The variation and the mean traits were examined for different populations. Among the studied populations of Salicornia, the non-fertile parts on the longest secondary branch (V29) (84.75%), the fertile parts on the longest secondary branch (V28) (81.49%0 and the flowering plants in the first lateral branch (V34) (66.13%) had the highest diversity (Table 3). According to the results, the highest and lowest number of primary lateral branches (V9) were observed in (P27, 43) and (P22, 13.4), respectively. Complete information about other variables is given in Table 3.
The first five of the 32 principal components (PCs), obtained have eigenvalues greater than 2. Together they accounted for about 67.28% of the total variance of Salicornia traits (Figure 3, Table 4). The first two PCs account for 42.32% of the total variability (25.76% and 16.56%, respectively) (Table 4). PC1 represent ration of V7, V8, V11, V14, V16, V19, V25, V26, V31, V32, V37, V38, V39, V40, V41, V42, V44, V45, V46, V53, and V55. PC2 describe the ration of V1, V10, V13, V23, V24, V27, V30, and V43. Figure 3 and Table 4 show that traits lie around PC1 and PC2 center. The large variability of the traits allows observation such as V10, V31, V39, V41, and V45, where the amount of length of longest 1 st primary branch, length of the terminal spike, height of central floret of 3rd fertile segment, height of side floret of 3rd fertile segment and distance between florets on 2nd fertile segment.
According to the results of cluster analyses by the Ward method, Salicornia populations were assigned to four groups (Figure 4). The first group contained 8.18% of populations (P16, P18, P24, P31, P20, and P22). In this group, populations with a short height, long spike, greater weight of 1000 seed, low number of Stomata, and the width across the apex on the third fertile segment were more abundant than other populations. The morphotype and inflorescences of this group was distinct from other groups. The second group covered 15.15% of the whole population (P3, P11, P23, P2, and P33), comprising populations that were within the average range of trait sizes for diverse traits. The third group hosted 15.15% of the population (P4, P6, P1, P8, and P10), and the fourth group included 51.51% of the population (P6, P30, P25, P27, P21, P26, P15, P12, P28, P28, P7, P14, P17, P5, P9, P19, P29, P13, and p32). These populations had a great height, more internodes, more lateral branches, more stomata, a great weight of 1000 seeds and the width of the third fertile segment on the terminal spike. The accurate number of groups was identified using the detection function.
Genetic diversity of Salicornia populations
We evaluated genetic diversity in 33 Salicornia populations using 42 ISSR primers. 23 primers out of 42 primers under study generated a polymorphic band design at the suitable resolution, which were employed for the subsequent analysis phases (Table 5). Totally, 204 locations (averagely 8.87 locations per primer) were produced by these primers, 134 of them were polymorphic (65.69%). The ratio of markers to primer was 1 to 14, averagely 5.82 (figure 5).
The number of effective (Ne) alleles in UBC849 was 1.25 and in PB it was 1.92, averagely 1.63 in each gene locus. Maximum value of this statistic shows that alleles have identical frequency in this location, and this statistic's minimum shows the rarity of other alleles and one allele's high frequency in samples.
In investigating allelic diversity, the highest observed heterozygosity was found in the B marker as 0.477, and the lowest observed heterozygosity was noticed in the UBC849 marker as 0.199. Besides, the highest expected heterozygosity was observed at approximately 0.484 in B marker, and the lowest expected heterozygosity was observed at approximately 0.203 in the UBC849 marker. Examining Shannon index (I) values showed that the highest value for this index was in marker B with a 0.670 and the lowest value was in UBC849 marker as 0.351 (Table 6).
The Jaccard similarity coefficient and UPGMA algorithm were used for dividing different populations into two separate groups. The first group contained 12.12% and the second group included 87.88% of the masses. Two subgroups were made in the first group, which the first one included P24, P22, P26 and P1. The second group contains the residual 29 populations (P13, P20, P18, P30, P32, P29, P19, P8, P15, P17, P5, P27, P12, P33, P28, P31, P16, P21, P23, P14, P4, P3, P2, P7, P25, P6, P10, P11, P9), which was classified into two subgroups. The first one is composed of just the P13 population. Also, this population was approximately different from other ones (Fig. 6).
Structure 2.3.1 software was used for analyzing genetic population structure and precise classifying individuals into proper subpopulations. As shown by a two-way diagram of optimal determination of K with ISSR indicator, the ISSR primer shows the best K as 2, i.e., two subpopulations (K = 2) in the cultivars under study. The group was specified (Table 7, Table 8).
The stabilization index (Fst) is a common and appropriate measure for genetic differentiation among groups and populations. When the Fst is higher, a better allele differentiation is obtained, with a higher allele stabilization rate. Potential subgroups in K = 2 show the difference among the populations under study in two potential groups. Besides, the individuals' matrix of the share in these groups (Table 4-5) indicated belonging populations with high coefficients to one group. Barplot results demonstrated inclusion of 26 Salicornia populations in the first group (Red), and 5 populations in the second group (Green), with 2 populations had a complex structure (Fig. 7).