Genotype by environment interaction (G×E) refers to the comparative performances of genotypes differing among environments, representing differences in genotype rankings or differences in the level of expression of genetic differences among environments. G×E can reduce heritability and overall genetic gain, unless breeding programmes are structured to address different categories of environments. Understanding the impact of G×E, the role of environments in generating G×E and the problems and opportunities is vital to efficient breeding programme design and deployment of genetic material. We review the current main analytical methods for identifying G×E: factor analytic models, biplot analysis and reaction norm. We also review biological and statistical evidence of G×E for growth, form and wood properties in forest species of global economic importance, including some pines, eucalypts, Douglas-fir, spruces and some poplars. Among these species, high levels of G×E tend to be reported for growth traits, with low levels of G×E for form traits and wood properties. Finally, we discuss possible ways of exploiting G×E to maximise genetic gain in forest tree breeding. Characterising the role of environments in generating interactions is seen as the basic platform, allowing efficient testing of candidate genotypes. We discuss the importance of level-of-expression interaction, relative to rank-change interaction, as being greater than in many past reports, especially for deployment decisions. We examine the impacts of G×E on tree breeding, some environmental factors that cause G×E and the strategies for dealing with G×E in tree breeding, and the future role of genomics.

Pinus radiata is the subject of much the largest and longest-running tree breeding program in New Zealand. Very intensive plus-tree selection began in the early 1950s, the first seed orchard planting was in 1958, and the first orchard seed was collected in 1968. By 1985 seed orchards could meet the national planting needs, which helped prompt an innovative seed certification scheme. Selection began in local, ‘land-race’ stands. That proved appropriate, but better knowledge led to research on material from the species’ entire natural range and then to further seed importations from natural stands, mainly for future genetic resources. Initial plus-tree selection largely addressed poor tree form in the central North Island. This led to selecting a ‘short-internode’ (‘multinodal’) branching ideotype, with great improvements in general tree form and substantial gains in growth rate. A distinctive portfolio of several breeds, representing different breeding goals, has since been developed for the diverse planting sites and the various end products. By contrast, the original regionalisation of seed orchards proved largely unjustified. Despite early research indicating good heritabilities, wood properties have figured significantly in breeding goals only since growers substantially reduced harvest ages, but they are now intensively researched. The breeding program was greatly expanded from 1968 to meet the needs of multi-generation breeding. This was one of the program's contributions to developing breeding strategy as a discipline, various of which were reflected in the 1986 Development Plan. Industry uptake of tree breeding has since posed challenges in maintaining genetic variability for a secure, long-term underpinning of the program. Special genetic-gain trials were first planted in 1978, to demonstrate and quantify genetic improvement, and to help update forest out-turn predictions. The gains achieved in tree form helped induce New South Wales State to affiliate with the New Zealand program, following significant earlier contributions of breeding material to Australian breeding programs. Improved technologies for seed orchards and vegetative propagation have facilitated development of the breed portfolio and more specific customisation of seedlots. Becoming able to deliver genetic gain through controlled pollination has allowed changes in the structure of the breeding population. Further changes are resulting from an increasing capability to reconstruct pedigrees after open pollination. Developments towards full clonal forestry have faced obstacles, but offer greater genetic gains and a more precise customisation, and clonal forestry is now being vigorously commercialised. Having a single, large Forest Service until 1987 undoubtedly favoured the development of a very strong tree improvement program. Since then, continuing institutional changes, which are reviewed, have posed numerous challenges.

Clonal trials of Pinus radiata D. Don (radiata pine), representing two populations (or breeds), one selected for growth and form (GF) and the other selected for high wood density as well as growth and form (HD), were replicated on two low-altitude New Zealand sites: Tarawera (pumice soil, 38°08′S) and Woodhill (coastal dune, 36°42′S). The GF material comprised 33 pair-crosses (19 parents) × 10 clones, and the HD material comprised 19 single-pair crosses (35 parents) × 10 clones, with six ramets per clone per site. Diameter (DBH), two tree-form variables, and needle retention (NRA) were assessed 8 years after planting, and wood density (DEN), acoustic velocity, and collapse were assessed 9 years after planting. The site differences were generally expressed more strongly in the GF population. Estimated genetic parameters were mostly similar for the two breeds, except that genotypic correlation between DBH and DEN was apparently zero in the HD population. Estimated broad-sense heritabilities (H²) were generally markedly higher than narrow-sense heritability estimates (h²), except with DEN. Estimated between-site type-B clonal genotypic correlations were generally high (>0.8) for wood properties. Overall, DBH showed adverse genetic correlations with wood properties. The Elite/Breed strategy appeared to be helpful in combating adverse genetic correlations.

Favourable interactions between genotype, site and silviculture are desired for growing radiata pine, site and silviculture being two components of environment. Our challenge is to characterise the interactions so as to exploit them to advantage. Four classes of interaction are involved: three first-order interactions between pairs of factors; and one second-order interaction involving all three factors. Also of interest is to partition the interactions into two types: involving rank change of genotypes (RC interaction) and differing level of expression of genotypic differences among environments (LoE interaction). Tree breeders’ immediate concern is with main effects of environment and interaction effects on genotypes’ performance, yet prime interest for research and policy decisions lies in the role of environments and specific environmental factors in generating interactions. Of the two types of interaction, RC seems generally more important for breeding operations with LoE more so for deployment decisions. Existing information, and provisional expectations of interactions, is to be reviewed, mostly qualitatively, in terms of the comparative importance of RC and varying LoE. This will be done, for the first-order interactions, for various traits. Second-order interactions are as yet almost entirely conjectural. A framework is offered for accumulating both input from others and future research findings. In radiata pine, growth traits, especially stem diameter, tend to show marked RC interaction with site. Tree-form traits, however, seem mainly subject to LoE interaction, and disease resistance is intensely so. Among wood properties, some show very little interaction, while LoE interaction seems to occur for some other properties.