A COMPARISON OF STABILITY METHODS TO ASSESS ALFALFA POPULATIONS FOR PERFORMANCE FOLLOWING FREEZING STRESS

To develop cultivars with greater tolerance to environmental stresses, an accurate assessment and ranking of experimental strains for stress environments is essential. This study was conducted to compare the regression, genotype mean — CV grouping, and superiority measure approaches for rating 16 alfalfa (Medicago saliva L.) populations for performance following freezing stress. Field-grown plants were excavated in November, subjected to a range of freezing stresses (−8 to −16 °C), and regrowth yield measured. Two plant ages (seeding year and first production) over 2-yr were assessed. The three stability analyses investigated did not give the same ranking of regrowth performance among the 16 alfalfa populations. The superiority index and mean–CV grouping techniques were better in detecting populations that performed well over the range of freezing stress compared with the regression method. The superiority index provided a continuous ranking index of entries which would be useful for selection applications; however, it has a potential bias towards more responsive entries in low stress environments. For selection purposes, the superiority index should be combined with a regression analysis.Key words: Alfalfa, Medicago sativa L., freezing stress, genotype × environment interaction, stability assessment.