Generating genomics-driven knowledge opens a way to accelerate the resistance breeding process by family or population mapping and genomic selection. Important prerequisites are large populations that are genomically analyzed by medium- to high-density marker arrays and extensive phenotyping across locations and years of the same populations. The latter is important to train a genomic model that is used to predict genomic estimated breeding values of phenotypically untested genotypes. After reviewing the specific features of quantitative resistances and the basic genomic techniques, the possibilities for genomics-assisted breeding are evaluated for six pathosystems with hemi-biotrophic fungi: Small-grain cereals/Fusarium head blight (FHB), wheat/Septoria tritici blotch (STB) and Septoria nodorum blotch (SNB), maize/Gibberella ear rot (GER) and Fusarium ear rot (FER), maize/Northern corn leaf blight (NCLB). Typically, all quantitative disease resistances are caused by hundreds of QTL scattered across the whole genome, but often available in hotspots as exemplified for NCLB resistance in maize. Because all crops are suffering from many diseases, multi-disease resistance (MDR) is an attractive aim that can be selected by specific MDR QTL. Finally, the integration of genomic data in the breeding process for introgression of genetic resources and for the improvement within elite materials is discussed.

Maize production is on risk by Gibberella ear rot (GER) caused by Fusarium graminearum. This is one of the most important ear rot diseases in temperate zones as it leads to yield losses and production of harmful mycotoxins. We investigated, for the first time, the potential use of Brazilian tropical maize to increase resistance levels to GER in temperate European flint germplasm by analyzing six interconnected biparental populations. We assessed GER symptoms in Brazil and in Europe in up to six environments (= location × year combinations) during the growing seasons of 2018 and 2019. We conducted multi-parent QTL and biparental QTL mapping, and identified four QTLs with additive gene action, each explaining 5.4 to 21.8% of the total genotypic variance for GER resistance. Among them, QTL q1 was stable across test environments, populations, and between inbred lines and testcrosses. The accuracies of genomic prediction ranged from 0.50 to 0.59 depending on the resistance donor and prediction model. Jointly, our study reveals the potential use of Brazilian resistance sources to increase GER resistance levels by genomics-assisted breeding.