|Abstract or Summary
- Parents and progeny populations from a cross of 'Stephens'x 'Tibet dwarf' were grown in a spaced-planted experiment to obtain information concerning the nature of inheritance and possible associations between grain yield and the components of yield. Agronomic traits measured on an individual plant basis were: heading date, days to maturity, grain filling period, plant height, biological yield, number of spikes per plant, total length of spikes, spikelets per spike, kernels per spikelet, total kernel number, grain yield, kernel weight, harvest index, and chaff percentage. The nature of gene action was estimated for each trait by comparing broad and narrow sense heritability estimates. Possible associations among traits were determined by phenotypic and genotypic correlations as well as by path coefficient analysis. This latter analysis was based on the genotypic correlations of total grain yield with five other traits. Narrow sense heritability estimates were high for kernel weight, spikelets per spike, heading date, kernels per spikelet, chaff percentage, and total kernel number. Intermediate estimates were found for plant grain yield, spikes per plant, plant height, harvest index, days to maturity and total length of spikes. A low estimate was realized for biological yield. Narrow sense heritability values were higher than broad sense heritability for spikelets per spike, kernel weight, chaff percentage and heading date. These inflated estimates are attributed to sampling errors and/or uneven environmental influence on the various generations. Path coefficient analyses suggested that among the yield components, the number of kernels per spikelet was most closely associated with grain yield, followed by the number of spikelets per plant, the number of spikelets per spike and kernel weight. Low direct associations with grain yield were found for both harvest index and plant height. A moderately high association between harvest index and grain yield was the result of an indirect association via spikelets per spike. As suggested by both narrow sense heritability estimates and path coefficient analysis, selection for number of kernels per spikelet among the F₂ populations of the cross of Stephens and Tibet dwarf would result in increased grain yield in succeeding generations. Path coefficient analysis indicated that the direct relation between grain yield and plant height was low, suggesting that short, high yielding plants could be selected within segregating populations resulting from this cross.