SELECTION FOR ANNUAL GROWTH CURVES IN NICOTIANA TABACUM L

G. Namkoong ¹ and D. F. Matzinger ¹

¹ Southeastern Forest Experiment Station, USDA Forest Service, Genetics Department, North Carolina State University, Raleigh, North Carolina 27607

The cumulative growth of a plant is the result of interrelated processes, and response to selection for changes in the annual growth curve requires many physiological adjustments. Selection to modify the entire annual growth curve may therefore not be as effective as linear models may predict. Periodic growth of a population of Nicotiana tabacum L. was estimated to have heritabilities increasing from 7% up to 31% for successive heights, with positive genetic and phenotypic correlations among all periods. Two selection experiments on this population indicate the difficulties of using simple index selection to raise the entire growth curve. A selection index of eight periodic heights resulted in a gain in all periods for the first cycle of selection but mixed losses and gains in subsequent cycles for a small net gain after four cycles of selection. A selection index of three parameters of a nonlinear height growth function resulted in a consistent change in the growth curves over the four cycles of selection but a net loss in early growth and a large net gain in late-season growth.