A standard area diagram (SAD) set to aid assessment of the severity of coffee leaf rust ( Hemileia vastatrix ) was prepared and validated, and the characteristics of the rust lesions on Coffea arabica and C. canephora leaves were compared. The results indicated that the lesions were similar, so one scale could be used to evaluate the severity of rust on both species. The proposed SAD set contains illustrations of leaves with six disease severities (2·5, 5, 10, 20, 40 and 80%). The SAD set was validated by 10 raters with no previous experience of disease evaluation. The severity of rust was first estimated without using the SADs, and the same raters assessed the leaves again using the SADs. Regression analysis and Lin’s concordance correlation ( ρ _c ) analysis of estimated against actual disease severity showed precision and accuracy was significantly better using the SADs, for most raters, with improvements in both the slope and intercept of the regressions, and the values of u , υ (components of accuracy and precision of the ρ _c ) and ρ _c improved to 60, 70 and 90% of raters, respectively, when the SADs were used. The absolute error was ≤38% without the SADs, and ≤22% with them. Severity estimates were more reliable using the SADs ( R ² = 0·70–1·00 unaided, and R ² = 0·80–1·00 using the SADs). The SAD set improved rater assessments for the estimation of arabica and conilon rust severity, and can be used for assessing rust severity for many purposes, including plant breeding, fungicide screening and pathotype characterization.

In this study, a survey of physiological races of Hemileia vastatrix was carried out on 34 coffee leaf rust populations collected at several coffee genotyp

Genomic selection (GS) emphasizes the simultaneous prediction of the genetic effects of thousands of scattered markers over the genome. Several statistical methodologies have been used in GS for the prediction of genetic merit. In general, such methodologies require certain assumptions about the data, such as the normality of the distribution of phenotypic values. To circumvent the non-normality of phenotypic values, the literature suggests the use of Bayesian Generalized Linear Regression (GBLASSO). Another alternative is the models based on machine learning, represented by methodologies such as Artificial Neural Networks (ANN), Decision Trees (DT) and related possible refinements such as Bagging, Random Forest and Boosting. This study aimed to use DT and its refinements for predicting resistance to orange rust in Arabica coffee. Additionally, DT and its refinements were used to identify the importance of markers related to the characteristic of interest. The results were compared with those from GBLASSO and ANN. Data on coffee rust resistance of 245 Arabica coffee plants genotyped for 137 markers were used. The DT refinements presented equal or inferior values of Apparent Error Rate compared to those obtained by DT, GBLASSO, and ANN. Moreover, DT refinements were able to identify important markers for the characteristic of interest. Out of 14 of the most important markers analyzed in each methodology, 9.3 markers on average were in regions of quantitative trait loci (QTLs) related to resistance to disease listed in the literature.

Coffee leaf rust is the most limiting disease for coffee cultivation in Brazil. Despite its importance, relatively little is known about the genetic diversity of Hemileia vastatrix , the rust causal agent. In this work, the DNA from 112 monopustule isolates from different geographic locations and coffee genotypes were analysed by amplified fragment length polymorphisms ( AFLP ). The objectives were to assess the influence of the host and geographic origin on the diversity and population differentiation in H .  vastatrix . The fungal population showed a low level of genotypic diversity. Gene diversity ( h ) was 0·027 and the hypothesis of random mating in the total population was rejected, but evidence for recombination was found for two subpopulations (São Paulo and Paraná). The analysis of molecular variance revealed that 90% of the genetic distribution of the pathogen occurs among isolates within the subpopulation (states or host of origin). There was no correlation between geographic and genetic distance ( r  = −0·024, P  =   0·74), which together with the high number of migrants and the low degree of differentiation in populations of H . vastatrix , is consistent with the fact that the inoculum is probably easily dispersed by wind over long distances, allowing dispersal of the pathogen among coffee growing areas in Brazil. Therefore, it is difficult to predict the durability of resistant sources to coffee rust. The recommendation for the breeding programmes is thus to incorporate multigenic resistance as a control strategy.