A nonlinear model applied to genotype x environment interaction analysis in maize
Under the assumption that environmental effects act multiplicatively on genotype effects, a nonlinear model was studied, which is multiplicative for the main effects but additive for deviations. Experimental corn (Zea mays L) yield data of cultivars were analysed, in order to determine the applicability of the nonlinear model. The following consequences could be demonstrated: (1): in better environments in which higher mean values of the trait are observed, discrimination among genotypes or cultivars is favored; (2): in the usual joint analysis the interaction sum of squares depends on these due to the main effects; (3): the adjustment of treatment means of different experiments designed such that only some checks are common over the group of experiments, should be made through multiplicative factors rather than additive ones; (4): the linear regression coefficients, commonly used for the stability analysis of cultivars, becomes positively correlated with the means of the cultivars, over environments, to the analysis of corn yield data of 64 experiments, it could be verified that: (1) the nonlinear model showed equal or higher efficiency than the linear one; (2): the multiplicative action of environment on genotypic effects explained on the average, 11,1% of the cultivars x locations interaction, varying from 102% to 16.2%.
| Main Authors: | , , |
|---|---|
| Format: | Digital revista |
| Language: | por |
| Published: |
Pesquisa Agropecuaria Brasileira
2014
|
| Online Access: | https://seer.sct.embrapa.br/index.php/pab/article/view/14278 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|