Population structure and migration of the witches’ broom pathogen Moniliophthora perniciosa from cacao and cultivated and wild solanaceous hosts in southeastern Brazil

Moniliophthora perniciosa, causal agent of witches’ broom disease in cacao plantations in South America and the Caribbean Islands, has co-evolved with its host cacao, but the pathogen has also emerged in many solanaceous hosts in Brazil, including economically important food crops and wild species. This study was carried out to: (i) determine the existence of host subpopulations of M. perniciosa in Brazil; (ii) estimate gene and genotypic diversity of M. perniciosa host subpopulations infecting solanaceous hosts in southeastern Bahia and Minas Gerais states, Brazil; and (iii) estimate the amount and directionality of historical migration of M. perniciosa subpopulations. Up to 203 M. perniciosa isolates collected from solanaceous hosts with symptoms from Bahia and Minas Gerais states in Brazil and from Theobroma spp. (cacao) and Herrania spp. were characterized with 11 microsatellite markers. Factorial correspondence analyses, minimum-spanning network and Bayesian clustering revealed genetic clusters associated with their host of origin. Significant subpopulation differentiation was evident (ΦST = 0.30, P ≤ 0.05) among M. perniciosa host subpopulations. Most of the multilocus microsatellite genotypes (MLMGs) were host-specific, with few MLMGs shared among subpopulations. Pairwise comparisons among M. perniciosa host subpopulations were significant, except between jurubeba (Solanum paniculatum) and cultivated solanaceous subpopulations. The combined analyses rejected the null hypothesis that M. perniciosa in Brazil is a single genetic population not structured by host. These findings support a scenario of introduction and subsequent adaptation to solanaceous hosts that should be taken into consideration to improve mitigation and management of M. perniciosa.