Genotype x region interaction for two-row barley yield in Canada

Barley (Hordeum vulgare L.) breeding programs recognize eastern and western Canada as separate target regions, but the extent of local adaptation to r egions and subregions within them has not been studied. Genotype x region a nd subregion interactions were estimated in 145 lines from the two-row barl ey cross Harrington/TR306 in 22 trials in 1992-1993, The trials were groupe d into five subregions (Maritimes-Quebec, Ontario, Manitoba-North Dakota, S askatchewan, and Alberta) and two regions (eastern Canada and western Canad a plus North Dakota). Variance components were estimated by a model in whic h the genotype x location (sigma(GL)(2)) variance was subdivided into a gen otype x region (or subregion) variance ((TGS)) and a within-region or -subr egion sigma(GL)(2). NO sigma(GS)(2) was observed within the eastern or west ern regions, and genotypic correlations across subregions within regions ap proached 1.0. Significant sigma(GS)(2) was observed for eastern versus west ern Canada, but the correlation between genotypic effects across these regi ons was 0.83, In a selection experiment, subdivision of the eastern or west ern regions did not increase response. Selection in the east produced great er yields in both the east and west. The same genotype ranked first for yie ld in both regions. There was little specific adaptation to subregions, and two-row barley genotypes were broadly adapted across northern North Americ a. Further subdivision of the regions is unwarranted, and selection in eith er region is likely to result in response in the other. The lack of local a daptation indicates that breeding programs that test broadly are likely to outperform ones that are narrowly targeted.