Multi-environment investigations using GGE biplot and regression model for higher spike yield and essential oil content in lavender (Lavandula angustifolia)

Abstract

Lavender is a highly valuable medicinal and aromatic perennial shrub lacking superior and stable varieties for varied environments in the western Himalayas. In the present experiment, estimates of genotypic effects and their interaction with the environment under various conditions of Himachal Pradesh were investigated to identify superior and stable genotypes. For that, seven selections with a check were evaluated at four different locations in randomized block design with three replications during 2020 and 2021. The pooled analysis of variance showed that genotypic and environmental main effects were highly significant for all the traits under study while, G×E interaction was significant (p≤0.05) for plant height, spike length and spike yield. The highest mean performance for all the traits revealed that location-4 has favourable conditions for lavender cultivation, and LOH15141 followed by LOH15040 were the best performers over all the locations. These high-performing genotypes were clustered together with the check in a major cluster. The stability parameters of the Eberhart and Russell model showed that LOH15141 is stable and superior for both spike yield and essential oil content, further confirmed by GGE biplot analysis. GGE biplot analysis revealed that all the tested locations formed a single mega-environment. The strong positive yield contribution described through Pearson’s correlation coefficients and genotypic path diagram states that plant height, number of spikes and spike length were significant contributors to the spike yield and could be used as selection criteria to improve spike yield due to their high heritability and genetic advance.