HARVEST INDEX AS A SELECTION CRITERION FOR GRAIN YIELD IN TWO SPRING WHEAT CROSSES GROWN AT TWO POPULATION DENSITIES
The use of harvest index as a selection criterion for grain yield in F2 populations of spring wheat (Triticum aestivum L.) grown at two population densities was investigated. Harvest index was useful in delineating yield differences between lines for both crosses. The F4 lines selected in F2 for a high harvest index yielded about 9% more per plot in 1978 than F4 lines having a low harvest index in F2. Generally, lines selected at the higher commercial seeding rate yielded more than lines selected at the lower plant density. In 1979, a heavy Fusarium infection reduced the mean grain yield of the F6 lines and suppressed any significant response to selection resulting from population density and harvest index in F2. While selection based on high harvest index at low population density can be used to select higher yielding plants it was not as effective as selection at high population density which more closely approximates commercial crop densities. Additional research is needed before the use of harvest index as a selection tool in wheat breeding programs can be recommended for use in Atlantic Canada.