Early Generation Selection for Grain Yield in Narrow-Leaf Lupin (Lupinus angustifolius L.)

Experiments were undertaken to identify an efficient method of selection for yield in early segregating generations of narrow-leaf lupin crosses. A first series of experiments involved simulation of F₂ populations with mixtures of ten cultivars and advanced breeding lines. The performance of genotypes in mixtures sown at different densities was compared with their monoculture [commercial density) yields m the same and the following year. These comparisons provided a basis for assessing the value of different planting designs and different selection criteria in early generation selection for yield. Correlated monoculture yield responses to simulated selection for single plant yield were similar for the honeycomb design and standard selection procedures at both low and commercial densities. These responses were much higher than that for grid selection in a low density population. Correlated monoculture yield responses to selection for other plant characters were substantial only for flowering time and number of pod bearing branches under certain conditions. These responses were greatest for selection in a low density mixture and in the year following selection. A procedure involving selection of the earliest flowering plants and selection within this group of plants with the highest number of pod bearing branches should be an effective approach to early generation yield selection in the narrow-leaf lupin.