Genotype by environment interaction in the southern Swedish breeding population of Picea abies using new climatic indices

Genotype-environment interaction (G x E) has a critical impact on the design of a breeding programme and the resulting genetic gains. The aim of this study was to assess the incidence and predictability of G x E with respect to growth traits in the breeding population of Picea abies (L.) Karst. in southern Sweden. We used across-site genetic correlations as a measure of G x E and multiple linear regression to detect patterns in the correlations. As independent variables in the regression analysis, we used site and geographic factors as well as a comprehensive set of new climatic variables. A main result was that G x E was low to moderate with respect to the growth traits considered and largely unpredictable based on any of the geographic or climatic variables analyzed. However, it seems that the effects of late spring frost damage are important as they created significant differences in across-site correlations and we suggest that models describing frost risks should be developed. One major implication from this study is that the same genetic material could be used over larger areas in southern Sweden than recommended today with only a marginal or no loss in genetic gain.