Abstract
Predicted genetic gains of polymix, cloned polymix, full-sib and cloned full-sib testing and selection options for a breeding population were estimated for Pinus taeda L. Heritabilities for volume from the clonal testing were considerably greater than the heritabilities from the seedling testing. Cloned options had higher expected genetic gains than seedling options, even after adjusting for their longer breeding cycles and higher cost of testing. This was mainly due to higher genetic gains from within-family selection of the cloned testing options. Adjusted genetic gain from within-family selection of polymix and cloned polymix were 0.27 and 0.36% per year, respectively. Increasing the number of ramets per clone did not increase the genetic gain, but gain was sensitive to the clone selection ratio within families. When the number of trees tested per family was fixed at 90 trees, all clone/ramet combinations gave greater within-family gains than the seedling option. A complementary breeding strategy consisting of polymix breeding for estimation of general combining ability and clonal selection from clonally replicated full-sib families appears to be more efficient for loblolly pine improvement than the currently used seedling-based testing system. For. Sci. 50(1):44–53.