|CLARK, JOHN - UNIV OF ARK
|CREECH, DAVID - STEPHEN F AUSTIN UNIV
|POWELL, ARLIE - AUBURN UNIV
|ROOKS, SUSAN - UNIV OF NC
Submitted to: American Society for Horticultural Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/21/1995
Publication Date: N/A
Interpretive Summary: The cultivar potential of blueberry selections is determined at several locations in the southern United States. It would be useful also to know how stable blueberry variety ripen date and yield remain from environment to environment. This study was conducted to determine which, if any, of the available techniques for estimating stability in different environments smight be useful in blueberry. Cluster analysis appeared to be most useful compared to other methods used in this study.
Technical Abstract: To determine if any of the available techniques for estimating stability in different environments is useful in blueberry (Vaccinium ashei Reade and V. corymbosum L.), 14 clones were evaluated in nine environments for ripening date and yield. Type 1 and 2 stability statistics, plots for each genotype mean vs. its coefficient of variability (CV) across environments (genotype grouping), environmental index regression, and cluster analyses were compared. The highest yielding rabbiteye and southern highbush clones across locations were not deemed stable by Type 1 and Type 2 stability statistics, genotype grouping, or environmental regression technique. No evidence of curvilinear response was found. The non-parametric cluster analysis with known cultivars included appears to be most useful compared to other methods of estimating stability used in this study.