Species or genotypes? Reassessment of four recently described species of the Ceratocystis wilt pathogen, Ceratocystis fimbriata, on Mangifera indica

Ceratocystis wilt is among the most important diseases on mango (Mangifera indica) in Brazil, Oman, and Pakistan. The causal agent was originally identified in Brazil as Ceratocystis fimbriata, which is considered by some as a complex of many cryptic species, and four new species on mango trees were distinguished from C. fimbriata based on variation in internal transcribed spacer sequences. In the present study, phylogenetic analyses using DNA sequences of mating type genes, TEF-1α, and β-tubulin failed to identify lineages corresponding to the four new species names. Further, mating experiments found that the mango isolates representing the new species were interfertile with each other and a tester strain from sweet potato (Ipomoea batatas), on which the name C. fimbriata is based, and there was little morphological variation among the mango isolates. Microsatellite markers found substantial differentiation among mango isolates at the regional and population levels, but certain microsatellite genotypes were commonly found in multiple populations, suggesting that these genotypes had been disseminated in infected nursery stock. The most common microsatellite genotypes corresponded to the four recently named species (C. manginecans, C. acaciivora, C. mangicola, and C. mangivora), which are considered synonyms of C. fimbriata. This study points to the potential problems of naming new species based on introduced genotypes of a pathogen, the value of an understanding of natural variation within and among populations, and the importance of phenotype in delimiting species.