Ethiopia

Septoria tritici blotch (STB) is a devastating fungal disease affecting durum and bread wheat cultivation worldwide. The search for resistance sources in untapped genetic resources may speed up breeding for STB resistance. Ethiopian durum wheat landraces represent a valuable source of allelic diversity for several traits, including disease resistance. In this study, we measure STB phenotypes under natural infection on two interconnected populations: i) a diversity panel comprising 318 Ethiopian durum wheat lines, mostly farmer varieties, and ii) a nested association mapping (NAM) population developed from a subset of the diversity panel. Phenology, yield and yield component traits were concurrently measured in the populations. We evaluated the distribution of STB resistance in Ethiopian genetic materials and the relationship existing between STB resistance and agronomic traits. STB resistance sources were found in landraces as well as in NAM lines. The genetic material was genotyped with more than 13 thousand genome-wide SNP markers to describe the linkage disequilibrium and genetic structure existing within the panels. The genotyping information was combined with phenotypes to identify marker-trait associations and loci involved in STB resistance. We identified several loci, each explaining up to 10% of the phenotypic variance for disease resistance. We developed KASP markers tagging the most interesting loci to allow the uptake of our results in a breeding perspective. Our results showed that the Ethiopian untapped allelic diversity bears a great value for studying the molecular basis of STB resistance and for breeding for resistance in local and international material.

Information about changes associated with advances in crop breeding is essential for understanding yield-limiting factors and developing new strategies for future breeding programmes. Thirty-six durum wheat varieties released since 1966 were evaluated in three replications of the Randomized Complete Design at Debre Zeit and Akaki, Ethiopia during the 2016 cropping season to estimate the amount of genetic gain made over time in grain yield potential, yield-associated traits and in protein content. Analysis of variance revealed significant differences among varieties for all 16 quantitative traits, protein content and protein harvest in Kg ha-1 at each of the locations. Grain yield varied between 1.66t ha-1 for Arendato released in 1966 to 3.90t ha-1 for Megenagna released in 2012 with mean of 2.952t ha-1 at Debre Zeit. At Akaki yield range was between 2.45 and 5.04t ha-1 with mean of 3.992t ha-1. 25 varieties surpassed Arendato (3.754t ha-1) at this location. In the combined ANOVA significant difference between the varieties was observed only for spike length, spikelets spike-1, grains spikelet, grains spike-1, plant height, days to flowering, thousand grain weight and hectoliter weight. Varieties specifically adapted to only one of the locations, widely adapted varieties and varieties not adapted to any of the locations were identified. Regression analysis revealed that grain yield has increased by 22kg ha-1 year-1 since 1966; an increase of 40.6% over yield in 1966. This was accompanied with a significant decline of 11.4% in spike length, 6.7% in spikelets spike-1, 17.9% in protein content and 31.2% in protein yield ha-1 and a significant increase of 41.1% in grains spikelet-1, 32.9% in number of grains spike-1, 22.3% in thousand grain weight, 17.8% in grain filling period, 23.9% in seed growth rate, 40.1% in grain yield production rate, 7.9% in harvest index.

Wheat stem rust is one of the major wheat yield limiting factors in Ethiopia. A stem rust epidemic occurred in the wheat belts of Arsi and Bale zones in the 2013-2014 crop season caused by Pgt race TKTTF that is virulent to the widely grown Ug99-resistant variety Digelu. This epidemic highlighted the need for wheat varieties with resistance to multiple Pgt races. This study was therefore, carried out to evaluate the reaction of the major Ethiopian varieties and advanced breeding lines against the dominant Pgt races in Ethiopia. Races TKTTF, TTKSK, TRTTF and JRCQC were isolated from field samples and multiplied on the susceptible cultivar McNair starting in May 2014. Four wheat stem rust nurseries, each inoculated with a single Pgt race, were established at Kulumsa and monitored from July through October, 2014. Each nursery included 34 entries in two replicates and 137 entries in a single replicate, augmented with six sets of five repeating checks. An additional nursery established at Debre Zeit, containing 551 entries in an augmented design, was evaluated with the epidemic Pgt race TKTTF. These entries included the most relevant Ethiopian bread and durum wheat breeding lines and cultivars, and 34 seedling-susceptible lines to evaluate the race-specificity of adult plant resistance. Stem rust severities for the four races ranged from trace to 80 %. Out of all entries evaluated, 10 were resistant to all four Pgt races, while 11 entries were effective to three of the four races. At Debre Zeit, 31.4% of the entries were resistant to Pgt race TKTTF. This study showed that rapid isolation and increase of Pgt races in Ethiopia is possible to facilitate field screening of breeding lines to select for candidate cultivars with resistance to multiple virulent races of Pgt.

Stem rust (SR) resistance is required for CIMMYT durum germplasm to keep relevance in Ethiopia, where Ug99 and other Pgt races are a major yield-limiting constraint, and in countries along the possible dissemination paths of these races. Resistance to Ug99 is widespread in most durum germplasm groups when tested in Kenya, but resistance is lost when exposed to Ethiopian races; hence selection at the Debre Zeit site in Ethiopia is essential for durum wheat. Due to difficulties with shuttling segregating populations between Mexico and Ethiopia, we have adopted a strategy involving the identification of resistant/moderately resistant lines at Debre- Zeit, and inter-crossing in Mexico followed by selection for resistance to leaf rust and agronomic type and finally screening for SR reaction in the resulting F6 lines at Debre-Zeit at the same time as they are tested for yield and quality in preliminary yield trials in Mexico. This has generated a significant increase in the proportion of resistant and moderately resistant genotypes within outgoing CIMMYT germplasm, from less than 3% at the onset of the initiative in 2008 to 16% in 2011, and 38% in 2013. SR-resistant germplasm was characterized by similar frequency distributions to other traits in the overall germplasm such as yield potential, drought tolerance and industrial quality parameters. Advances have also been realized using marker-assisted selection (MAS) to introgress Sr22 from bread wheat and to combine it with Sr25, producing advanced lines with 2-gene stacks with confirmed outstanding resistance and superior quality attributes. Since the two genes are closely linked but from different sources bringing them together required a very rare recombination event finally detected via MAS among thousands of plants. They are now essentially inherited together with a very low likelihood of generating recombinant individuals with either gene. The yield potential and stability of these lines are under evaluation in Ethiopia and the best lines are being used in a second round of breeding.