Towards delivery of suitably high yielding, stable, and rust resistant wheat genotypes in the stem rust hotspots of Kenya
Emergence of Pgt race Ug99 and rapid proliferation of lineal highly virulent races imminently threaten Kenyan wheat. Devastating epidemics have led to huge losses among smallholder farmers who invariably are unable to spray appropriately and in situations where susceptible varieties are grown. To combat stem rust, the Kenya wheat improvement program seeks to release high yielding stable genotypes with suitable levels of disease resistance. Moreover, detection of genotypes that are adapted to rain-fed environments is an overarching objective. Six hundred and seventeen genotypes from various CIMMYT nurseries (PCBW, EPCBW, PCHPLUS, and 9th SRRSN) were selected based on plant type and reaction to stem rust at Njoro. The reconstituted nursery-KSRON, was sown in the main season of 2016 at Njoro and Timau for further evaluation. Forty red grained lines depicting R-MR infection types, severity of 30% or less, and average Thousand Kernel Weight of >40g were then selected to constitute a yield trial. At each of eight diverse environments, trials also comprising four commercial varieties as checks, were designed in RCBD, three replicates laid out in contiguous array of 8 rows x 10 m plots. Genotype (G), Environment (E) and GE interactions effects were estimated by fitting the AMMI model to yield data, supported by a biplot visualization of the results. Analysis revealed significant (P ?0.01) genotype (G), environment (E), and GE interactions. The first three principal components (PC) explained ~78% of the observed variation. Environment was the predominant source contributing over 85% to total sum of squares. The biplot pointed to at least four environments that were highly correlated. By classifying genotypes based on Shukla's stability variance and Kang's stability rating, six genotypes (R1402, R1411, R1424, R1481, R1484, and R1486) were deemed high yielding and stable, and thus suitable candidates for further testing through the release pipeline.