Abstract:
The major objectives of this study were: (1) to determine the extent to which southwest Oregon populations of Douglas-fir [Pseudotsuga menziesii var menziesii (Mirb.) Franco) differ in quantitative genetic structure (QGS); (2) to determine whether differences in QGS are associated with the environments from which populations originate; and (3) to examine the implications of differences in QCS with regards Co expected responses of traits to selection. A two year seedling common garden study was conducted where eight populations (45 families/population), two from each of four major coniferous zones, were sampled
Variation among populations in QGS was predominately associated with differences in trait means and genetic correlations among traits. The largest differences occurred between the coastal region and the high
elevation inland region. The major differences in QGS were associated with differences in correlations between growth and phenology traits in the second growing season. Differences in QGS were positively associated with the extent of habitat divergence for those comparisons between trait means and genetic correlations. The paired populations within an ecological zone were more similar, in general than populations from different ecological zones
Three major homogeneous associations, as measured by cluster analysis of both trait means and phenotypic correlation matrices, were found and are geographically aligned to the coastal region, a lower elevation inland
region, and a higher elevation inland region of southwest Oregon. Natural selection within these general physiographic regions would appear to be one explanation for the observed patterns in this study. When selecting for increased height growth in these four major zones, differences among zones in both direct and correlated responses
are expected. Expected correlated responses in phenology traits when selection is directed at height growth, while variable among zones, does not appear large enough to adversely affect adaptability to a large
degree. Restriction (0% change) of phenology and/or shoot:root ratios in a restricted selection index would severely limit growth response in two of the four zones. Restricted selection indices should be used only when necessary, and their effects should be assessed prior to implementation in an applied breeding program. Knowledge of QGS of popuitions should be ascertained, so that biological impacts of breeding and/or movement of reproductive materials outside of their native habitats can be assessed.
