Introduction: Camelina, as an oilseed crop, possesses unique agronomic traits and a high potential for cultivation under dryland conditions. In the context of climate change and limited water resources, identifying high-yielding and well-adapted genotypes suitable for dryland farming is of particular importance. Selecting superior genotypes through the evaluation of interrelationships among traits and simultaneous assessment based on a set of traits represents an efficient approach in breeding programs. Therefore, the objective of this study was to evaluate yield-related traits and identify superior camelina genotypes under dryland conditions by employing Pearson’s correlation matrix, cluster analysis based on heatmap visualization, and the factor analysis ideotype-design-best linear unbiased prediction (FAI-BLUP) index.
Materials and methods: In this study, 28 camelina genotypes were evaluated at the National Agricultural Research Station and Dryland Seed Production of Gonbad-e Kavus over two cropping seasons (2021–2023) using a randomized complete block design (RCBD) with three replications. Sowing was performed manually and at shallow depth, and weed control was also carried out manually. The recorded traits included days to flowering, days to ripening, plant height, branching height, number of sub-branches, number of pods on the main branch, number of pods on the sub-branch, number of pods per plant, number of seeds per pod, thousand-kernel weight, seed yield. Statistical analyses comprised combined analysis of variance and mean comparison using SAS software (version 9.4). In addition, Pearson’s correlation matrix, cluster analysis based on heatmap visualization, and calculation of the FAI-BLUP index were performed using RStudio.
Results: The combined analysis of variance revealed that the effects of year, genotype, and the genotype × year interaction were significant (p < 0.01) for all traits. Mean comparison indicated substantial differences among camelina genotypes in all studied traits and seed yield. The highest seed yield was recorded for genotypes 3, 17, 18, 23, 19, 25, and 20. Genotype 10 exhibited the greatest number of siliques per plant, while genotypes 3 and 10 had the highest and lowest thousand-seed weights, respectively. Correlation analysis demonstrated that seed yield was strongly and positively associated with the total number of siliques/pods per plant, as well as the number of pods on both the main branch and sub-branch. Cluster analysis based on heatmap visualization identified three major groups, with the first cluster (genotypes 3, 4, 10, 11, 14, 15, 17, 18, 19, and 26) showing high values across most measured traits. According to the FAI-BLUP index, genotypes 19, 23, 12, and 17 were identified as superior.
Conclusion: Correlation analysis indicated that the number of pods on the main branch and sub-branch, and thousand-kernel weight can serve as indirect selection indices for improving seed yield under dryland conditions. Cluster analysis based on heatmap visualization and the FAI-BLUP index identified genotypes 3, 10, 12, 17, 18, 19, 23, and 25 as superior, with high seed yields of 2819, 2475, 2487, 2740, 2627, 2569, 2586, and 2539 kg·ha⁻¹, respectively. Overall, these genotypes exhibited favorable performance across multiple agronomic traits and can serve as valuable genetic resources in breeding programs. |
