Relative family performance and variance structure of open-pollinated Douglas-fir seedlings grown in three competitive environments.

Open-pollinated Douglas-fir (Pseudotsuga menziesii var 'menziesii' (Mirb.) Franco) families were tested in three contrasting competitive environments to test the hypothesis that relative performance as measured by total seedling dry weight is dependent upon distance or genotype of neighbors. The three competitive environmnts included (1) a mixture of individuals from all families planted at close spacing, (2) single (pure) family blocks planted at close spacing, and (3) individuals from all families planted at a wide, non-competitive spacing. Despite occasional large changes in rank between competitive environments and only moderate correlations of family means between competitive environments, the family x competitive environment interaction was non-significant. Furthermore, families did not differ significantly in competitive ability or density tolerance. The competitive environment in which seedlings were grown, however, had a large effect on estimates of variance components, which in turn led to large differences in estimates of heritability and genetic gain. Evaluation of families in mixture resulted in the largest estimates of heritability, while evaluation in pure family blocks resulted in the lowest. Analysis of correlated response to selection indicated that testing and selection in mixture result in the largest estimated gain, even if progeny of selected individuals are subsequently grown in a pure or non-competitive environment.