Yield Stability and Multi-Trait Evaluation of Groundnut (Arachis hypogaea L.) Cultivars Across Mediterranean-Type Environments in Türkiye

Abstract

Genotype × environment interaction (GEI) is a major constraint on reliable cultivar recommendation in groundnut (Arachis hypogaea L.), particularly across environments that contrast in temperature, rainfall, and crop duration. In this study, the magnitude and structure of GEI were characterized in six groundnut genotypes, comprising five cultivars registered before the trials and Ayşehanım (G3), an unregistered test genotype in 2016–2017 that was registered in 2018; genotypes combining high mean performance with stability were identified; and the cultivars closest to a multi-trait ideotype were examined, as a panel-specific decision aid, under the Mediterranean-type conditions of three Turkish production areas (Adana, Antalya, and Manisa). The cultivars were evaluated across six environments (three locations × two years, 2016–2017; each location × year combination treated as a macro-environment) in a randomized complete block design with four replications, and seven phenological, agronomic, and quality traits were recorded. Data were analyzed by combined analysis of variance and by AMMI, GGE biplot, WAASB/WAASBY, BLUP, and MGIDI; given heterogeneous error variances among environments, genotype BLUPs were obtained from a mixed model with environment-specific residual variances, and the homogeneous-variance model was retained as a sensitivity analysis. Environment and GEI were highly significant for all traits (p < 0.01). For seed yield, environment accounted for 64.1% of the variation, whereas GEI (22.4%) surpassed the genotype effect (13.5%); within the environmental term, the location × year interaction (28.8%) exceeded both location (19.6%) and year (15.7%), indicating location-specific year responses. WAASBY, BLUP, and GGE consistently identified Sultan (6483 kg ha⁻¹) and Halisbey (6369 kg ha⁻¹) as the highest-yielding cultivars with acceptable stability across the six tested environments, whereas the exploratory MGIDI analysis ranked NC-7 and Ayşehanım nearest the ideotype. The two criteria favored distinct cultivars, with higher seed yield and lower shelling percentage forming a descriptive pattern among the leading cultivars rather than an established relationship between the traits. These results indicate that GEI, rather than mean yield alone, governed cultivar performance in this trial network, and suggest a complementary framework for regional evaluations of this scale in which WAASBY/BLUP and GGE guide yield–stability assessment and MGIDI supports multi-trait screening. Cultivar recommendations are provisional and require validation under current climatic and management conditions.