al. (1996 unpubl). The two sequences differed by three bp from a sequence produced in another study by Witthuhn et al. (1999) COMPLEX 777 and an unpublished sequence of an isolate from date palm in Saudi Arabia (FIG. 3). The b-tubulin and TEF1a sequences of the epitype were clearly distinct from all other species in the complex (SUPPLEMENTARY FIGS. 3–4), with the TEF-1a sequence corresponding with the one from banana in Harrington (2009). Ceratocystis radicicola (Bliss) C. Moreau, Rev. Mycol. (Paris) Suppl. Col. 17:22. 1952. FIG. 8i–l MycoBank MB294235 ; Ceratostomella radicicola Bliss, Mycologia 33:468. 1941. (basionym) ; Ophiostoma radicicolum (Bliss) Arx, Antonie van Leeuwenhoek 18:211. 1952. 5 Chalaropsis punctulata Hennebert, Antonie van Leeuwenhoek 33:334. 1967. ; Thielaviopsis punctulata (Hennebert) A.E. Paulin, T.C. Harr. & McNew, Mycologia 94:70. 2002. ‘‘Ascomata perithecial. Ascomatal bases partially or completely submerged, faintly colored, nearly spherical, 180–320 mm diam. Ascomatal appendages variously branched, 35–90 mm long. Ascomatal necks dark, becoming hyaline at the apices, 440–980 mm long, 24–71 mm diam. Ostiolar hyphae hyaline and fimbriate. Asci deliquescent, not observed. Ascospores hyaline, ellipsoidal, sides unequally convex, continuous, 8–15 3 3–4 mm, covered with a mucous sheath’’ (Bliss 1941). Conidiophores hyaline, phialidic, lageniform, (108–)141–207(–252) mm long, (5–)7–8(–10) mm wide at bases and (4–)4–5(–6) mm wide at apices, mononematous with enteroblastic conidium ontogeny, solitary, emerging laterally or terminal on hyphae. Primary conidia hyaline, aseptate, cylindrical or rectangular, (6–)8–14(–16) 3 (3–)4–5(–5) mm. Secondary conidia aseptate, initially hyaline, turning umber (9) and thick-walled at maturity, mostly oblong, (6–)10–14(–16) 3 (3–)4–6(–6) mm. Aleurioconidia produced holoblastically, produced singly, initially hyaline, dark mouse gray (14’’’’’k) when mature, granulated, thick-walled, ovoid to subglobose with a flattened base, (10–)(11–)13–16(–18) 3 (9–)10–12 (–13) mm wide. Colonies on MEA initially hyaline to white, becoming dark gray (21’’’’f ) with age. Mating system: Heterothallic (Bliss 1941, El-Ani et al. 1957). Types: Holotype of Ceratostomella radicicola (MBT 378362): USA, California, Indio, root and trunk of Phoenix dactylifera, coll. D.E. Bliss, holotype (not seen) BPI 596268, isotype (not seen) IMI 036479, culture ex-holotype CBS 114.47 5 CMW 1032 5 MUCL 9526. Holotype of Chalaropsis punctulata (MBT 2699): MAURITANIA, Atar, in roots of Lawsonia inermis, Apr 1966, coll. J. Brun, holotype not seen (CBS), 778 MYCOLOGIA culture ex-holotype CBS 167.67 5 ATCC 18454 5 MUCL 8674 5 CMW 26389 5 IFAC H-A1. Additional culture examined: IRAQ, on Phoenix dactylifera, 1993, A. Bahadli, CMW 37776 5 IMI 316225. Notes: Ceratocystis radicicola was isolated first and described from the roots of dying Phoenix dactylifera in California (Bliss 1941). The species was treated subsequently in Ophiostoma (von Arx 1952) and until present in Ceratocystis (Moreau 1952, Hunt 1956, de Hoog 1974, Nag Raj and Kendrick 1975, Upadhyay 1981). When Hennebert (1967) described Chalaropsis punctulata, he recognized that it was morphologically similar to the asexual state of C. radicicola but distinguished the two species based on conidial sizes. Nag Raj and Kendrick (1975) suggested that the conidial sizes overlapped and that the two species might be synonyms. Paulin-Mahady et al. (2002) showed that ITS sequences of the two species were identical and that the ex-type isolate of Chalaropsis punctulata could mate with an isolate of C. radicicola. In a phylogenetic tree based on concatenated sequence data for ITS, b-tubulin and TEF-1a produced by van Wyk et al. (2009), the extype of Chalaropsis punctulata (CMW 26389 5 CBS 167.67) grouped not with C. radicicola but closest to C. fagacearum. When considered separately the TEF1a blasted to C. adiposa (HM569644, Harrington 2009) and was in conflict with the other