Predicting grain yield in Australian environments using data from CIMMYT international wheat performance trials 3. Testing predicted correlated response to selection

Abstract

The predictions from previous work evaluating the scope for exploiting indirect selection for grain yield of wheat of Australia using the results of CIMMYT international wheat trials were tested in a large sample of wheat production environments (a total of 44 collected over four years) in Queensland, Australia. The findings of this study supported the predictions of the earlier work which were based on a smaller sample of environments (a total of six collected over two years). Average line grain yield over the international environments showed a strong positive association (rp = 0.874, P < 0.01) with that over the Queensland environments for the 11 common lines. This indicates that Queensland wheat breeders can select for broad grain yield adaptation to their Australian production environments on the results of the international tests conducted by CIMMYT, with a high degree of confidence. There was a large component of variation associated with line by environment (L × E) interaction in both the Australian and international experiments. Subsets of the international environments were identified which could be used to select for specific yield adaptation to groups of the Australian environments identified for Queensland. It is proposed that the relatively large number of Australian environments (⩾ 8 environments) in two of the groups identified by classification is indicative that the aspects of L × E interaction associated with these groups may be repeatable and could therefore be selected for. The results of this study indicate that there is considerable overlap between the aspects of yield adaptation being selected for in the CIMMYT wheat breeding program and that required for Queensland, Australia. It is suggested that there is an urgent need for further investigation of the nature of this overlap for other regions in the Australian wheat belt and for other countries which utilise CIMMYT wheat germplasm in their national breeding programs. If similar relationships to those identified for Queensland exist elsewhere, other wheat breeders would be able to use indirect selection to speed up the acquisition of useful CIMMYT breeding lines. Characterising and quantifying the nature of the overlap in adaptation required for different wheat production regions around the world would greatly assist the transfer of wheat germplasm from the CIMMYT wheat breeding program to collaborating national programs.