Comparative Performance of Actual and Predicted Synthetic Varieties in Sweetclover 1

SELECTED clones or lines in forage crops must be utilized very largely in the production of synthetic varieties. Formulae for estimating the theoretical yields of advanced generations of synthetic varieties have been presented by Wright (8) and their application to forage breeding discussed by Tysdal, Kiesselbach and Westover (7). Estimated yields can be calculated from data on the average yields of the inbred lines or clones and their mean performance in Fx crosses on the assumption of random complete cross pollination. Several tests of combining ability have been used in forage crops including clonal, polycross, topcross, and openpollination progeny performance. But relatively few studies have been reported comparing the -results of these tests to the performance of synthetic varieties. Kinman and Sprague (5) in a study with maize emphasized that yields of synthetic varieties may be influenced to a greater degree by the general combining ability of inbred lines than by the number used in any specific combination. Jenkins (3) suggested the value of topcross performance in corn as a means of choosing lines in early generations of inbreeding for use in the production of synthetic varieties. Tysdal and Crandall (6) compared the yields of eight synthetic varieties with