two gene regions. The ITS and b-tubulin sequences were identical to those obtained in this study for the extype of C. radicicola, CMW 1032 5 CBS 114.47 and isolate CMW 37776 5 IMI 316225. Based on these discrepancies, we have resequenced the three gene regions for isolate CMW 26389 5 CBS 167.67 (ITS: KF953932, b-tubulin: KF953931, TEF-1a: KF917202). As found previously, the ITS and BT sequences for these isolates were identical. The TEF1-a sequence differed in one bp between CMW 26389 and CMW 1032 5 CBS 114.47 and eight bp from CMW 37776 5 IMI 316225, and grouped clearly within the C. radicicola clade (FIG. 2). Based on these results, we endorse the synonymy between T. punctulata and C. radicicola suggested by Paulin-Mahady et al. (2002). initially straw (21f ), unornamented, dark, covered with stellar or digitate appendages at maturity, (26–) 27–37(–39) 3 (17–)19–27(–27). Ascomatal necks dark brown, commonly erect or curled and branched as observed in the holotype, (280–)650–984(–1244) mm long, (24–)29–43(–56) mm wide at apices, (43–) 61–83(–101) mm wide at bases. Ostiolar hyphae hyaline, divergent, (47–)70–108(–142) mm long. Asci not observed. Ascospores in sheaths, hyaline, aseptate, ellipsoidal (6–)7–9(–13) 3 (3–)3–4(–6), accumulating in easily discharged, mucilaginous, buff (19’’f ) droplets at apex of ascomatal necks. Conidiophores hyaline, phialidic, lageniform, tapering toward apices, (63–)92–150(–215) mm long, (5–)6– 7(–8) mm wide at bases and (3–)4–5(–6) mm wide at apices, mononematous with enteroblastic conidium ontogeny, solitary. Primary conidia hyaline, aseptate, cylindrical, (7–)7–12(–19) 3 (3–)4–5(–6) mm. Secondary conidia not observed. Aleurioconidia produced holoblastically, singly or in short chains of 2–3 units, umber (9), granulated, thick-walled, subglobose, obovoid (7–)9–12(–16) 3 (4–)6–8(–10) mm. Colonies on MEA gray-olivaceous (21’’’’b) or olivaceous-gray (21’’’’’b), reverse gray-olivaceous (21’’’’b). Mycelium mostly aerial, hyphae smooth and septate, not constricted at septa. Optimal temperature around 30 C, fast growing, , 70 mm on average in 36 h, no growth at 10 C or 35 C after 10 d. Mating system: Homothallic. Types: Holotype of Ceratocystis cerberus: CAMEROON, southwestern region, near Tiko (N4 10.377 E9 25.419), on stump of felled Elaeis guineensis tree, 13 Oct 2010, coll. M. Mbenoun & J. Roux, holotype PREM 60770 (PREM), culture ex-holotype CBS 130765 5 CMW 36668. Paratypes of Ceratocystis cerberus: CAMEROON, littoral region, Dibamba (N3 54.510 E9 50.177) on cut end of Elaeis guineensis leaf, 15 Oct 2010, coll. M. Mbenoun & J. Roux, paratype PREM 60769 (PREM), culture ex-paratype CBS 130764 5 CMW 36653; central region, Bokito (N4 30.279 E11 04.748), on wound on Theobroma cacao, Dec 2009, coll. M. Mbenoun & J. Roux, paratype PREM 60768 (PREM), culture ex-paratype CBS 130763 5 CMW 35021. Ceratocystis cerberus Mbenoun, M.J. Wingf. & Jol. Roux, sp. nov. FIG. 11 MycoBank MB805508 Additional cultures examined: CAMEROON, central region, Bokito (N4 30.279 E11 04.748), from trunk wound on Theobroma cacao, Dec 2009, coll. M. Mbenoun & J. Roux, CMW 35024; littoral region, Dibamba (N3 54.510 E9 50.177) on cut end of Elaeis guineensis leaf, 15 Oct 2010, coll. M. Mbenoun & J. Roux, CMW 36641. Etymology: Epithet refers to the multi-necked ascomata observed in the ex-type isolate of this species. Ascomata perithecial, formed superficially on surface of substratum or suspended in aerial mycelium. Ascomatal bases mostly globose, (137–)260–340 (–370) mm high 3 (148–)268–348(–368) mm wide, Notes: Ceratocystis cerberus is the only known homothallic species in the C. paradoxa complex. It is also the only species that did not produce distinctive secondary conidia under the conditions used in the present study (TABLE III). Under certain MBENOUN ET AL.: THE CERATOCYSTIS PARADOXA COMPLEX 779 FIG. 11. Sexual and asexual structures from fresh cultures of Ceratocystis cerberus (isolates CMW 36668, CMW 36653, CMW 