Abstract: China is the largest wood importer and the second largest wood consumer in the world, and its dependence on external supply has exceeded 50% for several years. However, the average annual growth of forest per hectare in China is about half of that in the developed countries in forestry, which indicates that there is still larger gap in tree breeding in China, compared with the developed countries in forestry. Therefore, strengthening the large-scale experiments and accurate genetic evaluation of forest trees has great significance in improving the productivity of China's plantation forests through the precise selection and breeding of tree varieties. Genotype by environment interaction is one of the important contents of large-scale experiments and accurate genetic evaluation of forest trees. Genotype by environment interaction (G×E) refers to a lack of consistency in the relative performance of genotypes among different environments, and represents differences in genotype rankings or differences in performance inconstant among environments. Existing studies have confirmed that G×E is very common and often large in forest trees, and it is usually difficult to find consistently superior genotypes with broad adaptation. Since G×E can reduce heritability and genetic gain, understanding the G×E effects and their environmental drivers is vital to mating design, species/variety selection and genotype deployment. The paper reviews the current main analytical method for identifying G×E(including factor analytic method and BLUP-GGE joint analysis) and estimating heritability, and compares the strength and weakness of these analytical method (including stability analysis, type-B genetic correlation, AMMI, GGE biplot, factor analytic method and BLUP-GGE joint analysis), and also reviews the progress of G×E studies on growth traits (such as diameter at breast height, height and volume), form traits (such as stem straightness, branch angle and branch size) and wood properties (such as wood density and modulus of elasticity) in forest species (such as Pinus elliottii, Pinus taeda, Picea abies, Eucalyptus grandis, Pinus radiata and Pseudotsuga menziesii) of global economic importance. Moreover, the paper discusses the environmental drivers that cause G×E and strategies for dealing with G×E in tree breeding. Finally, the future research of G×E is proposed, alongside development of new analytical method, focusing on multi-variate model of G×E and integration of genomic selection with G×E. New genetic analysis model for forest trees should be adopted into G×E studies. The patterns and magnitude of G×E should be focused on multi-variate model for multi-environment trials. Accurate estimation of environment-specific genomic breeding values of forest trees should be performed.

Key words: genotype by environment interaction, G×E drivers, genetic correlation, heritability, genetic gain