Evaluation of variability in argan oil content through different environments and preselection of elite genotypes

This investigation was carried out to determine the variability in oil content with the aim to identify genotypes of argan tree expressing high oil yield. The 150 argan trees were collected from five provenances in south west of Morocco over 3 years (2008–2010) and were screened from their oil content using Soxhlet method based on the 840 samples. Univariate and multivariate analysis were used to study the genetic variation between and within provenances. According to the results on the mean of the 3 years, the oil content was ranged from 38.45 to 62.54 %. The genotypes from Aoulouz and Had Dra expressing high oil yield. Variance components for fruit, seed, kernel traits and oil content showed significant differences (P < 0.01) among years, provenances, genotypes and their interaction. Principal component analysis proved that fruit, seed, kernel, weight traits are correlated with oil content and are discriminate characters between the genotypes. The results of the cluster analysis support the results of the principal component analysis, showing no correlation between oil content with geographical localization parameters. For all the promising genotypes, at least 25 % were found to be better and exceed the oil mean of the provenance for 3 years. So, 31 promising elite genotypes were preselected, and open new ways for future comparative test of them in diverse environments.