Genotype ✕ Environment Interactions in Burley Tobacco Variety Tests1
- C. L. Gupton,
- P. D. Legg,
- L. A. Link and
- H. F. Ross2
The objectives of our study were to estimate the components of variance for burley tobacco (Nicotiana tabacure L.) grown at six locations for 3 years and to determine their implications on testing procedures.
Eleven commercial hurley tobacco cultivars were grown at 6 locations in 4 states for 3 years. For each experiment we recorded the number of days to flowering; number of leaves and plant height after topping; length and width of fifth leaf from the top; internode length; yield of cured leaf; percent total nitrogen; percent α-amino nitrogen; water-soluble acids; pH; and percent ash, K2O, nicotine, and secondary amine alkaloids. Estimates of variance components and correlations among years and among locations were computed for each variable.
Estimates of variance components for cultivars were highly significant for all traits. Cultivar ✕ location and cultivar ✕ year interactions were relatively small for all traits. Second-order interactions were large relative to first-order interactions for most variables, but in most instances were much smaller than the component for cultivars.
The cultivars source of variation for yield was subdivided to elucidate the variance among F1 hybrids and pure lines and their interactions with environments. The location ✕ hybrid mean square was statistically significant, but the one for location ✕ pure lines was not. These results suggest that some hybrids may not have as much internal buffering capacity as pure lines in burley tobacco.
Our results show that years and locations may be considered random environments and may be used interchangeably. Adding more years costs more in time and money than does adding locations; therefore, using several locations for 2 years seems best for testing burley tobacco.Please view the pdf by using the Full Text (PDF) link under 'View' to the left.
Copyright © . .