35021). a. Perithecium with globose base. b. Divergent ostiolar hyphae. c. Ellipsoidal ascospores invested in sheaths. d. Cylindrical primary conidia. e. Digitate ascomatal ornamentation. f. Thick-walled aleurioconidia in short chains. g. Perithecium with multiple dichotomous necks. h. Flasked-shaped phialidic conidiophore. Bars: a, g 5 100 mm; b–e, g–h 5 10 mm. 780 MYCOLOGIA conditions C. cerberus forms ascomata with multiple dichotomous necks. In Ceratocystis this feature has been observed only in C. mangivora M. van Wyk & M.J. Wingf. (van Wyk et al. 2011). DISCUSSION Before this study, only four species were recognized in the Ceratocystis paradoxa complex, including C. musarum, C. paradoxa, C. radicicola and Thielaviopsis euricoi (Harrington 2009, Wingfield et al. 2013). Multigene DNA phylogenies in combination with mating studies and careful morphological examination of type material of all species and their synonyms revealed that the complex includes more species than previously recognized. Five species, including C. paradoxa s. str., are redefined and their descriptions amended. Lectotypes are designated for C. ethacetica and Endoconidium fragrans (synonym of C. ethacetica), epitypes were designated for C. paradoxa s. str., C. ethacetica and C. musarum, and a neotype was designated for Catenularia echinata (synonym of C. ethacetica). Two of the species, previously treated in Thielaviopsis, are transferred to Ceratocystis following the one fungus one name principles. A sixth species from Cameroon is described as new. Few morphological differences could be used as a diagnostic tool to distinguish species within the C. paradoxa complex. Sexual structures are not known for all species and are not commonly observed in nature or in culture because of the predominantly heterothallic mating system in this group. Furthermore, the characteristics of these structures, where they have been observed, overlap among species (Dade 1928, Bliss 1941; see also TABLE III). Likewise, the shape and size of asexual structures mostly overlap. The only species with a potential diagnostic character is C. radicicola, which can be discriminated from all other known species based on its aleurioconidia that are borne singly. DNA sequences are therefore critically important for accurate identification of species in this complex. Of the three gene regions used in this study, the TEF1-a emerged as the marker providing the best resolution in the group while ITS showed the least resolution. In particular, ITS could not discriminate C. paradoxa s. str. and C. euricoi. The use of this marker as a default barcode system in the C. paradoxa complex therefore should be avoided, similar to what has been shown for members of the C. moniliformis sensu lato species complex (van Wyk et al. 2006; Kamgan Nkeukam et al. 2008, 2013). Based on the dual nomenclature system for fungi, the sexual and asexual states of Ceratocystis species were placed in different genera. Following the rules of the Melbourne Code, these genera are now considered synonyms of Ceratocystis (de Beer et al. 2013b). Three of these genera that previously were considered anamorph-form genera are typified by currently recognized members of the C. paradoxa complex. The type species of Thielaviopsis is T. ethacetica (5 C. ethacetica). Stilbochalara dimorpha (5 C. paradoxa s. str.) is the type species of Stilbochalara, and Hughesiella euricoi (5 C. euricoi) is the type species of Hughesiella. Following the Melbourne Code (Hawksworth 2011, McNeill et al. 2012) and the suggestions for a pragmatic approach to the naming of plant pathogens (Wingfield et al. 2012), these genus names are taxonomically convergent with Ceratocystis (de Beer et al. 2013b). However, if the genus Ceratocystis is split into smaller genera as proposed by Wingfield et al. (2013), Thielaviopsis could be reinstated based on priority to accommodate species in the C. paradoxa complex. Based on combined ITS, b-tubulin and TEF1-a gene phylogenies, the