Genetic diversity of Walnut (Juglans regia L.) genotypes using multivariate analysis in Lorestan province

Persian walnut (Juglans regia L.) is one of the most important species of Juglans genus and Iran is one of the most important centers of genetic diversity of this species. Thus evaluation of genetic diversity of Persian walnut in Lorestan province is very important. So if the results displayed a sufficient genetic diversity of nut and tree, we can use it for walnut breeding programs in future.
In this study 43 Persian walnut genotypes of the five populations were selected. 41 morphologic traits evaluated between these populations were including Alashtar, Doroud, Khorram-abad, Boroujerd and Nourabad. Assaying traits on 10 to 20 samples from each tree, and 5 to 10 leaves, that selected accidently, for 41 quantitative traits related to nuts and leaves such as protein and oil of nuts was done.
One-sided variance analysis of the data’s indicated a significant difference between regions for most traits. Means comparison with Duncan’s test at 5% level showed that most traits were significantly different in various regions of Lorestan province. The study of variety coefficients of different traits showed that the husk weight had the most variation among the traits (35.51%). After that, husk humidity (29.45%) and green seed weight (28.39%) respectively owned the major variation among the studied traits. In terms of high protein and oil, Alashtar (20.26 and 64.1%), Boroujerd (19.98 and 60.73%) and Doroud (19.14 and 59.11%) populations respectively had the highest contents of protein and oil than other populations. In terms of nut yield, Boroujerd population had seed weight with hard skin of 12.13gr and highest kernel weight (4.98gr). This population made it the most suitable population for selecting stocks with suitable nut yield. On other hands, the highest percentage of kernel with hard skin was observed in Nourabad population and the highest percentage of kernel with green skin was in Doroud population. Cluster analysis of morphologic traits based on Ward method, differentiated genotype in 3 clusters in distance of -41.21 and populations also in 3 clusters in distance of 67.71. In the first cluster Alashtar, Nourabad and Beyranshahr populations, in the second cluster Doroud population and in the third cluster Boroujerd populations were placed. Based on the results of principal component analysis (PCA), 3 main components justified total of 93.30% of total variance. Dispersion of populations based on 2 main components were corresponded with cluster analysis of walnut populations based on morphologic traits in Lorestan province and populations were put in 3 different areas on diagram.
Results showed that genetic relationships among populations that were geographically near each other, caused these populations to be in same groups in cluster analysis. Green seed volume and thickness, length and width of seed with hard skin, had the highest effects on dispersion of populations in this study. Doroud and Nourabad populations respectively with the highest percentage of kernel with green skin and kernel with hard skin are recommended for selection in terms of kernel yield and future walnut breeding program in Lorestan province. Results showed that walnut populations in Lorestan province had a high diversity and the selection must be done through mentioned traits.