Efficiency of early testing in Pinus sylvestris L. grown under two different spacings in growth chamber

An hypothesis in designing early tests was that mimicking the limiting factor occurring later in a rotation would improve the genetic correlations between early and late measurements. Seedlings belonging to 30 open pollinated families of Scots pine (Pinus sylvestris) were therefore grown for three growth periods at two spacings, wide and dense (9 and 36 plants per box with an area of 0.188 m 2 ), in the growth chamber to mimic the competition for light. The wide spacing was replicated in another experiment with the same parents, and the seedlings were grown for two growth periods. Eight growth chamber characters were jointly evaluated with 20-year-old tree growth characters in four field trials with the same parents as in the growth chamber trials. The field trials were located in central Sweden, and established with full-sib families. The additive genetic correlations between growth chamber characters and height in field tests were none or weak. The correspondence between the results from wide and dense spacing in the growth chamber was strong. Moderate genetic correlations between growth chamber characters estimated in two different experiments, established with seed collected in two different years, were probably due to differences in mating in the seed orchard. The additive genetic correlations for height between field test sites were in the range 0.26 to 0.99. The raised hypothesis that mimicking the competition for light by a dense spacing would improve the genetic correlation between early and late measurements was not supported for the tested range of growth chamber conditions. Therefore, this study does not give any support for use of early tests in operational forest tree breeding, although other experiments have resulted in high correlation between field and growth chamber experiments.