genealogical structure of the C. paradoxa complex includes three major clades. Clade 1 occupies a basal position and includes C. paradoxa s. str., C. euricoi and an unnamed species represented by the isolate CMW 28535 5 CBS 101054. The three species are conjointly characterized by the production of synnemata, a feature observed in the original description of de Seynes (1888) and subsequently reported only by Petch (1910). The latter author suggested that the production of synnemata is triggered by harsh conditions (e.g. impoverished and dry substrates). In the present study, only some of the isolates of C. paradoxa s. str. and C. euricoi produced synnemata. This occurred in paired cultures on WA supplemented with sugarcane chips originally set up to test for sexual compatibility. Of note, in the unassigned isolate CMW 28535 5 CBS 101054, synnemata formed readily and abundantly on MEA. The formation of synnemata might represent a distinctive homologous character to members of Clade 1. Because it is variably expressed among and between species, the possibility that this feature is more common across the C. paradoxa s. lat. lineage and only requires adequate conditions to express, however, cannot be overruled a priori. Clade 2 of C. paradoxa s. lat. is composed of C. ethacetica and C. musarum. These two species have similar morphological characteristics in their asexual states, shared by other members of the C. paradoxa complex. However, their sexual states differ markedly in morphology. For example, from the description of Riedl (1962), the sexual state of C. musarum includes the absence of ornamentation on ascomatal bases, convergent ostiolar hyphae and cylindrical ascospores without sheaths. These characteristics do not fit with the general descriptions of other members of C. MBENOUN ET AL.: THE CERATOCYSTIS PARADOXA paradoxa s. lat. Based on examination of the holotype specimen of C. musarum we concluded that the fungus was at an early stage of development when it was preserved and, hence, ascomata were immature. We did not recover any ascospores from this specimen despite the abundant presence of ascomata. This also supports the view that Riedl (1962) could have confused primary conidia for ascospores. Vovlas et al. (1994) reported ‘‘C. paradoxa’’ in association with the nematode Helicotylenchus multicinctus (Cobb), causing necrotic lesions on roots and rhizomes of declining bananas in Sao Tome and Principe. Although the identity of the fungus cannot be ascertained, their description of its sexual structures conforms with those expected for C. paradoxa s. lat. and suggests a heterothallic mating system. Although statistically well supported, Clade 3 of C. paradoxa s. lat. is heterogeneous in terms of morphological and biological characteristics of the fungi in this group. It could be separated into two subclades respectively represented by C. radicicola and C. cerberus. Ceratocystis radicicola is heterothallic (Bliss 1941) and the only species in which aleurioconidia are borne singly. On the other hand, C. cerberus forms aleurioconia in short chains and is the only species producing only one endoconidial form, in that no secondary conidia were observed in any of the isolates examined in this study. Moreover, C. cerberus is homothallic and produces large numbers of ascomata in isolates derived from single spores or hyphal tips on artificial media. This characteristic has not been observed in other species of C. paradoxa s. lat. with known sexual states. This suggests that homothallism is most likely a derived character in C. cerberus. Of note, Harrington (2009) mentioned a homothallic strain (C1753) of ‘‘C. paradoxa’’, which in this study was closely related with but distinct from C. cerberus based on TEF-1a. The putatively undescribed species represented by CMW 28536 5 CBS 116770 also falls in the same clade. However, this strain did not produce perithecia in our tests. This does not preclude the possibility that it