| Phenotypic plasticity, the effect of plastic traits on fitness, and the stability of genetic variance-covariance matrices were investigated in an artificial population of Arabidopsis thaliana in response to environmental stress. These questions were also investigated in a selection experiment, where plants were subjected to selection on reproductive fitness (fruit production) under flooded and non-flooded water regimes.;Forty-seven accessions of early flowering A. thaliana were grown in laboratory conditions and were subjected to environmental stress such as low and medium light intensity, flooded and non-flooded water regimes, and a factorial arrangement of these light and water stresses. In addition, soft selection was applied for three generations in order to investigate patterns of phenotypic plasticity and phenotypic integration expressed in response to these stressful environments. Data was analyzed using a mixed model for the analysis of variance, linear regression was used for calculating the effects of plastic traits on reproductive fitness, and principal component analysis and vector correlations were used to investigate the stability of genetic variance-covariance matrices.;Results indicated that there was little plasticity to the environmental stresses applied, but there was a high degree of genetic variation to these environmental factors among accessions. Plastic traits had varying effects on reproductive fitness, in general there was selection for the increase of trait values, except for the bolting time which was under negative selection. When environmental stresses were manipulated independently, genetic variance-covariance matrices tended to stable. In contrast, when the environmental stresses were manipulated in a factorial arrangement, these matrices were less stable. |
