Estudante: Diogo Felipe Milanesi. Data: 20/04/2021, às 16:00 horas pelo ZOOM. Orientador: Sérgio Hermínio Brommonschenkel.

Cotton is among the most important fiber crops in the world. Every year an average of 35 million hectares of cotton is planted across the planet. Brazil is fourth largest cotton producing country in the world and the second largest exporter, behind the United States. More than 99% of Brazilian cotton production is located at the Cerrado biome (MT and BA) where the acreage is extensive and dominated by a few cultivars. This large monoculture and the wet and warm environment of the region favors epidemics of fungal diseases such a Ramularia leaf spot (RLS), which is, nowadays, the most important cotton disease in Brazil. RLS causes premature defoliation with a consequent reduction in the yield potential; average productivity losses up to 30% have been reported. Until 2016, Ramulariopsis gossypii (syn. Ramularia areola) was the fungal species associated with RLS. In that year, it was reported a second species (R. pseudoglycines, sp. nov.) in cotton samples with RLS symptoms from Brazil. It was unknown, at the beginning of this present work, the distribution of these two species in Brazil and the importance of each species in RLS epidemics. Resistant varieties and chemical control are the most important strategies for RLS management. However, the increasingly early incidence of the disease and the often unsatisfactory response from chemical control suggests the occurrence of resistance of the fungus to the active principles used to manage RLS. The variable response of resistant genotypes and varieties between regions or states also indicates the existence of physiological variability in the pathogen population or prevalence of different species in different locations. In this work we addressed these unanswered questions using a combination of classical and cutting-edge genomic tools. Phylogenetic analyses of ITS and RPB2 regions from 110 Ramulariopsis sp. isolates collected at main cotton growing regions revealed the occurrence of R. pseudoglycines on 104 samples, and R. gossypii in 6. R. gossypii was only found in unsprayed breeding plots or isolated cotton plants in areas without large cotton plantations. The genome of two R. pseudoglycines and one R. gossypii isolates were sequenced and completely assembled and annotated. R. gossypii showed the largest genome as a result of expansion of transposable elements (TEs). Whole genome resequencing (WGS) of 71 isolates and their comparison to the reference genomes indicates that R. pseudoglycines populations sampled had a clonal structure. By assaying the virulence of isolates of R. pseudoglycines on susceptible and resistant genotypes it was possible to identify different races of this pathogen and to associate gains of virulence with mutations in candidate avirulence genes (designated RpAVR1 and RpAVR2). Intra-species polymorphisms (SNPs and deletions), as well as inter-species intrinsic differences (absence of RpAVR1 in R. gossypii) were related to virulence to the genotypes FMT705 and FMT07-248220 that carry different resistance genes. Differential sensitivity of both species to single-site fungicides were also associated with genomic changes. A combination of SNPs and expansion of the number of copies of the CYP51 gene (CNV) were associated with greater resistance to triazoles only in R. pseudoglycines. The G143A mutation associated with strobilurin resistance was also only observed on R. pseudoglycines, which could be related to differences in the intron position in the Cytb gene. Mutations associated with benzimidazole resistance were observed on isolates from both species. Fungicide resistance seems to drive the current predominance of R. pseudoglycines on large cultivation areas. Altogether, our findings will highlight the importance of the chemical control to maintain the durability of the main sources of RLS resistance used by breeding programs in Brazil.