is homothallic and might have lost its ability to produce ascomata in culture. The taxonomic reevaluation of C. paradoxa s. lat. prompts a reconsideration of the host range and geographic distribution of species in this complex. Ceratocystis paradoxa s. lat. has been considered cosmopolitan. The results of this limited study suggest that C. paradoxa s. str. is far less common than previously considered. It most commonly has been confused with C. ethacetica, which better fits the description of widespread distribution and broad host range generally associated with C. paradoxa s. lat. (Morgan-Jones 1967, Anonymous 1987). The host range of Ceratocystis paradoxa s. lat. includes economically important crops COMPLEX 781 such as banana, cacao, coconut palm, date palm, oil palm, pineapple and sugarcane. These plants have spread around the world, expanding the geographic range of their associated and possibly coevolved pathogens, and providing them with new opportunities for host expansion. Clarifying the taxonomy of species in the C. paradoxa complex will assist in more accurate identification of these pathogens. It also should provide a solid foundation for research to determine their centers of origin, pathways of movement and the management of the diseases that they cause. ACKNOWLEDGMENTS We thank Walter Gams, David Hawksworth and John McNeill for advice and comments regarding various taxonomic issues treated in this study. Aimé D. Begoude Boyogueno and Alain C. Misse provided assistance during field studies for which we are most grateful. We acknowledge the financial assistance of the Department of Corporate International Affairs of the University of Pretoria, members of the Tree Protection Cooperative Programme (TPCP) and the Department of Science and Technology (DST)/National Research Foundation (NRF) Center of Excellence in Tree Health Biotechnology (CTHB). We recognize logistical support of the Institute of Agricultural Research for Development (IRAD) during field studies in Cameroon. LITERATURE CITED Abdullah SK, Asensio L, Monfort E, Gomez-Vidal S, Salinas J, Lopez Lorca LV, Jansson HB. 2009. Incidence of the two date palm pathogens, Thielaviopsis paradoxa and T. punctulata, in soil from date palm plantations in Elx, southeast Spain. J Plant Prot Res 49:276–279, doi:10.2478/v10045-009-0043-z Àlvarez E, Llano GA, Loke JB, Chacon MI. 2012. Characterization of Thielaviopsis paradoxa isolates from oil palms in Colombia, Ecuador and Brazil. J Phytopathol 160:690–700, doi:10.1111/jph.12012 Ainsworth GC, Bisby GR. 1943. A dictionary of the fungi. 1st ed. Surrey, UK: IMI. 771 p. Anonymous. 1987, Ceratocystis paradoxa. Distribution maps of plant diseases No. 142. 5th ed. Wallingford, UK: CAB International. Baker CJ, Harrington TC, Krauss U, Alfenas AC. 2003. Genetic variability and host specialization in the Latin American clade of Ceratocystis fimbriata. Phytopathology 93:1274–1284, doi:10.1094/PHYTO.2003.93.10.1274 Baker Engelbrecht CJ, Harrington TC. 2005. Intersterility, morphology and taxonomy of Ceratocystis fimbriata on sweet potato, cacao and sycamore. Mycologia 97:57–69, doi:10.3852/mycologia.97.1.57 Batista AC, Vital AF. 1956. Hughesiella—novo e curioso gênero de fungos dematiaceae. Anais Soc Biol Pernamb 14:141–144. Bliss DE. 1941. A new species of Ceratostomella on the date palm. Mycologia 33:468–482, doi:10.2307/3754617 782 MYCOLOGIA Dade HA. 1928. Ceratostomella paradoxa, the perfect stage of Thielaviopsis paradoxa (de Seynes) von Höhnel. Trans Br Mycol Soc 13:184–194, plates X–XII, doi:10.1016/ S0007-1536(28)80017-9 Davidson RW. 1935. Fungi causing stain in logs and lumber in the southern states, including five new species. J Agric Res 50:789–807. de Beer ZW, Seifert KA, Wingfield MJ. 2013a. The ophiostomatoid fungi: their dual position in the Sordariomycetes. In: Seifert KA, de Beer ZW, Wingfield MJ, eds. The ophiostomatoid fungi: expanding frontiers. Utrecht, the Netherlands: CBS Biodiversity Series. p 1–19. ———, ———, ———. 