Determination of Genetic Diversity Depending on Quantitative Characters Among Genotypes of Triticale (X Trititcosecale Wittmac) Using Cluster Analysis

The study included twenty genotypes of triticale, whose seeds were sown during 2018-2019 season at the Research Station of the Faculty of Agriculture, University of Kirkuk in the Sayyadah region on three dates (5 November, 20 November and 5 December) using randomized complete block design according to split plot system with three replications. The data were recorded for traits: first, second and third developmental stages, number of days to 50% flowering, plant height, flag leaf area, number of tillers per plant, number of spikes per plant, length and weight of spike, number of spikelet's per spike, number of grains per spike, 1000 grains weight, biological yield, grain yield per plant, harvest index, protein percent, specific weight, gluten percent, flour strength, moisture percent and ash percentage, The data were analyzed to identify the nature of the differences between genotypes and planting dates. Because of the significant (genotypes x planting dates) interaction, a cluster analysis was conducted with the aim of grouping similar genotypes into homogeneous groups and estimating the degree of genetic diversity between them through the use of hierarchical clustering technology to estimate distances between groups of genotypes formed for each planting date separately. The results showed that the mean squares of genotypes' was highly significant 1% for all traits except harvest index, with a highly significant interaction with dates for all traits except number of spikelet's and protein percent. The stages of the cluster analysis showed that the genotypes were distributed into 13 groups for the first date and 14 groups for the second and third dates. Some groups included one genotype, indicating the difference of these genotypes from other due to the difference in their genetic origin, which was consequently reflected on their performance, while other groups includes two genotypes. It is concluded from the results of the clustering analysis that there is a strong convergence between the genotypes of stage 18 with the genotype LIRON at the first date and with POLLMER in the second and third dates because they have the lowest euclidean distances, and this requires avoiding crosses between these pairs, while the highest distance was between CMH80 and CMH82 in the first and third dates and CENT/1715 and POPP-CAAL in the second date indicated high genetic variation between them and other genotypes, which may be due to the variation in their genetic origin or to having preferred main genes, other genotypes devoid of them, which encourages their introduction into hybridization with genotypes that showed distinct genetic variation to take advantage of the phenomenon of heterosis and its segregations.