About Us | Help Videos | Contact Us | Subscriptions


Defining Sunflower Selection Strategies for a Highly Heterogeneous Target Population of Environments


This article in CS

  1. Vol. 46 No. 1, p. 136-144
    Received: Mar 1, 2005

    * Corresponding author(s): avega@waycom.com.ar
Request Permissions

  1. Abelardo J. de la Vega *a and
  2. Scott C. Chapmanb
  1. a Advanta Semillas S.A.I.C., Ruta Nac. 33 Km 636, CC 559, (2600) Venado Tuerto, Argentina
    b CSIRO Plant Industry, Queensland Bioscience Precinct, 306 Carmody Rd., St. Lucia, QLD 4067, Australia


Genotype × environment (G × E) interactions can be a major impediment to genetic progress in sunflower (Helianthus annuus L.) breeding for Argentina. Previous studies revealed that northern and central environments show repeatable differences in genotype discrimination, suggesting some G × E interactions could be accommodated by selecting for specific adaptation. In this study, a trial dataset of 10 hybrids grown over 46 environments was used to validate this megaenvironment definition, to determine the value of division of the sunflower region of Argentina, and to define optimal testing strategies to balance resources between subregions. Pattern analysis confirmed the northern and central megaenvironments. Subdivision of the target region and the testing resources increased the within-subregion genotype to G × E interaction ratios and did not decrease trial repeatabilities. The genetic correlation between target region and its subregions was 0.36. In contrast to studies for barley in Canada, the calculated ratios of correlated response in a subregion to indirect selection in the undivided target region relative to direct response in the subregion demonstrate that division of the sunflower region is 3× more effective than selecting for broad adaptation to the undivided target region. The unpredictable G × E interactions within subregions should be accommodated by selecting for broad adaptation. In the northern subregion, there is scope to redefine testing strategies by replacing years with locations with no cost in performance predictability. Testing resources can be balanced based on the market value of the two subregions.

  Please view the pdf by using the Full Text (PDF) link under 'View' to the left.

Copyright © 2006. Crop Science Society of AmericaCrop Science Society of America