2013b. A nomenclature for ophiostomatoid genera and species in the Ophiostomatales and Microascales. In: Seifert KA, de Beer ZW, Wingfield MJ, eds. The ophiostomatoid fungi: expanding frontiers. Utrecht, the Netherlands: CBS Biodiversity Series. p 245–322. de Hoog GS. 1974. The genera Blastobotrys, Sporothrix, Calcarisporium and Calcarisporiella gen. nov. Stud Mycol 7:1–84. Delacroix G. 1893. Espèces nouvelles observées au Laboratoire de Pathologie Végétale. Bull Soc Mycol Fr 9:184– 188. de Seynes J. 1886. Recherches pour servir à l’histoire naturelle des végétaux inférieurs III. Paris: Masson. ———. 1888. La moisissure de l’ananas. Bull Soc Mycol Fr 4:26–30. Edgar RC. 2004. MUSCLE: a multiple sequence alignment method with reduced time and space complexity. BMC Bioinformatics 5:113, doi:10.1186/1471-2105-5-113 El-Ani AS, Klotz LJ, Wilbur WD. 1957. Heterothallism, heterokaryosis and inheritance of brown perithecia in Ceratostomella radicicola. Mycologia 49:181–187, doi:10.2307/3755628 Ferdinandsen CCF, Winge O. 1910. Fungi from Prof. Warmings expedition to Venezuela and the West Indies. Bot Tidsskr 30:208–222. Garofalo JF, McMillan RT. 2004. Thielaviopsis diseases of palms. Proc Fla Hortic Soc 117:324–325. Girard JC, Rott P. 2004. Pineapple disease. In: Rott P, Bailey RA, Comstock JC, Croft BJ, Saumtally AS, eds. A guide to sugarcane diseases. Montpellier, France: Librairie du Cirad. p 131–135. Glass NL, Donaldson GC. 1995. Development of primer sets designed for use with PCR to amplify conserved genes from filamentous Ascomycetes. Appl Environ Microbiol 614:1323–1330. Guindon S, Dufayard JF, Lefort V, Anisimova M, Hordijk W, Gascuel O. 2010. New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0. Syst Biol 59:307–21, doi:10.1093/sysbio/syq010 Harrington TC. 2009. The genus Ceratocystis. Where does the oak wilt fungus fit? In: Appel DN, Billings RF, eds. Proc 2nd National Oak Wilt Sym. Austin, Texas: USDA Forest Service, Forest Health Protection. p 27–42. ———, Steimel JP, Wingfield MJ, Kile GA. 1996. Isozyme variation and species delimitation in the Ceratocystis coerulescens complex. Mycologia 88:104–113, doi:10. 2307/3760789 ———, Wingfield MJ. 1998. The Ceratocystis species on conifers. Can J Bot 76:1446–1457, doi:10.1139/b98-145 Hawksworth DL. 2011. A new dawn for the naming of fungi: impacts of decisions made in Melbourne in Jul 2011 on the future publication and regulation of fungal names. MycoKeys 1:7–20, doi:10.3897/mycokeys.1.2062 ———, Crous PW, Redhead SA, Reynolds DR, Samson RA, Seifert KA, Taylor JW, Wingfield MJ, Abaci Ö, Aime C, Asan A, Bai F-Y, de Beer ZW, Begerow D, Berikten D, Boekhout T, Buchanan PK, Burgess T, Buzina W, Cai L, Cannon PF, Crane JL, Damm U, Daniel H-M, van Diepeningen AD, Druzhinina I, Dyer PS, Eberhardt U, Fell JW, Frisvad JC, Geiser DM, Geml J, Glienke C, Gräfenhan T, Groenewald JZ, Groenewald M, de Gruyter J, Guého-Kellermann E, Guo L-D, Hibbett DS, Hong S-B, de Hoog GS, Houbraken J, Huhndorf SM, Hyde KD, Ismail A, Johnston PR, Kadaifciler DG, Kirk PM, Kõljalg U, Kurtzman CP, Lagneau PE, Lévesque CA, Liu X, Lombard L, Meyer W, Miller AN, Minter DW, Najafzadeh NJ, Norvell L, Ozerskaya SM, Öziç R, Pennycook SR, Peterson SW, Pettersson OV, Quaedvlieg W, Robert VA, Ruibal C, Schnürer J, Schroers HJ, Shivas R, Slippers B, Spierenburg H, Takashima M, Taşkın E, Thines M, Thrane U, Uztan AH, Van Raak M, Varga J, Vasco A, Verkley GJM, Videira SIR, De Vries RP, Weir BS, Yilmaz N, Yurkov A, Zhang N. 2011. The Amsterdam declaration on fungal nomenclature. IMA Fungus 2:105–112, doi:10.5598/ imafungus.2011.02.01.14 ———, McNeill J, de Beer ZW, Wingfield MJ. 2013. Names of fungal species with the same epithet applied to different morphs: how to treat them. IMA Fungus 4:53– 56, doi:10.5598/imafungus.2013.04.01.06 Heath RN, Wingfield MJ, Wingfield BD, Meke G, Mbaga A, Roux J. 2009. Ceratocystis species on Acacia mearnsii and Eucalyptus spp. in eastern and southern Africa including six new species. Fungal Divers 34:41–68. Hennebert GL. 1967. Chalaropsis punctulata, a new hyphomycete. Anton Leeuw 33:333–340, doi:10.1007/ BF02045578 Höhnel FXR. 1904. Zur kentniss einige fadenpilze. Hedwigia 43:295–299. ———. 