Genotype by environment interaction for index traits that combine growth and wood density in loblolly pine

Abstract

Genotype×environment interactions of individual traits have been assessed in numerous experiments with forest trees. However, since breeding programs rarely aim at the improvement of a single trait, the impact of G×E on index or composite traits must also be assessed. In a study with 12-year-old loblolly pine families in the southeastern U.S., G×E variance was of relatively little importance compared to genetic variance for wood density but was of greater significance for several growth traits. An index that combined stem volume and wood density to improve dry weight but maintain wood density constant (restricted selection index) resulted in substantially greater G×E variance compared to either of the component traits. The interaction variance of an index trait is shown to be a function of the index coefficients and the G×E variances and covariances for its constituent traits. As a result, for some conditions it surpasses the magnitude of G×E variance for each component trait.