Genotype×environment interactions and some considerations of their implications for wheat breeding in Australia This review is one of a series commissioned by the Advisory Committee of the Journal.

Genotype×environment (G×E) interactions complicate selection forbroad adaptation, while their nature and causes need to be understood toutilise and exploit them in selection for specific adaptation. This invitedreview combines an assessment of the literature with the experience we havegained from involvement in wheat breeding and associated research programs toassess (1) the implications of G×E interactions for wheat breeding inAustralia, (2) the impact that research into G E interactions has had onbreeding strategy, and (3) the evidence for impact from this research efforton genetic improvement of crop adaptation. The role of analytical methodologyin this process is considered and some important issues are discussed.There are sufficient examples drawn from wheat breeding in Australia tosuggest that progress in dealing with G×E interactions can be made andseveral of these are presented. They show that impact in plant breedingfollows from achieving an appropriate level of understanding of theenvironmental and genetic factors causing the interactions as well as anassessment of their importance in the target genotype-environment system. Anaccurate definition of the environmental factor(s) contributing to theG×E interactions has been particularly important in determining therelevance of observed differences in plant adaptation to the target populationof environments. From the combination of biological and statistical studies, amore comprehensive understanding of G×E interactions has emerged andcontributed to new concepts and procedures for dealing with them.Distinguishing between what are repeatable and non-repeatable interactions isa key step. Genuine cases of positive specific adaptation observed inmulti-environment trials (METs) can be exploited by appropriately targetedselection strategies, while non-repeatable interactions are accommodated byselection for broad adaptation.The investigation of G×E interactions for grain yield of wheat inAustralia has matured to the point where an understanding of some of theircauses has enabled wheat breeders to exploit positive components of specificadaptation. The experience that has been gained in achieving these advancesindicates the importance of establishing a MET system that is relevant to thetarget population of environments of the breeding program. The investment ofadequate resources into effective design, conduct, analysis, andinterpretation of METs remains critical to continued progress from selectionin complex genotype-environment systems that present large G× Einteractions. Wheat breeders who understand their genetic material and thetarget population of environments can then use the generated information baseto achieve impact from their breeding programs.