1909. Fragmente zur mykologie VI. Mitteilung, Nr. 182 bis 288. Sitzungsberichte der kaiserlichen akademie der wissenschaften in Wien. Math-Nat Klass 118:168. Hunt J. 1956. Taxonomy of the genus Ceratocystis. Lloydia 19:1–58. Jacobs K, Bergdahl DR, Wingfield MJ, Halik S, Seifert KA, Bright DE, Wingfield BD. 2004. Leptographium wingfieldii introduced into North America and found associated with exotic Tomicus piniperda and native bark beetles. Mycol Res 108:411–418, doi:10.1017/ S0953756204009748 Johnson JA, Harrington TC, Engelbrecht CJB. 2005. Phylogeny and taxonomy of the North American clade of the Ceratocystis fimbriata complex. Mycologia 97: 1067–1092, doi:10.3852/mycologia.97.5.1067 MBENOUN ET AL.: THE CERATOCYSTIS PARADOXA Kamgan Nkuekam G, Jacobs K, de Beer ZW, Wingfield MJ, Roux J. 2008. Ceratocystis and Ophiostoma species including three new taxa, associated with wounds on native South African trees. Fungal Divers 29:37–59. Kamgan GN, Wingfield MJ, Roux J. 2013. Ceratocystis species, including two new taxa from Eucalyptus trees in South Africa. Australas Pl Pathol 42:283–311, doi:10.1007/s13313-012-0192-9 Katoh K, Kuma K, Toh H, Miyata T. 2005. MAFFT 5: improvement in accuracy of multiple sequence alignment. Nucleic Acids Res 33:511–518, doi:10.1093/nar/ gki198 Kile GA. 1993. Plant diseases caused by species of Ceratocystis sensu stricto and Chalara. In: Wingfield MJ, Seifert KA, Webber JF, eds. Ceratocystis and Ophiostoma: taxonomy, ecology and pathogenicity. St Paul, Minnesota: APS press. p 173–183. McNeill J, Barrie FR, Buck WR, Demoulin V, Greuter W, Hawksworth DL, Herendeen PS, Knapp S, Marhold K, Prado J, Prud’homme van Reine WF, Smith GF, Wiersema JH, Turland NJ. 2012. International code of nomenclature for algae, fungi and plants (Melbourne Code). Regnum Vegetabile 154 ARG Gantner Verlag KG. Melin E, Nannfeldt JA. 1934. Research into the blueing of ground woodpulp. Svenska Skogsvårdsföreningens Tidskrift 32:397–616. Mitchell RS. 1937. Stem end rot of bananas with special reference to the physiological relationships of Thielaviopsis paradoxa (de Seynes) von Höhn. J Coun Sci Ind Res Aust 10:123–130. Möller EM, Bahnweg G, Sandermann H, Geiger HH. 1992. A simple and efficient protocol for isolation of high molecular weight DNA from filamentous fungi, fruit bodies and infected plant tissues. Nucleic Acids Res 20: 6115–6116, doi:10.1093/nar/20.22.6115 Moreau C. 1952. Coexistence des formes Thielaviopsis et Graphium chez une souche de Ceratocystis major (van Beyma) nov. comb. Remarques sur les variations des Ceratocystis. Rev Mycol (Suppl. Col) 17:17–25. Morgan-Jones G. 1967. Ceratocystis paradoxa: CMI descriptions of pathogenic fungi and bacteria No. 143. Surrey, UK: Commonwealth Mycological Institute. Nag Raj TR, Kendrick WB. 1975. A monograph of Chalara and allied genera. Laurier W, Waterloo, Ontario: Waterloo Univ. Press. 200 p. Paulin-Mahady AE, Harrington TC, McNew D. 2002. Phylogenetic and taxonomic evaluation of Chalara, Chalaropsis and Thielaviopsis anamorphs associated with Ceratocystis. Mycologia 94:62–72, doi:10.2307/ 3761846 Petch T. 1910. Thielaviopsis paradoxa (de Seynes) v. Höhnel. Ann Royal Bot Gardens, Peradeniya 5:511–574. Peyronel B. 1918. Micromiceti di Val Germanasca. Nuovo G Bot Ital 25:405–464. Posada D. 2008. jModelTest: phylogenetic model averaging. Mol Biol Evol 25:1253–1256, doi:10.1093/molbev/ msn083 Rambaut A, Drummond AJ. 2007. Tracer 1.5. (Available from http://beast.bio.ed.ac.uk/Tracer) COMPLEX 783 Rayner RW. 1970. A mycological color chart. Kew, UK: British Mycological Society. 34 p. Réblová M, Gams W, Seifert KA. 2011. Monilochaetes and allied genera of the Glomerellales, and a reconsideration of families in the Microascales. Stud Mycol 68:163– 191, doi:10.3114/sim.2011.68.07 Riedl H. 1962. Ceratocystis musarum sp. n., die heuptfruchtform der Thielaviopsis–Art von bananenstielen. Sydowia 15:247–251. Ronquist F, Huelsenbeck JP. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19:1572–1574, doi:10.1093/bioinformatics/ btg180 Saccardo PA. 1892. Supplementum universale, pars II. Syll Fung 10:213–216. ———. 1899. Catenularia Grev. Syll Fung 14:1076. ———. 1913. Thielaviopsis Went. Syll Fung 22:1341. Seifert KA, de Beer ZW, Wingfield MJ. 2013. The ophiostomatoid fungi: expanding frontiers. Utrecht, the Netherlands: CBS Biodiversity Ser 12. 337 p. ———, Morgan-Jones G, Gams W, Kendrick B. 2011. The genera of Hyphomycetes. Utrecht, the Netherlands: CBS Biodiversity Ser 9. 997 p. Swofford DL. 2002. PAUP* 4: Phylogenetic analysis using parsimony (*and other methods). Sunderland, Massachusetts: Sinauer Associates. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S. 2011. MEGA 5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance and maximum parsimony methods. Mol Biol Evol 28:2731–2739, doi:10.1093/molbev/msr121 Upadhyay HP. 1981. A monograph of Ceratocystis and Ceratocystiopsis. Athens: Univ. Georgia Press. 176 p. van Wyk M, Al Adawi AO, Khan IA, Deadman ML, Al Jahwari AA, Wingfield BD, Ploetz R, Wingfield MJ. 2007. Ceratocystis manginecans sp. nov., causal agent of a destructive mango wilt disease in Oman and Pakistan. Fungal Divers 27:213–230. ———, ———, ———, ———, Wingfield MJ. 2006. Molecular phylogeny of the Ceratocystis moniliformis complex and description of C. tribiliformis sp. nov. Fungal Divers 21:181–201. ———, Wingfield BD, Al-Adawi AO, Rossetto CJ, Ito MF, Wingfield MJ. 2011. Two new Ceratocystis species associated with mango disease in Brazil. Mycotaxon 117:381– 404. doi: 10.5248/117.381, doi:10.5248/117.381 ———, ———, Clegg PA, Wingfield MJ. 2009. Ceratocystis larium sp. nov., a new species from Styrax benzoin wounds associated with incense harvesting in Indonesia. Persoonia 22:75–82, doi:10.3767/003158509X439076 von Arx JA. 1952. Ueber die Ascomycetengattungen Ceratostomella Sacc., Ophiostoma Syd. und Rostrella Zimmermann. Anton Leeuw 18:201–213, doi:10.1007/BF02538608 Vovlas N, Frisullo S, Santo MSN de A, Abrantes IM de O, Espirito Santo SN. 1994. Ceratocystis paradoxa and Helicotylenchus multicinctus associated with root systems of declining bananas in the Republica Democratica de Sao Tome e Principe. Nematol Mediterr 22:119– 121. 784 MYCOLOGIA Wakker JH, Went FAFC. 1898. De ziekten van het suikerriet op Java, de niet door dieren Veroorzak Worden. Leiden: Brill. 217 p. Went FAFC. 1893. De ananasziekte van het suikerriet. Mededeelingen Proefstation West-Java 5:1–8. White TJ, Bruns T, Lee S, Taylor J. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand DH, Sninsky JJ, White TJ, eds. PCR protocols: a guide to methods and applications. New York: Academic Press. p 315–322. Wingfield BD, van Wyk M, Roos H, Wingfield MJ. 2013. Ceratocystis: emerging evidence for discrete generic boundaries. In: Seifert KA, de Beer ZW, Wingfield MJ, eds. The ophiostomatoid fungi: expanding frontiers. Utrecht, the Netherlands: CBS Biodiversity Series. p 57– 64. Wingfield MJ, de Beer ZW, Slippers B, Wingfield BD, Groenewald JZ, Lombard L, Crous PW. 2012. One fungus, one name promotes progressive plant pathology. Mol Plant Pathol 13:604–613, doi:10.1111/j.1364-3703. 2011.00768.x ———, Seifert KA, Webber JF. 1993. Ceratocystis and Ophiostoma: taxonomy, ecology and pathogenicity. St Paul, Minnesota: APS Press. 304 p. Witthuhn RC, Wingfield BD, Wingfield M J, Harrington TC. 1999. PCR-based identification and phylogeny of species of Ceratocystis sensu stricto. Mycol Res 103: 743–749, doi:10.1017/S0953756298007679

Log In

Reconsidering species boundaries in the Ceratocystis paradoxa complex, including a new species from oil palm and cacao in Cameroon

Reconsidering species boundaries in the Ceratocystis paradoxa complex, including a new species from oil palm and cacao in Cameroon

Related Papers

RELATED PAPERS